Tools/ChangeLog

   1 2011-11-30  Adam Roben  <aroben@apple.com>
   2
   3         Make NRWT show DRT/WTR build logs when the build fails or --verbose is passed
   4
   5         This will make it a lot easier to investigate build failures on the bots.
   6
   7         Fixes <http://webkit.org/b/71160> NRWT doesn't show build output when building DRT
   8
   9         Reviewed by Eric Seidel.
  10
  11         * Scripts/webkitpy/common/system/executive_mock.py:
  12         (MockExecutive.run_command): Include the mock output in the ScriptError we raise since it
  13         makes the expected output in WebKitPortTest.test_build_driver more closely mimic the output
  14         seen in practice.
  15
  16         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
  17         (ManagerTest.test_http_locking): Pass False for the configure_logging parameter to Printer.
  18         The Printer was modifying the root logger in way that would cause my new tests in
  19         WebKitPort.test_build_driver to fail. The value of the configure_logging parameter doesn't
  20         seem important to the functioning of this test, and passing False makes Printer not modify
  21         the root logger.
  22
  23         * Scripts/webkitpy/layout_tests/port/webkit.py:
  24         (WebKitPort._run_script): Log the output of the command at the DEBUG level so it will show
  25         up when --verbose is passed but not otherwise.
  26         (WebKitPort._build_driver): When an error occurs, log the error message and the script's
  27         output at the ERROR level so it will always show up.
  28
  29         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
  30         (WebKitPortTest.test_build_driver): Test that nothing is logged in the normal, successful
  31         cases. Test that the build scripts' output is logged when --verbose is passed or the build
  32         fails.
  33
  34 2011-11-30  Adam Roben  <aroben@apple.com>
  35
  36         Add a way to set OutputCapture's log level
  37
  38         This will be used by some unit tests I'm planning to add.
  39
  40         Fixes <http://webkit.org/b/73469> OutputCapture can't be used to capture DEBUG messages
  41
  42         Reviewed by Eric Seidel.
  43
  44         * Scripts/webkitpy/common/system/outputcapture.py:
  45         (OutputCapture.__init__): Initialize our internal logging level to INFO, which was the only
  46         level we ever used before this patch.
  47         (OutputCapture.set_log_level): Added. This is useful for testing what log level code that is
  48         being tested uses. Sets the log level for any current or future logging handlers.
  49         (OutputCapture.capture_output): Use our internal logging level instead of always using INFO.
  50
  51         * Scripts/webkitpy/common/system/outputcapture_unittest.py: Added.
  52         (OutputCaptureTest.setUp): Create an OutputCapture to test.
  53         (OutputCaptureTest.log_all_levels): Helper method to log a message at each log level.
  54         (OutputCaptureTest.assertLogged): Helper method to assert that a particular set of strings
  55         was logged.
  56         (OutputCaptureTest.test_initial_log_level): Test that we start out at the INFO level.
  57         (OutputCaptureTest.test_set_log_level): Test that set_log_level correctly modifies the log
  58         level.
  59
  60 2011-11-29  Beth Dakin  <bdakin@apple.com>
  61
  62         https://bugs.webkit.org/show_bug.cgi?id=72751
  63         WebKit2.MouseMoveAfterCrash API test is failing
  64
  65         Reviewed by Sam Weinig.
  66
  67         Like DumpRenderTree, the NSWindow should act as an active window.
  68         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
  69         (-[ActiveOffscreenWindow isKeyWindow]):
  70         (TestWebKitAPI::PlatformWebView::PlatformWebView):
  71
  72 2011-11-30  Gustavo Noronha Silva  <gns@gnome.org>
  73
  74         Add step to update gtk dependencies to the buildbot
  75         https://bugs.webkit.org/show_bug.cgi?id=73455
  76
  77         Reviewed by Martin Robinson.
  78
  79         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
  80
  81 2011-11-30  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
  82
  83         [Qt] Add a fullscreen option to MiniBrowser.
  84         https://bugs.webkit.org/show_bug.cgi?id=73445
  85
  86         Reviewed by Noam Rosenthal.
  87
  88         This bypasses the compositor on the N9 to increase performances
  89         and remove the huge useless gray bar part of the window frame
  90         at the bottom.
  91
  92         * MiniBrowser/qt/BrowserWindow.cpp:
  93         (BrowserWindow::BrowserWindow):
  94         * MiniBrowser/qt/MiniBrowserApplication.cpp:
  95         (printHelp):
  96         (MiniBrowserApplication::handleUserOptions):
  97         * MiniBrowser/qt/MiniBrowserApplication.h:
  98         (WindowOptions::setStartFullScreen):
  99         (WindowOptions::startFullScreen):
 100
 101 2011-11-30  Martin Robinson  <mrobinson@igalia.com>
 102
 103         [GTK] Add an initial jhbuild setup which installs fonts into the WebKitBuild
 104         https://bugs.webkit.org/show_bug.cgi?id=73425
 105
 106         Reviewed by Gustavo Noronha Silva.
 107
 108         Add initial jhbuild support to the GTK+ port. update-webkitgtk-libs will
 109         ensure that the jhbuild root is up-to-date. Currently the only module is
 110         the font module.
 111
 112         * Scripts/update-webkitgtk-libs: Added.
 113         * gtk/common.py: Added a helper to get the number of CPUs.
 114         * gtk/jhbuild.modules: Added.
 115         * gtk/jhbuildrc: Added.
 116
 117 2011-11-30  Simon Hausmann  <simon.hausmann@nokia.com>
 118
 119         [Qt] Remove dependency of Qt DRT on JavaScriptCore.
 120
 121         Reviewed by Csaba Osztrogonác.
 122
 123         * DumpRenderTree/WorkQueue.cpp: Include the correct header file we're using anyway,
 124         instead of one that pulls in JavaScriptCore headers.
 125         * DumpRenderTree/qt/DumpRenderTree.pro: There's no need to pull in the JSC headers
 126         with load(javascriptcore).
 127
 128 2011-11-30  Sheriff Bot  <webkit.review.bot@gmail.com>
 129
 130         Unreviewed, rolling out r101440 and r101442.
 131         http://trac.webkit.org/changeset/101440
 132         http://trac.webkit.org/changeset/101442
 133         https://bugs.webkit.org/show_bug.cgi?id=73429
 134
 135         multiple crashes on layout tests (Requested by hayato on
 136         #webkit).
 137
 138         * DumpRenderTree/chromium/TestShell.cpp:
 139         (TestShell::resetWebSettings):
 140         * DumpRenderTree/chromium/TestShell.h:
 141         * DumpRenderTree/chromium/WebPreferences.cpp:
 142         (WebPreferences::reset):
 143         (WebPreferences::applyTo):
 144         * DumpRenderTree/chromium/WebPreferences.h:
 145         * DumpRenderTree/chromium/WebViewHost.cpp:
 146         (WebViewHost::WebViewHost):
 147         (WebViewHost::~WebViewHost):
 148
 149 2011-11-30  Fady Samuel  <fsamuel@chromium.org>
 150
 151         [Chromium] Fix broken DRT build for Aura Linux
 152         https://bugs.webkit.org/show_bug.cgi?id=72667
 153
 154         Reviewed by Tony Chang.
 155
 156         DRT now builds for Aura Linux. This, more or less, shares the same code as Android.
 157
 158         The following changes have been made:
 159
 160           1. Renamed TestShellAndroid to TestShellLinux.
 161           2. Stub code moved added in TestShellStub that will only compile on non-GTK Linux builds.
 162           3. Code common to Gtk, Aura, and Android moved to TestShellLinux.
 163
 164         * DumpRenderTree/DumpRenderTree.gypi:
 165         * DumpRenderTree/chromium/EventSender.cpp:
 166         (EventSender::keyDown):
 167         * DumpRenderTree/chromium/TestShellAndroid.cpp: Removed.
 168         * DumpRenderTree/chromium/TestShellGtk.cpp:
 169         * DumpRenderTree/chromium/TestShellLinux.cpp: Copied from Tools/DumpRenderTree/chromium/TestShellGtk.cpp.
 170         (AlarmHandler):
 171         (TestShell::waitTestFinished):
 172         (setupFontconfig):
 173         (platformInit):
 174         * DumpRenderTree/chromium/TestShellStub.cpp: Added.
 175         (checkLayoutTestSystemDependencies):
 176         (openStartupDialog):
 177
 178 2011-11-30  Fady Samuel  <fsamuel@chromium.org>
 179
 180         [Chromium] Set Result Before Early Exit for Fixed Layout Methods in LayoutTestController
 181         https://bugs.webkit.org/show_bug.cgi?id=73328
 182
 183         Reviewed by Kent Tamura.
 184
 185         * DumpRenderTree/chromium/LayoutTestController.cpp:
 186         (LayoutTestController::enableFixedLayoutMode):
 187         (LayoutTestController::setFixedLayoutSize):
 188
 189 2011-11-29  Hayato Ito  <hayato@chromium.org>
 190
 191         Skip writing the result of diff_image since non-chromium ports don't implement diff_image.
 192         https://bugs.webkit.org/show_bug.cgi?id=73381
 193
 194         Reviewed by Ryosuke Niwa.
 195
 196         This is a quick fix for the breakage of tests on non-chromium port.
 197         We should implement diff_image later on non-chromium ports.
 198
 199         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
 200         (write_test_result):
 201
 202 2011-11-29  Balazs Ankes  <Ankes.Balazs@stud.u-szeged.hu>
 203
 204         [NRWT] Fix --platform=qt-5.0 --new-baseline combo
 205         https://bugs.webkit.org/show_bug.cgi?id=72489
 206
 207         Reviewed by Eric Seidel.
 208
 209         Before the this fix always added "qt" option for the option list
 210         * Scripts/run-webkit-tests:
 211         * Scripts/webkitpy/layout_tests/port/qt.py:
 212
 213 2011-11-29  David Levin  <levin@chromium.org>
 214
 215         Add a way to revert a variable to its previous value after leaving a scope.
 216         https://bugs.webkit.org/show_bug.cgi?id=73371
 217
 218         Reviewed by Adam Barth.
 219
 220         * DumpRenderTree/ForwardingHeaders/wtf/TemporarilyChange.h: Added.
 221         * TestWebKitAPI/TestWebKitAPI.gypi: Added test file to the build.
 222         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Ditto.
 223         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Ditto.
 224         * TestWebKitAPI/Tests/WTF/TemporarilyChange.cpp: Added.
 225         (TestWebKitAPI::TEST): Added a test for TemporarilyChange.
 226
 227 2011-11-28  Kentaro Hara  <haraken@chromium.org>
 228
 229         Implement [Supplemental] IDL and support it in run-bindings-tests
 230         https://bugs.webkit.org/show_bug.cgi?id=73162
 231
 232         Reviewed by Adam Barth.
 233
 234         run-bindings-tests supports the [Supplemental] IDL in the following way.
 235             (1) It creates a file |x| listing all IDL files.
 236             (2) It generates a supplemental dependency file |y| using resolve-supplemental.pl and |x|.
 237             (3) For each IDL file |z|, it runs generate-bindings.pl with |y|.
 238                 Then, generate-bindings.pl generates .h and .cpp files for the IDL file |z|,
 239                 including all attributes in IDL files that are supplementing the IDL file |z|.
 240
 241         * Scripts/run-bindings-tests:
 242         (generate_from_idl):
 243         (generate_supplemental_dependency):
 244         (detect_changes):
 245         (run_tests):
 246         (main):
 247
 248 2011-11-18  Nat Duca  <nduca@chromium.org>
 249
 250         [chromium] Enable threaded compositing via CCThreadProxy::hasThread only
 251         https://bugs.webkit.org/show_bug.cgi?id=70838
 252
 253         Reviewed by James Robinson.
 254
 255         * DumpRenderTree/chromium/TestShell.cpp:
 256         (TestShell::resetWebSettings):
 257         * DumpRenderTree/chromium/TestShell.h:
 258         (TestShell::threadedCompositingEnabled):
 259         * DumpRenderTree/chromium/WebPreferences.cpp:
 260         (WebPreferences::reset):
 261         (WebPreferences::applyTo):
 262         * DumpRenderTree/chromium/WebPreferences.h:
 263         * DumpRenderTree/chromium/WebViewHost.cpp:
 264         (WebViewHost::WebViewHost):
 265         (WebViewHost::~WebViewHost):
 266
 267 2011-11-29  Ojan Vafai  <ojan@chromium.org>
 268
 269         List of builders in the flakiness dashboard is out of date
 270         https://bugs.webkit.org/show_bug.cgi?id=73347
 271
 272         Reviewed by Adam Barth.
 273
 274         The chromium gpu bots no longer run layout tests, so kill that
 275         group entirely.
 276
 277         * TestResultServer/static-dashboards/builders.js:
 278         * TestResultServer/static-dashboards/dashboard_base.js:
 279
 280 2011-11-29  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
 281
 282         [Qt] Don't hard-code the list of WebKit2 generated sources
 283
 284         The generated sources are... wait for it... generated. So
 285         use the generator itself to figure out which sources we need
 286         to compile.
 287
 288         Reviewed by Simon Hausmann.
 289
 290         * qmake/mkspecs/features/default_post.prf:
 291
 292 2011-11-29  Martin Robinson  <mrobinson@igalia.com>
 293
 294         [GTK] Add a method to detect 'make dist' errors without running 'make dist'
 295         https://bugs.webkit.org/show_bug.cgi?id=73216
 296
 297         Reviewed by Philippe Normand.
 298
 299         Add a script that tries to sniff out 'make dist' problems without running
 300         'make dist.' 'make distcheck' takes a very long time to run and this should
 301         reduce the amount of times it needs to be run consecutively.
 302
 303         * gtk/common.py:
 304         (get_build_path.is_valid_build_directory): Guess the source directory
 305         by the existence of the GNUmakefile instead of the .libs directory. This
 306         allows one to run the script after running autogen.sh but before fully
 307         building.
 308         * gtk/find-make-dist-errors: Added.
 309
 310 2011-11-29  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
 311
 312         [Qt] Remove use of internal headers in the MiniBrowser
 313
 314         Reviewed by Simon Hausmann.
 315
 316         * MiniBrowser/qt/MiniBrowser.pro:
 317
 318 2011-11-29  Simon Hausmann  <simon.hausmann@nokia.com>
 319
 320         [Qt] Fix debug-shlib build without webkit2
 321
 322         Reviewed by Tor Arne Vestbø.
 323
 324         * qmake/mkspecs/features/qtwebkit.prf: Respect no_webkit2.
 325
 326 2011-11-28  Csaba Osztrogonác  <ossy@webkit.org>
 327
 328         [Qt][WK2] Unreviewed buildfix after r101307.
 329
 330         * qmake/mkspecs/features/webkit2.prf: Add missing includepath.
 331
 332 2011-11-29  Roland Steiner  <rolandsteiner@chromium.org>
 333
 334         <style scoped>: add ENABLE(STYLE_SCOPED) flag to WebKit
 335         https://bugs.webkit.org/show_bug.cgi?id=72848
 336
 337         Add option to build-webkit.
 338
 339         Reviewed by Dimitri Glazkov.
 340
 341         * Scripts/build-webkit:
 342
 343 2011-11-28  Ojan Vafai  <ojan@chromium.org>
 344
 345         gtest normalization in the test results server is sometimes wrong
 346         https://bugs.webkit.org/show_bug.cgi?id=73262
 347
 348         Reviewed by Tony Chang.
 349
 350         If a gtest has it's modified changed (e.g. add FLAKY_) then it will have two
 351         entries in the incremental JSON, one of the entries will have the correct data
 352         from the run and the other will have dummy no-data values. Make sure to
 353         always pick the one with real data.
 354
 355         * TestResultServer/model/jsonresults.py:
 356         (JsonResults._remove_gtest_modifiers):
 357         * TestResultServer/model/jsonresults_unittest.py:
 358         (JsonResultsTest.test_remove_gtest_modifiers):
 359
 360 2011-11-28  Ojan Vafai  <ojan@chromium.org>
 361
 362         Some of the results.json files have results/times entries at the directory level
 363         https://bugs.webkit.org/show_bug.cgi?id=73261
 364
 365         Reviewed by Tony Chang.
 366
 367         This is just a bug that got introduced in a temporary push of the results server.
 368         This patch repairs the broken files. After all the bots have cycled, we can simplify
 369         this code to just assert that results/times are not at the directory level.
 370
 371         Also, when catching exceptions, log the full stacktrace.
 372
 373         * TestResultServer/model/jsonresults.py:
 374         (_is_directory):
 375         (JsonResults._load_json):
 376         (JsonResults._merge_tests):
 377         (JsonResults.merge):
 378         * TestResultServer/model/jsonresults_unittest.py:
 379         (JsonResultsTest.test_merge_directory_hierarchy_extra_results_and_times):
 380
 381 2011-11-28  Tony Chang  <tony@chromium.org>
 382
 383         ews bots should pass --force to update-webkit-chromium
 384         https://bugs.webkit.org/show_bug.cgi?id=73230
 385
 386         Reviewed by Adam Barth.
 387
 388         This makes it less likely for gclient changes to break the bots (e.g.,
 389         if a DEPS repository moves or is switched to a branch).
 390
 391         Also pass --force when using |build-webkit --update-chromium| since this
 392         appears to only be used by the bots.
 393
 394         Take 2: Add Options.non_interactive to update.py's options() method.
 395
 396         * Scripts/update-webkit:
 397         * Scripts/webkitdirs.pm:
 398         (determineIsChromium): Add --force-update to update-webkit --chromium.
 399         (forceChromiumUpdate):
 400         (buildChromium): Pass --force to update-webkit-chromium.
 401         * Scripts/webkitpy/tool/steps/update.py:
 402         (Update.run): Add --force-update if non-interactive (i.e., bots).
 403         * Scripts/webkitpy/tool/steps/update_unittest.py:
 404
 405 2011-11-28  Tony Chang  <tony@chromium.org>
 406
 407         Revert r101279, broke the ews and cq bots.
 408
 409         * Scripts/update-webkit:
 410         * Scripts/webkitdirs.pm:
 411         (determineIsChromium):
 412         (buildChromium):
 413         * Scripts/webkitpy/common/config/ports.py:
 414         (WebKitPort.update_webkit_command):
 415         (ChromiumPort.update_webkit_command):
 416         * Scripts/webkitpy/common/config/ports_mock.py:
 417         (MockPort.update_webkit_command):
 418         * Scripts/webkitpy/tool/steps/update.py:
 419         * Scripts/webkitpy/tool/steps/update_unittest.py: Removed.
 420
 421 2011-11-28  Ojan Vafai  <ojan@chromium.org>
 422
 423         testlistjson on the test results server doesn't understand hierarchical results format
 424         https://bugs.webkit.org/show_bug.cgi?id=73246
 425
 426         Reviewed by Tony Chang.
 427
 428         * TestResultServer/model/jsonresults.py:
 429         (JsonResults._delete_results_and_times):
 430         Instead of just getting the top-level keys, we now walk the tests tree and
 431         delete the results and times values.
 432
 433         * TestResultServer/model/jsonresults_unittest.py:
 434         Simplify the test harness to take in the same JSON format for the tests
 435         that the actual results.json uses instead of something very similar but
 436         needlessly different. This also allows for testing the broken case of
 437         results and times values being at non-leaf level.
 438
 439 2011-11-28  Tony Chang  <tony@chromium.org>
 440
 441         ews bots should pass --force to update-webkit-chromium
 442         https://bugs.webkit.org/show_bug.cgi?id=73230
 443
 444         Reviewed by Adam Barth.
 445
 446         This makes it less likely for gclient changes to break the bots (e.g.,
 447         if a DEPS repository moves or is switched to a branch).
 448
 449         Also pass --force when using |build-webkit --update-chromium| since this
 450         appears to only be used by the bots.
 451
 452         * Scripts/update-webkit:
 453         * Scripts/webkitdirs.pm:
 454         (determineIsChromium): Add --force-update to update-webkit --chromium.
 455         (forceChromiumUpdate):
 456         (buildChromium): Pass --force to update-webkit-chromium.
 457         * Scripts/webkitpy/tool/steps/update.py:
 458         (Update.run): Add --force-update if non-interactive (i.e., bots).
 459         * Scripts/webkitpy/tool/steps/update_unittest.py:
 460
 461 2011-11-28  Michael Saboff  <msaboff@apple.com>
 462
 463         Fixed help message for --exclude-kraken to say that
 464         it excludes Kraken and not SunSpider.
 465
 466         Rubber-stamped by Filip Pizlo.
 467
 468         * Scripts/bencher:
 469
 470 2011-11-28  Sheriff Bot  <webkit.review.bot@gmail.com>
 471
 472         Unreviewed, rolling out r101273.
 473         http://trac.webkit.org/changeset/101273
 474         https://bugs.webkit.org/show_bug.cgi?id=73244
 475
 476         multiple test failures and timeouts (Requested by tony^work on
 477         #webkit).
 478
 479         * DumpRenderTree/DumpRenderTree.gypi:
 480         * DumpRenderTree/chromium/EventSender.cpp:
 481         (EventSender::keyDown):
 482         * DumpRenderTree/chromium/TestShellAndroid.cpp: Copied from Tools/DumpRenderTree/chromium/TestShellGtk.cpp.
 483         (AlarmHandler):
 484         (TestShell::waitTestFinished):
 485         (platformInit):
 486         (openStartupDialog):
 487         (checkLayoutTestSystemDependencies):
 488         * DumpRenderTree/chromium/TestShellGtk.cpp:
 489         (AlarmHandler):
 490         (setupFontconfig):
 491         (TestShell::waitTestFinished):
 492         (platformInit):
 493         * DumpRenderTree/chromium/TestShellLinux.cpp: Removed.
 494         * DumpRenderTree/chromium/TestShellStub.cpp: Removed.
 495
 496 2011-11-28  John Yani  <vanuan@gmail.com>
 497
 498         new-run-webkit-tests is locale dependent
 499         https://bugs.webkit.org/show_bug.cgi?id=68691
 500
 501         Reviewed and modified by Eric Seidel.
 502
 503         This is a simpler version of John's patch which I'm landing.
 504         This just engages the hacks from the Host constructor on a global
 505         basis.
 506
 507         * Scripts/webkitpy/common/host.py:
 508         (Host.__init__):
 509         (Host._engage_awesome_locale_hacks):
 510
 511 2011-11-28  Fady Samuel  <fsamuel@chromium.org>
 512
 513         [Chromium] Fix broken DRT build for Aura Linux
 514         https://bugs.webkit.org/show_bug.cgi?id=72667
 515
 516         Reviewed by Tony Chang.
 517
 518         DRT now builds for Aura Linux. This, more or less, shares the same code as Android.
 519
 520         The following changes have been made:
 521
 522           1. Renamed TestShellAndroid to TestShellLinux.
 523           2. Stub code moved added in TestShellStub that will only compile on non-GTK Linux builds.
 524           3. Code common to Gtk, Aura, and Android moved to TestShellLinux.
 525
 526         * DumpRenderTree/DumpRenderTree.gypi:
 527         * DumpRenderTree/chromium/EventSender.cpp:
 528         (EventSender::keyDown):
 529         * DumpRenderTree/chromium/TestShellAndroid.cpp: Removed.
 530         * DumpRenderTree/chromium/TestShellGtk.cpp:
 531         * DumpRenderTree/chromium/TestShellLinux.cpp: Copied from Tools/DumpRenderTree/chromium/TestShellGtk.cpp.
 532         (AlarmHandler):
 533         (TestShell::waitTestFinished):
 534         (setupFontconfig):
 535         (platformInit):
 536         * DumpRenderTree/chromium/TestShellStub.cpp: Added.
 537         (checkLayoutTestSystemDependencies):
 538         (openStartupDialog):
 539
 540 2011-11-28  Eric Seidel  <eric@webkit.org>
 541
 542         Add Environment object to Host and fix the GCC smartquotes trouble seen on the commit-queue
 543         https://bugs.webkit.org/show_bug.cgi?id=71983
 544
 545         Reviewed by Adam Barth.
 546
 547         We'll add more code to Environment overtime,
 548         allowing us to mock out more of our direct interactions with os.environ.
 549
 550         This patch also makes run_command print the passed in environment.
 551
 552         * Scripts/webkitpy/common/host_mock.py:
 553         * Scripts/webkitpy/common/system/environment.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
 554         * Scripts/webkitpy/common/system/environment_mock.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
 555         * Scripts/webkitpy/common/system/environment_unittest.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
 556         * Scripts/webkitpy/common/system/executive.py:
 557         * Scripts/webkitpy/common/system/executive_mock.py:
 558         * Scripts/webkitpy/layout_tests/port/webkit.py:
 559         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
 560         * Scripts/webkitpy/tool/commands/download_unittest.py:
 561         * Scripts/webkitpy/tool/steps/build.py:
 562
 563 2011-11-28  Simon Hausmann  <simon.hausmann@nokia.com>
 564
 565         [Qt] Build system fixes against V8.
 566
 567         Reviewed by Tor Arne Vestbø.
 568
 569         * qmake/mkspecs/features/webcore.prf: Get rid of old v8 cruft.
 570         * qmake/mkspecs/features/wtf.prf: If requested, pull in v8 from Qt and configure
 571         WTF accordingly to use v8. (WTF in the implementation needs it as well as users of WTF)
 572
 573 2011-11-28  Kenneth Rohde Christiansen  <kenneth@webkit.org>
 574
 575         Make sure the useFixedLayout feature is consistently handled
 576         https://bugs.webkit.org/show_bug.cgi?id=73212
 577
 578         Reviewed by Simon Hausmann.
 579
 580         The web page parameters are set to our Qt defaults and are now altered
 581         in the WebKitTestRunner.
 582
 583         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
 584         (WTR::WrapperWindow::WrapperWindow):
 585
 586 2011-11-28  Simon Hausmann  <simon.hausmann@nokia.com>
 587
 588         [Qt] WTF should be built as separate static library
 589         https://bugs.webkit.org/show_bug.cgi?id=73201
 590
 591         Reviewed by Tor Arne Vestbø.
 592
 593         * DumpRenderTree/qt/DumpRenderTree.pro: Require wtf.
 594         * MiniBrowser/qt/BrowserWindow.cpp: Remove unnecessary config.h inclusion and
 595         thus wtf dependency.
 596         * WebKitTestRunner/InjectedBundle/Target.pri: Require wtf.
 597         * WebKitTestRunner/Target.pri: Ditto.
 598         * qmake/mkspecs/features/default_post.prf: When re-ordering the static libs
 599         consider wtf, too.
 600         * qmake/mkspecs/features/javascriptcore.prf: Remove WTF specific build stuff,
 601         moved into wtf.prf.
 602         * qmake/mkspecs/features/qtwebkit.prf: For debug builds with shared libs, also
 603         link in libwtf.
 604         * qmake/mkspecs/features/wtf.prf: Added. Stuff needed for building and using WTF.
 605
 606 2011-11-28  Kentaro Hara  <haraken@chromium.org>
 607
 608         run-bindings-tests shows diff error for a newly added binding test
 609         https://bugs.webkit.org/show_bug.cgi?id=73179
 610
 611         Reviewed by Adam Barth.
 612
 613         Currently, when we add a new binding test, say bindings/scripts/test/TestSupplemental.idl,
 614         run-bindings-tests just shows diff error because reference files
 615         (i.e. {V8,JS,...}TestSupplemental.h and {V8,JS,...}TestSupplemental.cpp) do not yet exist:
 616
 617             FAIL: (V8) V8TestSupplemental.h
 618             diff: WebCore/bindings/scripts/test/V8/V8TestSupplemental.h: No such file or directory
 619
 620         Instead, this patch treats a non-existent reference file as an empty file and
 621         shows the result of the newly added binding test, like this:
 622
 623             FAIL: (V8) V8TestSupplemental.h
 624             --- WebCore/bindings/scripts/test/V8/V8TestSupplemental.h       1970-01-01 09:00:00.000000000 +0900
 625             +++ /tmp/tmpmX2i2F/V8TestSupplemental.h 2011-11-28 11:08:34.150707677 +0900
 626             @@ -0,0 +1,84 @@
 627             +/*
 628             +    This file is part of the WebKit open source project.
 629             +    This file has been generated by generate-bindings.pl. DO NOT MODIFY!
 630             +
 631             (omitted)
 632
 633         * Scripts/run-bindings-tests:
 634         (detect_changes): Just added an -N option to the diff command.
 635
 636 2011-11-28  Martin Robinson  <mrobinson@igalia.com>
 637
 638         [Gtk][gtkdoc] Syntax error in Tools/gtk/generate-gtkdoc
 639         https://bugs.webkit.org/show_bug.cgi?id=73165
 640
 641         Reviewed by Philippe Normand.
 642
 643         Fix some syntax errors in this script and remove an unused variable.
 644
 645         * gtk/generate-gtkdoc:
 646         (get_webkit2_options.src_path):
 647         (get_webkit2_options):
 648         (get_webkit1_options.src_path):
 649
 650 2011-11-26  Kevin Ollivier  <kevino@theolliviers.com>
 651
 652         [wx] Unreviewed build fix. Enable the inspector.
 653
 654         * waf/build/settings.py:
 655
 656 2011-11-26  Martin Robinson  <mrobinson@igalia.com>
 657
 658         Try to fix the build after r101174.
 659
 660         * gtk/generate-gtkdoc: Properly handle empty PKG_CONFIG_PATHs.
 661
 662 2011-11-17  Martin Robinson  <mrobinson@igalia.com>
 663
 664         [GTK] Integrate build-gtkdoc into build-webkit and make
 665         https://bugs.webkit.org/show_bug.cgi?id=72626
 666
 667         Reviewed by Philippe Normand.
 668
 669         * GNUmakefile.am: Added support for new scripts and "make docs".
 670         * Scripts/webkitdirs.pm: Call generate-gtkdoc when building.
 671         (buildAutotoolsProject):
 672         * gtk/generate-gtkdoc: Added.
 673         * gtk/gtkdoc.py: Renamed from Source/WebKit2/UIProcess/API/gtk/docs/gtkdoc.py.
 674
 675 2011-11-25  Ádám Kallai  <Kallai.Adam@stud.u-szeged.hu>
 676
 677         [Qt] Add support for using OWRT with Qt5
 678         https://bugs.webkit.org/show_bug.cgi?id=72947
 679
 680         Add qt-5.0 platform and don't pass -graphicssystem to QtTestBrowser if using Qt5.
 681
 682         Reviewed by Csaba Osztrogonác.
 683
 684         * Scripts/old-run-webkit-tests:
 685
 686 2011-11-25  Yury Semikhatsky  <yurys@chromium.org>
 687
 688         [Chromium] Web Inspector: get rid of WebDevToolsFrontendClient::sendFrontendLoaded method
 689         https://bugs.webkit.org/show_bug.cgi?id=73126
 690
 691         No need to have a separate client call for this. The front-end can store messages
 692         which arrive before front-end loaded and dispatch them later. This happens in layout
 693         tests only by the way as all other messages should be initiated by the front-end code.
 694
 695         Reviewed by Pavel Feldman.
 696
 697         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
 698         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
 699         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
 700         * DumpRenderTree/chromium/DRTDevToolsClient.h:
 701
 702 2011-11-24  Kentaro Hara  <haraken@chromium.org>
 703
 704         Replace subprocess.call() with subprocess.Popen()
 705         https://bugs.webkit.org/show_bug.cgi?id=73105
 706
 707         Reviewed by Adam Barth.
 708
 709         Currently, the stdout and stderr of the child processes of run-bindings-tests
 710         are mixed (or even gone away) with the stdout and stderr of run-bindings-tests itself.
 711         This is the problem of subprocess.call(), which does not synchronize the stdout and stderr
 712         between a parent process and a child process. This patch replaces subprocess.call() with
 713         subprocess.Popen() and synchronizes the stdout and stderr between them.
 714         This patch also makes output messages of run-bindings-tests more readable.
 715
 716         With this patch, the output messages of run-bindings-tests look like as follows:
 717
 718             PASS: (CPP) WebDOMTestObj.h
 719             PASS: (CPP) WebDOMTestObj.cpp
 720             FAIL: (CPP) WebDOMTestInterface.h
 721             --- Source/WebCore/bindings/scripts/test/CPP/WebDOMTestInterface.h      2011-11-25 13:34:09.313516268 +0900
 722             +++ /tmp/tmpDAV87G/WebDOMTestInterface.h        2011-11-25 13:44:42.712946812 +0900
 723             @@ -23,7 +23,7 @@
 724              #ifndef WebDOMTestInterface_h
 725              #define WebDOMTestInterface_h
 726
 727             -#if ENABLE(Condition1) || ENABLE(Condition2)
 728             +#if ENABLE(Condition1) || ENABLE(Condition3)
 729
 730              #include <WebDOMObject.h>
 731              #include <WebDOMString.h>
 732              @@ -53,5 +53,5 @@
 733              WebDOMTestInterface toWebKit(WebCore::TestInterface*);
 734
 735              #endif
 736             -#endif // ENABLE(Condition1) || ENABLE(Condition2)
 737             +#endif // ENABLE(Condition1) || ENABLE(Condition3)
 738             PASS: (CPP) WebDOMFloat64Array.cpp
 739             PASS: (CPP) WebDOMFloat64Array.h
 740
 741             Some tests FAIL! (To update the reference files, execute "run-bindings-tests --reset-results")
 742
 743         * Scripts/run-bindings-tests:
 744         (generate_from_idl):
 745         (detect_changes):
 746         (run_tests):
 747         (main):
 748
 749 2011-11-24  Kentaro Hara  <haraken@chromium.org>
 750
 751         Fix the current working directory of run-bindings-tests
 752         https://bugs.webkit.org/show_bug.cgi?id=73106
 753
 754         Reviewed by Adam Barth.
 755
 756         Currently, run-bindings-tests outputs the following error
 757         (although it outputs "all tests passed!" in the end):
 758
 759             Testing the ObjC generator on TestObj.idl
 760             gcc: WebCore/bindings/objc/PublicDOMInterfaces.h: No such file or directory
 761             gcc: warning: '-x objective-c' after last input file has no effect
 762             gcc: no input files
 763
 764         This is because CodeGeneratorObjC.pm assumes that the current working directory
 765         is WebKit/Source/, but run-bindings-tests sets the current working directory to WebKit/.
 766         Thus, this patch changes it to WebKit/Source/.
 767
 768         * Scripts/run-bindings-tests:
 769         (generate_from_idl):
 770         (main):
 771
 772 2011-11-24  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
 773
 774         [Qt] Get rid of the buildDirForSource() function in the build system
 775
 776         At some point the plan was to allow for running qmake on sub-trees
 777         of the sources, into the top level build directory, but this is no
 778         longer possible, so no reason to keep the convenience function around.
 779
 780         Reviewed by Simon Hausmann.
 781
 782         * WebKitTestRunner/DerivedSources.pri:
 783         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
 784         * qmake/mkspecs/features/default_post.prf:
 785         * qmake/mkspecs/features/default_pre.prf:
 786         * qmake/mkspecs/features/functions.prf:
 787         * qmake/mkspecs/features/javascriptcore.prf:
 788         * qmake/mkspecs/features/webcore.prf:
 789         * qmake/mkspecs/features/webkit2.prf:
 790
 791 2011-11-23  Sheriff Bot  <webkit.review.bot@gmail.com>
 792
 793         Unreviewed, rolling out r101107.
 794         http://trac.webkit.org/changeset/101107
 795         https://bugs.webkit.org/show_bug.cgi?id=73062
 796
 797         Breaks the commit-queue (Requested by abarth on #webkit).
 798
 799         * Scripts/webkitpy/common/host.py:
 800         (Host._engage_awesome_windows_hacks):
 801         * Scripts/webkitpy/common/host_mock.py:
 802         (MockHost.__init__):
 803         * Scripts/webkitpy/common/system/environment.py: Removed.
 804         * Scripts/webkitpy/common/system/environment_unittest.py: Removed.
 805         * Scripts/webkitpy/common/system/executive.py:
 806         (Executive._run_command_with_teed_output):
 807         (Executive.run_and_throw_if_fail):
 808         * Scripts/webkitpy/common/system/executive_mock.py:
 809         (MockExecutive.run_and_throw_if_fail):
 810         (MockExecutive.run_command):
 811         * Scripts/webkitpy/layout_tests/port/webkit.py:
 812         (WebKitPort._driver_build_script_name):
 813         (WebKitPort._run_script):
 814         (WebKitPort._build_driver):
 815         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
 816         (test_build_driver):
 817         * Scripts/webkitpy/tool/commands/download_unittest.py:
 818         * Scripts/webkitpy/tool/steps/build.py:
 819         (Build.build):
 820
 821 2011-11-23  Ojan Vafai  <ojan@chromium.org>
 822
 823         Unreviewed, rolling out r101110.
 824         http://trac.webkit.org/changeset/101110
 825         https://bugs.webkit.org/show_bug.cgi?id=73057
 826
 827         The bots have now cycled and this patch is no longer needed.
 828
 829         * Scripts/webkitpy/layout_tests/controllers/manager.py:
 830         (Manager._upload_json_files):
 831
 832 2011-11-23  Ojan Vafai  <ojan@chromium.org>
 833
 834         Temporarily save the incremental results json in order to identify bug
 835         https://bugs.webkit.org/show_bug.cgi?id=73057
 836
 837         Reviewed by Dean Jackson.
 838
 839         The test results server needed to be rolled back from tip of tree
 840         due to a JSON merging bug. It's hard to identify what the bug is without
 841         the raw data.
 842
 843         * Scripts/webkitpy/layout_tests/controllers/manager.py:
 844         (Manager._upload_json_files):
 845
 846 2011-11-23  Ojan Vafai  <ojan@chromium.org>
 847
 848         Merging two version 4 results hits an error in the test results server
 849         https://bugs.webkit.org/show_bug.cgi?id=73042
 850
 851         Reviewed by Adam Barth.
 852
 853         The old code assumed that if the aggregate results value was not a leaf
 854         value that the incremental results would exist, which is just totally
 855         wrong. Added a test for that case as well.
 856
 857         * TestResultServer/model/jsonresults.py:
 858         (JsonResults._merge_tests):
 859         * TestResultServer/model/jsonresults_unittest.py:
 860         (JsonResultsTest.test_merge_build_directory_hierarchy_old_version):
 861         (JsonResultsTest.test_merge_build_directory_hierarchy):
 862
 863 2011-11-23  Eric Seidel  <eric@webkit.org>
 864
 865         Add Environment object to Host and fix the GCC smartquotes trouble seen on the commit-queue
 866         https://bugs.webkit.org/show_bug.cgi?id=71983
 867
 868         Reviewed by Adam Barth.
 869
 870         We'll add more code to Environment overtime,
 871         allowing us to mock out more of our direct interactions with os.environ.
 872
 873         This patch also makes run_command print the passed in environment.
 874
 875         * Scripts/webkitpy/common/host_mock.py:
 876         * Scripts/webkitpy/common/system/environment.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
 877         * Scripts/webkitpy/common/system/environment_mock.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
 878         * Scripts/webkitpy/common/system/environment_unittest.py: Copied from Tools/Scripts/webkitpy/tool/steps/build.py.
 879         * Scripts/webkitpy/common/system/executive.py:
 880         * Scripts/webkitpy/common/system/executive_mock.py:
 881         * Scripts/webkitpy/layout_tests/port/webkit.py:
 882         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
 883         * Scripts/webkitpy/tool/commands/download_unittest.py:
 884         * Scripts/webkitpy/tool/steps/build.py:
 885
 886 2011-11-23  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
 887
 888         [Qt] Re-generate QtWebKit API forwarding headers when API changes
 889
 890         The forwarding headers are generated by syncqt, and syncqt is run by qmake,
 891         so we need sync.profile to live in the same directory as the project file
 892         that will be touched when adding/removing API (api.pri in our case).
 893
 894         Since the WebKit2 APIs live in a separate static library, we have to
 895         add the project file for WebKit2 as a dependency of the api.pri file
 896         as well, so that any changes to the WebKit2 API will still trigger a
 897         run of qmake (and then syncqt).
 898
 899         Lastly, we use the new QMAKE_SYNCQT_OUTDIR variable in Qt 5 to ensure
 900         that the forwarding headers are still generated in the root build dir,
 901         not in the Source dir along with the internal forwarding headers.
 902
 903         Reviewed by Andreas Kling.
 904
 905         * qmake/mkspecs/features/default_pre.prf:
 906         * qmake/mkspecs/features/functions.prf:
 907
 908 2011-11-23  Johnny Ding  <jnd@chromium.org>
 909
 910         Set right default value to baseURL in LayoutTestController::queueLoadHTMLString.
 911         https://bugs.webkit.org/show_bug.cgi?id=72950
 912
 913         Reviewed by Tony Chang.
 914
 915         * DumpRenderTree/chromium/LayoutTestController.cpp:
 916         (LayoutTestController::queueLoadHTMLString):
 917
 918 2011-11-23  Adam Barth  <abarth@webkit.org>
 919
 920         garden-o-matic should display a message if it can't find results to display
 921         https://bugs.webkit.org/show_bug.cgi?id=72991
 922
 923         Reviewed by Dimitri Glazkov.
 924
 925         Not all failures (e.g., TIMEOUT) have results that we can display.
 926         This patch adds a short message so that the user can see that we're
 927         done trying to load the results.
 928
 929         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
 930         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
 931         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
 932
 933 2011-11-23  Adam Barth  <abarth@webkit.org>
 934
 935         garden-o-matic should explain more clearly what the builder lists at the top of the page mean
 936         https://bugs.webkit.org/show_bug.cgi?id=72992
 937
 938         Reviewed by Dimitri Glazkov.
 939
 940         This patch also prepares us to list other sorts of failures, like
 941         performance and unit test.
 942
 943         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
 944         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
 945         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
 946         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
 947
 948 2011-11-23  Ryan Sleevi  <rsleevi@chromium.org>
 949
 950         Add new Chromium tests to the flakiness dashboard
 951         https://bugs.webkit.org/show_bug.cgi?id=72997
 952
 953         Reviewed by Ojan Vafai.
 954
 955         * TestResultServer/static-dashboards/dashboard_base.js:
 956         ():
 957
 958 2011-11-23  Adam Barth  <abarth@webkit.org>
 959
 960         garden-o-matic should support AUDIO failures
 961         https://bugs.webkit.org/show_bug.cgi?id=69477
 962
 963         Reviewed by Dimitri Glazkov.
 964
 965         This is a first iteration at supporting AUDIO failures in
 966         garden-o-matic.  We'll probably need to iterate a bit when we have a
 967         real audio failure in the tree to play with.  For example, we'll
 968         probably want to add and <audio> element to the "Examine" view to hear
 969         the differences, but that all can be done in a subsequent patch.
 970
 971         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
 972
 973 2011-11-23  Scott Graham  <scottmg@chromium.org>
 974
 975         Adding gamepad support
 976         https://bugs.webkit.org/show_bug.cgi?id=69451
 977
 978         Reviewed by Darin Fisher.
 979
 980         Add a 'gamepadController' to Chromium DRT. Allows for setting of mock
 981         gamepad data that will be retrieved when JS code pulls via API access.
 982
 983         * DumpRenderTree/DumpRenderTree.gypi:
 984         * DumpRenderTree/chromium/GamepadController.cpp: Added.
 985         (GamepadController::GamepadController):
 986         (GamepadController::bindToJavascript):
 987         (GamepadController::reset):
 988         (GamepadController::connect):
 989         (GamepadController::disconnect):
 990         (GamepadController::setId):
 991         (GamepadController::setButtonCount):
 992         (GamepadController::setButtonData):
 993         (GamepadController::setAxisCount):
 994         (GamepadController::setAxisData):
 995         (GamepadController::fallbackCallback):
 996         * DumpRenderTree/chromium/GamepadController.h: Added.
 997         * DumpRenderTree/chromium/TestShell.cpp:
 998         (TestShell::TestShell):
 999         (TestShell::resetTestController):
1000         (TestShell::bindJSObjectsToWindow):
1001         * DumpRenderTree/chromium/TestShell.h:
1002         (TestShell::gamepadController):
1003
1004 2011-11-23  Raphael Kubo da Costa  <kubo@profusion.mobi>
1005
1006         [CMake] Move the top-level logic to the top-level directory.
1007         https://bugs.webkit.org/show_bug.cgi?id=72685
1008
1009         Reviewed by Brent Fulgham.
1010
1011         * CMakeLists.txt: Added.
1012         * DumpRenderTree/efl/CMakeLists.txt: Keep building into Programs/.
1013         * EWebLauncher/CMakeLists.txt: Renamed from Tools/CMakeListsEfl.txt.
1014         * Scripts/webkitdirs.pm:
1015         (generateBuildSystemFromCMakeProject):
1016         * WinCELauncher/CMakeLists.txt: Renamed from Tools/CMakeListsWinCE.txt.
1017
1018 2011-11-17  Vincent Scheib  <scheib@chromium.org>
1019
1020         Pointer Lock: Initial Tests for navigator.webkitPonter
1021         https://bugs.webkit.org/show_bug.cgi?id=72659
1022
1023         Reviewed by Darin Fisher.
1024
1025         * DumpRenderTree/chromium/TestShell.cpp:
1026         (TestShell::TestShell):
1027             Enabling PointerLock in TestShell
1028
1029 2011-11-22  Ryan Sleevi  <rsleevi@chromium.org>
1030
1031         Update flakiness dashboard for new Chromium builders and tests
1032         https://bugs.webkit.org/show_bug.cgi?id=72931
1033
1034         Reviewed by Ojan Vafai.
1035
1036         * TestResultServer/static-dashboards/builders.js:
1037         * TestResultServer/static-dashboards/dashboard_base.js:
1038
1039 2011-11-22  Adam Roben  <aroben@apple.com>
1040
1041         Remove some debugging code I accidentally left in r101005
1042
1043         * Scripts/prepare-ChangeLog:
1044         (get_function_line_ranges_for_python):
1045
1046 2011-11-22  Adam Roben  <aroben@apple.com>
1047
1048         Teach prepare-ChangeLog how to find changed classes/methods/functions in Python files
1049
1050         Fixes <http://webkit.org/b/57008> prepare-ChangeLog doesn't find names of modified
1051         classes/methods in Python source files
1052
1053         Reviewed by Dan Bates.
1054
1055         * Scripts/prepare-ChangeLog:
1056         (get_function_line_ranges): Call get_function_line_ranges_for_python for files ending in .py
1057         or that use python as their interpreter. Modified the code that extracts the interpreter to
1058         ignore a leading "/usr/bin/env " in the interpreter line, as is common for our Python
1059         scripts.
1060         (get_function_line_ranges_for_python): Added. Does simple parsing of Python files to look
1061         for class/def lines and generate ranges based on them.
1062
1063 2011-11-21  Yuta Kitamura  <yutak@chromium.org>
1064
1065         [GTK] Enable WebSocket hybi tests
1066         https://bugs.webkit.org/show_bug.cgi?id=72870
1067
1068         Reviewed by Xan Lopez.
1069
1070         * DumpRenderTree/gtk/DumpRenderTree.cpp:
1071         (resetDefaultsToConsistentValues):
1072         Reset the setting value to the default value (true).
1073         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
1074         (LayoutTestController::overridePreference):
1075         Change the setting value through a function in DumpRenderTreeSupportGtk.
1076
1077 2011-11-21  Ryosuke Niwa  <rniwa@webkit.org>
1078
1079         ChangeLogEntry should be able to parse entries with multiple authors
1080         https://bugs.webkit.org/show_bug.cgi?id=72690
1081
1082         Reviewed by Eric Seidel.
1083
1084         Support multi-author change log entries in ChangeLogEntry. Split author names using
1085         the same logic as splitting reviewer names. _authors now stores list of dictionaries
1086         with "name" and "email" as keys. Unlike reviewers, we can't always resolve authors to
1087         accounts because some contributors are never listed in committers.py.
1088
1089         * Scripts/webkitpy/common/checkout/changelog.py:
1090         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
1091         * Scripts/webkitpy/common/config/committers.py: Fixed a bug that *_by_{name|email}
1092         throws an exception when the argument is None.
1093
1094 2011-11-21  Simon Hausmann  <simon.hausmann@nokia.com>
1095
1096         [Qt] Speed up debug builds.
1097         https://bugs.webkit.org/show_bug.cgi?id=72882
1098
1099         Reviewed by Tor Arne Vestbø.
1100
1101         Introduce a special type of build that replaces the static jsc/wc/wk2
1102         static libraries with shared libraries, that individually are much faster
1103         to link and don't require relinking into the big library when changing just
1104         one file in one module. It can be enabled with
1105         --qmakearg="CONFIG+=force_static_libs_as_shared"
1106
1107         * qmake/mkspecs/features/force_static_libs_as_shared.prf: Added. With this configuration,
1108         disable symbol visibility, static linkage and enable rpath.
1109         * qmake/mkspecs/features/functions.prf: Add a helper function to determine if
1110         we need to link the convenience libraries together or not and renamed linkStaticLibrary
1111         to just linkLibrary (because linkage is not necessarily static).
1112         * qmake/mkspecs/features/qtwebkit.prf: When using the trick, make sure to also
1113         link against jsc/wc/wk2 when linking app code against QtWebKit.
1114         * qmake/mkspecs/features/javascriptcore.prf: Use new +needToLinkLibrary helper
1115         function.
1116         * qmake/mkspecs/features/webcore.prf: Ditto.
1117         * qmake/mkspecs/features/webkit2.prf: Ditto.
1118
1119 2011-11-21  Adam Klein  <adamk@chromium.org>
1120
1121         Update flakiness dashboard linkification regexps for crbug.com/100000 and above
1122         https://bugs.webkit.org/show_bug.cgi?id=72898
1123
1124         Reviewed by Ojan Vafai.
1125
1126         Also removed a comment that was going to be more trouble to keep up to
1127         date than seemed worthwhile (the code is now the documentation).
1128
1129         * TestResultServer/static-dashboards/flakiness_dashboard.html:
1130
1131 2011-11-21  Dominic Mazzoni  <dmazzoni@google.com>
1132
1133         [Chromium] WebAccessibilityNotification should use AssertMatchingEnums.
1134         https://bugs.webkit.org/show_bug.cgi?id=72895
1135
1136         Reviewed by Chris Fleizach.
1137
1138         * DumpRenderTree/chromium/WebViewHost.cpp:
1139         (WebViewHost::postAccessibilityNotification):
1140
1141 2011-11-21  Philippe Normand  <pnormand@igalia.com>
1142
1143         Unreviewed, run-gtk-tests temporary workaround after r98500.
1144
1145         * Scripts/run-gtk-tests: Try to find the build directory like in
1146         build-gtkdoc, until webkit-build-directory --configuration gets fixed.
1147
1148 2011-11-21  Carlos Garcia Campos  <cgarcia@igalia.com>
1149
1150         [GTK] Add a way to skip unit tests in the bots
1151         https://bugs.webkit.org/show_bug.cgi?id=72879
1152
1153         Reviewed by Philippe Normand.
1154
1155         Rewrite the script in python to make it easier for everybody to
1156         change it and add support for skipping tests.
1157
1158         * Scripts/run-gtk-tests:
1159
1160 2011-11-21  Raphael Kubo da Costa  <kubo@profusion.mobi>
1161
1162         [EFL] Add more appcache functions to ewk_settings.
1163         https://bugs.webkit.org/show_bug.cgi?id=72143
1164
1165         Reviewed by Kenneth Rohde Christiansen.
1166
1167         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
1168         (LayoutTestController::clearAllApplicationCaches):
1169         (LayoutTestController::setAppCacheMaximumSize): Use the right
1170         function, this is not related to the memory cache, but rather to the
1171         appcache.
1172
1173 2011-11-18  Adam Roben  <aroben@apple.com>
1174
1175         Make NRWT consider all lines of output from parse-malloc-history, not just the first
1176
1177         Fixes <http://webkit.org/b/72652> REGRESSION (NRWT): NRWT always reports "1 unique leaks
1178         found" even when there are more than 1 unique leaks
1179
1180         Reviewed by Darin Adler.
1181
1182         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
1183         (LeakDetector.count_total_bytes_and_unique_leaks): Pass re.MULTILINE to re.findall so that
1184         "^" will be considered to match the beginning of any line, not only the beginning of the
1185         string.
1186
1187         * Scripts/webkitpy/layout_tests/port/leakdetector_unittest.py:
1188         (LeakDetectorTest.test_count_total_bytes_and_unique_leaks): Added a second callstack to the
1189         fake output and updated expectations to match.
1190
1191 2011-11-21  Andras Becsi  <andras.becsi@nokia.com>
1192
1193         [Qt] Clean up MiniBrowser project file after recent changes
1194         https://bugs.webkit.org/show_bug.cgi?id=72869
1195
1196         Reviewed by Simon Hausmann.
1197
1198         Remove nonexistent files from OTHER_FILES after our touch and desktop webviews were merged.
1199
1200         * MiniBrowser/qt/MiniBrowser.pro:
1201
1202 2011-11-21  Dominic Mazzoni  <dmazzoni@google.com>
1203
1204         Accessibility: Multiselect list boxes need to report the active option in addition to which items are selected.
1205         https://bugs.webkit.org/show_bug.cgi?id=72479
1206
1207         Reviewed by Chris Fleizach.
1208
1209         * DumpRenderTree/AccessibilityUIElement.cpp:
1210         (getIsSelectedOptionActiveCallback):
1211         (AccessibilityUIElement::getJSClass):
1212         * DumpRenderTree/AccessibilityUIElement.h:
1213         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
1214         (AccessibilityUIElement::AccessibilityUIElement):
1215         (AccessibilityUIElement::isSelectedOptionActiveGetterCallback):
1216         * DumpRenderTree/chromium/AccessibilityUIElement.h:
1217         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
1218         (AccessibilityUIElement::isSelectedOptionActive):
1219         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
1220         (AccessibilityUIElement::isSelectedOptionActive):
1221         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
1222         (AccessibilityUIElement::isSelectedOptionActive):
1223
1224 2011-11-21  Takashi Toyoshima  <toyoshim@chromium.org>
1225
1226         Use Host().port_factory instead of webkitpy.layout_tests.port.factory.
1227         https://bugs.webkit.org/show_bug.cgi?id=72859
1228
1229         Reviewed by Eric Seidel.
1230
1231         * Scripts/new-run-webkit-websocketserver:
1232
1233 2011-11-20  Shinya Kawanaka  <shinyak@google.com>
1234
1235         [Chromium] Layout test does not return correct misspelling positions if there are multiple misspelled words.
1236         https://bugs.webkit.org/show_bug.cgi?id=72655
1237
1238         Reviewed by Hajime Morita.
1239
1240         Fixed the calculation of offset.
1241
1242         * DumpRenderTree/chromium/WebViewHost.cpp:
1243         (WebViewHost::finishLastTextCheck):
1244
1245 2011-11-20  Ojan Vafai  <ojan@chromium.org>
1246
1247         Change the final place where we use version 3 of the results json output
1248         https://bugs.webkit.org/show_bug.cgi?id=72838
1249
1250         Reviewed by Adam Barth.
1251
1252         This converts the json from being a flat map of test name --> results
1253         to being hierarchical by directory. This will make the json files
1254         considerably smaller.
1255
1256         Also cleaned up some functions that were returning/checking boolean values
1257         that were always True.
1258
1259         * TestResultServer/model/jsonresults.py:
1260         * TestResultServer/model/jsonresults_unittest.py:
1261
1262 2011-11-20  Ojan Vafai  <ojan@chromium.org>
1263
1264         Include the master name when querying the test results server.
1265         This will soon be required since leaving the master name out mean
1266         that multiple files could match the query.
1267
1268         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1269
1270 2011-11-19  Ojan Vafai  <ojan@chromium.org>
1271
1272         Fix jsonp callback code to add the callback if json content
1273         is not jsonp.
1274         * TestResultServer/handlers/testfilehandler.py:
1275
1276 2011-11-19  Alice Boxhall  <aboxhall@chromium.org>
1277
1278         Modify dashboard_base.js to optionally accept version 4 results json.
1279
1280         Reviewed by Ojan Vafai.
1281
1282         * TestResultServer/static-dashboards/dashboard_base.js:
1283         (ADD_RESULTS):
1284         (flattenTrie):
1285         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
1286         (testFlattenTrie):
1287
1288 2011-11-19  Ojan Vafai  <ojan@chromium.org>
1289
1290         Stop storing results files as jsonp in the test results server
1291         https://bugs.webkit.org/show_bug.cgi?id=72814
1292
1293         Reviewed by Adam Barth.
1294
1295         * TestResultServer/model/jsonresults.py:
1296         * TestResultServer/model/jsonresults_unittest.py:
1297
1298 2011-11-19  Ojan Vafai  <ojan@chromium.org>
1299
1300         Remove the dependence on jsonp from more of new-run-webkit-tests and the test results server
1301         https://bugs.webkit.org/show_bug.cgi?id=72813
1302
1303         Reviewed by Adam Barth.
1304
1305         Once this lands, we can start storing pure json in the test results server and then
1306         we can delete the code with all the FIXMEs added here.
1307
1308         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1309         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1310         Only add jsonp for full_results.json.
1311         * TestResultServer/model/jsonresults.py:
1312         * TestResultServer/model/jsonresults_unittest.py:
1313         Accept pure json uploads.
1314         * TestResultServer/static-dashboards/dashboard_base.js:
1315         (appendJSONScriptElementFor):
1316         Use the callback parameter so that the server can start returning pure json if it's left out.
1317
1318 2011-11-19  Ojan Vafai  <ojan@chromium.org>
1319
1320         Allow json NRWT downloads to be pure json and not jsonp
1321         https://bugs.webkit.org/show_bug.cgi?id=72809
1322
1323         Reviewed by Adam Barth.
1324
1325         I'm moving the server to storing and serving up raw json instead of jsonp.
1326         You can still get the jsonp by passing a "callback" parameter, but there's no
1327         need for run-webkit-tests to get jsonp when all it wants is the raw json.
1328
1329         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1330         Make the wrapper stripping only happen if the json is actually wrapped.
1331         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
1332
1333 2011-11-18  Daniel Bates  <dbates@rim.com>
1334
1335         Add CMake build infrastructure for the BlackBerry port
1336         https://bugs.webkit.org/show_bug.cgi?id=72768
1337
1338         Reviewed by Antonio Gomes.
1339
1340         Add optional --blackberry command line argument to build-webkit to build the
1341         BlackBerry port.
1342
1343         * Scripts/build-webkit:
1344         * Scripts/webkitdirs.pm:
1345         (determineBaseProductDir):
1346         (argumentsForConfiguration):
1347         (builtDylibPathForName):
1348         (determineIsQt):
1349         (isBlackBerry): Added.
1350         (determineIsBlackBerry): Added.
1351         (blackberryTargetArchitecture): Added.
1352         (isAppleWebKit):
1353
1354 2011-11-18  James Robinson  <jamesr@chromium.org>
1355
1356         Unreviewed, rolling out r100835.
1357         http://trac.webkit.org/changeset/100835
1358         https://bugs.webkit.org/show_bug.cgi?id=72778
1359
1360         Broke some page scale tests
1361
1362         * DumpRenderTree/chromium/EventSender.cpp:
1363         (EventSender::scalePageBy):
1364
1365 2011-11-18  Adam Barth  <abarth@webkit.org>
1366
1367         Prepare to move manual-tests out of WebCore
1368         https://bugs.webkit.org/show_bug.cgi?id=72782
1369
1370         Reviewed by Eric Seidel.
1371
1372         Update this file to refer to the new location of ManualTests.
1373
1374         * Scripts/webkitpy/common/config/build.py:
1375
1376 2011-11-18  James Robinson  <jamesr@chromium.org>
1377
1378         [chromium] Set min/max page scale factors when calling eventSender.scalePageBy to avoid clamping
1379         https://bugs.webkit.org/show_bug.cgi?id=72778
1380
1381         Reviewed by Kenneth Russell.
1382
1383         Set the min/max page scale clamps when a test calls eventSender.scalePageBy() so that the scale isn't clamped.
1384
1385         * DumpRenderTree/chromium/EventSender.cpp:
1386         (EventSender::scalePageBy):
1387
1388 2011-11-18  Scott Graham  <scottmg@chromium.org>
1389
1390         IDL changes for gamepad support
1391         https://bugs.webkit.org/show_bug.cgi?id=71753
1392
1393         Reviewed by Adam Barth.
1394
1395         Runtime enable gamepad in chromium test shell.
1396
1397         * DumpRenderTree/chromium/TestShell.cpp:
1398         (TestShell::TestShell):
1399
1400 2011-10-27  Darin Fisher  <darin@chromium.org>
1401
1402         [chromium] Improve fullscreen API
1403         https://bugs.webkit.org/show_bug.cgi?id=70477
1404
1405         Reviewed by Adam Barth.
1406
1407         Implement enter/exitFullScreen.  The implementation is asynchronous,
1408         via postDelayedTask, to simulate how the browser actually behaves.
1409
1410         * DumpRenderTree/chromium/WebViewHost.cpp:
1411         (WebViewHost::enterFullScreen):
1412         (WebViewHost::exitFullScreen):
1413         (WebViewHost::setAddressBarURL):
1414         (WebViewHost::enterFullScreenNow):
1415         (WebViewHost::exitFullScreenNow):
1416         * DumpRenderTree/chromium/WebViewHost.h:
1417
1418 2011-11-18  Adam Roben  <aroben@apple.com>
1419
1420         Ignore an ANGLE leak that is not WebKit's fault
1421
1422         * Scripts/old-run-webkit-tests:
1423         (countAndPrintLeaks):
1424         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
1425         (LeakDetector._callstacks_to_exclude_from_leaks):
1426         Added ScanFromString to the call stacks to exclude on all OS versions.
1427
1428 2011-11-18  Adam Roben  <aroben@apple.com>
1429
1430         Ignore another Lion-specific leak that is not WebKit's fault
1431
1432         * Scripts/old-run-webkit-tests:
1433         (countAndPrintLeaks):
1434         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
1435         (LeakDetector._callstacks_to_exclude_from_leaks):
1436         Added SecTransformExecute to the call stacks to exclude.
1437
1438 2011-11-18  Yuta Kitamura  <yutak@chromium.org>
1439
1440         [Qt] Enable WebSocket hybi tests
1441         https://bugs.webkit.org/show_bug.cgi?id=72687
1442
1443         Reviewed by Simon Hausmann.
1444
1445         Let LayoutTestControllerQt be able to switch WebSocket protocols via overridePreference().
1446
1447         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1448         (WebCore::WebPage::resetSettings):
1449         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1450         (LayoutTestController::overridePreference):
1451
1452 2011-11-18  Sergio Villar Senin  <svillar@igalia.com>
1453
1454         [GTK][WK2] Invalid TestNetscapePlugin path when using new-run-webkit-tests
1455         https://bugs.webkit.org/show_bug.cgi?id=72707
1456
1457         Reviewed by Philippe Normand.
1458
1459         Make new-run-webkit-tests use the right path to locate the test
1460         plugin used by WebKitTestRunner.
1461
1462         * Scripts/webkitpy/layout_tests/port/gtk.py:
1463
1464 2011-11-18  Simon Hausmann  <simon.hausmann@nokia.com>
1465
1466         [Qt] WTR and InjectedBundle should not link statically against JSC/WTF
1467         https://bugs.webkit.org/show_bug.cgi?id=72697
1468
1469         Reviewed by Kenneth Rohde Christiansen.
1470
1471         * Tools.pro: Make WTR linux only for the moment.
1472         * WebKitTestRunner/InjectedBundle/Target.pri: Remove static jsc/wtf linkage.
1473         * WebKitTestRunner/Target.pri: Ditto.
1474         * qmake/mkspecs/features/default_post.prf: Don't build the main webkit sources
1475         with ELF visibility on Linux. Instead we're temporarily using an LD version script.
1476
1477 2011-11-18  Kenneth Rohde Christiansen  <kenneth@webkit.org>
1478
1479         [Qt] Change default size of the MiniBrowser
1480
1481         Reviewed by Simon Hausmann.
1482
1483         As we lay out desktop pages using the width of 980, change our default
1484         size to 980x735 so that we are not scaling down by default.
1485
1486         * MiniBrowser/qt/MiniBrowserApplication.h:
1487         (WindowOptions::WindowOptions):
1488
1489 2011-11-17  Raphael Kubo da Costa  <kubo@profusion.mobi>
1490
1491         [EFL] Clean up the use of DATA_DIR in the buildsystem
1492         https://bugs.webkit.org/show_bug.cgi?id=72681
1493
1494         Reviewed by Daniel Bates.
1495
1496         Instead of adding -DDATA_DIR in all scopes, only define it in
1497         WebCore/PlatformEFL.cmake, as WebCore is the only place that uses
1498         it.
1499
1500         We can then define DATA_DIR to the generated theme directory for
1501         EWebLauncher, since it's what we want here.
1502
1503         * CMakeListsEfl.txt:
1504
1505 2011-11-17  Raphael Kubo da Costa  <kubo@profusion.mobi>
1506
1507         [EFL] Move platform includes last in DRT's include directories.
1508         https://bugs.webkit.org/show_bug.cgi?id=72684
1509
1510         Reviewed by Daniel Bates.
1511
1512         This is the first step in fixing the build when a platform include (in
1513         this case, Valgrind 3.7.0 ships its config.h) and the wrong config.h
1514         is included instead of WebKit's.
1515
1516         * DumpRenderTree/efl/CMakeLists.txt:
1517
1518 2011-11-17  Ryosuke Niwa  <rniwa@webkit.org>
1519
1520         fuzzy_match doesn't recognize "Dan B" or "hyatt"
1521         https://bugs.webkit.org/show_bug.cgi?id=72636
1522
1523         Reviewed by Eric Seidel.
1524
1525         Make contributors_by_fuzz_match search string in shorthand names of contributors such as "Ryosuke N"
1526         and user names of email addreses such as rniwa in rniwa@webkit.org.
1527
1528         Also import legacy contributor names and mispell lists from Eric's script on the bug 26533 as a test.
1529
1530         * Scripts/webkitpy/common/config/committers.py:
1531         * Scripts/webkitpy/common/config/committers_unittest.py:
1532
1533 2011-11-17  Eric Seidel  <eric@webkit.org>
1534
1535         Remove support for running check-webkit-style w/o a webkit checkout
1536         https://bugs.webkit.org/show_bug.cgi?id=72678
1537
1538         Reviewed by Adam Barth.
1539
1540         After talking to Dave Levin we've decided to remove this
1541         code.  I know of no users of check-webkit-style w/o a webkit checkout
1542         and this needlessly adds complexity.
1543
1544         * Scripts/webkitpy/style/main.py:
1545
1546 2011-11-17  Eric Seidel  <eric@webkit.org>
1547
1548         Remove more direct uses of os.path
1549         https://bugs.webkit.org/show_bug.cgi?id=72677
1550
1551         Reviewed by Adam Barth.
1552
1553         Remove more uses of os.path, including refactoring
1554         CommitterValidator to use a Host object instead of
1555         manual hacks to emulate what SCM and FileSystem provide.
1556
1557         * Scripts/webkitpy/common/checkout/changelog.py:
1558         * Scripts/webkitpy/common/config/committervalidator.py:
1559         * Scripts/webkitpy/common/config/committervalidator_unittest.py:
1560         * Scripts/webkitpy/common/system/filesystem_mock.py:
1561          - Turns out none of our unittests actually looked at this path.
1562            now I've made it be a valid path.
1563         * Scripts/webkitpy/tool/bot/feeders.py:
1564         * Scripts/webkitpy/tool/commands/queues.py:
1565
1566 2011-11-17  Eric Seidel  <eric@webkit.org>
1567
1568         Teach TextFileReader about FileSystem
1569         https://bugs.webkit.org/show_bug.cgi?id=72673
1570
1571         Reviewed by Adam Barth.
1572
1573         Unfortunately TextFileReader doesn't use FileSystem
1574         everywhere yet, so we can't move the unittests to
1575         using MockFileSystem, but we're close.
1576
1577         * Scripts/webkitpy/style/filereader.py:
1578         * Scripts/webkitpy/style/filereader_unittest.py:
1579         * Scripts/webkitpy/style/main.py:
1580
1581 2011-11-17  Eric Seidel  <eric@webkit.org>
1582
1583         Give check-webkit-style a Host
1584         https://bugs.webkit.org/show_bug.cgi?id=72670
1585
1586         Reviewed by Adam Barth.
1587
1588         * Scripts/webkitpy/common/checkout/scm/detection.py:
1589         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
1590          - We don't run these unittests, but this one was looking for the wrong exception.
1591         * Scripts/webkitpy/style/main.py:
1592         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
1593          - This was create a new scm object even though it already had one?
1594
1595 2011-11-17  Eric Seidel  <eric@webkit.org>
1596
1597         Move check-webkit-style change_directory code to more modern mocks
1598         https://bugs.webkit.org/show_bug.cgi?id=72664
1599
1600         Reviewed by Adam Barth.
1601
1602         * Scripts/check-webkit-style:
1603         * Scripts/webkitpy/style/main.py:
1604         * Scripts/webkitpy/style/main_unittest.py:
1605
1606 2011-11-17  Zan Dobersek  <zandobersek@gmail.com>
1607
1608         [GTK] enable ref tests
1609         https://bugs.webkit.org/show_bug.cgi?id=72601
1610
1611         Re-enable ref tests. Run the driver under evenly numbered display if performing
1612         pixel tests or under oddly numbered display otherwise. This prevents the ref tests
1613         being run in an already used Xvfb instance if pixel tests are disabled.
1614
1615         Reviewed by Tony Chang.
1616
1617         * Scripts/webkitpy/layout_tests/port/gtk.py: Also set Xvfb variable to None
1618         after the process is killed
1619
1620 2011-11-17  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
1621
1622         Make check-webkit-style accept xxx_p.h as a primary header for xxx.cpp for Qt's sake
1623         https://bugs.webkit.org/show_bug.cgi?id=72620
1624
1625         Reviewed by David Levin.
1626
1627         Qt's convention of keeping private (but exposed) API suffixed by "_p.h" but not
1628         adding the suffix the cpp file. One example of false positive
1629         https://bugs.webkit.org/show_bug.cgi?id=72319#c12.
1630
1631         * Scripts/webkitpy/style/checkers/cpp.py: accept headers ending with _p as primary if its
1632         basename contains the cpp's basename.
1633         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
1634
1635 2011-11-17  Eric Seidel  <eric@webkit.org>
1636
1637         Move check-webkit-style's guts into webkitpy/style/main.py
1638         https://bugs.webkit.org/show_bug.cgi?id=72657
1639
1640         Reviewed by Adam Barth.
1641
1642         * Scripts/check-webkit-style:
1643         * Scripts/webkitpy/style/main.py:
1644
1645 2011-11-17  Eric Seidel  <eric@webkit.org>
1646
1647         new-run-webkit-tests is locale dependent
1648         https://bugs.webkit.org/show_bug.cgi?id=68691
1649
1650         Reviewed by Adam Barth.
1651
1652         Make NRWT use a clean environment, just like ORWT did.
1653
1654         * Scripts/webkitpy/layout_tests/port/base.py:
1655         * Scripts/webkitpy/layout_tests/port/chromium.py:
1656         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1657         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
1658         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
1659         * Scripts/webkitpy/layout_tests/port/efl.py:
1660         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
1661         * Scripts/webkitpy/layout_tests/port/mac.py:
1662         * Scripts/webkitpy/layout_tests/port/qt.py:
1663         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
1664         * Scripts/webkitpy/layout_tests/port/webkit.py:
1665         * Scripts/webkitpy/layout_tests/port/win.py:
1666
1667 2011-11-17  Dirk Pranke  <dpranke@chromium.org>
1668
1669         Chromium Mac 10.5 CG dbg bots are failing to run webkit_gpu_tests
1670         https://bugs.webkit.org/show_bug.cgi?id=72615
1671
1672         Reviewed by James Robinson.
1673
1674         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
1675
1676 2011-11-17  Simon Hausmann  <simon.hausmann@nokia.com>
1677
1678         [Qt] Layer violation: qt_runtime.cpp accesses QWebElement and QTDRTNode
1679         https://bugs.webkit.org/show_bug.cgi?id=72595
1680
1681         Reviewed by Noam Rosenthal.
1682
1683         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
1684         (WebCore::DumpRenderTree::DumpRenderTree): Use new initialize() function to
1685         ensure the registration of the DRTNode JS bindings.
1686         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
1687         (LayoutTestController::LayoutTestController): Remove unnecessary explicit qt meta type
1688         registration of QWebElement (now done implicitly through QWebPage ->
1689         QtWebElementRunTime::initialize code path)
1690
1691 2011-11-17  Adam Roben  <aroben@apple.com>
1692
1693         Fix a crash-inducing typo I introduced in r100648
1694
1695         Fixes <http://webkit.org/b/72642> REGRESSION (r100648): NRWT crashes on Chromium when
1696         handling a crashing test
1697
1698         Reviewed by Adam Barth.
1699
1700         * Scripts/webkitpy/layout_tests/port/chromium.py:
1701         (ChromiumDriver.run_test): Get the driver name from the port.
1702
1703         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
1704         (ChromiumDriverTest.test_crashed_process_name): Simulate a crash and check that we got the
1705         right crashed process name.
1706
1707 2011-11-17  Adam Roben  <aroben@apple.com>
1708
1709         Fix NRWT's parsing of the number of excluded leaks
1710
1711         Fixes <http://webkit.org/b/72635> NRWT crashes when parsing leaks files from which certain
1712         leaks were excluded
1713
1714         Reviewed by Adam Barth.
1715
1716         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
1717         (LeakDetector._parse_leaks_output): Extract just the number of excluded leaks from the
1718         regular expression match, not the entire matched string.
1719
1720         * Scripts/webkitpy/layout_tests/port/leakdetector_unittest.py:
1721         (LeakDetectorTest): Added some example output that shows how things look when leaks are
1722         excluded.
1723         (LeakDetectorTest.test_parse_leaks_output): Test that we parse the number of exclusions
1724         correctly.
1725
1726 2011-11-17  Adam Roben  <aroben@apple.com>
1727
1728         Make NRWT find crash logs for the crashed process, which may not necessarily be the driver process
1729
1730         Fixes <http://webkit.org/b/72526> REGRESSION (NRWT): WebKitTestRunner crash log gets saved
1731         when web process crashes, but WebProcess crash log should get saved instead
1732
1733         Reviewed by Eric Seidel.
1734
1735         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py:
1736         (write_test_result): Simplified a little to share more code between the case where we
1737         crashed when running the test and where we crashed when checking the expected rendering for
1738         a reftest. Changed to pass the crashed process name down to write_crash_report.
1739         (TestResultWriter.write_crash_report): Added crashed_process_name parameter, which we pass
1740         along to CrashLogs.find_newest_log rather than always using the driver name.
1741
1742         * Scripts/webkitpy/layout_tests/port/chromium.py:
1743         (ChromiumDriver.run_test): Pass the driver name as the name of the crashed process if we
1744         crashed.
1745
1746         * Scripts/webkitpy/layout_tests/port/test.py:
1747         (TestInstance.__init__): Initialize new web_process_crash attribute.
1748         (unit_test_list): Added new web-process-crash-with-stderr test, which is used by the new
1749         test this patch adds.
1750         (TestDriver.run_test): Pass a crashed process name to DriverOutput based on whether the test
1751         specifies that the driver crashed or the web process crashed.
1752
1753         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
1754         (MainTest.test_web_process_crash_log): Added. Similar to test_crash_log, but uses a test for
1755         which we pretend the web process crashed rather than the driver process.
1756
1757         (MainTest.test_run_singly_actually_runs_tests):
1758         (MainTest.test_unexpected_failures):
1759         (EndToEndTest.test_end_to_end):
1760         Updated magic numbers to account for new unexpected failing test added to unit_test_list.
1761
1762 2011-11-17  Tony Chang  <tony@chromium.org>
1763
1764         [chromium] stop spamming features warning on every build
1765         https://bugs.webkit.org/show_bug.cgi?id=72629
1766
1767         Reviewed by Ojan Vafai.
1768
1769         We toggle $blobSupport on if $mediaStreamSupport is on.  Since
1770         $mediaStreamSupport is on by default for chromium, make the default
1771         for $blobSupport also on.
1772
1773         * Scripts/build-webkit:
1774
1775 2011-11-17  Adam Roben  <aroben@apple.com>
1776
1777         Remove extra quoting of call stacks and types passed to run-leaks by NRWT
1778
1779         Fixes <http://webkit.org/b/72623> REGRESSION (NRWT): run-webkit-tests --leaks reports leaks
1780         that are supposed to be ignored
1781
1782         Reviewed by Adam Barth.
1783
1784         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
1785         (LeakDetector._leaks_args): Removed extra quoting of call stacks and types. Quoting is only
1786         required when passing arguments via the shell (which we aren't doing here). It is not
1787         required by Perl's option-parsing code (as the comment I removed mistakenly claimed).
1788
1789         * Scripts/webkitpy/layout_tests/port/leakdetector_unittest.py:
1790         (LeakDetectorTest.test_leaks_args): Updated expectations.
1791
1792 2011-11-17  Philippe Normand  <pnormand@igalia.com>
1793
1794         Unreviewed, run-bindings-tests fix after r91028.
1795
1796         * Scripts/run-bindings-tests: Use scm.detection module.
1797
1798 2011-11-17  Jochen Eisinger  <jochen@chromium.org>
1799
1800         Silence a warning about control reaching the end of a non-void function
1801         https://bugs.webkit.org/show_bug.cgi?id=72616
1802
1803         Reviewed by Tony Gentilcore.
1804
1805         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
1806         (AccessibilityUIElement::verticalScrollbar):
1807
1808 2011-11-17  Adam Roben  <aroben@apple.com>
1809
1810         Ignore some leaks in frameworks we link against on Lion
1811
1812         Fixes <http://webkit.org/b/72609> Lion Leaks bot complains about a bunch of leaks that
1813         aren't WebKit's fault
1814
1815         Reviewed by Antti Koivisto.
1816
1817         * Scripts/old-run-webkit-tests:
1818         (countAndPrintLeaks):
1819         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
1820         (LeakDetector._callstacks_to_exclude_from_leaks):
1821         Added some call stacks to exclude on Lion that represent leaks in lower-level frameworks.
1822
1823         * Scripts/webkitpy/layout_tests/port/mac.py:
1824         (MacPort.is_lion): Added.
1825
1826         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
1827         (MacPortTest.test_is_version_methods): Added tests for the is_leopard/is_snowleopard/is_lion
1828         methods.
1829
1830 2011-11-17  Philippe Normand  <pnormand@igalia.com>
1831
1832         [GTK] Disable ref tests
1833         https://bugs.webkit.org/show_bug.cgi?id=72599
1834
1835         Reviewed by Xan Lopez.
1836
1837         * Scripts/webkitpy/layout_tests/port/gtk.py: Enforce disabled ref
1838         tests to avoid massive flakiness.
1839
1840 2011-11-17  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
1841
1842         [Qt] Move WebKit2 C++ APIs to private API and build QML extension plugin on top of that
1843
1844         A new extension object has been added to QQuickWebView (the same approach should be used
1845         for other API classes that need experimental APIs). The QML extension mechanism is then
1846         built on top of the experimental object.
1847
1848         https://bugs.webkit.org/show_bug.cgi?id=72522
1849
1850         Reviewed by Simon Hausmann.
1851
1852         * MiniBrowser/qt/BrowserWindow.cpp:
1853         (BrowserWindow::BrowserWindow):
1854         (BrowserWindow::webView):
1855         * MiniBrowser/qt/BrowserWindow.h:
1856         * MiniBrowser/qt/UrlLoader.cpp:
1857         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
1858         * qmake/mkspecs/features/default_post.prf:
1859         * qmake/mkspecs/features/default_pre.prf:
1860         * qmake/mkspecs/features/qtwebkit-private.prf: Added.
1861         * qmake/mkspecs/features/qtwebkit.prf:
1862
1863 2011-11-17  Adam Barth  <abarth@webkit.org>
1864
1865         Move test_results_uploader.py out of layout_package
1866         https://bugs.webkit.org/show_bug.cgi?id=72590
1867
1868         Reviewed by Eric Seidel.
1869
1870         Most of the lines of code in this file are wrong, but I've restrained
1871         myself and only changed a few of them to generalized this class to the
1872         common package.
1873
1874         This is part of a series of patches to remove layout_package.
1875
1876         * Scripts/webkitpy/common/net/file_uploader.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results_uploader.py.
1877         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
1878         * Scripts/webkitpy/layout_tests/layout_package/test_results_uploader.py: Removed.
1879
1880 2011-11-17  Adam Barth  <abarth@webkit.org>
1881
1882         Unreviewed. Fix style nits in printing.py.
1883
1884         * Scripts/webkitpy/layout_tests/views/printing.py:
1885
1886 2011-11-17  Adam Barth  <abarth@webkit.org>
1887
1888         Minor style nits in run_webkit_tests.py
1889         https://bugs.webkit.org/show_bug.cgi?id=72583
1890
1891         Reviewed by Eric Seidel.
1892
1893         This patch just fixes some minor style issues as I work my way back
1894         into this code.
1895
1896         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1897
1898 2011-11-17  Adam Barth  <abarth@webkit.org>
1899
1900         Move test_result_writer out of layout_package
1901         https://bugs.webkit.org/show_bug.cgi?id=72586
1902
1903         Rubber-stamped by Eric Seidel.
1904
1905         This is part of a series of patches to remove layout_package.
1906
1907         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
1908         * Scripts/webkitpy/layout_tests/controllers/test_result_writer.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py.
1909         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py: Removed.
1910         * Scripts/webkitpy/tool/commands/rebaseline.py:
1911
1912 2011-11-17  Adam Barth  <abarth@webkit.org>
1913
1914         svn-apply shouldn't magically move files into Source anymore
1915         https://bugs.webkit.org/show_bug.cgi?id=72579
1916
1917         Reviewed by Eric Seidel.
1918
1919         The Source directory has existed for long enough that we don't need
1920         svn-apply to magically re-write old-style patches anymore.
1921
1922         * Scripts/VCSUtils.pm:
1923         (parseGitDiffHeader):
1924         (parseSvnDiffHeader):
1925
1926 2011-11-16  Ryosuke Niwa  <rniwa@webkit.org>
1927
1928         Add a list of contribution areas
1929         https://bugs.webkit.org/show_bug.cgi?id=72566
1930
1931         Reviewed by Eric Seidel.
1932
1933         Added ContributionAreas class.
1934
1935         * Scripts/webkitpy/common/config/contributionareas.py: Added.
1936         * Scripts/webkitpy/common/config/contributionareas_unittest.py: Added.
1937
1938 2011-11-16  Ryosuke Niwa  <rniwa@webkit.org>
1939
1940         Refactor ChangeLogTest.test_parse_reviewer_text
1941         https://bugs.webkit.org/show_bug.cgi?id=72572
1942
1943         Reviewed by Eric Seidel.
1944
1945         Refactored the test code by introduing two helper functions.
1946
1947         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
1948
1949 2011-11-16  Eric Seidel  <eric@webkit.org>
1950
1951         Add SCMDetector object to make scm detection mockable
1952         https://bugs.webkit.org/show_bug.cgi?id=72247
1953
1954         Reviewed by Adam Barth.
1955
1956         I think SCM detection may eventually be rolled into Checkout,
1957         but this patch at least makes it possible to mock code-paths
1958         which rely on scm detection.
1959         In the process of replacing callers of these free-functions
1960         I found that one of the functions was no longer used,
1961         and that one of the callers could instead just use the SCM
1962         object it already had access to through port.host.scm().
1963         I also discovered that I was not calling Host._initialize_scm()
1964         and thus host.scm() was always returning None!
1965
1966         * Scripts/check-webkit-style:
1967         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
1968         * Scripts/webkitpy/common/checkout/deps.py:
1969         * Scripts/webkitpy/common/checkout/scm/__init__.py:
1970         * Scripts/webkitpy/common/checkout/scm/detection.py:
1971         * Scripts/webkitpy/common/checkout/scm/git.py:
1972         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
1973         * Scripts/webkitpy/common/checkout/scm/svn.py:
1974         * Scripts/webkitpy/common/host.py:
1975         * Scripts/webkitpy/common/host_mock.py:
1976         * Scripts/webkitpy/layout_tests/controllers/manager.py:
1977         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
1978         * Scripts/webkitpy/layout_tests/port/base.py:
1979         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
1980         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
1981         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
1982         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
1983
1984 2011-11-16  Mark Rowe  <mrowe@apple.com>
1985
1986         Fix the build.
1987
1988         The C++ standard library interferes with Objective-C exceptions when built with -fno-exceptions.
1989         Apply the workaround that we use in other projects to WebKitTestRunner.
1990
1991         * WebKitTestRunner/Configurations/Base.xcconfig: Use the prefix header for all targets.
1992         * WebKitTestRunner/Configurations/WebKitTestRunner.xcconfig: Ditto.
1993         * WebKitTestRunner/WebKitTestRunnerPrefix.h: Work around the C++ standard library obnoxiousness.
1994
1995 2011-11-16  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
1996
1997         [Qt] Fix build after WTR AX support
1998         https://bugs.webkit.org/show_bug.cgi?id=72560
1999
2000         Reviewed by Geoffrey Garen.
2001
2002         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
2003         * WebKitTestRunner/InjectedBundle/Target.pri:
2004
2005 2011-11-16  Tony Chang  <tony@chromium.org>
2006
2007         repurpose the flexbox bot for grid layout
2008         https://bugs.webkit.org/show_bug.cgi?id=72557
2009
2010         Reviewed by Ryosuke Niwa.
2011
2012         Since new flexbox is enabled everywhere, we don't need this bot
2013         anymore.  However, work on grid layout has begun, so we can use this
2014         bot to compile with ENABLE_CSS_GRID_LAYOUT.
2015
2016         * BuildSlaveSupport/build.webkit.org-config/config.json: Just rename
2017         the bot on the waterfall, I will change the slave config to enable
2018         the compile flag and tests.
2019
2020 2011-11-11  Adrienne Walker  <enne@google.com>
2021
2022         [chromium] Expose mock scrollbars to window.internals
2023         https://bugs.webkit.org/show_bug.cgi?id=72195
2024
2025         Reviewed by James Robinson.
2026
2027         Because mock scrollbars are a global setting, reset between runs.
2028
2029         * DumpRenderTree/chromium/WebPreferences.cpp:
2030         (WebPreferences::applyTo):
2031
2032 2011-11-16  James Robinson  <jamesr@chromium.org>
2033
2034         [chromium] Enable video and canvas 2d compositing triggers by default for layout tests in the compositing directory
2035         https://bugs.webkit.org/show_bug.cgi?id=72562
2036
2037         Reviewed by Dirk Pranke.
2038
2039         * DumpRenderTree/chromium/TestShell.cpp:
2040         (TestShell::runFileTest):
2041
2042 2011-11-16  Tony Chang  <tony@chromium.org>
2043
2044         [NRWT] speculative fix for multiple subprocess test on cygwin
2045         https://bugs.webkit.org/show_bug.cgi?id=72518
2046
2047         Reviewed by Adam Barth.
2048
2049         * Scripts/webkitpy/layout_tests/port/chromium.py: cygwin probably
2050             tries to emulate posix fd behavior.
2051         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2052
2053 2011-11-16  Dirk Pranke  <dpranke@chromium.org>
2054
2055         Revert r100509; change was committed twice (first time as r100491).
2056         Unreviewed, build fix.
2057
2058         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2059
2060 2011-11-16  Dirk Pranke  <dpranke@chromium.org>
2061
2062         Run a dummy test in the gpu configurations if there's nothing to do.
2063         https://bugs.webkit.org/show_bug.cgi?id=72498
2064
2065         Reviewed by James Robinson.
2066
2067         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2068
2069 2011-11-16  Ryosuke Niwa  <rniwa@webkit.org>
2070
2071         contributors_by_fuzzy_match is super slow
2072         https://bugs.webkit.org/show_bug.cgi?id=72540
2073
2074         Reviewed by Eric Seidel.
2075
2076         Make contributors_by_name do case insensitive search using a dictionary.
2077
2078         Also call contributors_by_name first in contributors_by_fuzzy_match now that it's fast
2079         because that's the most common case.
2080
2081         * Scripts/webkitpy/common/config/committers.py:
2082         * Scripts/webkitpy/common/config/committers_unittest.py:
2083
2084 2011-11-16  David Levin  <levin@chromium.org>
2085
2086         check-webkit-style should recognize functions even if they have OVERRIDE after them.
2087         https://bugs.webkit.org/show_bug.cgi?id=72515
2088
2089         Reviewed by Adam Barth.
2090
2091         * Scripts/webkitpy/style/checkers/cpp.py:
2092         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
2093
2094 2011-11-16  Tony Chang  <tony@chromium.org>
2095
2096         [NRWT] refactor drivers to require pixel_tests param
2097         https://bugs.webkit.org/show_bug.cgi?id=72517
2098
2099         Reviewed by Ojan Vafai.
2100
2101         No tests, just a refactoring.
2102
2103         * Scripts/webkitpy/layout_tests/port/chromium.py:
2104         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2105         * Scripts/webkitpy/layout_tests/port/webkit.py:
2106         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
2107
2108 2011-11-16  Dirk Pranke  <dpranke@chromium.org>
2109
2110         Run a dummy test in the gpu configurations if there's nothing to do.
2111         https://bugs.webkit.org/show_bug.cgi?id=72498
2112
2113         Reviewed by James Robinson.
2114
2115         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2116
2117 2011-11-16  Chris Fleizach  <cfleizach@apple.com>
2118
2119         WebKitTestRunner needs to support accessibility-related DRT APIs
2120         https://bugs.webkit.org/show_bug.cgi?id=42131
2121
2122         Reviewed by Beth Dakin.
2123
2124         Port the AX code in DRT to WKTestRunner by using IDL.
2125
2126         * DumpRenderTree/AccessibilityUIElement.cpp:
2127         (horizontalScrollbarCallback):
2128         (verticalScrollbarCallback):
2129         (AccessibilityUIElement::horizontalScrollbar):
2130         (AccessibilityUIElement::verticalScrollbar):
2131         (AccessibilityUIElement::getJSClass):
2132         * DumpRenderTree/AccessibilityUIElement.h:
2133         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
2134         (AccessibilityUIElement::horizontalScrollbar):
2135         (AccessibilityUIElement::verticalScrollbar):
2136         * WebKitTestRunner/DerivedSources.make:
2137         * WebKitTestRunner/GNUmakefile.am:
2138         * WebKitTestRunner/InjectedBundle/AccessibilityController.cpp: Added.
2139         (WTR::AccessibilityController::create):
2140         (WTR::AccessibilityController::AccessibilityController):
2141         (WTR::AccessibilityController::~AccessibilityController):
2142         (WTR::AccessibilityController::makeWindowObject):
2143         (WTR::AccessibilityController::wrapperClass):
2144         (WTR::AccessibilityController::rootElement):
2145         (WTR::AccessibilityController::focusedElement):
2146         (WTR::AccessibilityController::elementAtPoint):
2147         * WebKitTestRunner/InjectedBundle/AccessibilityController.h: Added.
2148         (WTR::AccessibilityController::logFocusEvents):
2149         (WTR::AccessibilityController::logValueChangeEvents):
2150         (WTR::AccessibilityController::logScrollingStartEvents):
2151         (WTR::AccessibilityController::logAccessibilityEvents):
2152         (WTR::AccessibilityController::addNotificationListener):
2153         (WTR::AccessibilityController::notificationReceived):
2154         (WTR::AccessibilityController::resetToConsistentState):
2155         * WebKitTestRunner/InjectedBundle/AccessibilityTextMarker.cpp: Added.
2156         (WTR::AccessibilityTextMarker::create):
2157         (WTR::AccessibilityTextMarker::AccessibilityTextMarker):
2158         (WTR::AccessibilityTextMarker::~AccessibilityTextMarker):
2159         (WTR::AccessibilityTextMarker::platformTextMarker):
2160         (WTR::AccessibilityTextMarker::wrapperClass):
2161         * WebKitTestRunner/InjectedBundle/AccessibilityTextMarker.h: Added.
2162         (WTR::AccessibilityTextMarker::isEqual):
2163         * WebKitTestRunner/InjectedBundle/AccessibilityTextMarkerRange.cpp: Added.
2164         (WTR::AccessibilityTextMarkerRange::create):
2165         (WTR::AccessibilityTextMarkerRange::AccessibilityTextMarkerRange):
2166         (WTR::AccessibilityTextMarkerRange::~AccessibilityTextMarkerRange):
2167         (WTR::AccessibilityTextMarkerRange::platformTextMarkerRange):
2168         (WTR::AccessibilityTextMarkerRange::wrapperClass):
2169         * WebKitTestRunner/InjectedBundle/AccessibilityTextMarkerRange.h: Added.
2170         (WTR::AccessibilityTextMarkerRange::isEqual):
2171         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp: Added.
2172         (WTR::AccessibilityUIElement::create):
2173         (WTR::AccessibilityUIElement::wrapperClass):
2174         (WTR::AccessibilityUIElement::isValid):
2175         (WTR::AccessibilityUIElement::AccessibilityUIElement):
2176         (WTR::AccessibilityUIElement::~AccessibilityUIElement):
2177         (WTR::AccessibilityUIElement::isEqual):
2178         (WTR::AccessibilityUIElement::elementAtPoint):
2179         (WTR::AccessibilityUIElement::childAtIndex):
2180         (WTR::AccessibilityUIElement::indexOfChild):
2181         (WTR::AccessibilityUIElement::childrenCount):
2182         (WTR::AccessibilityUIElement::titleUIElement):
2183         (WTR::AccessibilityUIElement::parentElement):
2184         (WTR::AccessibilityUIElement::takeFocus):
2185         (WTR::AccessibilityUIElement::takeSelection):
2186         (WTR::AccessibilityUIElement::addSelection):
2187         (WTR::AccessibilityUIElement::removeSelection):
2188         (WTR::AccessibilityUIElement::allAttributes):
2189         (WTR::AccessibilityUIElement::attributesOfLinkedUIElements):
2190         (WTR::AccessibilityUIElement::linkedUIElementAtIndex):
2191         (WTR::AccessibilityUIElement::attributesOfDocumentLinks):
2192         (WTR::AccessibilityUIElement::attributesOfChildren):
2193         (WTR::AccessibilityUIElement::parameterizedAttributeNames):
2194         (WTR::AccessibilityUIElement::increment):
2195         (WTR::AccessibilityUIElement::decrement):
2196         (WTR::AccessibilityUIElement::showMenu):
2197         (WTR::AccessibilityUIElement::press):
2198         (WTR::AccessibilityUIElement::stringAttributeValue):
2199         (WTR::AccessibilityUIElement::uiElementAttributeValue):
2200         (WTR::AccessibilityUIElement::numberAttributeValue):
2201         (WTR::AccessibilityUIElement::boolAttributeValue):
2202         (WTR::AccessibilityUIElement::isAttributeSupported):
2203         (WTR::AccessibilityUIElement::isAttributeSettable):
2204         (WTR::AccessibilityUIElement::isActionSupported):
2205         (WTR::AccessibilityUIElement::role):
2206         (WTR::AccessibilityUIElement::subrole):
2207         (WTR::AccessibilityUIElement::roleDescription):
2208         (WTR::AccessibilityUIElement::title):
2209         (WTR::AccessibilityUIElement::description):
2210         (WTR::AccessibilityUIElement::language):
2211         (WTR::AccessibilityUIElement::stringValue):
2212         (WTR::AccessibilityUIElement::accessibilityValue):
2213         (WTR::AccessibilityUIElement::helpText):
2214         (WTR::AccessibilityUIElement::orientation):
2215         (WTR::AccessibilityUIElement::x):
2216         (WTR::AccessibilityUIElement::y):
2217         (WTR::AccessibilityUIElement::width):
2218         (WTR::AccessibilityUIElement::height):
2219         (WTR::AccessibilityUIElement::intValue):
2220         (WTR::AccessibilityUIElement::minValue):
2221         (WTR::AccessibilityUIElement::maxValue):
2222         (WTR::AccessibilityUIElement::valueDescription):
2223         (WTR::AccessibilityUIElement::insertionPointLineNumber):
2224         (WTR::AccessibilityUIElement::selectedTextRange):
2225         (WTR::AccessibilityUIElement::isEnabled):
2226         (WTR::AccessibilityUIElement::isRequired):
2227         (WTR::AccessibilityUIElement::isFocused):
2228         (WTR::AccessibilityUIElement::isFocusable):
2229         (WTR::AccessibilityUIElement::isSelected):
2230         (WTR::AccessibilityUIElement::isSelectable):
2231         (WTR::AccessibilityUIElement::isMultiSelectable):
2232         (WTR::AccessibilityUIElement::setSelectedChild):
2233         (WTR::AccessibilityUIElement::selectedChildrenCount):
2234         (WTR::AccessibilityUIElement::selectedChildAtIndex):
2235         (WTR::AccessibilityUIElement::isExpanded):
2236         (WTR::AccessibilityUIElement::isChecked):
2237         (WTR::AccessibilityUIElement::isVisible):
2238         (WTR::AccessibilityUIElement::isOffScreen):
2239         (WTR::AccessibilityUIElement::isCollapsed):
2240         (WTR::AccessibilityUIElement::isIgnored):
2241         (WTR::AccessibilityUIElement::hasPopup):
2242         (WTR::AccessibilityUIElement::hierarchicalLevel):
2243         (WTR::AccessibilityUIElement::clickPointX):
2244         (WTR::AccessibilityUIElement::clickPointY):
2245         (WTR::AccessibilityUIElement::documentEncoding):
2246         (WTR::AccessibilityUIElement::documentURI):
2247         (WTR::AccessibilityUIElement::url):
2248         (WTR::AccessibilityUIElement::speak):
2249         (WTR::AccessibilityUIElement::attributesOfColumnHeaders):
2250         (WTR::AccessibilityUIElement::attributesOfRowHeaders):
2251         (WTR::AccessibilityUIElement::attributesOfColumns):
2252         (WTR::AccessibilityUIElement::attributesOfRows):
2253         (WTR::AccessibilityUIElement::attributesOfVisibleCells):
2254         (WTR::AccessibilityUIElement::attributesOfHeader):
2255         (WTR::AccessibilityUIElement::indexInTable):
2256         (WTR::AccessibilityUIElement::rowIndexRange):
2257         (WTR::AccessibilityUIElement::columnIndexRange):
2258         (WTR::AccessibilityUIElement::rowCount):
2259         (WTR::AccessibilityUIElement::columnCount):
2260         (WTR::AccessibilityUIElement::selectedRowAtIndex):
2261         (WTR::AccessibilityUIElement::disclosedByRow):
2262         (WTR::AccessibilityUIElement::disclosedRowAtIndex):
2263         (WTR::AccessibilityUIElement::ariaOwnsElementAtIndex):
2264         (WTR::AccessibilityUIElement::ariaFlowToElementAtIndex):
2265         (WTR::AccessibilityUIElement::ariaIsGrabbed):
2266         (WTR::AccessibilityUIElement::ariaDropEffects):
2267         (WTR::AccessibilityUIElement::lineForIndex):
2268         (WTR::AccessibilityUIElement::rangeForLine):
2269         (WTR::AccessibilityUIElement::boundsForRange):
2270         (WTR::AccessibilityUIElement::setSelectedTextRange):
2271         (WTR::AccessibilityUIElement::stringForRange):
2272         (WTR::AccessibilityUIElement::attributedStringForRange):
2273         (WTR::AccessibilityUIElement::attributedStringRangeIsMisspelled):
2274         (WTR::AccessibilityUIElement::uiElementForSearchPredicate):
2275         (WTR::AccessibilityUIElement::cellForColumnAndRow):
2276         (WTR::AccessibilityUIElement::horizontalScrollbar):
2277         (WTR::AccessibilityUIElement::verticalScrollbar):
2278         (WTR::AccessibilityUIElement::addNotificationListener):
2279         (WTR::AccessibilityUIElement::removeNotificationListener):
2280         (WTR::AccessibilityUIElement::textMarkerRangeForElement):
2281         (WTR::AccessibilityUIElement::textMarkerRangeLength):
2282         (WTR::AccessibilityUIElement::textMarkerRangeForMarkers):
2283         (WTR::AccessibilityUIElement::startTextMarkerForTextMarkerRange):
2284         (WTR::AccessibilityUIElement::endTextMarkerForTextMarkerRange):
2285         (WTR::AccessibilityUIElement::accessibilityElementForTextMarker):
2286         (WTR::AccessibilityUIElement::textMarkerForPoint):
2287         (WTR::AccessibilityUIElement::previousTextMarker):
2288         (WTR::AccessibilityUIElement::nextTextMarker):
2289         (WTR::AccessibilityUIElement::stringForTextMarkerRange):
2290         * WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h: Added.
2291         (WTR::AccessibilityUIElement::platformUIElement):
2292         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityController.idl: Added.
2293         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityTextMarker.idl: Added.
2294         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityTextMarkerRange.idl: Added.
2295         * WebKitTestRunner/InjectedBundle/Bindings/AccessibilityUIElement.idl: Added.
2296         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
2297         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
2298         (WTR::InjectedBundle::beginTesting):
2299         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
2300         (WTR::InjectedBundle::accessibilityController):
2301         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
2302         (WTR::InjectedBundlePage::didClearWindowForFrame):
2303         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
2304         * WebKitTestRunner/InjectedBundle/gtk/LayoutTestControllerGtk.cpp:
2305         (WTR::LayoutTestController::platformName):
2306         * WebKitTestRunner/InjectedBundle/mac/AccessibilityTextMarkerMac.mm: Added.
2307         (WTR::AccessibilityTextMarker::isEqual):
2308         * WebKitTestRunner/InjectedBundle/mac/AccessibilityTextMarkerRangeMac.mm: Added.
2309         (WTR::AccessibilityTextMarkerRange::isEqual):
2310         * WebKitTestRunner/InjectedBundle/mac/AccessibilityUIElementMac.mm: Added.
2311         (+[NSString stringWithJSStringRef:]):
2312         (-[NSString createJSStringRef]):
2313         (-[AccessibilityNotificationHandler initWithPlatformElement:]):
2314         (-[AccessibilityNotificationHandler dealloc]):
2315         (-[AccessibilityNotificationHandler _notificationReceived:]):
2316         (-[AccessibilityNotificationHandler setCallback:]):
2317         (WTR::AccessibilityUIElement::AccessibilityUIElement):
2318         (WTR::AccessibilityUIElement::~AccessibilityUIElement):
2319         (WTR::AccessibilityUIElement::isEqual):
2320         (WTR::descriptionOfValue):
2321         (WTR::attributesOfElement):
2322         (WTR::concatenateAttributeAndValue):
2323         (WTR::convertNSArrayToVector):
2324         (WTR::descriptionOfElements):
2325         (WTR::AccessibilityUIElement::getLinkedUIElements):
2326         (WTR::AccessibilityUIElement::getDocumentLinks):
2327         (WTR::AccessibilityUIElement::getChildren):
2328         (WTR::AccessibilityUIElement::getChildrenWithRange):
2329         (WTR::AccessibilityUIElement::childrenCount):
2330         (WTR::AccessibilityUIElement::elementAtPoint):
2331         (WTR::AccessibilityUIElement::indexOfChild):
2332         (WTR::AccessibilityUIElement::childAtIndex):
2333         (WTR::AccessibilityUIElement::linkedUIElementAtIndex):
2334         (WTR::AccessibilityUIElement::ariaOwnsElementAtIndex):
2335         (WTR::AccessibilityUIElement::ariaFlowToElementAtIndex):
2336         (WTR::AccessibilityUIElement::disclosedRowAtIndex):
2337         (WTR::AccessibilityUIElement::selectedChildAtIndex):
2338         (WTR::AccessibilityUIElement::selectedChildrenCount):
2339         (WTR::AccessibilityUIElement::selectedRowAtIndex):
2340         (WTR::AccessibilityUIElement::titleUIElement):
2341         (WTR::AccessibilityUIElement::parentElement):
2342         (WTR::AccessibilityUIElement::disclosedByRow):
2343         (WTR::AccessibilityUIElement::attributesOfLinkedUIElements):
2344         (WTR::AccessibilityUIElement::attributesOfDocumentLinks):
2345         (WTR::AccessibilityUIElement::attributesOfChildren):
2346         (WTR::AccessibilityUIElement::allAttributes):
2347         (WTR::AccessibilityUIElement::stringAttributeValue):
2348         (WTR::AccessibilityUIElement::numberAttributeValue):
2349         (WTR::AccessibilityUIElement::uiElementAttributeValue):
2350         (WTR::AccessibilityUIElement::boolAttributeValue):
2351         (WTR::AccessibilityUIElement::isAttributeSettable):
2352         (WTR::AccessibilityUIElement::isAttributeSupported):
2353         (WTR::AccessibilityUIElement::parameterizedAttributeNames):
2354         (WTR::AccessibilityUIElement::role):
2355         (WTR::AccessibilityUIElement::subrole):
2356         (WTR::AccessibilityUIElement::roleDescription):
2357         (WTR::AccessibilityUIElement::title):
2358         (WTR::AccessibilityUIElement::description):
2359         (WTR::AccessibilityUIElement::orientation):
2360         (WTR::AccessibilityUIElement::stringValue):
2361         (WTR::AccessibilityUIElement::language):
2362         (WTR::AccessibilityUIElement::helpText):
2363         (WTR::AccessibilityUIElement::x):
2364         (WTR::AccessibilityUIElement::y):
2365         (WTR::AccessibilityUIElement::width):
2366         (WTR::AccessibilityUIElement::height):
2367         (WTR::AccessibilityUIElement::clickPointX):
2368         (WTR::AccessibilityUIElement::clickPointY):
2369         (WTR::AccessibilityUIElement::intValue):
2370         (WTR::AccessibilityUIElement::minValue):
2371         (WTR::AccessibilityUIElement::maxValue):
2372         (WTR::AccessibilityUIElement::valueDescription):
2373         (WTR::AccessibilityUIElement::insertionPointLineNumber):
2374         (WTR::AccessibilityUIElement::isActionSupported):
2375         (WTR::AccessibilityUIElement::isEnabled):
2376         (WTR::AccessibilityUIElement::isRequired):
2377         (WTR::AccessibilityUIElement::isFocused):
2378         (WTR::AccessibilityUIElement::isSelected):
2379         (WTR::AccessibilityUIElement::isExpanded):
2380         (WTR::AccessibilityUIElement::isChecked):
2381         (WTR::AccessibilityUIElement::hierarchicalLevel):
2382         (WTR::AccessibilityUIElement::speak):
2383         (WTR::AccessibilityUIElement::ariaIsGrabbed):
2384         (WTR::AccessibilityUIElement::ariaDropEffects):
2385         (WTR::AccessibilityUIElement::lineForIndex):
2386         (WTR::AccessibilityUIElement::rangeForLine):
2387         (WTR::AccessibilityUIElement::boundsForRange):
2388         (WTR::AccessibilityUIElement::stringForRange):
2389         (WTR::AccessibilityUIElement::attributedStringForRange):
2390         (WTR::AccessibilityUIElement::attributedStringRangeIsMisspelled):
2391         (WTR::AccessibilityUIElement::uiElementForSearchPredicate):
2392         (WTR::AccessibilityUIElement::attributesOfColumnHeaders):
2393         (WTR::AccessibilityUIElement::attributesOfRowHeaders):
2394         (WTR::AccessibilityUIElement::attributesOfColumns):
2395         (WTR::AccessibilityUIElement::attributesOfRows):
2396         (WTR::AccessibilityUIElement::attributesOfVisibleCells):
2397         (WTR::AccessibilityUIElement::attributesOfHeader):
2398         (WTR::AccessibilityUIElement::rowCount):
2399         (WTR::AccessibilityUIElement::columnCount):
2400         (WTR::AccessibilityUIElement::indexInTable):
2401         (WTR::AccessibilityUIElement::rowIndexRange):
2402         (WTR::AccessibilityUIElement::columnIndexRange):
2403         (WTR::AccessibilityUIElement::cellForColumnAndRow):
2404         (WTR::AccessibilityUIElement::horizontalScrollbar):
2405         (WTR::AccessibilityUIElement::verticalScrollbar):
2406         (WTR::AccessibilityUIElement::selectedTextRange):
2407         (WTR::AccessibilityUIElement::setSelectedTextRange):
2408         (WTR::AccessibilityUIElement::increment):
2409         (WTR::AccessibilityUIElement::decrement):
2410         (WTR::AccessibilityUIElement::showMenu):
2411         (WTR::AccessibilityUIElement::press):
2412         (WTR::AccessibilityUIElement::setSelectedChild):
2413         (WTR::AccessibilityUIElement::accessibilityValue):
2414         (WTR::AccessibilityUIElement::documentEncoding):
2415         (WTR::AccessibilityUIElement::documentURI):
2416         (WTR::AccessibilityUIElement::url):
2417         (WTR::AccessibilityUIElement::addNotificationListener):
2418         (WTR::AccessibilityUIElement::removeNotificationListener):
2419         (WTR::AccessibilityUIElement::isFocusable):
2420         (WTR::AccessibilityUIElement::isSelectable):
2421         (WTR::AccessibilityUIElement::isMultiSelectable):
2422         (WTR::AccessibilityUIElement::isVisible):
2423         (WTR::AccessibilityUIElement::isOffScreen):
2424         (WTR::AccessibilityUIElement::isCollapsed):
2425         (WTR::AccessibilityUIElement::isIgnored):
2426         (WTR::AccessibilityUIElement::hasPopup):
2427         (WTR::AccessibilityUIElement::takeFocus):
2428         (WTR::AccessibilityUIElement::takeSelection):
2429         (WTR::AccessibilityUIElement::addSelection):
2430         (WTR::AccessibilityUIElement::removeSelection):
2431         (WTR::AccessibilityUIElement::textMarkerRangeForElement):
2432         (WTR::AccessibilityUIElement::textMarkerRangeLength):
2433         (WTR::AccessibilityUIElement::previousTextMarker):
2434         (WTR::AccessibilityUIElement::nextTextMarker):
2435         (WTR::AccessibilityUIElement::stringForTextMarkerRange):
2436         (WTR::AccessibilityUIElement::textMarkerRangeForMarkers):
2437         (WTR::AccessibilityUIElement::startTextMarkerForTextMarkerRange):
2438         (WTR::AccessibilityUIElement::endTextMarkerForTextMarkerRange):
2439         (WTR::AccessibilityUIElement::textMarkerForPoint):
2440         (WTR::AccessibilityUIElement::accessibilityElementForTextMarker):
2441         * WebKitTestRunner/InjectedBundle/mac/LayoutTestControllerMac.mm:
2442         (WTR::LayoutTestController::platformName):
2443         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
2444         (WTR::LayoutTestController::platformName):
2445         * WebKitTestRunner/InjectedBundle/win/LayoutTestControllerWin.cpp:
2446         (WTR::LayoutTestController::platformName):
2447         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
2448         * WebKitTestRunner/win/InjectedBundle.vcproj:
2449
2450 2011-11-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2451
2452         [Qt] Fix use of config.tests
2453
2454         We were checking @buildArgs to determine if the optional argument was
2455         passed or not, but this is not the right way to do that in Perl. The
2456         right way is to not use a prototype for the function, and to pass
2457         the list as a reference (to avoid list expansion). That way we can
2458         check the number of arguments, and choose the right code path based
2459         on that.
2460
2461         We were also missing a default-value of 0 for any feature that's not
2462         set in features.prf.
2463
2464         Reviewed by Kenneth Rohde Christiansen.
2465
2466         * Scripts/build-webkit:
2467         * Scripts/webkitdirs.pm:
2468         (qtFeatureDefaults):
2469         (buildQMakeProject):
2470
2471 2011-11-16  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2472
2473         [Qt] Prospective fix for broken plugin tests
2474
2475         Don't hide symbols for the NPAPI test plugin
2476
2477         Rubber-stamped by Andreas Kling.
2478
2479         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
2480
2481 2011-11-16  Adam Roben  <aroben@apple.com>
2482
2483         Move filenames/line numbers off to the right edge in Leaks Viewer
2484
2485         Fixes <http://webkit.org/b/72510> Callstacks in Leaks Viewer are a little hard to read
2486         because filenames/line numbers are right next to function names
2487
2488         Reviewed by Andreas Kling.
2489
2490         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserImpl.js:
2491         (LeaksParserImpl.prototype._createNode): Separate the filename and line number out from the
2492         function name. Use the filename as the profile node's URL (we don't have enough info to
2493         compute a real URL currently) and the line number as the profile node's line number so the
2494         Inspector code will put them on the right side of the callstack graph.
2495
2496         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/WebInspectorShims.js:
2497         (WebInspector.displayNameForURL): Stubbed this out. For now we don't ever get any URLs
2498         passed here, just filenames.
2499
2500         (WebInspector.linkifyURLAsNode):
2501         (WebInspector.linkifyResourceAsNode):
2502         Copied and slightly modified from inspector.js.
2503
2504 2011-11-16  Adam Roben  <aroben@apple.com>
2505
2506         Make Leaks Viewer load builds from the new Lion Leaks bot
2507
2508         Fixes <http://webkit.org/b/72507> REGRESSION: Leaks Viewer no longer shows recent leaky builds
2509
2510         Reviewed by Andreas Kling.
2511
2512         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
2513         (LeaksViewer._displayURLPrompt): Use the Lion Leaks bot instead of the nonexistent
2514         SnowLeopard one.
2515
2516 2011-11-16  Raphael Kubo da Costa  <kubo@profusion.mobi>
2517
2518         [EFL] Clean up the offline apps-related functions in ewk_settings.
2519         https://bugs.webkit.org/show_bug.cgi?id=72132
2520
2521         Reviewed by Kenneth Rohde Christiansen.
2522
2523         `ewk_settings_cache' and `ewk_view_setting_cache' are too common
2524         prefixes for code that is related only to the HTML5 app cache
2525         features.
2526
2527         Use `ewk_settings_application_cache' and
2528         `ewk_view_setting_application_cache' instead.
2529
2530         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
2531         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
2532
2533 2011-11-15  Simon Hausmann  <simon.hausmann@nokia.com>
2534
2535         [Qt] Centralize hide_symbols and ensure all libs are built with symbol visibility & bsymbolic_functions
2536
2537         Reviewed by Tor Arne Vestbø.
2538
2539         * qmake/mkspecs/features/default_post.prf: Place the common hide_symbols and -Bsymbolic-functions
2540         magic here where we handle all TEMPLATE = lib .pro files (except plugins).
2541         * WebKitTestRunner/InjectedBundle/Target.pri: Mark the injected bundle as plugin, which also
2542         removes the .so.0.0.0 and just produces .so.
2543
2544 2011-11-16  Peter Beverloo  <peter@chromium.org>
2545
2546         Explicitly include unistd.h in TestNetscapePlugIn
2547         https://bugs.webkit.org/show_bug.cgi?id=72385
2548
2549         This is needed for the sleep() call in the indicateTestFailure method,
2550         which seems to be implicitly available for other platforms. This is not
2551         the case on Android.
2552
2553         Reviewed by Steve Block.
2554
2555         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
2556
2557 2011-11-16  Peter Beverloo  <peter@chromium.org>
2558
2559         [Chromium] Update the Android NDK version being pulled in by update-webkit-chromium
2560         https://bugs.webkit.org/show_bug.cgi?id=72384
2561
2562         Update to Android NDK r7, which was released last Friday. This adds
2563         support for RTTI and the -mimplicit-it gcc flag, which are required to
2564         respectively build ICU and Skia.
2565
2566         Reviewed by Steve Block.
2567
2568         * Scripts/update-webkit-chromium:
2569         * Scripts/webkitdirs.pm:
2570         (buildChromiumMakefile):
2571
2572 2011-11-16  Simon Hausmann  <simon.hausmann@nokia.com>
2573
2574         Unreviewed, rolling out r100266.
2575         http://trac.webkit.org/changeset/100266
2576
2577         Broke WTR.
2578
2579         * qmake/mkspecs/features/default_post.prf:
2580
2581 2011-11-16  Eric Seidel  <eric@webkit.org>
2582
2583         check-webkit-style broken by r99773: "Could not determine the port"
2584         https://bugs.webkit.org/show_bug.cgi?id=72275
2585
2586         Reviewed by Adam Barth.
2587
2588         I will continue engaging hacks, until windows improves.
2589
2590         * Scripts/webkitpy/common/host.py:
2591          - Host uses self.executive instead of self._executive like Port does.
2592          - I wanted to test this, but it mutates global state -- thus impossible to unittest at this time.
2593
2594 2011-11-16  Eric Seidel  <eric@webkit.org>
2595
2596         check-webkit-style broken by r99773: "Could not determine the port"
2597         https://bugs.webkit.org/show_bug.cgi?id=72275
2598
2599         Reviewed by Adam Barth.
2600
2601         Engage windows hacks harder.
2602
2603         This is a speculative fix for ChromiumWin.  We're now
2604         creating an SCM object much earlier, so we need to
2605         engage our "awesome" windows hacks earlier.
2606
2607         * Scripts/webkitpy/common/host.py:
2608         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
2609
2610 2011-11-15  Darin Adler  <darin@apple.com>
2611
2612         DRT's uiElementAttributeValueCallback function is leaky
2613         https://bugs.webkit.org/show_bug.cgi?id=72453
2614
2615         Reviewed by Dan Bernstein.
2616
2617         * DumpRenderTree/AccessibilityUIElement.cpp:
2618         (uiElementAttributeValueCallback): Use JSRetainPtr and don't leak.
2619
2620 2011-11-14  Ryosuke Niwa  <rniwa@webkit.org>
2621
2622         Implement edit-distance based reviewer recognition algorithm
2623         https://bugs.webkit.org/show_bug.cgi?id=72351
2624
2625         Reviewed by Eric Seidel.
2626
2627         Implement an algorithm to recognize reviewer's name based on its edit distance (or more precisely
2628         its Levenshtein distance) to each reviewer's full name, first, last and middle names, and IRC nicknames.
2629         Furthermore, we cap the maximum edit distance at len(name) - 1 to avoid matching a bogus string like
2630         "build fix" to a reviewer's name (e.g. with with edit distance 9).
2631
2632         This algorithm is implemented in CommitterList.contributors_by_fuzzy_match. The function to compute
2633         the edit distance is implemented in edit_distance.py.
2634
2635         Also moved _has_valid_reviewer from ValidateReviewer to ChangeLogEntry because we can no longer rely
2636         on the presence of ChangeLogEntry.reviewer() to verify that reviewer string is nicely formatted.
2637
2638         * Scripts/webkitpy/common/checkout/changelog.py:
2639         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
2640         * Scripts/webkitpy/common/config/committers.py:
2641         * Scripts/webkitpy/common/config/committers_unittest.py:
2642         * Scripts/webkitpy/common/editdistance.py: Added.
2643         * Scripts/webkitpy/common/editdistance_unittest.py: Added.
2644         * Scripts/webkitpy/tool/steps/validatereviewer.py:
2645         * Scripts/webkitpy/tool/steps/validatereviewer_unittest.py: Removed.
2646
2647 2011-11-15  Tony Chang  <tony@chromium.org>
2648
2649         Skip a failing webkitpy test on cygwin.
2650
2651         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
2652
2653 2011-11-15  Sam Weinig  <sam@webkit.org>
2654
2655         TestWebKitAPI should not put its resources in the root products directory
2656         https://bugs.webkit.org/show_bug.cgi?id=72446
2657
2658         Reviewed by Anders Carlsson.
2659
2660         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
2661         Change the copy resources phase to place the resources in a TestWebKitAPI.resources
2662         directory.
2663
2664         * TestWebKitAPI/Tests/WebKit2ObjC/WKBrowsingContextLoadDelegateTest.mm:
2665         (TEST_F):
2666         * TestWebKitAPI/Tests/mac/DOMRangeOfString.mm:
2667         (TestWebKitAPI::TEST):
2668         * TestWebKitAPI/Tests/mac/DeviceScaleFactorOnBack.mm:
2669         (TestWebKitAPI::DeviceScaleFactorOnBack::url):
2670         * TestWebKitAPI/Tests/mac/DynamicDeviceScaleFactor.mm:
2671         (TestWebKitAPI::DynamicDeviceScaleFactor::url):
2672         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
2673         (TestWebKitAPI::Util::createURLForResource):
2674         Update calls to retrieve files from the bundle to look in the newly
2675         created subdirectory.
2676
2677 2011-11-15  Tony Chang  <tony@chromium.org>
2678
2679         Fix a webkitpy test caused by renaming the accelerate-video flag in r100355.
2680
2681         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
2682
2683 2011-11-15  Tony Chang  <tony@chromium.org>
2684
2685         Unreviewed, fix test-webkitpy tests.  Move the check for image results
2686         earlier.
2687
2688         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2689         * Scripts/webkitpy/layout_tests/port/webkit.py:
2690
2691 2011-11-15  James Robinson  <jamesr@chromium.org>
2692
2693         [chromium] Fix flag name in chromium DumpRenderTree for accelerated video trigger
2694         https://bugs.webkit.org/show_bug.cgi?id=72444
2695
2696         Reviewed by Adam Barth.
2697
2698         * DumpRenderTree/chromium/DumpRenderTree.cpp:
2699
2700 2011-11-15  Tony Chang  <tony@chromium.org>
2701
2702         Unreviewed, better fix for ref tests failures on SL Intel bot.
2703         Revert the previous change and return None if pixel results from
2704         either the test page or the expected page is missing.
2705
2706         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2707         * Scripts/webkitpy/layout_tests/port/webkit.py:
2708
2709 2011-11-15  James Robinson  <jamesr@chromium.org>
2710
2711         [chromium] Merge chromium-gpu layout test configurations into non-gpu versions
2712         https://bugs.webkit.org/show_bug.cgi?id=72402
2713
2714         Reviewed by Dirk Pranke.
2715
2716         Allows the compositing layout tests to run in the normal layout test run. This sets the default compositing
2717         triggers in DumpRenderTree to more closely match what we use in Chromium. There are two exceptions, canvas 2d
2718         and video (media) are still controlled by an explicit trigger so that they are true when platform=chromium-gpu
2719         and false otherwise, so that we can run the same set of tests in both configurations. This isn't necessary for
2720         the normal compositing triggers since we currently run the compositing tests in only one configuration - with
2721         compositing enabled.
2722
2723         Since the compositing tests are currently marked WONTFIX SKIP on non-GPU platforms in test_expectations.txt this
2724         patch should not impact any tests.
2725
2726         * DumpRenderTree/chromium/DumpRenderTree.cpp:
2727         (main):
2728         * DumpRenderTree/chromium/TestShell.cpp:
2729         (TestShell::TestShell):
2730         (TestShell::resetWebSettings):
2731         * DumpRenderTree/chromium/TestShell.h:
2732         (TestShell::setAcceleratedCompositingForVideoEnabled):
2733         * DumpRenderTree/chromium/WebPreferences.cpp:
2734         (WebPreferences::reset):
2735         (WebPreferences::applyTo):
2736         * DumpRenderTree/chromium/WebPreferences.h:
2737         * Scripts/webkitpy/layout_tests/port/chromium.py:
2738         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2739         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
2740
2741 2011-11-15  Tony Chang  <tony@chromium.org>
2742
2743         Unreviewed, try to fix ref tests failures on SL Intel bot.
2744         It looks like ImageDiff isn't returning an image diff.
2745
2746         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
2747
2748 2011-11-15  Tony Chang  <tony@chromium.org>
2749
2750         Unreviewed fix of ref-tests on GTK+.
2751         If a driver is never started, _xvfb_process is never set.
2752
2753         * Scripts/webkitpy/layout_tests/port/gtk.py:
2754
2755 2011-11-14  Tony Chang  <tony@chromium.org>
2756
2757         [NRWT] Reftests should run even when pixel tests are disabled.
2758         https://bugs.webkit.org/show_bug.cgi?id=60605
2759
2760         Reviewed by Dirk Pranke.
2761
2762         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py: Only skip ref tests if --no-ref-tests is passed.
2763             Also add an assert to make sure we get image hashes back when running ref tests.
2764         * Scripts/webkitpy/layout_tests/controllers/worker.py: Use Driver.has_crashed() instead of poll().
2765         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2766         * Scripts/webkitpy/layout_tests/port/chromium.py: Use DriverProxy.
2767         * Scripts/webkitpy/layout_tests/port/driver.py: Add DriverProxy which does the work
2768             of starting a pixel driver if needed.  It handles the logic of sending the test
2769             to the correct driver.  Also renamed Driver.poll() to Driver.has_crashed().
2770         * Scripts/webkitpy/layout_tests/port/dryrun.py:
2771         * Scripts/webkitpy/layout_tests/port/test.py: Switch to using DriverProxy so we get test coverage.
2772         * Scripts/webkitpy/layout_tests/port/webkit.py:
2773         * Scripts/webkitpy/layout_tests/run_webkit_tests.py: Add --no-ref-tests.
2774         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py: Test --no-ref-tests.
2775
2776 2011-11-15  David Kilzer  <ddkilzer@apple.com>
2777
2778         Don't use File::Slurp for run-leaks unit tests
2779         <http://webkit.org/b/72356>
2780
2781         Reviewed by Daniel Bates.
2782
2783         * Scripts/webkitperl/run-leaks_unittest/RunLeaks.pm: Added.
2784         * Scripts/webkitperl/run-leaks_unittest/run-leaks-report-v1.0.pl:
2785         Extracted common package logic into RunLeaks.pm.  Fixed call to
2786         RunLeaks::parseLeaksOutput().
2787         * Scripts/webkitperl/run-leaks_unittest/run-leaks-report-v2.0-new.pl: Ditto.
2788         * Scripts/webkitperl/run-leaks_unittest/run-leaks-report-v2.0-old.pl: Ditto.
2789
2790 2011-11-15  Peter Kasting  <pkasting@google.com>
2791
2792         Handle svn 1.7 when detecting whether a directory is in an svn checkout.
2793         https://bugs.webkit.org/show_bug.cgi?id=72349
2794
2795         Reviewed by Adam Roben.
2796
2797         * Scripts/VCSUtils.pm:
2798         (isSVNDirectory):
2799
2800 2011-11-15  Simon Hausmann  <simon.hausmann@nokia.com>
2801
2802         [Qt] Centralize hide_symbols and ensure all libs are built with symbol visibility & bsymbolic_functions
2803
2804         Reviewed by Tor Arne Vestbø.
2805
2806         * qmake/mkspecs/features/default_post.prf: Place the common hide_symbols and -Bsymbolic-functions
2807         magic here where we handle all TEMPLATE = lib .pro files.
2808
2809 2011-11-15  Simon Hausmann  <simon.hausmann@nokia.com>
2810
2811         [Qt] Unreviewed prospective --minimal build fix.
2812
2813         * DumpRenderTree/qt/DumpRenderTree.pro: Avoid USE_SYSTEM_MALLOC=0
2814         and USE_SYSTEM_MALLOC=1 appearing in DEFINES at the same time.
2815
2816 2011-11-15  Simon Hausmann  <simon.hausmann@nokia.com>
2817
2818         [Qt] Unreviewed --minimal build fix.
2819
2820         * Scripts/build-webkit: Add to DEFINES+= from @options only if there is
2821         a define. The "coverage" option doesn't have one, causing "DEFINES+= =0"
2822         to make moc's preprocessor abort.
2823
2824 2011-11-14  Raphael Kubo da Costa  <kubo@profusion.mobi>
2825
2826         build-webkit: Accept --cmakearg to pass additional arguments to CMake.
2827         https://bugs.webkit.org/show_bug.cgi?id=72156
2828
2829         Reviewed by Daniel Bates.
2830
2831         Similar to --makearg and --qmakearg, --cmakearg lets one pass
2832         additional arguments to CMake (eg. --cmakearg="-DFOO=bar").
2833
2834         * Scripts/build-webkit:
2835         * Scripts/webkitdirs.pm:
2836         (generateBuildSystemFromCMakeProject):
2837
2838 2011-11-14  Ryosuke Niwa  <rniwa@webkit.org>
2839
2840         Improve ChangeLogEntry's reviewer parsing algorithm part 2
2841         https://bugs.webkit.org/show_bug.cgi?id=72340
2842
2843         Reviewed by Eric Seidel.
2844
2845         This patch improves the recognition of NOBODY, wrestler names, and parenthesized clauses,
2846         and prepares ChangeLogEntry to support edit-distance-based reviewer-name recognition.
2847
2848         * Scripts/webkitpy/common/checkout/changelog.py:
2849         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
2850
2851 2011-11-14  Eric Seidel  <eric@webkit.org>
2852
2853         check-webkit-style broken by r99773: "Could not determine the port"
2854         https://bugs.webkit.org/show_bug.cgi?id=72275
2855
2856         Reviewed by Adam Barth.
2857
2858         The TestExpectationsChecker was using a generic try/except block
2859         which caught all exceptions, so we didn't notice that failing
2860         to pass a Host to PortFactory was causing an exception in port instantiation.
2861         I've factored out the "lookup the port" logic into a separate function
2862         which I've now unittested.  This should fix the bug and prevent
2863         others like it from occuring the the future.
2864
2865         * Scripts/webkitpy/style/checkers/test_expectations.py:
2866         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
2867
2868 2011-11-14  Julien Chaffraix  <jchaffraix@webkit.org>
2869
2870         Add --css-grid-layout to build-webkit and the build systems
2871         https://bugs.webkit.org/show_bug.cgi?id=72320
2872
2873         Reviewed by Ojan Vafai.
2874
2875         * Scripts/build-webkit:
2876         Added the option.
2877
2878 2011-11-14  Sheriff Bot  <webkit.review.bot@gmail.com>
2879
2880         Unreviewed, rolling out r100192.
2881         http://trac.webkit.org/changeset/100192
2882         https://bugs.webkit.org/show_bug.cgi?id=72328
2883
2884         Caused infinite tests to fail (Requested by abarth on
2885         #webkit).
2886
2887         * Scripts/webkitpy/layout_tests/port/base.py:
2888         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
2889
2890 2011-11-14  Tony Chang  <tony@chromium.org>
2891
2892         Remove the CSS3_FLEXBOX compile time flag and enable on all ports
2893         https://bugs.webkit.org/show_bug.cgi?id=72196
2894
2895         Reviewed by Ojan Vafai.
2896
2897         * Scripts/build-webkit:
2898
2899 2011-11-14  John Yani  <vanuan@gmail.com>
2900
2901         new-run-webkit-tests should not be locale dependent https://bugs.webkit.org/show_bug.cgi?id=68691
2902         Override LOCALE to en_US.
2903
2904         Reviewed by Eric Seidel.
2905
2906         * Scripts/webkitpy/layout_tests/port/base.py:
2907
2908 2011-11-14  Julien Chaffraix  <jchaffraix@webkit.org>
2909
2910         [Chromium] build-webkit silently drops the features flags
2911         https://bugs.webkit.org/show_bug.cgi?id=72293
2912
2913         Reviewed by Tony Chang.
2914
2915         Chromium still ignores @features but now we give a warning!
2916
2917         * Scripts/build-webkit: Check if one of the features is different
2918         from the default and print a warning in this case.
2919
2920 2011-11-14  David Kilzer  <ddkilzer@apple.com>
2921
2922         run-leaks does not work on Lion?
2923         <http://webkit.org/b/71059>
2924         <rdar://problem/10428527>
2925
2926         Reviewed by Adam Roben.
2927
2928         The output of leaks(1) changed again in Lion to move the
2929         "leaks Report Version: 2.0" line from the first line of the
2930         output to just above the "Process " lines that run-leaks is
2931         interested in parsing.  This required using a more generic
2932         algorithm to find the start of the "Process " lines.
2933
2934         * Scripts/run-leaks:
2935         (parseLeaksOutput): Make the code to skip headers more generic.
2936         * Scripts/webkitperl/run-leaks_unittest/run-leaks-report-v1.0.pl: Added.
2937         * Scripts/webkitperl/run-leaks_unittest/run-leaks-report-v2.0-new.pl: Added.
2938         * Scripts/webkitperl/run-leaks_unittest/run-leaks-report-v2.0-old.pl: Added.
2939
2940 2011-11-14  Lei Zhang  <thestig@chromium.org>
2941
2942         Missing include in Tools/DumpRenderTree/chromium/AccessibilityUIElement.h:
2943         https://bugs.webkit.org/show_bug.cgi?id=72088
2944
2945         Reviewed by Tony Chang.
2946
2947         * DumpRenderTree/chromium/AccessibilityUIElement.h:
2948
2949 2011-11-14  Andrew Scherkus  <scherkus@chromium.org>
2950
2951         [Chromium] Skip media layout tests on chromium-gpu-cg-mac-leopard platform
2952         https://bugs.webkit.org/show_bug.cgi?id=72147
2953
2954         Reviewed by Adam Barth.
2955
2956         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
2957
2958 2011-11-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2959
2960         [Qt] Move the QtWebKit module file to match the layout of Qt's mkspecs
2961
2962         Reviewed by Simon Hausmann.
2963
2964         * qmake/mkspecs/modules/qt_webkit.pri: Renamed from Tools/qmake/qt_webkit.pri.
2965
2966 2011-11-14  Simon Hausmann  <simon.hausmann@nokia.com>
2967
2968         [Qt] Disable plugins on X11 for WK2
2969
2970         Reviewed by Kenneth Rohde Christiansen.
2971
2972         Currently plugins are in-process and they crash in the likely case of
2973         a plugin installation that uses Qt 4, which beautifully clashes with the
2974         Qt 5 based QtWebProcess.
2975
2976         * qmake/mkspecs/features/features.prf:
2977
2978 2011-11-14  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
2979
2980         [Qt] Add support for config tests for Qt 5
2981
2982         This allows us to build and run configure-tests at build
2983         time to determine if a given platform feature is available.
2984
2985         Reviewed by Simon Hausmann.
2986
2987         * Scripts/build-webkit:
2988         * Scripts/webkitdirs.pm:
2989         (qtFeatureDefaults):
2990         * qmake/configure.pro: Added.
2991         * qmake/mkspecs/features/features.prf:
2992         * qmake/sync.profile: Added.
2993
2994 2011-11-11  Pierre Rossi  <pierre.rossi@gmail.com>
2995
2996         [Qt] Remove the QStyle dependency in Qt's mobile theme
2997         https://bugs.webkit.org/show_bug.cgi?id=67773
2998
2999         This refactoring splits up RenderThemeQt with two
3000         subclasses, a QStyle-backed one, and a Mobile version
3001         that uses the old mobile theme for now.
3002
3003         QStyle availability is detected at compile time, and
3004         its use is determined by the QT_WEBKIT_USE_MOBILE_THEME
3005         environment variable.
3006
3007         Reviewed by Simon Hausmann.
3008
3009         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
3010         (WTR::activateFonts):
3011         * qmake/mkspecs/features/features.prf: HAVE_QSTYLE detection,
3012         based on availability of QtWidgets.
3013         * qmake/mkspecs/features/webcore.prf: cleanup.
3014
3015 2011-11-14  Tony Gentilcore  <tonyg@chromium.org>
3016
3017         Unreviewed, rolling out r100116.
3018         http://trac.webkit.org/changeset/100116
3019
3020         r100104 was rolled out, so no longer needed
3021
3022         * Scripts/run-bindings-tests:
3023
3024 2011-11-14  Tony Gentilcore  <tonyg@chromium.org>
3025
3026         Unreviewed, rolling out r100104.
3027         http://trac.webkit.org/changeset/100104
3028         https://bugs.webkit.org/show_bug.cgi?id=72247
3029
3030         broke windows builds
3031
3032         * Scripts/check-webkit-style:
3033         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
3034         * Scripts/webkitpy/common/checkout/deps.py:
3035         * Scripts/webkitpy/common/checkout/scm/__init__.py:
3036         * Scripts/webkitpy/common/checkout/scm/detection.py:
3037         * Scripts/webkitpy/common/checkout/scm/git.py:
3038         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
3039         * Scripts/webkitpy/common/checkout/scm/svn.py:
3040         * Scripts/webkitpy/common/host.py:
3041         * Scripts/webkitpy/common/host_mock.py:
3042         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3043         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
3044         * Scripts/webkitpy/layout_tests/port/base.py:
3045         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
3046         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3047         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
3048         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
3049
3050 2011-11-14  Philippe Normand  <pnormand@igalia.com>
3051
3052         Unreviewed, build fix after r100104. run-binding-tests broke.
3053
3054         * Scripts/run-bindings-tests: Explicitely use scm.detection.
3055
3056 2011-11-13  Eric Seidel  <eric@webkit.org>
3057
3058         Add SCMDetector object to make scm detection mockable
3059         https://bugs.webkit.org/show_bug.cgi?id=72247
3060
3061         Reviewed by Adam Barth.
3062
3063         I think SCM detection may eventually be rolled into Checkout,
3064         but this patch at least makes it possible to mock code-paths
3065         which rely on scm detection.
3066         In the process of replacing callers of these free-functions
3067         I found that one of the functions was no longer used,
3068         and that one of the callers could instead just use the SCM
3069         object it already had access to through port.host.scm().
3070         I also discovered that I was not calling Host._initialize_scm()
3071         and thus host.scm() was always returning None!
3072
3073         * Scripts/check-webkit-style:
3074         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
3075         * Scripts/webkitpy/common/checkout/deps.py:
3076         * Scripts/webkitpy/common/checkout/scm/__init__.py:
3077         * Scripts/webkitpy/common/checkout/scm/detection.py:
3078         * Scripts/webkitpy/common/checkout/scm/git.py:
3079         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
3080         * Scripts/webkitpy/common/checkout/scm/svn.py:
3081         * Scripts/webkitpy/common/host.py:
3082         * Scripts/webkitpy/common/host_mock.py:
3083         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3084         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
3085         * Scripts/webkitpy/layout_tests/port/base.py:
3086         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
3087         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
3088         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
3089         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
3090
3091 2011-11-13  Gavin Peters  <gavinp@chromium.org>
3092
3093         command line control of webcore log channels in chromium DumpRenderTree
3094         https://bugs.webkit.org/show_bug.cgi?id=72126
3095
3096         The option --webcore-log-channels=<foo>,<bar>,<quux> is the same as
3097         what chromium takes.
3098
3099         Reviewed by Kent Tamura.
3100
3101         * DumpRenderTree/chromium/DumpRenderTree.cpp:
3102         (main):
3103
3104 2011-11-12  Ariya Hidayat  <ariya@sencha.com>
3105
3106         [Qt] Fix minor config typo in the build tool documentation
3107         https://bugs.webkit.org/show_bug.cgi?id=72194
3108
3109         Reviewed by Andreas Kling.
3110
3111         * qmake/README:
3112
3113 2011-11-11  Stephen Chenney  <schenney@chromium.org>
3114
3115         [Chromium] DRT does not have LayoutTestController.pageSizeAndMarginsInPixels
3116         https://bugs.webkit.org/show_bug.cgi?id=49257
3117
3118         Adds the remaining missing layout test functionality for printing.
3119
3120         Reviewed by Darin Fisher.
3121
3122         * DumpRenderTree/chromium/LayoutTestController.cpp:
3123         (LayoutTestController::LayoutTestController): Add bindings for new JS methods
3124         (parsePageNumber): Method to parse JS arguments for a page number
3125         (parsePageNumberSizeMargins): Method to parse JS arguments for page size and margins
3126         (LayoutTestController::pageSizeAndMarginsInPixels): Get the page size and margins for a given page.
3127         (LayoutTestController::isPageBoxVisible): get the visibility status for printing
3128         (LayoutTestController::pageProperty): get a specific CSS property when in printing mode
3129         * DumpRenderTree/chromium/LayoutTestController.h: Add declarations for new methods
3130
3131 2011-11-11  Darin Adler  <darin@apple.com>
3132
3133         Remove all releaseRef implementations except for RetainPtr
3134         https://bugs.webkit.org/show_bug.cgi?id=71423
3135
3136         Reviewed by Julien Chaffraix.
3137
3138         * DumpRenderTree/cf/WebArchiveDumpSupport.cpp:
3139         (createXMLStringFromWebArchiveData): Use leakRef instead of releaseRef.
3140
3141 2011-11-11  Ryosuke Niwa  <rniwa@webkit.org>
3142
3143         Make ChangeLogEntry's reviewer parsing algorithm support last 4 WebCore change logs
3144         https://bugs.webkit.org/show_bug.cgi?id=72090
3145
3146         Reviewed by Eric Seidel.
3147
3148         Significantly improve ChangeLogEntry's reviewer parsing algorithm. This version can successfully parse
3149         ChangeLog, ChangeLog-2011-10-19, ChangeLog-2011-06-04, ChangeLog-2011-02-16, and ChangeLog-2010-12-06.
3150
3151         yay! r100002.
3152
3153         * Scripts/webkitpy/common/checkout/changelog.py:
3154         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
3155
3156 2011-11-11  Tony Chang  <tony@chromium.org>
3157
3158         Remove --force from gclient sync.  The bots seem to have made it past
3159         the update step.
3160
3161         * Scripts/update-webkit:
3162
3163 2011-11-11  Tony Chang  <tony@chromium.org>
3164
3165         Add --force to gclient sync. Once the cq bots cycle, we can remove
3166         this since it slows down the normal sync.
3167
3168         * Scripts/update-webkit:
3169
3170 2011-11-11  Marc-Antoine Ruel  <maruel@chromium.org>
3171
3172         When --force is used, also use --reset
3173         https://bugs.webkit.org/show_bug.cgi?id=72129
3174
3175         Reviewed by Dimitri Glazkov.
3176
3177         * Scripts/update-webkit-chromium:
3178
3179 2011-11-11  Tony Chang  <tony@chromium.org>
3180
3181         allow two chromium drivers to run at the same time in NRWT
3182         https://bugs.webkit.org/show_bug.cgi?id=72067
3183
3184         Reviewed by Dirk Pranke.
3185
3186         * Scripts/webkitpy/layout_tests/port/chromium.py: close_fds=True on
3187             posix prevents the subprocesses from sharing the same fd for stdin.
3188             Sharing the same fd prevents us from closing the files unless all references
3189             to the fd are closed.
3190         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py: In addition
3191             to adding a test for this, speed up a test (from 3sec to 0) by fixing
3192             the timeout override. It looks like it regressed in r95875.
3193
3194 2011-11-11  Sheriff Bot  <webkit.review.bot@gmail.com>
3195
3196         Unreviewed, rolling out r99964.
3197         http://trac.webkit.org/changeset/99964
3198         https://bugs.webkit.org/show_bug.cgi?id=72124
3199
3200         Broke too many Qt tests (Requested by tronical_ on #webkit).
3201
3202         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
3203         (WTR::activateFonts):
3204         * qmake/mkspecs/features/features.prf:
3205         * qmake/mkspecs/features/webcore.prf:
3206
3207 2011-11-11  Pierre Rossi  <pierre.rossi@gmail.com>
3208
3209         [Qt] Remove the QStyle dependency in Qt's mobile theme
3210         https://bugs.webkit.org/show_bug.cgi?id=67773
3211
3212         This refactoring splits up RenderThemeQt with two
3213         subclasses, a QStyle-backed one, and a Mobile version
3214         that uses the old mobile theme for now.
3215
3216         QStyle availability is detected at compile time, and
3217         its use is determined by the QT_WEBKIT_USE_MOBILE_THEME
3218         environment variable.
3219
3220         Reviewed by Simon Hausmann.
3221
3222         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
3223         (WTR::activateFonts):
3224         * qmake/mkspecs/features/features.prf: HAVE_QSTYLE detection,
3225         based on availability of QtWidgets.
3226         * qmake/mkspecs/features/webcore.prf: cleanup.
3227
3228 2011-11-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3229
3230         [Qt] Add missing include to PlatformWebViewQt
3231
3232         Reviewed by Simon Hausmann.
3233
3234         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
3235
3236 2011-11-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3237
3238         [Qt] Fix incremental builds on Windows after buildsystem refactor
3239
3240         The qmake rules to ensure that we run make on the derived sources
3241         before running qmake on the target were inside a unix-scope, which
3242         resulted in flakey builds on Windows.
3243
3244         We now generate a rule on all platforms that ensures that "make qmake"
3245         will always run make on the derived sources. In addition, on platforms
3246         that have GNU make (where "make incremental" does not run "make qmake",
3247         but only "make"), we add a similar rule that ensures that any time the
3248         target makefile is regenerated we first run qmake and make on the derived
3249         sources. This rule relies on GNU make's order-only-prerequisites to
3250         not trigger a qmake-run of the target every time we run make.
3251
3252         https://bugs.webkit.org/show_bug.cgi?id=71778
3253
3254         Reviewed by Simon Hausmann.
3255
3256         * qmake/mkspecs/features/functions.prf:
3257
3258 2011-11-11  Simon Hausmann  <simon.hausmann@nokia.com>
3259
3260         [Qt][WK2] Remove QWebNavigationController
3261         https://bugs.webkit.org/show_bug.cgi?id=72113
3262
3263         Reviewed by Andreas Kling.
3264
3265         Adapt to the move of the QWebNavigationController functions/properties to
3266         QQuickWebView.
3267
3268         * MiniBrowser/qt/MiniBrowserApplication.cpp:
3269         * MiniBrowser/qt/qml/BrowserWindow.qml:
3270
3271 2011-11-11  Simon Hausmann  <simon.hausmann@nokia.com>
3272
3273         [Qt] Build system tweak
3274
3275         Reviewed by Tor Arne Vestbø.
3276
3277         Be a good qmake citizen, return true from our linkStaticLibrary
3278         "test" function.
3279
3280         * qmake/mkspecs/features/functions.prf:
3281
3282 2011-11-11  Simon Hausmann  <simon.hausmann@nokia.com>
3283
3284         [Qt] Add support for QT += webkit-private
3285
3286         Reviewed by Tor Arne Vestbø.
3287
3288         * qmake/qt_webkit.pri: Declare private headers.
3289
3290 2011-11-11  Philippe Normand  <pnormand@igalia.com>
3291
3292         Unreviewed, build fix attempt after r99907. Since that revision
3293         GTK layout tests were not running at all...
3294
3295         * Scripts/webkitpy/layout_tests/port/gtk.py:
3296
3297 2011-11-10  Andrew Scherkus  <scherkus@chromium.org>
3298
3299         [Chromium] only run media GPU layout tests on platforms supporting accelerated compositing
3300         https://bugs.webkit.org/show_bug.cgi?id=72001
3301
3302         Reviewed by Dirk Pranke.
3303
3304         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
3305
3306 2011-11-05  Filip Pizlo  <fpizlo@apple.com>
3307
3308         bencher script should support remote benchmarking
3309         https://bugs.webkit.org/show_bug.cgi?id=71618
3310
3311         Rubber-stamped by Geoff Garen.
3312
3313         Added the following features:
3314
3315         - Benchmark report tells you the hostname and machine model (according to
3316           'sysctl hw.model') of the machine you did the run on, and svn revision number
3317           (according to 'svn info') of the VMs you're testing.
3318
3319         - The --remote option. You give this an ssh-style hostname (i.e. user@host is
3320           acceptable, in addition to just host, in which case your username is used)
3321           and it automatically packages the build directories of the VMs you're testing,
3322           sends them to the remote machine along with your copy of the bencher script
3323           and all of the options you passed.  This pretty much seamlessly allows you
3324           to run a set of benchmarks on any machine that is sufficiently compatible with
3325           yours (i.e. similar OS) and that has the appropriate ~/.bencher file, which
3326           must now include a tempPath setting that tells bencher where to put files
3327           when that machine is used as a slave in a remote benchmarking run.
3328
3329         - Added yet another way of specifying which benchmarks to run. This was sort of
3330           necessary for --remote to work as expected. Previously, you either said what
3331           suites to exclude (for example --exclude-sunspider) or that you only wanted
3332           to run one suite to the exclusion of others (--sunspider-only). But it's
3333           often better to just be able to specify the list of suites you want. This is
3334           now supported. You can say, for example, --v8 --sunspider, which means, just
3335           run V8 and SunSpider. This is equivalent to saying --exclude-kraken, which is
3336           still supported.
3337
3338         * Scripts/bencher:
3339
3340 2011-11-10  Tony Chang  <tony@chromium.org>
3341
3342         Lazily start DRT instances in NRWT
3343         https://bugs.webkit.org/show_bug.cgi?id=72038
3344
3345         Reviewed by Eric Seidel.
3346
3347         Just a refactoring, no new tests.
3348
3349         * Scripts/webkitpy/layout_tests/controllers/worker.py:
3350         * Scripts/webkitpy/layout_tests/port/chromium.py:
3351         * Scripts/webkitpy/layout_tests/port/dryrun.py:
3352         * Scripts/webkitpy/layout_tests/port/test.py:
3353         * Scripts/webkitpy/layout_tests/port/webkit.py:
3354
3355 2011-11-10  Ryosuke Niwa  <rniwa@webkit.org>
3356
3357         ChangeLog mistakenly recognizes any string between '*' and ':' as touched files
3358         https://bugs.webkit.org/show_bug.cgi?id=72057
3359
3360         Reviewed by Eric Seidel.
3361
3362         Make the regular expression more explicit.
3363
3364         * Scripts/webkitpy/common/checkout/changelog.py:
3365         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
3366
3367 2011-11-10  Balazs Kelemen  <kbalazs@webkit.org>
3368
3369         [Qt] X11 plugins need to be reworked for Qt5
3370         https://bugs.webkit.org/show_bug.cgi?id=70023
3371
3372         Reviewed by Simon Hausmann.
3373
3374         Rework our basic plugin support in a way that does
3375         not need a bridge between Qt and X. The solution is
3376         based on getting the content drawed by the plugin
3377         from the server as an image and creating a QImage
3378         from it.
3379
3380         * qmake/mkspecs/features/features.prf: Enable X11
3381         plugins if Qt is built with the xcb-xlib backend.
3382
3383 2011-11-10  Simon Hausmann  <simon.hausmann@nokia.com>
3384
3385         [Qt] Fix keyboard related layout tests after API refactoring
3386
3387         Reviewed by Tor Arne Vestbø.
3388
3389         Send key events to the handling page item, not the view.
3390
3391         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
3392         (WTR::PlatformWebView::sendEvent):
3393         (WTR::PlatformWebView::postEvent):
3394
3395 2011-11-10  Alexandre Mazari  <amazari@igalia.com>
3396
3397         [GTK] Make the ENABLE(FEATURE) macro work in DRT
3398         https://bugs.webkit.org/show_bug.cgi?id=71940
3399
3400         Reviewed by Philippe Normand.
3401
3402         * GNUmakefile.am:
3403         Add webcore_cppflags to Programs_DumpRenderTree_CPP so
3404         ENABLE_* variables are part of the environment.
3405
3406 2011-11-10  Simon Hausmann  <simon.hausmann@nokia.com>
3407
3408         [Qt] Fix focus related layout tests after API refactoring
3409
3410         Reviewed by Tor Arne Vestbø.
3411
3412         Make sure the focus is set on the page now, not on the view
3413         (which doesn't actually handle the focus).
3414
3415         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
3416         (WTR::WrapperWindow::WrapperWindow):
3417         (WTR::WrapperWindow::handleStatusChanged):
3418
3419 2011-11-10  Alexis Menard  <alexis.menard@openbossa.org>
3420
3421         [Qt] Merge QTouchWebView and QDesktopWebView into one class
3422         https://bugs.webkit.org/show_bug.cgi?id=71355
3423
3424         Reviewed by Kenneth Rohde Christiansen.
3425
3426         Make MiniBrowser and WebKitTestRunner work again after the
3427         merge.
3428
3429         * MiniBrowser/qt/BrowserWindow.cpp:
3430         (BrowserWindow::BrowserWindow):
3431         (BrowserWindow::webView):
3432         * MiniBrowser/qt/MiniBrowser.pro:
3433         * MiniBrowser/qt/MiniBrowser.qrc:
3434         * MiniBrowser/qt/MiniBrowserApplication.cpp:
3435         * MiniBrowser/qt/main.cpp:
3436         * MiniBrowser/qt/qml/BrowserWindow.qml:
3437         * MiniBrowser/qt/qml/DesktopView.qml: Removed.
3438         * MiniBrowser/qt/qml/TouchView.qml: Removed.
3439         * WebKitTestRunner/PlatformWebView.h:
3440         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
3441         (WTR::PlatformWebView::PlatformWebView):
3442
3443 2011-11-10  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3444
3445         [Qt] Change semantics of the haveQt() function to match API promises
3446
3447         Having Qt 5 does not imply that we have Qt 4, from an API point of
3448         view, as a major version can in theory break BC/SC. Our minimum
3449         version of Qt 4 will always be the latest released version of Qt 4,
3450         so the only place we should use haveQt(4, x) is when checking for
3451         the minimum requirements.
3452
3453         Reviewed by Simon Hausmann.
3454
3455         * qmake/mkspecs/features/default_pre.prf:
3456         * qmake/mkspecs/features/features.prf:
3457         * qmake/mkspecs/features/functions.prf:
3458
3459 2011-11-07  Kaustubh Atrawalkar  <kaustubh@motorola.com> and Sergio Villar Senin  <svillar@igalia.com>
3460
3461         [WK2] [GTK] Implement a MouseDown/MouseUp/MouseMoveTo/MouseScrollBy/LeapForward functions for WebKit2 EventSender
3462         https://bugs.webkit.org/show_bug.cgi?id=69411
3463
3464         Reviewed by Martin Robinson.
3465
3466         * WebKitTestRunner/EventSenderProxy.h:
3467         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
3468         * WebKitTestRunner/TestController.cpp:
3469         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
3470         use the EventSender instead of WebProcess Event Simulation.
3471         * WebKitTestRunner/gtk/EventSenderProxyGtk.cpp:
3472         (WTR::WTREventQueueItem::WTREventQueueItem):
3473         (WTR::EventSenderProxy::EventSenderProxy):
3474         (WTR::getMouseButtonModifiers):
3475         (WTR::eventSenderButtonToGDKButton):
3476         (WTR::EventSenderProxy::createMouseButtonEvent):
3477         (WTR::EventSenderProxy::updateClickCountForButton):
3478         (WTR::EventSenderProxy::replaySavedEvents):
3479         (WTR::EventSenderProxy::sendOrQueueEvent):
3480         (WTR::webkitModifiersToGDKModifiers):
3481         (WTR::getGDKKeySymForKeyRef):
3482         (WTR::EventSenderProxy::keyDown):
3483         (WTR::EventSenderProxy::mouseDown):
3484         (WTR::EventSenderProxy::mouseUp):
3485         (WTR::EventSenderProxy::mouseMoveTo):
3486         (WTR::EventSenderProxy::mouseScrollBy):
3487         (WTR::EventSenderProxy::leapForward):
3488
3489 2011-11-09  Simon Hausmann  <simon.hausmann@nokia.com>
3490
3491         [Qt] Enable exports and constructor functions in static libs
3492
3493         Reviewed by Tor Arne Vestbø.
3494
3495         * qmake/mkspecs/features/functions.prf: Use --whole-archive and their
3496         mac/win32-msvc equivalents when linking static libraries into a shared
3497         library or executable.
3498
3499 2011-11-09  Hayato Ito  <hayato@chromium.org>
3500
3501         [NRWT] Make results.html aware of reference filename in reftests.
3502         https://bugs.webkit.org/show_bug.cgi?id=71574
3503
3504         Reviewed by Ryosuke Niwa.
3505
3506         If a reference filename is different from the default one, include that in result summary
3507         so results.html can link to the correct reference file.
3508
3509         * Scripts/webkitpy/layout_tests/controllers/manager.py:
3510         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3511
3512 2011-11-09  Kevin Ollivier  <kevino@theolliviers.com>
3513
3514         [wx] Unreviewed build fix. Update project files and
3515         add LTC stubs for new methods.
3516
3517         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
3518         (LayoutTestController::addChromeInputField):
3519         (LayoutTestController::removeChromeInputField):
3520         (LayoutTestController::focusWebView):
3521         (LayoutTestController::setBackingScaleFactor):
3522         * waf/build/settings.py:
3523
3524 2011-11-09  Kevin Ollivier  <kevino@theolliviers.com>
3525
3526         [wx] Unreviewed build fix. Support XCode 4 when building deps.
3527
3528         * wx/install-unix-extras:
3529
3530 2011-11-09  Eric Seidel  <eric@webkit.org>
3531
3532         Remove the concept of platform-dependent unittests
3533         https://bugs.webkit.org/show_bug.cgi?id=71963
3534
3535         Unreviewed.  Fix test-webkitpy on cygwin.
3536
3537         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
3538
3539 2011-11-09  Dana Jansens  <danakj@chromium.org>
3540
3541         Add cc-bugs group to watch changes in chromium graphics
3542         https://bugs.webkit.org/show_bug.cgi?id=71690
3543
3544         Reviewed by David Levin.
3545
3546         * Scripts/webkitpy/common/config/committers.py: Make new Account superclass for non-contributor watch accounts
3547         * Scripts/webkitpy/common/config/committers_unittest.py: Test the Account superclass
3548         * Scripts/webkitpy/common/config/watchlist: Add cc-bugs@google.com to watch list
3549         * Scripts/webkitpy/common/watchlist/watchlistparser.py: Check for Accounts also, but require it to be the bugzilla email for any Account/Contributor/etc.
3550
3551 2011-11-09  Eric Seidel  <eric@webkit.org>
3552
3553         Remove more platform-dependent unittests
3554         https://bugs.webkit.org/show_bug.cgi?id=71971
3555
3556         Reviewed by Adam Barth.
3557
3558         Tests which only run on a couple platforms will break.
3559         All tests should run on all platforms where possible.
3560
3561         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
3562         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
3563         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
3564         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
3565
3566 2011-11-09  Ojan Vafai  <ojan@chromium.org>
3567
3568         Merge in results to the test results server even if build numbers conflict
3569         https://bugs.webkit.org/show_bug.cgi?id=71967
3570
3571         Reviewed by Tony Chang.
3572
3573         This was initially meant to be a sanity check that we don't
3574         enter duplicate runs, but it turns out that we can reasonably get
3575         in these situations when a builder is clobbered (e.g. the build
3576         numbers start over again from 0).
3577
3578         * TestResultServer/model/jsonresults.py:
3579         * TestResultServer/model/jsonresults_unittest.py:
3580
3581 2011-11-09  Eric Seidel  <eric@webkit.org>
3582
3583         Remove the concept of platform-dependent unittests
3584         https://bugs.webkit.org/show_bug.cgi?id=71963
3585
3586         Reviewed by Adam Barth.
3587
3588         These have been the source of never-ending sadness.
3589         We'd change behavior and forget to update results in
3590         some unittests because they were only run on certain platforms.
3591         This change removes a large source of these platform-dependent
3592         unittests, which was caused by the port_maker stuff.
3593
3594         It's possible that this change will break test-webkitpy
3595         on some platforms, but that will be a one-time cost.
3596         I will fix the breakage by removing the platform-dependantness
3597         of any such broken tests.
3598
3599         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3600         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
3601         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3602         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3603         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3604         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
3605         * Scripts/webkitpy/layout_tests/port/efl_unittest.py:
3606         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
3607         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3608         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3609         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3610         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3611         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
3612
3613 2011-11-09  Eric Seidel  <eric@webkit.org>
3614
3615         Make the Port object have a Host
3616         https://bugs.webkit.org/show_bug.cgi?id=71961
3617
3618         Reviewed by Adam Barth.
3619
3620         This has been a long time coming.
3621         Now that Port has a Host object, it's very easy
3622         to write unittests which use a properly mocked Port.
3623         Soon we will remove the executive, filesystem and user
3624         members of Port and get everything through the Host.
3625
3626         Note how in many callsites we're actually removing
3627         explicit mocks, since we're now always going to get the
3628         correct Executive/User/FileSystem from the
3629         Host and callsites don't need to be careful to pass each one.
3630
3631         This is prep-work for adding a new Environment member to
3632         Host and needing to be able to access the Environment
3633         member from Port functions w/o needing to pass an Environment
3634         object (or appropriate Mock) every place we create a Port object.
3635         Now that we have a single Host object to reach out through, its
3636         simple to add additional functionality (like Environment) while
3637         keeping our unittests properly mocked.
3638
3639         * Scripts/webkitpy/common/host.py:
3640         * Scripts/webkitpy/common/net/credentials_unittest.py:
3641         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
3642         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
3643         * Scripts/webkitpy/layout_tests/port/apple.py:
3644         * Scripts/webkitpy/layout_tests/port/base.py:
3645         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
3646         * Scripts/webkitpy/layout_tests/port/chromium.py:
3647         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
3648         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
3649         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
3650         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
3651         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
3652         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
3653         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
3654         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
3655         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
3656         * Scripts/webkitpy/layout_tests/port/dryrun.py:
3657         * Scripts/webkitpy/layout_tests/port/efl.py:
3658         * Scripts/webkitpy/layout_tests/port/factory.py:
3659         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
3660         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
3661         * Scripts/webkitpy/layout_tests/port/gtk.py:
3662         * Scripts/webkitpy/layout_tests/port/mac.py:
3663         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
3664         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
3665         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
3666         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
3667         * Scripts/webkitpy/layout_tests/port/qt.py:
3668         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
3669         * Scripts/webkitpy/layout_tests/port/test.py:
3670         * Scripts/webkitpy/layout_tests/port/webkit.py:
3671         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
3672         * Scripts/webkitpy/layout_tests/port/win.py:
3673         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
3674         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
3675         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
3676         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
3677         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
3678
3679 2011-11-09  Dominic Mazzoni  <dmazzoni@google.com>
3680
3681         [Chromium] Implement AccessibilityUIElement::insertionPointLineNumberGetterCallback
3682         https://bugs.webkit.org/show_bug.cgi?id=71561
3683
3684         Reviewed by Dimitri Glazkov.
3685
3686         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
3687         (AccessibilityUIElement::insertionPointLineNumberGetterCallback):
3688
3689 2011-11-09  Benjamin Poulain  <bpoulain@apple.com>
3690
3691         Implement dumpProgressFinishedCallback() for Mac layoutTestController
3692         https://bugs.webkit.org/show_bug.cgi?id=66772
3693
3694         Reviewed by Simon Fraser.
3695
3696         When dumpProgressFinishedCallback is enabled, DumpRenderTree should output
3697         "postProgressFinishedNotification" on FrameLoaderClient::postProgressFinishedNotification().
3698
3699         On Mac, the FrameLoaderClient post a notification. This patch add handling for this notification
3700         in the FrameLoadDelegate and print the output in response to the notification.
3701
3702         * DumpRenderTree/mac/FrameLoadDelegate.mm:
3703         (-[FrameLoadDelegate init]):
3704         (-[FrameLoadDelegate dealloc]):
3705         (-[FrameLoadDelegate webViewProgressFinishedNotification:]):
3706
3707 2011-11-09  Philippe Normand  <pnormand@igalia.com>
3708
3709         [GTK][DRT] window internals object is not reset after each test
3710         https://bugs.webkit.org/show_bug.cgi?id=71890
3711
3712         Reviewed by Martin Robinson.
3713
3714         Call WebCoreTestSupport::resetInternalsObject without going
3715         through DumpRenderTreeSupport.
3716
3717         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3718         (runTest):
3719
3720 2011-11-09  Beth Dakin  <bdakin@apple.com>
3721
3722         Build fix.
3723
3724         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
3725         (AccessibilityUIElement::uiElementAttributeValue):
3726
3727 2011-11-09  Beth Dakin  <bdakin@apple.com>
3728
3729         Speculative build fix.
3730
3731         * DumpRenderTree/AccessibilityUIElement.h:
3732
3733 2011-11-09  Julien Chaffraix  <jchaffraix@webkit.org>
3734
3735         Unreviewed build fix.
3736
3737         * DumpRenderTree/AccessibilityUIElement.cpp:
3738         (AccessibilityUIElement::makeJSAccessibilityUIElement):
3739         A bad merge added this check that is making the bots sad.
3740
3741 2011-11-09  Chris Fleizach  <cfleizach@apple.com>
3742
3743         AX: crash when accessing selectedTab in a tab list
3744         https://bugs.webkit.org/show_bug.cgi?id=70938
3745
3746         Reviewed by Beth Dakin.
3747
3748         Add the ability to retrieve an element through an arbitrary attribute.
3749
3750         * DumpRenderTree/AccessibilityUIElement.cpp:
3751         (uiElementAttributeValueCallback):
3752         (AccessibilityUIElement::uiElementAttributeValue):
3753         (AccessibilityUIElement::getJSClass):
3754         * DumpRenderTree/AccessibilityUIElement.h:
3755         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
3756         (AccessibilityUIElement::uiElementAttributeValue):
3757
3758 2011-11-09  Andy Wingo  <wingo@igalia.com>
3759
3760         Add webkitdirs.pm:getArchitecture implementation for GTK
3761         https://bugs.webkit.org/show_bug.cgi?id=71370
3762
3763         Reviewed by Xan Lopez.
3764
3765         * Scripts/webkitdirs.pm (determineArchitecture): Add an
3766         implementation for ports using autotools, so that a vanilla
3767         `run-javascriptcore-tests --gtk' invocation doesn't pass --32-bit
3768         to build-jsc on x86-64 machines.
3769
3770 2011-11-09  Simon Hausmann  <simon.hausmann@nokia.com>
3771
3772         Unreviewed, rolling out r99707.
3773         http://trac.webkit.org/changeset/99707
3774
3775         Re-enable until we can disable it together with tests in one
3776         shot.
3777
3778         * qmake/mkspecs/features/features.prf:
3779
3780 2011-11-09  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3781
3782         [Qt] Disable use of QtUiTools if the library/module is not found
3783
3784         We rely on the lookup order of prf files to inject our own uitools.prf
3785         that will verify that QtUiTools is in fact present before allowing the
3786         CONFIG += uitools to have any effect.
3787
3788         Reviewed by Simon Hausmann.
3789
3790         * DumpRenderTree/qt/DumpRenderTree.pro:
3791         * QtTestBrowser/QtTestBrowser.pro:
3792         * qmake/mkspecs/features/uitools.prf: Added.
3793
3794 2011-11-09  Simon Hausmann  <simon.hausmann@nokia.com>
3795
3796         Unreviewed, rolling out r99701.
3797         http://trac.webkit.org/changeset/99701
3798
3799         Breaks Qt mac build
3800
3801         * qmake/mkspecs/features/functions.prf:
3802
3803 2011-11-09  Simon Hausmann  <simon.hausmann@nokia.com>
3804
3805         [Qt] Prospective build fix for Win32/Mac with xmlpatterns
3806
3807         Reviewed by Tor Arne Vestbø.
3808
3809         * qmake/mkspecs/features/features.prf: Disable XSLT support initially.
3810         We default to XSLT support if qxmlpatterns is available. However qxmlpatterns
3811         is not up to the task of full XSLT support, so we should disable it.
3812
3813 2011-11-09  Simon Hausmann  <simon.hausmann@nokia.com>
3814
3815         [Qt] Enable exports and constructor functions in static libs
3816
3817         Reviewed by Tor Arne Vestbø.
3818
3819         * qmake/mkspecs/features/functions.prf: Use --whole-archive and their
3820         mac/win32-msvc equivalents when linking static libraries into a shared
3821         library or executable.
3822
3823 2011-11-09  Philippe Normand  <pnormand@igalia.com>
3824
3825         [GTK][DRT] window internals object is not reset after each test
3826         https://bugs.webkit.org/show_bug.cgi?id=71890
3827
3828         Reviewed by Xan Lopez.
3829
3830         * DumpRenderTree/gtk/DumpRenderTree.cpp:
3831         (runTest): reset the Internals object after each test.
3832
3833 2011-11-09  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3834
3835         [Qt] Detect presence of QtUiTools and enable it only when available
3836
3837         The disable_uitools option was a workaround for missing detection.
3838
3839         Reviewed by Simon Hausmann.
3840
3841         * QtTestBrowser/QtTestBrowser.pro:
3842         * qmake/mkspecs/features/default_post.prf:
3843         * qmake/mkspecs/features/default_pre.prf:
3844
3845 2011-11-09  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3846
3847         [Qt] Don't override config for the build_pass of debug_and_release
3848
3849         Qmake will parse the profile once for each configuration (debug and
3850         release), with the build_pass option set. In that situation we don't
3851         want to override what qmake is setting explicitly, otherwise the
3852         debug-step will end up hitting all the release-scopes.
3853
3854         Reviewed by Simon Hausmann.
3855
3856         * qmake/mkspecs/features/default_pre.prf:
3857
3858 2011-11-09  Simon Hausmann  <simon.hausmann@nokia.com>
3859
3860         [Qt] Unbreak the Qt build (copy & paste mistake - oops)
3861
3862         * qmake/mkspecs/features/functions.prf:
3863
3864 2011-11-09  Simon Hausmann  <simon.hausmann@nokia.com>
3865
3866         [Qt] Centralize duplicated code for linking the static libs
3867
3868         Reviewed by Tor Arne Vestbø.
3869
3870         * qmake/mkspecs/features/functions.prf:
3871         * qmake/mkspecs/features/javascriptcore.prf:
3872         * qmake/mkspecs/features/webcore.prf:
3873         * qmake/mkspecs/features/webkit2.prf:
3874
3875 2011-11-09  Simon Hausmann  <simon.hausmann@nokia.com>
3876
3877         [Qt] Build system cleanup
3878         https://bugs.webkit.org/show_bug.cgi?id=71815
3879
3880         Reviewed by Kenneth Rohde Christiansen.
3881
3882         * WebKitTestRunner/InjectedBundle/Target.pri: We don't need to link against the webkit2
3883         static library.
3884         * qmake/mkspecs/features/javascriptcore.prf: The wtf sources pull in a glib dependency
3885         (ownPtr functions for glib types). Propagate that to everyone using javascriptcore.
3886
3887 2011-11-08  Hayato Ito  <hayato@chromium.org>
3888
3889         [NRWT] Make single_test_runner aware of reference filename of test_input.
3890         https://bugs.webkit.org/show_bug.cgi?id=71567
3891
3892         Reviewed by Ryosuke Niwa.
3893
3894         Make single_test_runner honor a reference filename of given test_input so that
3895         we can use any file as reference html, instead of depending on implicit naming convention.
3896         The typical use case is to support w3c reftests.
3897
3898         No tests since no functional changes until there is a client.
3899         Test should be easily written with future clients.
3900
3901         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
3902         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
3903         * Scripts/webkitpy/layout_tests/models/test_failures.py:
3904         * Scripts/webkitpy/layout_tests/models/test_input.py:
3905
3906 2011-11-08  Lucas Forschler  <lforschler@apple.com>
3907
3908         https://bugs.webkit.org/show_bug.cgi?id=71839
3909         Update the leaks bot OS to Lion.
3910         Update the config.json to reflect the update.
3911
3912         Reviewed by Stephanie Lewis.
3913
3914         * BuildSlaveSupport/build.webkit.org-config/config.json:
3915
3916 2011-11-08  Sheriff Bot  <webkit.review.bot@gmail.com>
3917
3918         Unreviewed, rolling out r99626.
3919         http://trac.webkit.org/changeset/99626
3920         https://bugs.webkit.org/show_bug.cgi?id=71866
3921
3922         "C++ exceptions cannot be disabled without influencing the
3923         Objective-C exceptions" (Requested by benjaminp on #webkit).
3924
3925         * DumpRenderTree/mac/Configurations/DumpRenderTree.xcconfig:
3926
3927 2011-11-08  Benjamin Poulain  <bpoulain@apple.com>
3928
3929         [Mac] Disable C++ exceptions from DumpRenderTree
3930         https://bugs.webkit.org/show_bug.cgi?id=71834
3931
3932         Reviewed by Darin Adler.
3933
3934         DumpRenderTree Mac does not handle C++ exceptions. When such exception
3935         happens, it tends to cause the AutoReleasePool stack to be corrupted,
3936         which leads to crashes long after the exception was raised.
3937
3938         This patch disabled C++ exceptions in order to catch the problems where
3939         they happen when a C++ exception is raised.
3940
3941         * DumpRenderTree/mac/Configurations/DumpRenderTree.xcconfig:
3942
3943 2011-11-08  Julien Chaffraix  <jchaffraix@webkit.org>
3944
3945         Moved myself to the reviewer list (yay!).
3946
3947         * Scripts/webkitpy/common/config/committers.py:
3948
3949 2011-11-08  Zeno Albisser  <zeno@webkit.org>
3950
3951         Added myself as a committer.
3952
3953         * Scripts/webkitpy/common/config/committers.py:
3954
3955 2011-11-08  Andy Wingo  <wingo@igalia.com>
3956
3957         Skip some ECMAScript tests that fail on Linux in some time zones.
3958         https://bugs.webkit.org/show_bug.cgi?id=71371
3959
3960         Reviewed by Martin Robinson.
3961
3962         * Scripts/run-javascriptcore-tests: Skip some more tests.
3963
3964 2011-11-08  Csaba Osztrogonác  <ossy@webkit.org>
3965
3966         [Qt] Unreviewed trivial fix after build system refactoring.
3967
3968         * Scripts/run-qtwebkit-tests: Fix path of QtWebProcess for QML tests.
3969
3970 2011-11-08  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3971
3972         [Qt] Don't set OBJECTS_DIR and DEFINES in default_pre
3973
3974         If a pro file needs to access the OBJECTS_DIR it should use the
3975         form ${QMAKE_VAR_OBJECTS_DIR} to ensure the value is evaluated
3976         after defaults_post has been processed.
3977
3978         Setting DEFINES in defaults_pre is also bad, as any DEFINES=foo on
3979         the command line will override whatever we set in defaults_pre.
3980
3981         Reviewed by Simon Hausmann.
3982
3983         * qmake/mkspecs/features/default_post.prf:
3984         * qmake/mkspecs/features/default_pre.prf:
3985
3986 2011-11-08  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
3987
3988         [Qt] Prevent duplicated sources in SOURCES/OBJECTIVE_SOURCES
3989
3990         Reviewed by Simon Hausmann.
3991
3992         * qmake/mkspecs/features/default_post.prf:
3993
3994 2011-11-07  Zeno Albisser  <zeno.albisser@nokia.com>
3995
3996         [Qt][WK2] MiniBrowser should use urlFromUserInput()
3997         https://bugs.webkit.org/show_bug.cgi?id=71680
3998
3999         Use urlFromUserInput() / QUrl::fromUserInput() to automatically
4000         complete urls in MiniBrowser.
4001
4002         Reviewed by Tor Arne Vestbø.
4003
4004         * MiniBrowser/qt/BrowserWindow.cpp:
4005         (BrowserWindow::BrowserWindow):
4006         (BrowserWindow::load):
4007         * MiniBrowser/qt/qml/BrowserWindow.qml:
4008         * MiniBrowser/qt/utils.cpp:
4009         (Utils::urlFromUserInput):
4010         * MiniBrowser/qt/utils.h:
4011         (Utils::Utils):
4012
4013 2011-11-07  Yuqiang Xian  <yuqiang.xian@intel.com>
4014
4015         Added myself as a committer.
4016
4017         * Scripts/webkitpy/common/config/committers.py:
4018
4019 2011-11-07  Csaba Osztrogonác  <ossy@webkit.org>
4020
4021         Unreviewed webkitpy test fix after r99512.
4022
4023         * Scripts/webkitpy/tool/commands/suggestnominations_unittest.py:
4024         Xianzhu Wang is now a committer, so we can't nominate him as a committer anymore.
4025
4026 2011-11-07  Xianzhu Wang  <wangxianzhu@chromium.org>
4027
4028         Added myself as a committer.
4029
4030         * Scripts/webkitpy/common/config/committers.py:
4031
4032 2011-11-07  Sheriff Bot  <webkit.review.bot@gmail.com>
4033
4034         Unreviewed, rolling out r99470.
4035         http://trac.webkit.org/changeset/99470
4036         https://bugs.webkit.org/show_bug.cgi?id=71740
4037
4038         "Broke test-webkitpy" (Requested by eseidel2 on #webkit).
4039
4040         * Scripts/webkitpy/common/config/watchlist:
4041
4042 2011-11-07  Tony Chang  <tony@chromium.org>
4043
4044         [chromium] update flakiness dashboard to reflect current bots
4045
4046         Reviewed by Ojan Vafai.
4047
4048         * TestResultServer/static-dashboards/builders.js: Remove (deps) debug
4049             bots and merged linux debug bots into a single bot.
4050         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
4051         (setupExpectationsTest):
4052
4053 2011-11-07  Dana Jansens  <danakj@chromium.org>
4054
4055         Add cc-bugs group to watch changes in chromium graphics
4056         https://bugs.webkit.org/show_bug.cgi?id=71690
4057
4058         Reviewed by James Robinson.
4059
4060         * Scripts/webkitpy/common/config/watchlist:
4061
4062 2011-11-07  Balazs Ankes  <Ankes.Balazs@stud.u-szeged.hu>
4063
4064         [NRWT] Parsing of test_expectations.txt should be agnostic to newline at end
4065         https://bugs.webkit.org/show_bug.cgi?id=70912
4066
4067         Reviewed by Ojan Vafai.
4068
4069         * Scripts/webkitpy/layout_tests/port/webkit.py: Add newline at the end of test_expectations.txt.
4070         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py: Make up the test_test_expectations method for the updated webkit.py.
4071
4072 2011-11-07  Eric Seidel  <eric@webkit.org>
4073
4074         new-run-webkit-tests: support --repeat-each feature from old-run-webkit-tests
4075         https://bugs.webkit.org/show_bug.cgi?id=62199
4076
4077         Reviewed by Dirk Pranke.
4078
4079         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4080         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4081         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
4082
4083 2011-11-07  Fady Samuel  <fsamuel@chromium.org>
4084
4085         [Chromium] Rename WebView::scalePage to WebView::setPageScaleFactor to match WebCore
4086         https://bugs.webkit.org/show_bug.cgi?id=71485
4087
4088         Reviewed by Darin Fisher.
4089
4090         * DumpRenderTree/chromium/EventSender.cpp:
4091         (EventSender::scalePageBy):
4092         * DumpRenderTree/chromium/TestShell.cpp:
4093         (TestShell::resetTestController):
4094
4095 2011-11-07  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4096
4097         [Qt] Put the jsc binary in 'bin' instead of leaving it deep in the build tree
4098
4099         Allows us to not package up the whole Source/JavaScriptCore directory for the
4100         buildbots.
4101
4102         Reviewed-by Simon Hausmann.
4103
4104         * BuildSlaveSupport/built-product-archive:
4105         * Scripts/webkitdirs.pm:
4106         (jscProductDir):
4107
4108 2011-11-07  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4109
4110         [Qt] Ensure we always export symbols for the QtWebKit API when building WebKit
4111
4112         Reviewed-by Simon Hausmann.
4113
4114         * qmake/mkspecs/features/default_post.prf:
4115
4116 2011-11-07  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4117
4118         [Qt] Sanitize qmake paths correctly before comparing
4119
4120         The qmake function toSanitizedPath had a bug causing empty paths on MinGW,
4121         which caused us to always generate derived webcore sources, but never build
4122         them. The bug has been fixed, and we now also sanitize paths for the injected
4123         bundle derived sources.
4124
4125         Reviewed by Simon Hausmann.
4126
4127         * WebKitTestRunner/InjectedBundle/DerivedSources.pri:
4128         * qmake/mkspecs/features/functions.prf:
4129
4130 2011-11-07  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4131
4132         [Qt] Print right order for changing feature defines
4133
4134         Reviewed by Simon Hausmann.
4135
4136         * Scripts/webkitdirs.pm:
4137         (buildQMakeProject):
4138
4139 2011-11-06  Csaba Osztrogonác  <ossy@webkit.org>
4140
4141         [Qt] Unreviewed trivial fix after build system refactoring.
4142
4143         * Scripts/run-qtwebkit-tests: Fix qml_import_path.
4144
4145 2011-11-06  Eric Seidel  <eric@webkit.org>
4146
4147         Clarify how the symbol and runtime-feature based test exclusion works and cleanup the code a bit
4148         https://bugs.webkit.org/show_bug.cgi?id=66078
4149
4150         Unreviewed.  Add an assert about the type of "args"
4151         to our Executive Mocks and fix a piece of code
4152         from this patch which was hitting that assert
4153         when executing for real.
4154
4155         * Scripts/webkitpy/common/system/executive_mock.py:
4156         * Scripts/webkitpy/layout_tests/port/webkit.py:
4157
4158 2011-11-06  Eric Seidel  <eric@webkit.org>
4159
4160         Split chunk handling out of prepare_lists_and_print_output to make it more readable
4161         https://bugs.webkit.org/show_bug.cgi?id=71629
4162
4163         Reviewed by Adam Barth.
4164
4165         * Scripts/webkitpy/layout_tests/controllers/manager.py:
4166
4167 2011-11-06  Eric Seidel  <eric@webkit.org>
4168
4169         Clarify how the symbol and runtime-feature based test exclusion works and cleanup the code a bit
4170         https://bugs.webkit.org/show_bug.cgi?id=66078
4171
4172         Reviewed by Adam Barth.
4173
4174         The runtime feature detection was fixed to work in bug 64472.
4175         In this bug I moved the symbol-based feature detection from
4176         popen() to Executive.run_command and cleaned up the callers
4177         and unittests to make sure that we're correctly parsing the
4178         nm output correctly.  The old code happened to work even though
4179         the runtime-features path was using "str in list" and the
4180         symbol features path was using "str in str" and it just happened
4181         to do what we wanted to.  Now runtime features and symbol feature
4182         blacklists are computed separately (and with better documentation).
4183
4184         This system remains confusing, partially because these are black-lists
4185         which are amended to whatever static blacklist may exist for the
4186         port as part of a Skipped list file.
4187
4188         For example, notice how the runtime feature list only has directory
4189         blacklists for a couple features.  If all features are off,
4190         how do we skip enough tests with only 2 entries in the blacklist map?
4191         The answer is that Windows is the only port to use runtime feature
4192         detection, and the win/Skipped file turns off all the other features
4193         statically (like mathml, mhtml, wss, etc.) where as some other ports (like AppleMac)
4194         which use symbol-based feature detection turn of mathml, wcss, etc
4195         using the blacklists found in _missing_symbol_to_skipped_tests.
4196
4197         I also noticed a couple places where we still referenced xhtmlmp
4198         even though support for such has been removed from WebKit.  Removed those.
4199
4200         This should result in no functional change.
4201
4202         * Scripts/webkitpy/layout_tests/port/gtk.py:
4203          - Use self._filesystem instead of os.path
4204         * Scripts/webkitpy/layout_tests/port/webkit.py:
4205         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
4206
4207 2011-11-06  Dan Bernstein  <mitz@apple.com>
4208
4209         Made show-pretty-diff work when the input path contains spaces.
4210
4211         Reviewed by Anders Carlsson.
4212
4213         * Scripts/show-pretty-diff:
4214
4215 2011-11-05  Daniel Cheng  <dcheng@chromium.org>
4216
4217         [chromium] Use the security origin instead of the URL when checking notification permissions
4218         https://bugs.webkit.org/show_bug.cgi?id=71590
4219
4220         Reviewed by Adam Barth.
4221
4222         * DumpRenderTree/chromium/NotificationPresenter.cpp:
4223         (NotificationPresenter::checkPermission):
4224         * DumpRenderTree/chromium/NotificationPresenter.h:
4225
4226 2011-11-05  Darin Adler  <darin@apple.com>
4227
4228         Fix a small leak in Mac version of LayoutTestController
4229         https://bugs.webkit.org/show_bug.cgi?id=71615
4230
4231         Reviewed by Mark Rowe.
4232
4233         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
4234         (LayoutTestController::applicationCacheDiskUsageForOrigin): Release the origin.
4235         (LayoutTestController::localStorageDiskUsageForOrigin): Ditto.
4236
4237 2011-11-04  Eric Seidel  <eric@webkit.org>
4238
4239         Upgrade to the latest Mechanize
4240         https://bugs.webkit.org/show_bug.cgi?id=71594
4241
4242         Reviewed by Adam Barth.
4243
4244         I noticed our mechanize was somewhat out of date when removing Python 2.5 support.
4245         Mechanize is no longer two separate libraries, so importing it is much simpler.
4246
4247         * Scripts/webkitpy/thirdparty/__init__.py:
4248
4249 2011-11-04  Stephen Chenney  <schenney@chromium.org>
4250
4251         Crash in ScrollAnimator.cpp
4252         https://bugs.webkit.org/show_bug.cgi?id=69865
4253
4254         The code in ScrollAnimator assumes that horizontal per-page mouse
4255         wheel events cannot happen, which is not true. This patch adds layout
4256         tests for all paging wheel event situations and fixes the broken
4257         horizontal case.
4258
4259         Reviewed by Anders Carlsson
4260
4261         * DumpRenderTree/chromium/EventSender.cpp:
4262         (EventSender::handleMouseWheel): Added the ability to specify that a
4263         mouse wheel event should be paging.
4264
4265 2011-11-04  Eric Seidel  <eric@webkit.org>
4266
4267         Remove deprecated free functions in port.factory
4268         https://bugs.webkit.org/show_bug.cgi?id=71494
4269
4270         Unreviewed.  Fix exception when running new-run-webkit-httpd.
4271
4272         * Scripts/new-run-webkit-httpd:
4273
4274 2011-11-04  Adam Roben  <aroben@apple.com>
4275
4276         Remove Leopard-specific code from build.webkit.org
4277
4278         We don't have any Leopard bots anymore.
4279
4280         Fixes <http://webkit.org/b/71583> build.webkit.org configuration contains obsolete Leopard code
4281
4282         Reviewed by Mark Rowe.
4283
4284         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Removed now-unused StartATSServer
4285         and StopATSServer steps.
4286         (unitTestsSupported): Removed check for mac-leopard.
4287         (TestFactory.__init__): Removed Leopard-specific steps.
4288
4289 2011-11-04  Eric Seidel  <eric@webkit.org>
4290
4291         new-run-webkit-tests autoinstalls python-irclib even though it doesn't need to
4292         https://bugs.webkit.org/show_bug.cgi?id=71549
4293
4294         Reviewed by Adam Barth.
4295
4296         Several members of Host didn't really belong there
4297         and were instead specific to WebKitPatch, so I've moved
4298         them up onto that class.
4299
4300         * Scripts/webkitpy/common/host.py:
4301         * Scripts/webkitpy/common/host_mock.py:
4302         * Scripts/webkitpy/tool/main.py:
4303         * Scripts/webkitpy/tool/mocktool.py:
4304         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
4305
4306 2011-11-04  Tony Chang  <tony@chromium.org>
4307
4308         garden-o-matic: bring back party time!
4309         https://bugs.webkit.org/show_bug.cgi?id=71582
4310
4311         Reviewed by Adam Barth.
4312
4313         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html: Allow images from file:
4314         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/images/partytime.gif: Renamed from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/partytime.gif.
4315         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js: Add length() for UpdateTracker.
4316         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js: Plumb through length().
4317         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js: Toggle partytime.gif via css class.
4318         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/onebar.css: Add css styling for partytime.
4319
4320 2011-11-04  Raphael Kubo da Costa  <kubo@profusion.mobi>
4321
4322         [EFL] DRT: Fix the order of the items in the dumped history list.
4323         https://bugs.webkit.org/show_bug.cgi?id=71562
4324
4325         It did not make much sense to hold the history list in a map, as we do not need
4326         the keys. And the printed items must be sorted by target, which is more easily
4327         done with a vector.
4328
4329         Should make tests such as http/tests/navigation/error404-frames.html pass.
4330
4331         Reviewed by Antonio Gomes.
4332
4333         * DumpRenderTree/efl/DumpHistoryItem.cpp:
4334         (compareHistoryItemsByTarget):
4335         (dumpHistoryItem):
4336
4337 2011-11-04  Patrick Gansterer  <paroga@webkit.org>
4338
4339         [Qt] Remove ENABLE_SQLITE from qmake files
4340         https://bugs.webkit.org/show_bug.cgi?id=71546
4341
4342         Reviewed by Simon Hausmann.
4343
4344         * qmake/mkspecs/features/features.prf:
4345         * qmake/mkspecs/features/webcore.prf:
4346
4347 2011-11-04  Adam Barth  <abarth@webkit.org>
4348
4349         watchlist for WebIDL shouldn't trigger on Internals.idl
4350         https://bugs.webkit.org/show_bug.cgi?id=70657
4351
4352         Reviewed by David Levin.
4353
4354         Internals.idl isn't part of the Web-facing IDL in the project.
4355
4356         * Scripts/webkitpy/common/config/watchlist:
4357
4358 2011-11-04  Anders Carlsson  <andersca@apple.com>
4359
4360         Fix WKTR crash when running NPN_ConvertPoint test.
4361
4362         Call notifyDone in a timeout - otherwise we'll end up entering layout from layout.
4363
4364         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp:
4365         (ConvertPoint::NPP_SetWindow):
4366
4367 2011-11-04  Anders Carlsson  <andersca@apple.com>
4368
4369         NPN_ConvertPoint plug-in test should be called from NPP_SetWindow
4370         https://bugs.webkit.org/show_bug.cgi?id=71570
4371
4372         Reviewed by Simon Fraser.
4373
4374         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
4375         (PluginTest::NPP_SetWindow):
4376         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
4377         Remove the NPP parameter from NPP_SetWindow.
4378
4379         * DumpRenderTree/TestNetscapePlugIn/Tests/NPPSetWindowCalledDuringDestruction.cpp:
4380         (NPPSetWindowCalledDuringDestruction::NPP_SetWindow):
4381         Remove the NPP parameter from NPP_SetWindow.
4382
4383         * DumpRenderTree/TestNetscapePlugIn/Tests/PassDifferentNPPStruct.cpp:
4384         (PassDifferentNPPStruct::NPP_SetWindow):
4385         Remove the NPP parameter from NPP_SetWindow and use PluginTest::log for logging.
4386
4387         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp:
4388         (ConvertPoint::NPP_SetWindow):
4389         Run the test from NPP_SetWindow and call notifyDone() when done.
4390
4391         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
4392         (NPP_SetWindow):
4393         Remove unnecessary parameter.
4394
4395 2011-11-04  Simon Hausmann  <simon.hausmann@nokia.com>
4396
4397         [Qt] Fix run-qtwebkit-tests execution on the build bot(s).
4398
4399         Rubber-stamped by Tor Arne Vestbø.
4400
4401         * BuildSlaveSupport/build.webkit.org-config/master.cfg: The path to the tests
4402         changed slightly after the refactoring.
4403
4404 2011-11-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4405
4406         [Qt] Don't warn about disabling WebKit2 when qmake does recursive includes
4407
4408         Reviewed by Simon Hausmann.
4409
4410         * qmake/mkspecs/features/default_pre.prf:
4411
4412 2011-11-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4413
4414         [Qt] Don't error out on wrong Qt version when qmake does recursive includes
4415
4416         Reviewed by Simmon Hausmann.
4417
4418         * qmake/mkspecs/features/default_pre.prf:
4419
4420 2011-11-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4421
4422         [Qt] Disable STDIN check before cleaning build dir
4423
4424         It didn't work on the bots.
4425
4426         Reviewed by Ossy.
4427
4428         * Scripts/webkitdirs.pm:
4429
4430 2011-11-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4431
4432         [Qt] Don't build all of webkit when running build-jsc
4433
4434         In case the whole webkit project was built, but we then subsequently
4435         want to build only JSC, we need to run make in the proper subdirectory.
4436
4437         This also means the incremental target needs to be added to all
4438         makefiles.
4439
4440         Reviewed by Simmon Hausmann.
4441
4442         * Scripts/webkitdirs.pm:
4443         (buildQMakeProject):
4444         * qmake/mkspecs/features/default_post.prf:
4445
4446 2011-11-04  Simon Hausmann  <simon.hausmann@nokia.com>
4447
4448         Fix run-javascriptcore-tests for Qt when it's executed by the bot.
4449
4450         Reviewed by Tor Arne Vestbø.
4451
4452         * Scripts/webkitdirs.pm:
4453         (jscProductDir): jsc path changed.
4454
4455 2011-11-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4456
4457         [Qt] Error out early if trying to build against Qt < 4.8
4458
4459         Reviewed by Simon Hausmann.
4460
4461         * qmake/mkspecs/features/default_pre.prf:
4462
4463 2011-11-04  Simon Hausmann  <simon.hausmann@nokia.com>
4464
4465         Fix run-launcher for Qt and Gtk.
4466
4467         Rubber-stamped by Tor Arne Vestbø.
4468
4469         * Scripts/webkitdirs.pm:
4470         (isWK2): Re-add isWK2(), as it's used in run-launcher.
4471
4472 2011-11-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4473
4474         [Qt] Ignore '-2' if passed to build-webkit. WebKit2 is now built by default.
4475
4476         Reviewed by Simon Hausmann.
4477
4478         * Scripts/build-webkit:
4479
4480 2011-11-04  Carlos Garcia Campos  <cgarcia@igalia.com>
4481
4482         [GTK] Use web view title as window title in MiniBrowser
4483         https://bugs.webkit.org/show_bug.cgi?id=71544
4484
4485         Reviewed by Philippe Normand.
4486
4487         * MiniBrowser/gtk/BrowserWindow.c:
4488         (webViewURIChanged): Removed extra space.
4489         (webViewTitleChanged): Set window title using
4490         webkit_web_view_get_title().
4491         (browser_window_init): Set initial default window title.
4492         (browserWindowConstructed): Connect to WebView notify::title
4493         signal.
4494
4495 2011-11-04  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
4496
4497         [Qt] Refactor and clean up the qmake build system
4498
4499         The qmake build system has accumulated a bit of cruft and redundancy
4500         over time. There's also a fairly tight coupling between how to build
4501         the various targets, and _what_ to build, making it harder to add new
4502         rules or sources. This patch aims to elevate these issues somewhat.
4503
4504         This is a short-list of the changes:
4505
4506           * The rules for how to build targets are now mostly contained as
4507             prf-files in Tools/qmake/mkspecs/features. Using mkspecs also
4508             allows us to do pre- and post-processing of each project file,
4509             which helps to clean up the actual project files.
4510
4511           * Derived sources are no longer generated as a separate make-step
4512             but is part of each target's project file as a subdir. Makefile
4513             rules are used to ensure that we run make on the derived sources
4514             before running qmake on the actual target makefile. This makes
4515             it easier to keep a proper dependency between derived sources
4516             and the target.
4517
4518           * We use GNU make and the compiler to generate dependencies on
4519             UNIX-based systems running Qt 5. This allows us to lessen the
4520             need to run qmake, which should reduce compile time.
4521
4522           * WebKit2 is now build by default if building with Qt 5. It can
4523             be disabled by passing --no-webkit2 to build-webkit.
4524
4525         The result of these changes are hopefully a cleaner and easier
4526         build system to modify, and faster build times due to no longer
4527         running qmake on every single build. It's also a first step
4528         towards possibly generating the list of sources using another
4529         build system.
4530
4531         https://bugs.webkit.org/show_bug.cgi?id=71222
4532
4533         Reviewed by Simon Hausmann.
4534
4535         * DerivedSources.pro: Removed.
4536         * DumpRenderTree/qt/DumpRenderTree.pro:
4537         * DumpRenderTree/qt/ImageDiff.pro:
4538         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
4539         * MiniBrowser/qt/MiniBrowser.pro:
4540         * QtTestBrowser/QtTestBrowser.pro:
4541         * QtTestBrowser/launcherwindow.cpp:
4542         (LauncherWindow::initializeView):
4543         (LauncherWindow::createChrome):
4544         (LauncherWindow::screenshot):
4545         * QtTestBrowser/launcherwindow.h:
4546         (WindowOptions::WindowOptions):
4547         * Scripts/build-webkit:
4548         (unlinkZeroFiles):
4549         * Scripts/webkitdirs.pm:
4550         (qtFeatureDefaults):
4551         (promptUser):
4552         (buildQMakeProject):
4553         * Tools.pro:
4554         * WebKitTestRunner/DerivedSources.pri: Added.
4555         * WebKitTestRunner/DerivedSources.pro: Removed.
4556         * WebKitTestRunner/InjectedBundle/DerivedSources.pri: Added.
4557         * WebKitTestRunner/InjectedBundle/InjectedBundle.pro: Added.
4558         * WebKitTestRunner/InjectedBundle/Target.pri: Added.
4559         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro: Removed.
4560         * WebKitTestRunner/Target.pri: Added.
4561         * WebKitTestRunner/WebKitTestRunner.pro:
4562         * WebKitTestRunner/qt/TestInvocationQt.cpp:
4563         * WebKitTestRunner/qt/WebKitTestRunner.pro: Removed.
4564         * qmake/README: Added.
4565         * qmake/mkspecs/features/default_post.prf: Added.
4566         * qmake/mkspecs/features/default_pre.prf: Added.
4567         * qmake/mkspecs/features/features.prf: Renamed from Source/WebCore/features.pri.
4568         * qmake/mkspecs/features/functions.prf: Added.
4569         * qmake/mkspecs/features/javascriptcore.prf: Renamed from Source/JavaScriptCore/JavaScriptCore.pri.
4570         * qmake/mkspecs/features/mac/default_post.prf: Added.
4571         * qmake/mkspecs/features/mac/default_pre.prf: Added.
4572         * qmake/mkspecs/features/minimal_qt.prf: Added.
4573         * qmake/mkspecs/features/qtwebkit.prf: Added.
4574         * qmake/mkspecs/features/rpath.prf: Added.
4575         * qmake/mkspecs/features/unix/default_post.prf: Added.
4576         * qmake/mkspecs/features/unix/default_pre.prf: Added.
4577         * qmake/mkspecs/features/valgrind.prf: Added.
4578         * qmake/mkspecs/features/webcore.prf: Renamed from Source/WebCore/WebCore.pri.
4579         * qmake/mkspecs/features/webkit2.prf: Copied from Source/WebKit2/WebKit2.pri.
4580         * qmake/mkspecs/features/win32/default_post.prf: Added.
4581         * qmake/qt_webkit.pri: Renamed from Source/WebKit/qt/qt_webkit_version.pri.
4582         * qmake/syncqt-4.8: Added.
4583         (showUsage):
4584         ():
4585         (copyFile):
4586         (symlinkFile):
4587
4588 2011-11-04  Eric Seidel  <eric@webkit.org>
4589
4590         Remove deprecated free functions in port.factory
4591         https://bugs.webkit.org/show_bug.cgi?id=71494
4592
4593         Unreviewed.  Fixing exception seen when running test-webkitpy.
4594
4595         Sorry, I previously had deleted the relevant rebaseline-chromium-webkit-tests
4596         unittest, since that script is nearly ready to be deleted.  But I added
4597         it back at the last second and failed to run the tests. :(
4598         This fixes the exception the bots were seeing.
4599
4600         * Scripts/webkitpy/style/checkers/test_expectations.py:
4601         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
4602         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
4603
4604 2011-11-02  Xiaomei Ji  <xji@chromium.org>
4605
4606         Enable ctrl-arrow move cursor by word in visual order in cr-win by command line flag.
4607         https://bugs.webkit.org/show_bug.cgi?id=71163
4608
4609         Reviewed by Ryosuke Niwa.
4610
4611         * DumpRenderTree/chromium/WebPreferences.cpp: Set visualWordMovementEabled to false in DRT.
4612         (WebPreferences::applyTo):
4613
4614 2011-11-03  Eric Seidel  <eric@webkit.org>
4615
4616         Remove deprecated free functions in port.factory
4617         https://bugs.webkit.org/show_bug.cgi?id=71494
4618
4619         Reviewed by Adam Barth.
4620
4621         I removed port.factory.get, get_all, and all_port_names -- long-since deprecated
4622         free functions in factory.py.  To remove these required fixing all callsites
4623         to use MockHost, Host, or PortFactory where appropriate.  After this change
4624         we're only left with a handfull of places where we inappropraitely use a
4625         real Executive, FileSystem or User object during unittesting.
4626
4627         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
4628          - These PortFactory overrides were from an earlier edition of my previous patch
4629            in the end, I removed MockPortFactory so all tests ended up with this same
4630            real PortFactory() with a MockHost.  This was simply redundent code.
4631         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
4632          - Use MockHost to get a mocked-out PortFactory.
4633         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
4634          - This code has no access to a Host object, so we just use a default PortFactory
4635            for now.  This will still incorrectly create real Executive/FileSystem objects
4636            durring unittesting, which is wrong. :(
4637         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
4638          - Use a MockHost to get a mocked PortFactory.
4639         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py:
4640          - Use a MockHost to get a mocked PortFactory.
4641         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py:
4642          - ditto.
4643         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
4644          - ditto.
4645         * Scripts/webkitpy/layout_tests/port/__init__.py:
4646          - factory.get no longer exists.  Yay!
4647         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
4648          - Use a MockHost to get a mocked PortFactory.
4649            To do this efficiently, I created a helper function
4650            and combined all these (mostly redundant) tests into one test
4651            with several calls to the helper function.
4652         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
4653          - Pass MockUser and MockExecutive in these port testing subclasses.
4654          - I also removed a bunch of unecessary uses of mocktool. namespacing.
4655         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
4656          - Awesome windows hacks should only happen on win32.  Once we started
4657            using MockExecutives these 'svn help' commands started printing in
4658            other tests, which is how I found this bug.
4659         * Scripts/webkitpy/layout_tests/port/dryrun.py:
4660          - DryRunPort has no Host, so just use a PortFactory() directly.
4661            This will do the wrong thing if we're unittesting the DryRun port
4662            but it's no worse than the existing code.
4663         * Scripts/webkitpy/layout_tests/port/factory.py:
4664          - Remove the free functions.
4665          - Fix argument overriding to be able to override None.
4666         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
4667          - Use a mocked-out PortFactory during testing.
4668         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
4669          - Pass MockExecutive and MockUser to these manual Port instantiations.
4670         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
4671          - No access to a Host object here, so just using PortFactory directly.
4672            This is wrong, but no more wrong than the existing code was.
4673         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
4674          - Use a MockHost to get a mocked out PortFactory.
4675         * Scripts/webkitpy/layout_tests/port/test.py:
4676          - Use our fancy _set_default_overriding_none system to clean this code up and pass a MockExecutive.
4677         * Scripts/webkitpy/layout_tests/port/win.py:
4678          - Don't run cmd /c ver from WinPort.__init__ or it shows up in unittests which have logging MockExecutive
4679            now that we're actually passing MockExecutive to most Port instantiations during testing.
4680         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
4681          - Pass our new "unittesting" bool to _detect_version when actually unittesting.
4682         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
4683          - Host is the future.
4684         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
4685          - Use a MockHost to get a mocked out PortFactory.  Once we have a mocked PortFactory
4686            we don't need to pass MockUser explicitly, it does that automatically for us.
4687         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
4688          - Use a MockHost to get a mocked out PortFactory.
4689         * Scripts/webkitpy/style/checkers/test_expectations.py:
4690          - We don't have access to a Host object, so using PortFactory() directly.
4691         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
4692          - Remove unused variable.
4693         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
4694          - Host isn't accessible here, so using PortFactory() directly, which
4695            although unfortunate is no worse than before.
4696         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
4697          - Testing is much easier when you don't need to hack global state...
4698         * Scripts/webkitpy/tool/commands/expectations.py:
4699          - Use the port_factory on the tool.
4700         * Scripts/webkitpy/tool/commands/rebaseline.py:
4701          - ditto.
4702         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
4703          - This unittest was wrong.  It was trying to pass the tool object, but failing to pass
4704            the arguments in the right order.  Fixing that made it use a MockFileSystem
4705            (like it had been trying to), which then required us to actually populate
4706            that MockFileSystem with fake expectation files (instead of using the ones on the real disk).
4707         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
4708          - Get the port_factory from the tool.
4709         * Scripts/webkitpy/tool/servers/gardeningserver.py:
4710          - Ditto.
4711         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
4712          - Use a MockHost to get a mocked out Port object.
4713
4714 2011-11-03  Raphael Kubo da Costa  <kubo@profusion.mobi>
4715
4716         [EFL] DRT: Allow choosing which backing store to use.
4717         https://bugs.webkit.org/show_bug.cgi?id=70532
4718
4719         Reviewed by Antonio Gomes.
4720
4721         This change makes it possible to change the backing store used by DRT
4722         from Tiled Backing Store (TBS) to Single Backing Store (SBS) by
4723         setting the environment variable DRT_USE_SINGLE_BACKING_STORE to 1.
4724
4725         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
4726         (DumpRenderTreeChrome::createView):
4727         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
4728         (shouldUseSingleBackingStore):
4729         (chooseAndInitializeAppropriateSmartClass):
4730         (drtViewAdd):
4731         * DumpRenderTree/efl/DumpRenderTreeView.h:
4732
4733 2011-11-02  Erik Arvidsson  <arv@chromium.org>
4734
4735         JS Test Harness: Remove more link tags
4736         https://bugs.webkit.org/show_bug.cgi?id=71339
4737
4738         Reviewed by Ojan Vafai.
4739
4740         * Scripts/make-new-script-test:
4741         (writeTestFile): Don't output <link>
4742
4743 2011-11-03  Thouraya ANDOLSI  <thouraya.andolsi@st.com>
4744
4745         Reviewed by Martin Robinson.
4746
4747         [GTK] arguments passed to build-jsc are not taken into account
4748         https://bugs.webkit.org/show_bug.cgi?id=58333
4749
4750         * Scripts/build-jsc:
4751
4752 2011-11-03  Carlos Garcia Campos  <cgarcia@igalia.com>
4753
4754         [GTK] Show url of history items in a status bar in MiniBrowser
4755         https://bugs.webkit.org/show_bug.cgi?id=71474
4756
4757         Reviewed by Martin Robinson.
4758
4759         * MiniBrowser/gtk/BrowserWindow.c:
4760         (browserWindowSetStatusText): Set status text and show/hide the
4761         status label.
4762         (resetStatusText): Reset status text when history menu is hidden.
4763         (browserWindowHistoryItemSelected): Show url of currently selected
4764         history item.
4765         (browserWindowCreateBackForwardMenu): Connect to hide signal of
4766         menu to reset the status text.
4767         (browserWindowConstructed): Use GtkOverlay if available to show
4768         status text.
4769
4770 2011-11-03  Carlos Garcia Campos  <cgarcia@igalia.com>
4771
4772         [GTK] Add back/forward menus to MiniBrowser using WebKit2 GTK+ API
4773         https://bugs.webkit.org/show_bug.cgi?id=71466
4774
4775         Reviewed by Martin Robinson.
4776
4777         * MiniBrowser/gtk/BrowserWindow.c:
4778         (browserWindowHistoryItemActivated): Go to selected back formard
4779         list item using webkit_web_view_go_to_back_forward_list_item()
4780         (browserWindowCreateBackForwardMenu): Create a GtkMenu for the
4781         given GList of WebKitBackForwardListItems.
4782         (browserWindowUpdateNavigationActions): Enable/disable back and
4783         forward buttons depending on whether it's possible to go
4784         back/forward. Create a menu for back and forward toolbar buttons.
4785         (backForwadlistChanged): Call browserWindowUpdateNavigationActions().
4786         (browserWindowConstructed): Connect to WebKitBackForwardList
4787         changed signal.
4788
4789 2011-11-03  Simon Hausmann  <simon.hausmann@nokia.com>
4790
4791         [Qt] Remove Maemo specific code paths
4792         https://bugs.webkit.org/show_bug.cgi?id=71476
4793
4794         Reviewed by Kenneth Rohde Christiansen.
4795
4796         * QtTestBrowser/launcherwindow.cpp:
4797         (LauncherWindow::~LauncherWindow):
4798         (LauncherWindow::init):
4799         (LauncherWindow::showLinkHover):
4800         (LauncherWindow::selectElements):
4801         (LauncherWindow::showFPS):
4802         (LauncherWindow::updateFPS):
4803         * QtTestBrowser/launcherwindow.h:
4804         (WindowOptions::WindowOptions):
4805         * QtTestBrowser/locationedit.cpp:
4806         * QtTestBrowser/locationedit.h:
4807
4808 2011-11-03  Carlos Garcia Campos  <cgarcia@igalia.com>
4809
4810         [GTK] Show load progress information in MiniBrowser using WebKit2 GTK+ API
4811         https://bugs.webkit.org/show_bug.cgi?id=71461
4812
4813         Reviewed by Martin Robinson.
4814
4815         * MiniBrowser/gtk/BrowserWindow.c:
4816         (resetEntryProgress): Reset the entry progress after a while when
4817         load has been completed.
4818         (webViewLoadProgressChanged): Update location entry progress.
4819         (browserWindowConstructed): Connect to WebView
4820         notify::estimated-load-progress signal.
4821
4822 2011-11-03  Carlos Garcia Campos  <cgarcia@igalia.com>
4823
4824         [GTK] Remove WebKit2 C API from MiniBrowser
4825         https://bugs.webkit.org/show_bug.cgi?id=71459
4826
4827         Reviewed by Martin Robinson.
4828
4829         Use the GTK+ API instead to add minimum functionality. The other
4830         features will be ported to GTK+ API in following patches.
4831
4832         * MiniBrowser/gtk/BrowserWindow.c:
4833         (activateUriEntryCallback): Use webkit_web_view_load_uri().
4834         (goBackCallback): Use webkit_web_view_go_back().
4835         (goForwardCallback): Use webkit_web_view_go_forward().
4836         (webViewURIChanged): Update location entry with current uri using
4837         webkit_web_view_get_uri().
4838         (browserWindowFinalize):
4839         (browserWindowGetProperty):
4840         (browserWindowSetProperty):
4841         (browser_window_init):
4842         (browserWindowConstructed): Connect to notify::uri signal of
4843         WebView to be notified when the URI changes.
4844         (browser_window_class_init):
4845         (browser_window_new): Use WebKitWebView.
4846         (browser_window_get_view): Use WebKitWebView
4847         * MiniBrowser/gtk/BrowserWindow.h:
4848         * MiniBrowser/gtk/GNUmakefile.am:
4849         * MiniBrowser/gtk/WebBundle/WebBundleMain.c: Removed.
4850         * MiniBrowser/gtk/main.c:
4851         (loadURI): Use webkit_web_view_new().
4852         (main):
4853
4854 2011-11-03  Simon Hausmann  <simon.hausmann@nokia.com>
4855
4856         [Qt] Unable to start MiniBrowser after run-webkit-tests
4857         https://bugs.webkit.org/show_bug.cgi?id=71469
4858
4859         Reviewed by Csaba Osztrogonác.
4860
4861         * Scripts/run-launcher: Pass @ARGV instead of @args, which is stripped
4862         from things like -2. Neither Gtk or Qt MiniBrowser like unknown command
4863         line arguments.
4864
4865 2011-11-03  Philippe Normand  <pnormand@igalia.com>
4866
4867         [GTK] [WK2] ttf-liberation fonts moved to a new location (in Debian)
4868         https://bugs.webkit.org/show_bug.cgi?id=71445
4869
4870         Reviewed by Martin Robinson.
4871
4872         * DumpRenderTree/gtk/DumpRenderTree.cpp:
4873         (initializeFonts): set directoriesDescription only when needed.
4874         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
4875         (WTR::inititializeFontConfigSetting): Add the new font path for Debian fonts and
4876         refactored the font files loading code to avoid copy/pastes, ease
4877         maintenance and future font directories additions.
4878
4879 2011-11-03  Carlos Garcia Campos  <cgarcia@igalia.com>
4880
4881         [GTK] Remove GtkLauncher2
4882         https://bugs.webkit.org/show_bug.cgi?id=71449
4883
4884         Reviewed by Martin Robinson.
4885
4886         GtkLauncher2 is the GtkLauncher code compiled with webkit2. This
4887         made sense when we wanted to implement the exactly same wk1 API
4888         for wk2. Now the code would require an ifdef for most the api, so
4889         it's better to remove GtkLauncher2 and port MiniBrowser to the new
4890         GTK+ API.
4891
4892         * GNUmakefile.am:
4893         * GtkLauncher/main.c:
4894         (notifyProgressCb):
4895         (closeWebViewCb):
4896         (createBrowser):
4897         (addWebSettingsGroupToContext):
4898         (main):
4899
4900 2011-11-02  Philippe Normand  <pnormand@igalia.com>
4901
4902         [GTK] [DRT] ttf-liberation fonts moved to a new location (in Debian)
4903         https://bugs.webkit.org/show_bug.cgi?id=71359
4904
4905         Reviewed by Martin Robinson.
4906
4907         * DumpRenderTree/gtk/DumpRenderTree.cpp:
4908         (initializeFonts): Add the new font path for Debian fonts and
4909         refactored the font files loading code to avoid copy/pastes, ease
4910         maintenance and future font directories additions.
4911
4912 2011-11-02  Eric Seidel  <eric@webkit.org>
4913
4914         Move Mocks into _mock files near their implementations
4915         https://bugs.webkit.org/show_bug.cgi?id=71425
4916
4917         Reviewed by Adam Barth.
4918
4919         This entire change is splitting mocktool.py into separate _mock.py files
4920         and placing them next to their real implementations.
4921
4922         I also deleted MockPortFactory (since it was wrong) and was just hiding
4923         the fact that the "skipped-ports" command was broken (and has been for a long time).
4924         So I made MockHost use a real PortFactory (passed a MockHost) and changed
4925         the skipped-ports implementation to use modern PortFactory methods.
4926
4927         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
4928         * Scripts/webkitpy/common/checkout/checkout_mock.py: Added.
4929         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
4930         * Scripts/webkitpy/common/checkout/deps_mock.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/__init__.py.
4931         * Scripts/webkitpy/common/checkout/scm/scm_mock.py: Added.
4932         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
4933         * Scripts/webkitpy/common/config/ports_mock.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/gtk_unittest.py.
4934         * Scripts/webkitpy/common/host.py:
4935         * Scripts/webkitpy/common/host_mock.py: Added.
4936         * Scripts/webkitpy/common/net/bugzilla/bugzilla_mock.py: Added.
4937         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
4938         * Scripts/webkitpy/common/net/buildbot/buildbot_mock.py: Added.
4939         * Scripts/webkitpy/common/net/credentials_unittest.py:
4940         * Scripts/webkitpy/common/net/failuremap_unittest.py:
4941         * Scripts/webkitpy/common/net/irc/irc_mock.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/__init__.py.
4942         * Scripts/webkitpy/common/net/statusserver_mock.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/gtk_unittest.py.
4943         * Scripts/webkitpy/common/net/statusserver_unittest.py:
4944         * Scripts/webkitpy/common/net/web_mock.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/gtk_unittest.py.
4945         * Scripts/webkitpy/common/system/executive.py:
4946         * Scripts/webkitpy/common/system/executive_mock.py:
4947         * Scripts/webkitpy/common/system/platforminfo_mock.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/__init__.py.
4948         * Scripts/webkitpy/common/system/user_mock.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/gtk_unittest.py.
4949         * Scripts/webkitpy/common/system/workspace_mock.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/__init__.py.
4950         * Scripts/webkitpy/common/system/workspace_unittest.py:
4951         * Scripts/webkitpy/common/watchlist/watchlist_mock.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/__init__.py.
4952         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
4953         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py:
4954         * Scripts/webkitpy/layout_tests/port/__init__.py:
4955         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
4956         * Scripts/webkitpy/layout_tests/port/efl_unittest.py:
4957         * Scripts/webkitpy/layout_tests/port/factory.py:
4958         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
4959         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
4960         * Scripts/webkitpy/layout_tests/port/http_lock_unittest.py:
4961         * Scripts/webkitpy/layout_tests/port/leakdetector_unittest.py:
4962         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
4963         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
4964         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
4965         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
4966         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
4967         * Scripts/webkitpy/test/main.py:
4968         * Scripts/webkitpy/tool/bot/botinfo_unittest.py:
4969         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
4970         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
4971         * Scripts/webkitpy/tool/commands/queries_unittest.py:
4972         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
4973         * Scripts/webkitpy/tool/mocktool.py:
4974         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
4975
4976 2011-11-02  Dan Bernstein  <mitz@apple.com>
4977
4978         Fixed the build and removed another copy to ColorBits.ttf.
4979
4980         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
4981         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
4982         * WebKitTestRunner/fonts/ColorBits.ttf: Removed.
4983
4984 2011-11-02  Dan Bernstein  <mitz@apple.com>
4985
4986         REGRESSION (r98759-r98773): platform/mac/fonts/color-bitmap.html failing on Lion Intel Debug (Tests)
4987         https://bugs.webkit.org/show_bug.cgi?id=71424
4988
4989         Reviewed by Sam Weinig.
4990
4991         Removed the ColorBits font and the code to register it in the process
4992         running the tests. The font was only used by the aforementioned test,
4993         which is now being removed.
4994
4995         * DumpRenderTree/fonts/ColorBits-A.png: Removed.
4996         * DumpRenderTree/fonts/ColorBits.ttf: Removed.
4997         * DumpRenderTree/mac/DumpRenderTree.mm:
4998         (allowedFontFamilySet):
4999         (activateTestingFonts):
5000         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp:
5001         (WTR::inititializeFontConfigSetting):
5002         * WebKitTestRunner/InjectedBundle/mac/ActivateFonts.mm:
5003         (WTR::allowedFontFamilySet):
5004         (WTR::activateFonts):
5005
5006 2011-11-02  Dan Bernstein  <mitz@apple.com>
5007
5008         Fix an intermittent hang in RenderedImageFromDOMRange.
5009
5010         Reviewed by Anders Carlsson.
5011
5012         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
5013         (TestWebKitAPI::Util::run): Made the call to -runMode:beforeDate: return
5014         immediately if there are no sources by passing a date in the distant past.
5015
5016 2011-11-02  Adam Roben  <aroben@apple.com>
5017
5018         Add git-add-reviewer
5019
5020         I've been using this script for a while to add reviewers to git commits in my local
5021         repository before pushing them to the Subversion repository. Basic usage is:
5022
5023         git-add-reviewer HEAD^ "Joe Schmoe"
5024         or
5025         git-add-reviewer -i trunk
5026
5027         Fixes <http://webkit.org/b/71419> Would like an easy way to add reviewer names to existing
5028         git commits
5029
5030         Reviewed by Sam Weinig.
5031
5032         * Scripts/git-add-reviewer: Added.
5033         (top level): Parse options, check that the working tree is clean, then run in either
5034         interactive or non-interactive mode as appropriate.
5035         (interactive): Write out a temp file that contains each commit with "NOBODY" next to it and
5036         open it in the user's editor. The user can replace "NOBODY" with reviewer names. When the
5037         editor exits, we check out the commit just before the earliest one we're modifying. Then
5038         for each commit, we check it out, update the reviewer, and amend the commit to contain the
5039         new reviewer. Finally, we update the branch that was originally checked out to point to the
5040         last amended commit.
5041         (nonInteractive): Check out the commit to modify, write the current commit message to the
5042         MERGE_MSG file so addReviewer will update it, add the reviewer to the ChangeLog(s), amend
5043         the commit, and rebase the original branch on top of the amended commit. (usage): Print out
5044         a usage message and exit.
5045         (requireCleanWorkTree): Do the same steps git-rebase does to ensure that the working tree
5046         and index are clean.
5047         (fail): Print out the specified error message, if any, and return false. This function is
5048         used as a return value in various error cases throughout the script.
5049         (cherryPick): Cherry-pick the specified commit and return 1 if we succeeded.
5050         (addReviewer): Add the specified reviewer to all the ChangeLogs modified in this commit, and
5051         to .git/MERGE_MSG, which will be used as the commit message for the next commit.
5052         (commit): Call git-commit and use .git/MERGE_MSG as the commit message.
5053
5054         (addReviewerToChangeLog):
5055         (addReviewerToCommitMessage):
5056         These just call through to addReviewerToFile.
5057
5058         (addReviewerToFile): Read in the file, replacing any "NOBODY" text with the reviewer name(s)
5059         and replacing "Reviewed" with "Rubber-stamped" if specified. Write the resulting text to a
5060         temp file, then move the temp file over the original file and stage it for the next commit.
5061         (head): Returns the name of the currently-checked-out branch.
5062         (isAncestor): Returns true if the first commit is an ancestor of the second.
5063         (toCommit): Converts a commitish to a commit ID using git-rev-parse.
5064         (changeLogsForCommit): Returns a list of all the ChangeLogs modified in the given commit.
5065         (resetToCommit): Checks out the given commit, throwing away any local changes.
5066         (writeCommitMessageToFile): Gets the commit message for the current commit and writes it to
5067         the specified file.
5068         (rebaseOntoHead): Rebases the specified branch onto HEAD.
5069         (checkout): Checks out the given commit.
5070         (getConfigValue): Reads the specified config variable from git-config.
5071
5072 2011-11-02  Dean Jackson  <dino@apple.com>
5073
5074         Add ENABLE_CSS_SHADERS flag
5075         https://bugs.webkit.org/show_bug.cgi?id=71394
5076
5077         Reviewed by Sam Weinig.
5078
5079         * Scripts/build-webkit:
5080
5081 2011-10-31  Eric Seidel  <eric@webkit.org>
5082
5083         BaselineOptimizer tests should use mocks instead of real Executive/FileSystem objects
5084         https://bugs.webkit.org/show_bug.cgi?id=71237
5085
5086         Reviewed by Adam Barth.
5087
5088         Calling the static version of factory.get() with proper mocking
5089         requires passsing an explict filesystem, executive, etc.
5090         So instead, we use a PortFactory instance and pass it a Host pointer.
5091         I had to add a MockHost since we'd not needed a non-host tool before now.
5092
5093         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
5094         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
5095         * Scripts/webkitpy/tool/commands/rebaseline.py:
5096         * Scripts/webkitpy/tool/mocktool.py:
5097
5098 2011-11-02  Anders Carlsson  <andersca@apple.com>
5099
5100         Update for the WebKit2 API fixage.
5101
5102         * MiniBrowser/mac/BrowserWindowController.m:
5103         (-[BrowserWindowController awakeFromNib]):
5104         * WebKitTestRunner/TestController.cpp:
5105         (WTR::TestController::initialize):
5106
5107 2011-11-02  Anders Carlsson  <andersca@apple.com>
5108
5109         Fix build.
5110
5111         Add new functions to the page loader client.
5112
5113         * MiniBrowser/mac/BrowserWindowController.m:
5114         (-[BrowserWindowController awakeFromNib]):
5115         * WebKitTestRunner/TestController.cpp:
5116         (WTR::TestController::initialize):
5117
5118 2011-11-02  Dan Bernstein  <mitz@apple.com>
5119
5120         <rdar://problem/10336700> Add API to get rendered text image without having to select it
5121         https://bugs.webkit.org/show_bug.cgi?id=71407
5122
5123         Reviewed by Simon Fraser.
5124
5125         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
5126         * TestWebKitAPI/Tests/mac/RenderedImageFromDOMRange.mm: Added.
5127         (-[RenderedImageFromDOMRangeFrameLoadDelegate webView:didFinishLoadForFrame:]):
5128         (TestWebKitAPI::TEST):
5129
5130 2011-11-02  Benjamin Poulain  <benjamin@webkit.org>
5131
5132         Make the main frame's base path more explicit in _drt_descriptionSuitableForTestResult
5133         https://bugs.webkit.org/show_bug.cgi?id=71351
5134
5135         Reviewed by Darin Adler.
5136
5137         Add the delimiter "/" at the end of the path instead of adding +1 when using
5138         the path.
5139
5140         * DumpRenderTree/mac/ResourceLoadDelegate.mm:
5141         (-[NSURL _drt_descriptionSuitableForTestResult]):
5142
5143 2011-11-02  Tom Sepez  <tsepez@chromium.org>
5144
5145         XSSAuditor is silent
5146         https://bugs.webkit.org/show_bug.cgi?id=70973
5147
5148         Reviewed by Adam Barth.
5149
5150         * DumpRenderTree/chromium/WebViewHost.cpp:
5151         (WebViewHost::didDetectXSS):
5152         * DumpRenderTree/chromium/WebViewHost.h:
5153         * DumpRenderTree/mac/FrameLoadDelegate.mm:
5154         (-[FrameLoadDelegate webView:didDetectXSS:]):
5155         * MiniBrowser/mac/BrowserWindowController.m:
5156         (didDetectXSSForFrame):
5157         (-[BrowserWindowController awakeFromNib]):
5158         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
5159         (WTR::InjectedBundlePage::InjectedBundlePage):
5160         (WTR::InjectedBundlePage::didDetectXSSForFrame):
5161         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
5162
5163 2011-11-02  Simon Hausmann  <simon.hausmann@nokia.com>
5164
5165         [Qt] REGRESSION: 5 editing tests fail
5166         https://bugs.webkit.org/show_bug.cgi?id=71117
5167
5168         Reviewed by Andreas Kling.
5169
5170         Make sure that the view is properly focused, in order to get the
5171         didBeginEditing DRT calls.
5172
5173         This requires activating the window (the call will result in a FocusIn
5174         event to the window as well as setting QGuiApplication::focusWindow())
5175         as well as enabling the focus on the view item. The signature of
5176         setFocus has changed to take a boolean.
5177
5178         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
5179         (WTR::WrapperWindow::handleStatusChanged):
5180
5181 2011-11-02  Matt Falkenhagen  <falken@chromium.org>
5182
5183         [chromium] Add per-script font preferences support in overridePreference
5184         https://bugs.webkit.org/show_bug.cgi?id=71110
5185
5186         This allows a layout test to do something like:
5187
5188         // Set standard font for Arabic to Ahem.
5189         overridePreference("WebKitStandardFontMap", [ "Arab", "Ahem" ]);
5190
5191         This should make writing a layout test for per-script fonts easier.
5192
5193         This change only affects the Chromium DumpRenderTree, but the other
5194         platforms should probably also have a similar change eventually.
5195
5196         Reviewed by Tony Chang.
5197
5198         * DumpRenderTree/chromium/LayoutTestController.cpp: add per-script
5199         font preferences to overridePreference
5200         (LayoutTestController::cppVariantToWebStringArray):
5201         (setFontMap):
5202         (LayoutTestController::overridePreference):
5203         * DumpRenderTree/chromium/LayoutTestController.h:
5204         * DumpRenderTree/chromium/WebPreferences.cpp:
5205         (setStandardFontFamilyWrapper):
5206         (setFixedFontFamilyWrapper):
5207         (setSerifFontFamilyWrapper):
5208         (setSansSerifFontFamilyWrapper):
5209         (setCursiveFontFamilyWrapper):
5210         (setFantasyFontFamilyWrapper):
5211         (applyFontMap):
5212         (WebPreferences::applyTo):
5213         * DumpRenderTree/chromium/WebPreferences.h: add per-script font
5214         preference members
5215         (WebPreferences::UScriptCodeHashTraits::emptyValue):
5216         (WebPreferences::UScriptCodeHashTraits::constructDeletedValue):
5217         (WebPreferences::UScriptCodeHashTraits::isDeletedValue):
5218
5219 2011-11-02  Elliot Poger  <epoger@google.com>
5220
5221         until use_skia=1 by default on mac, default port should be chromium-cg-mac
5222         https://bugs.webkit.org/show_bug.cgi?id=71312
5223
5224         Reviewed by Adam Barth.
5225
5226         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
5227         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
5228         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
5229
5230 2011-11-02  Stephen Chenney  <schenney@chromium.org>
5231
5232         printing layout test failures after r98634
5233         https://bugs.webkit.org/show_bug.cgi?id=71080
5234
5235         Reviewed by Kent Tamura
5236
5237         * DumpRenderTree/chromium/WebViewHost.cpp:
5238         (WebViewHost::paintPagesWithBoundaries): Added transforms invert the CG Bitmap so the coord system
5239         matches that for Skia.
5240
5241 2011-11-01  Sam Weinig  <sam@webkit.org>
5242
5243         Can't complete ChangeLog when using resolve-Changelogs and webkit-tools-completion.sh
5244         https://bugs.webkit.org/show_bug.cgi?id=71346
5245
5246         Reviewed by Dan Bernstein.
5247
5248         * Scripts/webkit-tools-completion.sh:
5249         Add "-o default" which allows for completing files in addition to the word list.
5250
5251 2011-11-02  Simon Hausmann  <simon.hausmann@nokia.com>
5252
5253         [Qt][WK2] Add support for touch event testing to WebKitTestRunner
5254         https://bugs.webkit.org/show_bug.cgi?id=71310
5255
5256         Reviewed by Kenneth Rohde Christiansen.
5257
5258         * WebKitTestRunner/DerivedSources.pro: Build system foo to ensure the ENABLE_TOUCH_EVENTS
5259         define works in EventSendingController.idl
5260         * WebKitTestRunner/EventSenderProxy.h: Add declarations for Qt implementations of touch
5261         handlers.
5262         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl: Add touch functions
5263         required by tests.
5264         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp: Forward touch messages to the
5265         ui process.
5266         (WTR::EventSendingController::addTouchPoint):
5267         (WTR::EventSendingController::updateTouchPoint):
5268         (WTR::EventSendingController::setTouchModifier):
5269         (WTR::EventSendingController::touchStart):
5270         (WTR::EventSendingController::touchMove):
5271         (WTR::EventSendingController::touchEnd):
5272         (WTR::EventSendingController::clearTouchPoints):
5273         (WTR::EventSendingController::releaseTouchPoint):
5274         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
5275         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro: Include features.pri for feature
5276         defines.
5277         * WebKitTestRunner/TestController.cpp:
5278         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle): Forward incoming
5279         touch messages to the EventSenderProxy.
5280         * WebKitTestRunner/qt/EventSenderProxyQt.cpp: Implement touch test harness functions.
5281         (WTR::EventSenderProxy::EventSenderProxy):
5282         (WTR::EventSenderProxy::addTouchPoint):
5283         (WTR::EventSenderProxy::updateTouchPoint):
5284         (WTR::EventSenderProxy::setTouchModifier):
5285         (WTR::EventSenderProxy::touchStart):
5286         (WTR::EventSenderProxy::touchMove):
5287         (WTR::EventSenderProxy::touchEnd):
5288         (WTR::EventSenderProxy::clearTouchPoints):
5289         (WTR::EventSenderProxy::releaseTouchPoint):
5290         (WTR::EventSenderProxy::sendTouchEvent):
5291         * WebKitTestRunner/qt/WebKitTestRunner.pro: Include features.pri for feature defines.
5292
5293 2011-11-01  Sam Weinig  <sam@webkit.org>
5294
5295         resolve-ChangeLogs --help should be faster.
5296
5297         Reviewed by Adam Roben.
5298
5299         * Scripts/resolve-ChangeLogs:
5300         (usageAndExit):
5301         Add a call to usageAndExit() before doing work to find unmerged changelogs
5302         if --help or something illegal is provided on the command line.
5303
5304 2011-11-01  Benjamin Poulain  <bpoulain@apple.com>
5305
5306         [Mac] _drt_descriptionSuitableForTestResult generate an exception for absolute paths above the main frame
5307         https://bugs.webkit.org/show_bug.cgi?id=71342
5308
5309         Reviewed by Darin Adler.
5310
5311         [NSURL _drt_descriptionSuitableForTestResult] was generating a NSRangeException due to the call to
5312         [NSString substringFromIndex:] with an index out of the string range.
5313
5314         The source of this path is the test-loading-archive-subresource-null-mimetype.html which load a subresource
5315         from the root of the disk.
5316         The problem was hidden because the exception are catched when calling the delegate.
5317
5318         This patch changes _drt_descriptionSuitableForTestResult to return absolute string if the resource is not
5319         in a subdirectory of the main frame.
5320
5321         * DumpRenderTree/mac/ResourceLoadDelegate.mm:
5322         (-[NSURL _drt_descriptionSuitableForTestResult]):
5323
5324 2011-11-01  Alok Priyadarshi  <alokp@chromium.org>
5325
5326         [chromium] --enable-hardware-gpu flag is not recognized by run-webkit-tests
5327         https://bugs.webkit.org/show_bug.cgi?id=71318
5328
5329         Reviewed by James Robinson.
5330
5331         Fixed a typo for --enable-hardware-gpu flag.
5332
5333         * Scripts/webkitpy/layout_tests/port/chromium.py:
5334
5335 2011-11-01  Elliot Poger  <epoger@google.com>
5336
5337         chromium-gpu port should default to chromium-gpu-mac, not chromium-gpu-cg-mac
5338         https://bugs.webkit.org/show_bug.cgi?id=71286
5339
5340         Reviewed by Stephen White.
5341
5342         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
5343
5344 2011-11-01  Simon Hausmann  <simon.hausmann@nokia.com>
5345
5346         [Qt][WK2] Fix build of WebKitTestRunner with ELF symbol visibility
5347         https://bugs.webkit.org/show_bug.cgi?id=71299
5348
5349         Reviewed by Andreas Kling.
5350
5351         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp: Use the properly exported symbol for the Qt
5352         build to enable WebCoreTestSupport::injectInternalsObject.
5353         (WTR::InjectedBundlePage::didClearWindowForFrame):
5354         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro: Fix include path to DRTSupportQt
5355
5356 2011-11-01  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
5357
5358         [Qt] Merge common desktop and touch view APIs under a base class.
5359         https://bugs.webkit.org/show_bug.cgi?id=71280
5360
5361         Reviewed by Andreas Kling.
5362
5363         Remove unneded forwarding logic for TouchWebView since those properties are
5364         now in the view just as the desktop web view.
5365
5366         * MiniBrowser/qt/qml/TouchView.qml:
5367
5368 2011-11-01  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
5369
5370         check-webkit-style: Allow names starting with "_q_".
5371         https://bugs.webkit.org/show_bug.cgi?id=70625
5372
5373         Reviewed by Tor Arne Vestbø.
5374
5375         Used by the Qt port as the standard prefix for private slots.
5376
5377         * Scripts/webkitpy/style/checkers/cpp.py:
5378         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
5379
5380 2011-11-01  Simon Hausmann  <simon.hausmann@nokia.com>
5381
5382         [Qt][WK2] Fix valgrind error about uninitialized variable
5383         https://bugs.webkit.org/show_bug.cgi?id=71273
5384
5385         Reviewed by Kenneth Christiansen.
5386
5387         * MiniBrowser/qt/MiniBrowserApplication.cpp:
5388         (MiniBrowserApplication::notify): Initialize isPrimary.
5389
5390 2011-10-31  Adam Roben  <aroben@apple.com>
5391
5392         Make Leaks Viewer less strict when parsing leaks-related output from NRWT/ORWT
5393
5394         Fixes <http://webkit.org/b/71258> REGRESSION (r98639): Leaks Viewer doesn't show recent
5395         leaky builds
5396
5397         Reviewed by Anders Carlsson.
5398
5399         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
5400         (RecentBuildsLoader.prototype.start): Relax a regex to gloss over differences in ORWT's vs.
5401         NRWT's leaks-related output (i.e., "M total leaks found!" vs. "M total leaks found for a
5402         total of N bytes!").
5403
5404 2011-10-31  Raphael Kubo da Costa  <kubo@profusion.mobi>
5405
5406         [EFL] Rename ewk_view_setting_scripts_window_open_{get,set} after r93833
5407         https://bugs.webkit.org/show_bug.cgi?id=71228
5408
5409         Reviewed by Antonio Gomes.
5410
5411         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
5412         (DumpRenderTreeChrome::createNewWindow):
5413         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
5414
5415 2011-10-31  Eric Seidel  <eric@webkit.org>
5416
5417         webkitpy tests depend too much on the user's environment
5418         https://bugs.webkit.org/show_bug.cgi?id=71234
5419
5420         Reviewed by Dirk Pranke.
5421
5422         This change just makes a bunch of our older tests use
5423         more modern mocking to avoid trying to launch processes
5424         or read from the user's filesystem during unittesting.
5425
5426         I found many of these by adding an assert in Executive.run_command
5427         that we were not unittesting.  I can't add that assert always
5428         as there are some valid uses of Executive during unittesting.
5429         Once I fix more of these, I may find a way to add such an assert conditionally.
5430
5431         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
5432         * Scripts/webkitpy/common/net/credentials_unittest.py:
5433         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
5434         * Scripts/webkitpy/layout_tests/port/base.py:
5435         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
5436
5437 2011-10-31  Kenneth Rohde Christiansen  <kenneth@webkit.org>
5438
5439         [Qt] MiniBrowser doesn't resize as the size is always overridden
5440         https://bugs.webkit.org/show_bug.cgi?id=71212
5441
5442         Reviewed by Simon Hausmann.
5443
5444         Remove call overriden the size.
5445
5446         * MiniBrowser/qt/main.cpp:
5447         (main):
5448
5449 2011-10-30  Sheriff Bot  <webkit.review.bot@gmail.com>
5450
5451         Unreviewed, rolling out r98819, r98823, r98825, and r98830.
5452         http://trac.webkit.org/changeset/98819
5453         http://trac.webkit.org/changeset/98823
5454         http://trac.webkit.org/changeset/98825
5455         http://trac.webkit.org/changeset/98830
5456         https://bugs.webkit.org/show_bug.cgi?id=71199
5457
5458         Broke Chromium Windows bot (Requested by rniwa on #webkit).
5459
5460         * Scripts/webkitpy/layout_tests/port/base.py:
5461         * Scripts/webkitpy/layout_tests/port/chromium.py:
5462         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
5463         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
5464         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
5465         * Scripts/webkitpy/layout_tests/port/efl.py:
5466         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
5467         * Scripts/webkitpy/layout_tests/port/mac.py:
5468         * Scripts/webkitpy/layout_tests/port/qt.py:
5469         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
5470         * Scripts/webkitpy/layout_tests/port/webkit.py:
5471         * Scripts/webkitpy/layout_tests/port/win.py:
5472
5473 2011-10-30  Eric Seidel  <eric@webkit.org>
5474
5475         new-run-webkit-tests is locale dependent
5476         https://bugs.webkit.org/show_bug.cgi?id=68691
5477
5478         Unreviewed.
5479
5480         This is a temporary hack until someone from the cr-win port
5481         can take a look and determine which environment variable
5482         lighttpd is missing.
5483
5484         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
5485
5486 2011-10-30  Eric Seidel  <eric@webkit.org>
5487
5488         new-run-webkit-tests is locale dependent
5489         https://bugs.webkit.org/show_bug.cgi?id=68691
5490
5491         Unreviewed.  I would have preferred to have this reviewed,
5492         but relevant reviewers are asleep and bots are broken.
5493
5494         This was a regression from moving to a clean environment.
5495         ChromiumWin (and possibly other ports), need the "PATH"
5496         environment copied over.  This wasn't caught in my testing
5497         because although we had unittests to cover this, they
5498         weren't being run on anything but windows.  The vast majority
5499         of this change is just fixing the unittests to use our
5500         modern MockFileSystem/MockUser/MockExecutive so they can
5501         be run on any system (and removing the platform checks from
5502         the unittests so they are run everywhere).
5503
5504         The actual fix is the single line "PATH" string added to base.py.
5505         The rest of this change is just fixing the Chromium port unittests
5506         to run on all systems (including changing the Chromium port to
5507         use FileSystem.path_to_module instead of __file__).
5508
5509         * Scripts/webkitpy/layout_tests/port/base.py:
5510         * Scripts/webkitpy/layout_tests/port/chromium.py:
5511         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
5512         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
5513         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
5514         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
5515
5516 2011-10-29  Eric Seidel  <eric@webkit.org>
5517
5518         new-run-webkit-tests is locale dependent
5519         https://bugs.webkit.org/show_bug.cgi?id=68691
5520
5521         Unreviewed.
5522
5523         Fix the Qt port and unittest the function so we don't break it again.
5524
5525         * Scripts/webkitpy/layout_tests/port/qt.py:
5526         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
5527
5528 2011-10-29  Eric Seidel  <eric@webkit.org>
5529
5530         new-run-webkit-tests is locale dependent
5531         https://bugs.webkit.org/show_bug.cgi?id=68691
5532
5533         Reviewed by Adam Barth.
5534
5535         Make NRWT use a clean environment just like ORWT did.
5536         I've manually validated that NRWT now copies over
5537         the same environment variables that ORWT did.
5538         This new code is exercised by existing setup_environ_for_server tests.
5539
5540         * Scripts/webkitpy/layout_tests/port/base.py:
5541         * Scripts/webkitpy/layout_tests/port/efl.py:
5542         * Scripts/webkitpy/layout_tests/port/mac.py:
5543         * Scripts/webkitpy/layout_tests/port/qt.py:
5544         * Scripts/webkitpy/layout_tests/port/webkit.py:
5545         * Scripts/webkitpy/layout_tests/port/win.py:
5546
5547 2011-10-29  Eric Seidel  <eric@webkit.org>
5548
5549         new-run-webkit-tests: MASTER BUG: Switch all webkit.org bots over
5550         https://bugs.webkit.org/show_bug.cgi?id=34984
5551
5552         Reviewed by Adam Barth.
5553
5554         Move run-webkit-tests from using a white-list of supported NRWT ports
5555         to a blacklist of ports requiring ORWT.
5556
5557         I've also added the ability to opt-in to parallel testing on a per-port basis.
5558
5559         With these changes I'm going to officially declare this bug dead and
5560         work on remaining polish issues and individual developer complaints.
5561
5562         * Scripts/run-webkit-tests:
5563         (useNewRunWebKitTests):
5564         (platformIsReadyForParallelTesting):
5565
5566 2011-10-29  Sheriff Bot  <webkit.review.bot@gmail.com>
5567
5568         Unreviewed, rolling out r98780.
5569         http://trac.webkit.org/changeset/98780
5570         https://bugs.webkit.org/show_bug.cgi?id=71173
5571
5572         Broke Qt bots (Requested by rniwa on #webkit).
5573
5574         * Scripts/run-webkit-tests:
5575         (useNewRunWebKitTests):
5576
5577 2011-10-28  Rafael Weinstein  <rafaelw@chromium.org>
5578
5579         webkit.py gdb visualizer is broken after change to StringImpl
5580         https://bugs.webkit.org/show_bug.cgi?id=71154
5581
5582         Reviewed by Tony Chang.
5583
5584         r98624 changed StringImpl m_data to m_data32. This changes updates
5585         webkit.py to match.
5586
5587         * gdb/webkit.py:
5588
5589 2011-10-28  Joseph Pecoraro  <pecoraro@apple.com>
5590
5591         Fix typo in check-for-inappropriate-objc-class-names
5592         https://bugs.webkit.org/show_bug.cgi?id=71144
5593
5594         Reviewed by Dan Bernstein.
5595
5596         * Scripts/check-for-inappropriate-objc-class-names:
5597
5598 2011-10-28  Sadrul Habib Chowdhury  <sadrul@chromium.org>
5599
5600         Add support for sending scroll-update events from EventSender.
5601         https://bugs.webkit.org/show_bug.cgi?id=66272
5602
5603         Reviewed by Adam Barth.
5604
5605         * DumpRenderTree/chromium/EventSender.cpp:
5606         (EventSender::EventSender):
5607         (EventSender::reset):
5608         (EventSender::gestureScrollUpdate):
5609         (EventSender::gestureEvent):
5610         * DumpRenderTree/chromium/EventSender.h:
5611
5612 2011-10-28  Eric Seidel  <eric@webkit.org>
5613
5614         new-run-webkit-tests: MASTER BUG: Switch all webkit.org bots over
5615         https://bugs.webkit.org/show_bug.cgi?id=34984
5616
5617         Reviewed by Adam Barth.
5618
5619         Move run-webkit-tests from using a white-list of supported NRWT ports
5620         to a blacklist of ports requiring ORWT.
5621
5622         I've also added the ability to opt-in to parallel testing on a per-port basis.
5623
5624         With these changes I'm going to officially declare this bug dead and
5625         work on remaining polish issues and individual developer complaints.
5626
5627         * Scripts/run-webkit-tests:
5628         (useNewRunWebKitTests):
5629         (platformIsReadyForParallelTesting):
5630
5631 2011-10-28  Eric Seidel  <eric@webkit.org>
5632
5633         NRWT crashes when parsing leaks output due to UTF-8 decoding error
5634         https://bugs.webkit.org/show_bug.cgi?id=71112
5635
5636         Reviewed by Adam Barth.
5637
5638         Use read_binary_file instead of read_text_file and add a test
5639         case with some non-utf8 data to make sure we don't break this in the future.
5640
5641         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
5642         * Scripts/webkitpy/layout_tests/port/leakdetector_unittest.py:
5643
5644 2011-10-28  David Levin  <levin@chromium.org>
5645
5646         watchlist should not CC the bug reporter
5647         https://bugs.webkit.org/show_bug.cgi?id=71079
5648
5649         Reviewed by Adam Barth.
5650
5651         Tested by test_apply_watch_list_local (webkitpy.tool.steps.applywatchlist_unittest.ApplyWatchListTest)
5652         which remains upchanged and thus verifies that the reporter doesn't get added to the cc line.
5653
5654         * Scripts/webkitpy/tool/commands/applywatchlistlocal_unittest.py: Add the new email in the output.
5655         * Scripts/webkitpy/tool/mocktool.py: Change the reporter and make them one of the people to cc.
5656         * Scripts/webkitpy/tool/steps/applywatchlist.py: Remove the reporter from the cc list.
5657
5658 2011-10-28  Zoltan Horvath  <zoltan@webkit.org>
5659
5660         Add watchlist for QtWebKit2 API and for WK2's Qt specific files.
5661         https://bugs.webkit.org/show_bug.cgi?id=71116
5662
5663         Reviewed by David Levin.
5664
5665         * Scripts/webkitpy/common/config/watchlist:
5666
5667 2011-10-28  Simon Hausmann  <simon.hausmann@nokia.com>
5668
5669         [Qt] Adapt to QSG* to QQuick* API change in qtdeclarative
5670         https://bugs.webkit.org/show_bug.cgi?id=70494
5671
5672         Reviewed by Kenneth Christiansen.
5673
5674         Use QQuick* instead of QSG* as well as QStandardPaths
5675         instead of QDesktopLocation.
5676
5677         * MiniBrowser/qt/BrowserWindow.cpp:
5678         (BrowserWindow::BrowserWindow):
5679         * MiniBrowser/qt/BrowserWindow.h:
5680         * QtTestBrowser/cookiejar.cpp:
5681         (TestBrowserCookieJar::TestBrowserCookieJar):
5682         * QtTestBrowser/launcherwindow.cpp:
5683         (LauncherWindow::setDiskCache):
5684         * WebKitTestRunner/PlatformWebView.h:
5685         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
5686         (WTR::WrapperWindow::WrapperWindow):
5687         (WTR::WrapperWindow::handleStatusChanged):
5688
5689 2011-10-28  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
5690
5691         [Qt] Fix the Ctrl behavior for touch mocking in MiniBrowser.
5692         https://bugs.webkit.org/show_bug.cgi?id=71106
5693
5694         Reviewed by Simon Hausmann.
5695
5696         Holding Ctrl allows multiple touch points to be held on the screen using
5697         mouse buttons. It would previously only prevent TouchReleased to be sent
5698         and would require another MouseButtonRelease to be sent without holding Ctrl.
5699
5700         This patch makes sure that all held touch points are released when Ctrl
5701         is released if MouseButtonRelease was received.
5702         It also removes the touch QEvent::Type logic since it's currently handled
5703         by QtGui by observing the modified touch points.
5704
5705         * MiniBrowser/qt/MiniBrowserApplication.cpp:
5706         (MiniBrowserApplication::notify):
5707         (MiniBrowserApplication::sendTouchEvent):
5708         * MiniBrowser/qt/MiniBrowserApplication.h:
5709
5710 2011-10-28  Kenneth Rohde Christiansen  <kenneth@webkit.org>
5711
5712         [Qt] MiniBrowser needs a -window-size option
5713         https://bugs.webkit.org/show_bug.cgi?id=70999
5714
5715         Reviewed by Simon Hausmann.
5716
5717         Add an -window-size option useful for testing the viewport handling.
5718         Also clean up the option handling.
5719
5720         * MiniBrowser/qt/BrowserWindow.cpp:
5721         (BrowserWindow::BrowserWindow):
5722         * MiniBrowser/qt/MiniBrowserApplication.cpp:
5723         (MiniBrowserApplication::handleUserOptions):
5724         * MiniBrowser/qt/MiniBrowserApplication.h:
5725         (WindowOptions::WindowOptions):
5726         (WindowOptions::setRequestedWindowSize):
5727         (WindowOptions::requestedWindowSize):
5728         * MiniBrowser/qt/utils.cpp:
5729         (takeOptionValue):
5730         * MiniBrowser/qt/utils.h:
5731
5732 2011-10-28  Jochen Eisinger  <jochen@chromium.org>
5733
5734         Add allowScriptFromSource callback to FrameLoaderClient
5735         https://bugs.webkit.org/show_bug.cgi?id=71013
5736
5737         Reviewed by Darin Fisher.
5738
5739         * DumpRenderTree/chromium/LayoutTestController.cpp:
5740         (LayoutTestController::LayoutTestController):
5741         (LayoutTestController::setScriptsAllowed):
5742         * DumpRenderTree/chromium/LayoutTestController.h:
5743         * DumpRenderTree/chromium/WebPermissions.cpp:
5744         (WebPermissions::allowScriptFromSource):
5745         (WebPermissions::setScriptsAllowed):
5746         (WebPermissions::reset):
5747         * DumpRenderTree/chromium/WebPermissions.h:
5748
5749 2011-10-28  Alexandru Chiculita  <achicu@adobe.com>
5750
5751         Added myself as a committer.
5752
5753         * Scripts/webkitpy/common/config/committers.py:
5754
5755 2011-10-27  Adam Barth  <abarth@webkit.org>
5756
5757         Remove WinCairo from the set of core builders.  WinCairo rarely builds
5758         and is never green.
5759
5760         Rubber-stamped by Eric Seidel.
5761
5762         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
5763         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
5764
5765 2011-10-27  Yuta Kitamura  <yutak@chromium.org>
5766
5767         WebSocket: Connecting to localhost:8880 takes one second on Windows
5768         https://bugs.webkit.org/show_bug.cgi?id=64788
5769
5770         Reviewed by Dirk Pranke.
5771
5772         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
5773         Bind to "localhost" instead of "127.0.0.1" to let pywebsocket listen on both
5774         IPv4 and IPv6 addresses. This should prevent the test reserved-opcodes.html
5775         from timing out on Windows, because this test tries to open a lot of
5776         connections to localhost and each attempt takes one second to fall back from
5777         IPv6 to IPv4 on Windows (I have no idea why Windows works like this, though).
5778
5779 2011-10-27  Adam Barth  <abarth@webkit.org>
5780
5781         Dis-integrate run-bindings-tests with webkitpy
5782         https://bugs.webkit.org/show_bug.cgi?id=71092
5783
5784         Reviewed by Ryosuke Niwa.
5785
5786         These tests aren't really adding any value.
5787
5788         * Scripts/webkitpy/common/config/ports.py:
5789         * Scripts/webkitpy/tool/commands/download_unittest.py:
5790         * Scripts/webkitpy/tool/mocktool.py:
5791         * Scripts/webkitpy/tool/steps/runtests.py:
5792         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
5793         * Scripts/webkitpy/tool/steps/steps_unittest.py:
5794
5795 2011-10-27  Eric Seidel  <eric@webkit.org>
5796
5797         REGRESSION (r98639): NRWT crashes when parsing leaks output
5798         https://bugs.webkit.org/show_bug.cgi?id=71087
5799
5800         Unreviewed.  Sadly we have no good way to test logging.
5801
5802         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
5803
5804 2011-10-27  David Levin  <levin@chromium.org>
5805
5806         Test expectation errors should include the file name and platform in both the log and exception info.
5807         https://bugs.webkit.org/show_bug.cgi?id=71067
5808
5809         Reviewed by Ojan Vafai.
5810
5811         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
5812         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
5813
5814 2011-10-27  Eric Seidel  <eric@webkit.org>
5815
5816         new-run-webkit-tests doesn't understand --qt or --gtk
5817         https://bugs.webkit.org/show_bug.cgi?id=63970
5818
5819         Reviewed by Ojan Vafai.
5820
5821         Add support for --qt and --gtk to NRWT.
5822         I also removed some incorrect comments in the option parsing
5823         code (options which we already implement, but said we didn't)
5824         as well as moved --platform parsing out of the "results_options"
5825         section into the top configuration section.
5826
5827         Now run-webkit-tests no longer needs NRWT-specific code for passing
5828         --platform=qt, etc.
5829
5830         * Scripts/run-webkit-tests:
5831         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
5832
5833 2011-10-27  Eric Seidel  <eric@webkit.org>
5834
5835         REGRESSION (NRWT): build.webkit.org doesn't show the total number of leaks found during a test run on the Leaks bot
5836         https://bugs.webkit.org/show_bug.cgi?id=66227
5837
5838         Reviewed by Adam Roben.
5839
5840         I believe this should fix the bug.
5841
5842         * Scripts/run-webkit-tests: make NRWT default for --leaks
5843         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
5844         (LeakDetector._parse_leaks_output): removed the (unneeded) process_pid argument, and made the regexp use named groups (even though we don't ever grab them by name)
5845         (LeakDetector.count_total_bytes_and_unique_leaks): renamed from parse_leak_files
5846         (LeakDetector.count_total_leaks): new file (the guts of this change) which is used to re-parse the leaks output during the summarize leaks phase.
5847         * Scripts/webkitpy/layout_tests/port/leakdetector_unittest.py:
5848         Changes to reflect the rename of count_total_bytes_and_unique_leaks and a new test for count_total_leaks.
5849         * Scripts/webkitpy/layout_tests/port/mac.py:
5850         Use count_total_leaks to spit out the total leak count like ORWT did, and remove the FIXME on the subject.
5851
5852 2011-10-27  Stephen Chenney  <schenney@chromium.org>
5853
5854         [Chromium] Need setPrinting
5855         https://bugs.webkit.org/show_bug.cgi?id=46152
5856
5857         Reviewed by Hajime Morita.
5858
5859         * DumpRenderTree/chromium/LayoutTestController.cpp:
5860         (LayoutTestController::LayoutTestController): Added logic for handling setPrinting.
5861         (LayoutTestController::reset): Clearing flags for setPrinting.
5862         (LayoutTestController::setPrinting): Set flags for setPrinting.
5863         * DumpRenderTree/chromium/LayoutTestController.h:
5864         (LayoutTestController::setIsPrinting): Added the setPrinting methods and flags.
5865         (LayoutTestController::isPrinting): Method to return the setPrinting status.
5866         * DumpRenderTree/chromium/TestShell.cpp:
5867         (dumpFramesAsPrintedText): A method for creating a text dump in printed layout.
5868         (TestShell::dump): Modified to dump setPrinting-specific output.
5869         * DumpRenderTree/chromium/WebViewHost.cpp:
5870         (WebViewHost::paintPagesWithBoundaries): Handling for pixel tests when setPrinting is
5871         active.
5872         * DumpRenderTree/chromium/WebViewHost.h: Prototype for paintPagesWithBoundaries.
5873
5874 2011-10-27  Sam Weinig  <sam@webkit.org>
5875
5876         Add allowsPlugIns property to WKBrowsingContextGroup
5877         https://bugs.webkit.org/show_bug.cgi?id=70987
5878
5879         Reviewed by Anders Carlsson.
5880
5881         * TestWebKitAPI/Tests/WebKit2ObjC/WKBrowsingContextGroupTest.mm:
5882         Add basic test for WKBrowsingContextGroup.allowsPlugIns.
5883
5884 2011-10-27  Adam Roben  <aroben@apple.com>
5885
5886         Test WKBundlePageGetBackingScaleFactor
5887
5888         Test for <http://webkit.org/b/71025> <rdar://problem/10355037> REGRESSION (r97191): Clients
5889         can't reliably determine the scale factor of snapshots returned via
5890         WKBundlePageCreateSnapshot* API
5891
5892         Reviewed by Sam Weinig.
5893
5894         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new files.
5895
5896         * TestWebKitAPI/Tests/WebKit2/mac/GetBackingScaleFactor.mm: Added.
5897         (TestWebKitAPI::didReceiveMessageFromInjectedBundle): Record the passed-back scale factor.
5898         (TestWebKitAPI::setInjectedBundleClient): Hook up our didReceiveMessage callback.
5899         (TestWebKitAPI::createWindow): Conjure up a window.
5900         (TestWebKitAPI::TEST): Ask the injected bundle code to call the new
5901         WKBundlePageGetBackingScaleFactor API when the WKView is in windows with various scale
5902         factors and when using a custom scale factor, and check that we get those same scale factors
5903         back.
5904
5905         * TestWebKitAPI/Tests/WebKit2/mac/GetBackingScaleFactor_Bundle.mm: Added.
5906         (TestWebKitAPI::GetBackingScaleFactorTest::GetBackingScaleFactorTest): Just initialize our
5907         members.
5908         (TestWebKitAPI::GetBackingScaleFactorTest::didCreatePage): Store the page for later.
5909         (TestWebKitAPI::GetBackingScaleFactorTest::didReceiveMessage): Get the backing scale factor
5910         from the page using the new API and send it back.
5911
5912 2011-10-27  Balazs Kelemen  <kbalazs@webkit.org>
5913
5914         [Qt][WK2] MiniBrowser crashes when started with 0 arguments
5915
5916         Rubber-stamped by Simon Hausmann.
5917
5918         * MiniBrowser/qt/MiniBrowserApplication.cpp:
5919         (MiniBrowserApplication::handleUserOptions):
5920         Fix typo error.
5921
5922 2011-10-27  Takashi Toyoshima  <toyoshim@chromium.org>
5923
5924         [WebSocket] update pywebsocket to 0.7.
5925         https://bugs.webkit.org/show_bug.cgi?id=70992
5926
5927         Reviewed by Kent Tamura.
5928
5929         pywebsocket 0.7 fixed the bug on dual stack support on OS X.
5930
5931         * Scripts/webkitpy/thirdparty/__init__.py:
5932
5933 2011-10-27  Elliot Poger  <epoger@google.com>
5934
5935         make lion observe MAC lines in test_expectations.txt
5936         https://bugs.webkit.org/show_bug.cgi?id=70941
5937
5938         Reviewed by Ojan Vafai.
5939
5940         * Scripts/webkitpy/layout_tests/port/chromium.py:
5941         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
5942
5943 2011-10-27  John Knottenbelt  <jknotten@chromium.org>
5944
5945         [Chromium] Allow DRT to be built without notifications enabled.
5946         https://bugs.webkit.org/show_bug.cgi?id=70919
5947
5948         Reviewed by Tony Gentilcore.
5949
5950         * DumpRenderTree/chromium/LayoutTestController.cpp:
5951         (LayoutTestController::grantDesktopNotificationPermission):
5952         (LayoutTestController::simulateDesktopNotificationClick):
5953         * DumpRenderTree/chromium/NotificationPresenter.cpp:
5954         * DumpRenderTree/chromium/TestShell.cpp:
5955         (TestShell::TestShell):
5956         (TestShell::resetTestController):
5957
5958 2011-10-21  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
5959
5960         [Qt] Improve command line arguments handling for MiniBrowser.
5961         https://bugs.webkit.org/show_bug.cgi?id=70616
5962
5963         Reviewed by Kenneth Rohde Christiansen.
5964
5965         - Use two dashes for word arguments.
5966         - Return with an error when an unsupported argument is given.
5967         - Allow the URL to be placed before option arguments.
5968
5969         * MiniBrowser/qt/MiniBrowserApplication.cpp:
5970         (printHelp):
5971         (MiniBrowserApplication::handleUserOptions):
5972         * MiniBrowser/qt/utils.cpp:
5973         (takeOptionFlag):
5974         (takeOptionValue):
5975         * MiniBrowser/qt/utils.h:
5976
5977 2011-10-27  Csaba Osztrogonác  <ossy@webkit.org>
5978
5979         [Qt] NRWT doesn't work on qt-mac platform
5980         https://bugs.webkit.org/show_bug.cgi?id=64445
5981
5982         Reviewed by Adam Barth.
5983
5984         * Scripts/run-webkit-tests: Switch qt-mac platform to NRWT.
5985         (useNewRunWebKitTests):
5986         * Scripts/webkitpy/layout_tests/port/qt.py: Fix library path on Mac.
5987
5988 2011-10-26  Sam Weinig  <sam@webkit.org>
5989
5990         Add initial setting to WKBrowsingContextGroup
5991         https://bugs.webkit.org/show_bug.cgi?id=70942
5992
5993         Reviewed by Anders Carlsson.
5994
5995         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
5996         * TestWebKitAPI/Tests/WebKit2ObjC/WKBrowsingContextGroupTest.mm: Added.
5997         Add basic test that setting the javaScriptEnabled setting takes. It is not
5998         possible to test its effect quite yet.
5999
6000 2011-10-26  Alejandro G. Castro  <alex@igalia.com>
6001
6002         [GTK] [WK2] Add WebKit2 distcheck support
6003         https://bugs.webkit.org/show_bug.cgi?id=70933
6004
6005         Reviewed by Martin Robinson.
6006
6007         * WebKitTestRunner/GNUmakefile.am: Add missing files to the
6008         sources and EXTRA_DIST.
6009
6010 2011-10-26  Simon Hausmann  <simon.hausmann@nokia.com>
6011
6012         [Qt][WK2] Fix failing assertion on startup of MiniBrowser -touch.
6013
6014         Reviewed by Tor Arne Vestbø.
6015
6016         We're using QScroller since r98460, which requires QWidget/QApplication. So
6017         use QApplication and QtWidgets for the time being.
6018
6019         * MiniBrowser/qt/MiniBrowser.pro:
6020         * MiniBrowser/qt/MiniBrowserApplication.cpp:
6021         (MiniBrowserApplication::MiniBrowserApplication):
6022         (MiniBrowserApplication::notify):
6023         * MiniBrowser/qt/MiniBrowserApplication.h:
6024
6025 2011-10-25  Zeno Albisser  <zeno.albisser@nokia.com>
6026
6027         [Qt][WK2] Rewrite MiniBrowser in QML
6028         https://bugs.webkit.org/show_bug.cgi?id=70315
6029
6030         Add the first QML based implementation of MiniBrowser.
6031
6032         Patch by Jocelyn Turcotte <jocelyn.turcotte@nokia.com>,
6033                  Zeno Albisser <zeno.albisser@nokia.com>
6034
6035         Reviewed by Tor Arne Vestbø.
6036
6037         * MiniBrowser/qt/BrowserWindow.cpp:
6038         (BrowserWindow::BrowserWindow):
6039         (BrowserWindow::webView):
6040         (BrowserWindow::load):
6041         (BrowserWindow::keyPressEvent):
6042         (BrowserWindow::~BrowserWindow):
6043         * MiniBrowser/qt/BrowserWindow.h:
6044         * MiniBrowser/qt/MiniBrowser.pro:
6045         * MiniBrowser/qt/MiniBrowser.qrc:
6046         * MiniBrowser/qt/MiniBrowserApplication.cpp:
6047         (MiniBrowserApplication::MiniBrowserApplication):
6048         (MiniBrowserApplication::handleUserOptions):
6049         * MiniBrowser/qt/MiniBrowserApplication.h:
6050         (WindowOptions::WindowOptions):
6051         * MiniBrowser/qt/UrlLoader.cpp:
6052         (UrlLoader::UrlLoader):
6053         * MiniBrowser/qt/icons/next.png: Added.
6054         * MiniBrowser/qt/icons/previous.png: Added.
6055         * MiniBrowser/qt/icons/refresh.png: Added.
6056         * MiniBrowser/qt/icons/stop.png: Added.
6057         * MiniBrowser/qt/main.cpp:
6058         (main):
6059         * MiniBrowser/qt/qml/BrowserWindow.qml: Added.
6060
6061 2011-10-26  Zeno Albisser  <zeno.albisser@nokia.com>
6062
6063         [Qt][WK2] Remove QtWidgets dependency in MiniBrowser
6064         https://bugs.webkit.org/show_bug.cgi?id=70529
6065
6066         Remove obsolete code for a new QML based implementation
6067         of MiniBrowser.
6068         Change BrowserWindow to derive from QSGView instead of
6069         QMainWindow. QSGView is used as a top level window that
6070         provides the container for placing QML content.
6071         Make MiniBrowserApplication derive from QGuiApplication
6072         instead of QApplication.
6073
6074         Patch by Jocelyn Turcotte <jocelyn.turcotte@nokia.com>,
6075                  Zeno Albisser <zeno.albisser@nokia.com>
6076
6077         Reviewed by Simon Hausmann.
6078
6079         * MiniBrowser/qt/BrowserView.cpp: Removed.
6080         * MiniBrowser/qt/BrowserView.h: Removed.
6081         * MiniBrowser/qt/BrowserWindow.cpp:
6082         (BrowserWindow::BrowserWindow):
6083         (BrowserWindow::load):
6084         (BrowserWindow::screenshot):
6085         (BrowserWindow::loadURLListFromFile):
6086         (BrowserWindow::updateUserAgentList):
6087         (BrowserWindow::~BrowserWindow):
6088         * MiniBrowser/qt/BrowserWindow.h:
6089         * MiniBrowser/qt/MiniBrowser.pro:
6090         * MiniBrowser/qt/MiniBrowserApplication.cpp:
6091         (MiniBrowserApplication::MiniBrowserApplication):
6092         (MiniBrowserApplication::notify):
6093         * MiniBrowser/qt/MiniBrowserApplication.h:
6094         * MiniBrowser/qt/UrlLoader.cpp:
6095         (UrlLoader::UrlLoader):
6096         * MiniBrowser/qt/main.cpp:
6097
6098 2011-10-25  Zeno Albisser  <zeno.albisser@nokia.com>
6099
6100         [Qt][WK2] Remove QAction from MiniBrowser
6101         https://bugs.webkit.org/show_bug.cgi?id=70525
6102
6103         In order to rewrite MiniBrowser in QML we need a suitable
6104         mechanism for triggering navigation actions.
6105         Due to the QtWidgets dependency of QAction the current
6106         solution is not suitable anymore. Therefore we introduce
6107         invokable methods and export properties in QWebNavigationController.
6108
6109         Patch by Jocelyn Turcotte <jocelyn.turcotte@nokia.com>,
6110                  Zeno Albisser <zeno.albisser@nokia.com>
6111
6112         Reviewed by Simon Hausmann.
6113
6114         * MiniBrowser/qt/BrowserView.cpp:
6115         * MiniBrowser/qt/BrowserView.h:
6116         * MiniBrowser/qt/BrowserWindow.cpp:
6117         (BrowserWindow::BrowserWindow):
6118
6119 2011-10-26  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
6120
6121         [EFL] Unregister viewport,changed signal in EWebLauncher.
6122         https://bugs.webkit.org/show_bug.cgi?id=70478
6123
6124         Reviewed by Eric Seidel.
6125
6126         'viewport,changed' signal was registered by EWebLauncher in order to test viewport meta tag.
6127         However, it sets wrong layout to some web sites which have viewport meta tag. Because, host pc
6128         doesn't support proper DPI value though viewport is using DPI value to compute layout values.
6129         This signal is for mobile device.
6130
6131         * EWebLauncher/main.c:
6132         (browserCreate):
6133
6134 2011-10-25  Anders Carlsson  <andersca@apple.com>
6135
6136         Plug-ins have no way to find out when the device scale factor changes
6137         https://bugs.webkit.org/show_bug.cgi?id=67226
6138         <rdar://problem/10048276>
6139
6140         Reviewed by Sam Weinig.
6141
6142         Keep a cached copy of the contents scale and update it when it changes.
6143
6144         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ContentsScaleFactor.cpp:
6145         (ContentsScaleFactor::ContentsScaleFactor):
6146         (ContentsScaleFactor::cachedContentsScaleFactor):
6147         (ContentsScaleFactor::ScriptableObject::hasProperty):
6148         (ContentsScaleFactor::ScriptableObject::getProperty):
6149         (ContentsScaleFactor::NPP_New):
6150         (ContentsScaleFactor::NPP_SetValue):
6151
6152 2011-10-25  Sam Weinig  <sam@webkit.org>
6153
6154         Add WKBrowsingContextLoadDelegateTest test for a failed load
6155         https://bugs.webkit.org/show_bug.cgi?id=70859
6156
6157         Reviewed by Anders Carlsson.
6158
6159         * TestWebKitAPI/Tests/WebKit2ObjC/WKBrowsingContextLoadDelegateTest.mm:
6160         (TEST_F): Add TEST_F(WKBrowsingContextLoadDelegateTest, SimpleLoadFail).
6161
6162         (-[SimpleLoadFailDelegate browsingContextControllerDidFailProvisionalLoad:withError:]):
6163         Test that we are getting the correct NSError code and domain.
6164
6165 2011-10-25  Erik Arvidsson  <arv@chromium.org>
6166
6167         JS Test Harness: Make successfullyParsed optional
6168         https://bugs.webkit.org/show_bug.cgi?id=70784
6169
6170         Reviewed by Ojan Vafai.
6171
6172         * Scripts/make-new-script-test:
6173         (writeTestFile): Update the template for script tests.
6174
6175 2011-10-25  Anders Carlsson  <andersca@apple.com>
6176
6177         Plug-ins have to use JavaScript to find out the current device scale factor
6178         https://bugs.webkit.org/show_bug.cgi?id=67225
6179         <rdar://problem/10048258>
6180
6181         Reviewed by Darin Adler.
6182
6183         Add a plug-in test for getting the contents scale factor.
6184
6185         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
6186         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ContentsScaleFactor.cpp: Added.
6187         (ContentsScaleFactor::ContentsScaleFactor):
6188         (ContentsScaleFactor::contentsScaleFactor):
6189         (ContentsScaleFactor::ScriptableObject::hasProperty):
6190         (ContentsScaleFactor::ScriptableObject::getProperty):
6191         (ContentsScaleFactor::ScriptableObject::pluginTest):
6192         (ContentsScaleFactor::NPP_GetValue):
6193
6194 2011-10-25  Sam Weinig  <sam@webkit.org>
6195
6196         Flesh out WKBrowsingContextLoadDelegate a bit
6197         https://bugs.webkit.org/show_bug.cgi?id=70846
6198
6199         Reviewed by Anders Carlsson.
6200
6201         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
6202         * TestWebKitAPI/Tests/WebKit2ObjC: Added.
6203         * TestWebKitAPI/Tests/WebKit2ObjC/WKBrowsingContextLoadDelegateTest.mm: Added.
6204         (WKBrowsingContextLoadDelegateTest::WKBrowsingContextLoadDelegateTest):
6205         (WKBrowsingContextLoadDelegateTest::SetUp):
6206         (WKBrowsingContextLoadDelegateTest::TearDown):
6207         (-[SimpleLoadDelegate browsingContextControllerDidFinishLoad:]):
6208         (TEST_F):
6209         Add basic testing for WKBrowsingContextLoadDelegate.
6210
6211 2011-10-25  Alexey Proskuryakov  <ap@apple.com>
6212
6213         make-script-test-wrappers should not look outside fast/js
6214         https://bugs.webkit.org/show_bug.cgi?id=66357
6215
6216         Reviewed by Maciej Stachowiak.
6217
6218         * Scripts/make-script-test-wrappers: Don't look for template files everywhere, only support
6219         fast/js.
6220
6221 2011-10-25  Adam Barth  <abarth@webkit.org>
6222
6223         Add a --make option to update-webkit to engage the Makefile-based build system
6224         https://bugs.webkit.org/show_bug.cgi?id=70847
6225
6226         Reviewed by Dimitri Glazkov.
6227
6228         I can never remember what crazy environment variables I'm supposed to
6229         set to tell GYP what to do.
6230
6231         * Scripts/update-webkit:
6232
6233 2011-10-25  Raphael Kubo da Costa  <kubo@profusion.mobi>
6234
6235         [EFL] DRT: Force layout when sending events.
6236         https://bugs.webkit.org/show_bug.cgi?id=70357
6237
6238         Reviewed by Antonio Gomes.
6239
6240         Some tests such as editing/input/page-up-down-scrolls.html rely on the
6241         scrollbars being properly set up during the onload events.
6242         Document::implicitClose(), however, dispatches the onload events before
6243         calling FrameView::layout(), so these tests usually fail.
6244
6245         We now do the same as the GTK+ and Chromium ports and manually force the
6246         layout when dispatching events.
6247
6248         * DumpRenderTree/efl/EventSender.cpp:
6249         (sendMouseEvent):
6250         (keyDownCallback):
6251
6252 2011-10-25  Daniel Bates  <dbates@rim.com>
6253
6254         build-webkit --clean should only clean build directory for CMake-based ports
6255         https://bugs.webkit.org/show_bug.cgi?id=70834
6256
6257         Reviewed by Antonio Gomes.
6258
6259         Currently build-webkit --clean both cleans the build directory and builds WebKit for
6260         CMake-based ports. Instead, build-webkit --clean should only clean up the build directory
6261         as per its description in the usage message of build-webkit.
6262
6263         * Scripts/webkitdirs.pm:
6264         (buildCMakeProjectOrExit):
6265
6266 2011-10-25  Anna Cavender  <annacc@chromium.org>
6267
6268         Enable running of track layout tests for Chromium
6269         https://bugs.webkit.org/show_bug.cgi?id=70803
6270
6271         Reviewed by Adam Barth.
6272
6273         Make DumpRenderTree/TestShell always run with track enabled
6274         (equivalent of --enable-video-track).
6275
6276         * DumpRenderTree/chromium/TestShell.cpp:
6277         (TestShell::TestShell):
6278
6279 2011-10-25  Vamshikrishna Yellenki  <vamshi@motorola.com>
6280
6281         [WK2]Main window never gets focus using TAB key in WebKit2 GTK+.
6282         https://bugs.webkit.org/show_bug.cgi?id=70603
6283
6284         Reviewed by Gustavo Noronha Silva.
6285
6286         * MiniBrowser/gtk/BrowserWindow.c:
6287         (focus):
6288         (unFocus):
6289         (browserWindowUIClientInit):
6290
6291 2011-10-25  Dan Bernstein  <mitz@apple.com>
6292
6293         <rdar://problem/10337033> DOMRangeOfString:relativeTo:options has problems with -webkit-user-select: none
6294
6295         Reviewed by Adam Roben.
6296
6297         * TestWebKitAPI/Tests/mac/DOMRangeOfString.mm:
6298         (TestWebKitAPI::TEST):
6299
6300 2011-10-25  Eric Seidel  <eric@webkit.org>
6301
6302         WIN: editing tests fail under NRWT because editing delegate callbacks aren't stripped
6303         https://bugs.webkit.org/show_bug.cgi?id=64471
6304
6305         Reviewed by Adam Roben.
6306
6307         Speculative fix.
6308
6309         * Scripts/webkitpy/layout_tests/port/win.py:
6310         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
6311
6312 2011-10-24  Mikhail Naganov  <mnaganov@chromium.org>
6313
6314         Fix bug summary when rolling Chromium DEPS to LKGR.
6315         https://bugs.webkit.org/show_bug.cgi?id=69917
6316
6317         Reviewed by Adam Barth.
6318
6319         * Scripts/webkitpy/tool/bot/irc_command.py:
6320         * Scripts/webkitpy/tool/bot/sheriff.py:
6321         * Scripts/webkitpy/tool/commands/roll.py:
6322         * Scripts/webkitpy/tool/commands/roll_unittest.py:
6323
6324 2011-10-24  Ivan Briano  <ivan@profusion.mobi>
6325
6326         [EFL] Build fix. Add include path to find npapi.h
6327         https://bugs.webkit.org/show_bug.cgi?id=70730
6328
6329         Unreviewed build fix.
6330
6331         * DumpRenderTree/efl/CMakeLists.txt: Add include path for npapi.h
6332
6333 2011-10-24  Eric Seidel  <eric@webkit.org>
6334
6335         Plugin tests are failing after switching WK2 to NRWT
6336         https://bugs.webkit.org/show_bug.cgi?id=70760
6337
6338         Reviewed by Simon Fraser.
6339
6340         Perpetuate this horrible hack.
6341
6342         * Scripts/webkitpy/layout_tests/port/webkit.py:
6343
6344 2011-10-24  Eric Seidel  <eric@webkit.org>
6345
6346         NRWT doesn't support ORWT's --root option, but should
6347         https://bugs.webkit.org/show_bug.cgi?id=70416
6348
6349         Reviewed by Dirk Pranke.
6350
6351         Add support for ORWT's --root option which is used for
6352         running tests with a nightly build of WebKit.
6353
6354         * Scripts/old-run-webkit-tests:
6355         * Scripts/webkitpy/layout_tests/port/base.py:
6356          - webkit_build_path was never called by any code, removed.
6357         * Scripts/webkitpy/layout_tests/port/webkit.py:
6358         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
6359
6360 2011-10-24  David Levin  <levin@chromium.org>
6361
6362         check-webkit-style complains about WEBKIT_EXPORT in test/WebUnitTests.h
6363         https://bugs.webkit.org/show_bug.cgi?id=70747
6364
6365         Reviewed by Tony Chang.
6366
6367         * Scripts/webkitpy/style/checkers/cpp.py: Allow WEBKIT_EXPORT in tests dir.
6368         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Corresponding test.
6369
6370 2011-10-24  Balazs Kelemen  <kbalazs@webkit.org>
6371
6372         [Qt][WK2] Crash when running pixel tests
6373         https://bugs.webkit.org/show_bug.cgi?id=70725
6374
6375         Reviewed by Chang Shu.
6376
6377         * WebKitTestRunner/qt/TestInvocationQt.cpp:
6378         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
6379
6380 2011-10-24  Carlos Garcia Campos  <cgarcia@igalia.com>
6381
6382         [GTK] Remove g_thread_init() calls
6383         https://bugs.webkit.org/show_bug.cgi?id=70717
6384
6385         Reviewed by Xan Lopez.
6386
6387         With newer glib it's not needed anymore and with previous versions
6388         it was already called by g_type_init().
6389
6390         * DumpRenderTree/gtk/DumpRenderTree.cpp:
6391         (main):
6392         * GtkLauncher/main.c:
6393         (main):
6394         * MiniBrowser/gtk/main.c:
6395         (main):
6396
6397 2011-10-24  Alejandro G. Castro  <alex@igalia.com>
6398
6399         [WK2] [GTK] WebKitTestRunner crashes with heap corruption
6400         https://bugs.webkit.org/show_bug.cgi?id=69403
6401
6402         We did not added autotoolsconfig.h to the config.h file and we
6403         were adding Platform.h without the config.h in the
6404         WebKitTestRunnerPrefix.h. Added the autotoolsconfig.h include in
6405         the Prefix file.
6406
6407         Reviewed by Martin Robinson.
6408
6409         * WebKitTestRunner/WebKitTestRunnerPrefix.h:
6410
6411 2011-10-23  Tomasz Morawski  <t.morawski@samsung.com>
6412
6413         [EFL] Remove unused code from EWebLauncher
6414         https://bugs.webkit.org/show_bug.cgi?id=70590
6415
6416         Reviewed by Andreas Kling.
6417
6418         Removed unused function on_closeWindow from EWebLauncher
6419
6420         * EWebLauncher/main.c:
6421
6422 2011-10-23  Luke Macpherson   <macpherson@chromium.org>
6423
6424         Add watchlist for WebCore/css and put myself on it.
6425         https://bugs.webkit.org/show_bug.cgi?id=70701
6426
6427         Reviewed by Adam Barth.
6428
6429         * Scripts/webkitpy/common/config/watchlist:
6430         Add watchlist for WebCore/css and put myself on it.
6431
6432 2011-10-23  Kaustubh Atrawalkar  <kaustubh@motorola.com>
6433
6434         [GTK] Implement KeyDown function for WebKit2 EventSender.
6435         https://bugs.webkit.org/show_bug.cgi?id=69410
6436
6437         Reviewed by Martin Robinson.
6438
6439         Implement the KeyDown function on Gtk platform.
6440
6441         * WebKitTestRunner/GNUmakefile.am:
6442         * WebKitTestRunner/TestController.cpp:
6443         (WTR::TestController::TestController):
6444         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
6445         * WebKitTestRunner/gtk/EventSenderProxyGtk.cpp: Added.
6446         (WTR::EventSenderProxy::EventSenderProxy):
6447         (WTR::dispatchEvent):
6448         (WTR::getModifiers):
6449         (WTR::getGDKKeySymForKeyRef):
6450         (WTR::EventSenderProxy::keyDown):
6451
6452 2011-10-22  Zan Dobersek  <zandobersek@gmail.com>
6453
6454         [WK2][GTK] run-launcher script fails after testing WK2 tests
6455         https://bugs.webkit.org/show_bug.cgi?id=70676
6456
6457         Reviewed by Andreas Kling.
6458
6459         Pass the '-2' flag instead of '--webkit-test-runner' when executing
6460         run-launcher script after testing with WebKitTestRunner.
6461
6462         * Scripts/webkitpy/layout_tests/port/gtk.py:
6463
6464 2011-10-22  Zan Dobersek  <zandobersek@gmail.com>
6465
6466         [WK2][GTK] Missing env variables cause WKTR to crash
6467         https://bugs.webkit.org/show_bug.cgi?id=70674
6468
6469         Reviewed by Martin Robinson.
6470
6471         Set two environment variables that the Gtk's port of WebKitTestRunner
6472         depends on and crashes if they're not explicitly set by the user.
6473         Also modify the setting of the WEBKIT_INSPECTOR_PATH variable so it
6474         properly uses the _build_path method.
6475
6476         * Scripts/webkitpy/layout_tests/port/gtk.py: Set
6477         TEST_RUNNER_INJECTED_BUNDLE_FILENAME and TEST_RUNNER_TEST_PLUGIN_PATH
6478         environment variables.
6479
6480 2011-10-21  Filip Pizlo  <fpizlo@apple.com>
6481
6482         Bencher script doesn't measure GC times accurately
6483         https://bugs.webkit.org/show_bug.cgi?id=70588
6484
6485         Reviewed by Geoff Garen.
6486
6487         Added two new options which allow different ways of measuring GC times:
6488
6489         --measure-gc, which omits calls to gc() between benchmark invocations.
6490         This option takes an optional argument, which is the name of the
6491         VM in which to enable this feature. This allows comparing a single VM
6492         against itself, with and without GC.
6493
6494         --rerun <N>, which causes each sample measurement to include N invocations
6495         which do not have gc() calls between them. The default is N = 1, which
6496         results in the same behavior as before.
6497
6498         You can use either --measure-gc or --rerun <N> for N > 1 (preferably
6499         N >= 3) to get more of a contribution from GC to the measured times.
6500         --rerun results in tighter confidence intervals than --measure-gc, since
6501         it amortizes GC effects in each sample, while with --measure-gc some
6502         samples will see GC and some won't leading to a higher standard devation
6503         and thus requiring more samples to reduce confidence intervals to
6504         managable levels.
6505
6506         * Scripts/bencher:
6507
6508 2011-10-21  Sam Weinig  <sam@webkit.org>
6509
6510         Remove ability to create a WKView without a WKContextRef and WKPageGroupRef
6511         https://bugs.webkit.org/show_bug.cgi?id=70653
6512
6513         Reviewed by Simon Fraser.
6514
6515         * MiniBrowser/mac/AppDelegate.h:
6516         * MiniBrowser/mac/AppDelegate.m:
6517         (-[BrowserAppDelegate init]):
6518         (-[BrowserAppDelegate newWindow:]):
6519         (-[BrowserAppDelegate openPanelDidEnd:returnCode:contextInfo:]):
6520         * MiniBrowser/mac/BrowserWindowController.h:
6521         * MiniBrowser/mac/BrowserWindowController.m:
6522         (-[BrowserWindowController initWithContext:pageGroup:]):
6523         (-[BrowserWindowController windowWillClose:]):
6524         (createNewPage):
6525         (-[BrowserWindowController awakeFromNib]):
6526         * TestWebKitAPI/mac/WebKitAgnosticTest.mm:
6527         (TestWebKitAPI::WebKitAgnosticTest::runWebKit2Test):
6528         Update testing code to pass a PageGroup as necessary.
6529
6530 2011-10-21  Leandro Pereira  <leandro@profusion.mobi>
6531
6532         webkitpy: Teach NRWT about the EFL port
6533         https://bugs.webkit.org/show_bug.cgi?id=70637
6534
6535         Allows using ``efl'' as a platform when executing NRWT.
6536
6537         Reviewed by Eric Seidel.
6538
6539         * Scripts/webkitpy/layout_tests/port/efl.py: Added.
6540         * Scripts/webkitpy/layout_tests/port/efl_unittest.py: Added.
6541         * Scripts/webkitpy/layout_tests/port/factory.py: Adjust factory to make EflPort objects
6542         when using PortFactory.get(port_name='efl').
6543
6544 2011-10-21  Devdatta Deshpande  <pwjd73@motorola.com>
6545
6546         [Gtk] mousemove event always has metaKey == true
6547         https://bugs.webkit.org/show_bug.cgi?id=35299
6548
6549         Reviewed by Martin Robinson.
6550
6551         * DumpRenderTree/gtk/EventSender.cpp:
6552         (gdkModifierFromJSValue): GDK_META_MASK is used to represent metaKey for platform GTK mouse and
6553         keyboard events. So making it in sync with WebCore files.
6554         (mouseMoveToCallback): Sending modifier keys state for mouse move event.
6555
6556 2011-10-21  Carlos Garcia Campos  <cgarcia@igalia.com>
6557
6558         watchlist: Add a new entry for WebKit2 GTK+ public API
6559         https://bugs.webkit.org/show_bug.cgi?id=70601
6560
6561         Reviewed by Philippe Normand.
6562
6563         Subscribe myself to WebKit2 GTK+ API patches. Also add a message
6564         for patches that might include new API that points to the WebKit2
6565         GTK+ API guidelines.
6566
6567         * Scripts/webkitpy/common/config/watchlist:
6568
6569 2011-10-21  Simon Hausmann  <simon.hausmann@nokia.com>
6570
6571         Remove QtScript source code from WebKit.
6572         https://bugs.webkit.org/show_bug.cgi?id=64088
6573
6574         Reviewed by Tor Arne Vestbø.
6575
6576         Removed dead code that isn't developed anymore.
6577
6578         * Scripts/webkitpy/style/checker_unittest.py:
6579
6580 2011-10-20  Eric Seidel  <eric@webkit.org>
6581
6582         Switch webkit2 bot to NRWT
6583         https://bugs.webkit.org/show_bug.cgi?id=56729
6584
6585         Reviewed by Adam Barth.
6586
6587         Move the WK2 bot to using NRWT instead of ORWT, as announced on webkit-dev:
6588         https://lists.webkit.org/pipermail/webkit-dev/2011-October/018337.html
6589
6590         * Scripts/run-webkit-tests:
6591         (runningOnBuildBot):
6592         (useNewRunWebKitTests):
6593
6594 2011-10-20  Eric Seidel  <eric@webkit.org>
6595
6596         NRWT is asserting on DRT due to lack of newline when processing line from DumpRenderTree
6597         https://bugs.webkit.org/show_bug.cgi?id=70585
6598
6599         Reviewed by Adam Barth.
6600
6601         This really should be an assert, as this inidicates a programming error in either DRT or NRWT.
6602         However since DRT was functioning fine before in Qt, we'll make this an error log for now
6603         and once we find the root cause flip this back to an assert.
6604
6605         * Scripts/webkitpy/layout_tests/port/webkit.py:
6606
6607 2011-10-20  Yuta Kitamura  <yutak@chromium.org>
6608
6609         Unreviewed, rolling out r98064.
6610         http://trac.webkit.org/changeset/98064
6611         https://bugs.webkit.org/show_bug.cgi?id=64788
6612
6613         Broke Mac bots.
6614
6615         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
6616
6617 2011-10-20  Yuta Kitamura  <yutak@chromium.org>
6618
6619         WebSocket: Connecting to localhost:8880 takes one second on Windows
6620         https://bugs.webkit.org/show_bug.cgi?id=64788
6621
6622         Reviewed by Dirk Pranke.
6623
6624         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
6625         Bind to "localhost" instead of "127.0.0.1" to let pywebsocket listen on both
6626         IPv4 and IPv6 addresses. This should prevent the test reserved-opcodes.html
6627         from timing out on Windows, because this test tries to open a lot of
6628         connections to localhost and each attempt takes one second to fall back from
6629         IPv6 to IPv4 on Windows (I have no idea why Windows works like this, though).
6630
6631 2011-10-20  Ryosuke Niwa  <rniwa@webkit.org>
6632
6633         nrwt: newly generated results are put in cross-platform directory
6634         https://bugs.webkit.org/show_bug.cgi?id=68931
6635
6636         Reviewed by Dirk Pranke.
6637
6638         The bug was caused by SingleTestRunner._add_missing_baselines's always calling _save_baseline_data
6639         with generate_new_baseline set to False. Fixed the bug by always passing True when .png file is missing
6640         (because png images are typically different on each platform), and passing True when .txt file is missing
6641         and the actual result's first line matches the regular expression "layer at \(\d+,\d+\) size \d+x\d+".
6642
6643         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
6644         * Scripts/webkitpy/layout_tests/port/test.py:
6645         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py: Changed the expectation
6646         and added a test case.
6647
6648 2011-10-20  Kent Tamura  <tkent@chromium.org>
6649
6650         Unreviewed. Adding myself to watchlist.
6651
6652         * Scripts/webkitpy/common/config/watchlist:
6653         Add "ChromiumDumpRenderTree" and "Forms" definitions, and cc them to me.
6654
6655 2011-10-20  Nico Weber  <thakis@chromium.org>
6656
6657         [chromium/mac] Add support for building with make
6658         https://bugs.webkit.org/show_bug.cgi?id=70455
6659
6660         Reviewed by Adam Barth.
6661
6662         Depends on http://codereview.chromium.org/8352029/
6663         `GYP_GENERATORS=make Tools/Scripts/update-webkit --chromium` will
6664         create a toplevel Makefile.chromium.
6665
6666         If that file exists and is newer than Source/WebKit/chromium/WebKit.xcodeproj,
6667         `Tools/Scripts/build-webkit --chromium` will build with make instead of
6668         xcodebuild.
6669
6670         Also let nrwt use a binary from out/ if it exists.
6671
6672         * Scripts/webkitdirs.pm:
6673         (determineBaseProductDir):
6674         (isChromiumMacMake):
6675         (determineIsChromiumMacMake):
6676         (buildChromium):
6677
6678 2011-10-20  Eric Seidel  <eric@webkit.org>
6679
6680         Possible REGRESSION(97879): NRWT fails when DumpRenderTree crashes
6681         https://bugs.webkit.org/show_bug.cgi?id=70524
6682
6683         Reviewed by Adam Barth.
6684
6685         This is a speculative fix, since I do not use a platform
6686         which outputs crashlogs over stderr.
6687
6688         * Scripts/webkitpy/layout_tests/port/server_process.py:
6689         * Scripts/webkitpy/layout_tests/port/webkit.py:
6690
6691 2011-10-20  Eric Seidel  <eric@webkit.org>
6692
6693         REGRESSION(97879): Pixel tests no longer work with NRWT on Mac
6694         https://bugs.webkit.org/show_bug.cgi?id=70492
6695
6696         Reviewed by Adam Barth.
6697
6698         The bug turned out to be that I was assuming the block.content
6699         would be empty before the binary content following Content-Length
6700         was read inside _read_block.  Turns out its not, due to extra newlines
6701         and "ExpectedHash" header.
6702
6703         In the process of trying to figure out what was going wrong I ended up
6704         cleaning up our newline usage in DumpRenderTree a little.  Moved
6705         two error messages from stdout to stderr, and fixed a little code indent/whitespace.
6706
6707         I also fixed ServerProcess to use "deadline" everywhere instead of timeout
6708         per Adam's request in the original bug.
6709
6710         * DumpRenderTree/PixelDumpSupport.cpp:
6711         (dumpWebViewAsPixelsAndCompareWithExpected):
6712         * DumpRenderTree/cg/ImageDiffCG.cpp:
6713         (main):
6714         * DumpRenderTree/mac/DumpRenderTree.mm:
6715         (dump):
6716         * DumpRenderTree/mac/PixelDumpSupportMac.mm:
6717         (restoreMainDisplayColorProfile):
6718         (setupMainDisplayColorProfile):
6719         * Scripts/webkitpy/layout_tests/port/server_process.py:
6720         * Scripts/webkitpy/layout_tests/port/webkit.py:
6721         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
6722
6723 2011-10-20  Tony Chang  <tony@chromium.org>
6724
6725         [chromium] Remove <stdint.h> from ImageDiff and use
6726         unsigned int instead of uint32_t.
6727
6728         Unreviewed, fixing the chromium win build.
6729
6730         * DumpRenderTree/chromium/ImageDiff.cpp:
6731         (Image::pixelAt):
6732         (Image::setPixelAt):
6733         (maxOf3):
6734         (getRedComponent):
6735         (getGreenComponent):
6736         (getBlueComponent):
6737         (weightedPercentageDifferent):
6738         (createImageDiff):
6739
6740 2011-10-20  Hao Zheng  <zhenghao@chromium.org>
6741
6742         [Chromium] Reduce dependencies of ImageDiff to compile it for Android.
6743         https://bugs.webkit.org/show_bug.cgi?id=69997
6744
6745         Reviewed by Tony Chang.
6746
6747         This change is required to make ImageDiff for Android, and can benefit
6748         other platforms, too. On Android, most targets are built as
6749         toolsets:target. while ImageDiff needs to be built as toolsets:host.
6750         Currently, building the standalone target of ImageDiff depends on about
6751         85 other targets (try  make ImageDiff'), which is inefficient. After the
6752         change, ImageDiff only depends on webkit_support_gfx, which in turn
6753         depends on libpng and zlib.
6754
6755         * DumpRenderTree/chromium/ImageDiff.cpp:
6756         (main):
6757
6758 2011-10-20  Leandro Pereira  <leandro@profusion.mobi>
6759
6760         [EFL] Unreviewed DumpRenderTree build fix.
6761
6762         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
6763         (LayoutTestController::addChromeInputField): Add stub.
6764         (LayoutTestController::removeChromeInputField): Ditto.
6765         (LayoutTestController::focusWebView): Ditto.
6766         (LayoutTestController::setBackingScaleFactor): Ditto.
6767
6768 2011-10-20  Leandro Pereira  <leandro@profusion.mobi>
6769
6770         [EFL] Plug ImageDiff/DumpRenderTree on CMake build system
6771         https://bugs.webkit.org/show_bug.cgi?id=70142
6772
6773         Reviewed by Gustavo Noronha Silva.
6774
6775         * CMakeListsEfl.txt: Include DRT CMakeLists.
6776         * DumpRenderTree/efl/CMakeLists.txt: Added.
6777
6778 2011-10-20  Philippe Normand  <pnormand@igalia.com>
6779
6780         [style] Allow usage of NULL in gst_*
6781         https://bugs.webkit.org/show_bug.cgi?id=70498
6782
6783         Reviewed by David Levin.
6784
6785         * Scripts/webkitpy/style/checkers/cpp.py: Simplified the detection
6786         of gst_ calls. Now just ignore NULL in all of them.
6787         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Test for above change.
6788
6789 2011-10-20  Leandro Pereira  <leandro@profusion.mobi>
6790
6791         [EFL] Unreviewed. Build fix after r97043.
6792
6793         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
6794         (DumpRenderTreeChrome::resetDefaultsToConsistentValues): Use ewk_view_scale_set() instead of ewk_view_page_scale().
6795         * DumpRenderTree/efl/EventSender.cpp:
6796         (scalePageByCallback): Ditto.
6797
6798 2011-10-20  Philippe Normand  <pnormand@igalia.com>
6799
6800         Unreviewed. Adding myself to watchlists.
6801         * Scripts/webkitpy/common/config/watchlist:
6802
6803 2011-10-19  Eric Seidel  <eric@webkit.org>
6804
6805         Reviewed by Adam Barth.
6806
6807         Teach new-run-webkit-tests about #CRASHED and #CRASHED - WebProcess
6808         https://bugs.webkit.org/show_bug.cgi?id=63683
6809
6810         * Scripts/webkitpy/common/net/layouttestresults.py:
6811         * Scripts/webkitpy/common/net/resultsjsonparser.py:
6812         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
6813         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
6814         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
6815         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
6816         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
6817         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
6818         * Scripts/webkitpy/layout_tests/port/base.py:
6819         * Scripts/webkitpy/layout_tests/port/server_process.py:
6820         * Scripts/webkitpy/layout_tests/port/webkit.py:
6821         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
6822
6823 2011-10-19  Jochen Eisinger  <jochen@chromium.org>
6824
6825         Chromium doesn't support git branch builds, so don't add the branch to the product dir
6826         https://bugs.webkit.org/show_bug.cgi?id=70449
6827
6828         Reviewed by Eric Seidel.
6829
6830         * Scripts/webkitdirs.pm:
6831         (determineBaseProductDir):
6832
6833 2011-10-19  Ojan Vafai  <ojan@chromium.org>
6834
6835         Unreviewed. Adding myself to watchlists.
6836         * Scripts/webkitpy/common/config/watchlist:
6837
6838 2011-10-19  Tom Zakrajsek  <tomz@codeaurora.org>
6839
6840         Add a suggest-nominations command to webkit-patch for computing potential committer/reviewer nominations
6841         https://bugs.webkit.org/show_bug.cgi?id=62166
6842
6843         Reviewed by Eric Seidel.
6844
6845         Included options to control committer/reviewer patch count requirements,
6846         an age-limit on patches, and verbose output for more in-depth analysis.
6847
6848         * Scripts/webkitpy/common/checkout/changelog.py:
6849         * Scripts/webkitpy/tool/commands/__init__.py:
6850         * Scripts/webkitpy/tool/commands/suggestnominations.py: Added.
6851         * Scripts/webkitpy/tool/commands/suggestnominations_unittest.py: Added.
6852
6853 2011-10-19  Eric Seidel  <eric@webkit.org>
6854
6855         new-run-webkit-tests -2 was crashing due to both "java" and "java/" being listed in Skipped files.
6856         Make NRWT normalize directory names in Skipped files during parsing.
6857
6858         Unreviewed.  Very simple (tested) change.
6859
6860         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
6861         * Scripts/webkitpy/layout_tests/port/webkit.py:
6862
6863 2011-10-19  David Levin  <levin@chromium.org>
6864
6865         watchlist: Should be more robust to bad regex.
6866         https://bugs.webkit.org/show_bug.cgi?id=69486
6867
6868         Reviewed by Adam Barth.
6869
6870         * Scripts/webkitpy/common/config/watchlist: Change the instructions due to
6871           watchlist being checked by check-webkit-style (bug 69487) and remove the .* from file
6872           patterns since they are no longer anchored on the right hand side.
6873         * Scripts/webkitpy/common/watchlist/amountchangedpattern.py: Change *pattern to take a compiled regex directly.
6874         * Scripts/webkitpy/common/watchlist/amountchangedpattern_unittest.py: Ditto.
6875         * Scripts/webkitpy/common/watchlist/changedlinepattern.py: Ditto.
6876         * Scripts/webkitpy/common/watchlist/changedlinepattern_unittest.py: Ditto.
6877         * Scripts/webkitpy/common/watchlist/filenamepattern.py: Ditto.
6878         * Scripts/webkitpy/common/watchlist/filenamepattern_unittest.py: Ditto and change a test now that
6879         filenames are no longer anchored on the right hand side.
6880         * Scripts/webkitpy/common/watchlist/watchlistparser.py: Catch regex errors and log them as errors.
6881         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py: Add some tests to verify that bad regexes
6882         don't kill everything.
6883
6884 2011-10-19  David Levin  <levin@chromium.org>
6885
6886         watchlist: Add a stylecheck to do validity checks for the watchlist config.
6887         https://bugs.webkit.org/show_bug.cgi?id=69487
6888
6889         Reviewed by Adam Barth.
6890
6891         * Scripts/webkitpy/style/checker.py: Add the watchlist file type.
6892         * Scripts/webkitpy/style/checkers/watchlist.py: Added.
6893         * Scripts/webkitpy/style/checkers/watchlist_unittest.py: Added.
6894
6895 2011-10-19  Eric Seidel  <eric@webkit.org>
6896
6897         NRWT ServerProcess can't read lines from stderr and stdio separately
6898         https://bugs.webkit.org/show_bug.cgi?id=70435
6899
6900         Reviewed by Adam Barth.
6901
6902         This is the underlying bug behind our lack of WK2 support, and timeout detection on Windows.
6903         DRT expects RWT to read one line at a time from stderr and stdout independently.
6904         This allows DRT to communicate errors over stderr and have them responded to immediately
6905         regardless of what state the stdio stream may be in.  Previously NRWT's ServerProcess
6906         class only exposed a read_line function which would read a line from stdout
6907         and batch all output from stderr into an error buffer.  Callers would only be able
6908         to respond to commands over stderr if they happened to correspond with a stdout line boundary.
6909
6910         * Scripts/webkitpy/layout_tests/port/server_process.py:
6911         * Scripts/webkitpy/layout_tests/port/webkit.py:
6912
6913 2011-10-19  Chang Shu  <cshu@webkit.org>
6914
6915         [Qt] Support WK1-only Skipped list
6916         https://bugs.webkit.org/show_bug.cgi?id=70254
6917
6918         Reviewed by Csaba Osztrogonác.
6919
6920         Added a new search path qt-wk1 for skipping WK1-only tests.
6921
6922         * Scripts/webkitpy/layout_tests/port/qt.py: NRWT
6923         * Scripts/old-run-webkit-tests: ORWT
6924
6925 2011-10-19  Gabor Rapcsanyi  <rgabor@webkit.org>
6926
6927         [GTK] Enable to use built-product-archive on GTK platform
6928         https://bugs.webkit.org/show_bug.cgi?id=70319
6929
6930         Reviewed by Csaba Osztrogonác.
6931
6932         * BuildSlaveSupport/built-product-archive:
6933
6934 2011-10-18  David Levin  <levin@chromium.org>
6935
6936         watchlist: Should try to run if it can and not throw on mistakes.
6937         https://bugs.webkit.org/show_bug.cgi?id=70358
6938
6939         Reviewed by Adam Barth.
6940
6941         * Scripts/webkitpy/common/system/outputcapture.py: Add the ability to capture log output.
6942         * Scripts/webkitpy/common/watchlist/watchlistloader_unittest.py: Adapt to the logging of errors.
6943         * Scripts/webkitpy/common/watchlist/watchlistparser.py: Change to log problems and fix problems when found.
6944         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py: Adapt to the logging of errors.
6945         * Scripts/webkitpy/common/watchlist/watchlistrule.py: Expose a way to remove instructions.
6946         * Scripts/webkitpy/common/watchlist/watchlistrule_unittest.py: Test the new function.
6947
6948 2011-10-18  Sam Weinig  <sam@webkit.org>
6949
6950         Move uses of C-SPI out of WKView.h and into WKViewPrivate.h
6951         https://bugs.webkit.org/show_bug.cgi?id=70387
6952
6953         Reviewed by Dan Bernstein.
6954
6955         * MiniBrowser/mac/BrowserWindowController.m:
6956         * TestWebKitAPI/Tests/mac/DeviceScaleFactorOnBack.mm:
6957         * TestWebKitAPI/mac/JavaScriptTestMac.mm:
6958         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
6959         * TestWebKitAPI/mac/WebKitAgnosticTest.mm:
6960         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
6961         Update to include WKViewPrivate.h where necessary.
6962
6963 2011-10-18  Joseph Pecoraro  <joepeck@webkit.org>
6964
6965         Web Inspector: Some localizedStrings.js Updates
6966         https://bugs.webkit.org/show_bug.cgi?id=70365
6967
6968         Reviewed by Darin Adler.
6969
6970         * Scripts/check-inspector-strings:
6971         Updated to the new webkitpy way to determine checkout root.
6972
6973         * Scripts/webkitdirs.pm:
6974         (copyInspectorFrontendFiles):
6975         Made it so build-webkit --inspector-frontend copies over
6976         the localizedStrings.js file as well as the frontend files.
6977
6978 2011-10-18  Raphael Kubo da Costa  <kubo@profusion.mobi>
6979
6980         [EFL] DRT: Clear list of visited pages before each test.
6981         https://bugs.webkit.org/show_bug.cgi?id=70355
6982
6983         Reviewed by Antonio Gomes.
6984
6985         r97596 uncovered a bug in the implementation: the list of visited
6986         links was not cleared between tests, which caused tests such as
6987         fast/repaint/outline-repaint-glitch.html to be flaky (the link color
6988         changed depending on whether fast/repaint/border-repaint-glitch.html
6989         was visited before or not).
6990
6991         Call ewk_history_clear to make sure the visited links list does not
6992         persist between tests.
6993
6994         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
6995         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
6996
6997 2011-10-18  Dimitri Glazkov  <dglazkov@chromium.org>
6998
6999         Split TestExpectationsEditor into its own file.
7000         https://bugs.webkit.org/show_bug.cgi?id=70348
7001
7002         Reviewed by Adam Barth.
7003
7004         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor.py: Copied from Tools/Scripts/webkitpy/layout_tests/models/test_expectations.py.
7005         * Scripts/webkitpy/layout_tests/controllers/test_expectations_editor_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py.
7006         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Adjusted for change.
7007         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Ditto.
7008         * Scripts/webkitpy/tool/servers/gardeningserver.py: Ditto.
7009
7010 2011-10-18  Ryosuke Niwa  <rniwa@webkit.org>
7011
7012         [Chromium] buildbot runs nrwt with --chromium instead of --chromium-cg-mac
7013         https://bugs.webkit.org/show_bug.cgi?id=70313
7014
7015         Reviewed by Adam Barth.
7016
7017         Fixed the bug by passing --platform --chromium-cg-mac in RunWebKitTests. We can't always add
7018         this parameter because build-webkit and other scripts don't accept it.
7019
7020         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
7021
7022 2011-10-18  Adam Barth  <abarth@webkit.org>
7023
7024         Subscribe myself to WebIDL changes
7025         https://bugs.webkit.org/show_bug.cgi?id=70341
7026
7027         Reviewed by David Levin.
7028
7029         I've also added some documentation and refactored my previous
7030         subscriptions to make it eaiser for others to subscribe as well.
7031
7032         * Scripts/webkitpy/common/config/watchlist:
7033
7034 2011-10-18  Dimitri Glazkov  <dglazkov@chromium.org>
7035
7036         Add myself to the newfangled watchlist plumbing.
7037         https://bugs.webkit.org/show_bug.cgi?id=70338
7038
7039         Reviewed by David Levin.
7040
7041         * Scripts/webkitpy/common/config/watchlist: Added sheepishly.
7042
7043 2011-10-18  Dimitri Glazkov  <dglazkov@chromium.org>
7044
7045         Fix a unit test in TestFailures.
7046         https://bugs.webkit.org/show_bug.cgi?id=70337
7047
7048         Reviewed by Adam Barth.
7049
7050         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/NewBugForm_unittests.js: Turns out,
7051             the value of a method property should always be lowercase.
7052
7053 2011-10-18  Adam Barth  <abarth@webkit.org>
7054
7055         Always enable ENABLE(XPATH)
7056         https://bugs.webkit.org/show_bug.cgi?id=70217
7057
7058         Reviewed by Eric Seidel.
7059
7060         * Scripts/build-webkit:
7061
7062 2011-10-18  Dimitri Glazkov  <dglazkov@chromium.org>
7063
7064         garden-o-matic should not report flaking in expected failures as failures.
7065         https://bugs.webkit.org/show_bug.cgi?id=70279
7066
7067         Reviewed by Adam Barth.
7068
7069         On the bots, flakes that are already expected to fail aren't reported as failures (they keep the bot green).
7070         For example, if TEXT failure is expected, a TEXT TIMEOUT result will not cause the redness. We should do the
7071         same for garden-o-matic.
7072
7073         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js: Spuriously changed to use Array iterators rather than jQuery APIs,
7074             rewrote the logic for determining expected or unexpected failures as a central analysis tool.
7075         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js: Added unit tests.
7076
7077 2011-10-18  Zan Dobersek  <zandobersek@gmail.com>
7078
7079         [Gtk] Support for client-based geolocation
7080         https://bugs.webkit.org/show_bug.cgi?id=64970
7081
7082         Reviewed by Martin Robinson.
7083
7084         Add support for testing client-based geolocation tests in DumpRenderTree.
7085
7086         * DumpRenderTree/gtk/DumpRenderTree.cpp:
7087         (resetDefaultsToConsistentValues): Reset GeolocationClientMock.
7088         (createWebView): Create web view after declaring DumpRenderTree mode.
7089         That way a mock client for geolocation is used instead of a GeoClue
7090         client if a web view is created when testing.
7091         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Implement functions for
7092         geolocation layout tests.
7093         (LayoutTestController::setMockGeolocationPosition):
7094         (LayoutTestController::setMockGeolocationError):
7095         (LayoutTestController::setGeolocationPermission):
7096         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
7097         * Scripts/build-webkit: Note that the Gtk port is using client-based
7098         geolocation, even though changing this flag has no effect on compilation.
7099
7100 2011-10-07  Andrey Kosyakov  <caseq@chromium.org>
7101
7102         [Chromium] suppress console output from DevTools window
7103         https://bugs.webkit.org/show_bug.cgi?id=69632
7104
7105         Reviewed by Pavel Feldman.
7106
7107         * DumpRenderTree/chromium/TestShell.cpp:
7108         (TestShell::showDevTools):
7109         * DumpRenderTree/chromium/WebViewHost.cpp:
7110         (WebViewHost::didAddMessageToConsole):
7111         (WebViewHost::reset):
7112         (WebViewHost::setLogConsoleOutput):
7113         * DumpRenderTree/chromium/WebViewHost.h:
7114
7115 2011-10-18  Sheriff Bot  <webkit.review.bot@gmail.com>
7116
7117         Unreviewed, rolling out r97735.
7118         http://trac.webkit.org/changeset/97735
7119         https://bugs.webkit.org/show_bug.cgi?id=70317
7120
7121         some people apparently use build-jsc with the qt port without
7122         building webkit first (Requested by torarne on #webkit).
7123
7124         * Scripts/build-webkit:
7125         * Scripts/webkitdirs.pm:
7126         (buildQMakeProject):
7127         (buildQMakeQtProject):
7128
7129 2011-10-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
7130
7131         [Qt] Remove dead code in build-webkit
7132
7133         build-webkit will iterate over a list of subdirectories to build,
7134         but this feature is only used by the Apple ports. For QtWebKit we
7135         always pass an empty directory to buildQMakeProject, which makes
7136         the logic for dealing with sub-projects in buildQMakeProject just
7137         added noise.
7138
7139         If we do want this feature at some point, we should redo it in
7140         the context of the upcoming build system changes for Qt.
7141
7142         Reviewed by Simon Hausmann.
7143
7144         * Scripts/build-webkit:
7145         * Scripts/webkitdirs.pm:
7146         (buildQt):
7147
7148 2011-10-18  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
7149
7150         Remove traces of Symbian support from the build scripts
7151
7152         Reviewed by Simon Hausmann.
7153
7154         * Scripts/build-webkit:
7155         * Scripts/webkitdirs.pm:
7156         (determineBaseProductDir):
7157         (argumentsForConfiguration):
7158         (usesPerConfigurationBuildDirectory):
7159         (buildQMakeProject):
7160
7161 2011-10-18  Ryosuke Niwa  <rniwa@webkit.org>
7162
7163         Build fix after r97672.
7164
7165         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
7166
7167 2011-10-17  Kent Tamura  <tkent@chromium.org>
7168
7169         A follow-up of r97716.
7170         https://bugs.webkit.org/show_bug.cgi?id=68803
7171
7172         * DumpRenderTree/chromium/LayoutTestController.cpp:
7173         (LayoutTestController::reset): Initialize m_dumpCreateView.
7174
7175 2011-10-17  Jochen Eisinger  <jochen@chromium.org>
7176
7177         [chromium] add LayoutTestController.dumpCreateView() to monitor calls to WebViewClient::createView
7178         https://bugs.webkit.org/show_bug.cgi?id=68803
7179
7180         Reviewed by Nate Chapin.
7181
7182         * DumpRenderTree/chromium/LayoutTestController.cpp:
7183         (LayoutTestController::LayoutTestController):
7184         (LayoutTestController::dumpCreateView):
7185         * DumpRenderTree/chromium/LayoutTestController.h:
7186         (LayoutTestController::shouldDumpCreateView):
7187         * DumpRenderTree/chromium/WebViewHost.cpp:
7188         (WebViewHost::createView):
7189
7190 2011-10-17  Nate Chapin  <japhet@chromium.org>
7191
7192         Add myself to watchlists, and add a watch
7193         for Source/WebCore/loader/.
7194         https://bugs.webkit.org/show_bug.cgi?id=70263
7195
7196         Reviewed by David Levin.
7197
7198         * Scripts/webkitpy/common/config/watchlist:
7199
7200 2011-10-17  Raphael Kubo da Costa  <kubo@profusion.mobi>
7201
7202         [EFL] DRT: Make non-recognized keys work in events as well.
7203         https://bugs.webkit.org/show_bug.cgi?id=69860
7204
7205         Reviewed by Antonio Gomes.
7206
7207         fast/forms/select-type-ahead-non-latin.html is the canonical case we fix
7208         here: it has a select box with a few non-latin options, and one of them
7209         is selected via the keyboard.
7210
7211         In this specific case, we need to make EventHandler::keyEvent send a
7212         keyPressEvent to SelectElement, so that
7213         SelectElement::defaultEventHandler calls typeAheadFind(). This happens
7214         only when PlatformKeyEvent::text() is not empty, which was not the case
7215         so far.
7216
7217         * DumpRenderTree/efl/EventSender.cpp:
7218         (keyDownCallback):
7219
7220 2011-10-17  Alok Priyadarshi  <alokp@chromium.org>
7221
7222         Add layout tests for WebPlugin compositor path
7223         https://bugs.webkit.org/show_bug.cgi?id=69027
7224
7225         Reviewed by James Robinson.
7226
7227         * DumpRenderTree/DumpRenderTree.gypi:
7228         * DumpRenderTree/chromium/TestWebPlugin.cpp: Added.
7229         (premultiplyAlpha):
7230         (TestWebPlugin::TestWebPlugin):
7231         (TestWebPlugin::~TestWebPlugin):
7232         (TestWebPlugin::mimeType):
7233         (TestWebPlugin::initialize):
7234         (TestWebPlugin::destroy):
7235         (TestWebPlugin::updateGeometry):
7236         (TestWebPlugin::parsePrimitive):
7237         (TestWebPlugin::parseColor):
7238         (TestWebPlugin::parseOpacity):
7239         (TestWebPlugin::initScene):
7240         (TestWebPlugin::drawScene):
7241         (TestWebPlugin::destroyScene):
7242         (TestWebPlugin::initProgram):
7243         (TestWebPlugin::initPrimitive):
7244         (TestWebPlugin::drawPrimitive):
7245         (TestWebPlugin::loadShader):
7246         (TestWebPlugin::loadProgram):
7247         * DumpRenderTree/chromium/TestWebPlugin.h: Added.
7248         (TestWebPlugin::scriptableObject):
7249         (TestWebPlugin::paint):
7250         (TestWebPlugin::updateFocus):
7251         (TestWebPlugin::updateVisibility):
7252         (TestWebPlugin::acceptsInputEvents):
7253         (TestWebPlugin::handleInputEvent):
7254         (TestWebPlugin::didReceiveResponse):
7255         (TestWebPlugin::didReceiveData):
7256         (TestWebPlugin::didFinishLoading):
7257         (TestWebPlugin::didFailLoading):
7258         (TestWebPlugin::didFinishLoadingFrameRequest):
7259         (TestWebPlugin::didFailLoadingFrameRequest):
7260         (TestWebPlugin::Scene::Scene):
7261         * DumpRenderTree/chromium/WebViewHost.cpp:
7262         (WebViewHost::createPlugin):
7263
7264 2011-10-17  Ryosuke Niwa  <rniwa@webkit.org>
7265
7266         [nrwt] Add new category for MISSING
7267         https://bugs.webkit.org/show_bug.cgi?id=69990
7268
7269         Reviewed by Dirk Pranke.
7270
7271         Treat missing results as a new category in summerize_results.
7272
7273         Because Chromium port wants to turn bots red when there are tests with missing results,
7274         extracted the logic to compute the exit code as exit_code_from_summarized_results in
7275         base and chromium ports.
7276
7277         * Scripts/webkitpy/layout_tests/controllers/manager.py:
7278         * Scripts/webkitpy/layout_tests/port/base.py:
7279         * Scripts/webkitpy/layout_tests/port/chromium.py:
7280         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
7281
7282 2011-10-17  Dimitri Glazkov  <dglazkov@chromium.org>
7283
7284         garden-o-matic operations fail because of CSP.
7285         https://bugs.webkit.org/show_bug.cgi?id=70249
7286
7287         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html: Added gardening server to the policy.
7288
7289 2011-10-17  Ada Chan  <adachan@apple.com>
7290
7291         Add a test to call WKPageLoadAlternateHTMLString() with a non-directory URL.
7292         https://bugs.webkit.org/show_bug.cgi?id=70168
7293
7294         Reviewed by Alexey Proskuryakov.
7295
7296         * TestWebKitAPI/Tests/WebKit2/LoadAlternateHTMLStringWithNonDirectoryURL.cpp: Added.
7297         (TestWebKitAPI::didFinishLoadForFrame):
7298         (TestWebKitAPI::TEST): Call WKPageLoadAlternateHTMLString() with a file URL and make sure it finishes loading.
7299
7300         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
7301         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
7302         Add the new file.
7303
7304 2011-10-17  Darin Adler  <darin@apple.com>
7305
7306         Add some more *.pyc ignores in directories with Python scripts.
7307
7308         * QueueStatusServer: Added property svn:ignore.
7309         * Scripts/webkitpy/common/watchlist: Added property svn:ignore.
7310         * Scripts/webkitpy/layout_tests/controllers: Added property svn:ignore.
7311         * Scripts/webkitpy/layout_tests/views: Added property svn:ignore.
7312
7313 2011-10-17  Alexis Menard  <alexis.menard@openbossa.org>
7314
7315         [Qt][WK2][meta] Fix failing API tests
7316         https://bugs.webkit.org/show_bug.cgi?id=70236
7317
7318         Reviewed by Csaba Osztrogonác.
7319
7320         In order to run WebKit2 API tests we need the WebProcess to be in
7321         the path so let's add the bin directory when running the API tests of Qt.
7322
7323         * Scripts/run-qtwebkit-tests:
7324
7325 2011-10-17  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
7326
7327         [Qt][WK2] Build dependency problems
7328         https://bugs.webkit.org/show_bug.cgi?id=68456
7329
7330         Reviewed by Csaba Osztrogonác.
7331
7332         * Scripts/webkitdirs.pm:
7333         (buildQMakeProject):
7334
7335 2011-10-17  Holger Hans Peter Freyther  <holger@moiji-mobile.com>
7336
7337         [build] Introduce Qt/SH4 buildbot
7338         https://bugs.webkit.org/show_bug.cgi?id=70227
7339
7340         Reviewed by Adam Roben.
7341
7342         Add buildslave for Linux/Qt/SH4.
7343
7344         * BuildSlaveSupport/build.webkit.org-config/config.json:
7345
7346 2011-10-16  Adam Barth  <abarth@webkit.org>
7347
7348         Always enable ENABLE(DOM_STORAGE)
7349         https://bugs.webkit.org/show_bug.cgi?id=70189
7350
7351         Reviewed by Eric Seidel.
7352
7353         * Scripts/build-webkit:
7354         * waf/build/settings.py:
7355
7356 2011-10-15  Adam Barth  <abarth@webkit.org>
7357
7358         Rename ENABLE(TILED_BACKING_STORE) to USE(TILED_BACKING_STORE)
7359         https://bugs.webkit.org/show_bug.cgi?id=70194
7360
7361         Reviewed by Daniel Bates.
7362
7363         * Scripts/build-webkit:
7364
7365 2011-10-15  Ojan Vafai  <ojan@chromium.org>
7366
7367         run-webkit-tests should accept --child-processes
7368         https://bugs.webkit.org/show_bug.cgi?id=69713
7369
7370         Reviewed by Adam Barth.
7371
7372         Only set --child-processes=1 if --child-processes it not
7373         passed on the commandline.
7374
7375         * Scripts/run-webkit-tests:
7376
7377 2011-10-15  Daniel Bates  <dbates@webkit.org>
7378
7379         Extract common gdb code into its own function; Remove script gdb-safari
7380         https://bugs.webkit.org/show_bug.cgi?id=68499
7381
7382         Reviewed by David Kilzer.
7383
7384         Almost identical code is used in webkitdirs::{debugMiniBrowser, debugWebKitTestRunner,
7385         runSafari}() and in the script gdb-safari to launch gdb to debug an application.
7386         We should extract the common parts into a function that can be shared by all callers
7387         so as to remove duplicate code.
7388
7389         As a side effect of removing duplicate code we can remove the script gdb-safari since
7390         its code has been incorporated into webkitdirs::debugSafari() which is called by the
7391         script debug-safari.
7392
7393         * Scripts/debug-safari: Modified to call webkitdirs::debugSafari().
7394         * Scripts/gdb-safari: Removed.
7395         * Scripts/webkitdirs.pm:
7396         (execMacWebKitAppForDebugging): Added; Extracted common debugging code from
7397         debugMiniBrowser(), debugWebKitTestRunner(), runSafari() and
7398         script gdb-safari to here.
7399         (debugSafari): Added.
7400         (runSafari): Extracted debugging code to debugSafari() and debugMacWebKitApp();
7401         Also, return the exit status of WebKit launcher (WebKit.exe) on Windows. Currently,
7402         runSafari() always returns 1 on Windows and the script run-safari exits using this
7403         value as its exit status. Hence, the script run-safari always exits with a nonzero
7404         exit status (i.e. abnormal termination) on Windows. Instead, we should return the
7405         exit status of the WebKit launcher application.
7406         (debugMiniBrowser): Modified to call debugMacWebKitApp().
7407         (debugWebKitTestRunner): Ditto.
7408
7409 2011-10-15  Robert Hogan  <robert@webkit.org>
7410
7411         [webkit-patch] Put Source/Tools changes at the top of patches, LayoutTests at the bottom
7412         https://bugs.webkit.org/show_bug.cgi?id=70056
7413
7414         Reviewed by Adam Barth.
7415
7416         The preferred file-order for patches is LayoutTests at the end and
7417         ChangeLogs before everything else in a subdirectory. Specify this in a git 'orderfile' and use
7418         it when creating patches for review and landing.
7419
7420         * Scripts/webkitpy/common/checkout/scm/git.py:
7421         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
7422         * Scripts/webkitpy/common/config/orderfile: Added.
7423
7424 2011-10-15  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
7425
7426         [Qt] [Symbian] Remove support for the Symbian platform for the QtWebKit port
7427         https://bugs.webkit.org/show_bug.cgi?id=69920
7428
7429         Reviewed by Kenneth Rohde Christiansen.
7430
7431         * DumpRenderTree/qt/DumpRenderTree.pro:
7432         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
7433         (WebCore::DumpRenderTree::open):
7434         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
7435         * QtTestBrowser/QtTestBrowser.pro:
7436         * QtTestBrowser/launcherwindow.cpp:
7437         (LauncherWindow::screenshot):
7438         (LauncherWindow::showFPS):
7439         (LauncherWindow::updateFPS):
7440         * QtTestBrowser/launcherwindow.h:
7441         (WindowOptions::WindowOptions):
7442         * QtTestBrowser/mainwindow.cpp:
7443         (MainWindow::buildUI):
7444         * Scripts/webkitpy/style/checker.py: Remove rules for the
7445         WebKit/qt/symbian/platformplugin directory.
7446         * Scripts/webkitpy/style/checker_unittest.py:
7447         * Tools.pro:
7448         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
7449         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
7450         (WTR::LayoutTestController::platformInitialize):
7451         * WebKitTestRunner/qt/TestControllerQt.cpp:
7452         (WTR::isExistingLibrary):
7453         * WebKitTestRunner/qt/WebKitTestRunner.pro:
7454
7455 2011-10-15  Adam Barth  <abarth@webkit.org>
7456
7457         Fix typo in webkitpy help
7458         https://bugs.webkit.org/show_bug.cgi?id=70169
7459
7460         Reviewed by Eric Seidel.
7461
7462         * Scripts/webkitpy/tool/commands/upload.py:
7463
7464 2011-10-15  Tony Chang  <tony@chromium.org>
7465
7466         fix another typo in treemap.html
7467         https://bugs.webkit.org/show_bug.cgi?id=70149
7468
7469         Reviewed by Ojan Vafai.
7470
7471         Fixes a bug where #treemapfocus=LayoutTests%2Fhttp didn't focus the http tests.
7472
7473         * TestResultServer/static-dashboards/treemap.html:
7474
7475 2011-10-14  Adam Barth  <abarth@webkit.org>
7476
7477         Hide status bubbles from queues which are very far behind
7478         https://bugs.webkit.org/show_bug.cgi?id=70133
7479
7480         Reviewed by Eric Seidel.
7481
7482         The mac-ews is so far behind that we shouldn't show its status bubble.
7483         The status bubble just makes contributors sad.
7484
7485         * QueueStatusServer/handlers/statusbubble.py:
7486
7487 2011-10-14  Dimitri Glazkov  <dglazkov@chromium.org>
7488
7489         Write unit test to ensure style-checker-filter option is passed properly in webkit-patch
7490         https://bugs.webkit.org/show_bug.cgi?id=70143
7491
7492         Reviewed by Adam Barth.
7493
7494         * Scripts/webkitpy/tool/commands/download_unittest.py: Added a test for land-cowboy. Not so cowboy after all.
7495         * Scripts/webkitpy/tool/steps/checkstyle.py: Fixed a bug, discovered by the newly added test.
7496
7497 2011-10-14  David Levin  <levin@chromium.org>
7498
7499         watchlist: Add a validation check for the email names in the list.
7500         https://bugs.webkit.org/show_bug.cgi?id=70154
7501
7502         Reviewed by Adam Barth.
7503
7504         If an email is listed which doesn't have a bugzilla log in, then any emails
7505         added to the bug along with that email will fail to happen, so try to prevent this
7506         from happening by validating that emails added to the watchlist are known to be
7507         contributors.
7508
7509         * Scripts/webkitpy/common/config/committers.py: Add my other email aliases for bugzilla.
7510         * Scripts/webkitpy/common/watchlist/watchlistparser.py: Add the check.
7511         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py: A unit test for the check.
7512
7513 2011-10-14  David Levin  <levin@chromium.org>
7514
7515         watchlist: If the style check fails, then the watchlist will not be run.
7516         https://bugs.webkit.org/show_bug.cgi?id=69484
7517
7518         Reviewed by Adam Barth.
7519
7520         * Scripts/webkitpy/tool/commands/queues.py: Run the watch list even
7521         if the style part fails and don't allow watch list failures turn the
7522         bot run red.
7523         * Scripts/webkitpy/tool/commands/queues_unittest.py: Appropriate unit tests.
7524         * Scripts/webkitpy/tool/mocktool.py: Add support to make an executive command throw.
7525
7526 2011-10-14  Dimitri Glazkov  <dglazkov@chromium.org>
7527
7528         Plumb style-checker filter up to command options and make land-cowboy use it.
7529         https://bugs.webkit.org/show_bug.cgi?id=70119
7530
7531         Reviewed by Adam Barth.
7532
7533         * Scripts/webkitpy/style/checker.py: Added ChangeLogChecker categories to the big list.
7534         * Scripts/webkitpy/style/checkers/changelog.py: Declared categories, used in ChangeLogChecker.
7535         * Scripts/webkitpy/tool/commands/download.py: Made LandCowboy use filter.
7536         * Scripts/webkitpy/tool/commands/download_unittest.py: Tweaked the test.
7537         * Scripts/webkitpy/tool/commands/upload_unittest.py: Ditto.
7538         * Scripts/webkitpy/tool/steps/checkstyle.py: Made check_style_filter work.
7539         * Scripts/webkitpy/tool/steps/options.py: Added new option, check-style-filter.
7540
7541 2011-10-14  Sam Weinig  <sam@webkit.org>
7542
7543         Remove the Leopard slaves.
7544
7545         Reviewed by Adam Roben.
7546
7547         * BuildSlaveSupport/build.webkit.org-config/config.json:
7548
7549 2011-10-14  Peter Beverloo  <peter@chromium.org>
7550
7551         [Chromium] Inherit settings from Chromium's envsetup.sh, address a NDK todo
7552         https://bugs.webkit.org/show_bug.cgi?id=70028
7553
7554         Reviewed by Adam Barth.
7555
7556         * Scripts/update-webkit-chromium:
7557         * Scripts/webkitdirs.pm:
7558         (buildChromiumMakefile):
7559         (buildChromium):
7560
7561 2011-10-14  Leandro Pereira  <leandro@profusion.mobi>
7562
7563         Unreviewed; add missing #includes.
7564
7565         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
7566         * DumpRenderTree/efl/EventSender.cpp:
7567
7568 2011-10-14  Raphael Kubo da Costa  <kubo@profusion.mobi>
7569
7570         [EFL] Correctly get the absolute URL in LayoutTestController::queueLoad()
7571         https://bugs.webkit.org/show_bug.cgi?id=69770
7572
7573         Reviewed by Antonio Gomes.
7574
7575         We were just appending the given URL to the absolute URI of the main
7576         frame, which resulted in "file:///foo/bar/baz.htmltheother/url.html"
7577         instead of "file:///foo/bar/theother/url.html".
7578
7579         This should make fast/dom/navigation-type-navigate.html behave
7580         better (it still fails, but later).
7581
7582         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
7583         (LayoutTestController::queueLoad):
7584
7585 2011-10-13  Arko Saha  <arko@motorola.com>
7586
7587         Microdata: Basic implementation of document.getItems() method.
7588         https://bugs.webkit.org/show_bug.cgi?id=68610
7589
7590         Reviewed by Ryosuke Niwa.
7591
7592         Added ENABLE(MICRODATA) feature flag. Implement document.getItems() DOM API.
7593         Spec : http://www.whatwg.org/specs/web-apps/current-work/complete/microdata.html
7594
7595         * Scripts/build-webkit:
7596
7597 2011-10-14  Csaba Osztrogonác  <ossy@webkit.org>
7598
7599         new-run-webkit-tests does not support qt-4.8 results
7600         https://bugs.webkit.org/show_bug.cgi?id=64071
7601
7602         Roll-back second part of r97252. :)
7603
7604         * Scripts/webkitpy/layout_tests/port/qt.py:
7605
7606 2011-10-14  Raphael Kubo da Costa  <kubo@profusion.mobi>
7607
7608         [EFL] DRT: Do not use OwnFastMallocPtr to manage char*'s.
7609         https://bugs.webkit.org/show_bug.cgi?id=70106
7610
7611         Reviewed by Antonio Gomes.
7612
7613         OwnFastMallocPtr was being used as a smart pointer that automatically called
7614         free() on the strings returned by the EFL or by ewk.
7615
7616         However, when WTF is built in release mode, it uses its own memory management
7617         code instead of using the system malloc(), free() and friends. This means bad
7618         things will happen when one uses WTF's free() on memory allocated with system
7619         malloc() by the EFL or ewk.
7620
7621         The easiest way to solve this is to call free() ourselves.
7622
7623         * DumpRenderTree/efl/DumpRenderTree.cpp:
7624         (dumpFramesAsText):
7625         (getFinalTestURL):
7626
7627 2011-10-14  Raphael Kubo da Costa  <kubo@profusion.mobi>
7628
7629         [EFL] Add DumpRenderTreeSupportEfl
7630         https://bugs.webkit.org/show_bug.cgi?id=68458
7631
7632         Reviewed by Kenneth Rohde Christiansen.
7633
7634         Make the calls in DumpRenderTree use DumpRenderTreeSupportEfl.
7635
7636         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
7637         (DumpRenderTreeChrome::initialize): Turn on mock scrollbars.
7638         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
7639         (DumpRenderTreeChrome::onWindowObjectCleared):
7640         (DumpRenderTreeChrome::onDocumentLoadFinished):
7641         * DumpRenderTree/efl/GCControllerEfl.cpp:
7642         (GCController::collect):
7643         (GCController::collectOnAlternateThread):
7644         (GCController::getJSObjectCount):
7645         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
7646         (LayoutTestController::counterValueForElementById):
7647         (LayoutTestController::pageNumberForElementById):
7648         (LayoutTestController::numberOfPages):
7649         (LayoutTestController::workerThreadCount):
7650         (LayoutTestController::findString):
7651         (LayoutTestController::pauseAnimationAtTimeOnElementWithId):
7652         (LayoutTestController::pauseTransitionAtTimeOnElementWithId):
7653         (LayoutTestController::sampleSVGAnimationForElementAtTime):
7654         (LayoutTestController::numberOfActiveAnimations):
7655         (LayoutTestController::suspendAnimations):
7656         (LayoutTestController::resumeAnimations):
7657         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
7658         (createBitmapContextFromWebView):
7659
7660 2011-10-14  Csaba Osztrogonác  <ossy@webkit.org>
7661
7662         [Qt] Enable NRWT for Qt 4.8 version too, because it is supported after r97461.
7663
7664         Unreviewed.
7665
7666         * Scripts/run-webkit-tests:
7667         (useNewRunWebKitTests):
7668
7669 2011-10-14  Csaba Osztrogonác  <ossy@webkit.org>
7670
7671         new-run-webkit-tests does not support qt-4.8 results
7672         https://bugs.webkit.org/show_bug.cgi?id=64071
7673
7674         Roll-back r97252 with unit test fix.
7675
7676         Reviewed by Csaba Osztrogonác.
7677
7678         * Scripts/webkitpy/layout_tests/port/qt.py:
7679         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
7680
7681 2011-10-13  Ojan Vafai  <ojan@chromium.org>
7682
7683         Make filter-build-webkit exit 1 if the build fails.
7684         https://bugs.webkit.org/show_bug.cgi?id=70080
7685
7686         Reviewed by Eric Seidel.
7687
7688         This only works if you redirect stderr to stdout, but it at least makes it possible.
7689         This is useful if you want to run a followup command (e.g. run-webkit-tests) but
7690         only if the build succeeds.
7691
7692         * Scripts/filter-build-webkit:
7693
7694 2011-10-13  Dimitri Glazkov  <dglazkov@chromium.org>
7695
7696         land-cowboy should at least check style.
7697         https://bugs.webkit.org/show_bug.cgi?id=70073
7698
7699         Reviewed by Adam Barth.
7700
7701         * Scripts/webkitpy/tool/commands/download.py: Added CheckStyle step to LandCowboy command.
7702
7703 2011-10-13  Tony Chang  <tony@chromium.org>
7704
7705         fix regex in layout test times dashboard
7706         https://bugs.webkit.org/show_bug.cgi?id=70059
7707
7708         This fixes the treemapfocus query param.
7709
7710         Reviewed by Ojan Vafai.
7711
7712         * TestResultServer/static-dashboards/treemap.html:
7713
7714 2011-10-13  Beth Dakin  <bdakin@apple.com>
7715
7716         https://bugs.webkit.org/show_bug.cgi?id=70050
7717         DRT and WRT should have HiDPI testing capabilities
7718
7719         Reviewed by Darin Adler.
7720
7721         New layoutTestController function setBackingScaleFactor takes a double for a scale
7722         factor and a callback function to call once the backing scale factor has
7723         successfully been set for the view.
7724         * DumpRenderTree/LayoutTestController.cpp:
7725         (setBackingScaleFactorCallback):
7726         (LayoutTestController::staticFunctions):
7727
7728         On Mac, setBackingScaleFactor calls WebView _setCustomBackingScaleFactor. Other
7729         platforms do nothing.
7730         * DumpRenderTree/LayoutTestController.h:
7731         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
7732         (LayoutTestController::setBackingScaleFactor):
7733         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
7734         (LayoutTestController::setBackingScaleFactor):
7735         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
7736         (LayoutTestController::setBackingScaleFactor):
7737
7738         Reset the backing scale factor by calling _setCutsomBackingScaleFactor with a
7739         value of 0, which resets to the system default.
7740         * DumpRenderTree/mac/DumpRenderTree.mm:
7741         (resetWebViewToConsistentStateBeforeTesting):
7742
7743         When generating the pixel dump, create a Bitmap of the appropriate size based on
7744         the backing scale factor. If it is a HiDPI view, paint into the context using
7745         displayRectIgnoringOpacity.
7746         * DumpRenderTree/mac/PixelDumpSupportMac.mm:
7747         (createBitmapContextFromWebView):
7748
7749         And now the same stuff for WK2. Again, new layoutTestController function
7750         setBackingScaleFactor takes a double for a scale factor and a callback function.
7751         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
7752         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
7753         (WTR::LayoutTestController::setBackingScaleFactor):
7754         (WTR::LayoutTestController::callSetBackingScaleFactorCallback):
7755         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
7756
7757         The InjectedBundle sends a message to the UI process to set the backing scale
7758         factor. It also processes a message the UI process will send back once it
7759         completes this task, and at that time, the InjectedBundle calls the callback.
7760         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
7761         (WTR::InjectedBundle::didReceiveMessage):
7762         (WTR::InjectedBundle::postSetBackingScaleFactor):
7763         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
7764
7765         Again, reset the backing scale factor by calling WKPageSetCustomBackingScaleFactor
7766         with a value of 0, which resets to the system default.
7767         * WebKitTestRunner/TestController.cpp:
7768         (WTR::TestController::resetStateToConsistentValues):
7769
7770         Upon receiving the SetBackingScaleFactor message, call
7771         WKPageSetCustomBackingScaleFactor to the given scale factor, and post a message
7772         back to InjectedBundle to say it has been done.
7773         * WebKitTestRunner/TestInvocation.cpp:
7774         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
7775
7776         Don't use window snapshotting for the HiDPI tests. We don't have a way to fake a
7777         window's scale factor so we have to avoid that.
7778         * WebKitTestRunner/cg/TestInvocationCG.cpp:
7779         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
7780
7781 2011-10-13  John Knottenbelt  <jknotten@chromium.org>
7782
7783         [Chromium] Allow building without speech input enabled.
7784         https://bugs.webkit.org/show_bug.cgi?id=69867
7785
7786         Reviewed by Darin Fisher.
7787
7788         * DumpRenderTree/chromium/LayoutTestController.cpp:
7789         (LayoutTestController::addMockSpeechInputResult):
7790
7791 2011-10-13  Kent Tamura  <tkent@chromium.org>
7792
7793         [Chromium] A DRT fix for r97348.
7794
7795         * DumpRenderTree/chromium/LayoutTestController.cpp:
7796         (LayoutTestController::overridePreference):
7797         Ignore overridePreference() for "WebKitWebAudioEnabled", and don't
7798         print an error message.
7799
7800 2011-10-12  Lucas Forschler  <lforschler@apple.com>
7801
7802         Update kill-old-processes logic.
7803         https://bugs.webkit.org/show_bug.cgi?id=63651
7804
7805         Reviewed by Ryosuke Niwa.
7806
7807         Remove unused mac platform specific folder.
7808         Update master.config to use new kill-old-processes.
7809
7810         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
7811         * BuildSlaveSupport/kill-old-processes:
7812         * BuildSlaveSupport/mac: Removed.
7813
7814 2011-10-12  Ryosuke Niwa  <rniwa@webkit.org>
7815
7816         [NRWT] New tests without expected files are reported as flakey tests
7817         https://bugs.webkit.org/show_bug.cgi?id=67268
7818
7819         Reviewed by Dirk Pranke.
7820
7821         The bug was caused by summarize_results's regarding all tests that have an entry in result_summary
7822         and no entry in retry_summary as flaky. Since we don't retry missing tests, there will be no entries
7823         for missing results in retry_summary.
7824
7825         Fixed the bug by treating missing results separately.
7826
7827         * Scripts/webkitpy/layout_tests/controllers/manager.py:
7828         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
7829
7830 2011-10-12  Adam Barth  <abarth@webkit.org>
7831
7832         Remove ENABLE(XHTMLMP) and associated code
7833         https://bugs.webkit.org/show_bug.cgi?id=69729
7834
7835         Reviewed by David Levin.
7836
7837         * Scripts/build-webkit:
7838         * Scripts/old-run-webkit-tests:
7839         * Scripts/webkitperl/features.pm:
7840         * Scripts/webkitpy/layout_tests/port/webkit.py:
7841         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
7842
7843 2011-10-12  Ryosuke Niwa  <rniwa@webkit.org>
7844
7845         Kill old run-webkit-tests processes in kill-process on Mac
7846         https://bugs.webkit.org/show_bug.cgi?id=69976
7847
7848         Reviewed by Dirk Pranke.
7849
7850         Add a system call to remove all python processes running run-webkit-tests.
7851
7852         * BuildSlaveSupport/kill-old-processes:
7853         * BuildSlaveSupport/mac: Removed.
7854
7855 2011-10-12  Simon Fraser  <simon.fraser@apple.com>
7856
7857         Many WK2 snapshots are blank
7858         https://bugs.webkit.org/show_bug.cgi?id=69953
7859
7860         Reviewed by Beth Dakin.
7861
7862         Two fixes:
7863         1. negating a size_t in CGContextTranslateCTM() fails, because
7864         size_t is unsigned. Fix by casing to CGFloat before negation.
7865
7866         2. Turn on window autodisplay, so that compositing tests snapshot correctly.
7867         Autodisplay is necessary to kick off Core Animation rendering.
7868
7869         * WebKitTestRunner/cg/TestInvocationCG.cpp:
7870         (WTR::createCGContextFromImage):
7871         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
7872         (WTR::PlatformWebView::PlatformWebView):
7873
7874 2011-10-12  Adam Barth  <abarth@webkit.org>
7875
7876         commit-queue doesn't have a friendly error message when the reviewer line is messed up
7877         https://bugs.webkit.org/show_bug.cgi?id=69979
7878
7879         Reviewed by Eric Seidel.
7880
7881         Rather than combining the ChangeLog validation with a more complicated
7882         command, this patch has the commit-queue run it as a separate command,
7883         which will give us more control over the error message.
7884
7885         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
7886         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
7887         * Scripts/webkitpy/tool/commands/download.py:
7888         * Scripts/webkitpy/tool/commands/queues_unittest.py:
7889         * Scripts/webkitpy/tool/steps/validatechangelogs.py:
7890         * Scripts/webkitpy/tool/steps/validatereviewer.py:
7891
7892 2011-10-12  Eric Seidel  <eric@webkit.org>
7893
7894         Layout tests asserting in LayoutTestController::pathToLocalResource()
7895         https://bugs.webkit.org/show_bug.cgi?id=69898
7896
7897         Reviewed by Simon Fraser.
7898
7899         old-run-webkit-tests just didn't know to set LOCAL_RESOURCE_ROOT, oops.
7900
7901         * Scripts/old-run-webkit-tests:
7902         (openDumpTool):
7903
7904 2011-10-12  Adam Barth  <abarth@webkit.org>
7905
7906         The commit-queue doesn't tell me which tests my patch broke
7907         https://bugs.webkit.org/show_bug.cgi?id=69975
7908
7909         Reviewed by Eric Seidel.
7910
7911         We now print the list of newly failing tests, like we do for the EWS.
7912
7913         * Scripts/webkitpy/tool/commands/queues.py:
7914         * Scripts/webkitpy/tool/commands/queues_unittest.py:
7915
7916 2011-10-10  Alice Boxhall  <aboxhall@chromium.org>
7917
7918         Added myself as a committer.
7919
7920         * Scripts/webkitpy/common/config/committers.py:
7921
7922 2011-10-12  Adam Barth  <abarth@webkit.org>
7923
7924         _run_tests is a confusing variable name in webkitpy
7925         https://bugs.webkit.org/show_bug.cgi?id=69971
7926
7927         Reviewed by Eric Seidel.
7928
7929         Eric and I thought this was a function.
7930
7931         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py:
7932
7933 2011-10-12  Tony Chang  <tony@chromium.org>
7934
7935         Fix test-webkitpy after r97307 and r97293.
7936
7937         * Scripts/webkitpy/layout_tests/port/chromium.py: options.time_out_ms
7938           is None during tests so add back the check.
7939         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py: The
7940           output directory is now build/Debug.
7941
7942 2011-10-12  Tony Chang  <tony@chromium.org>
7943
7944         [chromium] remove old build dirs on chromium-win
7945         https://bugs.webkit.org/show_bug.cgi?id=69961
7946
7947         Reviewed by Adam Barth.
7948
7949         * Scripts/webkitpy/layout_tests/port/chromium_win.py: Output is now
7950           always in build and never in chrome or webkit.
7951
7952 2011-10-12  Dimitri Glazkov  <dglazkov@chromium.org>
7953
7954         garden-o-matic's "Expect Failure" button does not work correctly, so we should remove it until the plumbing is fixed to avoid landmines.
7955         https://bugs.webkit.org/show_bug.cgi?id=69954
7956
7957         Also renamed the action and associated event name to address an FIXME.
7958
7959         Reviewed by Adam Barth.
7960
7961         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js: Renamed event name.
7962         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js: Renamed the action.
7963         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js: Removed action temporarily.
7964
7965 2011-10-12  Tony Chang  <tony@chromium.org>
7966
7967         stop spamming 'stopping test driver timed out, killing it' when running rwt
7968         https://bugs.webkit.org/show_bug.cgi?id=69938
7969
7970         Reviewed by Adam Barth.
7971
7972         In http://trac.webkit.org/changeset/95875 , we made the timeout for
7973         stopping the driver based on time_out_ms, but we based it on the wrong
7974         default times.
7975
7976         * Scripts/webkitpy/layout_tests/port/chromium.py:
7977
7978 2011-10-12  Simon Fraser  <simon.fraser@apple.com>
7979
7980         3D transforms are flattened in WebKit2 snapshots
7981         https://bugs.webkit.org/show_bug.cgi?id=68276
7982
7983         Reviewed by Adam Roben.
7984
7985         Fix WKImage leak noticed by Adam.
7986
7987         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
7988         (WTR::PlatformWebView::windowSnapshotImage):
7989
7990 2011-10-12  Dimitri Glazkov  <dglazkov@chromium.org>
7991
7992         garden-o-matic should color-code test failure types.
7993         https://bugs.webkit.org/show_bug.cgi?id=69945
7994
7995         Reviewed by Adam Barth.
7996
7997         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js: Added an extra span for styling.
7998         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js: Updated tests.
7999         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Ditto.
8000         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css: Added color-coding.
8001
8002 2011-10-12  Adam Barth  <abarth@webkit.org>
8003
8004         webkitpy tests are failing on Mac bots
8005         https://bugs.webkit.org/show_bug.cgi?id=69929
8006
8007         Reviewed by Simon Fraser.
8008
8009         Admit that Lion exists.
8010
8011         * Scripts/webkitpy/layout_tests/port/chromium.py:
8012         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
8013
8014 2011-10-12  Simon Fraser  <simon.fraser@apple.com>
8015
8016         Improve output of run-api-tests to make time-outs more obvious
8017         https://bugs.webkit.org/show_bug.cgi?id=69788
8018
8019         Reviewed by Adam Roben.
8020
8021         List the tests that failed and timed out at the bottom of the output,
8022         if run with --verbose.
8023
8024         * Scripts/run-api-tests:
8025         (runAllTests):
8026         (runTest):
8027
8028 2011-10-12  Sheriff Bot  <webkit.review.bot@gmail.com>
8029
8030         Unreviewed, rolling out r97252.
8031         http://trac.webkit.org/changeset/97252
8032         https://bugs.webkit.org/show_bug.cgi?id=69944
8033
8034         Broke test-webkpy (Requested by abarth on #webkit).
8035
8036         * Scripts/webkitpy/layout_tests/port/qt.py:
8037         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
8038
8039 2011-10-12  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
8040
8041         new-run-webkit-tests does not support qt-4.8 results
8042         https://bugs.webkit.org/show_bug.cgi?id=64071
8043
8044         Reviewed by Csaba Osztrogonác.
8045
8046         * Scripts/webkitpy/layout_tests/port/qt.py:
8047         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
8048
8049 2011-10-10  Martin Robinson  <mrobinson@igalia.com>
8050
8051         [GTK] [WebKit2] Make adding another unit test easier
8052         https://bugs.webkit.org/show_bug.cgi?id=69409
8053
8054         Reviewed by Gustavo Noronha Silva.
8055
8056         * Scripts/run-gtk-tests: Update run-gtk-tests to be able to execute
8057         new WebKit2 tests.
8058
8059 2011-10-11  Takashi Toyoshima  <toyoshim@chromium.org>
8060
8061         [WebSocket] update pywebsocket to 0.6b6.
8062         https://bugs.webkit.org/show_bug.cgi?id=69824
8063
8064         Reviewed by Kent Tamura.
8065
8066         WebKit supports WebSocket with version 8 based protocol for now.
8067         We should update its protocol to version 13 based one like hybi-17.
8068         pywebsocket 0.6b6 supports both of version 8 and 13.
8069
8070         * Scripts/webkitpy/thirdparty/__init__.py:
8071
8072 2011-10-11  Simon Fraser  <simon.fraser@apple.com>
8073
8074         WebKitTestRunner needs to link with WebKit for NSURL category methods
8075         https://bugs.webkit.org/show_bug.cgi?id=69892
8076
8077         Reviewed by Sam Weinig.
8078
8079         WebDragClient::declareAndWriteDragImage() was throwing an Obj-C exception on 10.6 because
8080         the +_web_originalDataAsString selector on NSURL was not recognized. Link with
8081         WebKit.framework to fix this.
8082
8083         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
8084
8085 2011-10-11  Adam Barth  <abarth@webkit.org>
8086
8087         sheriffbot takes too long to acknowledge rollout commands
8088         https://bugs.webkit.org/show_bug.cgi?id=69871
8089
8090         Reviewed by Eric Seidel.
8091
8092         We used to update the working copy before acknowledging the command
8093         because we wanted to ping all the relevant IRC nicks.  That's caused a
8094         bunch of frustration because folks don't know whether the bot has heard
8095         their commands.
8096
8097         This patch makes the bot reply immediately before updating the working
8098         copy.  All the relevenat folks are still pinged when the bot finishes
8099         preparing the rollout.
8100
8101         * Scripts/webkitpy/tool/bot/irc_command.py:
8102         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
8103
8104 2011-10-11  Adam Barth  <abarth@webkit.org>
8105
8106         test-webkitpy fails on Lion
8107         https://bugs.webkit.org/show_bug.cgi?id=69873
8108
8109         Reviewed by Eric Seidel.
8110
8111         This patch fixes one of the two failures.  I'll need to track down the
8112         other failure another time.
8113
8114         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
8115
8116 2011-10-11  Adam Barth  <abarth@webkit.org>
8117
8118         garden-o-matic should warn me when run-webkit-tests errors out (e.g., due to python screw-ups)
8119         https://bugs.webkit.org/show_bug.cgi?id=65230
8120
8121         Reviewed by Dimitri Glazkov.
8122
8123         This patch causes us to treat crashed or hung run-webkit-tests runs
8124         similarly to how we treat compile failures.  In some sense, they're
8125         similar because they prevent us from getting test coverage.
8126
8127         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js:
8128         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js:
8129
8130 2011-10-11  Adam Barth  <abarth@webkit.org>
8131
8132         Failing builder display in garden-o-matic should have the same visual style as the failure grid
8133         https://bugs.webkit.org/show_bug.cgi?id=69876
8134
8135         Reviewed by Dimitri Glazkov.
8136
8137         This patch generalizes the view we use for failing builders in the
8138         FailureGrid so that we can use it for the list of builders that aren't
8139         compiling.
8140
8141         One downside of this patch is that it makes it harder to tell when
8142         debug builders are failing (as opposed to release builders) because
8143         that information isn't presented in this view.  A future patch will
8144         improve this aspect.
8145
8146         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js:
8147         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
8148         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
8149         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
8150         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/failures.css:
8151         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css:
8152
8153 2011-10-11  Filip Pizlo  <fpizlo@apple.com>
8154
8155         Some API tests are timing out in Release
8156         https://bugs.webkit.org/show_bug.cgi?id=69787
8157
8158         Reviewed by Mark Rowe.
8159
8160         NDEBUG must be set in release builds. Otherwise the world breaks. Also added some
8161         magic to ensure that GTest builds in release mode in clang with NDEBUG set.
8162
8163         * TestWebKitAPI/Configurations/Base.xcconfig:
8164         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
8165         * TestWebKitAPI/config.h:
8166
8167 2011-10-11  Adam Barth  <abarth@webkit.org>
8168
8169         garden-o-matic links to flakiness dashboard should open in a new tab
8170         https://bugs.webkit.org/show_bug.cgi?id=69864
8171
8172         Reviewed by Dimitri Glazkov.
8173
8174         This bug was obvious in the first five minutes of using this new
8175         feature.  :)
8176
8177         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
8178         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
8179
8180 2011-10-11  Adam Barth  <abarth@webkit.org>
8181
8182         Tweak the Content-Security-Policy for garden-o-matic.  It turns out
8183         that * doesn't really mean "all".  I've emailed the W3C working group
8184         about changing the spec.
8185
8186         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
8187
8188 2011-10-11  Adam Barth  <abarth@webkit.org>
8189
8190         garden-o-matic should link to the flakiness dashboard
8191         https://bugs.webkit.org/show_bug.cgi?id=69862
8192
8193         Reviewed by Dimitri Glazkov.
8194
8195         The flakiness dashboard has lots of great information about the failure
8196         history of each test.  This patch adds a link from garden-o-matic to
8197         the flakiness dashboard so the gardener can look at this information.
8198
8199         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
8200         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
8201         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
8202
8203 2011-10-11  David Levin  <levin@chromium.org>
8204
8205         watchlist: assertRaisesRegexp implementation should be compatible with python 2.7.
8206         https://bugs.webkit.org/show_bug.cgi?id=69789
8207
8208         Reviewed by Eric Seidel.
8209
8210         Made the signature of _assertRaisesRegexp match that of Python 2.7.
8211
8212         * Scripts/webkitpy/common/watchlist/watchlistloader_unittest.py: Adjusted to be compatible with Python 2.7.
8213         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py: Ditto.
8214         * Scripts/webkitpy/common/webkitunittest.py: Ditto.
8215         * Scripts/webkitpy/tool/commands/applywatchlistlocal_unittest.py: Ditto.
8216
8217 2011-10-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
8218
8219         [EFL] DRT: Do not abort() when an unknown key is pressed.
8220         https://bugs.webkit.org/show_bug.cgi?id=69856
8221
8222         Reviewed by Antonio Gomes.
8223
8224         A lot of tests crash due to the previous behaviour. Just using the
8225         received name of the pressed key works most of the time, and when it
8226         does not the failing tests should tell us.
8227
8228         * DumpRenderTree/efl/EventSender.cpp:
8229         (keyPadNameFromJSValue):
8230         (keyNameFromJSValue):
8231         (keyDownCallback):
8232
8233 2011-10-11  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
8234
8235         [Qt] [WK2] NRWT failed to launch MiniBrowser after test run
8236         https://bugs.webkit.org/show_bug.cgi?id=68016
8237
8238         Reviewed by Chang Shu.
8239
8240         * Scripts/webkitpy/layout_tests/port/qt.py:
8241
8242 2011-10-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
8243
8244         [EFL] DRT: Return the right window count in LayoutTestController.
8245         https://bugs.webkit.org/show_bug.cgi?id=69764
8246
8247         Reviewed by Antonio Gomes.
8248
8249         Follow-up to r96943: now that we have DumpRenderTreeChrome and
8250         DumpRenderTreeView in the tree, we can correctly count the number of
8251         open windows in DRT.
8252
8253         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
8254         (LayoutTestController::windowCount):
8255
8256 2011-10-11  Jer Noble  <jer.noble@apple.com>
8257
8258         REGRESSION (r96770-r96777): fast/dom/Window/window-properties.html, fast/dom/prototype-inheritance-2.html, fast/js/global-constructors.html failing on SnowLeopard Intel Release (Tests)
8259         https://bugs.webkit.org/show_bug.cgi?id=69800
8260
8261         Reviewed by Csaba Osztrogonác.
8262
8263         Enable WEB_AUDIO at runtime.
8264
8265         * DumpRenderTree/mac/DumpRenderTree.mm:
8266         (resetDefaultsToConsistentValues):
8267
8268 2011-10-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
8269
8270         [EFL] DRT: Make sure IconDatabase is closed before trying to open it again.
8271         https://bugs.webkit.org/show_bug.cgi?id=69769
8272
8273         Reviewed by Antonio Gomes.
8274
8275         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
8276         (LayoutTestController::setIconDatabaseEnabled):
8277
8278 2011-10-11  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
8279
8280         [Qt] Remove all references to QTDIR_build and standalone_package
8281
8282         Qt is now modularized, which means we no longer import WebKit into
8283         the Qt source tree. Instead we use git submodules, and building
8284         QtWebKit as "part of Qt" is really building QtWebKit as from trunk.
8285
8286         To decrease the number of buildsystem configurations we also remove
8287         the standalone_package code-path used when we were providing tarballs
8288         with the derived sources pre-generated.
8289
8290         Reviewed by Simon Hausmann.
8291
8292         * MiniBrowser/qt/MiniBrowser.pro:
8293         * QtTestBrowser/QtTestBrowser.pro:
8294         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
8295
8296 2011-10-11  Dominic Mazzoni  <dmazzoni@google.com>
8297
8298         Update Chromium accessibility support in DRT.
8299         https://bugs.webkit.org/show_bug.cgi?id=69676
8300
8301         Reviewed by Chris Fleizach.
8302
8303         Adds several missing properties and methods to enable more
8304         tests to pass, and in particular adds support for notification
8305         callbacks.
8306
8307         * DumpRenderTree/chromium/AccessibilityController.cpp:
8308         (AccessibilityController::AccessibilityController):
8309         (AccessibilityController::reset):
8310         (AccessibilityController::getFocusedElement):
8311         (AccessibilityController::shouldLogAccessibilityEvents):
8312         (AccessibilityController::notificationReceived):
8313         (AccessibilityController::logAccessibilityEventsCallback):
8314         * DumpRenderTree/chromium/AccessibilityController.h:
8315         * DumpRenderTree/chromium/AccessibilityUIElement.cpp:
8316         (getHelpText):
8317         (getStringValue):
8318         (getOrientation):
8319         (AccessibilityUIElement::AccessibilityUIElement):
8320         (AccessibilityUIElement::getChildAtIndex):
8321         (AccessibilityUIElement::isEqual):
8322         (AccessibilityUIElement::notificationReceived):
8323         (AccessibilityUIElement::roleGetterCallback):
8324         (AccessibilityUIElement::titleGetterCallback):
8325         (AccessibilityUIElement::descriptionGetterCallback):
8326         (AccessibilityUIElement::helpTextGetterCallback):
8327         (AccessibilityUIElement::stringValueGetterCallback):
8328         (AccessibilityUIElement::xGetterCallback):
8329         (AccessibilityUIElement::yGetterCallback):
8330         (AccessibilityUIElement::widthGetterCallback):
8331         (AccessibilityUIElement::heightGetterCallback):
8332         (AccessibilityUIElement::intValueGetterCallback):
8333         (AccessibilityUIElement::minValueGetterCallback):
8334         (AccessibilityUIElement::maxValueGetterCallback):
8335         (AccessibilityUIElement::childrenCountGetterCallback):
8336         (AccessibilityUIElement::insertionPointLineNumberGetterCallback):
8337         (AccessibilityUIElement::selectedTextRangeGetterCallback):
8338         (AccessibilityUIElement::isEnabledGetterCallback):
8339         (AccessibilityUIElement::isRequiredGetterCallback):
8340         (AccessibilityUIElement::isFocusedGetterCallback):
8341         (AccessibilityUIElement::isFocusableGetterCallback):
8342         (AccessibilityUIElement::isSelectedGetterCallback):
8343         (AccessibilityUIElement::isSelectableGetterCallback):
8344         (AccessibilityUIElement::isMultiSelectableGetterCallback):
8345         (AccessibilityUIElement::isExpandedGetterCallback):
8346         (AccessibilityUIElement::isCheckedGetterCallback):
8347         (AccessibilityUIElement::isVisibleGetterCallback):
8348         (AccessibilityUIElement::isOffScreenGetterCallback):
8349         (AccessibilityUIElement::isCollapsedGetterCallback):
8350         (AccessibilityUIElement::hasPopupGetterCallback):
8351         (AccessibilityUIElement::isValidGetterCallback):
8352         (AccessibilityUIElement::orientationGetterCallback):
8353         (AccessibilityUIElement::showMenuCallback):
8354         (AccessibilityUIElement::pressCallback):
8355         (AccessibilityUIElement::isEqualCallback):
8356         (AccessibilityUIElement::addNotificationListenerCallback):
8357         (AccessibilityUIElement::removeNotificationListenerCallback):
8358         (AccessibilityUIElement::takeFocusCallback):
8359         (AccessibilityUIElement::fallbackCallback):
8360         (RootAccessibilityUIElement::getChildAtIndex):
8361         (AccessibilityUIElementList::getOrCreate):
8362         * DumpRenderTree/chromium/AccessibilityUIElement.h:
8363         * DumpRenderTree/chromium/CppVariant.cpp:
8364         (CppVariant::invokeDefault):
8365         * DumpRenderTree/chromium/CppVariant.h:
8366         * DumpRenderTree/chromium/LayoutTestController.cpp:
8367         (LayoutTestController::LayoutTestController):
8368         (LayoutTestController::reset):
8369         * DumpRenderTree/chromium/LayoutTestController.h:
8370         * DumpRenderTree/chromium/WebViewHost.cpp:
8371         (WebViewHost::postAccessibilityNotification):
8372
8373 2011-10-10  Adam Klein  <adamk@chromium.org>
8374
8375         Support Distributed-CompileC as an action name in filter-build-webkit
8376         https://bugs.webkit.org/show_bug.cgi?id=69783
8377
8378         Reviewed by Ojan Vafai.
8379
8380         * Scripts/filter-build-webkit:
8381
8382 2011-10-10  Simon Fraser  <simon.fraser@apple.com>
8383
8384         3D transforms are flattened in WebKit2 snapshots
8385         https://bugs.webkit.org/show_bug.cgi?id=68276
8386
8387         Reviewed by Anders Carlsson.
8388
8389         Use a snapshot of the window, rather than an image from the web process
8390         for pixel testing in WebKitTestRunner. This correctly captures compositing
8391         layers, so works for 3D transform tests.
8392
8393         * WebKitTestRunner/PlatformWebView.h: Add a windowSnapshotImage() method.
8394         * WebKitTestRunner/cg/TestInvocationCG.cpp:
8395         (WTR::createCGContextFromImage): Add a param to optionally flip the image when drawing.
8396         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected): Call windowSnapshotImage(), and if it returns an image,
8397         use that instead of the image we get from the web process.
8398         * WebKitTestRunner/gtk/PlatformWebViewGtk.cpp:
8399         (WTR::PlatformWebView::windowSnapshotImage): Stub out the new method.
8400         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
8401         (WTR::PlatformWebView::windowSnapshotImage): Ditto.
8402         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
8403         (WTR::PlatformWebView::windowSnapshotImage): Ditto.
8404         * WebKitTestRunner/win/PlatformWebViewWin.cpp:
8405         (WTR::PlatformWebView::windowSnapshotImage): Ditto.
8406
8407 2011-10-10  David Levin  <levin@chromium.org>
8408
8409         Quick fix for watchlist unit test failures when using Python 2.7+
8410         Reviewed by Adam Barth.
8411
8412         Filed bug https://bugs.webkit.org/show_bug.cgi?id=69789 about a
8413         better fix.
8414
8415         * Scripts/webkitpy/common/webkitunittest.py: Always use our
8416         version of assertRaisesRegexp.
8417
8418 2011-10-10  Sheriff Bot  <webkit.review.bot@gmail.com>
8419
8420         Unreviewed, rolling out r97082.
8421         http://trac.webkit.org/changeset/97082
8422         https://bugs.webkit.org/show_bug.cgi?id=69784
8423
8424         broke Windows build (Requested by smfr on #webkit).
8425
8426         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
8427         * WebKitTestRunner/cg/TestInvocationCG.cpp: Renamed from Tools/WebKitTestRunner/cg/TestInvocationCG.mm.
8428         (WTR::createCGContextFromImage):
8429         (WTR::computeMD5HashStringForContext):
8430         (WTR::dumpBitmap):
8431         (WTR::paintRepaintRectOverlay):
8432         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
8433         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
8434         (WTR::PlatformWebView::PlatformWebView):
8435
8436 2011-10-10  Simon Fraser  <simon.fraser@apple.com>
8437
8438         3D transforms are flattened in WebKit2 snapshots
8439         https://bugs.webkit.org/show_bug.cgi?id=68276
8440
8441         Reviewed by Anders Carlsson.
8442
8443         The WKImage passed from the WebProcess when pixel dumping is created by painting compositing layers
8444         flattened, which flattens 3D transforms. Instead, use CGWindowListCreateImage() to get a snapshot of
8445         the window when doing pixel tests.
8446
8447         Also allow the window to auto-display, so that Core Animation layers are rendered, and animate.
8448
8449         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj: TestInvocationCG became a .mm file
8450         * WebKitTestRunner/cg/TestInvocationCG.mm: Renamed from Tools/WebKitTestRunner/cg/TestInvocationCG.cpp.
8451         (WTR::createBitmapCGContext): Utility function to create a bitmap context.
8452         (WTR::computeMD5HashStringForContext):
8453         (WTR::dumpBitmap):
8454         (WTR::paintRepaintRectOverlay):
8455         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected): Use CGWindowListCreateImage() to
8456         grap a snapshot of the window, rather than relying on the WKImage which was passed from the web process.
8457         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
8458         (WTR::PlatformWebView::PlatformWebView): Remove the line that turned off autoDisplay for the NSWindow.
8459         We need the window to autodisplay for Core Animation to render layers and start animations.
8460
8461 2011-10-10  Simon Fraser  <simon.fraser@apple.com>
8462
8463         WebKitTestRunner needs layoutTestController.setWindowIsKey
8464         https://bugs.webkit.org/show_bug.cgi?id=42688
8465
8466         Reviewed by Anders Carlsson.
8467
8468         Implement layoutTestController.setWindowIsKey() in WebKitTestRunner.
8469         This sets a flag that the platform layer uses to indicate that
8470         the window is the key window, which fixes focus ring drawing
8471         in pixel results.
8472
8473         Also focus the web view when we get a 'didCommitLoad' for the main
8474         frame, as DumpRenderTree does.
8475
8476         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
8477         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
8478         (WTR::InjectedBundle::postSetWindowIsKey):
8479         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
8480         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
8481         (WTR::cacheLayoutTestControllerCallback): Replace some magic numbers with an enum.
8482         (WTR::LayoutTestController::addChromeInputField):
8483         (WTR::LayoutTestController::removeChromeInputField):
8484         (WTR::LayoutTestController::focusWebView):
8485         (WTR::LayoutTestController::setWindowIsKey):
8486         (WTR::LayoutTestController::callAddChromeInputFieldCallback):
8487         (WTR::LayoutTestController::callRemoveChromeInputFieldCallback):
8488         (WTR::LayoutTestController::callFocusWebViewCallback):
8489         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
8490         * WebKitTestRunner/PlatformWebView.h: Add an NSWindow subclass
8491         so that we can override -isKeyWindow. Use a boolean member variable
8492         to track window key status.
8493         (WTR::PlatformWebView::setWindowIsKey):
8494         (WTR::PlatformWebView::windowIsKey):
8495         * WebKitTestRunner/TestController.cpp:
8496         (WTR::TestController::runModal): Change the window key status when showing a modal alert, as DRT does.
8497         (WTR::closeOtherPage): Use the same cast as other functions.
8498         (WTR::focus): Implement to call setWindowIsKey(true)
8499         (WTR::unfocus): Implement to call setWindowIsKey(false)
8500         (WTR::TestController::createOtherPage): Add page callbacks for focus and unfocus.
8501         (WTR::TestController::initialize):
8502         (WTR::TestController::didCommitLoadForFrame): Hook up callback to focus the web view, as DRT does.
8503         * WebKitTestRunner/TestController.h:
8504         * WebKitTestRunner/TestInvocation.cpp:
8505         (WTR::TestInvocation::didReceiveSynchronousMessageFromInjectedBundle):
8506         * WebKitTestRunner/gtk/PlatformWebViewGtk.cpp:
8507         (WTR::PlatformWebView::PlatformWebView):
8508         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
8509         (-[WebKitTestRunnerWindow isKeyWindow]):
8510         (WTR::PlatformWebView::PlatformWebView): Make a WebKitTestRunnerWindow.
8511         (WTR::PlatformWebView::~PlatformWebView): Clean up the back pointer on the NSWindow subclass, in case it outlives us.
8512         (WTR::PlatformWebView::focus): Set the view as the first responder, and the window as the key window, as DRT does.
8513         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
8514         (WTR::PlatformWebView::PlatformWebView):
8515         * WebKitTestRunner/win/PlatformWebViewWin.cpp:
8516         (WTR::PlatformWebView::PlatformWebView):
8517
8518 2011-10-10  Raphael Kubo da Costa  <kubo@profusion.mobi>
8519
8520         check-webkit-style: Add readability exception for Ewk_* data types.
8521         https://bugs.webkit.org/show_bug.cgi?id=69763
8522
8523         Reviewed by David Levin.
8524
8525         Many EFL-related files (such as EditorClientEfl.h or
8526         DumpRenderTreeSupportEfl.h) need to have declarations such as "typedef
8527         struct _Ewk_Foo Ewk_Foo", which currently produce false positives in
8528         check-webkit-style.
8529
8530         Add an exception for them, just like there are exceptions for
8531         Evas_*, Ecore_* and others.
8532
8533         * Scripts/webkitpy/style/checkers/cpp.py:
8534         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
8535
8536 2011-10-10  Dominic Mazzoni  <dmazzoni@google.com>
8537
8538         [Chromium] Get rid of WebAccessibilityCache.
8539         https://bugs.webkit.org/show_bug.cgi?id=68224
8540
8541         Reviewed by Dimitri Glazkov.
8542
8543         * DumpRenderTree/chromium/AccessibilityController.cpp:
8544         (AccessibilityController::bindToJavascript):
8545
8546 2011-10-09  Adam Barth  <abarth@webkit.org>
8547
8548         Remove "near miss" XSS vulnerabilities in garden-o-matic
8549         https://bugs.webkit.org/show_bug.cgi?id=69708
8550
8551         Reviewed by David Levin.
8552
8553         Collin Jackson did a security audit of garden-o-matic.  He hasn't found
8554         any actual vulnerabilities yet, but he did identify a couple "near
8555         misses" where a slight change in garden-o-matic could lead to a
8556         vulnerability.  This patch removes those potential issues.
8557
8558         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Bugzilla.js:
8559         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
8560
8561 2011-10-07  Adam Barth  <abarth@webkit.org>
8562
8563         Add a Content-Security-Policy to garden-o-matic
8564         https://bugs.webkit.org/show_bug.cgi?id=69668
8565
8566         Reviewed by David Levin.
8567
8568         This will help mitigate cross-site script and will also help us dogfood CSP.
8569
8570         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
8571
8572 2011-10-07  Raphael Kubo da Costa  <kubo@profusion.mobi>
8573
8574         [EFL] DRT: Add DumpRenderTreeChrome.{cpp,h}
8575         https://bugs.webkit.org/show_bug.cgi?id=66380
8576
8577         Reviewed by Antonio Gomes.
8578
8579         This encapsulates EWebKit into an object that manages multiple windows
8580         and provides a cleaner interface to EFL's DumpRenderTree.
8581
8582         * DumpRenderTree/efl/DumpRenderTreeChrome.cpp: Added.
8583         (DumpRenderTreeChrome::create):
8584         (DumpRenderTreeChrome::DumpRenderTreeChrome):
8585         (DumpRenderTreeChrome::~DumpRenderTreeChrome):
8586         (DumpRenderTreeChrome::createNewWindow):
8587         (DumpRenderTreeChrome::createView):
8588         (DumpRenderTreeChrome::removeWindow):
8589         (DumpRenderTreeChrome::initialize):
8590         (DumpRenderTreeChrome::extraViews):
8591         (DumpRenderTreeChrome::mainFrame):
8592         (DumpRenderTreeChrome::mainView):
8593         (DumpRenderTreeChrome::resetDefaultsToConsistentValues):
8594         (DumpRenderTreeChrome::onWindowObjectCleared):
8595         (DumpRenderTreeChrome::onLoadStarted):
8596         (DumpRenderTreeChrome::processWork):
8597         (DumpRenderTreeChrome::onLoadFinished):
8598         (DumpRenderTreeChrome::onStatusbarTextSet):
8599         (DumpRenderTreeChrome::onTitleChanged):
8600         (DumpRenderTreeChrome::onDocumentLoadFinished):
8601         * DumpRenderTree/efl/DumpRenderTreeChrome.h: Added.
8602         * DumpRenderTree/efl/EventSender.cpp:
8603         (sendClick): Use browser->mainFrame() instead of mainFrame.
8604         (mouseDownCallback): Ditto.
8605         (mouseUpCallback): Ditto.
8606         (mouseMoveToCallback): Ditto.
8607         (mouseScrollByCallback): Ditto.
8608         (keyDownCallback): Ditto.
8609         (scalePageByCallback): Ditto.
8610         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
8611         (LayoutTestController::counterValueForElementById): Use
8612         browser->mainFrame() instead of mainFrame.
8613         (LayoutTestController::pageNumberForElementById): Ditto.
8614         (LayoutTestController::numberOfPages): Ditto.
8615         (LayoutTestController::pauseAnimationAtTimeOnElementWithId): Ditto.
8616         (LayoutTestController::pauseTransitionAtTimeOnElementWithId): Ditto.
8617         (LayoutTestController::sampleSVGAnimationForElementAtTime): Ditto.
8618         (LayoutTestController::numberOfActiveAnimations): Ditto.
8619         (LayoutTestController::suspendAnimations): Ditto.
8620         (LayoutTestController::resumeAnimations): Ditto.
8621         (LayoutTestController::setWebViewEditable): Ditto.
8622         (LayoutTestController::webHistoryItemCount): Use browser->mainView()
8623         instead of browser.
8624         (LayoutTestController::queueLoad): Ditto.
8625         (LayoutTestController::setAcceptsEditing): Ditto.
8626         (LayoutTestController::setUserStyleSheetEnabled): Ditto.
8627         (LayoutTestController::setPrivateBrowsingEnabled): Ditto.
8628         (LayoutTestController::setFrameFlatteningEnabled): Ditto.
8629         (LayoutTestController::setSpatialNavigationEnabled): Ditto.
8630         (LayoutTestController::setPluginsEnabled): Ditto.
8631         (LayoutTestController::findString): Ditto.
8632         (LayoutTestController::setDeveloperExtrasEnabled): Ditto.
8633         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
8634         (createBitmapContextFromWebView): Use browser->mainFrame() instead of
8635         mainFrame.
8636         * DumpRenderTree/efl/WorkQueueItemEfl.cpp:
8637         (LoadItem::invoke): Ditto.
8638         (LoadHTMLStringItem::invoke): Ditto.
8639         (ScriptItem::invoke): Ditto.
8640         (ReloadItem::invoke): Use browser->mainView() instead of browser.
8641
8642 2011-10-07  Simon Hausmann  <simon.hausmann@nokia.com>
8643
8644         [Qt][WK2] Touch mocking is broken with Qt 5 post refactor merge
8645         https://bugs.webkit.org/show_bug.cgi?id=69617
8646
8647         Reviewed by Andreas Kling.
8648
8649         Injecting fake touch events using qt_translateRawTouchEvent does not work anymore, as it sends
8650         the touch events to widgets only. The QML view however is a QWindow. Therefore the fake touch
8651         events have to be dispatched through QWindowSystemInterface::handleTouchEvent. As a result the
8652         events appear to be spontaenous, which required adjusting the are-we-on-a-real-touch-device
8653         logic.
8654
8655         * MiniBrowser/qt/MiniBrowserApplication.cpp:
8656         (MiniBrowserApplication::MiniBrowserApplication):
8657         (MiniBrowserApplication::notify):
8658         * MiniBrowser/qt/MiniBrowserApplication.h:
8659
8660 2011-10-06  Adam Barth  <abarth@webkit.org>
8661
8662         Port PixelZoomer to work in garden-o-matic
8663         https://bugs.webkit.org/show_bug.cgi?id=69577
8664
8665         Reviewed by David Levin.
8666
8667         This implementation of PixelZoomer is basically a fork of the
8668         implementation used by results.html.  These two apps have different
8669         requirements on where this code lives.  results.html wants to be
8670         self-contained whereas garden-o-matic wants to live in the public_html
8671         folder of the buildbot.  We could do something fancy here to share the
8672         code, but that's not likely to be of much benefit.
8673
8674         I had to change a few things about pixelzoomer to accounter for the
8675         differences in DOM between results.html and garden-o-matic, but
8676         surprisingly little.  I also moved all the code into an anonymous
8677         function, which is the style we're suing for module separation in
8678         garden-o-matic.  Finally, some jQuery-ims snuck in because I didn't
8679         want to drag along too many extra functions from results.html.
8680
8681         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
8682         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
8683         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/pixelzoomer.js: Added.
8684         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
8685         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/pixelzoomer.css: Added.
8686         (.pixel-zoom-container):
8687         (.pixel-zoom-container > *):
8688         (.pixel-zoom-container .scaled-image-container):
8689         (.scaled-image-container > img):
8690
8691 2011-10-06  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
8692
8693         [EFL] Change data type of parameter in ewk_view_viewport_attributes_get().
8694         https://bugs.webkit.org/show_bug.cgi?id=69505
8695
8696         Reviewed by Andreas Kling.
8697
8698         * EWebLauncher/main.c: Use *int* variables for width and height of viewport meta tag.
8699         (on_viewport_changed):
8700
8701 2011-10-06  Brent Fulgham  <bfulgham@webkit.org>
8702
8703         [WinCairo] Unreviewed test correction.
8704
8705         * Scripts/webkitdirs.pm:
8706         (argumentsForConfiguration): Add 'wincairo' option to the set of
8707         configuration flags used by run-javascript-core tests, so that it
8708         does not attempt to run Apple's release version.
8709
8710 2011-10-06  Brent Fulgham  <bfulgham@webkit.org>
8711
8712         [WinCairo] Unreviewed test correction.
8713
8714         * BuildSlaveSupport/test-result-archive: Add 'wincairo' to the set
8715          of known build slave types so that test-result-archive won't fail
8716          when running tests.
8717
8718 2011-10-06  Martin Robinson  <mrobinson@igalia.com>
8719
8720         The GTK+ WebKit2 headers produce a lot of style warnings
8721         https://bugs.webkit.org/show_bug.cgi?id=69481
8722
8723         Reviewed by David Levin.
8724
8725         Prevent emitting so many style warnings for GTK+ API. We skip header
8726         files in the WebKit2 GTK+ API directory and also avoid warnings about
8727         identifier names that begin with "webkit_" in files that contain the
8728         string "gtk".
8729
8730         * Scripts/webkitpy/style/checker.py: Do not check header files in
8731         Source/WebKit2/UIProcess/API/gtk that do not end in Private.h. This required
8732         adding the ability to specify a regular expression in the skip list. Remove
8733         a few files from the skipped list that no longer exist.
8734         * Scripts/webkitpy/style/checker_unittest.py: Added a test for this behavior.
8735         * Scripts/webkitpy/style/checkers/cpp.py: If a path contains "gtk" don't warn
8736         about identifiers that begin with "webkit_".
8737         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added a test for this behavior.
8738
8739 2011-10-06  Brent Fulgham  <bfulgham@webkit.org>
8740
8741         [WinCairo] Correct config.json for WinCairo Test builds.
8742         https://bugs.webkit.org/show_bug.cgi?id=69272
8743
8744         Reviewed by Ryosuke Niwa.
8745
8746         * BuildSlaveSupport/build.webkit.org-config/config.json: Remove
8747          old 'WinCairo Debug (Build)' label and replace with correct
8748          'WinCairo Release' label.
8749
8750 2011-10-06  David Levin  <levin@chromium.org>
8751
8752         Add a style checker watchlist definition and add myself to it.
8753         https://bugs.webkit.org/show_bug.cgi?id=69564
8754
8755         Reviewed by Daniel Bates.
8756
8757         * Scripts/webkitpy/common/config/watchlist:
8758
8759 2011-10-06  Raphael Kubo da Costa  <kubo@profusion.mobi>
8760
8761         [EFL] DRT: Create icon database path in LayoutTestController::setIconDatabaseEnabled.
8762         https://bugs.webkit.org/show_bug.cgi?id=69450
8763
8764         Reviewed by Ryosuke Niwa.
8765
8766         This should avoid stderr messages when running
8767         webarchive/test-link-rel-icon.html.
8768
8769         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
8770         (LayoutTestController::setIconDatabaseEnabled):
8771
8772 2011-10-06  Alejandro G. Castro  <alex@igalia.com>
8773
8774         [WK2] [GTK] Add Skipped file for Webkit2 test runner
8775         https://bugs.webkit.org/show_bug.cgi?id=69517
8776
8777         Use the shared WK2 skipped file also in GTK port.
8778
8779         Reviewed by Martin Robinson.
8780
8781         * Scripts/old-run-webkit-tests:
8782         (readSkippedFiles):
8783
8784 2011-10-06  Elliot Poger  <epoger@google.com>
8785
8786         allow new-run-webkit-tests to run on Lion
8787         https://bugs.webkit.org/show_bug.cgi?id=69429
8788
8789         Add fallback lists for chromium-mac-lion, and add chromium-mac-snowleopard
8790         directory to fallback lists (even though that directory doesn't exist yet)
8791
8792         Reviewed by Adam Barth.
8793
8794         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
8795         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
8796
8797 2011-10-03  Hans Wennborg  <hans@chromium.org>
8798
8799         IndexedDB: Remove SQLite-LevelDB migration code
8800         https://bugs.webkit.org/show_bug.cgi?id=69264
8801
8802         Reviewed by Tony Chang.
8803
8804         Remove setOverrideIndexedDBBackingStore. It was used in the layout
8805         test for migration.
8806
8807         * DumpRenderTree/chromium/LayoutTestController.cpp:
8808         (LayoutTestController::LayoutTestController):
8809         * DumpRenderTree/chromium/LayoutTestController.h:
8810
8811 2011-10-05  James Robinson  <jamesr@chromium.org>
8812
8813         [chromium] Add a command line option to DumpRenderTree and run_webkit_tests.py to enable threaded compositing mode
8814         https://bugs.webkit.org/show_bug.cgi?id=69498
8815
8816         Reviewed by Adam Barth.
8817
8818         * DumpRenderTree/chromium/DumpRenderTree.cpp:
8819         (main):
8820         * DumpRenderTree/chromium/TestShell.cpp:
8821         (TestShell::TestShell):
8822         (TestShell::resetWebSettings):
8823         * DumpRenderTree/chromium/TestShell.h:
8824         (TestShell::setThreadedCompositingEnabled):
8825         * DumpRenderTree/chromium/WebPreferences.cpp:
8826         (WebPreferences::reset):
8827         (WebPreferences::applyTo):
8828         * DumpRenderTree/chromium/WebPreferences.h:
8829         * Scripts/webkitpy/layout_tests/port/chromium.py:
8830         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
8831
8832 2011-10-05  Erik Arvidsson  <arv@chromium.org>
8833
8834         [QT] Fix DRT after r96779
8835
8836         unreviewed.
8837
8838         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
8839         (WebCore::dumpHistoryItem):
8840
8841 2011-10-05  Erik Arvidsson  <arv@chromium.org>
8842
8843         window.location.href and others needlessly decodes URI-encoded characters
8844         https://bugs.webkit.org/show_bug.cgi?id=30225
8845
8846         Reviewed by Darin Adler.
8847
8848         * DumpRenderTree/chromium/TestShell.cpp:
8849         (normalizeLayoutTestURLInternal): Remove extra encode.
8850
8851 2011-10-05  Darin Adler  <darin@apple.com>
8852
8853         Fix build.
8854
8855         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest.cpp:
8856         (TestWebKitAPI::TEST): Remove underscore prefix from call to
8857         WKContextRegisterURLSchemeAsEmptyDocument.
8858
8859 2011-10-05  Brent Fulgham  <bfulgham@webkit.org>
8860
8861         [WinCairo] Switch to building Release target, and activate tests.
8862         https://bugs.webkit.org/show_bug.cgi?id=69272
8863
8864         Reviewed by Adam Barth.
8865
8866         * BuildSlaveSupport/build.webkit.org-config/config.json:
8867         Change from debug to release target. Switch from "Build" to
8868         "BuildAndTest"
8869
8870 2011-10-05  Simon Fraser  <simon.fraser@apple.com>
8871
8872         In WebKitTestRunner, text has font smoothing in pixel snapshots
8873         https://bugs.webkit.org/show_bug.cgi?id=69396
8874
8875         Reviewed by Darin Adler.
8876
8877         Call the new WKContext method that disables font smoothing in
8878         WebKitTestRunner, so that pixel snapshots don't have font smoothing
8879         enabled. Remove leading underscore from a WKContext function call.
8880
8881         * WebKitTestRunner/TestController.cpp:
8882         (WTR::TestController::resetStateToConsistentValues):
8883
8884 2011-10-05  David Levin  <levin@chromium.org>
8885
8886         Improve the watchlist for threading entries.
8887         https://bugs.webkit.org/show_bug.cgi?id=69463
8888
8889         Reviewed by Adam Barth.
8890
8891         * Scripts/webkitpy/common/config/watchlist:
8892
8893 2011-10-05  Adam Roben  <aroben@apple.com>
8894
8895         Add tests for using RetainPtrs inside HashMap and HashSet
8896
8897         Fixes <http://webkit.org/b/69414> <rdar://problem/10236833> Using RetainPtr as the key type
8898         in HashMap/HashSet fails to compile
8899
8900         Reviewed by John Sullivan.
8901
8902         * TestWebKitAPI/Tests/WTF/cf/RetainPtrHashing.cpp: Added.
8903         (TestWebKitAPI::TEST): Show that RetainPtr can be used inside HashSet and as both the key
8904         and value type of HashMap.
8905
8906         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
8907         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
8908         Added new file.
8909
8910 2011-10-05  Adam Barth  <abarth@webkit.org>
8911
8912         Mark as Expected in garden-o-matic is confusing
8913         https://bugs.webkit.org/show_bug.cgi?id=69452
8914
8915         Reviewed by David Levin.
8916
8917         krit didn't understand the difference between this button and the
8918         Rebaseline button because both of them seem to indicate that the new
8919         behavior is expected.  This patch renames "Mark as Expected" to "Expect
8920         Failure" to make it clear that we're just expecting the test to fail
8921         rather than accepting the new results as passing.  We might need to
8922         iterate on these names a bit more if folks continue to find them
8923         confusing.
8924
8925         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
8926
8927 2011-10-05  Adam Barth  <abarth@webkit.org>
8928
8929         Add watchlist entries for abarth
8930         https://bugs.webkit.org/show_bug.cgi?id=69461
8931
8932         Reviewed by David Levin.
8933
8934         I might have gone a bit overboard.  We'll see if this is too much spam
8935         for me.
8936
8937         * Scripts/webkitpy/common/config/watchlist:
8938
8939 2011-10-05  Brent Fulgham  <bfulgham@webkit.org>
8940
8941         [WinCairo] Teach old-run-webkit-tests about wincairo.
8942         https://bugs.webkit.org/show_bug.cgi?id=69280
8943
8944         Reviewed by Adam Roben.
8945
8946         * Scripts/old-run-webkit-tests: Handle wincairo case.
8947         (buildPlatformResultHierarchy):
8948         (readSkippedFiles): Ditto
8949         * Scripts/webkitdirs.pm: Deal with wincairo build case.
8950         (determineConfiguration):
8951
8952 2011-10-05  Chang Shu  <cshu@webkit.org>
8953
8954         [WK2] Support setMarkedText/hasMarkedText/unmarkText/insertText for WTR TextInputController
8955         https://bugs.webkit.org/show_bug.cgi?id=68924
8956
8957         Added TextInputController files and JS interfaces for the above functions.
8958
8959         Reviewed by Darin Adler.
8960
8961         * WebKitTestRunner/DerivedSources.make:
8962         * WebKitTestRunner/DerivedSources.pro:
8963         * WebKitTestRunner/GNUmakefile.am:
8964         * WebKitTestRunner/InjectedBundle/Bindings/TextInputController.idl: Added.
8965         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
8966         (WTR::InjectedBundle::beginTesting):
8967         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
8968         (WTR::InjectedBundle::textInputController):
8969         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
8970         (WTR::InjectedBundlePage::didClearWindowForFrame):
8971         * WebKitTestRunner/InjectedBundle/TextInputController.cpp: Added.
8972         (WTR::TextInputController::create):
8973         (WTR::TextInputController::TextInputController):
8974         (WTR::TextInputController::~TextInputController):
8975         (WTR::TextInputController::wrapperClass):
8976         (WTR::TextInputController::makeWindowObject):
8977         (WTR::TextInputController::setMarkedText):
8978         (WTR::TextInputController::hasMarkedText):
8979         (WTR::TextInputController::unmarkText):
8980         (WTR::TextInputController::insertText):
8981         * WebKitTestRunner/InjectedBundle/TextInputController.h: Added.
8982         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
8983         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
8984         * WebKitTestRunner/win/InjectedBundle.vcproj:
8985
8986 2011-10-04  Oliver Hunt  <oliver@apple.com>
8987
8988         Add rudimentary filtering to write barriers
8989         https://bugs.webkit.org/show_bug.cgi?id=69392
8990
8991         Reviewed by Filip Pizlo.
8992
8993         When we're recording gc phase times Heap.o picks up
8994         some exit time destructors, so we'll just ignore Heap.o
8995         in this check.
8996
8997         * Scripts/check-for-exit-time-destructors:
8998
8999 2011-10-05  Chang Shu  <cshu@webkit.org>
9000
9001         [WK2] WebKitTestRunner needs LayoutTestController.dumpConfigurationForViewport
9002         https://bugs.webkit.org/show_bug.cgi?id=69365
9003
9004         Reviewed by Darin Adler.
9005
9006         Added js API in WTR LayoutTestController.
9007
9008         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
9009         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
9010         (WTR::LayoutTestController::dumpConfigurationForViewport):
9011         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
9012
9013 2011-10-05  Raphael Kubo da Costa  <kubo@profusion.mobi>
9014
9015         [EFL] ImageDiff: Do not clear the image before it is processed and compared
9016         https://bugs.webkit.org/show_bug.cgi?id=69378
9017
9018         Reviewed by Antonio Gomes.
9019
9020         Commit 61836 to Evas added a few cleanup calls to some
9021         evas_object_image functions which effectively made ImageDiff consider
9022         all images to be equal.
9023
9024         It turns out that now evas_object_image_colorspace_set clears any
9025         data set, so the call to evas_object_image_data_set was not having any
9026         effect.
9027
9028         * DumpRenderTree/efl/ImageDiff.cpp:
9029         (readImageFromStdin): Call evas_object_image_data_set after
9030         evas_object_image_colorspace_set.
9031
9032 2011-10-05  Raphael Kubo da Costa  <kubo@profusion.mobi>
9033
9034         [EFL] Fix Evas_Object_Image reference counts and file descriptor leakage in ImageDiff.
9035         https://bugs.webkit.org/show_bug.cgi?id=68449
9036
9037         Reviewed by Antonio Gomes.
9038
9039         evas_render() must be called to close up temporary files and unmap unused
9040         resources. evas_object_image_data_set() must be called after an
9041         evas_object_image_data_get() to balance an internal reference count.
9042
9043         * DumpRenderTree/efl/ImageDiff.cpp:
9044         (differenceImageFromDifferenceBuffer):
9045         (calculateDifference):
9046         (printImage):
9047         (printImageDifferences):
9048
9049 2011-10-05  Balazs Kelemen  <kbalazs@webkit.org>
9050
9051         [Qt][WK2] Unreviewed build fix.
9052
9053         Don't use WebCore directly in WTR.
9054
9055         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
9056         (WTR::PlatformWebView::addChromeInputField):
9057         (WTR::PlatformWebView::removeChromeInputField):
9058         (WTR::PlatformWebView::makeWebViewFirstResponder):
9059
9060 2011-10-05  David Levin  <levin@chromium.org>
9061
9062         watchlist: Add more rules for chromium public api and sort the file.
9063         https://bugs.webkit.org/show_bug.cgi?id=69323
9064
9065         Reviewed by Adam Barth.
9066
9067         * Scripts/webkitpy/common/config/watchlist:
9068
9069 2011-10-05  David Levin  <levin@chromium.org>
9070
9071         watchlist: Don't add the same message to a bug more than once.
9072         https://bugs.webkit.org/show_bug.cgi?id=69303
9073
9074         Reviewed by Adam Barth.
9075
9076         * Scripts/webkitpy/common/net/bugzilla/bug.py: Added a way to determine
9077         if a message is in the comments already.
9078         * Scripts/webkitpy/common/net/bugzilla/bug_unittest.py: A test for the above.
9079         * Scripts/webkitpy/tool/commands/applywatchlistlocal_unittest.py:
9080         Pick a bug supplied by the mock. Change the test due to different output
9081         from the mock watch list tool.
9082         * Scripts/webkitpy/tool/mocktool.py: Change the mock watch list to return
9083         another email so it will be filtered out, fix bugs to have the cc and comment fields,
9084         and fix fetch_bug to handle bug_id's which are text (because that took me way too long
9085         to debug).
9086         * Scripts/webkitpy/tool/steps/applywatchlist.py: Change to filter out comments
9087         and/or cc's that are already in the bug.
9088         * Scripts/webkitpy/tool/steps/applywatchlist_unittest.py: Pick a bug supplied by the mock,
9089         and remove a comment that is filtered out.
9090
9091 2011-10-05  Balazs Kelemen  <kbalazs@webkit.org>
9092
9093         [Qt][WK2] Unreviewed build fix after r96643.
9094
9095         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
9096         (WTR::PlatformWebView::addChromeInputField):
9097         (WTR::PlatformWebView::removeChromeInputField):
9098         (WTR::PlatformWebView::makeWebViewFirstResponder):
9099
9100 2011-10-04  Kent Tamura  <tkent@chromium.org>
9101
9102         Introduce feature flags for incomplete input types
9103         https://bugs.webkit.org/show_bug.cgi?id=68971
9104
9105         Reviewed by Hajime Morita.
9106
9107         * Scripts/build-webkit:
9108         Add --input-type-{date,datetime,datetimelocal,month,time,week} flags.
9109         They are disabled by default.
9110
9111 2011-10-04  Adam Barth  <abarth@webkit.org>
9112
9113         garden-o-matic doesn't launch it's UI on Lion
9114         https://bugs.webkit.org/show_bug.cgi?id=69349
9115
9116         Reviewed by Darin Adler.
9117
9118         It turns out that Lion is a bit more picky about file URLs.  This patch
9119         switches us from using a path to using an actual file URL, silencing
9120         the warning and making the tool actually launch on Lion.
9121
9122         * Scripts/webkitpy/tool/servers/gardeningserver.py:
9123
9124 2011-10-04  Daniel Bates  <dbates@rim.com>
9125
9126         Fix Perl undefined variable warning when using run-safari after
9127         <http://trac.webkit.org/changeset/96448> (https://bugs.webkit.org/show_bug.cgi?id=68662)
9128
9129         Rubber-stamped by Alexey Proskuryakov.
9130
9131         * Scripts/webkitdirs.pm:
9132         (runMacWebKitApp): Check that the optional function argument $useOpenCommand
9133         is defined before using it in a comparison operation.
9134
9135 2011-10-04  Scott Graham  <scottmg@chromium.org>
9136
9137         Add GAMEPAD feature flag
9138         https://bugs.webkit.org/show_bug.cgi?id=66859
9139
9140         Reviewed by Darin Fisher.
9141
9142         * Scripts/build-webkit:
9143
9144 2011-10-04  Jon Lee  <jonlee@apple.com>
9145
9146         REGRESSION (WK2): (Shift-)option-tabbing skips over elements when transitioning from chrome to webview
9147         https://bugs.webkit.org/show_bug.cgi?id=68412
9148         <rdar://problem/9988252>
9149
9150         Reviewed by Darin Adler.
9151
9152         In order to create a test for the bug, I had to update DRT and WKTR to create some
9153         widget that allows first responder status to move away from the main web view.
9154
9155         Three methods were added to layoutTestController: addChromeInputField,
9156         removeChromeInputField, and focusWebView. addChromeInputField adds a text field
9157         that is a sibling to the web view, and sets up the key event loop between the two.
9158         removeChromeInputField removes that field. focusWebView moves first responder
9159         status to the web view.
9160
9161         The test makes the call via layoutTestController and passes a callback that it
9162         assumes will be executed once the task is completed. In DRT the callback is called
9163         synchronously. In WKTR this is handled with message passing between the two
9164         processes.
9165
9166         * DumpRenderTree/LayoutTestController.cpp:
9167         (addChromeInputFieldCallback):
9168         (removeChromeInputFieldCallback):
9169         (focusWebViewCallback):
9170         (LayoutTestController::staticFunctions):
9171         * DumpRenderTree/LayoutTestController.h:
9172         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
9173         (LayoutTestController::addChromeInputField):
9174         (LayoutTestController::removeChromeInputField):
9175         (LayoutTestController::focusWebView):
9176         * DumpRenderTree/mac/DumpRenderTree.mm:
9177         (resetWebViewToConsistentStateBeforeTesting): When resetting for the next test,
9178         make sure to remove the chrome input field.
9179         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
9180         (LayoutTestController::addChromeInputField):
9181         (LayoutTestController::removeChromeInputField):
9182         (LayoutTestController::focusWebView):
9183         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
9184         (LayoutTestController::addChromeInputField):
9185         (LayoutTestController::removeChromeInputField):
9186         (LayoutTestController::focusWebView):
9187         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
9188         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
9189         (WTR::InjectedBundle::didReceiveMessage):
9190         (WTR::InjectedBundle::postAddChromeInputField):
9191         (WTR::InjectedBundle::postRemoveChromeInputField):
9192         (WTR::InjectedBundle::postFocusWebView):
9193         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
9194         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
9195         (WTR::callbackMap): Create a hash map that keeps track of the callbacks provided
9196         through JS.
9197         (WTR::cacheLayoutTestControllerCallback):
9198         (WTR::callLayoutTestControllerCallback):
9199         (WTR::LayoutTestController::addChromeInputField):
9200         (WTR::LayoutTestController::removeChromeInputField):
9201         (WTR::LayoutTestController::focusWebView):
9202         (WTR::LayoutTestController::callAddChromeInputFieldCallback):
9203         (WTR::LayoutTestController::callRemoveChromeInputFieldCallback):
9204         (WTR::LayoutTestController::callFocusWebViewCallback):
9205         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
9206         * WebKitTestRunner/PlatformWebView.h:
9207         * WebKitTestRunner/TestController.cpp:
9208         (WTR::TestController::resetStateToConsistentValues):
9209         * WebKitTestRunner/TestInvocation.cpp:
9210         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
9211         * WebKitTestRunner/gtk/PlatformWebViewGtk.cpp:
9212         (WTR::PlatformWebView::addChromeInputField):
9213         (WTR::PlatformWebView::removeChromeInputField):
9214         (WTR::PlatformWebView::makeWebViewFirstResponder):
9215         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
9216         (WTR::PlatformWebView::addChromeInputField):
9217         (WTR::PlatformWebView::removeChromeInputField):
9218         (WTR::PlatformWebView::makeWebViewFirstResponder):
9219         * WebKitTestRunner/win/PlatformWebViewWin.cpp:
9220         (WTR::PlatformWebView::addChromeInputField):
9221         (WTR::PlatformWebView::removeChromeInputField):
9222         (WTR::PlatformWebView::makeWebViewFirstResponder):
9223
9224         * DumpRenderTree/mac/LayoutTestControllerMac.mm: These functions have nothing to do
9225         with the patch-- just cleaning up style.
9226         (LayoutTestController::addDisallowedURL):
9227         (originsArrayToJS):
9228         (LayoutTestController::queueLoad):
9229         (LayoutTestController::setMockDeviceOrientation):
9230         (LayoutTestController::setIconDatabaseEnabled):
9231         (LayoutTestController::setEditingBehavior):
9232
9233 2011-10-04  Simon Fraser  <simon.fraser@apple.com>
9234
9235         Move font-fixup code in WebKitTestRunner to a better place
9236         https://bugs.webkit.org/show_bug.cgi?id=69356
9237
9238         Reviewed by Sam Weinig.
9239
9240         Move code that swizzles NSFontManager methods to ActivateFonts.mm, which
9241         already contains font-related code.
9242
9243         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
9244         (WTR::InjectedBundle::initialize):
9245         * WebKitTestRunner/InjectedBundle/mac/ActivateFonts.mm:
9246         (WTR::allowedFontFamilySet):
9247         (WTR::drt_NSFontManager_availableFontFamilies):
9248         (WTR::drt_NSFontManager_availableFonts):
9249         (WTR::swizzleNSFontManagerMethods):
9250         (WTR::activateFonts):
9251         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
9252         (WTR::InjectedBundle::platformInitialize):
9253
9254 2011-10-04  Raphael Kubo da Costa  <kubo@profusion.mobi>
9255
9256         [EFL] DRT: Add DumpRenderTree.cpp
9257         https://bugs.webkit.org/show_bug.cgi?id=62034
9258
9259         Reviewed by Antonio Gomes.
9260
9261         This is the main implementation file for EFL's DumpRenderTree.
9262
9263         * DumpRenderTree/efl/DumpRenderTree.cpp: Added.
9264         (dumpFramesAsText):
9265         (dumpFrameScrollPosition):
9266         (shouldLogFrameLoadDelegates):
9267         (shouldDumpAsText):
9268         (sendPixelResultsEOF):
9269         (invalidateAnyPreviousWaitToDumpWatchdog):
9270         (onEcoreEvasResize):
9271         (onCloseWindow):
9272         (useLongRunningServerMode):
9273         (parseCommandLineOptions):
9274         (getFinalTestURL):
9275         (getExpectedPixelHash):
9276         (createLayoutTestController):
9277         (runTest):
9278         (runTestingServerLoop):
9279         (adjustOutputTypeByMimeType):
9280         (dumpFrameContentsAsText):
9281         (shouldDumpFrameScrollPosition):
9282         (shouldDumpPixelsAndCompareWithExpected):
9283         (shouldDumpBackForwardList):
9284         (initEfl):
9285         (shutdownEfl):
9286         (displayWebView):
9287         (dump):
9288         (initEcoreEvas):
9289         (main):
9290
9291 2011-10-03  Adam Barth  <abarth@webkit.org>
9292
9293         garden-o-matic should remove builders from the failure grid once they start to pass
9294         https://bugs.webkit.org/show_bug.cgi?id=69309
9295
9296         Unreviewed.  dglazkov is on vacation.  I need to interest someone else
9297         in reviewing these patches.
9298
9299         This patch follows the update/purge model from UpdateTracker.  This
9300         patch is less awesome than it could be because it causes the UI to
9301         flash slightly.  In a future patch, we'll want to use these
9302         update/purge notifications to update the UI without flashing.  However,
9303         that's not such a big deal because this UI is hidden in the default
9304         view.
9305
9306         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
9307         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js:
9308         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
9309         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
9310
9311 2011-10-03  David Levin  <levin@chromium.org>
9312
9313         webkitpy: Expose cc_emails and comments from bug.py
9314         https://bugs.webkit.org/show_bug.cgi?id=69308
9315
9316         Reviewed by Adam Barth.
9317
9318         * Scripts/webkitpy/common/net/bugzilla/bug.py: Expose cc_emails and comments .
9319         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py: Add parsing support for comments.
9320         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py: Fix the unit test due
9321         to comments being exposed.
9322
9323 2011-10-03  James Robinson  <jamesr@chromium.org>
9324
9325         Add myself to the watchlist for platform/graphics/chromium
9326         https://bugs.webkit.org/show_bug.cgi?id=69297
9327
9328         Reviewed by David Levin.
9329
9330         * Scripts/webkitpy/common/config/watchlist:
9331
9332 2011-10-03  Anders Carlsson  <andersca@apple.com>
9333
9334         Remove custom scrollbar painting hooks
9335         https://bugs.webkit.org/show_bug.cgi?id=69163
9336
9337         Reviewed by Alexey Proskuryakov.
9338
9339         * DumpRenderTree/chromium/WebPreferences.cpp:
9340         (WebPreferences::applyTo):
9341         No need to enable custom scrollbar painting.
9342
9343 2011-10-03  David Levin  <levin@chromium.org>
9344
9345         watchlist: Make watchlist run on a bot.
9346         https://bugs.webkit.org/show_bug.cgi?id=68974
9347
9348         Reviewed by Adam Barth.
9349
9350         * Scripts/webkitpy/tool/commands/queues.py: Add the watchlist to the style bot.
9351         * Scripts/webkitpy/tool/commands/queues_unittest.py:
9352         Change to log what is run so that we can see the style command
9353         and the watchlist command being run.
9354
9355 2011-10-03  David Levin  <levin@chromium.org>
9356
9357         watchlist: Allow specified a bug to attach the watchlist info to.
9358         https://bugs.webkit.org/show_bug.cgi?id=69288
9359
9360         Reviewed by Eric Seidel.
9361
9362         * Scripts/webkitpy/tool/commands/applywatchlistlocal.py: Add support for bugid.
9363         * Scripts/webkitpy/tool/commands/applywatchlistlocal_unittest.py:
9364         Add tests and fix an existing test because the unit tests do not go through the
9365         normal command line processing (so the -g argument remained when we determined bug ids).
9366         * Scripts/webkitpy/tool/commands/commandtest.py: Change the derivation to
9367         get the assert raises regex method.
9368         * Scripts/webkitpy/tool/commands/download.py: Fix the command help to be more unixy.
9369
9370 2011-10-03  Leandro Pereira  <leandro@profusion.mobi>
9371
9372         [EFL] DRT: Add JSStringUtils.{cpp,h}
9373         https://bugs.webkit.org/show_bug.cgi?id=66397
9374
9375         Reviewed by Ryosuke Niwa.
9376
9377         Adds helper functions to compare the equality of C-style strings and
9378         JSStringRef objects.
9379
9380         * DumpRenderTree/efl/JSStringUtils.cpp: Added.
9381         (equals):
9382         * DumpRenderTree/efl/JSStringUtils.h: Added.
9383
9384 2011-10-03  Adam Barth  <abarth@webkit.org>
9385
9386         garden-o-matic should work in Safari 5.1
9387         https://bugs.webkit.org/show_bug.cgi?id=69290
9388
9389         Reviewed by Sam Weinig.
9390
9391         My old implementation of bind was too clever by half.  This one seems
9392         to work better, at least according to this test.
9393
9394         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
9395         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
9396
9397 2011-09-29  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
9398
9399         Unreviewed: change my e-mail in commiters.py
9400
9401         (I'm leaving the company and my @openbossa.org will be no more)
9402
9403         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css:
9404
9405 2011-10-01  Brent Fulgham  <bfulgham@webkit.org>
9406
9407         Correct post-build step for WinCairo test runs.
9408         https://bugs.webkit.org/show_bug.cgi?id=69191
9409
9410         Reviewed by Brian Weinstein.
9411
9412         * DumpRenderTree/win/DumpRenderTreePostBuild.cmd: Add additional
9413         shell case to handle the CFLite build.
9414
9415 2011-10-01  Filip Pizlo  <fpizlo@apple.com>
9416
9417         Added myself to reviewers, because I'm now a reviewer.
9418
9419         * Scripts/webkitpy/common/config/committers.py:
9420
9421 2011-10-01  Filip Pizlo  <fpizlo@apple.com>
9422
9423         Bencher script makes it difficult to do automated performance testing
9424         https://bugs.webkit.org/show_bug.cgi?id=69207
9425
9426         Reviewed by Sam Weinig.
9427
9428         This adds two new features:
9429
9430         The ability to disable automatic VM detection, which is flaky if any
9431         profiling features are enabled in jsc.
9432
9433         The ability to compute, and report, a scaled result for all benchmark
9434         suites. It is the geometric mean of three numbers: SunSpider's
9435         arithmetic mean, V8's geometric mean, and Kraken's arithmetic mean.
9436         It is also possible to turn off all other output from bencher and just
9437         get this number with the --brief option.
9438
9439         * Scripts/bencher:
9440
9441 2011-10-01  Sam Weinig  <sam@webkit.org>
9442
9443         WTR is not successfully changing the NSUserDefaults
9444         https://bugs.webkit.org/show_bug.cgi?id=69204
9445
9446         Reviewed by Dan Bernstein.
9447
9448         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
9449         (WTR::InjectedBundle::platformInitialize):
9450         Use -setVolatileDomain:forName:NSArgumentDomain instead of registerDefaults to
9451         get our overrides picked up.
9452
9453 2011-09-30  Daniel Bates  <dbates@rim.com>
9454
9455         Cleanup: Extract code to launch a Mac WebKit application into a common function
9456         https://bugs.webkit.org/show_bug.cgi?id=68662
9457
9458         Reviewed by Adam Roben.
9459
9460         Extracts common code to launch a WebKit application on Mac OS X into a function
9461         so that we can remove duplicate code.
9462
9463         * Scripts/run-webkit-app: Modified to use webkitdirs::runMacWebKitApp().
9464         * Scripts/webkitdirs.pm: Sorted @EXPORT list and added constant USE_OPEN_COMMAND.
9465         (runMacWebKitApp): Added.
9466         (runSafari): Modified to use webkitdirs::runMacWebKitApp().
9467         (runMiniBrowser): Ditto.
9468         (runWebKitTestRunner): Ditto.
9469         (runTestWebKitAPI): Ditto.
9470
9471 2011-09-29  David Levin  <levin@chromium.org>
9472
9473         watchlist: Add webkit-patch command to run watchlist.
9474         https://bugs.webkit.org/show_bug.cgi?id=68973
9475
9476         Reviewed by Eric Seidel.
9477
9478         * Scripts/webkitpy/common/host.py: Exposed the watch_list.
9479         * Scripts/webkitpy/common/watchlist/watchlist.py: Changed to return
9480         the cc and messages as sorted lists. To allow for deterministic ordering.
9481         * Scripts/webkitpy/common/watchlist/watchlist_unittest.py: Ditto.
9482         * Scripts/webkitpy/tool/commands/__init__.py: Add ApplyWatchListLocal so
9483         that it will be exposed as a command by webkit-patch.
9484         * Scripts/webkitpy/tool/commands/applywatchlistlocal.py: Added.
9485         * Scripts/webkitpy/tool/commands/applywatchlistlocal_unittest.py: Added.
9486         * Scripts/webkitpy/tool/commands/download.py: Added ApplyWatchList which mimics CheckStyle.
9487         * Scripts/webkitpy/tool/commands/download_unittest.py: Added an appropriate test.
9488         * Scripts/webkitpy/tool/mocktool.py: Mock out the watch list.
9489         * Scripts/webkitpy/tool/steps/__init__.py: Added ApplyWatchList.
9490         * Scripts/webkitpy/tool/steps/applywatchlist.py: Added.
9491         * Scripts/webkitpy/tool/steps/applywatchlist_unittest.py: Added.
9492
9493 2011-09-30  David Levin  <levin@chromium.org>
9494
9495         watchlist: Add a way to detect a net increase or decrease of a pattern (in a file).
9496         https://bugs.webkit.org/show_bug.cgi?id=69031
9497
9498         Reviewed by Adam Barth.
9499
9500         * Scripts/webkitpy/common/config/watchlist: Added usage of the new pattern,
9501         a comment to explain something important about the email addresses, and
9502         a comment to let emacs know to treat the file as a python file.
9503         * Scripts/webkitpy/common/watchlist/amountchangedpattern.py: Added.
9504         Detects increases or decreases in a pattern.
9505         * Scripts/webkitpy/common/watchlist/amountchangedpattern_unittest.py: Added.
9506         * Scripts/webkitpy/common/watchlist/watchlist_unittest.py:
9507         Added unit tests for watchlist which has "more" or "less".
9508         * Scripts/webkitpy/common/watchlist/watchlistparser.py:
9509         Added support for "more" or "less".
9510
9511 2011-09-30  Sam Weinig  <sam@webkit.org>
9512
9513         Add support for eventSender.mouseScrollBy in WTR
9514         https://bugs.webkit.org/show_bug.cgi?id=69189
9515
9516         Add implementation of eventSender.mouseScrollBy for the mac.
9517
9518         Reviewed by Simon Fraser.
9519
9520         * WebKitTestRunner/EventSenderProxy.h:
9521         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
9522         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
9523         (WTR::EventSendingController::keyDown):
9524         (WTR::EventSendingController::mouseScrollBy):
9525         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
9526         * WebKitTestRunner/TestController.cpp:
9527         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
9528         * WebKitTestRunner/mac/EventSenderProxy.mm:
9529         (WTR::EventSenderProxy::mouseScrollBy):
9530
9531 2011-09-30  Simon Fraser  <simon.fraser@apple.com>
9532
9533         fast/forms/text-control-intrinsic-widths.html fails on Mac if Word/Office fonts are installed
9534         https://bugs.webkit.org/show_bug.cgi?id=35273
9535
9536         Reviewed by Dan Bernstein.
9537
9538         There are two types of font issue that can cause layout test failure.
9539         First, the user may have non-standard fonts installed that affect the layout
9540         of some tests. For example, some tests use the MS Gothic font if installed.
9541
9542         Secondly, the user may have duplicate fonts installed in ~/Library/Fonts,
9543         which override the system fonts (e.g. a copy of Arial from MS Office). For
9544         some tests, this can affect font metrics or fallback behavior.
9545
9546         This change addresses the first problem only. It limits the set of fonts
9547         available through NSFontManager, by swizzling some methods to return
9548         a list of fonts or font families which only include fonts from a hardcoded
9549         whitelist.
9550
9551         * DumpRenderTree/mac/DumpRenderTree.mm:
9552         (allowedFontFamilySet):
9553         (drt_NSFontManager_availableFontFamilies):
9554         (drt_NSFontManager_availableFonts):
9555         (swizzleNSFontManagerMethods):
9556         (activateTestingFonts):
9557         (adjustFonts):
9558         (createWebViewAndOffscreenWindow):
9559         (prepareConsistentTestingEnvironment):
9560         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
9561         (WTR::allowedFontFamilySet):
9562         (WTR::drt_NSFontManager_availableFontFamilies):
9563         (WTR::drt_NSFontManager_availableFonts):
9564         (WTR::swizzleNSFontManagerMethods):
9565         (WTR::InjectedBundle::platformInitialize):
9566
9567 2011-09-30  David Levin  <levin@chromium.org>
9568
9569         Need to skip webkitpy.common.checkout.baselineoptimizer_unittest.BaselineOptimizerTest on Windows..
9570         https://bugs.webkit.org/show_bug.cgi?id=69039
9571
9572         Reviewed by Adam Barth.
9573
9574         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
9575         Actually skip only one specific test that is failing on Windows.
9576
9577 2011-09-30  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
9578
9579         [Qt] Prevent qDebug() output from DRT and WTR unless --verbose
9580
9581         For DRT we didn't install the message handler early enough to
9582         catch output while constructing the QApplication. For WTR we
9583         didn't even install a message handler.
9584
9585         Since the UI process will forward any output from the web process
9586         we set an environment variable QT_WEBKIT_SUPPRESS_WEB_PROCESS_OUTPUT
9587         in WTR before the web process is started. This is picked up by the
9588         web process which installs its own message handler.
9589
9590         The environment variable can be overriden on the command line if you
9591         want to see output from the web process, or you can pass --verbose to
9592         WTR to see output from both processes.
9593
9594         https://bugs.webkit.org/show_bug.cgi?id=69132
9595
9596         Reviewed by Andreas Kling.
9597
9598         * DumpRenderTree/qt/main.cpp:
9599         * WebKitTestRunner/qt/main.cpp:
9600
9601 2011-09-30  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
9602
9603         [Qt] Make sure WTR sizes the window and item correctly
9604
9605         Revision 96345 changed the logic for how the view and
9606         window was created, but missed a vital part, setting
9607         the size.
9608
9609         We now use a QSGView for the window, that has a simple
9610         item as its root object that is always resized to fit
9611         within the window. The webview is then parented to the
9612         root object and set to anchors.fill: parent. That way
9613         any window geometry changes will propagate to the web
9614         view.
9615
9616         https://bugs.webkit.org/show_bug.cgi?id=69134
9617
9618         Reviewed by Andreas Kling.
9619
9620         * WebKitTestRunner/PlatformWebView.h:
9621         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
9622
9623 2011-09-30  Raphael Kubo da Costa  <kubo@profusion.mobi>
9624
9625         [EFL] Only save the current viewport in PixelDumpSupportEfl.
9626         https://bugs.webkit.org/show_bug.cgi?id=68450
9627
9628         Reviewed by Antonio Gomes.
9629
9630         The previous implementation saved the whole page contents in the final
9631         PNG, however the baselines and expectations in the tree only want the
9632         current viewport (with a scrollbar, if needed), so we now only save the
9633         currently visible contents.
9634
9635         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
9636         (createBitmapContextFromWebView): Use
9637         ewk_frame_visible_content_geometry_get instead of
9638         ewk_frame_contents_size_get.
9639
9640 2011-09-30  Raphael Kubo da Costa  <kubo@profusion.mobi>
9641
9642         [EFL] Support and keep track of extra windows (views).
9643         https://bugs.webkit.org/show_bug.cgi?id=68454
9644
9645         Reviewed by Antonio Gomes.
9646
9647         So far, window.open and related JavaScript code would not work, as
9648         DRTView did not implementation
9649         Ewk_View_Smart_Class::window_{open,close}. The default implementation
9650         in ewk just returns the current window/view, which broke any
9651         layouttest which relied on window.open.
9652
9653         We now properly create and remove these new views. Unfortunately, the
9654         current way ewk works (and due to the fact that we cannot pass class
9655         methods as function pointers to the EFL functions), DRTView needs to
9656         know that DRTChrome exists and call `browser' directly in some points.
9657
9658         * DumpRenderTree/efl/DumpRenderTreeView.cpp:
9659         (onWindowCreate):
9660         (onWindowCloseDelayed):
9661         (onWindowClose):
9662         (drtViewTiledAdd):
9663
9664 2011-09-30  Qi Zhang  <qi.2.zhang@nokia.com>
9665
9666         WebKitTestRunner needs layoutTestController.setAppCacheMaximumSize
9667         https://bugs.webkit.org/show_bug.cgi?id=42698
9668
9669         Reviewed by Andreas Kling.
9670
9671         Implemented layoutTestController.setAppCacheMaximumSize and layoutTestController.clearApplicationCache,
9672         and call clearApplicationCache when layoutTest start up.
9673
9674         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
9675         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
9676         (WTR::InjectedBundle::beginTesting):
9677         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
9678         (WTR::LayoutTestController::clearAllApplicationCaches):
9679         (WTR::LayoutTestController::setAppCacheMaximumSize):
9680         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
9681
9682 2011-09-30  Zeno Albisser  <zeno.albisser@nokia.com>
9683
9684         [Qt][WK2] Change WebKitTestRunner to use QMouseEvents instead of QGraphicsSceneMouseEvents
9685         https://bugs.webkit.org/show_bug.cgi?id=69145
9686
9687         This is a follow up to https://bugs.webkit.org/show_bug.cgi?id=69105
9688         Due to refactorings in Qt5, QMouseEvent now supports floating point coordinates
9689         and is perfectly suitable for SceneGraph as well. GraphicsScene events
9690         will not be available in QtGui anymore in future. Therefore we also have to change
9691         QtWebKit to use QMouseEvents instead of QGraphicsSceneMouseEvents.
9692
9693         Reviewed by Chang Shu.
9694
9695         * WebKitTestRunner/EventSenderProxy.h:
9696         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
9697         (WTR::EventSenderProxy::createMouseEvent):
9698         (WTR::EventSenderProxy::mouseDown):
9699         (WTR::EventSenderProxy::mouseUp):
9700         (WTR::EventSenderProxy::mouseMoveTo):
9701
9702 2011-09-30  James Robinson  <jamesr@chromium.org>
9703
9704         [chromium] Add WebKit API for sending input events to the compositor thread
9705         https://bugs.webkit.org/show_bug.cgi?id=69117
9706
9707         Reviewed by Darin Fisher.
9708
9709         Adds support for creating and setting a compositor thread in DRT.
9710
9711         * DumpRenderTree/chromium/WebViewHost.cpp:
9712         (WebViewHost::WebViewHost):
9713         * DumpRenderTree/chromium/WebViewHost.h:
9714
9715 2011-09-29  David Levin  <levin@chromium.org>
9716
9717         watchlist: Add cross-checks for WatchList once it is filled.
9718         https://bugs.webkit.org/show_bug.cgi?id=68975
9719
9720         Reviewed by Eric Seidel.
9721
9722         * Scripts/webkitpy/common/watchlist/watchlist.py: Made the data members public
9723         instead of having trivial getter and setters.
9724         * Scripts/webkitpy/common/watchlist/watchlist_unittest.py: Fix the unit tests to
9725         pass the validation checks.
9726         * Scripts/webkitpy/common/watchlist/watchlistparser.py: Add validation checks
9727         and fix a few style nits.
9728         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py: Add tests for the
9729         validation checks.
9730         * Scripts/webkitpy/common/watchlist/watchlistrule.py: Make definitions_to_match public.
9731
9732 2011-09-29  Xianzhu Wang  <wangxianzhu@chromium.org>
9733
9734         run-api-tests fails on chromium-win bot
9735         https://bugs.webkit.org/show_bug.cgi?id=69121
9736
9737         On Windows running a perl script needs the interpreter.
9738
9739         Reviewed by Adam Barth.
9740
9741         * Scripts/run-api-tests:
9742         (buildTestTool):
9743
9744 2011-09-29  Adam Barth  <abarth@webkit.org>
9745
9746         Attempt to make this tool work on Windows.
9747
9748         * Scripts/run-api-tests:
9749         (buildTestTool):
9750
9751 2011-09-29  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
9752
9753         [Qt] Try to fix the minial build
9754
9755         * QtTestBrowser/launcherwindow.cpp:
9756         * QtTestBrowser/webpage.cpp:
9757
9758 2011-09-23  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
9759
9760         [Qt] Fix build against Qt5 after refactor of widgets out of QtGUi
9761
9762         QWidget and friends now live in the QtWidgets library. We update
9763         includes in implementation files and private headers to us the
9764         non-module-prefixed path, and leave the lookup for the include
9765         path. For public headers we have to ifdef the includes as the
9766         user might now have the modules we need in his QT config.
9767
9768         Finally, QSGCanvas is no longer a QWidget but a QWindow, so we
9769         have to update our code and use windowHandle() for setting the
9770         parent relationships.
9771
9772         https://bugs.webkit.org/show_bug.cgi?id=68687
9773
9774         Reviewed by Andreas Kling.
9775
9776         * DumpRenderTree/qt/DumpRenderTree.pro:
9777         * DumpRenderTree/qt/ImageDiff.pro:
9778         * MiniBrowser/qt/BrowserView.cpp:
9779         * MiniBrowser/qt/BrowserView.h:
9780         * MiniBrowser/qt/BrowserWindow.cpp:
9781         * MiniBrowser/qt/BrowserWindow.h:
9782         * MiniBrowser/qt/MiniBrowser.pro:
9783         * MiniBrowser/qt/MiniBrowserApplication.h:
9784         * MiniBrowser/qt/main.cpp:
9785         * QtTestBrowser/QtTestBrowser.pro:
9786         * QtTestBrowser/launcherwindow.cpp:
9787         * QtTestBrowser/launcherwindow.h:
9788         * QtTestBrowser/locationedit.h:
9789         * QtTestBrowser/main.cpp:
9790         * QtTestBrowser/mainwindow.cpp:
9791         * QtTestBrowser/mainwindow.h:
9792         * QtTestBrowser/webinspector.h:
9793         * QtTestBrowser/webpage.cpp:
9794         * QtTestBrowser/webview.cpp:
9795         * WebKitTestRunner/PlatformWebView.h:
9796         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
9797
9798 2011-09-29  Ravi Phaneendra Kasibhatla  <ravi.kasibhatla@motorola.com>
9799
9800         Initial implementation of WebInspector for WebKit2 GTK port.
9801         https://bugs.webkit.org/show_bug.cgi?id=68235
9802
9803         Reviewed by Martin Robinson.
9804
9805         * MiniBrowser/gtk/BrowserWindow.c:
9806         (browserWindowConstructed): Set the developer extra flags to true to enable inspect element.
9807
9808 2011-09-29  Adam Barth  <abarth@webkit.org>
9809
9810         Unbreak the "update expectatations" button by using the existing
9811         machinery to create the failureInfo objects.
9812
9813         Unbreak the "close" button on the progress dialog by using |this|
9814         instead of the old name of the object.
9815
9816         Teach the controllers unit test not to talk to the actual network.
9817
9818         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
9819         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers_unittests.js:
9820         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
9821
9822 2011-09-29  Hans Wennborg  <hans@chromium.org>
9823
9824         IndexedDB: Use LevelDB also for in-memory databases
9825         https://bugs.webkit.org/show_bug.cgi?id=68903
9826
9827         Reviewed by Steve Block.
9828
9829         Remove the temporary dir that was necessary before LevelDB supported
9830         in-memory databases.
9831
9832         * DumpRenderTree/chromium/TestShell.cpp:
9833         (TestShell::TestShell):
9834         * DumpRenderTree/chromium/TestShell.h:
9835
9836 2011-09-28  Xianzhu Wang  <wangxianzhu@chromium.org>
9837
9838         Run TestWebKitAPI on Chromium buildbots
9839         https://bugs.webkit.org/show_bug.cgi?id=68678
9840
9841         Reviewed by Tony Chang.
9842
9843         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
9844
9845 2011-09-28  Dimitri Glazkov  <dglazkov@chromium.org>
9846
9847         garden-o-matic results should anticipate 0 unexpected failures.
9848         https://bugs.webkit.org/show_bug.cgi?id=69036
9849
9850         Also fixes controllers: UnexpectedFailures test.
9851
9852         Reviewed by Adam Barth.
9853
9854         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js: Tweaked.
9855
9856 2011-09-28  Dirk Pranke  <dpranke@chromium.org>
9857
9858         Modify new-run-webkit-tests to return the percentage difference in image diffs where possible.
9859
9860         https://bugs.webkit.org/show_bug.cgi?id=67253
9861
9862         Reviewed by Eric Seidel.
9863
9864         Nearly all of the work in this was done by Simon Fraser; I'm
9865         just repackaging it and cleaning it up a bit. This change
9866         modifies port.diff_image() to return a tuple of (pass/fail,
9867         %age), adds the value to the FailureImageHashMismatch, and
9868         writes the value into the full_results.json files to be
9869         displayed in the results page.
9870
9871         * Scripts/webkitpy/layout_tests/controllers/manager.py:
9872         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
9873         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
9874         * Scripts/webkitpy/layout_tests/models/test_failures.py:
9875         * Scripts/webkitpy/layout_tests/port/base.py:
9876         * Scripts/webkitpy/layout_tests/port/chromium.py:
9877         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
9878         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
9879         * Scripts/webkitpy/layout_tests/port/test.py:
9880         * Scripts/webkitpy/layout_tests/port/webkit.py:
9881         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
9882
9883 2011-09-25  Kentaro Hara  <haraken@chromium.org>
9884
9885         Unreviewed. Adding myself to committers.py.
9886
9887         * Scripts/webkitpy/common/config/committers.py:
9888
9889 2011-09-28  Adam Barth  <abarth@webkit.org>
9890
9891         garden-o-matic shouldn't depend on test-results.appspot.com
9892         https://bugs.webkit.org/show_bug.cgi?id=69019
9893
9894         Reviewed by Dimitri Glazkov.
9895
9896         This patch moves garden-o-matic from getting full_results.json from
9897         test-results.appspot.com to getting that information directly from the
9898         buildbot.  This patch improve the load time and accuracy of
9899         garden-o-matic because test-results.appspot.com is slow and is dropping
9900         a significant number of writes (see
9901         https://bugs.webkit.org/show_bug.cgi?id=68748).
9902
9903         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net.js:
9904         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
9905         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js:
9906
9907 2011-09-28  David Levin  <levin@chromium.org>
9908
9909         watchlist: Add a way to load the watchlist from config.
9910         https://bugs.webkit.org/show_bug.cgi?id=68991
9911
9912         Reviewed by Eric Seidel.
9913
9914         * Scripts/webkitpy/common/config/watchlist: Added.
9915         * Scripts/webkitpy/common/watchlist/watchlistloader.py: Added.
9916         Support for loading the watch list.
9917         * Scripts/webkitpy/common/watchlist/watchlistloader_unittest.py: Added.
9918         Tests for the loading.
9919         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py:
9920         Adjusted to use the common base class.
9921         * Scripts/webkitpy/common/webkitunittest.py: Added.
9922         Made a common base class for functionality used in more than one.
9923
9924 2011-09-28  David Levin  <levin@chromium.org>
9925
9926         Attempt fix for Leopard python unit test run.
9927
9928         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py:
9929
9930 2011-09-27  Dimitri Glazkov  <dglazkov@chromium.org>
9931
9932         REGRESSION(r95573): Crash when loading SVG documents in a flattened frame or any SVG document in Chromium/Mac.
9933         https://bugs.webkit.org/show_bug.cgi?id=68938
9934
9935         Made Chromium/Mac DRT better match how Chromium/Mac queries WebKit.
9936
9937         Reviewed by David Hyatt.
9938
9939         * DumpRenderTree/chromium/WebViewHost.cpp:
9940         (WebViewHost::didUpdateLayout): Added width/height queries.
9941         * DumpRenderTree/chromium/WebViewHost.h: Updated defs.
9942
9943 2011-09-28  Mihai Parparita  <mihaip@chromium.org>
9944
9945         Get rebaseline server running again
9946         https://bugs.webkit.org/show_bug.cgi?id=69020
9947
9948         Reviewed by Adam Barth.
9949
9950         Make rebaseline server run again after r90337 (missing format_date_time
9951         import, results_json wasn't outputting anything).
9952
9953         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
9954         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
9955
9956 2011-09-28  David Grogan  <dgrogan@chromium.org>
9957
9958         Move dgrogan from contributor to committer
9959         https://bugs.webkit.org/show_bug.cgi?id=69021
9960
9961         * Scripts/webkitpy/common/config/committers.py:
9962
9963 2011-09-28  Mihai Parparita  <mihaip@chromium.org>
9964
9965         Add Mac Skia GPU bots to builders.py
9966         https://bugs.webkit.org/show_bug.cgi?id=69016
9967
9968         Reviewed by Eric Seidel.
9969
9970         This allows wwebkit-patch rebaseline-expectations to work with Mac Skia
9971         GPU tests.
9972
9973         * Scripts/webkitpy/layout_tests/port/builders.py:
9974         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
9975
9976 2011-09-28  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
9977
9978         Second try at fixing the GTK+ build.
9979
9980         * GNUmakefile.am:
9981
9982 2011-09-28  Xianzhu Wang  <wangxianzhu@chromium.org>
9983
9984         Let TestWebKitAPI and webkit_unit_tests work for chromium-win
9985         https://bugs.webkit.org/show_bug.cgi?id=68680
9986
9987         Reviewed by Tony Chang.
9988
9989         * Scripts/run-chromium-webkit-unit-tests:
9990         * Scripts/webkitdirs.pm:
9991         (determineBaseProductDir):
9992
9993 2011-09-28  Kaustubh Atrawalkar  <kaustubh@motorola.com>
9994
9995         [Gtk] Implement textInputController.doCommand
9996         https://bugs.webkit.org/show_bug.cgi?id=66496
9997
9998         Reviewed by Martin Robinson.
9999
10000         * DumpRenderTree/gtk/TextInputController.cpp:
10001         (doCommandCallback): Callback function for textInputController.doCommand
10002
10003 2011-09-28  David Levin  <levin@chromium.org>
10004
10005         watchlist: Suggest corrections for typos and improve error message consistency.
10006         https://bugs.webkit.org/show_bug.cgi?id=68976
10007
10008         Reviewed by Adam Barth.
10009
10010         * Scripts/webkitpy/common/watchlist/watchlistparser.py: Add suggestions
10011         for typos and improve consistency of the error messages.
10012         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py: Add tests
10013         and improve the regex that were already present.
10014
10015 2011-09-28  David Levin  <levin@chromium.org>
10016
10017         watchlist: Add support for matching added or deleted lines.
10018         https://bugs.webkit.org/show_bug.cgi?id=68972
10019
10020         Reviewed by Adam Barth.
10021
10022         * Scripts/webkitpy/common/watchlist/changedlinepattern.py: Added.
10023         * Scripts/webkitpy/common/watchlist/changedlinepattern_unittest.py: Added.
10024         * Scripts/webkitpy/common/watchlist/watchlist.py: Comment fix up and fix input
10025         to the pattern match to only have the diff lines instead of the DiffFile.
10026         * Scripts/webkitpy/common/watchlist/watchlist_unittest.py:
10027         Add tests for the new patterns and combinations of the patterns.
10028         * Scripts/webkitpy/common/watchlist/watchlistparser.py:
10029         Sort imports. Add changeline support.
10030
10031 2011-09-28  Carlos Garcia Campos  <cgarcia@igalia.com>
10032
10033         [GTK] Add WebKitError to GTK+ WebKit2 API
10034         https://bugs.webkit.org/show_bug.cgi?id=68508
10035
10036         Reviewed by Martin Robinson.
10037
10038         * GNUmakefile.am: Add path to webkitgtk derived sources to the
10039         include path.
10040
10041 2011-09-28  Philippe Normand  <pnormand@igalia.com>
10042
10043         [GTK] missing WebKit2 support in run-gtk-tests
10044         https://bugs.webkit.org/show_bug.cgi?id=68992
10045
10046         Rubber-stamped by Andreas Kling.
10047
10048         * Scripts/run-gtk-tests: Get the WebKit2 tests to run from the
10049         webkit2 directory.
10050
10051 2011-09-27  Philippe Normand  <pnormand@igalia.com>
10052
10053         [GTK] build-webkit --no-webkit2 still builds WebKit2
10054         https://bugs.webkit.org/show_bug.cgi?id=68892
10055
10056         Reviewed by Martin Robinson.
10057
10058         Take the --no-webkit2 option in account when building WebKitGTK+.
10059
10060         * Scripts/build-webkit:
10061         * Scripts/webkitdirs.pm:
10062         (buildAutotoolsProject):
10063         (buildGtkProject):
10064
10065 2011-09-28  Ryosuke Niwa  <rniwa@webkit.org>
10066
10067         Make Chromium Win Release (Tests) a core builder
10068         https://bugs.webkit.org/show_bug.cgi?id=68906
10069
10070         Reviewed by Adam Barth.
10071
10072         Add Chromium Win Release (Tests) to the list of core builders.
10073
10074         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
10075         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
10076
10077 2011-09-27  Adam Barth  <abarth@webkit.org>
10078
10079         garden-o-matic should load faster and inform the user what is happening
10080         https://bugs.webkit.org/show_bug.cgi?id=68954
10081
10082         Reviewed by Dimitri Glazkov.
10083
10084         This test disables the expected failures tab.  The web inspector showed
10085         that loading the data for this tab was slowing down the loading time
10086         for the whole app.  We need to find a way to load that data lazily.
10087
10088         This patch also improves the update message to better describe the
10089         progress the app is making analyzing data from the buildbot.
10090
10091         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
10092         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
10093         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
10094
10095 2011-09-27  Adam Barth  <abarth@webkit.org>
10096
10097         garden-o-matic should show the first result when examining results
10098         https://bugs.webkit.org/show_bug.cgi?id=68960
10099
10100         Reviewed by Dimitri Glazkov.
10101
10102         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
10103
10104 2011-09-27  David Levin  <levin@chromium.org>
10105
10106         watchlist: Add support for cc and message rules.
10107         https://bugs.webkit.org/show_bug.cgi?id=68950
10108
10109         Reviewed by Adam Barth.
10110
10111         * Scripts/webkitpy/common/watchlist/watchlist.py: Added support to get
10112         the cc's and messages for a patch.
10113         * Scripts/webkitpy/common/watchlist/watchlist_unittest.py: Tests for the above.
10114         * Scripts/webkitpy/common/watchlist/watchlistparser.py: Parsing support
10115         for the rules.
10116         * Scripts/webkitpy/common/watchlist/watchlistrule.py: Copied from Tools/Scripts/webkitpy/common/watchlist/watchlist.py.
10117         A generic encapsulation of either a message list or a cc list.
10118         * Scripts/webkitpy/common/watchlist/watchlistrule_unittest.py: Copied from Tools/Scripts/webkitpy/common/watchlist/watchlist.py.
10119
10120 2011-09-27  Tom Zakrajsek  <tomz@codeaurora.org>
10121
10122         webkit-patch doesn't like UTF-8 characters in reviewers names
10123         https://bugs.webkit.org/show_bug.cgi?id=63452
10124
10125         Reviewed by Eric Seidel.
10126
10127         * Scripts/webkit-patch:
10128
10129 2011-09-27  Adam Barth  <abarth@webkit.org>
10130
10131         garden-o-matic examine buttons shows both expected and unexpected failures
10132         https://bugs.webkit.org/show_bug.cgi?id=68918
10133
10134         Reviewed by Dimitri Glazkov.
10135
10136         This was a copy/paste error when I refactored this classes to share
10137         more code.
10138
10139         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
10140
10141 2011-09-27  Ryosuke Niwa  <rniwa@webkit.org>
10142
10143         Add Kaustubh Atrawalkar to the list of contributors.
10144
10145         * Scripts/webkitpy/common/config/committers.py:
10146
10147 2011-09-27  Mihai Parparita  <mihaip@chromium.org>
10148
10149         [Chromium] Fixed layout mode tests break subsequent tests
10150         https://bugs.webkit.org/show_bug.cgi?id=68923
10151
10152         Reviewed by Adam Barth.
10153
10154         Reset fixed layout mode between tests, so that tests that run after
10155         platform/chromium/fast/repaint/fixed-layout-360x240.html still have the
10156         regular 800x600 non-fixed layout.
10157
10158         * DumpRenderTree/chromium/TestShell.cpp:
10159         (TestShell::resetTestController):
10160
10161 2011-09-27  David Levin  <levin@chromium.org>
10162
10163         watchlist: Add the filename pattern for definitions.
10164         https://bugs.webkit.org/show_bug.cgi?id=68917
10165
10166         Reviewed by Adam Barth.
10167
10168         * Scripts/webkitpy/common/watchlist/filenamepattern.py: Added.
10169         * Scripts/webkitpy/common/watchlist/filenamepattern_unittest.py: Added.
10170         * Scripts/webkitpy/common/watchlist/watchlist.py: Added the filename pattern
10171         for definitions.
10172         * Scripts/webkitpy/common/watchlist/watchlist_unittest.py: Added tests.
10173         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py: Typo fix.
10174
10175 2011-09-27  Adam Barth  <abarth@webkit.org>
10176
10177         garden-o-matic results view should sort test and builder names
10178         https://bugs.webkit.org/show_bug.cgi?id=68488
10179
10180         Reviewed by Andy Estes.
10181
10182         Previously, the test and builder names were displayed in an arbitrary
10183         order that changed from time to time.  That confused one user study
10184         participant.  This patch sorts the lists so that they occur in a
10185         predictable order.
10186
10187         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
10188
10189 2011-09-27  Sheriff Bot  <webkit.review.bot@gmail.com>
10190
10191         Unreviewed, rolling out r96108, r96111, r96113, and r96116.
10192         http://trac.webkit.org/changeset/96108
10193         http://trac.webkit.org/changeset/96111
10194         http://trac.webkit.org/changeset/96113
10195         http://trac.webkit.org/changeset/96116
10196         https://bugs.webkit.org/show_bug.cgi?id=68913
10197
10198         Wait for working Qt5 (Requested by ossy on #webkit).
10199
10200         * DumpRenderTree/qt/DumpRenderTree.pro:
10201         * DumpRenderTree/qt/ImageDiff.pro:
10202         * MiniBrowser/qt/BrowserView.cpp:
10203         (BrowserView::BrowserView):
10204         * MiniBrowser/qt/BrowserView.h:
10205         * MiniBrowser/qt/BrowserWindow.cpp:
10206         (BrowserWindow::BrowserWindow):
10207         (BrowserWindow::urlChanged):
10208         (BrowserWindow::~BrowserWindow):
10209         * MiniBrowser/qt/BrowserWindow.h:
10210         * MiniBrowser/qt/MiniBrowser.pro:
10211         * MiniBrowser/qt/MiniBrowserApplication.h:
10212         * MiniBrowser/qt/main.cpp:
10213         * QtTestBrowser/QtTestBrowser.pro:
10214         * QtTestBrowser/launcherwindow.cpp:
10215         * QtTestBrowser/launcherwindow.h:
10216         * QtTestBrowser/locationedit.h:
10217         * QtTestBrowser/main.cpp:
10218         * QtTestBrowser/mainwindow.cpp:
10219         * QtTestBrowser/mainwindow.h:
10220         * QtTestBrowser/webinspector.h:
10221         * QtTestBrowser/webpage.cpp:
10222         * QtTestBrowser/webview.cpp:
10223         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
10224         * WebKitTestRunner/PlatformWebView.h:
10225         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
10226         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
10227         (WTR::WebView::wkView):
10228         (WTR::WebView::pageRef):
10229         (WTR::WebView::~WebView):
10230         (WTR::WebView::WebView):
10231         (WTR::PlatformWebView::PlatformWebView):
10232         (WTR::PlatformWebView::sendEvent):
10233         (WTR::PlatformWebView::postEvent):
10234         * WebKitTestRunner/qt/WebKitTestRunner.pro:
10235
10236 2011-09-27  Carlos Garcia Campos  <cgarcia@igalia.com>
10237
10238         [GTK] Rename WebKit2 GTK+ API main header as webkit2.h
10239         https://bugs.webkit.org/show_bug.cgi?id=65178
10240
10241         Reviewed by Martin Robinson.
10242
10243         * GNUmakefile.am:
10244         * GtkLauncher/main.c: Include <webkit2/webkit2.h>.
10245
10246 2011-09-27  Csaba Osztrogonác  <ossy@webkit.org>
10247
10248         [Qt][WK2] Buildfix after r96108.
10249
10250         Rubber-stamped by Andreas Kling.
10251
10252         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
10253         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
10254         * WebKitTestRunner/qt/WebKitTestRunner.pro:
10255
10256 2011-09-27  David Levin  <levin@chromium.org>
10257
10258         watchlist: Change watchlistparser.py to be class based.
10259         https://bugs.webkit.org/show_bug.cgi?id=68869
10260
10261         Reviewed by Eric Seidel.
10262
10263         * Scripts/webkitpy/common/watchlist/watchlistparser.py:
10264         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py:
10265
10266 2011-09-27  David Levin  <levin@chromium.org>
10267
10268         watchlist: Break out the diff boilerplate to allow for re-use.
10269         https://bugs.webkit.org/show_bug.cgi?id=68871
10270
10271         Reviewed by Eric Seidel.
10272
10273         * Scripts/webkitpy/common/checkout/diff_parser_unittest.py:
10274         Break out the diff into a new file.
10275         * Scripts/webkitpy/common/checkout/diff_test_data.py: Added.
10276
10277 2011-09-27  Ilya Tikhonovsky  <loislo@chromium.org>
10278
10279         Web Inspector: UI performance: introduce heap size tracking stats.
10280         https://bugs.webkit.org/show_bug.cgi?id=68901
10281
10282         It is interesting how much the heap memory is used by Inspector in order of running the test.
10283
10284         Reviewed by Yury Semikhatsky.
10285
10286         * DumpRenderTree/chromium/TestShell.cpp:
10287         (TestShell::showDevTools):
10288         (TestShell::closeDevTools):
10289
10290 2011-09-23  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
10291
10292         [Qt] Fix build against Qt5 after refactor of widgets out of QtGUi
10293
10294         QWidget and friends now live in the QtWidgets library. We update
10295         includes in implementation files and private headers to us the
10296         non-module-prefixed path, and leave the lookup for the include
10297         path. For public headers we have to ifdef the includes as the
10298         user might now have the modules we need in his QT config.
10299
10300         Finally, QSGCanvas is no longer a QWidget but a QWindow, so we
10301         have to update our code and use windowHandle() for setting the
10302         parent relationships.
10303
10304         https://bugs.webkit.org/show_bug.cgi?id=68687
10305
10306         Reviewed by Andreas Kling.
10307
10308         * DumpRenderTree/qt/DumpRenderTree.pro:
10309         * DumpRenderTree/qt/ImageDiff.pro:
10310         * MiniBrowser/qt/BrowserView.cpp:
10311         * MiniBrowser/qt/BrowserView.h:
10312         * MiniBrowser/qt/BrowserWindow.cpp:
10313         * MiniBrowser/qt/BrowserWindow.h:
10314         * MiniBrowser/qt/MiniBrowser.pro:
10315         * MiniBrowser/qt/MiniBrowserApplication.h:
10316         * MiniBrowser/qt/main.cpp:
10317         * QtTestBrowser/QtTestBrowser.pro:
10318         * QtTestBrowser/launcherwindow.cpp:
10319         * QtTestBrowser/launcherwindow.h:
10320         * QtTestBrowser/locationedit.h:
10321         * QtTestBrowser/main.cpp:
10322         * QtTestBrowser/mainwindow.cpp:
10323         * QtTestBrowser/mainwindow.h:
10324         * QtTestBrowser/webinspector.h:
10325         * QtTestBrowser/webpage.cpp:
10326         * QtTestBrowser/webview.cpp:
10327         * WebKitTestRunner/PlatformWebView.h:
10328         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
10329
10330 2011-09-26  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
10331
10332         [Qt][WK2] Add support for hover API in Qt WebKit2
10333         https://bugs.webkit.org/show_bug.cgi?id=68369
10334
10335         Reviewed by Andreas Kling.
10336
10337         Change the statusbar to show the link URL when hovering links in
10338         MiniBrowser using QDesktopWebView.
10339
10340         * MiniBrowser/qt/BrowserWindow.cpp:
10341         (BrowserWindow::BrowserWindow):
10342         (BrowserWindow::onLinkHovered):
10343         * MiniBrowser/qt/BrowserWindow.h:
10344
10345 2011-09-26  Dimitri Glazkov  <dglazkov@chromium.org>
10346
10347         garden-o-matic's commit data on summary page should not crowd itself or twitch when hovered over.
10348         https://bugs.webkit.org/show_bug.cgi?id=68864
10349
10350         Reviewed by Adam Barth.
10351
10352         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary-mock.js: Updated mocks to work.
10353         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Changed the structure of commit data to keep commit revision apart from the rest of details.
10354         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Adjusted unit tests.
10355         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css: Made things look shiny.
10356
10357 2011-09-26  Csaba Osztrogonác  <ossy@webkit.org>
10358
10359         [Qt][WK2] One more unreviewed buildfix after r96005.
10360
10361         * WebKitTestRunner/qt/TestInvocationQt.cpp: Missing include added.
10362
10363 2011-09-26  Csaba Osztrogonác  <ossy@webkit.org>
10364
10365         [Qt][WK2] Unreviewed buildfix after r96005.
10366
10367         * WebKitTestRunner/qt/TestInvocationQt.cpp:
10368         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
10369
10370 2011-09-26  Dimitri Glazkov  <dglazkov@chromium.org>
10371
10372         garden-o-matic should be pretty in Open Sans.
10373         https://bugs.webkit.org/show_bug.cgi?id=68845
10374
10375         Reviewed by Adam Barth.
10376
10377         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css: Tweaked to bring back teh fonts.
10378
10379 2011-09-26  Dimitri Glazkov  <dglazkov@chromium.org>
10380
10381         garden-o-matic should not say "Just now" when it doesn't actually know when an event happened.
10382         https://bugs.webkit.org/show_bug.cgi?id=68841
10383
10384         Reviewed by Adam Barth.
10385
10386         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js: Changed ui.RelativeTime to not initialize with current time.
10387         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Adjusted unit tests.
10388         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js: Ditto.
10389
10390 2011-09-26  David Levin  <levin@chromium.org>
10391
10392         watchlist: Add parsing for definition section.
10393         https://bugs.webkit.org/show_bug.cgi?id=68850
10394
10395         Reviewed by Adam Barth.
10396
10397         * Scripts/webkitpy/common/watchlist/watchlist.py:
10398         * Scripts/webkitpy/common/watchlist/watchlistparser.py:
10399         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py:
10400
10401 2011-09-26  Simon Fraser  <simon.fraser@apple.com>
10402
10403         Fix WebKitTestRunner builds for Cairo, Windows and Qt.
10404
10405         * WebKitTestRunner/cairo/TestInvocationCairo.cpp:
10406         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
10407         * WebKitTestRunner/win/TestInvocationWin.cpp:
10408         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
10409
10410 2011-09-26  Simon Fraser  <simon.fraser@apple.com>
10411
10412         Repaint tests don't work in WebKit2
10413         https://bugs.webkit.org/show_bug.cgi?id=68453
10414
10415         Reviewed by Sam Weinig.
10416
10417         Convert DumpRenderTree to use use the repaint rects
10418         exposed via WebKit API when generating the repaint test
10419         overlay.
10420
10421         Fix WebKitTestRunner to work with repaint tests, again
10422         using the repaint rects exposed from WK2.
10423
10424         In both cases, layoutTestController.display() now just
10425         forces a display of the web view, and turns on repaint
10426         rect tracking on the main FrameView.
10427
10428         Later, when pixel results are requested, the alpha overlay
10429         that shows the repaint rects is generated using the list
10430         of rects supplied by WebCore.
10431
10432         * DumpRenderTree/mac/DumpRenderTree.mm:
10433         (resetDefaultsToConsistentValues):
10434         (resetWebViewToConsistentStateBeforeTesting):
10435         (displayWebView):
10436         * DumpRenderTree/mac/PixelDumpSupportMac.mm:
10437         (paintRepaintRectOverlay):
10438         (createBitmapContextFromWebView):
10439         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
10440         (WTR::InjectedBundle::beginTesting):
10441         (WTR::InjectedBundle::done):
10442         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
10443         (WTR::InjectedBundle::setRepaintRects):
10444         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
10445         (WTR::InjectedBundlePage::reset):
10446         (WTR::InjectedBundlePage::dump):
10447         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
10448         (WTR::LayoutTestController::display):
10449         * WebKitTestRunner/TestInvocation.cpp:
10450         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
10451         * WebKitTestRunner/TestInvocation.h:
10452         * WebKitTestRunner/cg/TestInvocationCG.cpp:
10453         (WTR::paintRepaintRectOverlay):
10454         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
10455
10456 2011-09-26  Nayan Kumar K  <nayankk@motorola.com>
10457
10458         Get hover'ed element URL from HitTest.
10459
10460         mouseDidMoveOverElement now gets the hover'ed element link using
10461         WKHitTestResult API's.
10462         https://bugs.webkit.org/show_bug.cgi?id=68426
10463
10464         Reviewed by Anders Carlsson.
10465
10466         * MiniBrowser/gtk/BrowserWindow.c:
10467         (mouseDidMoveOverElement):
10468
10469 2011-09-26  David Levin  <levin@chromium.org>
10470
10471         Add skeleton parsing for a WatchList.
10472         https://bugs.webkit.org/show_bug.cgi?id=68823
10473
10474         Reviewed by Adam Barth.
10475
10476         * Scripts/webkitpy/common/watchlist/watchlistparser.py: Added.
10477         Parses the top level watch list sections -- none are defined yet -- and
10478         throws an exception if an invalid one is listed.
10479         * Scripts/webkitpy/common/watchlist/watchlistparser_unittest.py: Added.
10480         Verify the exception for an invalid section in a watch list.
10481         * Scripts/webkitpy/common/watchlist/watchlist.py: Added.
10482
10483 2011-09-26  Adam Roben  <aroben@apple.com>
10484
10485         Clean up code imported from WebKitAPITest
10486
10487         Fixes <http://webkit.org/b/68799> WebViewDestruction tests and related code don't match
10488         TestWebKitAPI conventions
10489
10490         Reviewed by David Levin.
10491
10492         * TestWebKitAPI/Tests/WebKit/win/WebViewDestruction.cpp: Changed tests to use gtest macros
10493         and to share functionality via test fixtures. Prefixed Win32 API calls with ::. Updated
10494         namespace name.
10495
10496         (TestWebKitAPI::WebViewDestruction::SetUp):
10497         (TestWebKitAPI::WebViewDestruction::webViewCount):
10498         (TestWebKitAPI::WebViewDestructionWithHostWindow::SetUp):
10499         (TestWebKitAPI::WebViewDestruction::runMessagePump):
10500         (TestWebKitAPI::WebViewDestruction::TearDown):
10501         (TestWebKitAPI::WebViewDestructionWithHostWindow::TearDown):
10502         Moved functionality from free functions into these new test fixtures.
10503
10504         * TestWebKitAPI/win/HostWindow.cpp:
10505         * TestWebKitAPI/win/HostWindow.h:
10506         Prefixed Win32 API calls with ::. Updated namespace name.
10507
10508 2011-09-23  Adam Roben  <aroben@apple.com>
10509
10510         Merge WebKitAPITest into TestWebKitAPI
10511
10512         This also means that our formerly-WebKitAPITest tests will be run on the bots!
10513
10514         I made the minimal changes necessary to get the tests to run in TestWebKitAPI. I'll make
10515         them more idiomatic in a separate pass.
10516
10517         Fixes <http://webkit.org/b/66560> WebKitAPITest and TestWebKitAPI have duplicate
10518         functionality (and names!)
10519
10520         Reviewed by Anders Carlsson.
10521
10522         * TestWebKitAPI/Tests/WebKit/win/WebViewDestruction.cpp: Renamed from Tools/WebKitAPITest/tests/WebViewDestruction.cpp.
10523         * TestWebKitAPI/win/HostWindow.cpp: Renamed from Tools/WebKitAPITest/HostWindow.cpp.
10524         * TestWebKitAPI/win/HostWindow.h: Renamed from Tools/WebKitAPITest/HostWindow.h.
10525
10526         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Added new files.
10527
10528         * TestWebKitAPI/win/main.cpp:
10529         (main): Added a call to ::OleInitialize to make WebKit1 work. This came from WebKitAPITest.
10530
10531         * WebKitAPITest: Removed this directory and its contents.
10532
10533 2011-09-23  Adam Roben  <aroben@apple.com>
10534
10535         Remove two failing WebKitAPITest tests
10536
10537         The failures are covered by <http://webkit.org/b/33162> and <http://webkit.org/b/32868>.
10538
10539         Reviewed by Anders Carlsson.
10540
10541         * WebKitAPITest/tests/WebViewDestruction.cpp: Removed the NoCloseOrDestroyViewWindow and
10542         MainFrameAfterClose tests.
10543
10544 2011-09-25  Adam Barth  <abarth@webkit.org>
10545
10546         Finish removing PLATFORM(BREWMP) by removing associated code
10547         https://bugs.webkit.org/show_bug.cgi?id=68779
10548
10549         Reviewed by Sam Weinig.
10550
10551         * Scripts/webkitpy/common/config/build.py:
10552         * waf/build/settings.py:
10553
10554 2011-09-21  Ai Makabi  <makabi@google.com>
10555
10556         Provides a simple LRU cache class in Python.
10557
10558         https://bugs.webkit.org/show_bug.cgi?id=67823
10559
10560         Reviewed by Tony Chang.
10561
10562         * Scripts/webkitpy/common/lru_cache.py:
10563         * Scripts/webkitpy/common/lru_cache_unittest.py:
10564
10565 2011-09-25  Adam Barth  <abarth@webkit.org>
10566
10567         Remove PLATFORM(HAIKU) and associated code
10568         https://bugs.webkit.org/show_bug.cgi?id=68774
10569
10570         Reviewed by Sam Weinig.
10571
10572         * Scripts/webkitpy/common/config/build.py:
10573         * waf/build/settings.py:
10574
10575 2011-09-24  Adam Barth  <abarth@webkit.org>
10576
10577         Always enable ENABLE(OFFLINE_WEB_APPLICATIONS)
10578         https://bugs.webkit.org/show_bug.cgi?id=68767
10579
10580         Reviewed by Eric Seidel.
10581
10582         * Scripts/build-webkit:
10583
10584 2011-09-24  Adam Barth  <abarth@webkit.org>
10585
10586         Remove ENABLE(WCSS) and associated code
10587         https://bugs.webkit.org/show_bug.cgi?id=68759
10588
10589         Reviewed by Darin Adler.
10590
10591         * Scripts/build-webkit:
10592         * Scripts/old-run-webkit-tests:
10593         * Scripts/webkitperl/features.pm:
10594         (hasFeature):
10595         * Scripts/webkitpy/layout_tests/port/webkit.py:
10596         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
10597
10598 2011-09-24  Alejandro G. Castro  <alex@igalia.com>
10599
10600         Fixed GTK compilation after r95878, the operator== was defined
10601         twice when USE_WEBPROCESS_EVENT_SIMULATION is defined.
10602
10603         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
10604
10605 2011-09-23  Raphael Kubo da Costa  <kubo@profusion.mobi>
10606
10607         [DRT] Include the right config file for EFL's DRT.
10608         https://bugs.webkit.org/show_bug.cgi?id=67042
10609
10610         Reviewed by Martin Robinson.
10611
10612         Ports which use CMake as their buildsystem (such as the EFL one) also
10613         have config.h files, but they are named differently, so include the
10614         right one depending on the buildsystem being used.
10615
10616         * DumpRenderTree/config.h:
10617
10618 2011-09-23  Adam Klein  <adamk@chromium.org>
10619
10620         Add ENABLE_MUTATION_OBSERVERS feature flag
10621         https://bugs.webkit.org/show_bug.cgi?id=68732
10622
10623         Reviewed by Ojan Vafai.
10624
10625         This flag will guard an implementation of the "Mutation Observers" proposed in
10626         http://lists.w3.org/Archives/Public/public-webapps/2011JulSep/1622.html
10627
10628         * Scripts/build-webkit:
10629
10630 2011-09-23  Chang Shu  <cshu@webkit.org>
10631
10632         [WK2] [Qt] Implement MouseDown/MouseUp/MouseMoveTo functions for WebKit2 EventSender
10633         https://bugs.webkit.org/show_bug.cgi?id=68556
10634
10635         Implement the MouseDown/MouseUp/MouseMoveTo functions on Qt platform.
10636
10637         Reviewed by Darin Adler.
10638
10639         * WebKitTestRunner/EventSenderProxy.h:
10640         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
10641         * WebKitTestRunner/PlatformWebView.h:
10642         * WebKitTestRunner/mac/EventSenderProxy.mm:
10643         (WTR::EventSenderProxy::EventSenderProxy):
10644         (WTR::EventSenderProxy::leapForward):
10645         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
10646         (WTR::EventSenderProxy::EventSenderProxy):
10647         (WTR::getMouseButton):
10648         (WTR::getModifiers):
10649         (WTR::EventSenderProxy::updateClickCountForButton):
10650         (WTR::EventSenderProxy::createGraphicsSceneMouseEvent):
10651         (WTR::EventSenderProxy::mouseDown):
10652         (WTR::EventSenderProxy::mouseUp):
10653         (WTR::EventSenderProxy::mouseMoveTo):
10654         (WTR::EventSenderProxy::leapForward):
10655         (WTR::EventSenderProxy::sendOrQueueEvent):
10656         (WTR::EventSenderProxy::replaySavedEvents):
10657         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
10658         (WTR::PlatformWebView::postEvent):
10659         * WebKitTestRunner/qt/WebKitTestRunner.pro:
10660
10661 2011-09-23  Lei Zhang  <thestig@chromium.org>
10662
10663         [chromium] Make the layout test script's kill timeout proportional to --time-out-ms
10664         https://bugs.webkit.org/show_bug.cgi?id=68026
10665
10666         Reviewed by Dirk Pranke.
10667
10668         * Scripts/webkitpy/layout_tests/port/chromium.py:
10669
10670 2011-09-23  Elliot Poger  <epoger@google.com>
10671
10672         update layout_tests to account for new default of use_skia=1
10673         https://bugs.webkit.org/show_bug.cgi?id=68698
10674
10675         Rolls Source/WebKit/chromium/DEPS to pick up new default use_skia=1
10676         from http://src.chromium.org/viewvc/chrome?view=rev&revision=102532
10677         (Chromium-on-Mac now uses the Skia graphics library instead of
10678         Core Graphics).
10679
10680         Also updates layout_test code to work with that change.
10681
10682         Reviewed by Mihai Parparita.
10683
10684         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
10685         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
10686
10687 2011-09-23  Mark Rowe  <mrowe@apple.com>
10688
10689         Fix the build.
10690
10691         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp: Define operator<<(std::ostream&, const WTF::String&) inside
10692         the WTF namespace.
10693
10694 2011-09-22  Dirk Pranke  <dpranke@chromium.org>
10695
10696         enable NRWT for run-webkit-tests on Lion
10697         https://bugs.webkit.org/show_bug.cgi?id=68673
10698
10699         Reviewed by Eric Seidel.
10700
10701         * Scripts/run-webkit-tests:
10702         (useNewRunWebKitTests):
10703
10704 2011-09-22  Xianzhu Wang  <wangxianzhu@chromium.org>
10705
10706         Add unit test for existing StringBuilder
10707         https://bugs.webkit.org/show_bug.cgi?id=67080
10708
10709         Reviewed by Darin Adler.
10710
10711         * TestWebKitAPI/TestWebKitAPI.gypi:
10712         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
10713         * TestWebKitAPI/Tests/WTF/StringBuilder.cpp: Added.
10714         (operator<<):
10715         (TestWebKitAPI::expectBuilderContent):
10716         (TestWebKitAPI::expectEmpty):
10717         (TestWebKitAPI::TEST):
10718
10719 2011-09-22  Raphael Kubo da Costa  <kubo@profusion.mobi>
10720
10721         [EFL] Use ewk_view to navigate in history instead of ewk_history.
10722         https://bugs.webkit.org/show_bug.cgi?id=68455
10723
10724         Reviewed by Antonio Gomes.
10725
10726         BackForwardItem::invoke() was using ewk_history functions to navigate
10727         in history; however, this will only call WebCore::BackForwardListImpl
10728         and not move between pages at all. We now call ewk_view_navigate and
10729         its siblings to make sure the navigation actually happens.
10730
10731         This should make tests like fast/dom/navigation-type-back-forward.html
10732         stop timing out.
10733
10734         * DumpRenderTree/efl/WorkQueueItemEfl.cpp:
10735         (BackForwardItem::invoke):
10736
10737 2011-09-22  Dean Jackson  <dino@apple.com>
10738
10739         Add ENABLE_CSS_FILTERS
10740         https://bugs.webkit.org/show_bug.cgi?id=68652
10741
10742         Reviewed by Simon Fraser.
10743
10744         * Scripts/build-webkit:
10745
10746 2011-09-22  Anders Carlsson  <andersca@apple.com>
10747
10748         EventSenderProxy::mouseUp should call -[WKView mouseUp:] instead of -[WKView mouseDown:]
10749         https://bugs.webkit.org/show_bug.cgi?id=68660
10750
10751         Reviewed by Adam Barth.
10752
10753         * WebKitTestRunner/mac/EventSenderProxy.mm:
10754         (WTR::EventSenderProxy::mouseUp):
10755
10756 2011-09-22  Adam Roben  <aroben@apple.com>
10757
10758         Remove FindSafari
10759
10760         It isn't used or built anymore.
10761
10762         Fixes <http://webkit.org/b/68628>
10763
10764         Reviewed by Darin Adler.
10765
10766         * FindSafari: Removed this directory and its contents.
10767
10768 2011-09-22  Adam Roben  <aroben@apple.com>
10769
10770         Remove FindSafari from all our .sln files
10771
10772         It isn't used anymore, so there's no point in building it.
10773
10774         Part of <http://webkit.org/b/68628> Remove FindSafari
10775
10776         Reviewed by Steve Falkenburg.
10777
10778         * DumpRenderTree/DumpRenderTree.sln:
10779         * WebKitTestRunner/WebKitTestRunner.sln:
10780
10781 2011-09-22  Adam Roben  <aroben@apple.com>
10782
10783         Remove all uses of FindSafari
10784
10785         It isn't really needed anymore.
10786
10787         Fixes <http://webkit.org/b/68626>.
10788
10789         Reviewed by Darin Adler.
10790
10791         * Scripts/old-run-webkit-tests: Moved read/writeRegistryString from here to webkitdirs.pm.
10792
10793         * Scripts/run-webkit-nightly.cmd: Removed. This used to be used to launch nightly builds of
10794         WebKit, but now we use WebKit.exe for that.
10795
10796         * Scripts/webkitdirs.pm:
10797         (installedSafariPath): Changed to use readRegistryString instead of invoking FindSafari to
10798         do the same thing.
10799
10800         (readRegistryString):
10801         (writeRegistryString):
10802         Moved here from old-run-webkit-tests.
10803
10804 2011-09-22  Adam Roben  <aroben@apple.com>
10805
10806         Use the "direct object" form of system() to run WebKit.exe
10807
10808         This prevents the shell from parsing (and possibly splitting, if it contains spaces) the
10809         path to WebKit.exe.
10810
10811         Fixes <http://webkit.org/b/68623> run-safari doesn't work if there are spaces in the path to
10812         %WEBKITOUTPUTDIR%
10813
10814         Reviewed by Steve Falkenburg.
10815
10816         * Scripts/webkitdirs.pm:
10817         (runSafari):
10818
10819 2011-09-22  Adam Roben  <aroben@apple.com>
10820
10821         Quote the WEBKITLIBRARIESDIR path before passing it to the shell/cygpath
10822
10823         Fixes <http://webkit.org/b/68621> build-webkit goes totally haywire on Windows if there are
10824         spaces in the path to the WebKit source tree
10825
10826         Reviewed by John Sullivan.
10827
10828         * Scripts/webkitdirs.pm:
10829         (setupCygwinEnv):
10830
10831 2011-09-22  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
10832
10833         [Qt] Use same DPI for application font as rest of app in DRT and WTR
10834
10835         QApplication will initialize the default application font based
10836         on the application DPI at construction time, but we then override
10837         the application DPI using QX11Info (hard-coding it to 96 for
10838         consistency). This hard-coding is not reflected in the application
10839         font, so we explicitly have update the font ourselves.
10840
10841         The 6 test results that are updated were originally produced with
10842         a DPI of 75, as this is the default fallback DPI when a QFont is
10843         constructed before QApplication. This was wrong, and the results
10844         are updated to match a DPI of 96.
10845
10846         Reviewed by Csaba Osztrogonác.
10847
10848         * DumpRenderTree/qt/main.cpp:
10849         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
10850
10851 2011-09-21  Anna Cavender  <annacc@chromium.org>
10852
10853         Update committers.py with a new contributor contact
10854         https://bugs.webkit.org/show_bug.cgi?id=68565
10855
10856         Reviewed by Eric Carlson.
10857
10858         * Scripts/webkitpy/common/config/committers.py:
10859
10860 2011-09-21  Chang Shu  <cshu@webkit.org>
10861
10862         [WK2] [Mac] Implement a more-complete MouseDown/MouseUp/MouseMoveTo functions for WebKit2 EventSender
10863         https://bugs.webkit.org/show_bug.cgi?id=68108
10864
10865         This patch implements the event sender mouse events through WebKit2 UIProcess which is
10866         closer to the real simulation than a WebProcess-only approach. The patch only supports Mac
10867         platform as the first step so the existing code is still kept for other platforms.
10868
10869         Reviewed by Darin Adler.
10870
10871         * WebKitTestRunner/EventSenderProxy.h:
10872         (WTR::EventSenderProxy::EventSenderProxy):
10873         (WTR::EventSenderProxy::leapForward):
10874         (WTR::EventSenderProxy::currentEventTime):
10875         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
10876         (WTR::EventSendingController::EventSendingController):
10877         (WTR::EventSendingController::mouseDown):
10878         (WTR::EventSendingController::mouseUp):
10879         (WTR::EventSendingController::mouseMoveTo):
10880         (WTR::EventSendingController::leapForward):
10881         (WTR::EventSendingController::keyDown):
10882         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
10883         * WebKitTestRunner/TestController.cpp:
10884         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
10885         * WebKitTestRunner/mac/EventSenderProxy.mm:
10886         (WTR::eventTypeForMouseButtonAndAction):
10887         (WTR::operator==):
10888         (WTR::EventSenderProxy::updateClickCountForButton):
10889         (WTR::EventSenderProxy::mouseDown):
10890         (WTR::EventSenderProxy::mouseUp):
10891         (WTR::EventSenderProxy::mouseMoveTo):
10892         (WTR::EventSenderProxy::keyDown):
10893         * WebKitTestRunner/qt/EventSenderProxyQt.cpp:
10894         (WTR::EventSenderProxy::keyDown):
10895         (WTR::EventSenderProxy::mouseDown):
10896         (WTR::EventSenderProxy::mouseUp):
10897         (WTR::EventSenderProxy::mouseMoveTo):
10898
10899 2011-09-21  Dan Bernstein  <mitz@apple.com>
10900
10901         Tools part of: Prevent the WebKit frameworks from defining inappropriately-named Objective-C classes
10902         https://bugs.webkit.org/show_bug.cgi?id=68451
10903
10904         Reviewed by Darin Adler.
10905
10906         * Scripts/check-for-inappropriate-objc-class-names: Added. Checks for
10907         Objective-C classes with names not having one of a list of prefixes
10908         passed on the command line.
10909
10910 2011-09-21  Xianzhu Wang  <wangxianzhu@chromium.org>
10911
10912         Ensure TestWebKitAPI works on mac, win, chromium-mac and chromium-linux
10913         https://bugs.webkit.org/show_bug.cgi?id=68206
10914
10915         Reviewed by Tony Chang.
10916
10917         * Scripts/webkitdirs.pm:
10918         (determineBaseProductDir): add support for chromium-mac
10919         * TestWebKitAPI/Tests/WTF/MetaAllocator.cpp:
10920         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
10921
10922 2011-09-21  Adam Barth  <abarth@webkit.org>
10923
10924         garden-o-matic command should explain that it is starting a local server
10925         https://bugs.webkit.org/show_bug.cgi?id=68486
10926
10927         Reviewed by Dimitri Glazkov.
10928
10929         In the user study, both participants didn't understand why the
10930         garden-o-matic command appeared to hang.  This patch adds an
10931         explaination of what's going on.  It also adds a confirmation prompt
10932         before launching the browser to give the user a chance to read the
10933         text.  If that's too much of a speed bump, we can iterate.
10934
10935         * Scripts/webkitpy/tool/commands/gardenomatic.py:
10936
10937 2011-09-21  Chang Shu  <cshu@webkit.org>
10938
10939         [Qt] Implement KeyDown function for WebKit2 EventSender.
10940         https://bugs.webkit.org/show_bug.cgi?id=56485
10941
10942         Implement the KeyDown function on Qt platform.
10943
10944         Reviewed by Andreas Kling.
10945
10946         * WebKitTestRunner/PlatformWebView.h:
10947         * WebKitTestRunner/TestController.cpp:
10948         (WTR::TestController::TestController):
10949         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
10950         * WebKitTestRunner/qt/EventSenderProxyQt.cpp: Added.
10951         (WTR::getModifiers):
10952         (WTR::EventSenderProxy::keyDown):
10953         (buildModifierFlags):
10954         (EventSenderProxy::keyDown):
10955         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
10956         (WTR::PlatformWebView::sendEvent):
10957         * WebKitTestRunner/qt/WebKitTestRunner.pro:
10958
10959 2011-09-20  Adam Barth  <abarth@webkit.org>
10960
10961         garden-o-matic unit test fail because Safari 5.1 doesn't enumerate dictionaries in the same order as V8
10962         https://bugs.webkit.org/show_bug.cgi?id=68496
10963
10964         Reviewed by Dimitri Glazkov.
10965
10966         This patch adds a bunch of sorting to make sure things happen in a
10967         predictable order in both JavaScriptCore and V8.  We've talked before
10968         about whether these two JavaScript engines should enumerate
10969         dictionaries in the same order, but the consensus has been that they
10970         shouldn't.  Personally, I disagree with that consensus, but such is life.
10971
10972         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
10973         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
10974         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
10975         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
10976
10977 2011-09-20  Adam Barth  <abarth@webkit.org>
10978
10979         garden-o-matic doesn't work in Safari 5.1 because Safari 5.1 lacks Function.prototype.bind
10980         https://bugs.webkit.org/show_bug.cgi?id=68495
10981
10982         Reviewed by Dimitri Glazkov.
10983
10984         This patch adds an implementation of Function.prototype.bind if the browser lacks one.
10985
10986         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
10987
10988 2011-09-20  Eric Seidel  <eric@webkit.org>
10989
10990         [NRWT] REGRESSION: Local loader tests are failing on machines that lost /tmp/LayoutTests symlink
10991         https://bugs.webkit.org/show_bug.cgi?id=65781
10992
10993         Reviewed by Ryosuke Niwa.
10994
10995         Instead of making NRWT create the symlink, I made DumpRenderTree smart enough
10996         to resolve the passed in url relative to the absolute url for the test.
10997         For http tests, since the test url is an http url, we can't resolve relative
10998         to the test path, and thus use a new LOCAL_RESOURCE_ROOT environment variable
10999         for resolving.
11000
11001         I believe this is a better approach than the on used in the Qt and Chromium DRT's
11002         (which resolves the path relative to the built location of the DRT executable)
11003         and we should move this new code into a shared location in a follow-up patch.
11004
11005         It turns out that there was a second use for pathToLocalResource, used by one
11006         test (http/tests/plugins/post-url-file.html) used for getting a path to /tmp.
11007         To support this test I made the new pathToLocalResource smart enough to map
11008         /tmp to DUMPRENDERTREE_TEMP (which is already defined for all ports).
11009
11010         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
11011         (LayoutTestController::pathToLocalResource):
11012         * Scripts/webkitpy/layout_tests/port/webkit.py: Set LOCAL_RESOURCE_ROOT for use by DRT.
11013
11014 2011-09-20  Tom Zakrajsek  <tomz@codeaurora.org>
11015
11016         Need to specify a django version, instead of using default
11017         https://bugs.webkit.org/show_bug.cgi?id=68398
11018
11019         Reviewed by Eric Seidel.
11020
11021         * QueueStatusServer/__init__.py:
11022         * QueueStatusServer/main.py:
11023
11024 2011-09-20  Tom Zakrajsek  <tomz@codeaurora.org>
11025
11026         Update committers.py with a few more contributor contacts
11027         https://bugs.webkit.org/show_bug.cgi?id=68487
11028
11029         Reviewed by Eric Seidel.
11030
11031         * Scripts/webkitpy/common/config/committers.py:
11032
11033 2011-09-20  Leandro Pereira  <leandro@profusion.mobi>
11034
11035         Fix webkitpy unit tests after 95547 and 95549
11036         https://bugs.webkit.org/show_bug.cgi?id=68483
11037
11038         Reviewed by Adam Barth.
11039
11040         Fix tests for ChangeLog.touched_files() and ChangeLog.parse_entries_from_file().
11041
11042         * Scripts/webkitpy/common/checkout/changelog_unittest.py: While testing
11043         parse_entries_from_file(), use StringIO and entry.reviewer_text().
11044         While testing parse_latest_entry_from_file(), expect the correct
11045         touched_files() list.
11046
11047 2011-09-20  Adam Barth  <abarth@webkit.org>
11048
11049         garden-o-matic makes it hard to compare images when scroll bars appear
11050         https://bugs.webkit.org/show_bug.cgi?id=68482
11051
11052         Reviewed by Eric Seidel.
11053
11054         Because we autoscale image width, the images rescale if we overflow
11055         vertically and get a scrollbar.  That makes it hard to compare two sets
11056         of images.  This patch reserves space for the scrollbar so we won't
11057         have the layout change.
11058
11059         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css:
11060
11061 2011-09-20  Sheriff Bot  <webkit.review.bot@gmail.com>
11062
11063         Unreviewed, rolling out r95565.
11064         http://trac.webkit.org/changeset/95565
11065         https://bugs.webkit.org/show_bug.cgi?id=68470
11066
11067         Broke builds (Requested by andersca on #webkit).
11068
11069         * MiniBrowser/gtk/BrowserWindow.c:
11070         (mouseDidMoveOverElement):
11071
11072 2011-09-20  Nayan Kumar K  <nayankk@motorola.com>
11073
11074         Get hover'ed element URL from HitTest.
11075
11076         mouseDidMoveOverElement now gets the hover'ed element link using
11077         WKHitTestResult API's.
11078         https://bugs.webkit.org/show_bug.cgi?id=68426
11079
11080         Reviewed by Anders Carlsson.
11081
11082         * MiniBrowser/gtk/BrowserWindow.c:
11083         (mouseDidMoveOverElement):
11084
11085 2011-09-20  Leandro Pereira  <leandro@profusion.mobi>
11086
11087         Changelog class should have a method to return all entries
11088         https://bugs.webkit.org/show_bug.cgi?id=68399
11089
11090         Implement ChangeLog.parse_entries_from_file(). This method returns a generator
11091         of ChangeLogEntry objects, ordered from the latest to the oldest entry in the file.
11092
11093         Reviewed by Ryosuke Niwa.
11094
11095         * Scripts/webkitpy/common/checkout/changelog.py: Copy parse_latest_entry_from_file()
11096         and adapt it to become a generator.
11097         * Scripts/webkitpy/common/checkout/changelog_unittest.py: Add test case.
11098
11099 2011-09-20  Jarred Nicholls  <jarred@sencha.com>
11100
11101         [Qt] Permit qrc resources to load in QWebSettings::setUserStyleSheetUrl()
11102         https://bugs.webkit.org/show_bug.cgi?id=51159
11103
11104         Permit any URL with a local scheme to set a user style sheet. Allow qrc
11105         resource URLs as arguments to QWebSettings::setUserStyleSheetUrl().
11106
11107         Add DRT/WKTR qrc resource to test loading user style sheets from a qrc URL.
11108
11109         Reviewed by Kenneth Rohde Christiansen.
11110
11111         * DumpRenderTree/qt/DumpRenderTree.pro:
11112         * DumpRenderTree/qt/DumpRenderTree.qrc: Added.
11113         * DumpRenderTree/qt/resources/user.css: Added.
11114         (body):
11115         * WebKitTestRunner/qt/WebKitTestRunner.pro:
11116         * WebKitTestRunner/qt/WebKitTestRunner.qrc: Added.
11117         * WebKitTestRunner/qt/resources/user.css: Added.
11118         (body):
11119
11120 2011-09-20  Leandro Pereira  <leandro@profusion.mobi>
11121
11122         ChangeLogEntry class should contain a list of touched files
11123         https://bugs.webkit.org/show_bug.cgi?id=68115
11124
11125         Reviewed by Eric Seidel.
11126
11127         Add method to return the list of files touched in a ChangeLog entry.
11128
11129         * Scripts/webkitpy/common/checkout/changelog.py: Add
11130         touched_files_regexp and touched_files() method.
11131         * Scripts/webkitpy/common/checkout/changelog_unittest.py: Implement
11132         unit tests for touched_files() method.
11133
11134 2011-09-20  Sheriff Bot  <webkit.review.bot@gmail.com>
11135
11136         Unreviewed, rolling out r95526 and r95541.
11137         http://trac.webkit.org/changeset/95526
11138         http://trac.webkit.org/changeset/95541
11139         https://bugs.webkit.org/show_bug.cgi?id=68443
11140
11141         It broke 6 tests (Requested by ossy_ on #webkit).
11142
11143         * DumpRenderTree/qt/main.cpp:
11144         (main):
11145         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
11146         (WTR::activateFonts):
11147
11148 2011-09-20  Tom Zakrajsek  <tomz@codeaurora.org>
11149
11150         Fix patch and bug numbering in the mocktool simulation code
11151         https://bugs.webkit.org/show_bug.cgi?id=68374
11152
11153         Reviewed by Adam Barth.
11154
11155         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
11156         * Scripts/webkitpy/tool/bot/feeders_unittest.py:
11157         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
11158         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
11159         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
11160         * Scripts/webkitpy/tool/bot/queueengine_unittest.py:
11161         * Scripts/webkitpy/tool/commands/download_unittest.py:
11162         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
11163         * Scripts/webkitpy/tool/commands/queries_unittest.py:
11164         * Scripts/webkitpy/tool/commands/queues_unittest.py:
11165         * Scripts/webkitpy/tool/commands/queuestest.py:
11166         * Scripts/webkitpy/tool/commands/upload_unittest.py:
11167         * Scripts/webkitpy/tool/mocktool.py:
11168         * Scripts/webkitpy/tool/steps/steps_unittest.py:
11169         * Scripts/webkitpy/tool/steps/updatechangelogswithreview_unittest.py:
11170
11171 2011-09-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
11172
11173         [Qt] Don't override default QApplication font for DRT and WTR
11174
11175         Revision 95526 moved the relevant code below the QApplication
11176         constructor, to fix a potential crash, and that made the code
11177         take effect, breaking 6 tests.
11178
11179         We should rely on the QWebSettings to set default fonts, and
11180         this QApplication::setFont() is a relic from when we didn't
11181         have that option, so we remove the code.
11182
11183         Reviewed by Csaba Osztrogonác.
11184
11185         * DumpRenderTree/qt/main.cpp:
11186         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
11187
11188 2011-09-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
11189
11190         WebKitTestRunner: Provide usage if run without arguments
11191
11192         If WTR is run without either one or more filenames, or
11193         a '-' to indicate server mode, we don't need to start
11194         up the test machinery. Doing so might actually crash,
11195         as we'll exit the application immedeatly after anyways.
11196
11197         Reviewed by Simon Hausmann.
11198
11199         * WebKitTestRunner/TestController.cpp:
11200         (WTR::TestController::initialize):
11201
11202 2011-09-20  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
11203
11204         [Qt] Don't use QFont before initializing QApplication
11205
11206         Could potentially result in crash.
11207
11208         Reviewed by Simon Hausmann.
11209
11210         * DumpRenderTree/qt/main.cpp:
11211         (main):
11212
11213 2011-09-19  Adam Barth  <abarth@webkit.org>
11214
11215         Always enable ENABLE(EVENTSOURCE)
11216         https://bugs.webkit.org/show_bug.cgi?id=68414
11217
11218         Reviewed by Eric Seidel.
11219
11220         * Scripts/build-webkit:
11221
11222 2011-09-19  David Levin  <levin@chromium.org>
11223
11224         Sheriffbot rollout should be more intuitive.
11225         https://bugs.webkit.org/show_bug.cgi?id=68415
11226
11227         Reviewed by Adam Barth.
11228
11229         * Scripts/webkitpy/tool/bot/irc_command.py: Add support for revert and comma separated args.
11230         * Scripts/webkitpy/tool/bot/irc_command_unittest.py: Add parsing tests for comma separated args
11231           and a few others cases.
11232         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py: Verify that revert works.
11233
11234 2011-09-19  David Levin  <levin@chromium.org>
11235
11236         check-webkit-style generates bogus warning for StructuredExceptionHandlerSupressor.h
11237         https://bugs.webkit.org/show_bug.cgi?id=68391
11238
11239         Reviewed by Darin Adler.
11240
11241         * Scripts/webkitpy/style/checkers/cpp.py: Skip __asm lines when doing style checks on a line.
11242         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added a test for this.
11243
11244 2011-09-19  Alexis Menard  <alexis.menard@openbossa.org>
11245
11246         Style checker should not check public Qt WK2 API for parameter_name.
11247         https://bugs.webkit.org/show_bug.cgi?id=68359
11248
11249         Reviewed by Andreas Kling.
11250
11251         The public Qt WK2 APIs have different rules than WebKit for parameters of methods in h files.
11252         Add an exception in the style checker to disable the check. Also added the test for it.
11253
11254         * Scripts/webkitpy/style/checker.py:
11255         * Scripts/webkitpy/style/checker_unittest.py:
11256
11257 2011-09-19  Leandro Pereira  <leandro@profusion.mobi>
11258
11259         GTK DRT leaks memory while comparing history items
11260         https://bugs.webkit.org/show_bug.cgi?id=68086
11261
11262         Reviewed by Martin Robinson.
11263
11264         webkit_web_history_item_get_target() returns a g_strdup()'d string,
11265         so memory must be freed before compareHistoryItems() returns.
11266
11267         * DumpRenderTree/gtk/DumpRenderTree.cpp:
11268         (compareHistoryItems): Use GOwnPtr to manage memory automatically.
11269
11270 2011-09-19  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
11271
11272         Buildbot marks a nrwt bot red when tests are missing results
11273         https://bugs.webkit.org/show_bug.cgi?id=64812
11274
11275         Reviewed by Ryosuke Niwa.
11276
11277         * Scripts/webkitpy/layout_tests/controllers/manager.py:
11278         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
11279
11280 2011-09-19  Jessie Berlin  <jberlin@apple.com>
11281
11282         Sites that use history pushState or replaceState are recorded in history in Private Browsing
11283         mode.
11284         https://bugs.webkit.org/show_bug.cgi?id=68208
11285
11286         Reviewed by Brady Eidson.
11287
11288         Add a test that the WKContextDidNavigateWithNavigationDataCallback (which is invoked by
11289         updateGlobalHistory) is not called in Private Browsing when the page does history.pushState.
11290
11291         * TestWebKitAPI/Tests/WebKit2/PrivateBrowsingPushStateNoHistoryCallback.cpp: Added.
11292         (TestWebKitAPI::didNavigateWithNavigationData):
11293         This is a context history client callback, so it should not be called when in Private
11294         Browsing. Add a call to FAIL().
11295         (TestWebKitAPI::didSameDocumentNavigationForFrame):
11296         This is a page load client callback that happens after the didNavigateWithNavigationData
11297         callback. End the test.
11298         (TestWebKitAPI::TEST):
11299         Enable Private Browsing and load a page that uses history.pushState.
11300
11301         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
11302         Add PrivateBrowsingPushStateNoHistoryCallback.cpp and push-state.html.
11303         * TestWebKitAPI/Tests/WebKit2/push-state.html: Added.
11304         Ditto.
11305         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
11306         Ditto.
11307         * TestWebKitAPI/win/copy-resources.cmd:
11308         Ditto.
11309
11310 2011-09-16  Elliot Poger  <epoger@google.com>
11311
11312         add non-CG Webkit-Mac builds to flakiness dashboard
11313         https://bugs.webkit.org/show_bug.cgi?id=68265
11314
11315         Reviewed by Adam Barth.
11316
11317         As Chromium-on-Mac approaches the switchover from Core Graphics to
11318         Skia, we need to add the Skia version to the flakiness dashboard.
11319
11320         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js:
11321         * TestResultServer/static-dashboards/builders.js:
11322         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
11323
11324 2011-09-16  Jochen Eisinger  <jochen@chromium.org>
11325
11326         Rename FrameLoaderClient::allowImages to FrameLoaderClient::allowImage and include the image URL as parameter
11327         https://bugs.webkit.org/show_bug.cgi?id=68071
11328
11329         Reviewed by Adam Barth.
11330
11331         * DumpRenderTree/chromium/WebPermissions.cpp:
11332         (WebPermissions::allowImage):
11333         * DumpRenderTree/chromium/WebPermissions.h:
11334
11335 2011-09-16  Tom Zakrajsek  <tomz@codeaurora.org>
11336
11337         update-webkit-chromium should retry the gclient call if it fails once
11338         https://bugs.webkit.org/show_bug.cgi?id=68252
11339
11340         Reviewed by Adam Barth.
11341
11342         * Scripts/update-webkit-chromium:
11343
11344 2011-09-16  Rob Buis  <rbuis@rim.com>
11345
11346         Add Eli's rim address and change the nick.
11347
11348         * Scripts/webkitpy/common/config/committers.py:
11349
11350 2011-09-16  Anders Carlsson  <andersca@apple.com>
11351
11352         Add back the WKHitTestResultRef parameter to mouseDidMoveOverElement in an ABI preserving way
11353         https://bugs.webkit.org/show_bug.cgi?id=68269
11354
11355         Reviewed by Geoffrey Garen.
11356
11357         Update client function signatures.
11358
11359         * MiniBrowser/gtk/BrowserWindow.c:
11360         (browserWindowUIClientInit):
11361         * MiniBrowser/mac/BrowserWindowController.m:
11362         (mouseDidMoveOverElement):
11363         (-[BrowserWindowController awakeFromNib]):
11364         * MiniBrowser/win/BrowserView.cpp:
11365         (BrowserView::create):
11366         * WebKitTestRunner/TestController.cpp:
11367         (WTR::TestController::createOtherPage):
11368         (WTR::TestController::initialize):
11369
11370 2011-09-16  Anders Carlsson  <andersca@apple.com>
11371
11372         Temporarily remove the WKHitTestResultRef parameter to avoid breaking the WK2 ABI/API
11373         https://bugs.webkit.org/show_bug.cgi?id=68266
11374
11375         Reviewed by John Sullivan.
11376
11377         * MiniBrowser/mac/BrowserWindowController.m:
11378         (mouseDidMoveOverElement):
11379
11380 2011-09-16  Vincent Scheib  <scheib@chromium.org>
11381
11382         Add Vincent Scheib to the committers list.
11383
11384         * Scripts/webkitpy/common/config/committers.py:
11385
11386 2011-09-16  Sheriff Bot  <webkit.review.bot@gmail.com>
11387
11388         Unreviewed, rolling out r95285.
11389         http://trac.webkit.org/changeset/95285
11390         https://bugs.webkit.org/show_bug.cgi?id=68247
11391
11392         It broke a webkitpy test (Requested by ossy on #webkit).
11393
11394         * Scripts/webkitpy/layout_tests/controllers/manager.py:
11395
11396 2011-09-16  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
11397
11398         Buildbot marks a nrwt bot red when tests are missing results
11399         https://bugs.webkit.org/show_bug.cgi?id=64812
11400
11401         Reviewed by Dirk Pranke.
11402
11403         * Scripts/webkitpy/layout_tests/controllers/manager.py:
11404
11405 2011-09-16  Martin Robinson  <mrobinson@igalia.com>
11406
11407         [GTK] Disable WebKit2 by default in configure.ac, but enable it with build-webkit
11408         https://bugs.webkit.org/show_bug.cgi?id=68178
11409
11410         Reviewed by Xan Lopez.
11411
11412         Add a work-around to enable WebKit2 when using build-webkit.
11413
11414         * Scripts/webkitdirs.pm:
11415         (buildAutotoolsProject):
11416
11417 2011-09-15  James Robinson  <jamesr@chromium.org>
11418
11419         Add Antoine Labour and Shawn Singh to the contributors-but-not-committers list.
11420
11421         * Scripts/webkitpy/common/config/committers.py:
11422
11423 2011-09-15  Tom Zakrajsek  <tomz@codeaurora.org>
11424
11425         UpdateChangeLogsWithReviewer should be able to determine reviewer of latest reviewed patch in a bug
11426         https://bugs.webkit.org/show_bug.cgi?id=68003
11427
11428         Reviewed by Adam Barth.
11429
11430         * Scripts/webkitpy/tool/mocktool.py:
11431         * Scripts/webkitpy/tool/steps/updatechangelogswithreview_unittest.py:
11432         * Scripts/webkitpy/tool/steps/updatechangelogswithreviewer.py:
11433
11434 2011-09-15  Tom Zakrajsek  <tomz@codeaurora.org>
11435
11436         Python version check is confusing in test-webkitpy
11437         https://bugs.webkit.org/show_bug.cgi?id=68004
11438
11439         Reviewed by Adam Barth.
11440
11441         * Scripts/test-webkitpy:
11442
11443 2011-09-15  Eric Seidel  <eric@webkit.org>
11444
11445         Reviewed by Adam Barth.
11446
11447         webkit-patch should be able to find users and add them to bugzilla groups
11448         https://bugs.webkit.org/show_bug.cgi?id=63351
11449
11450         These are both very basic commands.  But it's now possible to find
11451         all users matching a regexp, as well as add all users matching a regexp
11452         to a set of groups.
11453
11454         bugzilla.py already knew how to find users (for validate-committer-lists)
11455         but now it has the ability to modify the user records.
11456
11457         I split some of the logic out into a new EditUsersParser class
11458         to try and reduce the amount of code in Bugzilla/BugzillaQueries.
11459
11460         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
11461         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
11462         * Scripts/webkitpy/tool/commands/__init__.py:
11463         * Scripts/webkitpy/tool/commands/adduserstogroups.py: Added.
11464         * Scripts/webkitpy/tool/commands/findusers.py: Added.
11465
11466 2011-09-15  Eric Seidel  <eric@webkit.org>
11467
11468         Remove ENABLE(SVG_AS_IMAGE) since all major ports have it on by default
11469         https://bugs.webkit.org/show_bug.cgi?id=68182
11470
11471         Reviewed by Adam Barth.
11472
11473         * Scripts/build-webkit:
11474         * waf/build/settings.py:
11475
11476 2011-09-15  Eric Seidel  <eric@webkit.org>
11477
11478         Unreviewed.  Make contributor email look-up case-insensitive.
11479
11480         validate-committer-list was incorrectly reporting that "chang.shu@nokia.com"
11481         was missing from committers.py due to case sensitivity.
11482
11483         This also includes a test expectations update to committers_unittest.py
11484         after my previous change.
11485
11486         * Scripts/webkitpy/common/config/committers.py:
11487         * Scripts/webkitpy/common/config/committers_unittest.py:
11488         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
11489
11490 2011-09-15  Eric Seidel  <eric@webkit.org>
11491
11492         Unreviewed.  Updated this based on webkit-patch suggest-nominations and validate-committer-list output.
11493
11494         Add a bunch of email aliases to committers.py for regular contributors
11495         who are committing using other email addresses than they have listed in the file.
11496
11497         * Scripts/webkitpy/common/config/committers.py:
11498
11499 2011-09-15  Eric Seidel  <eric@webkit.org>
11500
11501         Remove ENABLE_SVG_ANIMATION as all major ports have it on by default
11502         https://bugs.webkit.org/show_bug.cgi?id=68022
11503
11504         Reviewed by Ryosuke Niwa.
11505
11506         * Scripts/build-webkit:
11507         * waf/build/settings.py:
11508
11509 2011-09-15  Anders Carlsson  <andersca@apple.com>
11510
11511         Fix Mac build by including config.h where needed.
11512
11513         * TestWebKitAPI/Tests/TestWebKitAPI/mac/InstanceMethodSwizzler.mm:
11514         * TestWebKitAPI/Tests/WTF/MetaAllocator.cpp:
11515         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp:
11516         * TestWebKitAPI/Tests/WebKit2/WebArchive.cpp:
11517         * TestWebKitAPI/Tests/WebKit2/WebArchive_Bundle.cpp:
11518         * TestWebKitAPI/Tests/mac/DOMRangeOfString.mm:
11519         * TestWebKitAPI/Tests/mac/DeviceScaleFactorOnBack.mm:
11520         * TestWebKitAPI/Tests/mac/DynamicDeviceScaleFactor.mm:
11521         * TestWebKitAPI/Tests/mac/StringByEvaluatingJavaScriptFromString.mm:
11522         * TestWebKitAPI/mac/InstanceMethodSwizzler.mm:
11523         * TestWebKitAPI/mac/JavaScriptTestMac.mm:
11524         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
11525         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
11526         * TestWebKitAPI/mac/SyntheticBackingScaleFactorWindow.m:
11527         * TestWebKitAPI/mac/WebKitAgnosticTest.mm:
11528         * TestWebKitAPI/mac/main.mm:
11529
11530 2011-09-15  Xianzhu Wang  <wangxianzhu@chromium.org>
11531
11532         Let TestWebKitAPI work for chromium
11533         https://bugs.webkit.org/show_bug.cgi?id=67756
11534
11535         Reviewed by Sam Weinig.
11536
11537         * Scripts/build-api-tests:
11538         * Scripts/run-api-tests:
11539         (isSupportedPlatform):
11540         (runTest):
11541         (populateTests):
11542         * Scripts/webkitdirs.pm:
11543         (determineBaseProductDir):
11544         * TestWebKitAPI/Configurations/InjectedBundle.xcconfig:
11545         * TestWebKitAPI/Configurations/TestWebKitAPI.xcconfig:
11546         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
11547         * TestWebKitAPI/InjectedBundleController.cpp:
11548         * TestWebKitAPI/InjectedBundleMain.cpp:
11549         * TestWebKitAPI/JavaScriptTest.cpp:
11550         * TestWebKitAPI/PlatformUtilities.cpp:
11551         * TestWebKitAPI/TestWebKitAPI.gypi: Added.
11552         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
11553         * TestWebKitAPI/TestWebKitAPIPrefix.h: Removed.
11554         * TestWebKitAPI/Tests/WTF/CheckedArithmeticOperations.cpp:
11555         * TestWebKitAPI/Tests/WTF/StringOperators.cpp:
11556         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
11557         * TestWebKitAPI/Tests/WTF/VectorReverse.cpp:
11558         * TestWebKitAPI/Tests/WebKit2/AboutBlankLoad.cpp:
11559         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest.cpp:
11560         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest_Bundle.cpp:
11561         * TestWebKitAPI/Tests/WebKit2/CookieManager.cpp:
11562         * TestWebKitAPI/Tests/WebKit2/DocumentStartUserScriptAlertCrash.cpp:
11563         * TestWebKitAPI/Tests/WebKit2/DocumentStartUserScriptAlertCrash_Bundle.cpp:
11564         * TestWebKitAPI/Tests/WebKit2/DownloadDecideDestinationCrash.cpp:
11565         * TestWebKitAPI/Tests/WebKit2/EvaluateJavaScript.cpp:
11566         * TestWebKitAPI/Tests/WebKit2/FailedLoad.cpp:
11567         * TestWebKitAPI/Tests/WebKit2/Find.cpp:
11568         * TestWebKitAPI/Tests/WebKit2/ForceRepaint.cpp:
11569         * TestWebKitAPI/Tests/WebKit2/FrameMIMETypeHTML.cpp:
11570         * TestWebKitAPI/Tests/WebKit2/FrameMIMETypePNG.cpp:
11571         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp:
11572         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle_Bundle.cpp:
11573         * TestWebKitAPI/Tests/WebKit2/InjectedBundleBasic.cpp:
11574         * TestWebKitAPI/Tests/WebKit2/InjectedBundleBasic_Bundle.cpp:
11575         * TestWebKitAPI/Tests/WebKit2/LoadCanceledNoServerRedirectCallback.cpp:
11576         * TestWebKitAPI/Tests/WebKit2/LoadCanceledNoServerRedirectCallback_Bundle.cpp:
11577         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash.cpp:
11578         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash_Bundle.cpp:
11579         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
11580         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
11581         * TestWebKitAPI/Tests/WebKit2/PreventEmptyUserAgent.cpp:
11582         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp:
11583         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly_Bundle.cpp:
11584         * TestWebKitAPI/Tests/WebKit2/RestoreSessionStateContainingFormData.cpp:
11585         * TestWebKitAPI/Tests/WebKit2/SpacebarScrolling.cpp:
11586         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
11587         * TestWebKitAPI/Tests/WebKit2/WKString.cpp:
11588         * TestWebKitAPI/Tests/WebKit2/WKStringJSString.cpp:
11589         * TestWebKitAPI/Tests/WebKit2/win/AltKeyGeneratesWMSysCommand.cpp:
11590         * TestWebKitAPI/Tests/WebKit2/win/DoNotCopyANullCFURLResponse.cpp:
11591         * TestWebKitAPI/Tests/WebKit2/win/HideFindIndicator.cpp:
11592         * TestWebKitAPI/Tests/WebKit2/win/ResizeViewWhileHidden.cpp:
11593         * TestWebKitAPI/Tests/WebKit2/win/TranslateMessageGeneratesWMChar.cpp:
11594         * TestWebKitAPI/Tests/WebKit2/win/WMCloseCallsUIClientClose.cpp:
11595         * TestWebKitAPI/Tests/WebKit2/win/WMPrint.cpp:
11596         * TestWebKitAPI/TestsController.cpp:
11597         * TestWebKitAPI/config.h: Copied from TestWebKitAPI/TestWebKitAPIPrefix.h.
11598         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp:
11599         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
11600         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
11601         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
11602         * TestWebKitAPI/win/main.cpp:
11603
11604 2011-09-15  Jochen Eisinger  <jochen@chromium.org>
11605
11606         [chromium] extract normalization of LayoutTests URLs to a method of TestShell
11607         https://bugs.webkit.org/show_bug.cgi?id=68145
11608
11609         This will allow for other classes in DumpRenderTree to reuse this functionality
11610
11611         Reviewed by David Levin.
11612
11613         * DumpRenderTree/chromium/TestShell.cpp:
11614         (normalizeLayoutTestURLInternal):
11615         (dumpHistoryItem):
11616         (TestShell::normalizeLayoutTestURL):
11617         * DumpRenderTree/chromium/TestShell.h:
11618
11619 2011-09-14  Ada Chan  <adachan@apple.com>
11620
11621         Add test for WKBundleFrameCopyWebArchive().
11622         https://bugs.webkit.org/show_bug.cgi?id=67857
11623
11624         Reviewed by Anders Carlsson.
11625
11626         * TestWebKitAPI/Tests/WebKit2/WebArchive.cpp: Added.
11627         (TestWebKitAPI::didReceiveMessageFromInjectedBundle): Do some basic sanity checks to make sure
11628         the returned data is a valid webarchive.
11629         (TestWebKitAPI::setInjectedBundleClient):
11630         (TestWebKitAPI::didFinishLoadForFrame):
11631         (TestWebKitAPI::TEST): Load simple.html in a WebView.  Wait for it to finish loading, and then ask the
11632         injected bundle for the webarchive and check the returned webarchive.
11633         * TestWebKitAPI/Tests/WebKit2/WebArchive_Bundle.cpp: Added.
11634         (TestWebKitAPI::WebArchiveTest::WebArchiveTest):
11635         (TestWebKitAPI::WebArchiveTest::didReceiveMessage): Handle the "GetWebArchive" message and return
11636         the web archive.
11637
11638         Add files to project.
11639         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Also let Visual Studio reorder a file.
11640         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
11641         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
11642
11643 2011-09-14  Ryosuke Niwa  <rniwa@webkit.org>
11644
11645         committers.py should support multiple IRC nicknames
11646         https://bugs.webkit.org/show_bug.cgi?id=68110
11647
11648         Reviewed by Eric Seidel.
11649
11650         Add multiple IRC nickname support to committers.py
11651
11652         * Scripts/webkitpy/common/config/committers.py:
11653         * Scripts/webkitpy/common/config/committers_unittest.py:
11654         * Scripts/webkitpy/tool/bot/irc_command.py:
11655         * Scripts/webkitpy/tool/bot/sheriff.py:
11656
11657 2011-09-14  Jon Lee  <jonlee@apple.com>
11658
11659         Implement globalFlag for WebKitTestRunner
11660         https://bugs.webkit.org/show_bug.cgi?id=68033
11661         <rdar://problem/10119204>
11662
11663         Reviewed by Sam Weinig.
11664
11665         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Porting globalFlag from DRT
11666         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
11667         (WTR::LayoutTestController::LayoutTestController): initialize m_globalFlag to false
11668         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: add m_globalFlag
11669         (WTR::LayoutTestController::globalFlag): simple getter
11670         (WTR::LayoutTestController::setGlobalFlag): simple setter
11671
11672 2011-09-13  Eric Seidel  <eric@webkit.org>
11673
11674         Remove ENABLE_SVG_USE as <use> is required by HTML5
11675         https://bugs.webkit.org/show_bug.cgi?id=68019
11676
11677         Reviewed by Ryosuke Niwa.
11678
11679         * Scripts/build-webkit:
11680         * waf/build/settings.py:
11681
11682 2011-09-14  Ryosuke Niwa  <rniwa@webkit.org>
11683
11684         Remove Jan Erik Hanssen (added in my r95078) from the list for now
11685         since I can't confirm acceptable his committership.
11686
11687         * Scripts/webkitpy/common/config/committers.py:
11688
11689 2011-09-14  Jochen Eisinger  <jochen@chromium.org>
11690
11691         [chromium] Remove obsolete WebViewClient methods from DumpRenderTree.
11692         https://bugs.webkit.org/show_bug.cgi?id=68066
11693
11694         Reviewed by Adam Barth.
11695
11696         * DumpRenderTree/chromium/WebViewHost.cpp:
11697         * DumpRenderTree/chromium/WebViewHost.h:
11698
11699 2011-09-13  Ryosuke Niwa  <rniwa@webkit.org>
11700
11701         committers.py is missing some contributors
11702         https://bugs.webkit.org/show_bug.cgi?id=68059
11703
11704         Reviewed by Adam Barth.
11705
11706         Add David Barr, Jan Erik Hanssen, and Kim Grönholm.
11707
11708         Also add nicks of Alejandro G. Castro, Collin Jackson, and Matt Lilek.
11709
11710         * Scripts/webkitpy/common/config/committers.py:
11711
11712 2011-09-13  Ryosuke Niwa  <rniwa@webkit.org>
11713
11714         Fix webkit python test after r95061.
11715
11716         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
11717
11718 2011-09-13  Xianzhu Wang  <wangxianzhu@chromium.org>
11719
11720         Add myself as a contributor. No review needed.
11721
11722         * Scripts/webkitpy/common/config/committers.py:
11723
11724 2011-09-13  Adam Barth  <abarth@webkit.org>
11725
11726         garden-o-matic has unfriendly error messages when the local server is not available
11727         https://bugs.webkit.org/show_bug.cgi?id=68042
11728
11729         Reviewed by Dimitri Glazkov.
11730
11731         This patch moves us from using alert to showing the error inline in the
11732         progress dialog.
11733
11734         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
11735         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
11736         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
11737         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
11738         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
11739
11740 2011-09-13  Adam Barth  <abarth@webkit.org>
11741
11742         FailureGrid in garden-o-matic should link to the builder's waterfall page
11743         https://bugs.webkit.org/show_bug.cgi?id=68036
11744
11745         Reviewed by Dimitri Glazkov.
11746
11747         One of the users in the user study thought the configurations names in
11748         the failure grid would link to the underlying bots.  That makes sense,
11749         so this patch adds the behavior.
11750
11751         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js:
11752         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
11753         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
11754         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css:
11755
11756 2011-09-13  Adam Barth  <abarth@webkit.org>
11757
11758         garden-o-matic should be able to mark failures as expected
11759         https://bugs.webkit.org/show_bug.cgi?id=68027
11760
11761         Reviewed by Dimitri Glazkov.
11762
11763         All the back-end infrastructure exists for this operation.  This patch
11764         just adds some UI to the front-end and wires that UI up to the back
11765         end.  This exact UI will likely need some more polish, but this should
11766         be a reasonable first cut.
11767
11768         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
11769         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
11770         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
11771         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
11772         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
11773
11774 2011-09-13  Adam Barth  <abarth@webkit.org>
11775
11776         garden-o-matic should tell you about the progress of your rebaseline
11777         https://bugs.webkit.org/show_bug.cgi?id=67967
11778
11779         Reviewed by Dimitri Glazkov.
11780
11781         This patch adds ui.MessageBox, which is a light wrapper around jQuery
11782         UI's dialog widget.  We then use ui.MessageBox to show status messages
11783         about in-progress rebaselines.  As each baseline is downloaded, the
11784         MessageBox updates.  This UI replaces our temporary "alert"-based UI.
11785
11786         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
11787         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
11788         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
11789         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
11790         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
11791         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
11792
11793 2011-09-13  Leandro Pereira  <leandro@profusion.mobi>
11794
11795         GTK DumpRenderTree uses inefficient idioms to iterate over G[S]Lists
11796         https://bugs.webkit.org/show_bug.cgi?id=68024
11797
11798         Reviewed by Gustavo Noronha Silva.
11799
11800         Using g_list_count() and g_list_nth_data() together on a loop is
11801         inneficient since they're both O(n). Iterate over lists in a saner
11802         way.
11803
11804         * DumpRenderTree/gtk/DumpRenderTree.cpp:
11805         (dumpHistoryItem): Reduce the scope for the 'kids' variable, and
11806         iterate on it using g_list_next(). Free the list after done with it.
11807         (dumpBackForwardListForWebView): Instead of appending (which is
11808         expensive in GLists) history items and then iterating from the tail
11809         of the itemsToPrint list, prepend items and walk forwards as usual.
11810         (dumpBackForwardListForAllWebViews): Walk the list in a saner way,
11811         remove the (unneeded) viewList variable.
11812
11813 2011-09-13  Ryosuke Niwa  <rniwa@webkit.org>
11814
11815         Add Eric's IRC nick.
11816
11817         * Scripts/webkitpy/common/config/committers.py:
11818
11819 2011-09-13  Anders Carlsson  <andersca@apple.com>
11820
11821         Disable C++ exceptions when building with clang
11822         https://bugs.webkit.org/show_bug.cgi?id=68031
11823         <rdar://problem/9556880>
11824
11825         Reviewed by Mark Rowe.
11826
11827         * MiniBrowser/Configurations/Base.xcconfig:
11828         * TestWebKitAPI/Configurations/Base.xcconfig:
11829         * WebKitTestRunner/Configurations/Base.xcconfig:
11830
11831 2011-09-13  Ojan Vafai  <ojan@chromium.org>
11832
11833         [chromium] fix flakiness dashboard for chromium interactive tests
11834         https://bugs.webkit.org/show_bug.cgi?id=68029
11835
11836         Reviewed by Tony Chang.
11837
11838         * TestResultServer/static-dashboards/flakiness_dashboard.html:
11839         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
11840
11841 2011-09-13  Ryosuke Niwa  <rniwa@webkit.org>
11842
11843         Add Alice and Aryeh to the list of contributors.
11844
11845         * Scripts/webkitpy/common/config/committers.py:
11846
11847 2011-09-13  Ryosuke Niwa  <rniwa@webkit.org>
11848
11849         sheriffbot whois should also tell us email addresses
11850         https://bugs.webkit.org/show_bug.cgi?id=67975
11851
11852         Reviewed by Eric Seidel and David Levin.
11853
11854         Taught sheriffbot how to tell us contributor's email addresses.
11855
11856         * Scripts/webkitpy/tool/bot/irc_command.py:
11857         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
11858
11859 2011-09-13  Eric Seidel  <eric@webkit.org>
11860
11861         Remove ENABLE_SVG_FOREIGN_OBJECT as it is a required part of HTML5
11862         https://bugs.webkit.org/show_bug.cgi?id=68018
11863
11864         Reviewed by Ryosuke Niwa.
11865
11866         * Scripts/build-webkit:
11867         * waf/build/settings.py:
11868
11869 2011-09-13  Chang Shu  <cshu@webkit.org>
11870
11871         [WK2] [Mac] Implement KeyDown function for WebKit2 EventSender.
11872         https://bugs.webkit.org/show_bug.cgi?id=57515
11873
11874         The code change in WebKitTestRunner first sends keydown event from InjectedBundle to
11875         TestController synchronously. The latter then sends simulated event to WebKit2 module,
11876         which is synchronous, too.
11877
11878         Reviewed by Darin Adler.
11879
11880         * WebKitTestRunner/EventSenderProxy.h: Added.
11881         (WTR::EventSenderProxy::EventSenderProxy):
11882         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
11883         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
11884         (WTR::EventSendingController::keyDown):
11885         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
11886         * WebKitTestRunner/PlatformWebView.h:
11887         (WTR::PlatformWebView::platformWindow):
11888         * WebKitTestRunner/TestController.cpp:
11889         (WTR::TestController::TestController):
11890         (WTR::TestController::didReceiveSynchronousMessageFromInjectedBundle):
11891         * WebKitTestRunner/TestController.h:
11892         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
11893         * WebKitTestRunner/mac/EventSenderProxy.mm: Added.
11894         (WTR::buildModifierFlags):
11895         (WTR::EventSenderProxy::keyDown):
11896
11897 2011-09-12  Leandro Pereira  <leandro@profusion.mobi>
11898
11899         [EFL] DRT: Add DumpRenderTreeEfl.h
11900         https://bugs.webkit.org/show_bug.cgi?id=63993
11901
11902         Reviewed by Eric Seidel.
11903
11904         This header contains some global variables used by EFL's
11905         DumpRenderTree implementation.
11906
11907         * DumpRenderTree/DumpRenderTree.h: Include DumpRenderTreeEfl.h when
11908         appropriate.
11909         * DumpRenderTree/efl/DumpRenderTreeEfl.h: Added.
11910
11911 2011-09-12  Raphael Kubo da Costa  <kubo@profusion.mobi>
11912
11913         [EFL] Send the right key names for PageUp and PageDown.
11914         https://bugs.webkit.org/show_bug.cgi?id=67044
11915
11916         Reviewed by Kenneth Rohde Christiansen.
11917
11918         PlatformKeyboardEventEfl uses "Prior" and "Next" to refer to PageUp and
11919         PageDown (which is close to the defines in WindowsKeyboardCodes.h).
11920         However, EventSender was sending "Page_Up" and "Page_Down", so tests
11921         such as editing/input/page-up-down-scrolls.html were timing out due to
11922         the right key down events not being sent.
11923
11924         * DumpRenderTree/efl/EventSender.cpp:
11925         (keyPadNameFromJSValue):
11926         (keyNameFromJSValue):
11927
11928 2011-09-12  Eric Seidel  <eric@webkit.org>
11929
11930         sheriffbot whois is case sensitive
11931         https://bugs.webkit.org/show_bug.cgi?id=67972
11932
11933         Reviewed by Adam Barth.
11934
11935         I also fixed a typo in the "More than 5 contributors" message
11936         from my previous commit.
11937
11938         * Scripts/webkitpy/common/config/committers.py:
11939         * Scripts/webkitpy/common/config/committers_unittest.py:
11940         * Scripts/webkitpy/tool/bot/irc_command.py:
11941         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
11942
11943 2011-09-12  Eric Seidel  <eric@webkit.org>
11944
11945         sheriffbot whois should be smarter
11946         https://bugs.webkit.org/show_bug.cgi?id=67971
11947
11948         Reviewed by Adam Barth.
11949
11950         This makes sheriffbot whois basically grep the committers list
11951         making it much easier to find a contributor by name.
11952
11953         Unfortunately the current unittests depend on the real committers.py
11954         file, so as we add more eric's to the project, this unittest result will
11955         change.  I think that's OK for now.  We'll fix it when it breaks.
11956
11957         * Scripts/webkitpy/common/config/committers.py:
11958         * Scripts/webkitpy/tool/bot/irc_command.py:
11959         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
11960
11961 2011-09-12  Sheriff Bot  <webkit.review.bot@gmail.com>
11962
11963         Unreviewed, rolling out r94976.
11964         http://trac.webkit.org/changeset/94976
11965         https://bugs.webkit.org/show_bug.cgi?id=67953
11966
11967         Broke a bunch of security/load& tests (Requested by rniwa on
11968         #webkit).
11969
11970         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
11971         (LayoutTestController::pathToLocalResource):
11972
11973 2011-09-12  Eric Seidel  <eric@webkit.org>
11974
11975         Rubber-stamped by Adam Barth.
11976
11977         Remove the "the tree is burning!" warning previously shown
11978         when landing patches.  It was a dream.  But it's just noise now.
11979
11980         Some bots will require a restart to notice the removal of the option
11981         so I've left the --ignore-builders option in for now.
11982
11983         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
11984         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
11985         * Scripts/webkitpy/tool/commands/download.py:
11986         * Scripts/webkitpy/tool/commands/download_unittest.py:
11987         * Scripts/webkitpy/tool/commands/queues_unittest.py:
11988         * Scripts/webkitpy/tool/steps/__init__.py:
11989         * Scripts/webkitpy/tool/steps/commit.py:
11990         * Scripts/webkitpy/tool/steps/ensurebuildersaregreen.py: Removed.
11991         * Scripts/webkitpy/tool/steps/postdiffforrevert.py:
11992
11993 2011-09-12  Eric Seidel  <eric@webkit.org>
11994
11995         [NRWT] REGRESSION: Local loader tests are failing on machines that lost /tmp/LayoutTests symlink
11996         https://bugs.webkit.org/show_bug.cgi?id=65781
11997
11998         Reviewed by Ryosuke Niwa.
11999
12000         Instead of making NRWT create the symlink, I just made DumpRenderTree smart enough
12001         to resolve the passed in url relative to the absolute url for the test.
12002
12003         I believe this is a better approach than the on used in the Qt and Chromium DRT's
12004         (which resolves the path relative to the built location of the DRT executable)
12005         and we should move this new code into a shared location in a follow-up patch.
12006
12007         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
12008         (LayoutTestController::pathToLocalResource):
12009
12010 2011-09-12  Eric Seidel  <eric@webkit.org>
12011
12012         Reshuffle some code in WebKitDriver._read_block in preparation for reading stderr/stdout separately
12013         https://bugs.webkit.org/show_bug.cgi?id=67530
12014
12015         Reviewed by Adam Barth.
12016
12017         I believe I've fixed the bug in the original patch which prompted the rollout.
12018         The previous patch was using the wrong deadline for the initial read,
12019         subtracting time.time() twice from the deadline value.
12020
12021         * Scripts/webkitpy/layout_tests/port/webkit.py:
12022         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
12023
12024 2011-09-12  Balazs Kelemen  <kbalazs@webkit.org>
12025
12026         [Qt][WK2] WebKitTestRunner does not produce crash logs
12027         https://bugs.webkit.org/show_bug.cgi?id=67714
12028
12029         Reviewed by Chang Shu.
12030
12031         Added a simple way of generating backtrace on crash
12032         to the web process. The implementation is similar what
12033         we have in DRT. It depends on GNU libc functionality
12034         so it is only enabled where we are running in such an environment.
12035
12036         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
12037         (WTR::printBacktrace):
12038         (WTR::crashHandler):
12039         (WTR::InjectedBundle::platformInitialize):
12040
12041 2011-09-11  Filip Pizlo  <fpizlo@apple.com>
12042
12043         Added my IRC nick to the committers file.
12044
12045         * Scripts/webkitpy/common/config/committers.py:
12046
12047 2011-09-01  Filip Pizlo  <fpizlo@apple.com>
12048
12049         The executable allocator makes it difficult to free individual
12050         chunks of executable memory
12051         https://bugs.webkit.org/show_bug.cgi?id=66363
12052
12053         Reviewed by Oliver Hunt.
12054
12055         Introduced a best-fit, balanced-tree based allocator. The allocator
12056         required a balanced tree that does not allocate memory and that
12057         permits the removal of individual nodes directly (as opposed to by
12058         key); neither AVLTree nor WebCore's PODRedBlackTree supported this.
12059         Changed all references to executable code to use a reference counted
12060         handle.
12061
12062         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
12063         * TestWebKitAPI/Tests/WTF/MetaAllocator.cpp: Added.
12064         (TestWebKitAPI::TEST_F):
12065         * TestWebKitAPI/Tests/WTF/RedBlackTree.cpp: Added.
12066         (TestWebKitAPI::Pair::findExact):
12067         (TestWebKitAPI::Pair::remove):
12068         (TestWebKitAPI::Pair::findLeastGreaterThanOrEqual):
12069         (TestWebKitAPI::Pair::assertFoundAndRemove):
12070         (TestWebKitAPI::Pair::assertEqual):
12071         (TestWebKitAPI::Pair::assertSameValuesForKey):
12072         (TestWebKitAPI::Pair::testDriver):
12073         (TestWebKitAPI::TEST_F):
12074
12075 2011-09-10  Andy Estes  <aestes@apple.com>
12076
12077         Move myself from committers_unable_to_review to reviewers_list.
12078
12079         * Scripts/webkitpy/common/config/committers.py:
12080
12081 2011-09-10  Kevin Ollivier  <kevino@theolliviers.com>
12082
12083         [wx] Unreviewed build fix. MSW build fixes.
12084
12085         * DumpRenderTree/config.h:
12086         * waf/build/wxpresets.py:
12087
12088 2011-09-10  Andreas Kling  <kling@webkit.org>
12089
12090         Crash beneath WKURLIsEqual() when reloading in MiniBrowser.
12091         https://bugs.webkit.org/show_bug.cgi?id=59328
12092
12093         Reviewed by Kenneth Rohde Christiansen.
12094
12095         * MiniBrowser/mac/BrowserWindowController.m:
12096         (-[BrowserWindowController updateProvisionalURLForFrame:]): Add missing null check.
12097
12098 2011-09-10  Adam Barth  <abarth@webkit.org>
12099
12100         Build fix for Android.  DumpRenderTree is a client of the WebKit API
12101         and does not have access to notImplemented().
12102
12103         * DumpRenderTree/chromium/TestShellAndroid.cpp:
12104         (openStartupDialog):
12105
12106 2011-09-09  Alexey Proskuryakov  <ap@apple.com>
12107
12108         Add a script to create a monolithic script-test
12109         https://bugs.webkit.org/show_bug.cgi?id=67747
12110
12111         Reviewed by Darin Adler.
12112
12113         * Scripts/make-new-script-test: Added.
12114
12115 2011-09-09  Luiz Agostini  <luiz@webkit.org>
12116
12117         Adding myself to the reviewers list.
12118
12119         * Scripts/webkitpy/common/config/committers.py:
12120
12121 2011-09-09  Ryosuke Niwa  <rniwa@webkit.org>
12122
12123         Buildbot may use RunWebKitTest's evaluateCommand on bots that run nrwt
12124         https://bugs.webkit.org/show_bug.cgi?id=67855
12125
12126         Reviewed by Dirk Pranke.
12127
12128         The bug was caused by the fact some ports use RunWebKitTests instead of NewRunWebKitTests to run tests,
12129         and using RunWebKitTests.evaluateCommand after calling _parseNewRunWebKitTestsOutput.
12130
12131         Fixed the bug by merging evaluateCommand of RunWebKitTests and NewRunWebKitTests.
12132
12133         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
12134
12135 2011-09-09  Mark Hahnenberg  <mhahnenberg@apple.com>
12136
12137         Add myself to committers.py
12138         https://bugs.webkit.org/show_bug.cgi?id=67853
12139
12140         Reviewed by Darin Adler.
12141
12142         * Scripts/webkitpy/common/config/committers.py:
12143
12144 2011-09-09  Laszlo Gombos  <laszlo.1.gombos@nokia.com>
12145
12146         [Qt] Remove common.pri
12147         https://bugs.webkit.org/show_bug.cgi?id=67814
12148
12149         Reviewed by Andreas Kling.
12150
12151         * Tools.pro:
12152
12153 2011-09-09  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
12154
12155         [NRWT] the nrwt check httpd with --no-http option
12156         https://bugs.webkit.org/show_bug.cgi?id=67835
12157
12158         Reviewed by Tony Chang.
12159
12160         * Scripts/webkitpy/layout_tests/controllers/manager.py:
12161         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
12162
12163 2011-09-09  Hao Zheng  <zhenghao@chromium.org>
12164
12165         [chromium] Implement DRT support for Android.
12166         https://bugs.webkit.org/show_bug.cgi?id=67590
12167
12168         Reviewed by Tony Chang.
12169
12170         Add 2 mechanism to make DRT interaction more robust on Android:
12171         - Send 'QUIT' to DRT to make it stop.
12172         - Wait until receiving '#READY' from DRT, so that DRT won't miss any
12173         input commands before it starts.
12174         And LayoutTestController.cpp needs to include <cctype> for 'tolower'
12175         and <clocale> for 'LC_ALL' to compile on Android.
12176
12177         Also change to use sigaction instead of signal for TestShellGtk.cpp,
12178         as signal() man page recommends using sigaction().
12179
12180         * DumpRenderTree/DumpRenderTree.gypi:
12181         * DumpRenderTree/chromium/DumpRenderTree.cpp:
12182         (main):
12183         * DumpRenderTree/chromium/LayoutTestController.cpp:
12184         * DumpRenderTree/chromium/TestShellAndroid.cpp: Added.
12185         (AlarmHandler):
12186         (TestShell::waitTestFinished):
12187         (platformInit):
12188         (openStartupDialog):
12189         (checkLayoutTestSystemDependencies):
12190         * DumpRenderTree/chromium/TestShellGtk.cpp:
12191         (AlarmHandler):
12192         (TestShell::waitTestFinished):
12193
12194 2011-09-09  Kevin Ollivier  <kevino@theolliviers.com>
12195
12196         [wx] Unreviewed build fix. Add Source/JavaScriptCore/dfg directory to the build.
12197
12198         * waf/build/settings.py:
12199
12200 2011-09-09  Sheriff Bot  <webkit.review.bot@gmail.com>
12201
12202         Unreviewed, rolling out r94851.
12203         http://trac.webkit.org/changeset/94851
12204         https://bugs.webkit.org/show_bug.cgi?id=67847
12205
12206         It broke a webkitpy test (Requested by ossy on #webkit).
12207
12208         * Scripts/webkitpy/layout_tests/controllers/manager.py:
12209
12210 2011-09-09  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
12211
12212         [NRWT] the nrwt check httpd with --no-http option
12213         https://bugs.webkit.org/show_bug.cgi?id=67835
12214
12215         Reviewed by Csaba Osztrogonác.
12216
12217         * Scripts/webkitpy/layout_tests/controllers/manager.py:
12218
12219 2011-09-08  Brian Weinstein  <bweinstein@apple.com>
12220
12221         WebKit2: Assertion when calling didPerform*Redirect on null source/destination URL string
12222         https://bugs.webkit.org/show_bug.cgi?id=67794
12223         <rdar://problem/9892024>
12224
12225         Add a test that we don't send a didPerformServerRedirect when a load is cancelled from willSendRequest.
12226
12227         Reviewed by Brady Eidson.
12228
12229         * TestWebKitAPI/Tests/WebKit2/LoadCanceledNoServerRedirectCallback.cpp: Added.
12230         (TestWebKitAPI::didFinishLoadForFrame): Set flags on whether we've loaded the main frame, iframe, or both.
12231         (TestWebKitAPI::didPerformServerRedirect): Set a flag for whether we've performed a server redirect.
12232         (TestWebKitAPI::TEST): Set up a WKPageLoaderClient and WKContextHistoryClient, and load simple-iframe.html.
12233             Once both frames have loaded (main frame and blocked iframe), make sure that we have never called
12234             didPerformServerRedirect.
12235
12236         * TestWebKitAPI/Tests/WebKit2/LoadCanceledNoServerRedirectCallback_Bundle.cpp: Added.
12237         (TestWebKitAPI::LoadCanceledNoServerRedirectCallbackTest::LoadCanceledNoServerRedirectCallbackTest): Initialize
12238             our InjectedBundleTest.
12239         (TestWebKitAPI::LoadCanceledNoServerRedirectCallbackTest::willSendRequestForFrame): If we are dealing with a request
12240             from the main frame (the main resource in simple-iframe.html), allow it. If not (dealing with the iframe), block
12241             the request.
12242         (TestWebKitAPI::LoadCanceledNoServerRedirectCallbackTest::didCreatePage): Set up the WKBundlePageResourceLoadClient.
12243
12244         * TestWebKitAPI/Tests/WebKit2/simple-iframe.html: Added.
12245
12246         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Add new files.
12247         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Ditto.
12248         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Ditto.
12249         * TestWebKitAPI/win/copy-resources.cmd: Ditto.
12250
12251 2011-09-08  David Levin  <levin@chromium.org>
12252
12253         check-webkit-style: Ref|Own Ptr& should be allowed as a parameter.
12254         https://bugs.webkit.org/show_bug.cgi?id=67797
12255
12256         Reviewed by Eric Seidel.
12257
12258         * Scripts/webkitpy/style/checkers/cpp.py: Allow Ref/Own Ptr when it ends with &.
12259         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added test cases for this.
12260
12261 2011-09-08  Fady Samuel  <fsamuel@chromium.org>
12262
12263         Fixed Layout Mode should be adjustable from layoutTestController for testing on Chromium platforms
12264         https://bugs.webkit.org/show_bug.cgi?id=67723
12265
12266         Reviewed by Adam Barth.
12267
12268         * DumpRenderTree/chromium/EventSender.h:
12269         * DumpRenderTree/chromium/LayoutTestController.cpp:
12270         (LayoutTestController::LayoutTestController):
12271         (LayoutTestController::enableFixedLayoutMode):
12272         (LayoutTestController::setFixedLayoutSize):
12273         * DumpRenderTree/chromium/LayoutTestController.h:
12274
12275 2011-09-07  Matthew Delaney  <mdelaney@apple.com>
12276
12277         Add standalone script that filters the output of build-webkit to be more human-readable
12278         https://bugs.webkit.org/show_bug.cgi?id=44081
12279
12280         Reviewed by Daniel Bates.
12281
12282         * Scripts/filter-build-webkit: Added. Just need to feed it the output of
12283         a build script (e.g. build-webkit) and it spits out a prettier version.
12284         * Scripts/VCSUtils.pm: Refactored "possiblyColored" out of run-api-tests into VCSUtils.
12285         * Scripts/run-api-tests: Now includes VCSUtils to use "possiblyColored"
12286
12287 2011-09-07  Noel Gordon  <noel.gordon@gmail.com>
12288
12289         Add myself to committers.py
12290         https://bugs.webkit.org/show_bug.cgi?id=67696
12291
12292         Reviewed by Kent Tamura.
12293
12294         * Scripts/webkitpy/common/config/committers.py:
12295
12296 2011-09-07  Adam Barth  <abarth@webkit.org>
12297
12298         rebaseline button in garden-o-matic details view should work
12299         https://bugs.webkit.org/show_bug.cgi?id=67738
12300
12301         Reviewed by Dimitri Glazkov.
12302
12303         This patch wires up the rebaseline button in garden-o-matic results
12304         details view.  The button is current per-test, so there isn't a way to
12305         rebaseline individual bots.  This is a place to start.  If we decide we
12306         want per-bot rebaselining, we'll need to think of some better UI.
12307
12308         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
12309         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
12310         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
12311         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
12312
12313 2011-09-07  Eric Seidel  <eric@webkit.org>
12314
12315         Teach the PortFactory object how to pass along executive/user/filesystem to port objects (which fixes the failing rebaseline_test on bots)
12316         https://bugs.webkit.org/show_bug.cgi?id=67737
12317
12318         Reviewed by Adam Barth.
12319
12320         I got a little over-eager when deprecating host.port_factory in the previous commit.
12321         Turns out that Host.port_factory was already the "modern" port/factory.py
12322         it was just the Module object.  I un-deprecated the variable, and changed it from
12323         being the module to being an actual PortFactory, passing it a host object.
12324         I also taught PortFactory how to set user/executive/filesystem in the kwargs
12325         when instantiating a new Port object.
12326         This magically fixed the test_rebaseline test, since its now no longer talking
12327         to the real filesystem.
12328
12329         * Scripts/webkitpy/common/host.py:
12330         * Scripts/webkitpy/layout_tests/port/factory.py:
12331         * Scripts/webkitpy/tool/commands/queries.py:
12332         * Scripts/webkitpy/tool/mocktool.py:
12333
12334 2011-09-07  Eric Seidel  <eric@webkit.org>
12335
12336         Add a new PortFactory class for creating new-style ports and deprecate the old PortFactory
12337         https://bugs.webkit.org/show_bug.cgi?id=67734
12338
12339         Reviewed by Adam Barth.
12340
12341         Right now every place that creates a Port object has to be very careful
12342         to pass in a MockSCM/MockUser/MockFilesystem when appropriate.
12343         The correct fix is for port/factory.py (which creates most port objects)
12344         to know how to pass these in automatically.  However, currently all port/factory.py
12345         functions are free functions instead of being on an object.  This patch
12346         adds a new PortFactory object and deprecates the PortFactory for the old-style Port objects.
12347
12348         In a follow-up patch I will add PortFactory to the Host object and teach PortFactory how to
12349         pass along the proper scm/user/filesystem to the Port object.
12350
12351         * Scripts/webkitpy/common/host.py:
12352         * Scripts/webkitpy/layout_tests/port/factory.py:
12353         * Scripts/webkitpy/tool/commands/queries.py:
12354         * Scripts/webkitpy/tool/mocktool.py:
12355
12356 2011-09-07  Sheriff Bot  <webkit.review.bot@gmail.com>
12357
12358         Unreviewed, rolling out r94441.
12359         http://trac.webkit.org/changeset/94441
12360         https://bugs.webkit.org/show_bug.cgi?id=67731
12361
12362         "MacEWSes are backed up due to new machines not able to reach
12363         git.webkit.org.  Disabling tests for now" (Requested by
12364         eseidel2 on #webkit).
12365
12366         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
12367
12368 2011-09-07  Elliot Poger  <epoger@google.com>
12369
12370         filter test_expectations properly for chromium-mac vs chromium-cg-mac
12371         https://bugs.webkit.org/show_bug.cgi?id=67324
12372
12373         Reviewed by Dirk Pranke.
12374
12375         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
12376         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
12377         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
12378         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
12379         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
12380
12381 2011-09-07  Eric Boren  <borenet@gmail.com>
12382
12383         [Chromium] Modify WebTouchEvent structure to match WebCore::TouchEvent
12384         https://bugs.webkit.org/show_bug.cgi?id=66800
12385
12386         Reviewed by Darin Fisher.
12387
12388         * DumpRenderTree/chromium/EventSender.cpp:
12389         (EventSender::sendCurrentTouchEvent):
12390
12391 2011-09-07  Jarred Nicholls  <jarred@sencha.com>
12392
12393         [Qt][DRT] Normalize file:///tmp/LayoutTests in LayoutTestController::pathToLocalResource()
12394         https://bugs.webkit.org/show_bug.cgi?id=67254
12395
12396         Reviewed by Csaba Osztrogonác.
12397
12398         Translate file:///tmp/LayoutTests/* urls into the repository LayoutTests directory, which is
12399         derived from the running location of DumpRenderTree binary.
12400
12401         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
12402         (LayoutTestController::pathToLocalResource):
12403         (LayoutTestController::setUserStyleSheetLocation): pass the url through pathToLocalResource
12404
12405 2011-09-07  Csaba Osztrogonác  <ossy@webkit.org>
12406
12407         REGRESSION(r94288) build-webkit doesn't work on Symbian
12408         https://bugs.webkit.org/show_bug.cgi?id=67706
12409
12410         * Scripts/webkitdirs.pm: Disable automatic clean build on Symbian, because it doesn't support shadow build.
12411         (buildQMakeProject):
12412
12413 2011-09-07  Csaba Osztrogonác  <ossy@webkit.org>
12414
12415         REGRESSION(r94288) build-webkit doesn't work on Windows
12416         https://bugs.webkit.org/show_bug.cgi?id=67646
12417
12418         Roll back r94655 with a small fix.
12419
12420         * Scripts/webkitdirs.pm:
12421         (buildQMakeProject):
12422
12423 2011-09-07  Sheriff Bot  <webkit.review.bot@gmail.com>
12424
12425         Unreviewed, rolling out r94635.
12426         http://trac.webkit.org/changeset/94635
12427         https://bugs.webkit.org/show_bug.cgi?id=67704
12428
12429         Linux Touch build is failing to compile this change.
12430         (Requested by loislo on #webkit).
12431
12432         * DumpRenderTree/chromium/EventSender.cpp:
12433         (EventSender::sendCurrentTouchEvent):
12434
12435 2011-09-07  Sheriff Bot  <webkit.review.bot@gmail.com>
12436
12437         Unreviewed, rolling out r94655.
12438         http://trac.webkit.org/changeset/94655
12439         https://bugs.webkit.org/show_bug.cgi?id=67703
12440
12441         It broke Leopard debug bot (Requested by ossy on #webkit).
12442
12443         * Scripts/webkitdirs.pm:
12444
12445 2011-09-07  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
12446
12447         REGRESSION(r94288) build-webkit doesn't work on Windows
12448         https://bugs.webkit.org/show_bug.cgi?id=67646
12449
12450         Reviewed by Csaba Osztrogonác.
12451
12452         * Scripts/webkitdirs.pm:
12453
12454 2011-09-06  Eric Boren  <borenet@gmail.com>
12455
12456         [Chromium] Modify WebTouchEvent structure to match WebCore::TouchEvent
12457         https://bugs.webkit.org/show_bug.cgi?id=66800
12458
12459         Reviewed by Darin Fisher.
12460
12461         * DumpRenderTree/chromium/EventSender.cpp:
12462         (EventSender::sendCurrentTouchEvent):
12463
12464 2011-09-06  Adam Barth  <abarth@webkit.org>
12465
12466         garden-o-matic details view should having working rebaseline and next/previous buttons
12467         https://bugs.webkit.org/show_bug.cgi?id=67659
12468
12469         Reviewed by Dimitri Glazkov.
12470
12471         This patch wires up basic back/forward buttons that let you traverse
12472         through the results we're examining in the details view.  This ended up
12473         being more code than I expected, but I wanted to keep all the state
12474         information in the DOM itself.
12475
12476         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
12477         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
12478         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
12479
12480 2011-09-01  Dirk Pranke  <dpranke@chromium.org>
12481
12482         fix MockFileSystem.glob(), refactor filesystem tests
12483         https://bugs.webkit.org/show_bug.cgi?id=67462
12484
12485         Reviewed by Eric Seidel.
12486
12487         In debugging why my proposed change for MockFileSystem.glob()
12488         in bug 66228 didn't work, I stumbled across two real bugs
12489         that cancelled each other out for the unit tests. I fixed those
12490         (glob shouldn't recurse into subdirs, and we weren't populating
12491         self.dirs from self.files properly).
12492
12493         I have also created a "mixin" class for tests that can be shared
12494         between the real filesystem and the mock filesystem - MockFileSystem
12495         is being used enough it needs its own tests.
12496
12497         * Scripts/webkitpy/common/system/filesystem_mock.py:
12498         * Scripts/webkitpy/common/system/filesystem_mock_unittest.py: Added.
12499         * Scripts/webkitpy/common/system/filesystem_unittest.py:
12500
12501 2011-09-06  Anders Carlsson  <andersca@apple.com>
12502
12503         Move NPAPI headers in bridge to plugins
12504         https://bugs.webkit.org/show_bug.cgi?id=67661
12505
12506         Update forwarding headers.
12507
12508         * DumpRenderTree/unix/TestNetscapePlugin/ForwardingHeaders/WebKit/npapi.h:
12509         * DumpRenderTree/unix/TestNetscapePlugin/ForwardingHeaders/WebKit/npruntime.h:
12510
12511 2011-09-06  Tony Chang  <tony@chromium.org>
12512
12513         [chromium] Check in 'build' for compiled binaries first
12514         https://bugs.webkit.org/show_bug.cgi?id=67549
12515
12516         Reviewed by Dirk Pranke.
12517
12518         I'm working on moving the windows build output to a common location
12519         rather than relative to the sln file.  Because of that, we need to
12520         check those build locations first for DumpRenderTree.exe.
12521
12522         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
12523
12524 2011-09-06  Adam Barth  <abarth@webkit.org>
12525
12526         Use a hierarchial structure for the garden-o-matic results view
12527         https://bugs.webkit.org/show_bug.cgi?id=67620
12528
12529         Reviewed by Dimitri Glazkov.
12530
12531         This patch re-organizes the results view to use a hierachial structure
12532         for selection.  At the top-level are sections for each test.  Expanding
12533         the test section shows a tab selector for builders.
12534
12535         This structure prepares us for wiring up various actions to this view.
12536
12537         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
12538         (.):
12539         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
12540         ():
12541         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
12542         (.):
12543         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
12544         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
12545         (.results-view>.toolbar):
12546         (.results-view>.toolbar ul.actions):
12547         (.results-view>.toolbar ul.actions li):
12548         (.results-grid table):
12549         (.results-grid table td, .results-view table th):
12550         (.results-grid table th):
12551         (.results-grid .text-result):
12552         (.results-grid .image-result):
12553
12554 2011-09-06  Adam Barth  <abarth@webkit.org>
12555
12556         garden-o-matic's base.probe generates spam in the JavaScript console
12557         https://bugs.webkit.org/show_bug.cgi?id=67622
12558
12559         Reviewed by Dimitri Glazkov.
12560
12561         Now that we have CORS set up on the buildbot, we can use XMLHttpRequest
12562         instead of HTMLScriptElement to test whether results exist.  This
12563         reduces the console.log spam and saves bandwidth because we can use a
12564         HEAD request instead of a GET request.
12565
12566         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net.js:
12567
12568 2011-09-05  Adam Barth  <abarth@webkit.org>
12569
12570         Add a tab to garden-o-matic for viewing expected failures
12571         https://bugs.webkit.org/show_bug.cgi?id=67619
12572
12573         Reviewed by Dimitri Glazkov.
12574
12575         This patch adds a basic tab to garden-o-matic that shows all the tests
12576         failing on the bots, including the expected failures.  This patch
12577         mostly involves refactoring existing classes to allow new subclasses.
12578
12579         This view is useful for when gardeners mark tests as expected to fail
12580         and then rebaseline them later.
12581
12582         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
12583         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js:
12584         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
12585         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary-mock.js:
12586         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
12587         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
12588         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
12589
12590 2011-09-05  Adam Barth  <abarth@webkit.org>
12591
12592         Fix global variable leak in garden-o-matic
12593         https://bugs.webkit.org/show_bug.cgi?id=67617
12594
12595         Reviewed by Dimitri Glazkov.
12596
12597         Caught by running the unit tests with noglobals turned on.
12598
12599         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
12600
12601 2011-09-05  Balazs Kelemen  <kbalazs@webkit.org>
12602
12603         [Qt][WK2] Add pixel test support
12604         https://bugs.webkit.org/show_bug.cgi?id=66283
12605
12606         Reviewed by Andreas Kling.
12607
12608         Implemented pixel result generation by reusing the logic
12609         that we have in DumpRenderTree.
12610         * WebKitTestRunner/qt/TestInvocationQt.cpp:
12611         (WTR::dumpImage):
12612         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
12613
12614 2011-09-04  Robin Dunn  <robin@alldunn.com>
12615
12616         [wx] Enable wxWebKit to run using the wxGC Cairo backend on platforms other than GTK.
12617         https://bugs.webkit.org/show_bug.cgi?id=67577
12618
12619         Reviewed by Kevin Ollivier.
12620
12621         * waf/build/settings.py:
12622
12623 2011-09-03  Kevin Ollivier  <kevino@theolliviers.com>
12624
12625         [wx] Unreviewed build fix. Add wtf/dtoa directory to build.
12626
12627         * waf/build/settings.py:
12628
12629 2011-09-03  Dan Bernstein  <mitz@apple.com>
12630
12631         32-bit build fix.
12632
12633         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
12634         (TestWebKitAPI::Util::toSTD):
12635
12636 2011-09-03  Dan Bernstein  <mitz@apple.com>
12637
12638         Move testStringByEvaluatingJavaScriptFromString() from DumpRenderTree to TestWebKitAPI
12639         https://bugs.webkit.org/show_bug.cgi?id=67559
12640
12641         Reviewed by Darin Adler.
12642
12643         * DumpRenderTree/mac/DumpRenderTree.mm:
12644         (dumpRenderTree): Moved testStringByEvaluatingJavaScriptFromString() from here and removed
12645         the call to it.
12646         * TestWebKitAPI/PlatformUtilities.h: Declared Util::toSTD(NSString *)
12647         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added StringByEvaluatingJavaScriptFromString.mm.
12648         * TestWebKitAPI/Tests/mac/StringByEvaluatingJavaScriptFromString.mm: Added.
12649         (TestWebKitAPI::TEST): Added. Moved testStringByEvaluatingJavaScriptFromString() to here.
12650         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
12651         (TestWebKitAPI::Util::toSTD): Added.
12652
12653 2011-09-03  Adam Barth  <abarth@webkit.org>
12654
12655         garden-o-matic results view should have a better results selector
12656         https://bugs.webkit.org/show_bug.cgi?id=67542
12657
12658         Reviewed by Dimitri Glazkov.
12659
12660         This patch adds a basic grid-based results selector to the results
12661         view.  Currently, this grid just shows how the tests fail on the
12662         various bots.  A future patch will let you select which failure to view
12663         and to mark the failures for rebaselining.
12664
12665         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
12666         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
12667           - Add some more basic functions for manipulating dictionaries.
12668         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
12669         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
12670         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
12671
12672 2011-09-03  Adam Barth  <abarth@webkit.org>
12673
12674         garden-o-matic results view should show Actual and Expected results for text failures
12675         https://bugs.webkit.org/show_bug.cgi?id=67536
12676
12677         Reviewed by Eric Seidel.
12678
12679         In the past, we didn't show actual and expected text results (we just
12680         showed the diff) because we didn't have the awesome six-by-two results
12681         grid.  Now that we do, we can move show these results.
12682
12683         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
12684         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
12685           - Fixed some typos that were preventing these unit tests from running.
12686         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
12687         (.):
12688         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js:
12689           - Report -actual.txt and -expected.txt as possible results URLs.
12690
12691 2011-09-03  Adam Barth  <abarth@webkit.org>
12692
12693         garden-o-matic shows "ghost" failures that are already fixed
12694         https://bugs.webkit.org/show_bug.cgi?id=67535
12695
12696         Reviewed by Eric Seidel.
12697
12698         About 10% of writes fail to stick on test-results.appspot.com.  This
12699         patch causes us to fetch the full_results.json blob directly from
12700         build.chromium.org rather than relying upon AppEngine.
12701
12702         Unforuntately, we still need to get historical information from
12703         AppEngine.  That will take several patches to fix becuase I need to
12704         change the bots to write the full_results.json file to the
12705         build-specific directory in addition to the zip file.
12706
12707         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
12708         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
12709         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
12710
12711 2011-09-03  Sheriff Bot  <webkit.review.bot@gmail.com>
12712
12713         Unreviewed, rolling out r94458.
12714         http://trac.webkit.org/changeset/94458
12715         https://bugs.webkit.org/show_bug.cgi?id=67558
12716
12717         Caused appcache test to fail on various bots (Requested by
12718         rniwa on #webkit).
12719
12720         * Scripts/webkitpy/layout_tests/port/webkit.py:
12721         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
12722
12723 2011-09-03  Sheriff Bot  <webkit.review.bot@gmail.com>
12724
12725         Unreviewed, rolling out r94453.
12726         http://trac.webkit.org/changeset/94453
12727         https://bugs.webkit.org/show_bug.cgi?id=67557
12728
12729         Caused appcache test to fail on various bots (Requested by
12730         rniwa on #webkit).
12731
12732         * Scripts/webkitpy/layout_tests/port/webkit.py:
12733
12734 2011-09-02  Michael Saboff  <msaboff@apple.com>
12735
12736         Add JSC:RegExp functional tests
12737         https://bugs.webkit.org/show_bug.cgi?id=67339
12738
12739         New perl script to build JavaScriptCore and run the RegExp tests.
12740
12741         Reviewed by Gavin Barraclough.
12742
12743         * Scripts/run-regexp-tests: Added.
12744
12745 2011-09-02  Michael Saboff  <msaboff@apple.com>
12746
12747         Add JSC:RegExp functional tests
12748         https://bugs.webkit.org/show_bug.cgi?id=67339
12749
12750         New perl script to build JavaScriptCore and run the RegExp tests.
12751
12752         Reviewed by Gavin Barraclough.
12753
12754         * Scripts/run-regexp-tests: Added.
12755
12756 2011-09-02  Eric Seidel  <eric@webkit.org>
12757
12758         Reshuffle some code in WebKitDriver._read_block in preparation for reading stderr/stdout separately
12759         https://bugs.webkit.org/show_bug.cgi?id=67530
12760
12761         Unreviewed.  Fixing typo from previous commit.
12762
12763         Turns out there was *no* unittesting of WebKitDriver.  Added a basic test of _read_block
12764         which exercises the code I previously made a typo in.
12765
12766         * Scripts/webkitpy/layout_tests/port/webkit.py:
12767         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
12768
12769 2011-09-02  Eric Seidel  <eric@webkit.org>
12770
12771         Reshuffle some code in WebKitDriver._read_block in preparation for reading stderr/stdout separately
12772         https://bugs.webkit.org/show_bug.cgi?id=67530
12773
12774         Reviewed by Adam Barth.
12775
12776         No functional change, just reshuffling code.
12777
12778         * Scripts/webkitpy/layout_tests/port/webkit.py:
12779
12780 2011-09-02  Darin Adler  <darin@apple.com>
12781
12782         Added Subversion ignore for .pyc generated files.
12783
12784         * Scripts/webkitpy/layout_tests/models: Added property svn:ignore.
12785         * Scripts/webkitpy/layout_tests/servers: Added property svn:ignore.
12786         * Scripts/webkitpy/tool/servers: Added property svn:ignore.
12787
12788 2011-09-02  Adam Barth  <abarth@webkit.org>
12789
12790         Enable testing on mac-ews
12791         https://bugs.webkit.org/show_bug.cgi?id=67505
12792
12793         Reviewed by Eric Seidel.
12794
12795         This patch enables testing on the Mac EWS, which should help
12796         contributors not turn the Mac bots red.
12797
12798         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
12799
12800 2011-09-01  Beth Dakin  <bdakin@apple.com>
12801
12802         https://bugs.webkit.org/show_bug.cgi?id=67451
12803         Add a test for going back to a cached page after changing the device scale factor
12804
12805         Reviewed by Adam Roben.
12806
12807         New test DeviceScaleFactorOnBack.mm inherits from WebKitAgnosticTest.
12808         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
12809         * TestWebKitAPI/Tests/mac/DeviceScaleFactorOnBack.mm: Added.
12810         (TestWebKitAPI::DeviceScaleFactorOnBack::url):
12811         (TestWebKitAPI::DeviceScaleFactorOnBack::didLoadURL):
12812         (TestWebKitAPI::DeviceScaleFactorOnBack::createWindow):
12813         (TestWebKitAPI::DeviceScaleFactorOnBack::initializeView):
12814         (TestWebKitAPI::DeviceScaleFactorOnBack::runTest):
12815         (TestWebKitAPI::TEST_F):
12816
12817         Added new function to WebKitAgnosticTest to goBack. Separated
12818         waitForLoadToFinish() into its own function (instead of being a part of
12819         loadAndWaitUntilFinished()) so that it can be called from goBack as
12820         well. Also added initializeView() so we can set the cache model.
12821         * TestWebKitAPI/mac/WebKitAgnosticTest.h:
12822         (TestWebKitAPI::WebKitAgnosticTest::initializeView):
12823         * TestWebKitAPI/mac/WebKitAgnosticTest.mm:
12824         (TestWebKitAPI::WebKitAgnosticTest::WebKitAgnosticTest):
12825         (TestWebKitAPI::WebKitAgnosticTest::runWebKit1Test):
12826         (TestWebKitAPI::WebKitAgnosticTest::runWebKit2Test):
12827         (TestWebKitAPI::WebKitAgnosticTest::loadURL):
12828         (TestWebKitAPI::WebKitAgnosticTest::goBack):
12829         (TestWebKitAPI::WebKitAgnosticTest::waitForLoadToFinish):
12830
12831 2011-09-02  Anders Carlsson  <andersca@apple.com>
12832
12833         Build fix.
12834
12835         * DumpRenderTree/TestNetscapePlugIn/Tests/PrivateBrowsing.cpp:
12836         (PrivateBrowsing::ScriptableObject::pluginTest):
12837
12838 2011-09-02  Anders Carlsson  <andersca@apple.com>
12839
12840         Move private browsing test to a PluginTest subclass
12841         https://bugs.webkit.org/show_bug.cgi?id=67498
12842
12843         Reviewed by Darin Adler.
12844
12845         * DumpRenderTree/DumpRenderTree.gypi:
12846         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
12847         * DumpRenderTree/TestNetscapePlugIn/PluginObject.cpp:
12848         (pluginGetProperty):
12849         * DumpRenderTree/TestNetscapePlugIn/PluginObject.h:
12850         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
12851         (PluginTest::NPP_SetValue):
12852         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
12853         * DumpRenderTree/TestNetscapePlugIn/Tests/PrivateBrowsing.cpp: Added.
12854         (PrivateBrowsing::PrivateBrowsing):
12855         (PrivateBrowsing::privateBrowsingEnabled):
12856         (PrivateBrowsing::cachedPrivateBrowsingEnabled):
12857         (PrivateBrowsing::ScriptableObject::hasProperty):
12858         (PrivateBrowsing::ScriptableObject::getProperty):
12859         (PrivateBrowsing::ScriptableObject::pluginTest):
12860         (PrivateBrowsing::NPP_New):
12861         (PrivateBrowsing::NPP_GetValue):
12862         (PrivateBrowsing::NPP_SetValue):
12863         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
12864         (NPP_New):
12865         (NPP_SetValue):
12866         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
12867         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
12868         * GNUmakefile.am:
12869
12870 2011-09-02  Csaba Osztrogonác  <ossy@webkit.org>
12871
12872         [Qt] Build dependency problems
12873         https://bugs.webkit.org/show_bug.cgi?id=38054
12874
12875         Trivial fix after r94288.
12876
12877         Rubber-stamped by Andras Becsi.
12878
12879         * Scripts/webkitdirs.pm: We should read the whole defaults.txt, not only the first two lines.
12880         (buildQMakeProject):
12881
12882 2011-09-02  Andras Becsi  <andras.becsi@nokia.com>
12883
12884         Remove my old e-mail address and add my new one.
12885
12886         Rubber-stamped by Tor Arne Vestbø.
12887
12888         * Scripts/webkitpy/common/config/committers.py:
12889
12890 2011-09-02  Philippe Normand  <pnormand@igalia.com>
12891
12892         [GTK] run-gtk-tests leaves a zombie Xvfb process
12893         https://bugs.webkit.org/show_bug.cgi?id=67389
12894
12895         Reviewed by Martin Robinson.
12896
12897         * Scripts/run-gtk-tests: Properly close STDIN and STDERR on the
12898         child process instead of doing shell redirection.
12899
12900 2011-09-01  Mark Rowe  <mrowe@apple.com>
12901
12902         Fix the build by adding some more missing return types.
12903
12904         * DumpRenderTree/mac/ResourceLoadDelegate.mm:
12905         (-[ResourceLoadDelegate webView:identifierForInitialRequest:fromDataSource:]):
12906
12907 2011-09-01  Eric Seidel  <eric@webkit.org>
12908
12909         REGRESSION (NRWT): Leaks Viewer can't load leaks from test runs that used NRWT
12910         https://bugs.webkit.org/show_bug.cgi?id=66228
12911
12912         Reviewed by Dirk Pranke.
12913
12914         ORWT used $(PROCESS)$(NUMBER)-leaks.txt for leaks filenames.
12915         The LeaksViewer tool globs for all *-leaks.txt files.
12916         This patch makes NRWT output $(PROCESS)-$(PID)-leaks.txt which should match.
12917
12918         In order to test this I had to fix a limitation in our MockFileSystem.glob
12919         method.  However, doing so uncovered a typo and bug in the integration tests.
12920         I've disabled the offending integration tests.
12921
12922         * Scripts/webkitpy/common/system/filesystem_mock.py:
12923         * Scripts/webkitpy/layout_tests/port/leakdetector.py:
12924         * Scripts/webkitpy/layout_tests/port/leakdetector_unittest.py:
12925
12926 2011-09-01  Dominic Mazzoni  <dmazzoni@google.com>
12927
12928         Adds a LayoutTestController method to make it possible to test
12929         starting speech input via an API call.
12930         https://bugs.webkit.org/show_bug.cgi?id=60170
12931
12932         Reviewed by Dimitri Glazkov.
12933
12934         * DumpRenderTree/LayoutTestController.h:
12935         * DumpRenderTree/chromium/LayoutTestController.cpp:
12936         (LayoutTestController::LayoutTestController):
12937         (LayoutTestController::startSpeechInput):
12938         * DumpRenderTree/chromium/LayoutTestController.h:
12939         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
12940         (LayoutTestController::startSpeechInput):
12941         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
12942         (LayoutTestController::startSpeechInput):
12943         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
12944         (LayoutTestController::startSpeechInput):
12945         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
12946         (LayoutTestController::startSpeechInput):
12947         * DumpRenderTree/qt/LayoutTestControllerQt.h:
12948         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
12949         (LayoutTestController::startSpeechInput):
12950         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
12951         (LayoutTestController::startSpeechInput):
12952
12953 2011-09-01  Tim Horton  <hortont424@gmail.com>
12954
12955         Fix TestWebKitAPI build on clang-2.0
12956
12957         * TestWebKitAPI/Tests/WTF/CheckedArithmeticOperations.cpp:
12958
12959 2011-09-01  Adam Barth  <abarth@webkit.org>
12960
12961         Remove unused "extension" parameter from rebaseline gardeningserver handler
12962         https://bugs.webkit.org/show_bug.cgi?id=67384
12963
12964         Reviewed by James Robinson.
12965
12966         This parameter hasn't done anything in a while.
12967
12968         * Scripts/webkitpy/tool/commands/rebaseline.py:
12969         * Scripts/webkitpy/tool/servers/gardeningserver.py:
12970         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
12971
12972 2011-09-01  Eric Seidel  <eric@webkit.org>
12973
12974         Unreviewed.
12975
12976         Fix a typo in unittests.  Unfortunately fixing it
12977         caused several assertions.  Just disabling the test
12978         instead, since it was providing no value before due to the typo.
12979
12980         * Scripts/webkitpy/layout_tests/port/test.py:
12981
12982 2011-09-01  Adam Barth  <abarth@webkit.org>
12983
12984         Add a "rebaseline" button to the garden-o-matic summary page
12985         https://bugs.webkit.org/show_bug.cgi?id=67382
12986
12987         Reviewed by Dimitri Glazkov.
12988
12989         As much as we might wish folks to not rebaseline entire test groups at
12990         once, the truth is that doing so is a common operation.  Adding a
12991         button for this operation is likely to make folks happy.
12992
12993         This patch also cleans up a bunch of plumbing surrounding failureInfo
12994         objects.
12995
12996         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
12997         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
12998         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
12999         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
13000         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
13001         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js:
13002         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js:
13003         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js:
13004         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
13005         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
13006
13007 2011-09-01  Adam Barth  <abarth@webkit.org>
13008
13009         garden-o-matic should let the gardener assign blame for a group of test failures
13010         https://bugs.webkit.org/show_bug.cgi?id=67379
13011
13012         Reviewed by Dimitri Glazkov.
13013
13014         If a group of test failures has a large revision range, the gardener
13015         might wish to blame a specific revision for the failure.  Currently,
13016         this only effect of assigning blame in this way is to hide the other
13017         suspect commits.
13018
13019         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
13020         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
13021         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
13022
13023 2011-09-01  Adam Barth  <abarth@webkit.org>
13024
13025         Hide garden-o-matic failure table until hover
13026         https://bugs.webkit.org/show_bug.cgi?id=67376
13027
13028         Reviewed by Dimitri Glazkov.
13029
13030         This patch removes the failure table for build failures (because it
13031         does not make sense) and hides the failure table for test groups until
13032         the user hovers over the test failure block, reducing visual noise.
13033
13034         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
13035         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
13036         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css:
13037
13038 2011-09-01  Adam Barth  <abarth@webkit.org>
13039
13040         Switch garden-o-matic from commitData.title to commitData.summary
13041         https://bugs.webkit.org/show_bug.cgi?id=67375
13042
13043         Reviewed by Dimitri Glazkov.
13044
13045         commitData.title is a string provided by Trac, which isn't great in all
13046         cases.  commitData.summary is my attempt to find the most important
13047         one-line summary of the commit.  This patch switches garden-o-matic
13048         over to summary.
13049
13050         Also, this patch tweaks a bunch of CSS to tighten up the summary page
13051         and reduce visual noise.
13052
13053         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
13054         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
13055         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css:
13056
13057 2011-09-01  Adam Barth  <abarth@webkit.org>
13058
13059         Group failing tests in garden-o-matic if there are more than three in a folder
13060         https://bugs.webkit.org/show_bug.cgi?id=67373
13061
13062         Reviewed by Dimitri Glazkov.
13063
13064         Sometimes a failure will cause many tests in a given directory to fail.
13065         When that happens, it's better to list the directory as a single item
13066         (with a count) rather than spamming all the test names.
13067
13068         (This patch also updates a bunch of unit test expected results after my last commit.)
13069
13070         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
13071         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
13072         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js:
13073         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
13074         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
13075
13076 2011-09-01  Adam Barth  <abarth@webkit.org>
13077
13078         Lower the visual noise from the failing builders display
13079         https://bugs.webkit.org/show_bug.cgi?id=67369
13080
13081         Reviewed by Dimitri Glazkov.
13082
13083         This patch ensures that the "Building" row in the failing builders
13084         display is always last and dims it by 50% to reduce visual noise.  This
13085         lets the gardener focus on the actual failures.
13086
13087         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
13088         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js:
13089         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/failures.css: Added.
13090
13091 2011-09-01  Kristóf Kosztyó  <kkristof@inf.u-szeged.hu>
13092
13093         [Qt] Build dependency problems
13094         https://bugs.webkit.org/show_bug.cgi?id=38054
13095
13096         Make clean build after the defines are changed.
13097
13098         Reviewed by Csaba Osztrogonác.
13099
13100         * Scripts/webkitdirs.pm:
13101         (buildQMakeProject):
13102
13103 2011-08-31  Adam Barth  <abarth@webkit.org>
13104
13105         Move DumpRenderTree to WebKitPlatformSupport
13106         https://bugs.webkit.org/show_bug.cgi?id=67331
13107
13108         Reviewed by Darin Fisher.
13109
13110         This patch update DumpRenderTree to account for the WebKitClient =>
13111         WebKitPlatformSupport rename.
13112
13113         * DumpRenderTree/chromium/NotificationPresenter.cpp:
13114         (NotificationPresenter::show):
13115         * DumpRenderTree/chromium/Task.cpp:
13116         (postTask):
13117
13118 2011-08-31  Eric Seidel  <eric@webkit.org>
13119
13120         Add queues.webkit.org/active-bots page to make it easy to see which bots are down
13121         https://bugs.webkit.org/show_bug.cgi?id=67314
13122
13123         Reviewed by Adam Barth.
13124
13125         Very simple page.  Just lists all the bots, what queue they belong to and what we last heard from them.
13126         If we haven't heard from a bot in the last 500 messages (last day or so), then it won't show up here.
13127         This makes it very easy for me to see which of my cr-jail bots might be down at any one time.
13128
13129         I also moved a bunch of code into dashboard.css which belonged there.
13130
13131         * QueueStatusServer/handlers/activebots.py: Added.
13132         * QueueStatusServer/main.py:
13133         * QueueStatusServer/stylesheets/dashboard.css:
13134         (.queue_bubble):
13135         (.queue_name):
13136         (.last_heard_from):
13137         (.status_text):
13138         (.alive):
13139         (.behind):
13140         (.dead):
13141         * QueueStatusServer/templates/activebots.html: Added.
13142         * QueueStatusServer/templates/dashboard.html:
13143         * QueueStatusServer/templates/recentstatus.html:
13144
13145 2011-08-31  Oliver Hunt  <oliver@apple.com>
13146
13147         Move CheckedArithmeticOperations.cpp to the intended location
13148
13149         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
13150         * TestWebKitAPI/Tests/WTF/CheckedArithmeticOperations.cpp: Renamed from Tools/TestWebKitAPI/Tests/CheckedArithmeticOperations.cpp.
13151
13152 2011-08-31  Elliot Poger  <epoger@google.com>
13153
13154         add unittest to validate the complete set of Chromium configurations
13155         https://bugs.webkit.org/show_bug.cgi?id=67211
13156
13157         Reviewed by Adam Barth.
13158
13159         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
13160
13161 2011-08-31  Nico Weber  <thakis@chromium.org>
13162
13163         Make ScrollAnimator(Chromium)?Mac check scrollAnimatorEnabled()
13164         https://bugs.webkit.org/show_bug.cgi?id=67297
13165
13166         Reviewed by Tony Chang.
13167
13168         Explicitly disable smooth scrolling in chromium's DRT.
13169
13170         * DumpRenderTree/chromium/WebPreferences.cpp:
13171         (WebPreferences::applyTo):
13172
13173 2011-08-31  Oliver Hunt  <oliver@apple.com>
13174
13175         Add Checked<> tests to windows build
13176
13177         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
13178
13179 2011-08-30  Oliver Hunt  <oliver@apple.com>
13180
13181         Add support for checked arithmetic
13182         https://bugs.webkit.org/show_bug.cgi?id=67095
13183
13184         Reviewed by Sam Weinig.
13185
13186         Add test cases for Checked<>
13187
13188         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
13189         * TestWebKitAPI/Tests/CheckedArithmeticOperations.cpp: Added.
13190
13191 2011-08-31  Adam Barth  <abarth@webkit.org>
13192
13193         Move summary.html to garden-o-matic.html.
13194
13195         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html.
13196         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/garden-o-matic.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js.
13197         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js: Removed.
13198         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Removed.
13199
13200 2011-08-31  Noel Gordon  <noel.gordon@gmail.com>
13201
13202         [chromium] DRT linux should pass nativeKeyCode to plugins
13203         https://bugs.webkit.org/show_bug.cgi?id=66679
13204
13205         Reviewed by Tony Chang.
13206
13207         * DumpRenderTree/chromium/EventSender.cpp:
13208         (EventSender::keyDown):  Call a Chromium Linux GTK helper routine that returns the
13209         (simulated) nativeKeyCode for the given input keyboard code.
13210
13211 2011-08-31  Adam Barth  <abarth@webkit.org>
13212
13213         Tighen up UI for garden-o-matic tab strip
13214         https://bugs.webkit.org/show_bug.cgi?id=67266
13215
13216         Reviewed by Dimitri Glazkov.
13217
13218         This patch add CSS to move the tab strip to the edges of the page and
13219         to remove some gratuitous border-radius.
13220
13221         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css:
13222         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
13223
13224 2011-08-31  Adam Barth  <abarth@webkit.org>
13225
13226         garden-o-matic results view shouldn't be horrifically ugly
13227         https://bugs.webkit.org/show_bug.cgi?id=67264
13228
13229         Reviewed by Dimitri Glazkov.
13230
13231         Minor tweak to the results view CSS to make it fit in better with the new tabbed UI.
13232
13233         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
13234         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
13235         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css:
13236
13237 2011-08-31  Adam Barth  <abarth@webkit.org>
13238
13239         garden-o-matic needs a tabstrip
13240         https://bugs.webkit.org/show_bug.cgi?id=67263
13241
13242         Reviewed by Dimitri Glazkov.
13243
13244         This patch adds a basic tabstrip to garden-o-matic.  The results
13245         details view now opens in the "Results" tab, and the user can flip back
13246         and forth between the Summary and the Results tab quickly.  There's
13247         also a stub for a "Commits" tab, which is disabled.
13248
13249         This patch isn't the most beautiful in the world, but it gets the ball
13250         rolling on a multi-modal UI.
13251
13252         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
13253         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js:
13254         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
13255         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css:
13256         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
13257
13258 2011-08-31  Ai Makabi  <makabi@google.com>
13259
13260         Extract reference links from reftest test file.
13261         https://bugs.webkit.org/show_bug.cgi?id=66838
13262
13263         Reviewed by Shinichiro Hamaji.
13264
13265         * Scripts/webkitpy/layout_tests/reftests/__init__.py: Added.
13266         * Scripts/webkitpy/layout_tests/reftests/extract_reference_link.py: Added.
13267         * Scripts/webkitpy/layout_tests/reftests/extract_reference_link_unittest.py: Added.
13268
13269 2011-08-03  Philippe Normand  <pnormand@igalia.com>
13270
13271         [webkitpy] missing log import in common/net/resultsjsonparser.py
13272         https://bugs.webkit.org/show_bug.cgi?id=65606
13273
13274         Reviewed by Martin Robinson.
13275
13276         * Scripts/webkitpy/common/net/resultsjsonparser.py: Import the logging module.
13277
13278 2011-08-30  Philippe Normand  <pnormand@igalia.com>
13279
13280         [stylechecker] complains about NULL usage in some gst_structure_ calls
13281         https://bugs.webkit.org/show_bug.cgi?id=67194
13282
13283         Reviewed by Martin Robinson.
13284
13285         Don't warn about NULL in some gst_structure_ calls. 0 can't be
13286         used for these, a proper sentinel is required.
13287
13288         * Scripts/webkitpy/style/checkers/cpp.py:
13289         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
13290
13291 2011-08-30  Adam Barth  <abarth@webkit.org>
13292
13293         Remove old-style UI from garden-o-matic
13294         https://bugs.webkit.org/show_bug.cgi?id=67257
13295
13296         Reviewed by Dimitri Glazkov.
13297
13298         This code is nowhere near as awesome as the new hotness.  Let's remove
13299         it so it doesn't weigh us down.
13300
13301         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html: Removed.
13302         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/main.js: Removed.
13303         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
13304         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
13305         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/unexpected-passes.js: Removed.
13306         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.html: Removed.
13307
13308 2011-08-30  Takashi Toyoshima  <toyoshim@chromium.org>
13309
13310         [WebSocket] update pywebsocket to 0.6b5.
13311         https://bugs.webkit.org/show_bug.cgi?id=66924
13312
13313         Reviewed by Kent Tamura.
13314
13315         pywebsocket 0.6b2 supported minimum features of Hybi-08 specification.
13316         Now, pywebsocket 0.6b5 supports full features of Hybi-10 specification.
13317         0.6b3 or later supports a closing handshake handler and it is useful to
13318         verify client's closing handshake implementation.
13319         0.6b4 supports AbortedByUserException. It is useful to eliminate
13320         unnecessary server warning messages on intended abnormal closures.
13321         0.6b5 supports Python 2.5.1 or Mac 10.5.
13322
13323         * Scripts/webkitpy/thirdparty/__init__.py:
13324         Update download URL and directory path.
13325
13326 2011-08-30  Dan Bernstein  <mitz@apple.com>
13327
13328         Test for <rdar://problem/9281695> Add text search API for getting the DOM range of a text match
13329         https://bugs.webkit.org/show_bug.cgi?id=67230
13330
13331         Reviewed by Darin Adler.
13332
13333         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
13334         * TestWebKitAPI/Tests/mac/DOMRangeOfString.html: Added.
13335         * TestWebKitAPI/Tests/mac/DOMRangeOfString.mm: Added.
13336         (-[DOMRangeOfStringFrameLoadDelegate webView:didFinishLoadForFrame:]):
13337         (TestWebKitAPI::TEST):
13338
13339 2011-08-30  Aaron Colwell  <acolwell@chromium.org>
13340
13341         Add MediaSource API to HTMLMediaElement
13342         https://bugs.webkit.org/show_bug.cgi?id=64731
13343
13344         Reviewed by Eric Carlson.
13345
13346         * Scripts/build-webkit:
13347
13348 2011-08-30  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
13349
13350         Emit last progress notification before calling dispatchDidFinishLoad
13351         https://bugs.webkit.org/show_bug.cgi?id=28851
13352
13353         Reviewed by Adam Barth.
13354
13355         Implement dumpProgressFinishedCallback() in the layoutTestController for Qt,
13356         Chromium and GTK ports. The first two were simple changes.
13357
13358         Some refactoring was needed in GTK port to assure that the dumps are emitted
13359         correctly and the processing after the page load finished is in the right place.
13360
13361         To detect the load finished, we use the callback for the "load-status" notifier
13362         signal of the top loading frame. And we use the deprecated "load-finished" signal
13363         for dumping (if enabled) the progress finished expected output.
13364
13365         * DumpRenderTree/LayoutTestController.cpp:
13366         (LayoutTestController::LayoutTestController):
13367         (dumpProgressFinishedCallbackCallback):
13368         (LayoutTestController::staticFunctions):
13369         * DumpRenderTree/LayoutTestController.h:
13370         (LayoutTestController::dumpProgressFinishedCallback):
13371         (LayoutTestController::setDumpProgressFinishedCallback):
13372         * DumpRenderTree/chromium/LayoutTestController.cpp:
13373         (LayoutTestController::LayoutTestController):
13374         (LayoutTestController::dumpProgressFinishedCallback):
13375         (LayoutTestController::reset):
13376         * DumpRenderTree/chromium/LayoutTestController.h:
13377         (LayoutTestController::shouldDumpProgressFinishedCallback):
13378         (LayoutTestController::setShouldDumpProgressFinishedCallback):
13379         * DumpRenderTree/chromium/WebViewHost.cpp:
13380         (WebViewHost::didStopLoading):
13381         * DumpRenderTree/gtk/DumpRenderTree.cpp:
13382         (webViewLoadFinished):
13383         (topLoadingFrameLoadFinished):
13384         (webFrameLoadStatusNotified):
13385         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
13386         (LayoutTestController::reset):
13387         (LayoutTestController::dumpProgressFinishedCallback):
13388         * DumpRenderTree/qt/LayoutTestControllerQt.h:
13389
13390 2011-08-30  Filip Pizlo  <fpizlo@apple.com>
13391
13392         JavaScriptCore does not have a good mechanism for performance
13393         regression testing
13394         https://bugs.webkit.org/show_bug.cgi?id=67177
13395
13396         Reviewed by Darin Adler.
13397
13398         Added a new script, which allows for testing one, two, or more VMs
13399         at once on SunSpider, V8, and Kraken benchmarks.  Benchmark/VM runs
13400         are interleaved at random to minimize systematics.  The results
13401         that bencher prints includes all possible relevant information about
13402         how bencher was invoked.  Bencher can be easily used to measure
13403         performance changes down to fractions of a percent.
13404
13405         Bencher also includes experimental support for running benchmarks in
13406         a DumpRenderTree release build.  This works in SunSpider and V8, but
13407         not yet in Kraken.  Running benchmarks in DumpRenderTree more
13408         closely mimics the performance when running in browser.
13409
13410         * Scripts/bencher: Added.
13411
13412 2011-08-29  Alexey Proskuryakov  <ap@apple.com>
13413
13414         DumpRenderTree should begin each test with an empty cookie store
13415         https://bugs.webkit.org/show_bug.cgi?id=63545
13416         <rdar://problem/5666907>
13417
13418         Reviewed by Darin Adler.
13419
13420         * DumpRenderTree/mac/DumpRenderTree.mm:
13421         (resetDefaultsToConsistentValues):
13422         (setDefaultsToConsistentValuesForTesting):
13423         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
13424         (LayoutTestController::setAlwaysAcceptCookies):
13425         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
13426         (WTR::InjectedBundle::beginTesting):
13427         Use new SPIs to manipulate sessions.
13428
13429 2011-08-30  Sheriff Bot  <webkit.review.bot@gmail.com>
13430
13431         Unreviewed, rolling out r94055 and r94057.
13432         http://trac.webkit.org/changeset/94055
13433         http://trac.webkit.org/changeset/94057
13434         https://bugs.webkit.org/show_bug.cgi?id=67191
13435
13436         Many failures on Mac 10.5 (Requested by toyoshim on #webkit).
13437
13438         * Scripts/webkitpy/thirdparty/__init__.py:
13439
13440 2011-08-30  Adam Barth  <abarth@webkit.org>
13441
13442         Chromium Android should not be a core builder
13443         https://bugs.webkit.org/show_bug.cgi?id=67184
13444
13445         Reviewed by Ryosuke Niwa.
13446
13447         This bot hasn't been consistently green yet because it has never actually run!
13448
13449         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
13450         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
13451
13452 2011-08-30  Takashi Toyoshima  <toyoshim@chromium.org>
13453
13454         [WebSocket] update pywebsocket to 0.6b4.
13455         https://bugs.webkit.org/show_bug.cgi?id=66924
13456
13457         Reviewed by Kent Tamura.
13458
13459         pywebsocket 0.6b2 supported minimum features of Hybi-08 specification.
13460         Now, pywebsocket 0.6b4 supports full features of Hybi-10 specification.
13461         0.6b3 or later supports a closing handshake handler and it is useful to
13462         verify client's closing handshake implementation.
13463         0.6b4 supports AbortedByUserException. It is useful to eliminate
13464         unnecessary server warning messages on intended abnormal closures.
13465
13466         * Scripts/webkitpy/thirdparty/__init__.py:
13467         Update download URL and directory path.
13468
13469 2011-08-29  Tom Zakrajsek  <tomz@codeaurora.org>
13470
13471         sheriffbot "usage" help statements are wrong
13472         https://bugs.webkit.org/show_bug.cgi?id=67170
13473
13474         Reviewed by Adam Barth.
13475
13476         * Scripts/webkitpy/tool/bot/irc_command.py:
13477         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
13478         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
13479
13480 2011-08-29  Adam Barth  <abarth@webkit.org>
13481
13482         Add support for chomium-mac (non-CoreGraphics) baselines
13483         https://bugs.webkit.org/show_bug.cgi?id=67163
13484
13485         Reviewed by Eric Seidel.
13486
13487         This patch teaches the non-CG chromium-mac bots to look at non-CG
13488         baselines.  However, the patch still leaves CG as the default set of
13489         baselines.
13490
13491         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
13492         * Scripts/webkitpy/common/config/build.py:
13493         * Scripts/webkitpy/common/config/build_unittest.py:
13494         * Scripts/webkitpy/layout_tests/port/builders.py:
13495         * Scripts/webkitpy/layout_tests/port/chromium.py:
13496         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
13497         * Scripts/webkitpy/layout_tests/port/factory.py:
13498         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
13499         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
13500         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
13501         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
13502
13503 2011-08-29  Shawn Singh  <shawnsingh@chromium.org>
13504
13505         Small fix in chromium port of new-run-webkit-tests.
13506         https://bugs.webkit.org/show_bug.cgi?id=67078
13507
13508         Reviewed by Dirk Pranke.
13509
13510         * Scripts/webkitpy/layout_tests/port/chromium.py:
13511
13512 2011-08-29  Sheriff Bot  <webkit.review.bot@gmail.com>
13513
13514         Unreviewed, rolling out r93987, r93992, r93995, r93998, and
13515         r93999.
13516         http://trac.webkit.org/changeset/93987
13517         http://trac.webkit.org/changeset/93992
13518         http://trac.webkit.org/changeset/93995
13519         http://trac.webkit.org/changeset/93998
13520         http://trac.webkit.org/changeset/93999
13521         https://bugs.webkit.org/show_bug.cgi?id=67147
13522
13523         Many failing tests (Requested by ap on #webkit).
13524
13525         * DumpRenderTree/mac/DumpRenderTree.mm:
13526         (setDefaultsToConsistentValuesForTesting):
13527         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
13528         (WTR::InjectedBundle::beginTesting):
13529
13530 2011-08-26  Alexey Proskuryakov  <ap@apple.com>
13531
13532         DumpRenderTree should begin each test with an empty cookie store
13533         https://bugs.webkit.org/show_bug.cgi?id=63545
13534         <rdar://problem/5666907>
13535
13536         Reviewed by Darin Adler.
13537
13538         * DumpRenderTree/mac/DumpRenderTree.mm:
13539         (setDefaultsToConsistentValuesForTesting):
13540         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
13541         (WTR::InjectedBundle::beginTesting):
13542         Tell WebCore to use a temporary session, starting with a clean cookie store.
13543
13544 2011-08-23  Chris Marrin  <cmarrin@apple.com>
13545
13546         [mac] requestAnimationFrame support for mac port
13547         https://bugs.webkit.org/show_bug.cgi?id=59146
13548
13549         Reviewed by Simon Fraser.
13550
13551         Add support for request-animation-callback flag
13552
13553         * Scripts/build-webkit:
13554
13555 2011-08-27  Peter Beverloo  <peter@chromium.org>
13556
13557         Add a build bot for Chromium with OS(ANDROID)
13558         https://bugs.webkit.org/show_bug.cgi?id=66687
13559
13560         Reviewed by Adam Barth.
13561
13562         * BuildSlaveSupport/build.webkit.org-config/config.json:
13563
13564 2011-08-27  Peter Beverloo  <peter@chromium.org>
13565
13566         Update the master build bot configuration script to properly
13567         deal with the --chromium-android switch.
13568         https://bugs.webkit.org/show_bug.cgi?id=66687
13569
13570         The --chromium-android switch is needed when running
13571         update-webkit-chromium and build-webkit, as it'll tell the
13572         scripts to pull in the right dependencies and cross-compile
13573         WebKit using the Android NDK.
13574
13575         Reviewed by Adam Barth.
13576
13577         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
13578
13579 2011-08-27  Peter Beverloo  <peter@chromium.org>
13580
13581         Teach the update and build scripts how to handle dependencies
13582         and cross-compiling rules for the Chromium port on OS(ANDROID).
13583         https://bugs.webkit.org/show_bug.cgi?id=66687
13584
13585         Reviewed by Adam Barth.
13586
13587         * Scripts/build-webkit:
13588         * Scripts/update-webkit:
13589         * Scripts/update-webkit-chromium:
13590         * Scripts/webkitdirs.pm:
13591         (argumentsForConfiguration):
13592         (isChromium):
13593         (isChromiumAndroid):
13594         (determineIsChromiumAndroid):
13595         (buildChromium):
13596
13597 2011-08-26  Adam Barth  <abarth@webkit.org>
13598
13599         garden-o-matic should let me know that I need to run the local server
13600         https://bugs.webkit.org/show_bug.cgi?id=66985
13601
13602         Reviewed by Dimitri Glazkov.
13603
13604         The patch triggers an alert dialog whenever the user tries to use a
13605         feature that requires the local server.
13606
13607         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
13608         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js:
13609         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js:
13610
13611 2011-08-26  Dimitri Glazkov  <dglazkov@chromium.org>
13612
13613         garden-o-matic summary view should have pretty commit data information.
13614         https://bugs.webkit.org/show_bug.cgi?id=67069
13615
13616         Reviewed by Adam Barth.
13617
13618         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Added support for styling commit data.
13619         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Changed tests.
13620         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css: Defined global anchor color.
13621         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css: Styled commit data.
13622
13623 2011-08-26  Dimitri Glazkov  <dglazkov@chromium.org>
13624
13625         Tweak the use of Open Sans in garden-o-matic summary view to make it look decent on Windows and Linux.
13626         https://bugs.webkit.org/show_bug.cgi?id=67066
13627
13628         Reviewed by Adam Barth.
13629
13630         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary-mock.html: Tweaked.
13631         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Ditto.
13632
13633 2011-08-26  Raphael Kubo da Costa  <kubo@profusion.mobi>
13634
13635         [DRT] Include cairo.h in PixelDumpSupportCairo.h for EFL too.
13636         https://bugs.webkit.org/show_bug.cgi?id=67041
13637
13638         Reviewed by Martin Robinson.
13639
13640         This is a follow-up to r89053: EFL's PixelDumpSupportEfl.cpp includes
13641         PixelDumpSupportCairo.h, but the latter was #including cairo.h only
13642         when the GTK+ port was being built.
13643
13644         * DumpRenderTree/cairo/PixelDumpSupportCairo.h:
13645
13646 2011-08-26  Dimitri Glazkov  <dglazkov@chromium.org>
13647
13648         Clean up and line up layout in garden-o-matic summary view.
13649         https://bugs.webkit.org/show_bug.cgi?id=67057
13650
13651         1) Renamed ui.notifications.TestFailures to ui.notifications.TestsFailing to match BuildersFailing;
13652         2) Rearranged DOM structure a bit to accommodate recent changes;
13653         3) Added minimal CSS to line things up.
13654
13655         Reviewed by Adam Barth.
13656
13657         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js: Renamed TestFailures to TestsFailing.
13658         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary-mock.js: Added mocks for BuildersFailing.
13659         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js: Renamed TestFailures to TestsFailing.
13660         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js: Updated tests.
13661         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Changed DOM structure.
13662         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Updated tests.
13663         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css: Tweaked common appearance.
13664         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css: Added minimal layout/style.
13665         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary-mock.html: Added Google Web Fonts hook.
13666         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Ditto.
13667
13668 2011-08-26  Alice Liu  <alice.liu@apple.com>
13669
13670         https://bugs.webkit.org/show_bug.cgi?id=66823
13671         Add the request info to the WebPageProxy::CreateNewPage message
13672
13673         Reviewed by Anders Carlsson.
13674
13675         Update function declarations with additional parameter:
13676         * MiniBrowser/gtk/BrowserWindow.c:
13677         (createNewPage):
13678         * MiniBrowser/mac/BrowserWindowController.m:
13679         (createNewPage):
13680         * MiniBrowser/win/BrowserView.cpp:
13681         (createNewPage):
13682         * WebKitTestRunner/TestController.cpp:
13683         (WTR::TestController::createOtherPage):
13684         * WebKitTestRunner/TestController.h:
13685
13686 2011-08-26  Noel Gordon  <noel.gordon@gmail.com>
13687
13688         Make show-pretty-diff work on linux
13689         https://bugs.webkit.org/show_bug.cgi?id=67017
13690
13691         Reviewed by Adam Barth.
13692
13693         * Scripts/show-pretty-diff:
13694
13695 2011-08-25  Dimitri Glazkov  <dglazkov@chromium.org>
13696
13697         Reorganize garden-o-matic CSS files to match JS-built views.
13698         https://bugs.webkit.org/show_bug.cgi?id=66998
13699
13700         Reviewed by Adam Barth.
13701
13702         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/common.css: Added.
13703         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/notifications.css: Added.
13704         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/results.css: Added.
13705         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/summary.css: Removed.
13706         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary-mock.html: Changed to use new CSS files.
13707         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Ditto.
13708
13709 2011-08-25  Adam Barth  <abarth@webkit.org>
13710
13711         garden-o-matic should be able to roll out patches
13712         https://bugs.webkit.org/show_bug.cgi?id=66982
13713
13714         Reviewed by Dimitri Glazkov.
13715
13716         This patch just wires the rollout event through the controller to the
13717         local server, which is already set up to perform rollouts.
13718
13719         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js:
13720         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
13721         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js:
13722         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js:
13723
13724 2011-08-25  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
13725
13726         [Qt] [WK2] Assert in computeViewportAttributes when MiniBrowser started with -touch
13727         https://bugs.webkit.org/show_bug.cgi?id=66951
13728
13729         Reviewed by Andreas Kling.
13730
13731         Function computeViewportAttributes() gets called when height is still 0 because MiniBrowser sets
13732         width and height one by one and setWidth cause geometryChange event in QTouchWebView.
13733
13734         * MiniBrowser/qt/BrowserView.cpp:
13735         (BrowserView::resizeEvent):
13736
13737 2011-08-25  Dimitri Glazkov  <dglazkov@chromium.org>
13738
13739         garden-o-matic summary view should provide information on builders still in flight.
13740         https://bugs.webkit.org/show_bug.cgi?id=66992
13741
13742         Reviewed by Adam Barth.
13743
13744         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js: Added grabbing the list of in-flight builders.
13745         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js: Added buildersInFlightForRevision.
13746         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js: Added tests.
13747         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Split off updating builder results to own method.
13748         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Added tests.
13749
13750 2011-08-25  Adam Barth  <abarth@webkit.org>
13751
13752         webkit-patch rebaseline-expectations don't work when run outside the root
13753         https://bugs.webkit.org/show_bug.cgi?id=66966
13754
13755         Reviewed by Eric Seidel.
13756
13757         Turns out tool.path() used to be relative, which caused problems when
13758         running webkit-patch outside the root directory.  This patch makes it
13759         absolute.  I checked all callers, and they all appear to want this
13760         behavior.
13761
13762         It was unclear to me how to test this change because we only really
13763         have a testing strategy for webkitpy, not for the main script.
13764
13765         * Scripts/webkit-patch:
13766
13767 2011-08-25  Dimitri Glazkov  <dglazkov@chromium.org>
13768
13769         garden-o-matic summary needs a mock view that could be used for working on styling
13770         https://bugs.webkit.org/show_bug.cgi?id=66971
13771
13772         Reviewed by Adam Barth.
13773
13774         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary-mock.js: Added.
13775         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary-mock.html: Added.
13776
13777 2011-08-25  Anders Carlsson  <andersca@apple.com>
13778
13779         Return the correct value for NPNVsupportsCarbonBool
13780         https://bugs.webkit.org/show_bug.cgi?id=66964
13781
13782         Add a test.
13783
13784         Reviewed by Adam Roben.
13785
13786         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
13787         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/SupportsCarbonEventModel.cpp: Added.
13788         (SupportsCarbonEventModel::SupportsCarbonEventModel):
13789         (SupportsCarbonEventModel::testConvert):
13790         (SupportsCarbonEventModel::runTest):
13791         (SupportsCarbonEventModel::NPP_New):
13792
13793 2011-08-25  Dimitri Glazkov  <dglazkov@chromium.org>
13794
13795         garden-o-matic summary view should not be broken.
13796
13797         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/summary.css:
13798         (table.comparison): Added more specificity.
13799
13800 2011-08-25  Adam Barth  <abarth@webkit.org>
13801
13802         This command has been broken for a while.  This patch updates the
13803         callsite to the new API.
13804
13805         * Scripts/webkitpy/tool/commands/rebaseline.py:
13806
13807 2011-08-25  Dimitri Glazkov  <dglazkov@chromium.org>
13808
13809         garden-o-matic needs a way to report where and how tests are failing in the summary view.
13810         https://bugs.webkit.org/show_bug.cgi?id=66955
13811
13812         Introduce FailureGrid, your one-stop failure-indicating destination.
13813
13814         Reviewed by Adam Barth.
13815
13816         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html: Added new scripts.
13817         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures.js: Added.
13818         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/failures_unittests.js: Added.
13819         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Updated to use FailureGrid.
13820         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Updated tests.
13821         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Added new scripts.
13822
13823 2011-08-25  Igor Oliveira  <igor.oliveira@openbossa.org>
13824
13825         Add myself to the committers list
13826         https://bugs.webkit.org/show_bug.cgi?id=66941
13827
13828         Reviewed by Andreas Kling.
13829
13830         * Scripts/webkitpy/common/config/committers.py:
13831
13832 2011-08-25  Sheriff Bot  <webkit.review.bot@gmail.com>
13833
13834         Unreviewed, rolling out r93767.
13835         http://trac.webkit.org/changeset/93767
13836         https://bugs.webkit.org/show_bug.cgi?id=66931
13837
13838         Broke Qt bot. (Requested by yutak on #webkit).
13839
13840         * Scripts/webkitpy/thirdparty/__init__.py:
13841
13842 2011-08-25  Takashi Toyoshima  <toyoshim@chromium.org>
13843
13844         [WebSocket] update pywebsocket to 0.6b3 which supports closing handshake test
13845         https://bugs.webkit.org/show_bug.cgi?id=66924
13846
13847         Reviewed by Kent Tamura.
13848
13849         * Scripts/webkitpy/thirdparty/__init__.py:
13850
13851 2011-08-24  Adam Barth  <abarth@webkit.org>
13852
13853         The user can't close the details view in garden-o-matic
13854         https://bugs.webkit.org/show_bug.cgi?id=66911
13855
13856         Reviewed by Dimitri Glazkov.
13857
13858         In addition to adding a close button, this patch changes the test
13859         selector to use a <select> element and refactors the test selector to
13860         share code with the builder selector.
13861
13862         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
13863         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
13864         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions_unittests.js:
13865         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
13866
13867 2011-08-24  James Robinson  <jamesr@chromium.org>
13868
13869         [chromium] Stacktrace not in test output when a test crashes
13870         https://bugs.webkit.org/show_bug.cgi?id=66806
13871
13872         Reviewed by Dirk Pranke.
13873
13874         Appends the .error and .text output when a test crashes since on chromium the .text contains the actual stack
13875         trace.
13876
13877         * Scripts/webkitpy/layout_tests/port/chromium.py:
13878
13879 2011-08-24  Adam Barth  <abarth@webkit.org>
13880
13881         Remove empty directory.
13882
13883         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui: Removed.
13884
13885 2011-08-24  Adam Barth  <abarth@webkit.org>
13886
13887         Finish writing up details-view in garden-o-matic
13888         https://bugs.webkit.org/show_bug.cgi?id=66891
13889
13890         Reviewed by Dimitri Glazkov.
13891
13892         This patch mostly just copies a bunch of CSS from main.css to
13893         summary.css.  Presumably we'll delete main.css soon.
13894
13895         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
13896         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js:
13897         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions_unittests.js:
13898         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/summary.css:
13899         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
13900
13901 2011-08-24  James Robinson  <jamesr@chromium.org>
13902
13903         Some compositing LayoutTests flakily crashing on Mac 10.6
13904         https://bugs.webkit.org/show_bug.cgi?id=66740
13905
13906         Reviewed by Adam Barth.
13907
13908         Null-check obj->pluginTest before dereferencing it. This can be null in chromium if the plugin load fails, for
13909         example if the test attempts to load a PDF plugin the load will fail since we don't have such a thing in the
13910         chromium configuration.
13911
13912         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
13913         (NPP_Destroy):
13914
13915 2011-08-24  Anton D'Auria  <adauria@apple.com>
13916
13917         No review necessary.
13918
13919         Adding myself to list of committers.
13920
13921         * Scripts/webkitpy/common/config/committers.py:
13922
13923 2011-08-24  Avi Drissman  <avi@google.com>
13924
13925         Crash when restoring NSGraphicsContext when running the Chromium GPU layout tests on Mac
13926         https://bugs.webkit.org/show_bug.cgi?id=66875
13927
13928         Reviewed by James Robinson.
13929
13930         * DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
13931         (WebThemeEngineDRTMac::paintNSScrollerScrollbarThumb):
13932
13933 2011-08-23  Dimitri Glazkov  <dglazkov@chromium.org>
13934
13935         Make notifications appear in time order again.
13936         https://bugs.webkit.org/show_bug.cgi?id=66815
13937
13938         Reviewed by Adam Barth.
13939
13940         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Added setting index of the notification.
13941         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Updated the test.
13942
13943 2011-08-23  James Robinson  <jamesr@chromium.org>
13944
13945         [chromium] Need a way to test lost compositor context recovery
13946         https://bugs.webkit.org/show_bug.cgi?id=66820
13947
13948         Reviewed by Kenneth Russell.
13949
13950         Exposes a LayoutTestController interface in chromium to simulate a
13951         lost compositor context.
13952
13953         * DumpRenderTree/chromium/LayoutTestController.cpp:
13954         (LayoutTestController::LayoutTestController):
13955         (LayoutTestController::loseCompositorContext):
13956         * DumpRenderTree/chromium/LayoutTestController.h:
13957
13958 2011-08-23  Dimitri Glazkov  <dglazkov@chromium.org>
13959
13960         Change garden-o-matic frontend builder array to be a configuration map.
13961         https://bugs.webkit.org/show_bug.cgi?id=66816
13962
13963         Reviewed by Adam Barth.
13964
13965         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js: Made kBuilders a map, with value being a configuration object.
13966         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js: Adjusted callsites to expect a map.
13967         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/main.js: Ditto.
13968         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js: Ditto.
13969         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js: Ditto.
13970
13971 2011-08-23  Cary Clark  <caryclark@google.com>
13972
13973         Turn WebCanvas into CGContext before drawing scrollbar (DRT using Skia on Mac)
13974         https://bugs.webkit.org/show_bug.cgi?id=66803
13975
13976         This fixes the WebKit Canaries, in particular avoids a crash on tests
13977         like css1/basic/comments.html.
13978
13979         Reviewed by James Robinson.
13980
13981         * DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
13982         (WebThemeEngineDRTMac::paintNSScrollerScrollbarThumb):
13983         The NSGraphicsContext call graphicsContextWithGraphicsPort takes a void*
13984         parameter, so it took a WebCanvas* (mapped to SkCanvas* in Skia-land)
13985         and treated it as if it were a CGContextRef.
13986
13987 2011-08-23  Patrick Gansterer  <paroga@webkit.org>
13988
13989         Fix mergeChangeLogs test on native windows
13990         https://bugs.webkit.org/show_bug.cgi?id=66797
13991
13992         Reviewed by Adam Roben.
13993
13994         Windows has no /tmp/ directory and the "TMPDIR" environment variable isn't set in the default setup.
13995         Add the "TEMP" environment variable as additional source for the temporary directory used in writeTempFile.
13996
13997         * Scripts/webkitperl/VCSUtils_unittest/mergeChangeLogs.pl:
13998         (writeTempFile):
13999
14000 2011-08-23  Dimitri Glazkov  <dglazkov@chromium.org>
14001
14002         garden-o-matic relative time is wrong and updates too frequently.
14003         https://bugs.webkit.org/show_bug.cgi?id=66802
14004
14005         Reviewed by Adam Barth.
14006
14007         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js: Changed to account for rounding correctly.
14008         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js: Updated tests.
14009         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js: Upped the interval to 1 minute.
14010
14011 2011-08-23  Patrick Gansterer  <paroga@webkit.org>
14012
14013         Add missing quotes around filenames in mergeChangeLogs
14014         https://bugs.webkit.org/show_bug.cgi?id=66794
14015
14016         Reviewed by Adam Roben.
14017
14018         * Scripts/VCSUtils.pm:
14019         (mergeChangeLogs):
14020
14021 2011-08-23  Patrick Gansterer  <paroga@webkit.org>
14022
14023         webkit-perl tests fail on win32 Perl due to lack of list form of pipe open implementation
14024         https://bugs.webkit.org/show_bug.cgi?id=49080
14025
14026         Reviewed by Adam Roben.
14027
14028         Use the string form of pipe open instead (like we do at all other places).
14029
14030         * Scripts/VCSUtils.pm:
14031         (mergeChangeLogs):
14032
14033 2011-08-23  Adam Barth  <abarth@webkit.org>
14034
14035         garden-o-matic summary should be able to examine failures
14036         https://bugs.webkit.org/show_bug.cgi?id=66748
14037
14038         Reviewed by Dimitri Glazkov.
14039
14040         This patch refactors a bunch of code from summary.js to independent
14041         controllers in controllers.js.  This patch also contains the first
14042         iteration of binding the "examine" event to a the behavior showing the
14043         details view for the tests.  There's more work to do here to make this
14044         work properly (and elegantly), but this patch is a start.
14045
14046         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js:
14047         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js:
14048         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js:
14049         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers/rebaseline.js:
14050         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js:
14051         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
14052         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
14053         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js:
14054         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
14055
14056 2011-08-23  Adam Barth  <abarth@webkit.org>
14057
14058         webkitpy/common/config/build.py shouldn't mention non-existant Android directory
14059         https://bugs.webkit.org/show_bug.cgi?id=66737
14060
14061         Reviewed by Steve Block.
14062
14063         This directory doesn't exist anymore now that the Android port shares
14064         more code with Chromium.
14065
14066         * Scripts/webkitpy/common/config/build.py:
14067
14068 2011-08-22  Eric Seidel  <eric@webkit.org>
14069
14070         Hack start-queue.sh in an attempt to resurrect the cr-jail-1 cr-mac-ews bot.
14071         I won't have ssh access to the bot again for a couple days, so
14072         attempting to fix by changing this script.  If this doesn't
14073         work we'll just live with the it hung for now.
14074
14075         Unreviewed.
14076
14077         * EWSTools/start-queue.sh:
14078
14079 2011-08-22  Dimitri Glazkov  <dglazkov@chromium.org>
14080
14081         webkit-patch rebaseline-expectations command should not rearrange the entire test_expectations.txt file.
14082         https://bugs.webkit.org/show_bug.cgi?id=66727
14083
14084         Reviewed by Adam Barth.
14085
14086         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Made it possible to run TestExpectationSerializer without a TestConfigurationConverter,
14087             in which case the serializer treats all expectations as unparsed.
14088         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added unit tests.
14089
14090 2011-08-22  Dimitri Glazkov  <dglazkov@chromium.org>
14091
14092         Update unit test expectations after r93530.
14093
14094         * Scripts/webkitpy/layout_tests/port/factory_unittest.py: Updated.
14095         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py: Ditto.
14096
14097 2011-08-22  Dimitri Glazkov  <dglazkov@chromium.org>
14098
14099         [Chromium] Expand CPU/GPU/Skia/CG into a matrix in graphics_type.
14100         https://bugs.webkit.org/show_bug.cgi?id=66705
14101
14102         Reviewed by Adam Barth.
14103
14104         * Scripts/webkitpy/layout_tests/port/chromium.py: Added plumbing for CG graphics types.
14105         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Adjusted unit test.
14106
14107 2011-08-22  James Robinson  <jamesr@chromium.org>
14108
14109         [chromium] Update builder names for chromium gpu bots due to CG switchover
14110         https://bugs.webkit.org/show_bug.cgi?id=66693
14111
14112         Reviewed by Dirk Pranke.
14113
14114         Update the port and builder names for the chromium mac gpu bots to
14115         account for the CG transition.
14116
14117         * Scripts/webkitpy/layout_tests/port/builders.py:
14118
14119 2011-08-22  Dirk Pranke  <dpranke@chromium.org>
14120
14121         fix the GPU bot names on the layout test dashboard
14122         https://bugs.webkit.org/show_bug.cgi?id=66697
14123
14124         Reviewed by Dimitri Glazkov.
14125
14126         * TestResultServer/static-dashboards/builders.js:
14127
14128 2011-08-22  Dimitri Glazkov  <dglazkov@chromium.org>
14129
14130         garden-o-matic's relative time display should be updated regularly.
14131         https://bugs.webkit.org/show_bug.cgi?id=66682
14132
14133         Reviewed by Adam Barth.
14134
14135         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js: Moved Time to ui.Time, added code for regular updates.
14136         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Changed to use ui.Time.
14137         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js: Added tests.
14138
14139 2011-08-22  Xianzhu Wang  <wangxianzhu@chromium.org>
14140
14141         Pass --makeargs of build-webkit to chromium linux builder
14142         https://bugs.webkit.org/show_bug.cgi?id=66361
14143
14144         Reviewed by Tony Chang.
14145
14146         * Scripts/build-webkit:
14147         * Scripts/webkitdirs.pm:
14148         (buildChromiumMakefile):
14149         (buildChromium):
14150
14151 2011-08-22  Peter Beverloo  <peter@chromium.org>
14152
14153         Add myself to the contributor list.
14154         https://bugs.webkit.org/show_bug.cgi?id=66674
14155
14156         Reviewed by Tony Gentilcore.
14157
14158         * Scripts/webkitpy/common/config/committers.py:
14159
14160 2011-08-22  Balazs Kelemen  <kbalazs@webkit.org>
14161
14162         [Qt] ImageDiff does not consider --tolerance
14163         https://bugs.webkit.org/show_bug.cgi?id=66299
14164
14165         Reviewed by Csaba Osztrogonác.
14166
14167         * DumpRenderTree/qt/ImageDiff.cpp:
14168         (main):
14169
14170 2011-08-22  Csaba Osztrogonác  <ossy@webkit.org>
14171
14172         Update contributor list.
14173
14174         Reviewed by Zoltan Herczeg.
14175
14176         * Scripts/webkitpy/common/config/committers.py:
14177
14178 2011-08-22  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
14179
14180         [ORWT] Fix --reset-results --add-platform-exceptions combo
14181         https://bugs.webkit.org/show_bug.cgi?id=65464
14182
14183         Reviewed by Csaba Osztrogonác.
14184
14185         * Scripts/old-run-webkit-tests:
14186
14187 2011-08-20  Adam Barth  <abarth@webkit.org>
14188
14189         garden-o-matic summary.html should have UI for examining failures
14190         https://bugs.webkit.org/show_bug.cgi?id=66625
14191
14192         Reviewed by Dimitri Glazkov.
14193
14194         This patch introduces the notion of UI actions and converts our
14195         existing manual <button> manipulation with this new abstraction.  The
14196         patch also adds an Examine button to failure notifications.  In a
14197         future patch, I'll add a controller that wires this button up to
14198         something useful.
14199
14200         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
14201         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
14202         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions.js: Added.
14203         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/actions_unittests.js: Added.
14204         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
14205         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
14206         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js:
14207         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
14208
14209 2011-08-20  Adam Barth  <abarth@webkit.org>
14210
14211         garden-o-matic summary page should indiciate compile failures
14212         https://bugs.webkit.org/show_bug.cgi?id=66624
14213
14214         Reviewed by Dimitri Glazkov.
14215
14216         This patch creates a notification-based UI for build failures.  The new
14217         UI displays the same information that used to be displayed in the
14218         infobar.  There's lots of room for improvement in surfacing more
14219         information, but this is a start.
14220
14221         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js:
14222         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js:
14223         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js:
14224
14225 2011-08-19  James Robinson  <jamesr@chromium.org>
14226
14227         [chromium] Default to experimental-fully-parallel mode for chromium gpu port
14228         https://bugs.webkit.org/show_bug.cgi?id=66606
14229
14230         Reviewed by Dirk Pranke.
14231
14232         The chromium GPU configuration does not (currently) contain any tests
14233         incompatible with --experimental-fully-parallel, and this flag is a
14234         significan speedup because of the directory canvas/philip/tests which
14235         contains 802 tests in the same directory. Setting this option by
14236         default on this configuration is a 2x speed in release, 44s to 22s,
14237         and a 3x speedup in debug, 223s to 77s, on linux on a 16 core machine.
14238
14239         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
14240         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
14241         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
14242
14243 2011-08-19  Sheriff Bot  <webkit.review.bot@gmail.com>
14244
14245         Unreviewed, rolling out r93426.
14246         http://trac.webkit.org/changeset/93426
14247         https://bugs.webkit.org/show_bug.cgi?id=66607
14248
14249         Broke the Mac build by not building gtest.framework anymore
14250         (Requested by andersca on #webkit).
14251
14252         * TestWebKitAPI/Tests/WTF/StringOperators.cpp:
14253
14254 2011-08-19  James Robinson  <jamesr@chromium.org>
14255
14256         [chromium] Chromium GPU layout tests should use the normal process number selection, not force 1 worker
14257         https://bugs.webkit.org/show_bug.cgi?id=66603
14258
14259         Reviewed by Dirk Pranke.
14260
14261         The various ChromiumGpu ports should use the same child worker
14262         selection logic as other ports so they can run with full
14263         parallelization on multicore systems.
14264
14265         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
14266         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
14267
14268 2011-08-19  Alok Priyadarshi  <alokp@chromium.org>
14269
14270         [chromium] Remove dependency of WebThemeControlDRTWin on skia::PlatformCanvas
14271         https://bugs.webkit.org/show_bug.cgi?id=66570
14272
14273         Reviewed by Mihai Parparita.
14274
14275         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
14276         (WebThemeControlDRTWin::draw):
14277         (WebThemeControlDRTWin::drawTextField):
14278         (WebThemeControlDRTWin::drawProgressBar):
14279         * DumpRenderTree/chromium/WebThemeControlDRTWin.h:
14280
14281 2011-08-18  Dmitry Lomov  <dslomov@google.com>
14282
14283         TestWebKitAPI breaks on Windows due to fast malloc incompatibility.
14284         https://bugs.webkit.org/show_bug.cgi?id=66521
14285
14286         This patch ensures that gtest uses new and delete operators that are defined in JavaScriptCore.
14287
14288         Reviewed by David Levin.
14289
14290         * TestWebKitAPI/Tests/WTF/StringOperators.cpp: Removed redefinition of JS_EXPORTDATA.
14291
14292 2011-08-19  Adam Roben  <aroben@apple.com>
14293
14294         Add project file changes I left out in r93417
14295
14296         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
14297
14298 2011-08-19  Adam Roben  <aroben@apple.com>
14299
14300         Add a helper class for swizzling ObjC instance methods to TestWebKitAPI
14301
14302         Fixes <http://webkit.org/b/66565> TestWebKitAPI needs a way to override instance methods of
14303         NSScreen
14304
14305         Reviewed by Anders Carlsson.
14306
14307         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new files.
14308
14309         * TestWebKitAPI/Tests/TestWebKitAPI/mac/InstanceMethodSwizzler.mm: Added.
14310         (-[SimpleObject setValue:]): Original implementation.
14311
14312         (TestWebKitAPI::setValue2):
14313         (TestWebKitAPI::setValue3):
14314         Swizzled implementations.
14315
14316         (TestWebKitAPI::TEST): Test that we can swizzle and re-swizzle and un-swizzle an ObjC
14317         instance method.
14318
14319         * TestWebKitAPI/mac/InstanceMethodSwizzler.h: Added.
14320         * TestWebKitAPI/mac/InstanceMethodSwizzler.mm: Added.
14321         (TestWebKitAPI::InstanceMethodSwizzler::InstanceMethodSwizzler): Swizzle the method.
14322         (TestWebKitAPI::InstanceMethodSwizzler::~InstanceMethodSwizzler): Unswizzle the method.
14323
14324 2011-08-19  Tom Zakrajsek  <tomz@codeaurora.org>
14325
14326         Fix _path_to_apache() error reporting
14327         https://bugs.webkit.org/show_bug.cgi?id=66486
14328
14329         Reviewed by Ryosuke Niwa.
14330
14331         * Scripts/webkitpy/layout_tests/port/webkit.py:
14332
14333 2011-08-19  Adam Roben  <aroben@apple.com>
14334
14335         Windows build fix after r93404
14336
14337         * TestWebKitAPI/Tests/WebKit2/win/HideFindIndicator.cpp: Added missing #include.
14338
14339 2011-08-19  Tony Gentilcore  <tonyg@chromium.org>
14340
14341         Add leandrogracia to the committers list
14342         https://bugs.webkit.org/show_bug.cgi?id=66552
14343
14344         Also fix line endings in file.
14345
14346         No reviewed required.
14347
14348         * Scripts/webkitpy/common/config/committers.py:
14349
14350 2011-08-19  Amruth Raj  <amruthraj@motorola.com>
14351
14352         Fix compilation issue in WebKitTestRunner
14353         https://bugs.webkit.org/show_bug.cgi?id=66554
14354
14355         Reviewed by Martin Robinson.
14356
14357         * WebKitTestRunner/PixelDumpSupport.cpp:
14358
14359 2011-08-19  Adam Roben  <aroben@apple.com>
14360
14361         Pull some of DynamicDeviceScaleFactor's code up into a shareable base class
14362
14363         This will make it easier to write other WebKit-agnostic tests
14364
14365         Fixes <http://webkit.org/b/66558> Would like to be able to reuse DynamicDeviceScaleFactor's
14366         code in other WebKit-agnostic tests
14367
14368         Reviewed by Anders Carlsson.
14369
14370         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new files.
14371
14372         * TestWebKitAPI/Tests/mac/DynamicDeviceScaleFactor.mm: Fixed copyright. Moved a bunch of
14373         code to WebKitAgnosticTest.h/mm. Changed to inherit from WebKitAgnosticTest.
14374         (TestWebKitAPI::DynamicDeviceScaleFactor::url): Simple getter.
14375         (TestWebKitAPI::DynamicDeviceScaleFactor::didLoadURL): Calls through to runTest.
14376         (TestWebKitAPI::DynamicDeviceScaleFactor::runTest): Removed loading code which
14377         WebKitAgnosticTest handles for us.
14378         (TestWebKitAPI::TEST_F): Removed view-creation code which WebKitAgnosticTest handles for us.
14379
14380         * TestWebKitAPI/mac/WebKitAgnosticTest.h: Added. Code came from DynamicDeviceScaleFactor.
14381         (TestWebKitAPI::WebKitAgnosticTest::loadAndWaitUntilDone): Code came from
14382         DynamicDeviceScaleFactor::runTest.
14383
14384         * TestWebKitAPI/mac/WebKitAgnosticTest.mm: Copied from Tools/TestWebKitAPI/Tests/mac/DynamicDeviceScaleFactor.mm.
14385         (TestWebKitAPI::WebKitAgnosticTest::runWebKit1Test): Code came from
14386         DynamicDeviceScaleFactor.WebKit.
14387         (TestWebKitAPI::WebKitAgnosticTest::runWebKit2Test): Code came from
14388         DynamicDeviceScaleFactor.WebKit2.
14389
14390 2011-08-19  Adam Roben  <aroben@apple.com>
14391
14392         Clean up #includes in TestWebKitAPI
14393
14394         Fixes <http://webkit.org/b/66556> TestWebKitAPI's prefix header isn't very helpful
14395
14396         Reviewed by Anders Carlsson.
14397
14398         * TestWebKitAPI/TestWebKitAPIPrefix.h: Added wtf/Platform.h. Added gtest/gtest.h for C++
14399         files and WebKit/WebKit.h for ObjC files.
14400
14401         * TestWebKitAPI/InjectedBundleController.cpp:
14402         * TestWebKitAPI/JavaScriptTest.h:
14403         * TestWebKitAPI/PlatformUtilities.h:
14404         * TestWebKitAPI/PlatformWebView.h:
14405         * TestWebKitAPI/Test.h:
14406         * TestWebKitAPI/Tests/WTF/StringOperators.cpp:
14407         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
14408         * TestWebKitAPI/Tests/WTF/VectorReverse.cpp:
14409         * TestWebKitAPI/Tests/WebKit2/AboutBlankLoad.cpp:
14410         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest.cpp:
14411         * TestWebKitAPI/Tests/WebKit2/CookieManager.cpp:
14412         * TestWebKitAPI/Tests/WebKit2/DocumentStartUserScriptAlertCrash.cpp:
14413         * TestWebKitAPI/Tests/WebKit2/DocumentStartUserScriptAlertCrash_Bundle.cpp:
14414         * TestWebKitAPI/Tests/WebKit2/DownloadDecideDestinationCrash.cpp:
14415         * TestWebKitAPI/Tests/WebKit2/EvaluateJavaScript.cpp:
14416         * TestWebKitAPI/Tests/WebKit2/FailedLoad.cpp:
14417         * TestWebKitAPI/Tests/WebKit2/Find.cpp:
14418         * TestWebKitAPI/Tests/WebKit2/ForceRepaint.cpp:
14419         * TestWebKitAPI/Tests/WebKit2/FrameMIMETypeHTML.cpp:
14420         * TestWebKitAPI/Tests/WebKit2/FrameMIMETypePNG.cpp:
14421         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp:
14422         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle_Bundle.cpp:
14423         * TestWebKitAPI/Tests/WebKit2/InjectedBundleBasic.cpp:
14424         * TestWebKitAPI/Tests/WebKit2/InjectedBundleBasic_Bundle.cpp:
14425         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash.cpp:
14426         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
14427         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
14428         * TestWebKitAPI/Tests/WebKit2/PreventEmptyUserAgent.cpp:
14429         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp:
14430         * TestWebKitAPI/Tests/WebKit2/RestoreSessionStateContainingFormData.cpp:
14431         * TestWebKitAPI/Tests/WebKit2/SpacebarScrolling.cpp:
14432         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
14433         * TestWebKitAPI/Tests/WebKit2/WKString.cpp:
14434         * TestWebKitAPI/Tests/WebKit2/WKStringJSString.cpp:
14435         * TestWebKitAPI/Tests/WebKit2/win/AltKeyGeneratesWMSysCommand.cpp:
14436         * TestWebKitAPI/Tests/WebKit2/win/DoNotCopyANullCFURLResponse.cpp:
14437         * TestWebKitAPI/Tests/WebKit2/win/HideFindIndicator.cpp:
14438         * TestWebKitAPI/Tests/WebKit2/win/ResizeViewWhileHidden.cpp:
14439         * TestWebKitAPI/Tests/WebKit2/win/TranslateMessageGeneratesWMChar.cpp:
14440         * TestWebKitAPI/Tests/WebKit2/win/WMCloseCallsUIClientClose.cpp:
14441         * TestWebKitAPI/Tests/WebKit2/win/WMPrint.cpp:
14442         * TestWebKitAPI/Tests/mac/DynamicDeviceScaleFactor.mm:
14443         * TestWebKitAPI/TestsController.cpp:
14444         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
14445         Removed now-unnecessary #includes. Moved #includes of Test.h in with the rest of the
14446         #includes since it is no longer a special header.
14447
14448 2011-08-18  Dirk Pranke  <dpranke@chromium.org>
14449
14450         new-run-webkit-tests hung while acquiring http lock on snow leopard bots
14451         https://bugs.webkit.org/show_bug.cgi?id=64886
14452
14453         Temporarily disable the http locking to work around the issue.
14454         I'm not actually sure if this is going to work or improve things
14455         much.
14456
14457         Reviewed by Ryosuke Niwa.
14458
14459         * Scripts/webkitpy/layout_tests/port/mac.py:
14460
14461 2011-08-18  Tony Chang  <tony@chromium.org>
14462
14463         add embedded png checksums to WebKitTestRunner
14464         https://bugs.webkit.org/show_bug.cgi?id=66494
14465
14466         Looks like WebKitTestRunner has never supported embedded checksums.  This copies
14467         some code from DRT and adds it to the WebKitTestRunner.
14468
14469         Reviewed by Darin Adler.
14470
14471         * WebKitTestRunner/CyclicRedundancyCheck.cpp: Copied from Tools/DumpRenderTree
14472         * WebKitTestRunner/CyclicRedundancyCheck.h: Copied from Tools/DumpRenderTree
14473         * WebKitTestRunner/GNUmakefile.am: Add new files
14474         * WebKitTestRunner/PixelDumpSupport.cpp: Copied from Tools/DumpRenderTree
14475         * WebKitTestRunner/PixelDumpSupport.h: Copied from Tools/DumpRenderTree
14476         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj: Add new files
14477         * WebKitTestRunner/cairo/TestInvocationCairo.cpp:
14478         (WTR::dumpBitmap): Refactor to use PixelDumpSupport.
14479         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
14480         * WebKitTestRunner/cg/TestInvocationCG.cpp:
14481         (WTR::dumpBitmap): Refactor to use PixelDumpSupport.
14482         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
14483         * WebKitTestRunner/win/WebKitTestRunner.vcproj: Add new files
14484
14485 2011-08-18  Shawn Singh  <shawnsingh@chromium.org>
14486
14487         https://bugs.webkit.org/show_bug.cgi?id=47240
14488
14489         Fixed a cygwin path problem in the chromium port of diff_image;
14490         Also made the return values of the diff_image function more consistent.
14491
14492         Reviewed by Dirk Pranke.
14493
14494         * Scripts/webkitpy/layout_tests/port/chromium.py:
14495
14496 2011-08-17  Alejandro G. Castro  <alex@igalia.com>
14497
14498         [GTK] Fix compilation problems with deprecations in gtk+
14499         https://bugs.webkit.org/show_bug.cgi?id=66073
14500
14501         Reviewed by Martin Robinson.
14502
14503         * DumpRenderTree/gtk/DumpRenderTree.cpp:
14504         (main): G_CONST_RETURN was deprecated
14505         * GtkLauncher/main.c:
14506         (createWindow): Added gtk_box_new conditional compilation for
14507         gtk+-3.
14508         * MiniBrowser/gtk/BrowserWindow.c:
14509         (browser_window_init): Replaced gtk_vbox_new with gtk_box_new, we
14510         are just supporting gtk+-3 for WebKit2.
14511
14512 2011-08-18  Anders Carlsson  <andersca@apple.com>
14513
14514         Fix libc++ C++0x build
14515         https://bugs.webkit.org/show_bug.cgi?id=66479
14516
14517         Reviewed by Adam Roben.
14518
14519         Add missing includes.
14520
14521         * DumpRenderTree/LayoutTestController.cpp:
14522         * DumpRenderTree/mac/CheckedMalloc.cpp:
14523
14524 2011-08-18  Adam Roben  <aroben@apple.com>
14525
14526         Test that WebKit updates style when a WebView is moved between differently-scaled windows
14527
14528         Test for <http://webkit.org/b/66412> <rdar://problem/9971958> WebKit doesn't react when a
14529         WebView is moved between windows with different backing scale factors
14530
14531         Reviewed by Anders Carlsson.
14532
14533         * TestWebKitAPI/JavaScriptTest.cpp:
14534         (TestWebKitAPI::runJSTest): Moved a little bit of code from here...
14535         (TestWebKitAPI::compareJSResult): ...to here. Also made the error message more similar to
14536         gtest's built-in error messages.
14537
14538         * TestWebKitAPI/JavaScriptTest.h: Added overloads of runJSTest that take a WebView * and
14539         WKView * for convenience on Mac. Added compareJSResult helper function for implementing
14540         those overloads.
14541
14542         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Added new files.
14543
14544         * TestWebKitAPI/Tests/mac/DynamicDeviceScaleFactor.mm: Added.
14545         (-[FrameLoadDelegate initWithDidFinishLoadBoolean:]): Simple initializer.
14546         (-[FrameLoadDelegate webView:didFinishLoadForFrame:]): Record that the load finished.
14547         (TestWebKitAPI::didFinishLoadForFrame): Record that the load finished.
14548         (TestWebKitAPI::setPageLoaderClient): Set up the client.
14549         (TestWebKitAPI::DynamicDeviceScaleFactor::DynamicDeviceScaleFactor): Simple constructor.
14550         (TestWebKitAPI::DynamicDeviceScaleFactor::createWindow): Creates a
14551         SyntheticBackingScaleFactorWindow and returns it.
14552         (TestWebKitAPI::DynamicDeviceScaleFactor::runTest): Loads devicePixelRatio.html and checks
14553         that WebKit uses the correct scale factor when the WebView is not in a window, is put in a
14554         window, is moved to a differently-scaled window, and is taken out of the window.
14555         (TestWebKitAPI::DynamicDeviceScaleFactor::loadURL): Helper function with overloads for
14556         WebKit1 and WebKit2.
14557         (TestWebKitAPI::TEST_F): Runs the test, with overloads for WebKit1 and WebKit2.
14558
14559         * TestWebKitAPI/Tests/mac/devicePixelRatio.html: Added.
14560
14561         * TestWebKitAPI/mac/JavaScriptTestMac.mm: Added.
14562         (TestWebKitAPI::runJSTest): Fairly simple overloads for WebView * and WKView *.
14563
14564         * TestWebKitAPI/mac/SyntheticBackingScaleFactorWindow.h: Added.
14565         * TestWebKitAPI/mac/SyntheticBackingScaleFactorWindow.m: Added.
14566         (-[SyntheticBackingScaleFactorWindow initWithContentRect:styleMask:backing:defer:]): Simple
14567         initializer.
14568         (-[SyntheticBackingScaleFactorWindow setBackingScaleFactor:]): Simple setter.
14569
14570         (-[SyntheticBackingScaleFactorWindow backingScaleFactor]):
14571         (-[SyntheticBackingScaleFactorWindow userSpaceScaleFactor]):
14572         Overrides of NSWindow methods that WebKit uses to determine the device scale factor.
14573
14574 2011-08-18  Dmitry Lomov  <dslomov@google.com>
14575
14576         https://bugs.webkit.org/show_bug.cgi?id=66425
14577         check-webkit-style should detect incorrectly formatted file change descriptions.
14578         Description, if present, should be separated from colon with a space.
14579
14580         Reviewed by David Levin.
14581
14582         * Scripts/webkitpy/style/checkers/changelog.py: Fix.
14583         * Scripts/webkitpy/style/checkers/changelog_unittest.py: Unit tests.
14584
14585 2011-08-18  Adam Barth  <abarth@webkit.org>
14586
14587         webkit-patch pretty-diff should be shown in main help
14588         https://bugs.webkit.org/show_bug.cgi?id=66404
14589
14590         Reviewed by Dimitri Glazkov.
14591
14592         As requested by dglazkov.
14593
14594         * Scripts/webkitpy/tool/commands/prettydiff.py:
14595
14596 2011-08-17  Dmitry Lomov  <dslomov@google.com>
14597
14598         https://bugs.webkit.org/show_bug.cgi?id=66400
14599         Fix TestWebKitAPI build on Mac OS.
14600
14601         Reviewed by David Levin.
14602
14603         * TestWebKitAPI/Configurations/Base.xcconfig: Added ICU headers to the HEADER_SEARCH_PATH.
14604
14605 2011-08-17  Dimitri Glazkov  <dglazkov@chromium.org>
14606
14607         garden-o-matic spring cleaning: move scripts, styles, and images in the respective directories.
14608         https://bugs.webkit.org/show_bug.cgi?id=66427
14609
14610         Reviewed by Adam Barth.
14611
14612         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js: Removed.
14613         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js: Removed.
14614         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot_unittests.js: Removed.
14615         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js: Removed.
14616         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js: Removed.
14617         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js: Removed.
14618         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js: Removed.
14619         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js: Removed.
14620         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js: Removed.
14621         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm.js: Removed.
14622         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm_unittests.js: Removed.
14623         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js: Removed.
14624         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js: Removed.
14625         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NRWTResultsParser.js: Removed.
14626         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js: Removed.
14627         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js: Removed.
14628         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ORWTResultsParser.js: Removed.
14629         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/PersistentCache.js: Removed.
14630         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css: Removed.
14631         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm.js: Removed.
14632         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm_unittests.js: Removed.
14633         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js: Removed.
14634         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js: Removed.
14635         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities_unittests.js: Removed.
14636         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js: Removed.
14637         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBugzilla.js: Removed.
14638         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js: Removed.
14639         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Removed.
14640         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Removed.
14641         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/bugzilla_unittests.js: Removed.
14642         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js: Removed.
14643         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders_unittests.js: Removed.
14644         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js: Removed.
14645         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js: Removed.
14646         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js: Removed.
14647         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers: Removed.
14648         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers.js: Removed.
14649         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers/rebaseline.js: Removed.
14650         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-green.png: Removed.
14651         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-red.png: Removed.
14652         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
14653         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/images: Added.
14654         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/images/favicon-green.png: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-green.png.
14655         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/images/favicon-red.png: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-red.png.
14656         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
14657         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css: Removed.
14658         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js: Removed.
14659         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js: Removed.
14660         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model_unittests.js: Removed.
14661         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net.js: Removed.
14662         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net_unittests.js: Removed.
14663         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/rebaseline.html:
14664         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js: Removed.
14665         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js: Removed.
14666         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
14667         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Bugzilla.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js.
14668         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Buildbot.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js.
14669         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Buildbot_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot_unittests.js.
14670         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Builder.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js.
14671         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Builder_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js.
14672         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FailingTestsBugForm.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js.
14673         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FailingTestsBugForm_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js.
14674         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FlakyLayoutTestDetector.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js.
14675         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FlakyLayoutTestDetector_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js.
14676         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FlakyTestBugForm.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm.js.
14677         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/FlakyTestBugForm_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm_unittests.js.
14678         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/LayoutTestHistoryAnalyzer.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js.
14679         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/LayoutTestResultsLoader.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js.
14680         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/NRWTResultsParser.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NRWTResultsParser.js.
14681         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/NewBugForm.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js.
14682         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/NewBugForm_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js.
14683         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ORWTResultsParser.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ORWTResultsParser.js.
14684         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/PersistentCache.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/PersistentCache.js.
14685         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/TestRelatedBugForm.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm.js.
14686         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/TestRelatedBugForm_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm_unittests.js.
14687         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Trac.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js.
14688         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Utilities.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js.
14689         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/Utilities_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities_unittests.js.
14690         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ViewController.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js.
14691         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/WebKitBugzilla.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBugzilla.js.
14692         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/WebKitBuildbot.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js.
14693         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js.
14694         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/base_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js.
14695         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/bugzilla_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/bugzilla_unittests.js.
14696         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js.
14697         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/builders_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders_unittests.js.
14698         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js.
14699         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/checkout_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js.
14700         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/config.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js.
14701         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers.
14702         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/controllers.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers.js.
14703         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/main.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js.
14704         ():
14705         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js.
14706         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/model_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model_unittests.js.
14707         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net.js.
14708         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/net_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net_unittests.js.
14709         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js.
14710         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/results_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js.
14711         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/trac_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/trac_unittests.js.
14712         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js.
14713         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results.js.
14714         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/results_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results_unittests.js.
14715         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui_unittests.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js.
14716         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/unexpected-passes.js: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.js.
14717         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/TestFailures.css: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css.
14718         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/main.css: Copied from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css.
14719         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html:
14720         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/trac_unittests.js: Removed.
14721         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js: Removed.
14722         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results.js: Removed.
14723         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results_unittests.js: Removed.
14724         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js: Removed.
14725         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.html:
14726         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.js: Removed.
14727
14728 2011-08-17  Jochen Eisinger  <jochen@chromium.org>
14729
14730         Make SVNTest.test_svn_apply use the same timezone logic as VCSUtils.pm
14731         https://bugs.webkit.org/show_bug.cgi?id=65877
14732
14733         Reviewed by Eric Seidel.
14734
14735         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
14736
14737 2011-08-17  Dimitri Glazkov  <dglazkov@chromium.org>
14738
14739         Remove base.keys and replace it with Object.keys.
14740         https://bugs.webkit.org/show_bug.cgi?id=66422
14741
14742         Reviewed by Adam Barth.
14743
14744         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Removed base.keys.
14745         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Removed unit tests for it.
14746         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js: Replaced with Object.keys
14747
14748 2011-08-17  Dimitri Glazkov  <dglazkov@chromium.org>
14749
14750         garden-o-matic Summary view should have items in descending chronological order.
14751         https://bugs.webkit.org/show_bug.cgi?id=66403
14752
14753         Reviewed by Adam Barth.
14754
14755         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js: Renamed "push" to "add" for clarity, cleaned up some stuff.
14756         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Rewrote add to insert DOM elements in order.
14757         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Added unit tests.
14758
14759 2011-08-17  Dimitri Glazkov  <dglazkov@chromium.org>
14760
14761         Fix a few errors in garden-o-matic frontend unit tests.
14762         https://bugs.webkit.org/show_bug.cgi?id=66421
14763
14764         Reviewed by Adam Barth.
14765
14766         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Changed my excellent test to view time as variable continuum.
14767         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders_unittests.js: Tweaked the test to accommodate new builder names.
14768         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Hurriedly covered up my prior sins.
14769
14770 2011-08-17  Adam Barth  <abarth@webkit.org>
14771
14772         Update Tools to account for (CG) being added to Chromium builder names
14773         https://bugs.webkit.org/show_bug.cgi?id=66410
14774
14775         Reviewed by Adam Barth.
14776
14777         Some of these names will change back once we turn on Skia bots.
14778
14779         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders_unittests.js:
14780         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
14781         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
14782         * Scripts/webkitpy/layout_tests/port/builders.py:
14783         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
14784         * TestResultServer/static-dashboards/builders.js:
14785         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
14786
14787 2011-08-16  Dimitri Glazkov  <dglazkov@chromium.org>
14788
14789         Wire up relative time to garden-o-matic summary view.
14790         https://bugs.webkit.org/show_bug.cgi?id=66343
14791
14792         Reviewed by Adam Barth.
14793
14794         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js: Add time to the commitData object.
14795         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model_unittests.js: Adjusted unit tests.
14796         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Wired up relative date.
14797         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Updated unit tests.
14798         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/trac_unittests.js: Adjusted unit tests.
14799
14800 2011-08-17  Adam Roben  <aroben@apple.com>
14801
14802         Make run-api-tests run tests in alphabetical order
14803
14804         Fixes <http://webkit.org/b/66401> run-api-tests runs tests in a semi-random order
14805
14806         Reviewed by David Kilzer.
14807
14808         * Scripts/run-api-tests:
14809         (runAllTests):
14810         (runAllTestsInSuite):
14811         Sort the lists of tests and suites before iterating over them.
14812
14813 2011-08-17  Adam Barth  <abarth@webkit.org>
14814
14815         garden-o-matic should be able to rebaseline expected failures
14816         https://bugs.webkit.org/show_bug.cgi?id=66204
14817
14818         Reviewed by Dimitri Glazkov.
14819
14820         This patch introduces a new view for displaying expect, actual, and
14821         differences between results.  The new view also has a fledgling
14822         controller that handles some basic operations.  The new view is
14823         exercised by rebaseline.html, which lets you rebaselines expected
14824         failures, but the new view is not fully integrated into
14825         garden-o-matic.html
14826
14827         More patches will be required to make this stuff fully working, but
14828         this patch was already somewhat spiraling out of control.  Hopefully
14829         this patch will serve as a good starting point for further development.
14830
14831         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers: Added.
14832         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers.js: Added.
14833         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/controllers/rebaseline.js: Added.
14834         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
14835         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
14836         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
14837         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/rebaseline.html: Added.
14838         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
14839         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
14840         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
14841         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results.js:
14842         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results_unittests.js: Added.
14843
14844 2011-08-16  Chang Shu  <cshu@webkit.org>
14845
14846         Support reset in WebCore::Internals
14847         https://bugs.webkit.org/show_bug.cgi?id=66307
14848
14849         Reviewed by Dimitri Glazkov.
14850
14851         Added call to resetInternalsObject in DRT.
14852
14853         * DumpRenderTree/chromium/TestShell.cpp:
14854         (TestShell::resetTestController):
14855         * DumpRenderTree/mac/DumpRenderTree.mm:
14856         (resetWebViewToConsistentStateBeforeTesting):
14857         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
14858         (WebCore::WebPage::resetSettings):
14859
14860 2011-08-17  Alexis Menard  <alexis.menard@openbossa.org>
14861
14862         Add a new build slave for the Qt port on Mac OS SnowLeopard.
14863         https://bugs.webkit.org/show_bug.cgi?id=66392
14864
14865         Modify the config file to include the new build slave.
14866
14867         Reviewed by Adam Roben.
14868
14869         * BuildSlaveSupport/build.webkit.org-config/config.json:
14870
14871 2011-08-17  Sam White  <samuel.white@rochester.edu>
14872
14873         AccessibilityObject levels are inconsistent
14874         https://bugs.webkit.org/show_bug.cgi?id=66180
14875
14876         Added the ability to get numeric attribute values using numberAttributeValue. This
14877         function complements the existing stringAttributeValue and boolAttributeValue functions.
14878         The addition of numberAttributeValue was necessary because the stringAttributeValue
14879         function does a type check and will only return strings. This limitation made it
14880         impossible to get values for attributes that returned an NSNumber.
14881
14882         Reviewed by Chris Fleizach.
14883
14884         * DumpRenderTree/AccessibilityUIElement.cpp:
14885         (numberAttributeValueCallback):
14886         (AccessibilityUIElement::getJSClass):
14887         * DumpRenderTree/AccessibilityUIElement.h:
14888         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
14889         (AccessibilityUIElement::numberAttributeValue):
14890         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
14891         (AccessibilityUIElement::numberAttributeValue):
14892         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
14893         (AccessibilityUIElement::numberAttributeValue):
14894
14895 2011-08-17  Adam Roben  <aroben@apple.com>
14896
14897         Tell Trac logs always to display all specified revisions in a single page
14898
14899         Trac imposes a default limit of 100 revisions per log page. We need to tell it to allow more
14900         revisions than that.
14901
14902         Fixes <http://webkit.org/b/66373> Links to Trac from TestFailures page don't always show all
14903         relevant revisions
14904
14905         Reviewed by Daniel Bates.
14906
14907         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
14908         (trac.logURL): Add a limit parameter to the log URL that allows all revisions to be seen in
14909         a single page.
14910
14911         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js:
14912         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/trac_unittests.js:
14913         Updated tests for the above.
14914
14915 2011-08-17  Adam Roben  <aroben@apple.com>
14916
14917         Fix TestFailures's link to file a bug about TestFailures itself
14918
14919         Fixes <http://webkit.org/b/66387> REGRESSION (r92135): TestFailures page's link to file a
14920         bug about itself is broken
14921
14922         Reviewed by Dan Bates.
14923
14924         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
14925         (ViewController.prototype._domForAuxiliaryUIElements): Added a path separator to separate
14926         the host from the path.
14927
14928 2011-08-17  Adam Roben  <aroben@apple.com>
14929
14930         Teach TestFailures to ignore unbelievably short test runs
14931
14932         Fixes <http://webkit.org/b/66385> TestFailures page thinks all tests passed in
14933         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14956
14934
14935         Reviewed by Dan Bates.
14936
14937         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
14938         (Builder.prototype.getNumberOfFailingTests): If it looks like all tests passed, but
14939         run-webkit-tests took less than 10 seconds to run, assume that some weird error occurred
14940         that caused it not to run any tests at all (as happened for a while due to
14941         <http://webkit.org/b/64988>). Bumped the cache version to evict old, buggy cached data.
14942
14943         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
14944         Test for the above.
14945
14946         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
14947         (LayoutTestResultsLoader.prototype.start): Bumped the cache version to evict old, buggy cached data.
14948
14949 2011-08-16  Adam Barth  <abarth@webkit.org>
14950
14951         Add a CG qualifier similar to the GPU qualifier
14952         https://bugs.webkit.org/show_bug.cgi?id=66359
14953
14954         Reviewed by Eric Seidel.
14955
14956         Once we have a Skia configuration on Chromium Mac, we'll use this CG
14957         qualifier to note that expectations apply only to the CG configuration.
14958
14959         * Scripts/webkitpy/layout_tests/port/chromium.py:
14960         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
14961
14962 2011-08-16  Eric Seidel  <eric@webkit.org>
14963
14964         REGRESSION(92717): WinPort needs to parse version information from port_name
14965         https://bugs.webkit.org/show_bug.cgi?id=66325
14966
14967         Unreviewed.  Fixing an exception/typo seen on the bots (and testing to make sure it works).
14968
14969         * Scripts/webkitpy/layout_tests/port/mac.py:
14970         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
14971
14972 2011-08-16  Adam Barth  <abarth@webkit.org>
14973
14974         Change webkit-patch analyze-baselines to print in a more unix-friendly way
14975         https://bugs.webkit.org/show_bug.cgi?id=66349
14976
14977         Reviewed by James Robinson.
14978
14979         As requested by jamesr.
14980
14981         * Scripts/webkitpy/tool/commands/rebaseline.py:
14982
14983 2011-08-16  Eric Seidel  <eric@webkit.org>
14984
14985         REGRESSION(92717): WinPort needs to parse version information from port_name
14986         https://bugs.webkit.org/show_bug.cgi?id=66325
14987
14988         Reviewed by Adam Barth.
14989
14990         WinPort didn't have any port_name parsing logic.
14991         Seemed it was best to just share the logic from MacPort.
14992         Sharing the logic from MacPort seemed easiest with a shared superclass.
14993         Once I created ApplePort (the superclass) there was a bunch of logic to share.
14994         I found LeakDetector was just in the way when switching back and forth
14995         between win.py and mac.py so I moved it into its own file.
14996
14997         The only logic change here is that WinPort should now be
14998         able to parse port_name values like MacPort (and other ports)
14999         can.  The fact that ports do this in their constructors is
15000         wrong and a source of many bugs.  But at least now WinPort
15001         matches the status-quo.
15002
15003         * Scripts/webkitpy/layout_tests/port/apple.py: Added.
15004         * Scripts/webkitpy/layout_tests/port/leakdetector.py: Added.
15005         * Scripts/webkitpy/layout_tests/port/leakdetector_unittest.py: Added.
15006         * Scripts/webkitpy/layout_tests/port/mac.py:
15007         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
15008         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
15009         * Scripts/webkitpy/layout_tests/port/win.py:
15010         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
15011
15012 2011-08-16  Adam Barth  <abarth@webkit.org>
15013
15014         Add webkit-patch analyze-baselines, as requested by jamesr.
15015
15016         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
15017         * Scripts/webkitpy/tool/commands/rebaseline.py:
15018
15019 2011-08-16  Adam Barth  <abarth@webkit.org>
15020
15021         Fold bulk-optimize-baselines into optimize-baselines
15022         https://bugs.webkit.org/show_bug.cgi?id=66345
15023
15024         Reviewed by James Robinson.
15025
15026         bulk-optimize-baselines is now a superset of optimize-baselines, so
15027         they can be combined.
15028
15029         * Scripts/webkitpy/tool/commands/rebaseline.py:
15030
15031 2011-08-16  Adam Barth  <abarth@webkit.org>
15032
15033         Move chromium-mac to chromium-cg-mac to prepare for Skia switchover
15034         https://bugs.webkit.org/show_bug.cgi?id=66334
15035
15036         Reviewed by Tony Chang.
15037
15038         This patch moves us closer to switching the Chromium port from
15039         CoreGraphics to Skia.  This patch teaches our tools that the current
15040         Chromium Mac configuration is actually chromium-cg-mac, which makes
15041         room for a future chromium-mac configuration that uses Skia.
15042
15043         Shortly after this patch lands, I'll do a server-side move of the
15044         chromium-mac and chromium-mac-leopard directories to chromium-cg-mac
15045         and chromium-cg-mac-leopard.
15046
15047         * BuildSlaveSupport/build.webkit.org-config/config.json:
15048         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
15049         * Scripts/webkitpy/common/config/build.py:
15050         * Scripts/webkitpy/common/config/build_unittest.py:
15051         * Scripts/webkitpy/layout_tests/port/base.py:
15052         * Scripts/webkitpy/layout_tests/port/builders.py:
15053         * Scripts/webkitpy/layout_tests/port/chromium.py:
15054         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
15055         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
15056         * Scripts/webkitpy/layout_tests/port/dryrun.py:
15057         * Scripts/webkitpy/layout_tests/port/factory.py:
15058         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
15059         * Scripts/webkitpy/layout_tests/port/google_chrome_unittest.py:
15060         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
15061         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
15062         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
15063         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
15064         * TestResultServer/static-dashboards/flakiness_dashboard.html:
15065
15066 2011-08-16  Dimitri Glazkov  <dglazkov@chromium.org>
15067
15068         garden-o-matic frontend needs a friendly relative time description method.
15069         https://bugs.webkit.org/show_bug.cgi?id=66324
15070
15071         Reviewed by Adam Barth.
15072
15073         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Added base.relativizeTime method.
15074         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Added unit tests.
15075         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Wired up Time widget to use it.
15076
15077 2011-08-16  Dimitri Glazkov  <dglazkov@chromium.org>
15078
15079         Summary view should not add multiple entries for the same test failure.
15080         https://bugs.webkit.org/show_bug.cgi?id=66318
15081
15082         Reviewed by Adam Barth.
15083
15084         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js: Added a simplistic updating button.
15085         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Added ways to check for
15086             equality of ui.notifications.FailingTest, and whether ui.notifications.TestFailures already contains a given failure analysis.
15087         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Added tests.
15088
15089 2011-08-16  Adam Roben  <aroben@apple.com>
15090
15091         Teach TestFailures that tests that have failed or passed many times in a row are not flaky
15092
15093         The basic strategy here is that once a test has failed or passed many times in a row we
15094         never again consider it as a possibly flaky test. It's a simple strategy but seems to result
15095         in many fewer false positives than our current behavior.
15096
15097         Fixes <http://webkit.org/b/66327> TestFailures page considers far too many tests to be
15098         flaky, including tests that failed for a while but then were fixed
15099
15100         Reviewed by Dan Bates.
15101
15102         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
15103         (FlakyLayoutTestDetector): Initialize new _buildCount property, which is used to track the
15104         number of non-too-many-failure builds we've seen.
15105         (FlakyLayoutTestDetector.prototype.incorporateTestResults): Don't track tests which haven't
15106         failed in the _maximumFailOrPassCount most recent builds. For other not-yet-considered-flaky
15107         tests, keep track of how many times they have passed or failed in a row. If they pass or
15108         fail more than _maximumFailOrPassCount times, consider them to be non-flaky. (Once a test is
15109         considered flaky it doesn't matter how many times it passes or fails.)
15110
15111         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js:
15112         Added tests for the above.
15113
15114 2011-08-16  Dimitri Glazkov  <dglazkov@chromium.org>
15115
15116         garden-o-matic needs a summary view with actions for each problem.
15117         https://bugs.webkit.org/show_bug.cgi?id=66144
15118
15119         Reviewed by Adam Barth.
15120
15121         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html: Added notifications tests.
15122         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/summary.js: Added.
15123         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications.js: Added.
15124         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/scripts/ui/notifications_unittests.js: Added.
15125         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/styles/summary.css: Added.
15126         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/summary.html: Added.
15127
15128 2011-08-16  Dean Jackson  <dino@apple.com>
15129
15130         Adding Ted "hober" O'Connor as a non-committing contributor.
15131
15132         * Scripts/webkitpy/common/config/committers.py:
15133
15134 2011-08-16  Adam Roben  <aroben@apple.com>
15135
15136         Teach TestFailures how to interpret unfinished test runs
15137
15138         Fixes <http://webkit.org/b/66309> TestFailures thinks all tests passed in
15139         http://build.webkit.org/builders/Lion%20Intel%20Debug%20(Tests)/builds/136
15140
15141         Reviewed by David Kilzer.
15142
15143         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
15144         (Builder.prototype.getNumberOfFailingTests): Look for "isFinished" in the layout-test step
15145         rather than "isStarted" so that we don't count builds for which the test run never finished.
15146         Bumped the cache version to evict old, buggy cached data.
15147
15148         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
15149         Added a test for the above, and made a somewhat synthetic test more realistic by including
15150         more of the actual JSON data from build.webkit.org.
15151
15152         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
15153         (LayoutTestResultsLoader.prototype.start): Bumped the cache version to evict old, buggy
15154         cached data.
15155
15156 2011-08-16  Adam Roben  <aroben@apple.com>
15157
15158         Make Apple's Windows port fall back to Lion results instead of SnowLeopard
15159
15160         Apple's Windows port now uses Lion-era versions of CoreFoundation, ICU, etc., so in theory
15161         should match Lion's behavior.
15162
15163         Fixes <http://webkit.org/b/66302> Apple's Windows port matches Lion for some test results,
15164         but doesn't use Lion results
15165
15166         Reviewed by David Kilzer.
15167
15168         * Scripts/old-run-webkit-tests:
15169         (expectedDirectoryForTest):
15170         * Scripts/webkitpy/layout_tests/port/win.py:
15171         (WinPort.baseline_search_path):
15172         * Scripts/webkitpy/layout_tests/port/win_unittest.py:
15173         (WinPortTest.test_baseline_search_path):
15174         Replaced mac-snowleopard with mac-lion.
15175
15176         * Scripts/webkitpy/common/config/build.py:
15177         (_should_file_trigger_build): Added an entry for the LayoutTests/platform/mac-lion
15178         directory, which all Apple Mac platforms and Apple Win platforms could potentially use
15179         results from. Removed Apple Win from being affected by mac-snowleopard changes, since it
15180         will no longer look there for results.
15181
15182         * Scripts/webkitpy/common/config/build_unittest.py:
15183         (ShouldBuildTest.test_should_build): Tests for the above changes.
15184
15185 2011-08-16  Dean Jackson  <dino@apple.com>
15186
15187         Move me from contributor to reviewer.
15188
15189         * Scripts/webkitpy/common/config/committers.py:
15190
15191 2011-08-15  Dimitri Glazkov  <dglazkov@chromium.org>
15192
15193         garden-o-matic frontend needs model.commitDataForRevisionRange function.
15194         https://bugs.webkit.org/show_bug.cgi?id=66252
15195
15196         Reviewed by Adam Barth.
15197
15198         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js: Added commitDataForRevisionRange method.
15199         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model_unittests.js: Added tests.
15200
15201 2011-08-15  Dimitri Glazkov  <dglazkov@chromium.org>
15202
15203         garden-o-matic frontend needs a generic way to track updates.
15204         https://bugs.webkit.org/show_bug.cgi?id=66245
15205
15206         Reviewed by Adam Barth.
15207
15208         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Added UpdateTracker.
15209         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Added tests.
15210
15211 2011-08-15  Adam Roben  <aroben@apple.com>
15212
15213         Teach TestWebKitAPI/gtest how to print JavaScript failures nicely
15214
15215         Failures now give output of the form: foo should be bar but is baz
15216
15217         Fixes <http://webkit.org/b/66240> It's hard to tell what the actual result of a failed JS
15218         test is in TestWebKitAPI's output
15219
15220         Reviewed by David Levin.
15221
15222         * TestWebKitAPI/JavaScriptTest.cpp:
15223         (TestWebKitAPI::JavaScriptCallbackContext::JavaScriptCallbackContext): Removed now-unused
15224         members.
15225         (TestWebKitAPI::javaScriptCallback): Changed to just store the result string in the context
15226         object, rather than doing any testing of it here.
15227         (TestWebKitAPI::runJSTest): Made this function a gtest predicate-formatter. This allows us
15228         to use a pretty error message when the test fails.
15229
15230         * TestWebKitAPI/JavaScriptTest.h: Changed runJSTest to a predicate-formatter, and added nice
15231         gtest-style macros that wrap it.
15232
15233         * TestWebKitAPI/Test.h: Removed now-unused TEST_ASSERT_RETURN.
15234
15235         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash.cpp:
15236         (TestWebKitAPI::TEST): Changed to use the new macros.
15237
15238         * TestWebKitAPI/Tests/WebKit2/RestoreSessionStateContainingFormData.cpp:
15239         (TestWebKitAPI::createSessionStateContainingFormData): Ditto. Note that this function no
15240         longer returns 0 when the JS test fails. That shouldn't have any effect on whether or not
15241         the test passes, though. Returning early seems to have been an unnecessary optimization.
15242
15243         * TestWebKitAPI/Tests/WebKit2/SpacebarScrolling.cpp:
15244         (TestWebKitAPI::TEST): Changed to use the new macros.
15245
15246 2011-08-15  Eric Seidel  <eric@webkit.org>
15247
15248         Move the Leaks bot back to ORWT until I can fix LeaksViewer regressions filed by Adam Roben in bug 66227 and 66228.
15249
15250         Unreviewed.
15251
15252         * Scripts/run-webkit-tests:
15253         (usingLeaks):
15254         (useNewRunWebKitTests):
15255
15256 2011-08-14  Adam Barth  <abarth@webkit.org>
15257
15258         Switch results detail view over to new-style object-oriented UI widgets
15259         https://bugs.webkit.org/show_bug.cgi?id=66200
15260
15261         Reviewed by Dimitri Glazkov.
15262
15263         This patch replaces my goofy template-based UI for the results
15264         comparison screen with new object-oriented UI widgets.
15265
15266         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
15267         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
15268         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
15269         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
15270         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
15271         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui/results.js: Added.
15272         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
15273
15274 2011-08-13  Dimitri Glazkov  <dglazkov@chromium.org>
15275
15276         garden-o-matic's analyzeUnexpectedFailures needs a completion callback.
15277         https://bugs.webkit.org/show_bug.cgi?id=66166
15278
15279         Also changed base.RequestTracker to:
15280         a) fire callback immediately if requestsInFlight is 0;
15281         b) not barf if callback is not supplied.
15282
15283         Reviewed by Adam Barth.
15284
15285         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js: Started using completion callback.
15286         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js: Added completion callback.
15287         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Changed RequestTracker.
15288         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittest.js: Added unit tests.
15289
15290 2011-08-12  Mark Rowe  <mrowe@apple.com>
15291
15292         Be more forward-looking in the choice of compiler.
15293
15294         Rubber-stamped by Jon Honeycutt.
15295
15296         * DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig:
15297         * MiniBrowser/Configurations/CompilerVersion.xcconfig:
15298         * TestWebKitAPI/Configurations/CompilerVersion.xcconfig:
15299         * WebKitTestRunner/Configurations/CompilerVersion.xcconfig:
15300
15301 2011-08-12  Sadrul Habib Chowdhury  <sadrul@chromium.org>
15302
15303         DRT: Add support for gesture events.
15304         https://bugs.webkit.org/show_bug.cgi?id=66105
15305
15306         Reviewed by Darin Fisher.
15307
15308         * DumpRenderTree/chromium/EventSender.cpp:
15309         (EventSender::EventSender):
15310         (EventSender::gestureScrollBegin):
15311         (EventSender::gestureScrollEnd):
15312         (EventSender::gestureTap):
15313         (EventSender::gestureEvent):
15314         * DumpRenderTree/chromium/EventSender.h:
15315
15316 2011-08-12  Andy Estes  <aestes@apple.com>
15317
15318         allowRoundingHacksCallback uses the wrong argument for 'thisObject'.
15319         https://bugs.webkit.org/show_bug.cgi?id=66158
15320
15321         Reviewed by Dan Bernstein.
15322
15323         * DumpRenderTree/LayoutTestController.cpp:
15324         (allowRoundingHacksCallback): The third argument is the 'this'
15325         JSObjectRef, not the second.
15326
15327 2011-08-12  Andy Estes  <aestes@apple.com>
15328
15329         Cancel in onbeforeunload dialog sometime causes a button to stop working.
15330         https://bugs.webkit.org/show_bug.cgi?id=26211
15331
15332         Reviewed by Alexey Proskuryakov.
15333
15334         Implement a new LayoutTestController method that a test can call to set
15335         the return value of DumpRenderTree's beforeunload UI delegate.
15336
15337         * DumpRenderTree/LayoutTestController.cpp:
15338         (LayoutTestController::LayoutTestController): Initialize
15339         m_shouldStayOnPageAfterHandlingBeforeUnload.
15340         (setShouldStayOnPageAfterHandlingBeforeUnloadCallback): Call
15341         LayoutTestController::setShouldStayOnPageAfterHandlingBeforeUnload().
15342         (LayoutTestController::staticFunctions): Register
15343         'setShouldStayOnPageAfterHandlingBeforeUnload' as a static function.
15344         * DumpRenderTree/LayoutTestController.h:
15345         (LayoutTestController::shouldStayOnPageAfterHandlingBeforeUnload):
15346         (LayoutTestController::setShouldStayOnPageAfterHandlingBeforeUnload):
15347         * DumpRenderTree/chromium/LayoutTestController.cpp:
15348         (LayoutTestController::LayoutTestController): Initialize
15349         m_shouldStayOnPageAfterHandlingBeforeUnload and bind
15350         'setShouldStayOnPageAfterHandlingBeforeUnload' to its c++ setter.
15351         (LayoutTestController::reset): Reset
15352         m_shouldStayOnPageAfterHandlingBeforeUnload to false.
15353         (LayoutTestController::setShouldStayOnPageAfterHandlingBeforeUnload):
15354         Set m_shouldStayOnPageAfterHandlingBeforeUnload to the value passed in
15355         from JavaScript.
15356         * DumpRenderTree/chromium/LayoutTestController.h:
15357         (LayoutTestController::shouldStayOnPageAfterHandlingBeforeUnload):
15358         * DumpRenderTree/chromium/WebViewHost.cpp:
15359         (WebViewHost::runModalBeforeUnloadDialog): Add the correct logging and
15360         return the inverse of
15361         LayoutTestController::shouldStayOnPageAfterHandlingBeforeUnload().
15362         * DumpRenderTree/mac/UIDelegate.mm:
15363         (-[UIDelegate webView:runBeforeUnloadConfirmPanelWithMessage:initiatedByFrame:]): Ditto.
15364         * DumpRenderTree/win/UIDelegate.cpp:
15365         (UIDelegate::runBeforeUnloadConfirmPanelWithMessage): Ditto.
15366         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
15367         Define setShouldStayOnPageAfterHandlingBeforeUnload().
15368         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
15369         (WTR::InjectedBundle::postNewBeforeUnloadReturnValue): Post a message
15370         to the Test Controller telling it what value it should return in its
15371         onbeforeunload UI delegate.
15372         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
15373         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
15374         (WTR::LayoutTestController::setShouldStayOnPageAfterHandlingBeforeUnload):
15375         Call InjectedBundle::postNewBeforeUnloadReturnValue().
15376         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
15377         * WebKitTestRunner/TestController.cpp:
15378         (WTR::TestController::TestController): Initialize
15379         m_beforeUnloadReturnValue to true.
15380         (WTR::runBeforeUnloadConfirmPanel): Add the correct logging and return
15381         the value of TestController::beforeUnloadReturnValue().
15382         (WTR::TestController::resetStateToConsistentValues): Reset
15383         m_beforeUnloadReturnValue to true.
15384         * WebKitTestRunner/TestController.h:
15385         (WTR::TestController::beforeUnloadReturnValue):
15386         (WTR::TestController::setBeforeUnloadReturnValue):
15387         * WebKitTestRunner/TestInvocation.cpp:
15388         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle): Handle the
15389         message posted from the injected bundle by retrieving the message body
15390         as a WKBoolean and calling TestController::setBeforeUnloadReturnValue().
15391
15392 2011-08-12  Sam Weinig  <sam@webkit.org>
15393
15394         Move compiler specific macros to their own header
15395         https://bugs.webkit.org/show_bug.cgi?id=66119
15396
15397         Reviewed by Anders Carlsson.
15398
15399         * DumpRenderTree/ForwardingHeaders/wtf/Compiler.h: Added.
15400
15401 2011-08-11  Dimitri Glazkov  <dglazkov@chromium.org>
15402
15403         Add a way to extend DOM objects in garden-o-matic.
15404         https://bugs.webkit.org/show_bug.cgi?id=66096
15405
15406         Reviewed by Adam Barth.
15407
15408         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Added base.extends.
15409         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Added tests.
15410
15411 2011-08-11  MORITA Hajime  <morrita@google.com>
15412
15413         [Chromium][DRT] editing/spelling/spelling-backspace-between-lines.html fails
15414         https://bugs.webkit.org/show_bug.cgi?id=45438
15415
15416         Reviewed by Kent Tamura.
15417
15418         MockSpellCheck was assuming the input of spellCheckWord() is a word.
15419         But it can be a series of word. This change made spellCheckWord()
15420         understand simple word boundary which appears in tests.
15421
15422         * DumpRenderTree/chromium/MockSpellCheck.cpp:
15423         (MockSpellCheck::spellCheckWord):
15424
15425 2011-08-11  Mark Rowe  <mrowe@apple.com>
15426
15427         Rework some Makefile logic to remove a double-negative.
15428
15429         Reviewed by Jon Honeycutt.
15430
15431         * DumpRenderTree/mac/PerlSupport/Makefile:
15432
15433 2011-08-11  Anders Carlsson  <andersca@apple.com>
15434
15435         Add a FIXME.
15436
15437         * TestWebKitAPI/TestsController.cpp:
15438         (TestWebKitAPI::TestsController::TestsController):
15439
15440 2011-08-11  Anders Carlsson  <andersca@apple.com>
15441
15442         Initialize threading before running any tests
15443         https://bugs.webkit.org/show_bug.cgi?id=66112
15444
15445         Reviewed by Adam Barth.
15446
15447         * TestWebKitAPI/TestsController.cpp:
15448         (TestWebKitAPI::TestsController::TestsController):
15449
15450 2011-08-11  Adam Barth  <abarth@webkit.org>
15451
15452         Add unexpected-passes.html to TestFailures for marking tests as passing
15453         https://bugs.webkit.org/show_bug.cgi?id=66102
15454
15455         Reviewed by Dimitri Glazkov.
15456
15457         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
15458         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
15459             - Move some code into the library so it can be shared.
15460         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
15461         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
15462         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
15463             - Add round-trip unit tests.
15464         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.html: Added.
15465         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/unexpected-passes.js: Added.
15466             - I'm not entirely sure this feature should remain a separate HTML
15467               file, but it seemed better than cluttering up the main HTML file
15468               with too much extra stuff.
15469         * Scripts/webkitpy/tool/servers/gardeningserver.py:
15470         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
15471             - The gardening server couldn't handle adding PASS expectations.  Now it can.
15472
15473 2011-08-11  Eric Seidel  <eric@webkit.org>
15474
15475         NRWT has wrong fallback order for Mac now that Lion exists
15476         https://bugs.webkit.org/show_bug.cgi?id=66093
15477
15478         Reviewed by Adam Barth.
15479
15480         ORWT used a different system for fallback orders than NRWT did.
15481         I moved win.py to match ORWT fallback orders in bug 64486.
15482         This bug moves mac.py to match ORWT fallback order
15483         and adds 'lion' as a supported OS version.
15484
15485         Because ChromiumMac shares OS version detection code with AppleMac
15486         this also added support for 'lion' to chromium mac.  Hopefully that's a good thing.
15487
15488         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
15489         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
15490         * Scripts/webkitpy/layout_tests/port/mac.py:
15491         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
15492         * Scripts/webkitpy/layout_tests/port/win.py:
15493
15494 2011-08-11  Adam Barth  <abarth@webkit.org>
15495
15496         Update test results now that we use absolute URLs.
15497
15498         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js:
15499
15500 2011-08-11  Adam Barth  <abarth@webkit.org>
15501
15502         garden-o-matic should be able to determine when tests unexpectedly pass
15503         https://bugs.webkit.org/show_bug.cgi?id=66092
15504
15505         Reviewed by Dimitri Glazkov.
15506
15507         This patch just contains the library methods for extracting this
15508         information.  I'll add some UI in a subsequent patch.
15509
15510         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
15511         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js:
15512
15513 2011-08-11  Dimitri Glazkov  <dglazkov@chromium.org>
15514
15515         New expectation entries should be placed next to existing ones.
15516         https://bugs.webkit.org/show_bug.cgi?id=66066
15517
15518         Reviewed by Adam Barth.
15519
15520         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added the notion of insertion point,
15521             which is computed as the last found reference to the same test.
15522         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added and adjusted unit tests.
15523
15524 2011-08-11  Dimitri Glazkov  <dglazkov@chromium.org>
15525
15526         Unreviewed, rolling out r92847.
15527         http://trac.webkit.org/changeset/92847
15528         https://bugs.webkit.org/show_bug.cgi?id=65986
15529
15530         Broke check-webkit-style
15531
15532         * Scripts/webkitpy/style/checker.py:
15533         * Scripts/webkitpy/style/checker_unittest.py:
15534
15535 2011-08-11  Dimitri Glazkov  <dglazkov@chromium.org>
15536
15537         Unreviewed, rolling out r92847.
15538         http://trac.webkit.org/changeset/92847
15539         https://bugs.webkit.org/show_bug.cgi?id=65986
15540
15541         Broke check-webkit-style
15542
15543         * Scripts/webkitpy/style/checker.py:
15544         * Scripts/webkitpy/style/checker_unittest.py:
15545
15546 2011-08-11  Pierre Rossi  <pierre.rossi@gmail.com>
15547
15548         [Qt] autotests shouldn't require config.h in the style check
15549         https://bugs.webkit.org/show_bug.cgi?id=65986
15550
15551         Remove the include rules from the style-check for Qt tests.
15552         Since for the most part it's just a matter of using the
15553         API provided by QtWebKit, enforcing the WebCore set of rules
15554         only seems to make sense in the case of MIMESniffing.
15555
15556         Reviewed by Benjamin Poulain.
15557
15558         * Scripts/webkitpy/style/checker.py:
15559         * Scripts/webkitpy/style/checker_unittest.py:
15560
15561 2011-08-11  Zsolt Fehér  <h490267@stud.u-szeged.hu>
15562
15563         [WK2] Implement waitForPolicyDelegate
15564         https://bugs.webkit.org/show_bug.cgi?id=42330
15565
15566         Reviewed by Csaba Osztrogonác.
15567
15568         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
15569         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
15570         (WTR::navigationTypeToString):
15571         (WTR::InjectedBundlePage::decidePolicyForNavigationAction):
15572         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
15573         (WTR::LayoutTestController::LayoutTestController):
15574         (WTR::LayoutTestController::setCustomPolicyDelegate):
15575         (WTR::LayoutTestController::waitForPolicyDelegate):
15576         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
15577         (WTR::LayoutTestController::isPolicyDelegateEnabled):
15578         (WTR::LayoutTestController::isPolicyDelegatePermissive):
15579
15580 2011-08-11  Adam Barth  <abarth@webkit.org>
15581
15582         rebaseline-expectations should use the release bots, not the debug
15583         bots.  The debug bots don't all run all the tests.  Also, delay
15584         updating test_expectations a long as possible to avoid clearing out
15585         lines that we'll need for other ports.
15586
15587         * Scripts/webkitpy/layout_tests/port/builders.py:
15588         * Scripts/webkitpy/tool/commands/rebaseline.py:
15589         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
15590
15591 2011-08-10  Ryosuke Niwa  <rniwa@webkit.org>
15592
15593         r92792 inadvertently removed PlatformSpecificScheduler for all Snow Leopard bots.
15594         Revert it partially so that the remaining Snow Leopard bots can fulfill their duty.
15595
15596         * BuildSlaveSupport/build.webkit.org-config/config.json:
15597
15598 2011-08-10  Eric Seidel  <eric@webkit.org>
15599
15600         new-run-webkit-test's WinPort has no fallback logic
15601         https://bugs.webkit.org/show_bug.cgi?id=64486
15602
15603         Unreviewed.
15604
15605         Turns out that we've long had a subtle bug whereby if you didn't
15606         already have DumpRenderTree built, we might end up running unsupported tests
15607         because the DumpRenderTree --print-supported-features check would silently fail.
15608
15609         When I moved the call from using popen to using Executive.run_command, this silent
15610         failure became noisy.  (Breaking some bots.)
15611
15612         This re-orders our steps so that we always build DRT before collecting expectations,
15613         as well as adds a warning message when DumpRenderTree can't be found.
15614
15615         * Scripts/webkitpy/layout_tests/port/webkit.py:
15616         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
15617
15618 2011-08-10  Eric Seidel  <eric@webkit.org>
15619
15620         new-run-webkit-test's WinPort has no fallback logic
15621         https://bugs.webkit.org/show_bug.cgi?id=64486
15622
15623         Unreviewed.  Fix test-webkitpy under Python 2.5.
15624
15625         * Scripts/webkitpy/layout_tests/port/win.py:
15626
15627 2011-08-10  Benjamin Poulain  <benjamin@webkit.org>
15628
15629         [Qt][WK2] MiniBrowser is firing twice the QDesktopWebView::mousePressEvent method
15630         https://bugs.webkit.org/show_bug.cgi?id=65875
15631
15632         Reviewed by Kenneth Rohde Christiansen.
15633
15634         For some reason, QWidget send fake mouse event on some platform, under certains conditions.
15635
15636         This causes mouse events to be sent twice in some cases. We work around the problem
15637         by defining the first touch point to be a primary touch point (case which hit some
15638         weird condition in QWidget, skipping the generation of fake mouse events).
15639
15640         * MiniBrowser/qt/MiniBrowserApplication.cpp:
15641         (MiniBrowserApplication::notify):
15642
15643 2011-08-10  David Hyatt  <hyatt@apple.com>
15644
15645         https://bugs.webkit.org/show_bug.cgi?id=66004
15646
15647         Remove the regions and exclusions bots and also the flags in build-webkit, since I'm just turning
15648         the code on by default.
15649
15650         Reviewed by Adam Roben.
15651
15652         * BuildSlaveSupport/build.webkit.org-config/config.json:
15653         * Scripts/build-webkit:
15654
15655 2011-08-10  Oliver Hunt  <oliver@apple.com>
15656
15657         JSEvaluteScript does not return the correct object when given JSONP data
15658         https://bugs.webkit.org/show_bug.cgi?id=66003
15659
15660         Reviewed by Gavin Barraclough.
15661
15662         Add a test case.
15663
15664         * DumpRenderTree/mac/DumpRenderTree.mm:
15665         (testStringByEvaluatingJavaScriptFromString):
15666
15667 2011-08-10  Eric Seidel  <eric@webkit.org>
15668
15669         WIN: NRWT runs compositing tests on configurations that don't support compositing
15670         https://bugs.webkit.org/show_bug.cgi?id=64472
15671
15672         Reviewed by Adam Barth.
15673
15674         I think the code was just wrong.  It was never splitting the string
15675         into list pieces before.
15676         I've now tested the supported_features code and theoretically it
15677         should now work with Windows DRT.
15678
15679         I also made the list-lookup functions always return lists, to make
15680         it possible to clean up the list-transform code in the future
15681         (I opted not to do that cleanup in this patch, but at least now
15682         both types of feature lookup functions match return types.)
15683
15684         * Scripts/webkitpy/layout_tests/port/webkit.py:
15685         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
15686
15687 2011-08-10  Eric Seidel  <eric@webkit.org>
15688
15689         Clean up ChromiumDriver a little
15690         https://bugs.webkit.org/show_bug.cgi?id=65995
15691
15692         Unreviewed.
15693
15694         Fix NRWT for chromium.  Silly python.
15695
15696         * Scripts/webkitpy/layout_tests/port/chromium.py:
15697
15698 2011-08-10  Eric Seidel  <eric@webkit.org>
15699
15700         Clean up ChromiumDriver a little
15701         https://bugs.webkit.org/show_bug.cgi?id=65995
15702
15703         Reviewed by Adam Barth.
15704
15705         - We no longer support threading, so remove threading specific code.
15706         - Add FIXMEs about using ServerProcess instead.
15707         - Condense option-mapping if-cascade into a for loop.
15708         - Unindent long if blocks by using early return.
15709         - Unwrap lines which are needlessly wrapped.
15710
15711         There should be no functional changes here, just code cleanup/dead-code removal.
15712
15713         * Scripts/webkitpy/layout_tests/port/chromium.py:
15714
15715 2011-08-10  Dimitri Glazkov  <dglazkov@chromium.org>
15716
15717         Unmuddle construction options for TestConfiguration.
15718         https://bugs.webkit.org/show_bug.cgi?id=65988
15719
15720         The TestConfiguration instance is now constructed only using actual values.
15721         Port-based construction is split into a "from_port" class method.
15722
15723         Reviewed by Eric Seidel.
15724
15725         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added TestConfiguration.from_port
15726         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Adjusted code to accommodate the change.
15727         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Ditto.
15728         * Scripts/webkitpy/layout_tests/port/base.py: Ditto.
15729         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py: Ditto.
15730
15731 2011-08-10  Eric Seidel  <eric@webkit.org>
15732
15733         Leaks bot shows python logging prefixes as part of summary display
15734         https://bugs.webkit.org/show_bug.cgi?id=65931
15735
15736         Reviewed by Adam Barth.
15737
15738         The big part of this change is adding a unittesting system for
15739         master.cfg, so mere-mortals can make changes to build.webkit.org
15740         with greatly-reduced risk of breaking the world.
15741
15742         Turns out that in typing up my trival logging fix I made 2 typos,
15743         so hopefully the hour spent getting master.cfg to load was worth it. :)
15744
15745         I also tried to make mastercfg_unittest run as part of test-webkitpy
15746         but right now test-webkitpy requires that unittest files be located in
15747         a module-name-compatible directory.  'build.webkit.org-config' is not
15748         a valid module name.
15749
15750         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
15751          - Fixed _parseNewRunWebKitTestsOutput to strip NRWT's python formatter prefix
15752            by using a new _strip_python_logging_prefix function.
15753          - Also fixed this function to file to work with a more modern
15754            simplejson version (which returns unicode objects instead of str objects)
15755            and filed a related bug with buildbot.net due to their handling of unicode builder names.
15756         * BuildSlaveSupport/build.webkit.org-config/mastercfg_unittest.py: Added.
15757          - BuildBotConfigLoader is a huge pile of code just to get master.cfg to load.
15758          - The actual tiny unittest of the new _parseNewRunWebKitTestsOutput logic.
15759         * Scripts/webkitpy/thirdparty/__init__.py:
15760          - Add autoinstallation of buildbot (and jinja2) for use by mastercfg_unittest.py
15761
15762 2011-08-09  Dimitri Glazkov  <dglazkov@chromium.org>
15763
15764         garden-o-matic should extrapolate failures given a set of known results
15765         https://bugs.webkit.org/show_bug.cgi?id=65882
15766
15767         Reviewed by Adam Barth.
15768
15769         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Spuriously covered my old sins.
15770         * Scripts/webkitpy/layout_tests/port/builders.py: Augmented builders dictionary to contain build coverage specifiers,
15771             adjusted all call sites.
15772         * Scripts/webkitpy/tool/servers/gardeningserver.py: Added BuildCoverageExtrapolator, a thing that
15773             makes wider net for specific failures.
15774         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added unit tests.
15775
15776 2011-08-10  Adam Barth  <abarth@webkit.org>
15777
15778         The URL we were computing for retrieving baselines was missing a slash
15779         character because our mocks had an extra slash character.  I've
15780         corrected both the code and the mock.
15781
15782         * Scripts/webkitpy/tool/commands/rebaseline.py:
15783         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
15784         * Scripts/webkitpy/tool/mocktool.py:
15785
15786 2011-08-09  Adam Barth  <abarth@webkit.org>
15787
15788         Move the quote mark to the right place.
15789
15790         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
15791
15792 2011-08-09  Adam Barth  <abarth@webkit.org>
15793
15794         garden-o-matic should use huerstics to narrow regression range
15795         https://bugs.webkit.org/show_bug.cgi?id=65948
15796
15797         Reviewed by Dimitri Glazkov.
15798
15799         When analyzing a regression range, if one of the commit messages
15800         mentions the test in question, we now hueristically assume that
15801         revision caused the failure.  (I'm shamelessly stealing this idea from
15802         TestFailures.)
15803
15804         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
15805
15806 2011-07-13  Eric Seidel  <eric@webkit.org>
15807
15808         new-run-webkit-test's WinPort has no fallback logic
15809         https://bugs.webkit.org/show_bug.cgi?id=64486
15810
15811         Reviewed by Adam Roben.
15812
15813         I've tried to write a patch for bug 64439 twice now, and both times
15814         I've ended up re-writing half the port system.  So I'm breaking
15815         things up into smaller pieces, this being the first.
15816
15817         WinPort still does not have any port_name parsing, so when instantiated
15818         with the name "win-xp" (i.e. by the rebaseline server) it will just behave as the 'win' port.
15819         I'll fix this in a second pass when I standardize port_name parsing for all webkit ports.
15820
15821         Otherwise this should "just work" for windows.  I've not been able to test the
15822         version detection on my mac, but the unit tests show the code behaving as designed.
15823
15824         * Scripts/webkitpy/layout_tests/port/win.py:
15825         * Scripts/webkitpy/layout_tests/port/win_unittest.py: Added.
15826
15827 2011-08-09  Adam Barth  <abarth@webkit.org>
15828
15829         Teach build.webkit.org's garden-o-matic how to talk to the local server
15830         https://bugs.webkit.org/show_bug.cgi?id=65940
15831
15832         Reviewed by Dimitri Glazkov.
15833
15834         After this patch, the version of garden-o-matic on build.webkit.org is
15835         fully functional.  It can interact with the local server via CORS.
15836
15837         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
15838         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
15839         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
15840         * Scripts/webkitpy/tool/commands/gardenomatic.py:
15841         * Scripts/webkitpy/tool/servers/gardeningserver.py:
15842
15843 2011-08-09  Adam Barth  <abarth@webkit.org>
15844
15845         Enable CORS for garden-o-matic
15846         https://bugs.webkit.org/show_bug.cgi?id=65936
15847
15848         Reviewed by Dimitri Glazkov.
15849
15850         This patch will allow the hosted instance of garden-o-matic to
15851         communicate with the local server.
15852
15853         * Scripts/webkitpy/tool/servers/gardeningserver.py:
15854         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
15855
15856 2011-08-09  Adam Barth  <abarth@webkit.org>
15857
15858         Upgrade jquery in TestFailures
15859         https://bugs.webkit.org/show_bug.cgi?id=65935
15860
15861         Reviewed by Dimitri Glazkov.
15862
15863         Upgrade to a newer version of jQuery with better support for CORS and CSP.
15864
15865         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
15866         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
15867         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
15868
15869 2011-08-09  Sheriff Bot  <webkit.review.bot@gmail.com>
15870
15871         Unreviewed, rolling out r92683.
15872         http://trac.webkit.org/changeset/92683
15873         https://bugs.webkit.org/show_bug.cgi?id=65921
15874
15875          "Breaks Qt --minimal build. Revert pending update of the Qt
15876         minimal configuration" (Requested by simathur on #webkit).
15877
15878         * QtTestBrowser/cookiejar.cpp:
15879         (TestBrowserCookieJar::TestBrowserCookieJar):
15880         * QtTestBrowser/launcherwindow.cpp:
15881         (LauncherWindow::createChrome):
15882         (LauncherWindow::setDiskCache):
15883         * QtTestBrowser/main.cpp:
15884         (LauncherApplication::handleUserOptions):
15885         * QtTestBrowser/webpage.cpp:
15886         (WebPage::openUrlInDefaultBrowser):
15887
15888 2011-08-09  Siddharth Mathur  <siddharth.mathur@nokia.com>
15889
15890         [Qt] Simplify code by removing QT_NO_DESKTOPSERVICES and QT_NO_NETWORKDISKCACHE
15891         https://bugs.webkit.org/show_bug.cgi?id=65880
15892
15893         Reviewed by Andreas Kling.
15894
15895         QDesktopServices and QNetworkDiskCache have been around for a while now. Kill flags which
15896         check for their feature flags. Assume that any reasonable Qt build provides them.
15897         * QtTestBrowser/cookiejar.cpp:
15898         (TestBrowserCookieJar::TestBrowserCookieJar):
15899         * QtTestBrowser/launcherwindow.cpp:
15900         (LauncherWindow::createChrome):
15901         (LauncherWindow::setDiskCache):
15902         * QtTestBrowser/main.cpp:
15903         (LauncherApplication::handleUserOptions):
15904         * QtTestBrowser/webpage.cpp:
15905         (WebPage::openUrlInDefaultBrowser):
15906
15907 2011-08-09  Adam Barth  <abarth@webkit.org>
15908
15909         BaselineOptimizer created the wrong baseline for fast/js/regexp-overflow.html
15910         https://bugs.webkit.org/show_bug.cgi?id=65891
15911
15912         Reviewed by Eric Seidel.
15913
15914         The problem was that platform/chromium contained a bogus expectation
15915         file that needed to be removed, but by the time we got around to
15916         removing it, we'd already moved the correct baseline into its place.
15917         After this patch, we copy the gold results into memory before
15918         reshuffling things on disk (and we delete bad things before adding good
15919         things).
15920
15921         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
15922         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
15923
15924 2011-08-08  Adam Barth  <abarth@webkit.org>
15925
15926         Chromium Windows bots can't figure out what SVN revision they're running
15927         https://bugs.webkit.org/show_bug.cgi?id=65893
15928
15929         Reviewed by Eric Seidel.
15930
15931         The comment in _engage_awesome_windows_hacks explains why we're making
15932         this change.  It's ugly and rediculous, but this approach seems better
15933         than using shell=True when calling popen.
15934
15935         * Scripts/webkitpy/common/checkout/scm/svn.py:
15936         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
15937
15938 2011-08-08  Adam Barth  <abarth@webkit.org>
15939
15940         Remove deduplicate-tests
15941         https://bugs.webkit.org/show_bug.cgi?id=65886
15942
15943         Reviewed by Tony Chang.
15944
15945         This script is redundant with "webkit-patch bulk-optimize-baselines",
15946         which can find many more duplicates than deduplicate-tests because
15947         bulk-optimize-baselines can move results around whereas
15948         deduplicate-tests can only remove results.
15949
15950         * Scripts/deduplicate-tests: Removed.
15951         * Scripts/webkitpy/to_be_moved/deduplicate_tests.py: Removed.
15952         * Scripts/webkitpy/to_be_moved/deduplicate_tests_unittest.py: Removed.
15953
15954 2011-08-08  Adam Barth  <abarth@webkit.org>
15955
15956         scm.delete should delete empty parent directories as well
15957         https://bugs.webkit.org/show_bug.cgi?id=65878
15958
15959         Reviewed by Eric Seidel.
15960
15961         This behavior makes SVN match GIT (which has no concept of empty
15962         directories).  This bug comes up a lot when optimizing baselines, which
15963         often create empty directories when we're able to optimize everything
15964         out of a given folder.
15965
15966         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
15967         * Scripts/webkitpy/common/checkout/scm/svn.py:
15968
15969 2011-08-08  Lucas Forschler  <lforschler@apple.com>
15970
15971         Add triggers for Lion WK2 test bots.
15972
15973         Reviewed by Stephanie Lewis.
15974
15975         * BuildSlaveSupport/build.webkit.org-config/config.json:
15976
15977 2011-08-08  Adam Barth  <abarth@webkit.org>
15978
15979         bulk-optimize-baselines should take directories on the command line
15980         https://bugs.webkit.org/show_bug.cgi?id=65879
15981
15982         Reviewed by Dimitri Glazkov.
15983
15984         Passing lists of tests via stdin is a pain.  It's easier if the command
15985         takes a list of tests on the command line, like run-webkit-tests.
15986
15987         * Scripts/webkitpy/tool/commands/rebaseline.py:
15988
15989 2011-08-08  Adam Barth  <abarth@webkit.org>
15990
15991         Deprecate rebaseline-chromium-webkit-tests
15992         https://bugs.webkit.org/show_bug.cgi?id=65873
15993
15994         Reviewed by Dimitri Glazkov.
15995
15996         This script never worked correctly and now is redundant with more
15997         modern infrastructure for rebaselining tests.  This patch deprecates
15998         the script in order to flush out any issues with the new tools.  Once
15999         folks are happy with the new tools, we'll delete this script.
16000
16001         * Scripts/rebaseline-chromium-webkit-tests:
16002
16003 2011-08-08  Dimitri Glazkov  <dglazkov@chromium.org>
16004
16005         garden-o-matic should only touch expectations that need updating.
16006         https://bugs.webkit.org/show_bug.cgi?id=65876
16007
16008         Reviewed by Adam Barth.
16009
16010         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added reconstitute_only_these to TestExpectationSerializer.list_to_string,
16011             so that a caller could influence which lines to serialize from values.
16012         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests.
16013         * Scripts/webkitpy/tool/servers/gardeningserver.py: Started using reconstitute_only_these.
16014         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added tests.
16015
16016 2011-08-08  Dimitri Glazkov  <dglazkov@chromium.org>
16017
16018         TestExpectationsEditor should return a list of updated expectations.
16019         https://bugs.webkit.org/show_bug.cgi?id=65872
16020
16021         Reviewed by Adam Barth.
16022
16023         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added a list to store lines that are being updated.
16024         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed tests to account for updated lines.
16025
16026 2011-08-08  Jochen Eisinger  <jochen@chromium.org>
16027
16028         Fix SCM webkitpy unit test failures
16029         https://bugs.webkit.org/show_bug.cgi?id=65823
16030
16031         Reviewed by Eric Seidel.
16032
16033         * Scripts/webkitpy/common/checkout/scm/git.py:
16034         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
16035
16036 2011-08-08  Eric Seidel  <eric@webkit.org>
16037
16038         Move the --leaks bot back over to NRWT now that I'm back
16039         from vacation and able to debug it.
16040
16041         Unreviewed.
16042
16043         * Scripts/run-webkit-tests:
16044         (usingWebKit2):
16045         (useNewRunWebKitTests):
16046
16047 2011-08-05  Dimitri Glazkov  <dglazkov@chromium.org>
16048
16049         Wire up updating expectations in garden-o-matic.
16050         https://bugs.webkit.org/show_bug.cgi?id=65794
16051
16052         Reviewed by Adam Barth.
16053
16054         * Scripts/webkitpy/layout_tests/port/factory.py: Moved _port_for_builder from rebaseline.py to here, rename to get_from_builder_name.
16055         * Scripts/webkitpy/tool/commands/rebaseline.py: Adapted callsites to new location.
16056         * Scripts/webkitpy/tool/servers/gardeningserver.py: Added GardeningExpectationsUpdater, which knows how to update expectations.
16057         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added loads of tests.
16058
16059 2011-08-08  Jochen Eisinger  <jochen@chromium.org>
16060
16061         webkit-patch doesn't get along with renamed files
16062         https://bugs.webkit.org/show_bug.cgi?id=48075
16063
16064         Possibly a bit heavy handed - I removed all instances of -C and
16065         changed every instance of -M with '--no-renames' in git.py. This
16066         forces git to not try to tell us about renames at all, which is
16067         ultimately the behaviour we want. The old file is shown deleted,
16068         then the new file is shown added, followed by any changes that
16069         occurred. Also gets rid of the problem where deleting one file
16070         and adding another file which has similar content would
16071         unexpectedly show up as a rename, and fall out of a diff.
16072
16073         Based on a patch by Wyatt Carss.
16074
16075         Reviewed by Eric Seidel.
16076
16077         * Scripts/webkitpy/common/checkout/scm/git.py:
16078
16079 2011-08-07  Sam White  <samuel.white@rochester.edu>
16080
16081         Add the ability to search the AccessibilityObject cache
16082         https://bugs.webkit.org/show_bug.cgi?id=64994
16083
16084         Added testing support for AccessibilityObject cache searching. Currently,
16085         only the mac platform is full supported and has had API exposed. Other
16086         platforms have only have this new functionality stubbed. Full implementation
16087         is a job suited only for an accessibility expert of each respective platform.
16088
16089         Reviewed by Chris Fleizach.
16090
16091         * DumpRenderTree/AccessibilityUIElement.cpp:
16092         (uiElementForSearchPredicateCallback):
16093         (AccessibilityUIElement::getJSClass):
16094         * DumpRenderTree/AccessibilityUIElement.h:
16095         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
16096         (AccessibilityUIElement::uiElementForSearchPredicate):
16097         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
16098         (AccessibilityUIElement::uiElementForSearchPredicate):
16099         * DumpRenderTree/win/AccessibilityUIElementWin.cpp:
16100         (AccessibilityUIElement::uiElementForSearchPredicate):
16101
16102 2011-08-06  Adam Barth  <abarth@webkit.org>
16103
16104         Remove dead code in webkitpy
16105         https://bugs.webkit.org/show_bug.cgi?id=65818
16106
16107         Reviewed by Dimitri Glazkov.
16108
16109         This code was added to support the rebaseline2 command six months ago,
16110         but the rebaseline2 command was never finished and no one else has used
16111         this code since it was landed.  This patch removes the dead code.  If
16112         we revive rebaseline2, then we can revert this patch and recover the
16113         code.  Until then, this code is just dead weight.
16114
16115         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
16116         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
16117         * Scripts/webkitpy/common/net/testoutput.py: Removed.
16118         * Scripts/webkitpy/common/net/testoutput_unittest.py: Removed.
16119         * Scripts/webkitpy/common/net/testoutputset.py: Removed.
16120         * Scripts/webkitpy/common/net/testoutputset_unittest.py: Removed.
16121         * Scripts/webkitpy/common/system/directoryfileset.py: Removed.
16122         * Scripts/webkitpy/common/system/directoryfileset_unittest.py: Removed.
16123         * Scripts/webkitpy/layout_tests/port/base.py:
16124         * Scripts/webkitpy/layout_tests/port/chromium.py:
16125         * Scripts/webkitpy/layout_tests/port/webkit.py:
16126
16127 2011-08-06  Adam Barth  <abarth@webkit.org>
16128
16129         webkit-patch optimize-baselines can't handle promotions to base results directory
16130         https://bugs.webkit.org/show_bug.cgi?id=65819
16131
16132         Reviewed by Dimitri Glazkov.
16133
16134         This bug came up when testing the optimizer on some editing baselines.
16135
16136         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
16137         * Scripts/webkitpy/common/checkout/baselineoptimizer_unittest.py:
16138
16139 2011-08-06  Adam Barth  <abarth@webkit.org>
16140
16141         Add a small command to optimize baselines in bulk.
16142
16143         * Scripts/webkitpy/tool/commands/rebaseline.py:
16144
16145 2011-08-05  Dimitri Glazkov  <dglazkov@chromium.org>
16146
16147         Fix unit test breakage by plumbing convert_404_to_None in MockWeb.
16148
16149         * Scripts/webkitpy/tool/mocktool.py: Added missing parameter to MockWeb.get_binary
16150
16151 2011-08-05  Dimitri Glazkov  <dglazkov@chromium.org>
16152
16153         Garden-o-matic updateexpectations needs a unit test.
16154         https://bugs.webkit.org/show_bug.cgi?id=65780
16155
16156         Reviewed by Adam Barth.
16157
16158         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added unit test.
16159
16160 2011-08-05  Anders Carlsson  <andersca@apple.com>
16161
16162         Future proof an Objective-C test
16163         https://bugs.webkit.org/show_bug.cgi?id=65788
16164
16165         Reviewed by Dan Bernstein.
16166
16167         Add isObjectInstanceOf to ObjCController.
16168
16169         * DumpRenderTree/mac/ObjCController.m:
16170         (+[ObjCController isSelectorExcludedFromWebScript:]):
16171         (+[ObjCController webScriptNameForSelector:]):
16172         (-[ObjCController isObject:instanceOf:]):
16173
16174 2011-08-05  Adam Barth  <abarth@webkit.org>
16175
16176         trac.js needs unittests
16177         https://bugs.webkit.org/show_bug.cgi?id=65673
16178
16179         Reviewed by Dimitri Glazkov.
16180
16181         This tests also have somewhat large fixtures.  I've trimmed the example
16182         XML down a bunch while still hitting interesting cases in the parsing.
16183
16184         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
16185         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/trac_unittests.js: Added.
16186
16187 2011-08-05  Adam Barth  <abarth@webkit.org>
16188
16189         Rebuild rebaseline-chromium-webkit-tests on top of modern infrastructure
16190         https://bugs.webkit.org/show_bug.cgi?id=65759
16191
16192         Reviewed by Dimitri Glazkov.
16193
16194         rebaseline-chromium-webkit-tests doesn't really fit into webkitpy's
16195         architecture, is poorly tested, and has a bunch of quirks.  This patch
16196         rebuilds the core functionality of rebaseline-chromium-webkit-tests on
16197         top of more modern infrastructure.  In the process, we get more code
16198         re-use and better testing.
16199
16200         Once we're satisfied with this new implementation, we can delete the
16201         old implementation.
16202
16203         * Scripts/webkitpy/common/checkout/baselineoptimizer.py:
16204         * Scripts/webkitpy/common/net/web.py:
16205         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
16206         * Scripts/webkitpy/layout_tests/port/builders.py:
16207         * Scripts/webkitpy/layout_tests/port/chromium.py:
16208         * Scripts/webkitpy/layout_tests/port/factory.py:
16209         * Scripts/webkitpy/layout_tests/port/webkit.py:
16210         * Scripts/webkitpy/tool/commands/rebaseline.py:
16211         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
16212
16213 2011-08-05  Anders Carlsson  <andersca@apple.com>
16214
16215         Remove PluginHalter
16216         https://bugs.webkit.org/show_bug.cgi?id=65729
16217
16218         Reviewed by Darin Adler.
16219
16220         Remove call to set the allowed plug-in run time.
16221
16222         * DumpRenderTree/mac/DumpRenderTree.mm:
16223         (resetDefaultsToConsistentValues):
16224
16225 2011-08-05  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
16226
16227         Unreviewed build fix for Qt-Minimal buildbot
16228
16229         Use ~/.QtTestBrowser for the cookiejar when QDesktopServices
16230         is disabled.
16231
16232         Rubber-stamped by Holger Freyther.
16233
16234         * QtTestBrowser/cookiejar.cpp:
16235         (TestBrowserCookieJar::TestBrowserCookieJar):
16236
16237 2011-08-05  Ademar de Souza Reis Jr  <ademar.reis@openbossa.org>
16238
16239         [Qt] QtTestBrowser: Add support for saving cookies on disk
16240         https://bugs.webkit.org/show_bug.cgi?id=65328
16241
16242         Simple cookiejar that saves cookies on disk.
16243
16244         Added menu options to clear cookies (Edit->Clear cookies) and
16245         disable disk cookies (Settings->[ ]Save Cookies on Disk).
16246
16247         Option -no-disk-cookies is also available.
16248
16249         Reviewed by Andreas Kling.
16250
16251         * QtTestBrowser/QtTestBrowser.pro:
16252         * QtTestBrowser/cookiejar.cpp: Added.
16253         * QtTestBrowser/cookiejar.h: Added.
16254         * QtTestBrowser/launcherwindow.cpp:
16255         * QtTestBrowser/launcherwindow.h:
16256         * QtTestBrowser/main.cpp:
16257
16258 2011-08-04  Adam Barth  <abarth@webkit.org>
16259
16260         model.js needs some unit testing
16261         https://bugs.webkit.org/show_bug.cgi?id=65730
16262
16263         Reviewed by Dimitri Glazkov.
16264
16265         These tests do not cover all the functions exported by the model
16266         module.  We'll need to add the remaining ones in the future.
16267
16268         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model_unittests.js: Added.
16269         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
16270
16271 2011-08-04  Adam Barth  <abarth@webkit.org>
16272
16273         Another speculative fix for the "svn revision on Windows" problem.
16274
16275         * Scripts/webkitpy/common/checkout/scm/git.py:
16276
16277 2011-08-04  Mark Rowe  <mrowe@apple.com>
16278
16279         Future-proof Xcode configuration settings.
16280
16281         * DumpRenderTree/mac/Configurations/Base.xcconfig:
16282         * DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig:
16283         * DumpRenderTree/mac/Configurations/DebugRelease.xcconfig:
16284         * MiniBrowser/Configurations/Base.xcconfig:
16285         * MiniBrowser/Configurations/CompilerVersion.xcconfig:
16286         * MiniBrowser/Configurations/DebugRelease.xcconfig:
16287         * TestWebKitAPI/Configurations/Base.xcconfig:
16288         * TestWebKitAPI/Configurations/CompilerVersion.xcconfig:
16289         * TestWebKitAPI/Configurations/DebugRelease.xcconfig:
16290         * WebKitTestRunner/Configurations/Base.xcconfig:
16291         * WebKitTestRunner/Configurations/CompilerVersion.xcconfig:
16292         * WebKitTestRunner/Configurations/DebugRelease.xcconfig:
16293
16294 2011-08-04  Adam Barth  <abarth@webkit.org>
16295
16296         Fancier debug logging.
16297
16298         * Scripts/webkitpy/layout_tests/controllers/manager.py:
16299
16300 2011-08-04  David Levin  <levin@chromium.org>
16301
16302         [chromium] Fix assertion failure in ThreadIdentifierData::identifier in TestShell.
16303         https://bugs.webkit.org/show_bug.cgi?id=65685
16304
16305         Reviewed by Dmitry Titov.
16306
16307         * DumpRenderTree/chromium/WebViewHost.cpp:
16308         (WebViewHost::WebViewHost): Intialize WTF threading before
16309         anything in WTF is used. Usually this is done by WebViewImpl::WebViewImpl
16310         but in some cases code may use WTF without creating WebViewImpl and
16311         WTF code functionality may call WTF::currentThread, which requires
16312         this initialization now.
16313
16314 2011-08-04  Adam Barth  <abarth@webkit.org>
16315
16316         An educated guess at causing the Windows bots on build.webkit.org to
16317         find the SVN revision number.
16318
16319         * Scripts/webkitpy/layout_tests/controllers/manager.py:
16320         * Scripts/webkitpy/layout_tests/port/base.py:
16321
16322 2011-08-04  Adam Barth  <abarth@webkit.org>
16323
16324         More logging to help debug the "no SVN rev on Chromium Windows bots"
16325         bug.  My current theory is that we're trying to detect the SCM system
16326         in the wrong directory.
16327
16328         * Scripts/webkitpy/layout_tests/port/base.py:
16329
16330 2011-08-04  Dimitri Glazkov  <dglazkov@chromium.org>
16331
16332         Speed up webkit-patch optimize-expectations significantly by not collecting test files.
16333         https://bugs.webkit.org/show_bug.cgi?id=65697
16334
16335         I neglected to realize that collecting test names is not necessary for this operation.
16336
16337         Reviewed by Adam Barth.
16338
16339         * Scripts/webkitpy/tool/commands/expectations.py: Changed to pass an empty list of test files.
16340
16341 2011-08-04  Adam Barth  <abarth@webkit.org>
16342
16343         checkout_unittest.js should have better coverage of checkout.js
16344         https://bugs.webkit.org/show_bug.cgi?id=65671
16345
16346         Reviewed by Dimitri Glazkov.
16347
16348         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
16349             - Remove unused functions.
16350         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js:
16351             - Test remaining public functions in the checkout module.
16352
16353 2011-08-04  Dmitry Lomov  <dslomov@google.com>
16354
16355         https://bugs.webkit.org/show_bug.cgi?id=65706
16356         Run run-unit-tests on release mode test bots.
16357         Reenabling after 61812 is fixed.
16358
16359         Reviewed by Adam Roben.
16360
16361         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
16362
16363 2011-08-04  Adam Barth  <abarth@webkit.org>
16364
16365         builders.js needs unit tests
16366         https://bugs.webkit.org/show_bug.cgi?id=65670
16367
16368         Reviewed by Dimitri Glazkov.
16369
16370         This test requires somewhat large test fixtures, but using real
16371         examples seemed better than trimming them down too much.
16372
16373         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js:
16374         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders_unittests.js: Added.
16375         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
16376             - Note: trac_unittests.js doesn't exist (yet!).
16377
16378 2011-08-04  Alexis Menard  <alexis.menard@openbossa.org>
16379
16380         [Qt] Make navigation actions properly usable in QML.
16381         https://bugs.webkit.org/show_bug.cgi?id=65624
16382
16383         Fix compilation after the introduction of QWebNavigationController.
16384
16385         Reviewed by Benjamin Poulain.
16386
16387         * MiniBrowser/qt/BrowserView.cpp:
16388         (BrowserView::navigationAction):
16389
16390 2011-08-04  Adam Barth  <abarth@webkit.org>
16391
16392         Another subtle bug due to lack of testing in main.js.  :(
16393
16394         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
16395
16396 2011-08-04  Adam Barth  <abarth@webkit.org>
16397
16398         Fix types.  Sadly, main.js has no test coverage.  (I need to think
16399         about how to test this part of the code.)
16400
16401         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
16402
16403 2011-08-04  Adam Barth  <abarth@webkit.org>
16404
16405         bugzilla.js is missing unit tests
16406         https://bugs.webkit.org/show_bug.cgi?id=65656
16407
16408         Reviewed by Adam Roben.
16409
16410         Testing makes perfect.
16411
16412         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/bugzilla_unittests.js: Added.
16413         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net_unittests.js:
16414         (NetworkSimulator.prototype.runTest.get if):
16415         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
16416
16417 2011-08-04  Adam Barth  <abarth@webkit.org>
16418
16419         Add missing license blocks to garden-o-matic
16420         https://bugs.webkit.org/show_bug.cgi?id=65654
16421
16422         Reviewed by Dimitri Glazkov.
16423
16424         This patch is entirely boilerplate.
16425
16426         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
16427         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
16428         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js:
16429         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js:
16430         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
16431         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js:
16432         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
16433         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
16434         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
16435         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
16436         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
16437         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net.js:
16438         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net_unittests.js:
16439         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
16440         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js:
16441         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
16442         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
16443         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
16444
16445 2011-08-04  Adam Barth  <abarth@webkit.org>
16446
16447         garden-o-matic needs a way to mock out the network
16448         https://bugs.webkit.org/show_bug.cgi?id=65653
16449
16450         Reviewed by Dimitri Glazkov.
16451
16452         This patch introduces the net module, which is a useful mock point for
16453         abstracting away the network.  Future patches will use this
16454         infrastructure to test some currently untested code.
16455
16456         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js:
16457         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
16458         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
16459         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js:
16460         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
16461         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
16462         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
16463         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net.js: Added.
16464         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/net_unittests.js: Added.
16465         (NetworkSimulator):
16466         (NetworkSimulator.prototype.scheduleCallback):
16467         (NetworkSimulator.prototype.runTest):
16468         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
16469         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js:
16470         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
16471         * Scripts/webkitpy/tool/servers/gardeningserver.py:
16472
16473 2011-08-03  Adam Barth  <abarth@webkit.org>
16474
16475         Use failureInfo more pervasively in garden-o-matic
16476         https://bugs.webkit.org/show_bug.cgi?id=65650
16477
16478         Reviewed by Dimitri Glazkov.
16479
16480         These three variable appears in concert together many places.  This
16481         patch unifies the remaining uses into being failureInfo objects, which
16482         saves some marshalling.
16483
16484         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
16485         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
16486         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js:
16487         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
16488         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
16489
16490 2011-08-03  Adam Barth  <abarth@webkit.org>
16491
16492         garden-o-matic should be able to update expecations from the details pane
16493         https://bugs.webkit.org/show_bug.cgi?id=65648
16494
16495         Reviewed by Dimitri Glazkov.
16496
16497         The new button queues up the updates, which get executed all at once.
16498
16499         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
16500         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
16501         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
16502
16503 2011-08-03  Adam Barth  <abarth@webkit.org>
16504
16505         Add UI to garden-o-matic for updating expectations
16506         https://bugs.webkit.org/show_bug.cgi?id=65644
16507
16508         Reviewed by Dimitri Glazkov.
16509
16510         The patch plumbs the UI back to the gardening server, but the gardening
16511         server endpoint is just a stub at this point.
16512
16513         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
16514         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
16515         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
16516         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
16517         * Scripts/webkitpy/tool/servers/gardeningserver.py:
16518         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
16519
16520 2011-08-03  Adam Barth  <abarth@webkit.org>
16521
16522         garden-o-matic details view should queue rebaselines for later execution
16523         https://bugs.webkit.org/show_bug.cgi?id=65636
16524
16525         Reviewed by Dimitri Glazkov.
16526
16527         This patch doesn't contain any confirmation UI, but we'll probably want
16528         that at some point.
16529
16530         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
16531         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
16532         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
16533         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
16534
16535 2011-08-03  Kentaro Hara  <haraken@google.com>
16536
16537         Implement EventSender.scalePageBy()
16538         https://bugs.webkit.org/show_bug.cgi?id=58013
16539
16540         Reviewed by Darin Fisher.
16541
16542         Implemented EventSender.scalePageBy(f, x, y), which scales a page by a factor of f
16543         and then sets a scroll position to (x, y). Enabled the tests that had been waiting
16544         for the implementation of EventSender.scalePageBy(f, x, y).
16545
16546         Tests: compositing/scaling/tiled-layer-recursion.html
16547                fast/repaint/scale-page-shrink.html
16548                fast/dom/Element/scale-page-client-rects.html
16549                fast/dom/Range/scale-page-client-rects.html
16550                fast/events/scroll-in-scaled-page-with-overflow-hidden.html
16551                fast/dom/Element/scale-page-bounding-client-rect.html
16552                fast/dom/Range/scale-page-bounding-client-rect.html
16553
16554         * DumpRenderTree/chromium/EventSender.cpp:
16555         (EventSender::EventSender): Added bindings for scalePageBy().
16556         (EventSender::scalePageBy): A wrapper method for scalePage() in WebView.
16557         * DumpRenderTree/chromium/EventSender.h:
16558         * DumpRenderTree/chromium/TestShell.cpp:
16559         (TestShell::resetTestController): Resets the scale factor to 1.
16560
16561 2011-08-03  David Levin  <levin@chromium.org>
16562
16563         Rename WEBKIT_API to WEBKIT_EXPORT in check-webkit-style.
16564         https://bugs.webkit.org/show_bug.cgi?id=65652
16565
16566         Reviewed by Adam Barth.
16567
16568         * Scripts/webkitpy/style/checkers/cpp.py:
16569         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
16570
16571 2011-08-03  Filip Pizlo  <fpizlo@apple.com>
16572
16573         Adding Filip Pizlo to committer list.
16574
16575         Unreviewed.
16576
16577         * Scripts/webkitpy/common/config/committers.py:
16578
16579 2011-08-03  Dimitri Glazkov  <dglazkov@chromium.org>
16580
16581         Add "optimize-expectations" command to webkit-patch.
16582         https://bugs.webkit.org/show_bug.cgi?id=65633
16583
16584         For now, only operates on Chromium test_expectations.txt and does not report errors/warnings.
16585
16586         Reviewed by Adam Barth.
16587
16588         * Scripts/webkitpy/tool/commands/__init__.py: Added command.
16589         * Scripts/webkitpy/tool/commands/expectations.py: Added.
16590
16591 2011-08-03  Dimitri Glazkov  <dglazkov@chromium.org>
16592
16593         Allow multiple bug identifiers in test expectations.
16594         https://bugs.webkit.org/show_bug.cgi?id=65642
16595
16596         Reviewed by Adam Barth.
16597
16598         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Tweaked to allow multiple bug ids.
16599         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests.
16600
16601 2011-08-03  Dimitri Glazkov  <dglazkov@chromium.org>
16602
16603         Teach BuilderToPort to distinguish between Release and Debug builders.
16604         https://bugs.webkit.org/show_bug.cgi?id=65628
16605
16606         Reviewed by Adam Barth.
16607
16608         * Scripts/webkitpy/tool/commands/rebaseline.py: Added a smart options object and passed it to port factory.
16609         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py: Added tests.
16610
16611 2011-08-03  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
16612
16613         [Qt] Fix running qmltests for WK2 in the bot
16614         https://bugs.webkit.org/show_bug.cgi?id=65621
16615
16616         Reviewed by Csaba Osztrogonác.
16617
16618         The path for QML_IMPORT_PATH was assuming we were running the WK1 tests,
16619         but now there's the case for WK2 UIProcess tests as well.
16620
16621         * Scripts/run-qtwebkit-tests: add both possible paths to the QML_IMPORT_PATH.
16622
16623 2011-08-02  Dimitri Glazkov  <dglazkov@chromium.org>
16624
16625         Add a way to edit test expectations.
16626         https://bugs.webkit.org/show_bug.cgi?id=64922
16627
16628         Implements a TestExpectationsEditor, which provides a standard way to
16629         update and remove test expectations.
16630
16631         Reviewed by Adam Barth.
16632
16633         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed TestExpectationSerializer to skip lines that match no test configurations,
16634             added TestExpectationLine.is_flaky, added BugManager, a simple interface to be fleshed out later, and a TestExpectationsEditor, the big
16635             enchilada.
16636         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Loads of tests to capture desired behavior.
16637
16638 2011-08-03  Andreas Kling  <kling@webkit.org>
16639
16640         [Qt] MiniBrowser: Unbreak load progress indication.
16641         https://bugs.webkit.org/show_bug.cgi?id=65614
16642
16643         Reviewed by Kenneth Rohde Christiansen.
16644
16645         loadProgress() was renamed to loadProgressChanged().
16646
16647         * MiniBrowser/qt/BrowserWindow.cpp:
16648         (BrowserWindow::BrowserWindow):
16649         (BrowserWindow::onLoadProgressChanged):
16650         * MiniBrowser/qt/BrowserWindow.h:
16651
16652 2011-08-03  Martin Robinson  <mrobinson@igalia.com>
16653
16654         Build fix for WebKit2GTK+.
16655
16656         * WebKitTestRunner/GNUmakefile.am: Include libInternals as one of
16657         of the library dependencies of the injected bundle.
16658
16659 2011-08-03  Carlos Garcia Campos  <cgarcia@igalia.com>
16660
16661         [GTK] Install web and plugin processes in libexecdir instead of bindir
16662         https://bugs.webkit.org/show_bug.cgi?id=65600
16663
16664         Reviewed by Martin Robinson.
16665
16666         Set WEBKIT_EXEC_PATH environment variable to use web and plugin
16667         process from current directory instead of the intalled ones.
16668
16669         * MiniBrowser/gtk/GNUmakefile.am:
16670         * MiniBrowser/gtk/main.c:
16671         (main):
16672         * WebKitTestRunner/GNUmakefile.am:
16673         * WebKitTestRunner/gtk/main.cpp:
16674         (main):
16675
16676 2011-08-02  Anders Carlsson  <andersca@apple.com>
16677
16678         Add mac-lion platform to old-run-webkit-tests
16679         https://bugs.webkit.org/show_bug.cgi?id=65577
16680
16681         * Scripts/old-run-webkit-tests:
16682         * Scripts/webkitdirs.pm:
16683         (isLion):
16684
16685 2011-08-02  Alexandru Chiculita  <achicu@adobe.com>
16686
16687         Buildbot: The Properties object has no "hasProperty" method in buildbot 8.3
16688         https://bugs.webkit.org/show_bug.cgi?id=65554
16689
16690         Using "has_key" instead of "hasProperty" in order to support buildbot 8.3.
16691
16692         Reviewed by Adam Roben.
16693
16694         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
16695
16696 2011-08-02  Jochen Eisinger  <jochen@chromium.org>
16697
16698         Implement WebPermissionClient::allowPlugins for layout tests
16699         https://bugs.webkit.org/show_bug.cgi?id=33991
16700
16701         Reviewed by Adam Barth.
16702
16703         * DumpRenderTree/chromium/LayoutTestController.cpp:
16704         (LayoutTestController::LayoutTestController):
16705         (LayoutTestController::setPluginsAllowed):
16706         * DumpRenderTree/chromium/LayoutTestController.h:
16707         * DumpRenderTree/chromium/WebPermissions.cpp:
16708         (WebPermissions::allowPlugins):
16709         (WebPermissions::setPluginsAllowed):
16710         (WebPermissions::reset):
16711         * DumpRenderTree/chromium/WebPermissions.h:
16712
16713 2011-08-02  Adam Barth  <abarth@webkit.org>
16714
16715         TestFailures unit tests leak a global variable
16716         https://bugs.webkit.org/show_bug.cgi?id=65500
16717
16718         Reviewed by Adam Roben.
16719
16720         qunit detects the leak if you check the "noglobals" box on run-unittests.html.
16721
16722         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
16723
16724 2011-08-02  Alexandru Chiculita  <achicu@adobe.com>
16725
16726         CSS Regions build bot should archive and upload output files
16727         https://bugs.webkit.org/show_bug.cgi?id=65200
16728
16729         Until now the buildbots would only upload the result archive if the 'trigger' parameters was specified. The
16730         upload was needed, because the testbots needed to download and use the archive. For CSS Regions we don't have
16731         a testbot yet, but we want to save the archive on the server for manual testing. Added new parameter
16732         in config.json, called "upload" that when set to "true" will force the upload to the buildmaster server.
16733
16734         The "features" argument is now taken into account when determining the name of the archive on the server. That
16735         was needed in order to avoid the CSS Regions Mac buildbot overwrite the SnowLeopard Mac build.
16736
16737         Reviewed by Adam Roben.
16738
16739         * BuildSlaveSupport/build.webkit.org-config/config.json:
16740         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
16741
16742 2011-08-02  Patrick Gansterer  <paroga@webkit.org>
16743
16744         Fix _win32_check_running_pid on 64bit windows python
16745         https://bugs.webkit.org/show_bug.cgi?id=64950
16746
16747         Reviewed by Adam Roben.
16748
16749         Use the correct types for the PROCESSENTRY32 struct.
16750         th32DefaultHeapID is declared as ULONG_PTR and not as DWORD.
16751         Using 32bit for pointers instead of 64bit results in an ERROR_BAD_LENGTH
16752         failure, when calling the Process32First function.
16753         This is covered by the executive unittest.
16754
16755         * Scripts/webkitpy/common/system/executive.py:
16756
16757 2011-08-01  Adam Barth  <abarth@webkit.org>
16758
16759         garden-o-matic should call optimize-baselines when rebaselining tests
16760         https://bugs.webkit.org/show_bug.cgi?id=65499
16761
16762         Reviewed by Dimitri Glazkov.
16763
16764         I took the opportunity to modernize this code to use some of our more
16765         powerful primitives from base.
16766
16767         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
16768         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js:
16769         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js:
16770         * Scripts/webkitpy/tool/servers/gardeningserver.py:
16771         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
16772
16773 2011-08-01  Stephanie Lewis  <slewis@apple.com>
16774
16775         Finish reconfiguration started by Lucas Forschler
16776         Reconfigured slaves:
16777         Removed Leopard Intel Release (Build)
16778         Removed Leopard Intel Release (Tests)
16779         Remove apple-pixel-1
16780         Added Lion slaves from disabled Leopard slaves
16781
16782         Added builders:
16783         Lion Intel Release (Build)
16784         Lion Intel Debug (Build)
16785         Lion Intel Release (Tests)
16786         Lion Intel Release WK2 (Tests)
16787         Lion Intel Debug (Tests)
16788         Lion Intel Debug WK2 (Tests)
16789
16790         Added triggers:
16791         lion-intel-debug-tests
16792         lion-intel-debug-test-wk2
16793         lion-intel-release-tests
16794         lion-intel-release-tests-wk2
16795
16796         Removed trigger:
16797         leopard-intel-release-tests
16798
16799         Added PlatformSpecificScheduler:
16800         mac-lion
16801
16802         Added "mac-lion" to build.py and build_unitttest.py
16803
16804         Reviewed by Mark Rowe.
16805
16806         * BuildSlaveSupport/build.webkit.org-config/config.json:
16807         * Scripts/webkitpy/common/config/build.py:
16808         * Scripts/webkitpy/common/config/build_unittest.py:
16809
16810
16811 2011-08-01  Adam Barth  <abarth@webkit.org>
16812
16813         webkit-patch needs to be able to "optimize" the storage of baselines on disk
16814         https://bugs.webkit.org/show_bug.cgi?id=65418
16815
16816         Reviewed by Dimitri Glazkov.
16817
16818         If we're not careful when rebaselining tests, we can end up with lots
16819         of duplicate expected results files in the tree.  This patch adds a
16820         webkit-patch command that optimizes the storage of expected results on
16821         disk.
16822
16823         This command is similar to deduplicate-tests, except that it can move
16824         test results around rather than just remove duplicate results.
16825
16826         Unfortunately, this problem is very tricky because the baseline search
16827         structure is a hypergraph.  This patch include a huerstic optimizer
16828         that appears to work on a bunch of examples I've tried.  We'll likely
16829         need to refine it as gain experience.
16830
16831         * Scripts/webkitpy/common/system/filesystem.py:
16832         * Scripts/webkitpy/tool/commands/rebaseline.py:
16833
16834 2011-08-01  Dimitri Glazkov  <dglazkov@chromium.org>
16835
16836         Teach TestExpectationSerializer about parsed expectations.
16837         https://bugs.webkit.org/show_bug.cgi?id=65444
16838
16839         This is a bit largish in size, but the essentials are:
16840         * TestExpectationSerializer, given a parsed TestExpectationLine, can now spit out a pretty string.
16841         * We now store original content of the expectation line on TestExpectationLine.
16842         * Lots of tests.
16843
16844         Reviewed by Adam Barth.
16845
16846         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Fixed an issue where we didn't
16847             populate macros correctly, also fixed an issue where we incorrectly returned an empty list
16848             for all configurations.
16849         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added tests for those.
16850         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Removed vestiges of code long gone,
16851             refactored TestExpectationsSerializer to recognize and correctly serialize parsed TestExpectationLine instances,
16852             changed to store and use TestExpectationLine.original_string.
16853         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests.
16854
16855 2011-08-01  Adam Barth  <abarth@webkit.org>
16856
16857         Remove unused functionality from gardeningserver
16858         https://bugs.webkit.org/show_bug.cgi?id=65479
16859
16860         Reviewed by Dimitri Glazkov.
16861
16862         We now handle these operations directly from the front-end.  That means
16863         they work even without the gardening server running!
16864
16865         Also, add some missing files to the list of static files.
16866
16867         * Scripts/webkitpy/tool/servers/gardeningserver.py:
16868         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
16869
16870 2011-08-01  Dimitri Glazkov  <dglazkov@chromium.org>
16871
16872         Introduce SpecifierSorter, a thing that knows how specifiers should be ordered.
16873         https://bugs.webkit.org/show_bug.cgi?id=65413
16874
16875         Reviewed by Adam Barth.
16876
16877         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added SpecifierSorter and converted
16878             existing code to use it.
16879         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added unit tests and cleaned up a bit.
16880
16881 2011-08-01  Adam Barth  <abarth@webkit.org>
16882
16883         Refactor bugzilla.js for use by garden-o-matic
16884         https://bugs.webkit.org/show_bug.cgi?id=65450
16885
16886         Reviewed by Dimitri Glazkov.
16887
16888         This patch refactors bugzilla.js to use the AsynchronousCache and
16889         updates the style to use a module instead of an object.  This patch
16890         then fixes all the existing code that uses this class to use the new
16891         API style.
16892
16893         This main benefit of this patch is we remove the tricky manual caching
16894         and this code is now available to use in garden-o-matic (since the
16895         dependency on Utilities.js is now gone).
16896
16897         I ran all the unit tests and poked around in TestFailures a bit to see
16898         that everything seemed to be working properly.
16899
16900         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js:
16901         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js:
16902         (FailingTestsBugForm):
16903         (FailingTestsBugForm.prototype._createBugTitle):
16904         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js:
16905         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm.js:
16906         (FlakyTestBugForm):
16907         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm_unittests.js:
16908         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js:
16909         (NewBugForm):
16910         (NewBugForm.prototype.domElement):
16911         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js:
16912         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm.js:
16913         (TestRelatedBugForm):
16914         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm_unittests.js:
16915         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
16916         (ViewController.prototype._displayBuilder.start):
16917         (ViewController.prototype._displayBuilder):
16918         (ViewController.prototype._domForAuxiliaryUIElements):
16919         (ViewController.prototype._domForNewAndExistingBugs.bugzilla.quickSearch):
16920         (ViewController.prototype._domForPossiblyFlakyTests.flakyList.appendChildren):
16921         (ViewController.prototype._domForPossiblyFlakyTests):
16922         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
16923         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
16924         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
16925
16926 2011-08-01  Adam Barth  <abarth@webkit.org>
16927
16928         garden-o-matic shouldn't require local server to determine if compile failed
16929         https://bugs.webkit.org/show_bug.cgi?id=65446
16930
16931         Reviewed by Dimitri Glazkov.
16932
16933         This patch adds some basic bindings for the buildbot JSON API.  I
16934         wanted to use Buildbot.js an Builder.js from the original TestFailures,
16935         but they make a number of WebKit-specific assumptions.  We'll
16936         eventually need to reconcile these pieces of code.
16937
16938         I should really add tests for this code, but that will require me to
16939         build a better mock network.  I'll be happy to do that in a follow-up
16940         patch.
16941
16942         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/builders.js: Added.
16943             - It's very unclear to me whether didFail is correct.  Supposedly
16944               the first entry here is the exit code of the step, which is 0 on
16945               success.  I suspect we'll need to improve this function over time.
16946         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
16947         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
16948         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
16949
16950 2011-08-01  Adam Barth  <abarth@webkit.org>
16951
16952         garden-o-matic should show all failures, even ones that it can't find a regression range for
16953         https://bugs.webkit.org/show_bug.cgi?id=65447
16954
16955         Reviewed by Dimitri Glazkov.
16956
16957         Previously, we just wouldn't display test failures that we couldn't
16958         identify a regression range for (or whose regression range had fallen
16959         off the page).  This patch just displays them at the end of the list,
16960         which is slightly (but not much) better.
16961
16962         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
16963
16964 2011-08-01  Jeff Miller  <jeffm@apple.com>
16965
16966         Windows build fix after r92096.
16967
16968         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Use #include, not #import, and don't include Foundation.h.
16969         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Add InjectedBundleControllerWin.cpp.
16970
16971 2011-07-31  Anders Carlsson  <andersca@apple.com>
16972
16973         DownloadDecideDestinationCrash WebKit2 API tests crashes
16974         https://bugs.webkit.org/show_bug.cgi?id=65442
16975
16976         Reviewed by Sam Weinig.
16977
16978         Copy the 18-characters.html file to the destination directory so we won't
16979         crash trying to make an URL for it.
16980
16981         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
16982
16983 2011-07-31  Anders Carlsson  <andersca@apple.com>
16984
16985         Fix SpacebarScrolling WebKit2 API test
16986         https://bugs.webkit.org/show_bug.cgi?id=65431
16987
16988         Reviewed by Sam Weinig.
16989
16990         Add a way to do platform-specific bundle initialization, and use it to disable smooth
16991         scrolling. Also make it possible for a test to load the injected bundle without actually running
16992         a test from it.
16993
16994         * TestWebKitAPI/InjectedBundleController.cpp:
16995         (TestWebKitAPI::InjectedBundleController::initialize):
16996         * TestWebKitAPI/InjectedBundleController.h:
16997         * TestWebKitAPI/PlatformUtilities.cpp:
16998         (TestWebKitAPI::Util::createContextWithInjectedBundle):
16999         * TestWebKitAPI/PlatformUtilities.h:
17000         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
17001         * TestWebKitAPI/Tests/WebKit2/SpacebarScrolling.cpp:
17002         (TestWebKitAPI::TEST):
17003         * TestWebKitAPI/mac/InjectedBundleControllerMac.mm: Added.
17004         (TestWebKitAPI::InjectedBundleController::platformInitialize):
17005         * TestWebKitAPI/win/InjectedBundleControllerWin.cpp: Added.
17006         (TestWebKitAPI::InjectedBundleController::platformInitialize):
17007         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
17008
17009 2011-07-31  Dimitri Glazkov  <dglazkov@chromium.org>
17010
17011         Use set comparison to find duplicate or overlapping specifiers in test expectations.
17012         https://bugs.webkit.org/show_bug.cgi?id=65430
17013
17014         Reviewed by Adam Barth.
17015
17016         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed to use set comparisons, because length
17017             of a set is wrong.
17018         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Added some unit tests to actually catch
17019             regressions.
17020
17021 2011-07-31  Xan Lopez  <xlopez@igalia.com>
17022
17023         Group dependencies CFLAGS together.
17024
17025         * GNUmakefile.am: ditto.
17026
17027 2011-07-31  Daniel Bates  <dbates@webkit.org>
17028
17029         Attempt to fix the GTK build after <http://trac.webkit.org/changeset/92059>
17030         (https://bugs.webkit.org/show_bug.cgi?id=65419).
17031
17032         * GNUmakefile.am: Add $(CAIRO_CFLAGS) to libWebCoreInternals_la_CPPFLAGS.
17033
17034 2011-07-30  Patrick Gansterer  <paroga@webkit.org>
17035
17036         Fix ports_unittest.py on native windows
17037         https://bugs.webkit.org/show_bug.cgi?id=64959
17038
17039         Reviewed by Adam Barth.
17040
17041         Replace WebKitPort.script_path with WebKitPort.script_shell_command.
17042         The latter includes the interpreter executable required on native windows.
17043
17044         * Scripts/webkitpy/common/config/ports_unittest.py:
17045
17046 2011-07-30  Adam Barth  <abarth@webkit.org>
17047
17048         Properly escape regular expressions.  (Maybe these shouldn't be regular
17049         expressions?)
17050
17051         * Scripts/webkitpy/tool/commands/rebaseline.py:
17052
17053 2011-07-30  Adam Barth  <abarth@webkit.org>
17054
17055         Add support for Win7 to garden-o-matic.
17056
17057         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
17058         * Scripts/webkitpy/tool/commands/rebaseline.py:
17059
17060 2011-07-30  Adam Barth  <abarth@webkit.org>
17061
17062         garden-o-matic regression ranges expand after fast bots start passing
17063         https://bugs.webkit.org/show_bug.cgi?id=65406
17064
17065         Reviewed by Dimitri Glazkov.
17066
17067         Once the gardener has fixed a problem, the fast bots pick up the fix,
17068         which means we end up using the slow bots to compute the regression
17069         range.  Unfortunately, the slow bots give us large regression ranges,
17070         leading to this strange effect where the regression range quality
17071         degrades as fixes roll in.
17072
17073         This patch causes us to remember the previous regression ranges we've
17074         seen and to use that information in the future, as long as it's
17075         consistent with the new information we receive.
17076
17077         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
17078
17079 2011-07-30  Adam Barth  <abarth@webkit.org>
17080
17081         garden-o-matic should have a Refresh button to update its display
17082         https://bugs.webkit.org/show_bug.cgi?id=65403
17083
17084         Reviewed by Dimitri Glazkov.
17085
17086         The refresh button should be faster and less disruptive than the browser's reload button.
17087
17088         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
17089         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
17090
17091 2011-07-30  Adam Barth  <abarth@webkit.org>
17092
17093         garden-o-matic should look for the "patch by" line in revisions ostensibly authored by the commit-queue
17094         https://bugs.webkit.org/show_bug.cgi?id=65231
17095
17096         Reviewed by Dimitri Glazkov.
17097
17098         The human readable name is easier to understand, especially when the patch is landed by the commit-queue.
17099
17100         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
17101
17102 2011-07-30  Adam Barth  <abarth@webkit.org>
17103
17104         garden-o-matic should strikeout revisions that have been rolled out
17105         https://bugs.webkit.org/show_bug.cgi?id=65227
17106
17107         Reviewed by Dimitri Glazkov.
17108
17109         This helps gardeners keep track of which problems have been solved.
17110
17111         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
17112         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
17113         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
17114         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
17115
17116 2011-07-30  Adam Barth  <abarth@webkit.org>
17117
17118         Add missing files to the server.
17119
17120         * Scripts/webkitpy/tool/servers/gardeningserver.py:
17121
17122 2011-07-29  Adam Barth  <abarth@webkit.org>
17123
17124         Actually import the logging module!
17125
17126         * Scripts/webkitpy/common/checkout/scm/git.py:
17127
17128 2011-07-29  Adam Barth  <abarth@webkit.org>
17129
17130         More debug logging.  Maybe we're running the git version of this
17131         function somehow?  (That would explain why it doesn't work!)
17132
17133         * Scripts/webkitpy/common/checkout/scm/git.py:
17134
17135 2011-07-29  Adam Barth  <abarth@webkit.org>
17136
17137         Add some temporary debug logging messages to try to understand why the
17138         windows bots can't figure out their current SVN revision.
17139
17140         * Scripts/webkitpy/common/checkout/scm/svn.py:
17141
17142 2011-07-29  Adam Barth  <abarth@webkit.org>
17143
17144         Chromium try servers should use downstream test_expectations
17145         https://bugs.webkit.org/show_bug.cgi?id=65390
17146
17147         Reviewed by Dirk Pranke.
17148
17149         Otherwise the try servers will report bogus failures.
17150
17151         * Scripts/webkitpy/layout_tests/port/chromium.py:
17152
17153 2011-07-29  Dimitri Glazkov  <dglazkov@chromium.org>
17154
17155         Fix an error in macro collapsing, refactor code to be pretty.
17156         https://bugs.webkit.org/show_bug.cgi?id=65387
17157
17158         My set math was wrong.
17159
17160         Reviewed by Adam Barth.
17161
17162         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Moved collapsing into a classmethod, fixed the math to be correct.
17163         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added unit tests.
17164
17165 2011-07-27  Ryuan Choi  <ryuan.choi@samsung.com>
17166
17167         Style bot complains about EFL-specific forward declaration naming.
17168         https://bugs.webkit.org/show_bug.cgi?id=64680
17169
17170         Exclude EFL-specific forward declarations from indentifiers with underscores checks.
17171
17172         Reviewed by Adam Barth.
17173
17174         * Scripts/webkitpy/style/checkers/cpp.py:
17175         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
17176
17177 2011-07-27  Adam Roben  <aroben@apple.com>
17178
17179         Use wacky !svn/bc URLs consistently in Leaks Viewer
17180
17181         This ensures that any resources loaded via relative URLs will still get pulled from the same
17182         revision of Subversion.
17183
17184         Fixes <http://webkit.org/b/65267> Focus/exclude button glyphs don't load on Leaks Viewer
17185
17186         Reviewed by David Kilzer.
17187
17188         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html: Use the
17189         !svn/bc-style URL for inspector.css, too, just like we do for all other resources we pull
17190         from Subversion.
17191
17192 2011-07-27  Raphael Kubo da Costa  <kubo@profusion.mobi>
17193
17194         [EFL] Drop WorkQueueItemEfl.h in favour of the general WorkQueueItem.h
17195         https://bugs.webkit.org/show_bug.cgi?id=65269
17196
17197         Reviewed by Kent Tamura.
17198
17199         WorkQueueItemEfl.h was being included by efl's
17200         LayoutTestControllerEfl.cpp, however the top-level
17201         LayoutTestController still included WorkQueueItem.h, causing a lot of
17202         type mismatches when running the WorkQueueItem code.
17203
17204         We can stick to the original WorkQueueItem header by making a few
17205         changes to EFL's LayoutTestController.
17206
17207         This should make tests such as
17208         fast/loader/non-deferred-substitute-load.html stop crashing.
17209
17210         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
17211         (LayoutTestController::queueLoad):
17212         * DumpRenderTree/efl/WorkQueueItemEfl.cpp:
17213         (LoadItem::invoke):
17214         (LoadHTMLStringItem::invoke):
17215         (ScriptItem::invoke):
17216         * DumpRenderTree/efl/WorkQueueItemEfl.h: Removed.
17217
17218 2011-07-27  Dimitri Glazkov  <dglazkov@chromium.org>
17219
17220         Re-add unit tests I accidentally zapped in r91878.
17221
17222         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Re-added unit tests.
17223
17224 2011-07-27  Dimitri Glazkov  <dglazkov@chromium.org>
17225
17226         Replace SpecificityCalculator with TestConfiguration-driven logic.
17227         https://bugs.webkit.org/show_bug.cgi?id=65206
17228
17229         Use our newly-acquired ability to expand modifiers into a set of
17230         matching TestConfiguration instances to calculate specificity and
17231         determine whether expectation applies to a given test configuration.
17232
17233         Also, store bug modifier on TestExpectationsLine.
17234
17235         Reviewed by Adam Barth.
17236
17237         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Replaced the logic of calculating specificity with
17238             test configuration matching, removed a bunch of code.
17239         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Removed a bunch of tests that aren't useful anymore.
17240
17241
17242 2011-07-27  Dimitri Glazkov  <dglazkov@chromium.org>
17243
17244         Allow ports to specify their own test expectation specifier macros.
17245         https://bugs.webkit.org/show_bug.cgi?id=65291
17246
17247         Reviewed by Adam Barth.
17248
17249         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed SpecificityCalculator to use port macros.
17250         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Ditto.
17251         * Scripts/webkitpy/layout_tests/port/base.py: Added configuration_specifier_macros to retrieve the macros.
17252         * Scripts/webkitpy/layout_tests/port/chromium.py: Overrode it to return Chromium's macros.
17253         * Scripts/webkitpy/layout_tests/port/test.py: Added test macros.
17254
17255 2011-07-27  Dimitri Glazkov  <dglazkov@chromium.org>
17256
17257         Allow TestConfigurationConverter report conversion errors.
17258         https://bugs.webkit.org/show_bug.cgi?id=65287
17259
17260         Also, fix a few miscellaneous testing issues.
17261
17262         Reviewed by Adam Barth.
17263
17264         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added a way to report unknown modifier errors.
17265         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added a test for reporting errors.
17266         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py: Fixed an error where an invalid TestConfiguration instance
17267             was created. 'default' is not a valid build type.
17268         * Scripts/webkitpy/layout_tests/port/test.py: Ditto. '' is not a valid version.
17269
17270 2011-07-27  Dimitri Glazkov  <dglazkov@chromium.org>
17271
17272         Teach all ports about their test configurations.
17273         https://bugs.webkit.org/show_bug.cgi?id=65278
17274
17275         Each Port object should return a valid set of all TestConfiguration instances,
17276         available for this port.
17277
17278         Reviewed by Philippe Normand.
17279
17280         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Removed TestConfiguration.all_test_configurations,
17281             since it is different for each port.
17282         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Removed related test.
17283         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed SpecificityCalculator to take a list of
17284             all test configurations as a param.
17285         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Adjusted arguments accordingly.
17286         * Scripts/webkitpy/layout_tests/port/base.py: Forced all ports to override generation of all test configurations.
17287         * Scripts/webkitpy/layout_tests/port/chromium.py: Implemented _generate_all_test_configurations for the port.
17288         * Scripts/webkitpy/layout_tests/port/gtk.py: Ditto.
17289         * Scripts/webkitpy/layout_tests/port/mac.py: Ditto.
17290         * Scripts/webkitpy/layout_tests/port/qt.py: Ditto.
17291         * Scripts/webkitpy/layout_tests/port/win.py:
17292         * Scripts/webkitpy/layout_tests/port/port_testcase.py: Added an extra test to ensure we're generating a good set.
17293         * Scripts/webkitpy/layout_tests/port/test.py: Implemented a fixed-in-time all_test_configurations for testing.
17294         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py: Added a mock all_test_configurations implementation.
17295         * Scripts/webkitpy/layout_tests/port/win_unittest.py: Added a unit test for the port.
17296
17297 2011-07-27  Benjamin Poulain  <benjamin@webkit.org>
17298
17299         [WK2][Qt] Move from QGraphicsView to Qt Scene Graph
17300         https://bugs.webkit.org/show_bug.cgi?id=62969
17301
17302         Reviewed by Andreas Kling.
17303
17304         Co-authored with Simon Hausmann  <simon.hausmann@nokia.com>
17305
17306         With the scene graph, the interface to have a window on screen is
17307         QSGCanvas. QGraphicsView is not used, and openGL is an enforced
17308         dependency.
17309
17310         Both the BrowserView and the WebKitTest runner are changed from
17311         QGraphicsView to QSGCanvas.
17312
17313         * MiniBrowser/qt/BrowserView.cpp:
17314         (BrowserView::BrowserView):
17315         (BrowserView::resizeEvent):
17316         (BrowserView::view):
17317         * MiniBrowser/qt/BrowserView.h:
17318         * MiniBrowser/qt/BrowserWindow.cpp:
17319         (BrowserWindow::BrowserWindow):
17320         (BrowserWindow::urlChanged):
17321         (BrowserWindow::~BrowserWindow):
17322         * MiniBrowser/qt/BrowserWindow.h:
17323         * MiniBrowser/qt/MiniBrowser.pro:
17324         * MiniBrowser/qt/main.cpp:
17325         (main):
17326         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
17327         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
17328         (WTR::WebView::WebView):
17329         (WTR::PlatformWebView::PlatformWebView):
17330         * WebKitTestRunner/qt/WebKitTestRunner.pro:
17331
17332 2011-07-27  Sheriff Bot  <webkit.review.bot@gmail.com>
17333
17334         Unreviewed, rolling out r91792.
17335         http://trac.webkit.org/changeset/91792
17336         https://bugs.webkit.org/show_bug.cgi?id=65242
17337
17338         It broke NRWT on Qt and GTK (Requested by Ossy_night on
17339         #webkit).
17340
17341         * Scripts/webkitpy/layout_tests/models/test_configuration.py:
17342         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py:
17343         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
17344         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
17345         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
17346         * Scripts/webkitpy/layout_tests/port/test.py:
17347
17348 2011-07-26  Adam Barth  <abarth@webkit.org>
17349
17350         Switch to an older idiom that works on Python 2.5.
17351
17352         * Scripts/webkitpy/layout_tests/models/test_configuration.py:
17353
17354 2011-07-26  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
17355
17356         Add an exception to coding style check for EFL port.
17357         https://bugs.webkit.org/show_bug.cgi?id=65162
17358
17359         Add an exception of whitespace/declaration for EFL port. Because,
17360         '*' operator is next to variable in EFL coding style.
17361
17362         Reviewed by Kent Tamura.
17363
17364         * Scripts/webkitpy/style/checker.py:
17365
17366 2011-07-26  Keith Kyzivat  <keith.kyzivat@nokia.com>
17367
17368         [Qt] Add option to turn on disk caching in QtTestBrowser
17369         https://bugs.webkit.org/show_bug.cgi?id=65007
17370
17371         Reviewed by Noam Rosenthal.
17372
17373         Add menu item and command line option to turn on disk caching in
17374         QtTestBrowser. Simple QNetworkDiskCache added to the page's
17375         QNetworkAccssManager.
17376
17377         * QtTestBrowser/launcherwindow.cpp:
17378         (LauncherWindow::initializeView):
17379         (LauncherWindow::createChrome):
17380         (LauncherWindow::setDiskCache):
17381         * QtTestBrowser/launcherwindow.h:
17382         (WindowOptions::WindowOptions):
17383         * QtTestBrowser/main.cpp:
17384         (LauncherApplication::handleUserOptions):
17385
17386 2011-07-26  Sadrul Habib Chowdhury  <sadrul@chromium.org>
17387
17388         Add support for download='filename' attribute in anchors.
17389         https://bugs.webkit.org/show_bug.cgi?id=64580
17390
17391         Reviewed by Adam Barth.
17392
17393         * DumpRenderTree/chromium/WebViewHost.cpp:
17394         (WebViewHost::loadURLExternally):
17395         * DumpRenderTree/chromium/WebViewHost.h:
17396
17397 2011-07-26  Dan Bernstein  <mitz@apple.com>
17398
17399         Rubber-stamped by Anders Carlsson.
17400
17401         Update DumpRenderTree and API tests for the pictograph font family preference.
17402
17403         * DumpRenderTree/mac/DumpRenderTree.mm:
17404         (resetDefaultsToConsistentValues):
17405         * DumpRenderTree/win/DumpRenderTree.cpp:
17406         (resetDefaultsToConsistentValues):
17407         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
17408         (TestWebKitAPI::TEST):
17409         * WebKitTestRunner/TestController.cpp:
17410         (WTR::TestController::resetStateToConsistentValues):
17411
17412 2011-07-26  Dimitri Glazkov  <dglazkov@chromium.org>
17413
17414         Replace SpecificityCalculator with TestConfiguration-driven logic.
17415         https://bugs.webkit.org/show_bug.cgi?id=65206
17416
17417         Use our newly-acquired ability to expand modifiers into a set of
17418         matching TestConfiguration instances to calculate specificity and
17419         determine whether expectation applies to a given test configuration.
17420
17421         Also, store bug modifier on TestExpectationsLine.
17422
17423         Reviewed by Adam Barth.
17424
17425         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added a way to report unknown modifier errors.
17426         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added a test for reporting errors.
17427         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Replaced the logic of calculating specificity with
17428             test configuration matching, removed a bunch of code.
17429         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Removed a bunch of tests that aren't useful anymore.
17430         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py: Fixed an error where an invalid TestConfiguration instance
17431             was created. 'default' is not a valid build type.
17432         * Scripts/webkitpy/layout_tests/port/test.py: Ditto. '' is not a valid version.
17433
17434 2011-07-26  Adam Barth  <abarth@webkit.org>
17435
17436         Increase the information density in garden-o-matic
17437         https://bugs.webkit.org/show_bug.cgi?id=65141
17438
17439         Reviewed by Dimitri Glazkov.
17440
17441         This patch changes lets us show each revision on a single line,
17442         allowing use to show an entire day's worth of commits at once.
17443
17444         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
17445         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/model.js:
17446         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
17447
17448 2011-07-26  Dimitri Glazkov  <dglazkov@chromium.org>
17449
17450         TestConfigurationConverter should not barf on invalid specifiers.
17451         https://bugs.webkit.org/show_bug.cgi?id=65194
17452
17453         Reviewed by Adam Barth.
17454
17455         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Refactored TestConfigurationConverter.to_config_set a bit,
17456             made it return empty set whenever an invalid specifier is encountered.
17457         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added corresponding tests.
17458
17459 2011-07-26  Dimitri Glazkov  <dglazkov@chromium.org>
17460
17461         Introduce a way to convert between specifier lists and TestConfiguration sets.
17462         https://bugs.webkit.org/show_bug.cgi?id=65094
17463
17464         Add TestConfigurationConverter, a facility to convert:
17465         1) a set of specifiers into a set of TestConfiguration instances;
17466         2) a set of TestConfiguration instances into a list of sets of specifiers.
17467
17468         Reviewed by Adam Barth.
17469
17470         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added TestConfigurationConverter.
17471         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added tests for it.
17472
17473 2011-07-26  Alexandru Chiculita  <achicu@adobe.com>
17474
17475         CSS Regions build bot should build with "--css-regions" and "--css-exclusion" flags
17476         https://bugs.webkit.org/show_bug.cgi?id=65171
17477
17478         Added "features" parameter in config.json for "builder" objects. The parameter is optional and must be an array. All the items
17479         of the array will be passed to the build-webkit script. The items will automatically be prefixed with "--".
17480
17481         Reviewed by Adam Roben.
17482
17483         * BuildSlaveSupport/build.webkit.org-config/config.json:
17484         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
17485
17486 2011-07-26  Chris Rogers  <crogers@google.com>
17487
17488         Avoid outputting #EOF twice for web audio tests in chromium DRT
17489         https://bugs.webkit.org/show_bug.cgi?id=65201
17490
17491         Reviewed by Dirk Pranke.
17492
17493         * DumpRenderTree/chromium/TestShell.cpp:
17494         (TestShell::dump):
17495
17496 2011-07-26  Adam Barth  <abarth@webkit.org>
17497
17498         Windows bots on build.chromium.org can't figure out the head_svn_revision
17499         https://bugs.webkit.org/show_bug.cgi?id=65156
17500
17501         Reviewed by Tony Chang.
17502
17503         This patch is a bit of a guess as to the issue based on the
17504         investigation in https://bugs.webkit.org/show_bug.cgi?id=65148.  I'm
17505         not sure whether it will fix the issue, but it seems worth a try.
17506
17507         * Scripts/webkitpy/common/checkout/scm/svn.py:
17508
17509 2011-07-26  Carlos Garcia Campos  <cgarcia@igalia.com>
17510
17511         Unreviewed. Fix WebKitGtk build with GTK+ 3.x after r91707.
17512
17513         * DumpRenderTree/gtk/SelfScrollingWebKitWebView.cpp:
17514         (getPreferredSizeMethod):
17515
17516 2011-07-26  MORITA Hajime  <morrita@google.com>
17517
17518         [Chromium][DRT] MockSpellCheck.cpp should be clang clean
17519         https://bugs.webkit.org/show_bug.cgi?id=65166
17520
17521         Reviewed by Kent Tamura.
17522
17523         * DumpRenderTree/chromium/MockSpellCheck.cpp:
17524         (MockSpellCheck::spellCheckWord): -Wsign-compare warning.
17525
17526 2011-07-25  Patrick Gansterer  <paroga@webkit.org>
17527
17528         Ignore line endings in ChangeLogTests
17529         https://bugs.webkit.org/show_bug.cgi?id=64956
17530
17531         Reviewed by Adam Barth.
17532
17533         Split the lines of the ChangeLog text before comparing them.
17534         This fixes 3 tests on windows, since only the content importent.
17535
17536         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
17537         * Scripts/webkitpy/tool/steps/preparechangelog_unittest.py:
17538
17539 2011-07-25  Adam Barth  <abarth@webkit.org>
17540
17541         Add some debug logging to NRWT to diagnose why the windows bots can't determine their SVN revision
17542         https://bugs.webkit.org/show_bug.cgi?id=65148
17543
17544         Reviewed by Tony Chang.
17545
17546         * Scripts/webkitpy/layout_tests/controllers/manager.py:
17547
17548 2011-07-25  Ojan Vafai  <ojan@chromium.org>
17549
17550         Tweak whitespace wrapping to improve readability.
17551         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17552
17553 2011-07-25  Ojan Vafai  <ojan@chromium.org>
17554
17555         Minor tweak to wrap label elements as atomic units.
17556
17557         * TestResultServer/static-dashboards/dashboard_base.js:
17558         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17559
17560 2011-07-25  Ojan Vafai  <ojan@chromium.org>
17561
17562         always show flakiness details on the builder page
17563         https://bugs.webkit.org/show_bug.cgi?id=65139
17564
17565         Reviewed by Tony Chang.
17566
17567         In order to simplify the UI, we hide the flakiness details by default,
17568         but there's no actual use-case for this page without the flakiness details.
17569         So get rid of that option and always show the flakiness details.
17570
17571         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17572
17573 2011-07-25  Ojan Vafai  <ojan@chromium.org>
17574
17575         simplify the UI of the header above the test list table
17576         https://bugs.webkit.org/show_bug.cgi?id=65136
17577
17578         Reviewed by Adam Barth.
17579
17580         Use checkboxes instead of links since this is really about which tests to include.
17581         Ends up using less text and put the checkboxes to the right of the header instead of below it.
17582
17583         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17584         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
17585
17586 2011-07-25  Chris Rogers  <crogers@google.com>
17587
17588         Update new-run-webkit-tests to accomodate binary audio data
17589         https://bugs.webkit.org/show_bug.cgi?id=65135
17590
17591         Reviewed by Dirk Pranke.
17592
17593         * Scripts/webkitpy/layout_tests/port/chromium.py:
17594
17595 2011-07-25  Chris Rogers  <crogers@google.com>
17596
17597         Update Chromium DRT to output binary (instead of base64-encoded) data for web audio testing
17598         https://bugs.webkit.org/show_bug.cgi?id=65039
17599
17600         Reviewed by Tony Chang.
17601
17602         * DumpRenderTree/chromium/LayoutTestController.cpp:
17603         (LayoutTestController::LayoutTestController):
17604         (LayoutTestController::setAudioData):
17605         * DumpRenderTree/chromium/LayoutTestController.h:
17606         (LayoutTestController::audioData):
17607         * DumpRenderTree/chromium/TestEventPrinter.cpp:
17608         (DRTPrinter::handleAudioHeader):
17609         (TestShellPrinter::handleAudioHeader):
17610         * DumpRenderTree/chromium/TestShell.cpp:
17611         (TestShell::dump):
17612
17613 2011-06-07  Martin Robinson  <mrobinson@igalia.com>
17614
17615         Reviewed by Gustavo Noronha Silva.
17616
17617         [GTK] Use WebCore scrollbars for the DRT instead of GtkScrolledWindow
17618         https://bugs.webkit.org/show_bug.cgi?id=62252
17619
17620         Add a new subclass of WebKitWebView that overrides the size request to
17621         return 1 pixel by 1 pixel. This ensure that it operates properly when
17622         not packed into a GtkScrolledWindow. Also construct the WebKitWebView
17623         with self-scrolling set to true.
17624
17625         * DumpRenderTree/gtk/DumpRenderTree.cpp:
17626         (webInspectorInspectWebView): Use the new SelfScrollingWebKitWebview class here.
17627         (createWebView): Ditto.
17628         (main): Pack the widget into a GtkBox instead of a GtkScrolledWindow.
17629         * DumpRenderTree/gtk/SelfScrollingWebKitWebView.cpp: Added.
17630         (self_scrolling_webkit_web_view_class_init):
17631         (self_scrolling_webkit_web_view_init):
17632         (self_scrolling_webkit_web_view_new):
17633         (sizeRequestMethod):
17634         (getPreferredSizeMethod):
17635         * DumpRenderTree/gtk/SelfScrollingWebKitWebView.h: Added.
17636         * GNUmakefile.am: Added SelfScrollingWebKitWebView files to the source list.
17637
17638 2011-07-25  Ojan Vafai  <ojan@chromium.org>
17639
17640         by default hide tests that have passes for all recorded runs
17641         https://bugs.webkit.org/show_bug.cgi?id=65127
17642
17643         Reviewed by Adam Barth.
17644
17645         This is part of simplifying the default views of the dashboard to
17646         make it show less information for the common use cases.
17647
17648         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17649         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
17650
17651 2011-07-25  Ojan Vafai  <ojan@chromium.org>
17652
17653         stop generating expectations.json now that it's unused
17654         https://bugs.webkit.org/show_bug.cgi?id=65130
17655
17656         Reviewed by Adam Barth.
17657
17658         * Scripts/webkitpy/layout_tests/controllers/manager.py:
17659         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
17660         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
17661         * TestResultServer/handlers/menu.py:
17662         * TestResultServer/static-dashboards/dashboard_base.js:
17663         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17664
17665 2011-07-25  Ojan Vafai  <ojan@chromium.org>
17666
17667         don't show builders in the flakiness dashboard that have stopped running a given test suite
17668         https://bugs.webkit.org/show_bug.cgi?id=65126
17669
17670         Reviewed by Mihai Parparita.
17671
17672         * TestResultServer/static-dashboards/dashboard_base.js:
17673
17674 2011-07-25  Adam Barth  <abarth@webkit.org>
17675
17676         Overhaul garden-o-matic frontend to integrate revision history and failures
17677         https://bugs.webkit.org/show_bug.cgi?id=65089
17678
17679         Reviewed by Dimitri Glazkov.
17680
17681         This is a major overhaul of the garden-o-matic frontend.  The new UI
17682         integrates the revision history information with failures.  Failures
17683         appear in context adjacent to the revisions that caused them (which are
17684         highlighted).  Each buildbot receives a collumn to the left, which
17685         indicates which revisions that bot has tested and whether that bot
17686         showed the indicated failures.
17687
17688         Failures are selectable with checkboxes, and can then either be
17689         examined or rebaselined.
17690
17691         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
17692         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
17693         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
17694         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
17695         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
17696         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
17697         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js:
17698         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
17699         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
17700
17701 2011-07-25  Holger Hans Peter Freyther  <zecke@selfish.org>
17702
17703         [buildbot] Add Qt Linux MIPSEL Configuration
17704         https://bugs.webkit.org/show_bug.cgi?id=65116
17705
17706         Reviewed by Adam Roben.
17707
17708         Introduce a Qt Linux MIPS EL build slave. The configuration
17709         is a build only configuration.
17710
17711         * BuildSlaveSupport/build.webkit.org-config/config.json: Add MIPS buildbot.
17712
17713 2011-07-25  Tor Arne Vestbø  <tor.arne.vestbo@nokia.com>
17714
17715         [Qt][WK2] Add multi-point touch mocking to MiniBrowser
17716         https://bugs.webkit.org/show_bug.cgi?id=64374
17717
17718         Reviewed by Benjamin Poulain.
17719
17720         Each mouse button generate a touch point. The touch point stay
17721         virtually on screen if the mouse release happened while the Ctrl
17722         modifier was pressed.
17723
17724         * MiniBrowser/qt/MiniBrowserApplication.cpp:
17725         (MiniBrowserApplication::notify):
17726         * MiniBrowser/qt/MiniBrowserApplication.h:
17727
17728 2011-07-25  Balazs Kelemen  <kbalazs@webkit.org>
17729
17730         [Qt][WK2] Use NRWT for Qt-WK2
17731         https://bugs.webkit.org/show_bug.cgi?id=65097
17732
17733         Reviewed by Andreas Kling.
17734
17735         * Scripts/run-webkit-tests:
17736         (useNewRunWebKitTests):
17737
17738 2011-07-24  Sheriff Bot  <webkit.review.bot@gmail.com>
17739
17740         Unreviewed, rolling out r91635.
17741         http://trac.webkit.org/changeset/91635
17742         https://bugs.webkit.org/show_bug.cgi?id=65083
17743
17744         It broke the minimal build (Requested by Ossy___OFFLINE on
17745         #webkit).
17746
17747         * QtTestBrowser/launcherwindow.cpp:
17748         (LauncherWindow::initializeView):
17749         (LauncherWindow::createChrome):
17750         * QtTestBrowser/launcherwindow.h:
17751         (WindowOptions::WindowOptions):
17752         * QtTestBrowser/main.cpp:
17753         (LauncherApplication::handleUserOptions):
17754
17755 2011-07-24  Adam Barth  <abarth@webkit.org>
17756
17757         Rename garden-o-matic "alert" to "infobar" and increase the width to 100%.
17758
17759         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
17760         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
17761         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
17762         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
17763         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
17764         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js:
17765
17766 2011-07-23  Dimitri Glazkov  <dglazkov@chromium.org>
17767
17768         TestConfiguration should be hashable and comparable for equality.
17769         https://bugs.webkit.org/show_bug.cgi?id=65070
17770
17771         Reviewed by Adam Barth.
17772
17773         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added __hash__ and __eq__ methods.
17774         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added unit tests.
17775
17776 2011-07-23  Dimitri Glazkov  <dglazkov@chromium.org>
17777
17778         Extract TestConfiguration from base and write a few unit tests to capture current behavior.
17779         https://bugs.webkit.org/show_bug.cgi?id=65066
17780
17781         Reviewed by Adam Barth.
17782
17783         * Scripts/webkitpy/layout_tests/models/test_configuration.py: Added.
17784         * Scripts/webkitpy/layout_tests/models/test_configuration_unittest.py: Added.
17785         * Scripts/webkitpy/layout_tests/port/base.py: Moved code to test_configuration.py
17786
17787 2011-07-23  Alok Priyadarshi  <alokp@chromium.org>
17788
17789         Switching off acceleration for small canvas broke gpu tests
17790         https://bugs.webkit.org/show_bug.cgi?id=65053
17791
17792         Reviewed by Stephen White.
17793
17794         * DumpRenderTree/chromium/WebPreferences.cpp:
17795         (WebPreferences::reset):
17796         (WebPreferences::applyTo):
17797         * DumpRenderTree/chromium/WebPreferences.h:
17798
17799 2011-07-23  Keith Kyzivat  <keith.kyzivat@nokia.com>
17800
17801         [Qt] Add option to turn on disk caching in QtTestBrowser
17802         https://bugs.webkit.org/show_bug.cgi?id=65007
17803
17804         Reviewed by Noam Rosenthal.
17805
17806         Add menu item and command line option to turn on disk caching in
17807         QtTestBrowser. Simple QNetworkDiskCache added to the page's
17808         QNetworkAccssManager.
17809
17810         * QtTestBrowser/launcherwindow.cpp:
17811         (LauncherWindow::initializeView):
17812         (LauncherWindow::createChrome):
17813         (LauncherWindow::setDiskCache):
17814         * QtTestBrowser/launcherwindow.h:
17815         (WindowOptions::WindowOptions):
17816         * QtTestBrowser/main.cpp:
17817         (LauncherApplication::handleUserOptions):
17818
17819 2011-07-22  Ojan Vafai  <ojan@chromium.org>
17820
17821         Don't match 'DEBUG' when trying to filter BUG modifiers.
17822         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17823         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
17824
17825 2011-07-22  Ojan Vafai  <ojan@chromium.org>
17826
17827         Fix issue loading gtests since there's no g_expectations.
17828         Also, make filing a bug file a webkit.org bug, not a crbug since
17829         the link is only shown for layout tests.
17830         * TestResultServer/static-dashboards/dashboard_base.js:
17831         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17832
17833 2011-07-22  Ojan Vafai  <ojan@chromium.org>
17834
17835         pull test expectations into the dashboard from wvn.webkit.org
17836         https://bugs.webkit.org/show_bug.cgi?id=65054
17837
17838         Reviewed by Adam Barth.
17839
17840         Now that svn.webkit.org supports cors headers, we can just grab the txt
17841         file from there instead of generating a JSON file that we upload
17842         to appengine. We'll stop generating the JSON file in a followup patch.
17843
17844         * TestResultServer/static-dashboards/dashboard_base.js:
17845         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17846         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
17847
17848 2011-07-22  Adam Barth  <abarth@webkit.org>
17849
17850         Fix typo so the tool works on a case-sensitive file system.
17851
17852         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
17853
17854 2011-07-22  Adam Barth  <abarth@webkit.org>
17855
17856         garden-o-matic should know what patches are landing
17857         https://bugs.webkit.org/show_bug.cgi?id=64978
17858
17859         Reviewed by Dimitri Glazkov.
17860
17861         This patch adds a display of the recently committed patches to
17862         garden-o-matic.  The UI needs iteration, but this patch establishes the
17863         plumbing.
17864
17865         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
17866             - Teach trac how to extract more information from the commit log.
17867         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
17868             - Add a data attribute for storing the revision.
17869         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html:
17870             - Add a container for the recent commit history.
17871         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css:
17872             - CSS for displaying non-ugly commit history.
17873         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js:
17874             - Wire up trac to the UI container.
17875         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js:
17876             - A basic, compact display for a list of commits.
17877
17878 2011-07-22  Ojan Vafai  <ojan@chromium.org>
17879
17880         Fix bug when switching from layout-tests to another test-type if
17881         ToT webkit.org is the selected group.
17882         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17883
17884 2011-07-22  Ojan Vafai  <ojan@chromium.org>
17885
17886         Add back in line accidentally deleted in http://trac.webkit.org/changeset/91612.
17887         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17888
17889 2011-07-22  Ojan Vafai  <ojan@chromium.org>
17890
17891         simplify the flakiness dashboard to only show modifiers/expectations for the current builder
17892         https://bugs.webkit.org/show_bug.cgi?id=65049
17893
17894         Reviewed by Tony Chang.
17895
17896         This communicates less information, but probably makes the dashboard more usable.
17897         Also make it show that bug lines that don't point to bug numbers still
17898         show up in the bugs column (e.g. BUG_TONY) and stop showing the extra/missing
17899         columns in any view. The data is still used, but only in the view that updates
17900         the test_expectations.txt file.
17901
17902         * TestResultServer/static-dashboards/dashboard_base.js:
17903         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17904         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
17905
17906 2011-07-21  Ojan Vafai  <ojan@chromium.org>
17907
17908         hook in the rest of the bots to the dashboard now they run NRWT
17909         https://bugs.webkit.org/show_bug.cgi?id=64985
17910
17911         Reviewed by Adam Barth.
17912
17913         * TestResultServer/static-dashboards/builders.js:
17914
17915 2011-07-22  Alexey Proskuryakov  <ap@apple.com>
17916
17917         Adding Tim Horton to contributor list for Bugzilla auto-completion. He is CC'ed quite frequently.
17918
17919         Unreviewed.
17920
17921         * Scripts/webkitpy/common/config/committers.py:
17922
17923 2011-07-22  Michal Pakula vel Rutka  <m.pakula@samsung.com>
17924
17925         [EFL] ewk_frame_hit_test_new enchancement
17926         https://bugs.webkit.org/show_bug.cgi?id=64260
17927
17928         Apply changes done in ewk_frame_hit_test_new in EWebLauncher:
17929         Replacing Ewk_Hit_Test_Result_Context structure 'flags'
17930         in hit test output to by enum 'context'.
17931
17932         Reviewed by Antonio Gomes.
17933
17934         * EWebLauncher/main.c:
17935         (on_key_down):
17936
17937 2011-07-21  Adam Roben  <aroben@apple.com>
17938
17939         Fix typo in TestFailures
17940
17941         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
17942
17943 2011-07-21  Ojan Vafai  <ojan@chromium.org>
17944
17945         simplify gtest display now that we strip modifiers from the JSON
17946         https://bugs.webkit.org/show_bug.cgi?id=64990
17947
17948         Reviewed by Adam Barth.
17949
17950         Now that names are normalized, we can remove all the code that
17951         handles gtest name changes (e.g. for adding modifiers like DISABLED_).
17952         Instead, if you try to list a test with a modifier in it, we need
17953         to strip the modifier so we get the normalized value.
17954
17955         We also get rid of the concept of extra/missing expectations for gtests.
17956         In a patch soon, we'll stop showing extra/missing expectations from the UI
17957         entirely and only leave it for the special updating test_expectations.txt
17958         view of the dashboard, which doesn't apply to gtests.
17959
17960         * TestResultServer/static-dashboards/flakiness_dashboard.html:
17961         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
17962
17963 2011-07-21  Adam Barth  <abarth@webkit.org>
17964
17965         Refactor Trac.js for use in garden-o-matic
17966         https://bugs.webkit.org/show_bug.cgi?id=64998
17967
17968         Reviewed by Adam Roben.
17969
17970         This patch refactors Trac.js to match the architecture and style of
17971         garden-o-matic.  I've also factored AsynchronousCache out of Trac (and
17972         moved it to base) because it was redundant with some other code we
17973         already had.
17974
17975         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js:
17976         (FailingTestsBugForm):
17977         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js:
17978         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
17979         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
17980         (ViewController.prototype._displayBuilder.start):
17981         (ViewController.prototype._displayBuilder):
17982         (ViewController.prototype._domForRegressionRange.trac.getCommitDataForRevisionRange):
17983         (ViewController.prototype._domForRegressionRange):
17984         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js:
17985         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js:
17986         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html:
17987         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js:
17988         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
17989
17990 2011-07-21  Ojan Vafai  <ojan@chromium.org>
17991
17992         Fix issue throwing a 500 to the bots. We were modifying
17993         the dictionary during an iteration over it's keys and would get
17994         RuntimeError: dictionary changed size during iteration
17995
17996         * TestResultServer/model/jsonresults.py:
17997
17998 2011-07-21  Adam Barth  <abarth@webkit.org>
17999
18000         ORWT fails to consume some NRWT arguments
18001         https://bugs.webkit.org/show_bug.cgi?id=64988
18002
18003         Reviewed by Adam Roben.
18004
18005         * Scripts/old-run-webkit-tests:
18006
18007 2011-07-21  Adam Barth  <abarth@webkit.org>
18008
18009         Merge TestFailures and garden-o-matic unit tests harness
18010         https://bugs.webkit.org/show_bug.cgi?id=64984
18011
18012         Reviewed by Adam Roben.
18013
18014         One unit testing framework is better than two.
18015
18016         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-more-unittests.html: Removed.
18017         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
18018
18019 2011-07-21  Adam Barth  <abarth@webkit.org>
18020
18021         Rubber-stamped by Adam Roben.
18022
18023         Move garden-o-matic files into the same directory as TestFailures to
18024         make it easier to integrate the codebases.
18025
18026         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/base.js.
18027         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/base_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js.
18028         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js.
18029         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/checkout_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js.
18030         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/config.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/config.js.
18031         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-green.png: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-green.png.
18032         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/favicon-red.png: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-red.png.
18033         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/garden-o-matic.html: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/index.html.
18034         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.css: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/main.css.
18035         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/main.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/main.js.
18036         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/partytime.gif: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/partytime.gif.
18037         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/results.js.
18038         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/results_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js.
18039         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-more-unittests.html: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html.
18040         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js.
18041         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ui_unittests.js: Copied from Tools/Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js.
18042         * Scripts/webkitpy/tool/commands/abstractlocalservercommand.py:
18043         * Scripts/webkitpy/tool/commands/gardenomatic.py:
18044         * Scripts/webkitpy/tool/servers/data/gardeningserver: Removed.
18045         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js: Removed.
18046         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js: Removed.
18047         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js: Removed.
18048         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js: Removed.
18049         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js: Removed.
18050         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-green.png: Removed.
18051         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-red.png: Removed.
18052         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Removed.
18053         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css: Removed.
18054         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Removed.
18055         * Scripts/webkitpy/tool/servers/data/gardeningserver/partytime.gif: Removed.
18056         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js: Removed.
18057         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js: Removed.
18058         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html: Removed.
18059         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js: Removed.
18060         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js: Removed.
18061         * Scripts/webkitpy/tool/servers/gardeningserver.py:
18062
18063 2011-07-19  Ojan Vafai  <ojan@chromium.org>
18064
18065         fix links on the flakiness dashboard
18066         https://bugs.webkit.org/show_bug.cgi?id=64832
18067
18068         Reviewed by Tony Chang.
18069
18070         Make them actual anchor tags and linkify the test name instead of
18071         adding an extra link. Also simplify the update view since we now
18072         show the linkified test name.
18073
18074         * TestResultServer/static-dashboards/flakiness_dashboard.html:
18075
18076 2011-07-21  Adam Barth  <abarth@webkit.org>
18077
18078         webkit.org builders should upload results to test-results AppEngine instance
18079         https://bugs.webkit.org/show_bug.cgi?id=64979
18080
18081         Reviewed by Adam Roben.
18082
18083         The master just needs to pass some flags to NRWT.  This patch also lets
18084         ORWT consume the flags (ignoring them) so we can pass the flags
18085         unconditionally.
18086
18087         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
18088         * Scripts/old-run-webkit-tests:
18089
18090 2011-07-21  Ojan Vafai  <ojan@chromium.org>
18091
18092         normalize gtest names in the result json
18093         https://bugs.webkit.org/show_bug.cgi?id=64971
18094
18095         Reviewed by Eric Seidel.
18096
18097         This will make the resulting JSON smaller and make the
18098         flakiness dashboard much more sane (e.g. we won't need to
18099         show you the FLAKY_ version and the normal version).
18100
18101         * TestResultServer/model/jsonresults.py:
18102         * TestResultServer/model/jsonresults_unittest.py:
18103
18104 2011-07-21  Eric Seidel  <eric@webkit.org>
18105
18106         Unreviewed.  Reenable ORWT for the --leaks bot.
18107         Something is still wrong with the leaks summarizing
18108         but I'm about to head out on vacation for 2 weeks,
18109         so it's best to just revert back to ORWT until I'm back.
18110
18111         * Scripts/run-webkit-tests:
18112         (usingLeaks):
18113         (useNewRunWebKitTests):
18114
18115 2011-07-21  Ojan Vafai  <ojan@chromium.org>
18116
18117         cleanup jsonresults.py style in preparation for making some changes
18118         https://bugs.webkit.org/show_bug.cgi?id=64968
18119
18120         Reviewed by Adam Barth.
18121
18122         No code changes. Unittest still passes. Mostly just removing
18123         useless comments and 80 character wrapping.
18124
18125         * TestResultServer/model/jsonresults.py:
18126         * TestResultServer/model/jsonresults_unittest.py:
18127
18128 2011-07-21  Adam Barth  <abarth@webkit.org>
18129
18130         Add missing column header.
18131
18132         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
18133
18134 2011-07-21  Eric Seidel  <eric@webkit.org>
18135
18136         Unreviewed.  Just fixing (and unittesting) a previous typo.
18137
18138         Fix typo in print_leaks_summary regexp which was causing
18139         leaks summarizing to fail on the --leaks bot.
18140
18141         * Scripts/webkitpy/layout_tests/port/mac.py:
18142         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
18143
18144 2011-07-21  Adam Barth  <abarth@webkit.org>
18145
18146         garden-o-matic should be able to rebaseline many tests at once
18147         https://bugs.webkit.org/show_bug.cgi?id=64963
18148
18149         Reviewed by Dimitri Glazkov.
18150
18151         This patch contains some UI for rebaselining tests in bulk.  I don't
18152         expect this UI to be the final UI, but the underlying infrastructure
18153         will be useful.
18154
18155         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js:
18156         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
18157         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
18158         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
18159         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
18160
18161 2011-07-21  Adam Barth  <abarth@webkit.org>
18162
18163         garden-o-matic has trouble when the same test fails in different ways on diffrent bots
18164         https://bugs.webkit.org/show_bug.cgi?id=64830
18165
18166         Unreviewed.  (Missing the patch is blocking me from fixing the tree.)
18167
18168         This patch is just some minor bug fixes.  One of the bugs results from
18169         the sadness of having two representations (a list and a space-separated
18170         string) for the set of failure types.
18171
18172         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
18173         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
18174         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
18175         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
18176
18177 2011-07-20  Ryosuke Niwa  <rniwa@webkit.org>
18178
18179         Buildbot marks a nrwt bot red when tests are missing results
18180         https://bugs.webkit.org/show_bug.cgi?id=64812
18181
18182         Reviewed by Adam Barth.
18183
18184         The bug was caused by multiple expressions matching on the single output.
18185         Fixed it by exiting the loop as soon as one expression matches.
18186
18187         Because the regular expression for 'failures' is most general,
18188         moved it to the end of the list to avoid it catching other cases.
18189
18190         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
18191
18192 2011-07-20  Chang Shu  <cshu@webkit.org>
18193
18194         Adding myself to the reviewers list; No review needed.
18195
18196         * Scripts/webkitpy/common/config/committers.py:
18197
18198 2011-07-20  Eric Seidel  <eric@webkit.org>
18199
18200         new-run-webkit-tests --leaks does not restart DumpRenderTree often enough
18201         https://bugs.webkit.org/show_bug.cgi?id=64912
18202
18203         Reviewed by Adam Barth.
18204
18205         The bug was that set_default_option only works if the option's value
18206         is "None".  So I fixed --batch-size to default to None instead of 0
18207         (which kinda makes sense).
18208
18209         * Scripts/webkitpy/layout_tests/controllers/worker.py:
18210         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
18211
18212 2011-07-20  Eric Seidel  <eric@webkit.org>
18213
18214         Make kill-old-processes kill httpd on mac
18215         https://bugs.webkit.org/show_bug.cgi?id=64898
18216
18217         Reviewed by Adam Barth.
18218
18219         * BuildSlaveSupport/kill-old-processes:
18220          - The line endings seem to have flipped again,
18221            Thus making this diff gigantic.  The only line
18222            changed is the "httpd" line.  I suspect we'll need
18223            to set the line ending style from an svn checkout.
18224
18225 2011-07-20  Ojan Vafai  <ojan@chromium.org>
18226
18227         show a list of average test times in the treemap
18228         https://bugs.webkit.org/show_bug.cgi?id=64899
18229
18230         Reviewed by Adam Roben.
18231
18232         For now, you click a link and it replaces the treemap
18233         with a reverse sorted list of directories based on
18234         average test runtime within that directory.
18235
18236         To cut some of the noise, we don't show directories
18237         that only have one test in them or where the average
18238         runtime is less than 100ms.
18239
18240         * TestResultServer/static-dashboards/treemap.html:
18241
18242 2011-07-20  Eric Seidel  <eric@webkit.org>
18243
18244         Teach build.webkit.org how to identify leaks in NRWT output
18245         https://bugs.webkit.org/show_bug.cgi?id=64891
18246
18247         Reviewed by Adam Barth.
18248
18249         This was the last remaining piece blocking turning on
18250         NRWT for the leaks bot, so I did that in this change too.
18251         If for some reason we see trouble on the bots, I'll
18252         revert that part of this change.
18253
18254         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
18255         * Scripts/run-webkit-tests:
18256
18257 2011-07-20  Ojan Vafai  <ojan@chromium.org>
18258
18259         make the treemap display more human friendly
18260         https://bugs.webkit.org/show_bug.cgi?id=64896
18261
18262         Reviewed by Tony Chang.
18263
18264         * TestResultServer/static-dashboards/treemap.html:
18265
18266 2011-07-20  Eric Seidel  <eric@webkit.org>
18267
18268         Move HttpLock to using a FileSystem object
18269         https://bugs.webkit.org/show_bug.cgi?id=64885
18270
18271         Reviewed by Adam Barth.
18272
18273         I made a typo in my previous change, taking the value
18274         of read_text_file(pid_file) and setting it to lock_pid_file
18275         instead of current_pid.  Fixed now. :)
18276
18277         In order to test my new change I had to overhaul the unittests
18278         for this class to create a separate set of tests which work off
18279         of Mock objects instead of the real filesystem.
18280
18281         Since Executive doesn't yet wrap os.getpid() I added a FIXME
18282         in several places where we're currently calling os.getpid().  I
18283         felt adding Executive.getpid was outside of the scope of this change
18284         but once it exists some of this code will be much simpler to mock.
18285
18286         * Scripts/webkitpy/layout_tests/port/http_lock.py:
18287         * Scripts/webkitpy/layout_tests/port/http_lock_unittest.py:
18288         * Scripts/webkitpy/tool/mocktool.py:
18289
18290 2011-07-20  Adam Roben  <aroben@apple.com>
18291
18292         Fix typo in TestFailures's Bugzilla constants
18293
18294         This was preventing the "Version" field from being pre-filled when filing new bugs.
18295
18296         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBugzilla.js:
18297
18298 2011-07-20  Eric Seidel  <eric@webkit.org>
18299
18300         Move HttpLock to using a FileSystem object
18301         https://bugs.webkit.org/show_bug.cgi?id=64885
18302
18303         Reviewed by Adam Barth.
18304
18305         There is no functional change here.  Just moving off
18306         of direct os.foo calls to their filesystem equivalent.
18307
18308         This makes it possible to now unittest HttpLock using a
18309         MockFileSystem instead of needing to talk to the filesystem directly.
18310
18311         This also will have changed any port test cases which were using a
18312         mock file system to now have an HttpLock based on a mock filesystem
18313         instead.  I saw no change in the unittests so it's possible
18314         we're not testing that configuration (or more likely that it "just works").
18315
18316         * Scripts/webkitpy/layout_tests/port/base.py:
18317         * Scripts/webkitpy/layout_tests/port/http_lock.py:
18318         * Scripts/webkitpy/layout_tests/port/http_lock_unittest.py:
18319
18320 2011-07-20  Adam Barth  <abarth@webkit.org>
18321
18322         NRWT results uses too much space per run
18323         https://bugs.webkit.org/show_bug.cgi?id=64845
18324
18325         Reviewed by Ojan Vafai.
18326
18327         Removing these files saves about 750K per build.
18328
18329         * Scripts/webkitpy/layout_tests/controllers/manager.py:
18330
18331 2011-07-20  Zan Dobersek  <zandobersek@gmail.com>
18332
18333         [Gtk] [NRWT] Xvfb produces a lot of stderr output
18334         https://bugs.webkit.org/show_bug.cgi?id=64669
18335
18336         Reviewed by Philippe Normand.
18337
18338         Redirect the stderr output of the driver's Xvfb process
18339         to /dev/null to suppress unneeded output.
18340
18341         * Scripts/webkitpy/layout_tests/port/gtk.py:
18342
18343 2011-07-19  Eric Seidel  <eric@webkit.org>
18344
18345         parse-malloc-history always exits 1, causing NRWT to fail
18346         https://bugs.webkit.org/show_bug.cgi?id=64835
18347
18348         Reviewed by Adam Barth.
18349
18350         ORWT just never checked the return code. :)
18351
18352         * Scripts/parse-malloc-history:
18353         (main):
18354         * Scripts/webkitpy/layout_tests/port/mac.py:
18355
18356 2011-07-19  Kevin Ollivier  <kevino@theolliviers.com>
18357
18358         [wx] Unreviewed build fix, don't add the debug prefix as 2.9 no longer uses one.
18359
18360         * waf/build/wxpresets.py:
18361
18362 2011-07-19  Eric Seidel  <eric@webkit.org>
18363
18364         new-run-webkit-tests should only enable MallocStackLogging for DRT
18365         https://bugs.webkit.org/show_bug.cgi?id=64792
18366
18367         Reviewed by Adam Barth.
18368
18369         The previous code would enable it for all servers launched
18370         by the port, which included Apache, the python websocket server
18371         as well as ImageDiff.  Now only DumpRenderTree will have
18372         MallocStackLogging enabled or the GuardMalloc library injected.
18373
18374         I also cleaned up the websocket_server code to use filesystem
18375         while I was in it.
18376
18377         I also made DRT restart every 1000 tests when running with
18378         --leaks enabled.  I believe this made the --leaks run slightly
18379         faster, but it still takes over an hour on my machine. :(
18380
18381         * Scripts/webkitpy/layout_tests/controllers/worker.py:
18382         * Scripts/webkitpy/layout_tests/port/base.py:
18383         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
18384         * Scripts/webkitpy/layout_tests/port/gtk.py:
18385         * Scripts/webkitpy/layout_tests/port/mac.py:
18386         * Scripts/webkitpy/layout_tests/port/qt.py:
18387         * Scripts/webkitpy/layout_tests/port/server_process.py:
18388         * Scripts/webkitpy/layout_tests/port/webkit.py:
18389         * Scripts/webkitpy/layout_tests/servers/http_server.py:
18390         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
18391
18392 2011-07-19  Ojan Vafai  <ojan@chromium.org>
18393
18394         remove the concept of platform fallbacks
18395         https://bugs.webkit.org/show_bug.cgi?id=64829
18396
18397         Reviewed by Adam Barth.
18398
18399         In the process, wrote tests for this code and fixed a pretty major
18400         bug (now covered by the test).
18401
18402         * TestResultServer/static-dashboards/flakiness_dashboard.html:
18403         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
18404
18405 2011-07-19  Ojan Vafai  <ojan@chromium.org>
18406
18407         cleanup some of the PLATFORM logic in the flakiness dashboard
18408         https://bugs.webkit.org/show_bug.cgi?id=64821
18409
18410         Reviewed by Adam Barth.
18411
18412         * TestResultServer/static-dashboards/flakiness_dashboard.html:
18413         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
18414
18415 2011-07-19  Ojan Vafai  <ojan@chromium.org>
18416
18417         cleanup some of the PLATFORM logic in the flakiness dashboard
18418         https://bugs.webkit.org/show_bug.cgi?id=64821
18419
18420         Reviewed by Adam Barth.
18421
18422         * TestResultServer/static-dashboards/flakiness_dashboard.html:
18423         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
18424
18425 2011-07-19  Dimitri Glazkov  <dglazkov@chromium.org>
18426
18427         Store line number on TestExpectationLine.
18428         https://bugs.webkit.org/show_bug.cgi?id=64800
18429
18430         Reviewed by Adam Barth.
18431
18432         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Refactored to store line number on TestExpectationLine instances.
18433
18434 2011-07-19  Ojan Vafai  <ojan@chromium.org>
18435
18436         fix flakiness dashboard for XP bots with XP in the name instead of WIN
18437         https://bugs.webkit.org/show_bug.cgi?id=64826
18438
18439         Reviewed by Mihai Parparita.
18440
18441         * TestResultServer/static-dashboards/flakiness_dashboard.html:
18442         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
18443
18444 2011-07-19  Cary Clark  <caryclark@google.com>
18445
18446         add myself to committers.py
18447         https://bugs.webkit.org/show_bug.cgi?id=64816
18448
18449         No review needed.
18450
18451         * Scripts/webkitpy/common/config/committers.py:
18452
18453 2011-07-19  Daniel Bates  <dbates@webkit.org>
18454
18455         Rename variable isEmptyTextPatch in svn-{apply, unapply}
18456         https://bugs.webkit.org/show_bug.cgi?id=64648
18457
18458         Reviewed by Adam Roben.
18459
18460         Rename the variable isEmptyTextPatch to hasTextChunks, which is
18461         more descriptive and makes its usage read well in control statements.
18462
18463         * Scripts/svn-apply:
18464         (patch):
18465         * Scripts/svn-unapply:
18466         (patch):
18467
18468 2011-07-19  Adam Roben  <aroben@apple.com>
18469
18470         Make TestFailures show existing bugs and a new bug link for flaky tests
18471
18472         Fixes <http://webkit.org/b/63728> TestFailures page should make it easy to file bugs about
18473         flaky tests
18474         and
18475         <http://webkit.org/b/63830> TestFailures page doesn't show related bugs for possibly-flaky
18476         tests, but should
18477
18478         Reviewed by Sam Weinig.
18479
18480         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm.js:
18481         Added. This is what's used to file new bugs about flaky tests.
18482         (FlakyTestBugForm): Calls up to the base class, stores the arguments, and sets our
18483         title, description, and URL.
18484
18485         (FlakyTestBugForm.prototype._createBugDescription):
18486         (FlakyTestBugForm.prototype._createBugTitle):
18487         These do what they say.
18488
18489         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyTestBugForm_unittests.js:
18490         Added. Tests of the above.
18491
18492         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
18493         (#failure-history > li):
18494         (#possibly-flaky-tests > li):
18495         (#failure-history > li, #possibly-flaky-tests > li):
18496         (#possibly-flaky-tests > li > :first-child):
18497         (.flakiness-examples-list):
18498         Styling to account for the new elements.
18499
18500         (.expandable):
18501         (.expanded > .expandable):
18502         Generalized the .flakiness-example-list code to a generic .expandable class. Removed the
18503         transition properties because we can't transition to/from 'auto' (but really would like
18504         to!).
18505
18506         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
18507         (ViewController.prototype._displayBuilder): Moved code to create the FailingTestsBugForm
18508         here from _domForNewAndExistingBugs. Updated for changes to _domForPossiblyFlakyTests.
18509         (ViewController.prototype._domForNewAndExistingBugs): Changed to take a NewBugForm as an
18510         argument instead of creating one ourselves.
18511         (ViewController.prototype._domForPossiblyFlakyTests): Changed to take all analyzed builds as
18512         an argument instead of just the number of analyzed builds. Put the examples list inside a
18513         container <div>, which is also used to hold the new/existing bugs UI. Removed code to deal
18514         with animating the height of the examples list; we don't do this anymore because it's hard
18515         to make it work correctly with the asynchronous loading of existing bugs. When we populate
18516         the examples list, also set up the new/existing bug UI.
18517
18518         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
18519         FlakyTestBugForm.
18520
18521         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
18522         Pull in FlakyTestBugForm and its tests.
18523
18524 2011-07-19  Ryosuke Niwa  <rniwa@webkit.org>
18525
18526         Buildbot marks a nrwt bot red when tests are missing results
18527         https://bugs.webkit.org/show_bug.cgi?id=64812
18528
18529         Reviewed by Tony Chang.
18530
18531         Added "missing results" label for tests that are missing results.
18532
18533         buildbot Now reports shows a label "X missing results" on a run when X tests were missing
18534         expected results on that run.
18535
18536         Also rephrased "X failed" and "X were flaky" to "X failures" and "X flakes".
18537
18538         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
18539
18540 2011-07-19  Adam Roben  <aroben@apple.com>
18541
18542         Extract some of FailingTestsBugForm's code into a base class
18543
18544         Prep work for fixing <http://webkit.org/b/63728> TestFailures page should make it easy to
18545         file bugs about flaky tests
18546
18547         Reviewed by Sam Weinig.
18548
18549         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js:
18550         Moved BugzillaConstants to new WebKitBugzilla file. Moved a bunch of other code from here to
18551         TestRelatedBugForm.js.
18552
18553         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js:
18554         Moved some tests to TestRelatedBugForm_unittests.js.
18555
18556         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm.js: Added.
18557         (TestRelatedBugForm):
18558         (TestRelatedBugForm.prototype.domElement):
18559         (TestRelatedBugForm.prototype._computeOperatingSystem):
18560         (TestRelatedBugForm.prototype._computePlatform):
18561         Code came from FailingTestsBugForm.
18562
18563         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestRelatedBugForm_unittests.js:
18564         Added. Tests came from FailingTestsBugForm_unittests.js
18565
18566         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBugzilla.js:
18567         Added. Code came from FailingTestsBugForm.js.
18568
18569         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
18570         TestRelatedBugForm.js and WebKitBugzilla.js. Moved Bugzilla.js out of the list of files that
18571         need to be pulled in early for parsing reasons.
18572
18573         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
18574         Pull in TestRelatedBugForm and tests and WebKitBugzilla.
18575
18576 2011-07-19  Sam Weinig  <sam@webkit.org>
18577
18578         Remove obsolete Mac NRWT test bots
18579         https://bugs.webkit.org/show_bug.cgi?id=64806
18580
18581         Reviewed by Adam Roben.
18582
18583         Remove "Leopard Intel Release (NRWT)" and "SnowLeopard Intel Release (NRWT)"
18584         which have been disconnected for some time and have been obsoleted by
18585         the main testers using NRWT.
18586
18587         * BuildSlaveSupport/build.webkit.org-config/config.json:
18588
18589 2011-07-19  Sam Weinig  <sam@webkit.org>
18590
18591         webkit-patch rebaseline crashes on use
18592         https://bugs.webkit.org/show_bug.cgi?id=64775
18593
18594         Reviewed by Adam Roben.
18595
18596         Fix incorrectly spelled attribute name.
18597
18598         * Scripts/webkitpy/common/net/layouttestresults.py:
18599         The attribute is test_name, not filename.
18600
18601         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
18602         Add test for tests_matching_failure_types.
18603
18604 2011-07-18  Eric Seidel  <eric@webkit.org>
18605
18606         new-run-webkit-tests should support --leaks
18607         https://bugs.webkit.org/show_bug.cgi?id=63832
18608
18609         Reviewed by Dirk Pranke.
18610
18611         This may not be sufficient to actually transition over the leaks bot,
18612         but this is a huge step in the right direction.
18613
18614         I had to make parse-malloc-history understand being passed more than
18615         one file (to avoid the silly cat | nonsense in old-run-webkit-tests).
18616
18617         I removed some dead code relating to previous iterations of our crash detection.
18618
18619         I created a new class "LeakDetector" to encapsulate all this logic.
18620         Eventually we should consider pulling that class out of mac.py and
18621         sharing with other ports.  However given that ORWT has had
18622         --leaks support on Mac for almost 7 years and no other port has added
18623         it, leaves me to believe we're in no rush to move LeakDetector.
18624
18625         I've tested --leaks locally.  I suspect there are more bugs to shake out
18626         but it seems to work well enough to start.
18627
18628         I also added support for --guard-malloc, but have not tested it much.  It
18629         should be viewed as experimental at this time.
18630
18631         I also fixed various os.path uses to self._filesystem as I was reading
18632         through the various files to understand how best to fix this bug.
18633
18634         * Scripts/old-run-webkit-tests:
18635         (parseLeaksandPrintUniqueLeaks):
18636         * Scripts/parse-malloc-history:
18637         (main):
18638         * Scripts/webkitpy/common/system/crashlogs.py:
18639         * Scripts/webkitpy/layout_tests/controllers/manager.py:
18640         * Scripts/webkitpy/layout_tests/controllers/worker.py:
18641         * Scripts/webkitpy/layout_tests/port/base.py:
18642         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
18643         * Scripts/webkitpy/layout_tests/port/gtk.py:
18644         * Scripts/webkitpy/layout_tests/port/mac.py:
18645         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
18646         * Scripts/webkitpy/layout_tests/port/server_process.py:
18647         * Scripts/webkitpy/layout_tests/port/webkit.py:
18648         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
18649
18650 2011-07-19  Adam Barth  <abarth@webkit.org>
18651
18652         garden-o-matic should have a "Triage Failures" button for iterating over failures
18653         https://bugs.webkit.org/show_bug.cgi?id=64769
18654
18655         Reviewed by Eric Seidel.
18656
18657         This patch adds some global static state and refactors things a bit so
18658         that we can iterate over all the failures in the details pane.
18659
18660         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
18661             - Add a generic callback iterator to iterate through a series of
18662               callbacks. We use this to iterate through the failures we want to
18663               display the details of.
18664         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
18665             - Add a toolbar for the results summary and change the buttons to
18666               be real buttons.
18667         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
18668             - Make the detail pane 75% of the window.
18669             - Make the buttons pretty.
18670             - Allow the failure type badges to apply to the details pane as well.
18671         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
18672             - Restructure how we bring up the details pane so we can iterate
18673               through a bunch of failures.
18674             - Remove the transition between results details because it's
18675               annoying when you want to click through a bunch of failures.
18676         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
18677             - Add some more structure to the title bar for the results details
18678               so it's clearer which results we're currently displaying.
18679
18680 2011-07-19  Adam Barth  <abarth@webkit.org>
18681
18682         garden-o-matic should live update as conditions change on the buildbot
18683         https://bugs.webkit.org/show_bug.cgi?id=64646
18684
18685         Reviewed by Eric Seidel.
18686
18687         This patch causes the garden-o-matic display to update every 10 minutes
18688         to show the latest failures.  This patch required a moderately large
18689         refactoring of main.js to make the display incremental instead of
18690         all-at-once.
18691
18692         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
18693         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
18694         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
18695         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
18696
18697 2011-07-18  Tab Atkins  <jackalmage@gmail.com>
18698
18699         Tab Atkins is not in committers.py (as a non-committer)
18700         https://bugs.webkit.org/show_bug.cgi?id=64770
18701
18702         Reviewed by Eric Seidel.
18703
18704         * Scripts/webkitpy/common/config/committers.py:
18705
18706 2011-07-18  Ojan Vafai  <ojan@chromium.org>
18707
18708         Load ToT chromium bots by default in the flakiness dashboard
18709         https://bugs.webkit.org/show_bug.cgi?id=64756
18710
18711         Reviewed by Adam Barth.
18712
18713         For now, these are the most useful bots to look at. Eventually,
18714         we may want to load the build.webkit.org bots by default once
18715         most of the bots there upload results.
18716
18717         * TestResultServer/static-dashboards/dashboard_base.js:
18718
18719 2011-07-13  Jon Honeycutt  <jhoneycutt@apple.com>
18720
18721         Focus and selection events are not fired when a <select>'s selection
18722         changes
18723         https://bugs.webkit.org/show_bug.cgi?id=64504
18724         <rdar://problem/9319881>
18725
18726         Reviewed by Alice Liu.
18727
18728         * DumpRenderTree/AccessibilityController.h:
18729         Added m_notificationsEventHook for addNotificationListener().
18730         m_allEventsHook will now be used for setLogAccessibilityEvents().
18731
18732         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
18733         (AccessibilityController::AccessibilityController):
18734         Initialize m_notificationsEventHook.
18735         (AccessibilityController::~AccessibilityController):
18736         Turn off logging of all accessibility events. If
18737         m_notificationsEventHook is non-null, unhook it.
18738         (logEventProc):
18739         Add handling of EVENT_OBJECT_SELECTION.
18740         (AccessibilityController::setLogAccessibilityEvents):
18741         If the state of logging is not changing, return early. If we're turning
18742         off logging, unhook m_allEventsHook, and zero it out. Otherwise, add a
18743         hook for all events.
18744         (AccessibilityController::addNotificationListener):
18745         Use m_notificationsEventHook rather than m_allEventsHook.
18746
18747 2011-07-18  Eric Seidel  <eric@webkit.org>
18748
18749         webkit-patch apply-attachment does not work when not called from the root of the checkout
18750         https://bugs.webkit.org/show_bug.cgi?id=64751
18751
18752         Unreviewed.  Just updating unit test results after bug 64751.
18753
18754         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
18755         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
18756
18757 2011-07-18  Adam Barth  <abarth@webkit.org>
18758
18759         simplejson has trouble on chromium-linux
18760         https://bugs.webkit.org/show_bug.cgi?id=64757
18761
18762         Reviewed by Eric Seidel.
18763
18764         Use the native JSON, if available.
18765
18766         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
18767
18768 2011-07-18  Ojan Vafai  <ojan@chromium.org>
18769
18770         update the flakiness dashboard to understand the new platforms/formats in test_expectations
18771         https://bugs.webkit.org/show_bug.cgi?id=64743
18772
18773         Reviewed by Adam Barth.
18774
18775         The test expectations format changed a long time ago and the flakiness dashboard
18776         was never updated to match.
18777
18778         * TestResultServer/static-dashboards/dashboard_base.js:
18779         * TestResultServer/static-dashboards/flakiness_dashboard.html:
18780         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
18781         * TestResultServer/static-dashboards/timeline_explorer.html:
18782
18783 2011-07-18  Eric Seidel  <eric@webkit.org>
18784
18785         webkit-patch apply-attachment does not work when not called from the root of the checkout
18786         https://bugs.webkit.org/show_bug.cgi?id=64751
18787
18788         Reviewed by Adam Barth.
18789
18790         Last week I removed a os.chdir in EnsureWorkingDirectoryClean, which
18791         is a step that we run for most commands (as one of the earliest steps).
18792         EnsureWorkingDirectoryClean was incorrectly calling os.chdir to change
18793         the CWD to the checkout root.  This magically made a bunch of otherwise
18794         wrong code work.
18795
18796         When I realized that apply-attachment no longer worked as expected today,
18797         I went and fixed the bug, then realizing that we were not testing
18798         what the cwd was when running various commands.  I fixed our MockExecutive
18799         to always log what the cwd is and fixed a whole bunch of places
18800         where we needed to be setting the cwd.
18801
18802         Hopefully this will solve our cwd problems once and for-all, and webkit-patch
18803         will again correctly work when called from any directory (including outside
18804         of a webkit checkout).
18805
18806         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
18807         * Scripts/webkitpy/common/system/executive.py:
18808         * Scripts/webkitpy/common/system/workspace_unittest.py:
18809         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
18810         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py:
18811         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
18812         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
18813         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
18814         * Scripts/webkitpy/tool/bot/irc_command.py:
18815         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
18816         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
18817         * Scripts/webkitpy/tool/commands/download_unittest.py:
18818         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
18819         * Scripts/webkitpy/tool/commands/queues.py:
18820         * Scripts/webkitpy/tool/commands/queues_unittest.py:
18821         * Scripts/webkitpy/tool/commands/queuestest.py:
18822         * Scripts/webkitpy/tool/mocktool.py:
18823         * Scripts/webkitpy/tool/servers/gardeningserver.py:
18824         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
18825         * Scripts/webkitpy/tool/steps/build.py:
18826         * Scripts/webkitpy/tool/steps/checkstyle.py:
18827         * Scripts/webkitpy/tool/steps/editchangelog.py:
18828         * Scripts/webkitpy/tool/steps/preparechangelog.py:
18829         * Scripts/webkitpy/tool/steps/preparechangelogforrevert.py:
18830         * Scripts/webkitpy/tool/steps/runtests.py:
18831         * Scripts/webkitpy/tool/steps/steps_unittest.py:
18832         * Scripts/webkitpy/tool/steps/update.py:
18833
18834 2011-07-18  Adam Barth  <abarth@webkit.org>
18835
18836         Increase information garden-o-matic information density by switching to a table
18837         https://bugs.webkit.org/show_bug.cgi?id=64642
18838
18839         Reviewed by Dimitri Glazkov.
18840
18841         The old layout couldn't handle the large number of failures we had
18842         today.  This patch switches garden-o-matic to a table-based layout,
18843         which can handle many more failures gracefully.  (I expect we'll need
18844         more UI iterations.)
18845
18846         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
18847         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
18848         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
18849         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
18850
18851 2011-07-18  Adam Barth  <abarth@webkit.org>
18852
18853         garden-o-matic should be able to determine when compile breaks
18854         https://bugs.webkit.org/show_bug.cgi?id=64190
18855
18856         Reviewed by Dimitri Glazkov.
18857
18858         This patch adds a red-ish box to the top of the page whenever there is
18859         a compile error on the bots.  The box automatically opens and closes as
18860         appropriate and links to the waterfall display.  In the future, we
18861         might want to compute a regression range.
18862
18863         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
18864             - Add the build-only bots to the config.  We use these to check
18865               whether the build failed, which is faster than waiting for the
18866               tester bots to cycle.
18867         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
18868             - Add DOM for the alert bar.
18869         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
18870             - CSS to support the alert bar.
18871         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
18872             - Wiring up events to poll the buildbot to see whether compile has
18873               failed.
18874         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
18875             - Infrastructure for fetching and parsing the buildbot status JSON
18876               blob.  This code could be better factored for testability.  :(
18877         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
18878             - UI for displaying compile errors.
18879         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
18880
18881 2011-07-18  Sheriff Bot  <webkit.review.bot@gmail.com>
18882
18883         Unreviewed, rolling out r91132 and r91135.
18884         http://trac.webkit.org/changeset/91132
18885         http://trac.webkit.org/changeset/91135
18886         https://bugs.webkit.org/show_bug.cgi?id=64681
18887
18888         Broke GTK and Chromium debug bots (Requested by rniwa on
18889         #webkit).
18890
18891         * DumpRenderTree/AccessibilityController.h:
18892         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
18893         (AccessibilityController::AccessibilityController):
18894         (AccessibilityController::~AccessibilityController):
18895         (logEventProc):
18896         (AccessibilityController::setLogAccessibilityEvents):
18897         (AccessibilityController::addNotificationListener):
18898
18899 2011-07-17  Philippe Normand  <pnormand@igalia.com>
18900
18901         test-webkitpy failing with Python 2.5
18902         https://bugs.webkit.org/show_bug.cgi?id=64594
18903
18904         Reviewed by Eric Seidel.
18905
18906         Updated simplejson to version 2.1.6 that fixes an issue when
18907         dumping slashes. Also use cgi.parse_qs instead of the urlparse
18908         version to be compatible with Python2.5. Verified the fix with
18909         test-webkitpy on python 2.5 and python 2.6.
18910
18911         * Scripts/webkitpy/thirdparty/simplejson/README.txt:
18912         * Scripts/webkitpy/thirdparty/simplejson/__init__.py:
18913         * Scripts/webkitpy/thirdparty/simplejson/_speedups.c:
18914         (json_PyOS_string_to_double):
18915         (_convertPyInt_AsSsize_t):
18916         (_convertPyInt_FromSsize_t):
18917         (ascii_escape_char):
18918         (ascii_escape_unicode):
18919         (ascii_escape_str):
18920         (raise_errmsg):
18921         (join_list_unicode):
18922         (join_list_string):
18923         (_build_rval_index_tuple):
18924         (scanstring_str):
18925         (scanstring_unicode):
18926         (py_scanstring):
18927         (py_encode_basestring_ascii):
18928         (scanner_dealloc):
18929         (scanner_traverse):
18930         (scanner_clear):
18931         (_parse_object_str):
18932         (_parse_object_unicode):
18933         (_parse_array_str):
18934         (_parse_array_unicode):
18935         (_parse_constant):
18936         (_match_number_str):
18937         (_match_number_unicode):
18938         (scan_once_str):
18939         (scan_once_unicode):
18940         (scanner_call):
18941         (scanner_new):
18942         (scanner_init):
18943         (encoder_new):
18944         (encoder_init):
18945         (encoder_call):
18946         (_encoded_const):
18947         (encoder_encode_float):
18948         (encoder_encode_string):
18949         (_steal_list_append):
18950         (encoder_listencode_obj):
18951         (encoder_listencode_dict):
18952         (encoder_listencode_list):
18953         (encoder_dealloc):
18954         (encoder_traverse):
18955         (encoder_clear):
18956         (init_speedups):
18957         * Scripts/webkitpy/thirdparty/simplejson/decoder.py:
18958         * Scripts/webkitpy/thirdparty/simplejson/encoder.py:
18959         * Scripts/webkitpy/thirdparty/simplejson/ordered_dict.py: Added.
18960         * Scripts/webkitpy/thirdparty/simplejson/scanner.py:
18961         * Scripts/webkitpy/thirdparty/simplejson/tool.py: Added.
18962         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
18963
18964 2011-07-17  Dimitri Glazkov  <dglazkov@chromium.org>
18965
18966         Rename ModifierMatcher to SpecificityCalculator.
18967         https://bugs.webkit.org/show_bug.cgi?id=64660
18968
18969         It's a little longer than before, but it is much clearer.
18970
18971         Reviewed by Adam Barth.
18972
18973         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Renamed and adjusted
18974             all callsites, also rewrote the comments.
18975         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Adjusted callsites
18976             and renamed tests.
18977
18978 2011-07-15  Adam Barth  <abarth@webkit.org>
18979
18980         gardening server should proxy buildbot status for garden-o-matic
18981         https://bugs.webkit.org/show_bug.cgi?id=64588
18982
18983         Reviewed by Eric Seidel.
18984
18985         This information will be used to detect build breaks.  I took this
18986         opportunity to make ChromiumBuildBot non-static.
18987
18988         * Scripts/webkitpy/common/host.py:
18989         * Scripts/webkitpy/tool/commands/rebaseline.py:
18990         * Scripts/webkitpy/tool/commands/rebaseline_unittest.py:
18991         * Scripts/webkitpy/tool/mocktool.py:
18992         * Scripts/webkitpy/tool/servers/gardeningserver.py:
18993         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
18994
18995 2011-07-15  Julien Chaffraix  <jchaffraix@webkit.org>
18996
18997         [NRWT] Add support for --no-http
18998         https://bugs.webkit.org/show_bug.cgi?id=64564
18999
19000         Reviewed by Dirk Pranke.
19001
19002         Added support for --no-http, which disables both HTTP and websockets tests.
19003         It also matches the old-run-webkit-tests behavior if --force is used.
19004
19005         * Scripts/webkitpy/layout_tests/controllers/manager.py:
19006         Fixed HTTP_SUBDIR and WEBSOCKET_SUBDIR as tests do not start with a leading separator.
19007         We check if --no-http is set and add the HTTP / websockets tests to the skipped list prior to looking
19008         at the expectation file. Fixed the  _test_requires_lock function to use the same code path to determine
19009         what is worth have an HTTP lock as --no-http to avoid badness.
19010
19011         * Scripts/webkitpy/layout_tests/port/test.py: Added 2 new tests to our mock filesystem to validate that
19012         we do skip properly HTTP / websocket tests inside platform/.
19013
19014         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
19015         Added tests that we properly skip all the tests.
19016
19017         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
19018         Added checks for the command line arguments.
19019
19020 2011-07-13  Jon Honeycutt  <jhoneycutt@apple.com>
19021
19022         Focus and selection events are not fired when a <select>'s selection
19023         changes
19024         https://bugs.webkit.org/show_bug.cgi?id=64504
19025         <rdar://problem/9319881>
19026
19027         Reviewed by Alice Liu.
19028
19029         * DumpRenderTree/AccessibilityController.h:
19030         Added m_notificationsEventHook for addNotificationListener().
19031         m_allEventsHook will now be used for setLogAccessibilityEvents().
19032
19033         * DumpRenderTree/win/AccessibilityControllerWin.cpp:
19034         (AccessibilityController::AccessibilityController):
19035         Initialize m_notificationsEventHook.
19036         (AccessibilityController::~AccessibilityController):
19037         Turn off logging of all accessibility events. If
19038         m_notificationsEventHook is non-null, unhook it.
19039         (logEventProc):
19040         Add handling of EVENT_OBJECT_SELECTION.
19041         (AccessibilityController::setLogAccessibilityEvents):
19042         If the state of logging is not changing, return early. If we're turning
19043         off logging, unhook m_allEventsHook, and zero it out. Otherwise, add a
19044         hook for all events.
19045         (AccessibilityController::addNotificationListener):
19046         Use m_notificationsEventHook rather than m_allEventsHook.
19047
19048 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
19049
19050         Refactor TestExpectationModel to use TestExpectationLine as data item.
19051         https://bugs.webkit.org/show_bug.cgi?id=64635
19052
19053         This is a bit largish in scope. Does the following things:
19054
19055         1) Adds "path" member to TestExpectationLine to hold normalized path to test, computed at parsing,
19056            and changes code that used Port.normalize_test_name to rely on TestExpectationLine.path. As a result, TestExpectationModel no longer
19057            needs to have any port knowledge.
19058
19059         2) Adds "create_passing_expectation" class method to TestExpectationLine to generate a pristine passing expectation out of a test name,
19060            and changes TestExpectations._process_tests_without_expectations to use it, thus eliminating the need for a special API entry point.
19061            Now all expectations are added to the model in the same way!
19062
19063         3) Changes TestExpectationModel's main test index to store a tuple consisting of line number and TestExpectationLine instance.
19064
19065         Reviewed by Adam Barth.
19066
19067         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Refactored code.
19068
19069 2011-07-15  Adam Roben  <aroben@apple.com>
19070
19071         Teach TestFailures how to detect interrupted build steps
19072
19073         Fixes <http://webkit.org/b/64619> TestFailures page thinks all tests passed in
19074         http://build.webkit.org/builders/SnowLeopard%20Intel%20Release%20(WebKit2%20Tests)/builds/13401
19075
19076         Reviewed by Daniel Bates.
19077
19078         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
19079         (Builder.prototype.getNumberOfFailingTests): If the build step has a result code of 4, the
19080         build step was interrupted. Treat it as an error (by returning a failureCount of -1).
19081
19082         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
19083         Added a test for the above.
19084
19085         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
19086         (LayoutTestResultsLoader.prototype.start): Bumped the cache version to evict old, buggy
19087         cache data that was afflicted by the above bug.
19088
19089 2011-07-15  Eric Seidel  <eric@webkit.org>
19090
19091         new-run-webkit-tests crashes on Apple's Windows port when trying to launch Apache
19092         https://bugs.webkit.org/show_bug.cgi?id=64533
19093
19094         Reviewed by Adam Roben.
19095
19096         Blind removal of Chromium-specific code (now that Chromium does not use this file).
19097
19098         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
19099
19100 2011-07-15  Adam Roben  <aroben@apple.com>
19101
19102         Make TestFailures's list of possibly-flaky tests not so tall
19103
19104         Fixes <http://webkit.org/b/64618> TestFailures page's list of flaky tests takes up way too
19105         much room
19106
19107         Reviewed by Daniel Bates.
19108
19109         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
19110         Removed a no-longer-used .flakiness-example-separator selector.
19111
19112         (#failure-history > li):
19113         (#possibly-flaky-tests > li):
19114         (#failure-history > li, #possibly-flaky-tests > li):
19115         Use a lot less padding for the possibly-flaky tests.
19116
19117 2011-07-15  Tony Chang  <tony@chromium.org>
19118
19119         [chromium] land linux selection color layout test
19120         https://bugs.webkit.org/show_bug.cgi?id=64631
19121
19122         Reviewed by Ojan Vafai.
19123
19124         I had removed this code in r76620 because I thought it wasn't used.
19125         Turns out there was a layout test in the chromium tree that never
19126         got migrated that used it.  Re-add the code and land the layout
19127         test (I will remove it from the chromium tree soon).
19128
19129         * DumpRenderTree/chromium/LayoutTestController.cpp:
19130         (LayoutTestController::LayoutTestController):
19131         (LayoutTestController::forceRedSelectionColors):
19132         * DumpRenderTree/chromium/LayoutTestController.h:
19133
19134 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
19135
19136         Revert http://trac.webkit.org/changeset/91091, since Python 2.5
19137         does not support enumerate() start argument.
19138
19139         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Reverted.
19140
19141 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
19142
19143         Clean up test_expectations.py after refactorings.
19144         https://bugs.webkit.org/show_bug.cgi?id=64620
19145
19146         * Renamed all variables holding TestExpectationLine to expectation_line to avoid confusion with
19147           its sub-part, the actual expectation.
19148         * Renamed all references to options to modifier to eliminate dual terminology.
19149         * Made a bunch of parser constants and changed all callsites to use them.
19150         * Various other minor clean-ups.
19151
19152         Reviewed by Adam Barth.
19153
19154         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Cleaned up stuff.
19155         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed callsites after cleanup.
19156         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Ditto.
19157
19158 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
19159
19160         Move expectation parsing code to TestExpectationParser.
19161         https://bugs.webkit.org/show_bug.cgi?id=64605
19162
19163         This is a somewhat mechanical move, with two interesting bits:
19164
19165         1) TestExpectationParser.parse methods renamed to tokenize, to better
19166         reflect what they do
19167
19168         2) TestExpectationLine now carries all of its info, from tokens to parsed data,
19169         and even the list of tests that it matches (a line may refer to more than one test).
19170
19171         Reviewed by Adam Barth.
19172
19173         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Moved parsing-related TestExpectaions methods
19174             to TestExpectationParser, added more members to TestExpectationLine to carry parsed info, renamed existing parse methods
19175             to "tokenize", changed callsites to use new code.
19176         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed unit tests to reflect renames.
19177
19178 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
19179
19180         Switch to use Python enumerate function to enumerate line numbers.
19181         https://bugs.webkit.org/show_bug.cgi?id=64602
19182
19183         Reviewed by Adam Barth.
19184
19185         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Changed to use enumerate.
19186
19187 2011-07-15  Martin Robinson  <mrobinson@igalia.com>
19188
19189         Build fixes for WebKit2. Ensure that all generated sources are
19190         on nodist primaries, that they are on forward declared variables
19191         so that BUILT_SOURCES is calculated properly and that zlib is
19192         included during linking (for WOFF support).
19193
19194         * WebKitTestRunner/GNUmakefile.am:
19195
19196 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
19197
19198         Store error and warning information on TestExpectationLine.
19199         https://bugs.webkit.org/show_bug.cgi?id=64565
19200
19201         Keeping errors and warnings on the TestExpectationLine instance allows us to decouple storing errors
19202         from various parsing and validation mechanisms and have more flexibility in reporting and understanding the origin of the errors.
19203
19204         Reviewed by Adam Barth.
19205
19206         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added TestExpectationLine.warnings list to keep track of non-fatal errors,
19207             converted the code to add errors and warnings to corresponding TestExpectationLine instances, removed the code that used to store
19208             this info on TestExpectations. In the process, had to refactor ModifierMatcher a bit to take in a TestExpdectationLine instance.
19209         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed ModifierMatcher callsite.
19210
19211 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
19212
19213         Plumb the use of TestExpectationLine deeper, clean up.
19214         https://bugs.webkit.org/show_bug.cgi?id=64559
19215
19216         Instead of carrying various bits of TestExpectationLine, plumb it down to its consumers,
19217         also cleaning up names and remove an unused TestExpectations._get_options_list member.
19218
19219         Reviewed by Adam Barth.
19220
19221         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
19222
19223 2011-07-15  Dimitri Glazkov  <dglazkov@chromium.org>
19224
19225         Remove the notion of TestExpectationLine.valid, start storing parsing errors in expectations themselves.
19226         https://bugs.webkit.org/show_bug.cgi?id=64554
19227
19228         This moves us toward the world where errors are collected on the expectations, which allows us to
19229         easily enumerate them, keep association with the point of origin, and freely pass TestExpectationLine instances around.
19230
19231         Also eliminate the validator idea, since validation is a context-dependent concept and has to be decoupled from initial
19232         parsing.
19233
19234         Reviewed by Adam Barth.
19235
19236         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Removed TestExpectation.valid, validator,
19237             changed TestExpectationParser to collect errors in TestExpectationLine, refactored surrounding code.
19238         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Changed tests to accommodate changes.
19239         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Ditto.
19240
19241 2011-07-15  Adam Roben  <aroben@apple.com>
19242
19243         Rename TestFailureBugForm to FailingTestsBugForm
19244
19245         The new name will match better with a forthcoming FlakyTestBugForm class.
19246
19247         Fixes <http://webkit.org/b/64598> TestFailures page's TestFailureBugForm class has a bad
19248         name
19249
19250         Reviewed by Daniel Bates.
19251
19252         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm.js: Renamed from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm.js.
19253         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FailingTestsBugForm_unittests.js: Renamed from Tools/BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm_unittests.js.
19254
19255         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
19256         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
19257         Updated for renames.
19258
19259         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Ditto, and
19260         reordered <script>s to put files which others depend on for parsing first.
19261
19262 2011-07-15  Adam Barth  <abarth@webkit.org>
19263
19264         NRWT stores the Chromium revision number in full_results.json
19265         https://bugs.webkit.org/show_bug.cgi?id=64586
19266
19267         I have no earthly idea how to test this change.
19268
19269         * Scripts/webkitpy/layout_tests/controllers/manager.py:
19270         * Scripts/webkitpy/layout_tests/port/base.py:
19271
19272 2011-07-14  Ojan Vafai  <ojan@chromium.org>
19273
19274         fix remaining style issues in the static-dashboards directory
19275         https://bugs.webkit.org/show_bug.cgi?id=64561
19276
19277         Reviewed by Adam Barth.
19278
19279         * TestResultServer/static-dashboards/aggregate_results.html:
19280         * TestResultServer/static-dashboards/builders.js:
19281         * TestResultServer/static-dashboards/dashboard_base.js:
19282         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
19283         * TestResultServer/static-dashboards/timeline_explorer.html:
19284         * TestResultServer/static-dashboards/treemap.html:
19285
19286 2011-07-14  Yuta Kitamura  <yutak@chromium.org>
19287
19288         WebSocket: Introduce pywebsocket-0.6b2
19289         https://bugs.webkit.org/show_bug.cgi?id=64534
19290
19291         Reviewed by Kent Tamura.
19292
19293         * Scripts/webkitpy/thirdparty/__init__.py:
19294
19295 2011-07-14  Dimitri Glazkov  <dglazkov@chromium.org>
19296
19297         Introduce TestExpectationsModel, split out of TestExpectations.
19298         https://bugs.webkit.org/show_bug.cgi?id=64531
19299
19300         This is a simple split-and-make-work refactoring, a first step among many.
19301
19302         Reviewed by Adam Barth.
19303
19304         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Moved all model-related members
19305             out of TestExpectations and into TestExpectationsModel.
19306
19307 2011-07-14  Eric Seidel  <eric@webkit.org>
19308
19309         Move webkitpy off of loose mocks
19310         https://bugs.webkit.org/show_bug.cgi?id=64508
19311
19312         Unreviewed.  Fixing a test which fails under test-webkitpy --all (but not in a normal run).
19313
19314         * Scripts/webkitpy/common/checkout/scm/git.py:
19315
19316 2011-07-14  Adam Barth  <abarth@webkit.org>
19317
19318         garden-o-matic should have a "rebaseline" button
19319         https://bugs.webkit.org/show_bug.cgi?id=64446
19320
19321         Reviewed by Ojan Vafai.
19322
19323         This patch adds a basic Rebaseline button that copies the baselines
19324         displayed in the results pane into the appropriate directory in your
19325         working copy.
19326
19327         There are two main limitations:
19328
19329         1) There is no UI for actually committing the baselines.
19330
19331         2) The baselines are not optimized for redundancy (meaning you can have
19332            identical baselines in both chromium-mac and chromium-win).
19333
19334         * Scripts/webkitpy/tool/commands/rebaseline.py:
19335             - Turns out we need to create the directory for the baseline if it doesn't exist yet.
19336         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js:
19337             - Add an programatic API to call the server.
19338         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
19339             - Add the rebaseline button itself.
19340         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
19341             - Change the CSS so that the Rebaseline and Close buttons can
19342               appear in the normal order in the DOM.
19343         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
19344             - Bind the event and translate the parameters.
19345             - Hide/show the rebaseline button, as appropriate.
19346         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
19347             - Add some helpful utility functions for manipulating failure types.
19348         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
19349
19350 2011-07-14  Ryosuke Niwa  <rniwa@webkit.org>
19351
19352         REGRESSION: webkit-patch roll-chromium-deps is broken
19353         https://bugs.webkit.org/show_bug.cgi?id=64568
19354
19355         Reviewed by Dirk Pranke
19356
19357         Use _filesystem.join.
19358
19359         * Scripts/webkitpy/common/checkout/checkout.py:
19360         * Scripts/webkitpy/common/checkout/checkout_unittests.py:
19361
19362 2011-07-14  Noam Rosenthal  <noam.rosenthal@nokia.com>
19363
19364         Adding myself to the reviewers list; No review needed.
19365
19366         * Scripts/webkitpy/common/config/committers.py:
19367
19368 2011-07-14  Adam Roben  <aroben@apple.com>
19369
19370         Don't use Element.prototype.classList in TestFailures
19371
19372         Safari 5 doesn't support it.
19373
19374         Fixes <http://webkit.org/b/64550> Can't expand flaky tests on TestFailures page in Safari 5
19375
19376         Reviewed by Daniel Bates.
19377
19378         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
19379         (Element.prototype.hasStyleClass):
19380         (Element.prototype.addStyleClass):
19381         (Element.prototype.removeStyleClass):
19382         (Element.prototype.toggleStyleClass):
19383         Added these helper functions which simulate classList functionality.
19384
19385         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities_unittests.js:
19386         Added. Tests for the above.
19387
19388         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
19389         (ViewController.prototype._domForPossiblyFlakyTests): Changed to use
19390         toggleStyleClass/hasStyleClass instead of classList.
19391
19392         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
19393         Added Utilities_unittests.js, and reordered the tested files to be in
19394         roughly dependency order (i.e., the lowest-level files are imported
19395         and tested first).
19396
19397 2011-07-14  Eric Seidel  <eric@webkit.org>
19398
19399         NRWT doesn't store the svn revision in full_results.json on chromium-win
19400         https://bugs.webkit.org/show_bug.cgi?id=64492
19401
19402         Unreviewed.  Just fixing my test-webkitpy regression.
19403
19404         Fix the unit tests.  Unfortunately scm does not use a filesystem
19405         object so we can't control the result of detect_scm_system.
19406         When detect_scm_system would fail, we would log, which would
19407         cause all passing_run integration tests to fail.
19408
19409         * Scripts/webkitpy/common/checkout/scm/scm.py:
19410         * Scripts/webkitpy/layout_tests/controllers/manager.py:
19411         * Scripts/webkitpy/layout_tests/port/base.py:
19412         * Scripts/webkitpy/layout_tests/port/test.py:
19413         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
19414
19415 2011-07-14  Eric Seidel  <eric@webkit.org>
19416
19417         NRWT doesn't store the svn revision in full_results.json on chromium-win
19418         https://bugs.webkit.org/show_bug.cgi?id=64492
19419
19420         Reviewed by Ojan Vafai.
19421
19422         This should fix the bug.  Unfortunately this code is currently
19423         impossible to test since it's impossible to mock detect_scm_system at the
19424         moment.  I started re-writing scm.detection.py to be mockable, but decided that
19425         was best left for another day.
19426
19427         * Scripts/webkitpy/layout_tests/controllers/manager.py:
19428         * Scripts/webkitpy/layout_tests/port/base.py:
19429
19430 2011-07-14  Ojan Vafai  <ojan@chromium.org>
19431
19432         fix coding style of dashboard_base.js
19433         https://bugs.webkit.org/show_bug.cgi?id=64545
19434
19435         Reviewed by Adam Roben.
19436
19437         The code changes to the files other than dashboard_base.js
19438         are just fallout from renaming globals in dashboard_base.js.
19439
19440         * TestResultServer/static-dashboards/aggregate_results.html:
19441         * TestResultServer/static-dashboards/builders.js:
19442         * TestResultServer/static-dashboards/dashboard_base.js:
19443         * TestResultServer/static-dashboards/flakiness_dashboard.html:
19444         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js:
19445         * TestResultServer/static-dashboards/timeline_explorer.html:
19446         * TestResultServer/static-dashboards/treemap.html:
19447
19448 2011-07-14  Carlos Garcia Campos  <cgarcia@igalia.com>
19449
19450         Unreviewed. Fix WebKit2 GTK build after r90953.
19451
19452         * WebKitTestRunner/GNUmakefile.am:
19453
19454 2011-07-14  Adam Barth  <abarth@webkit.org>
19455
19456         garden-o-matic results pane should be more discoverable
19457         https://bugs.webkit.org/show_bug.cgi?id=64513
19458
19459         Reviewed by Eric Seidel.
19460
19461         This patch causes us to trigger the results pane on mouse clicks
19462         instead of mousein.  Also, handle the case of no results details more
19463         elegantly.
19464
19465         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
19466         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
19467             - Change the binding to "click".
19468         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
19469             - Handle the "no results URLs" case explicitly instead of never
19470               calling the callback.
19471         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
19472         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
19473             - Show a message when there are no result details to show.
19474         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
19475
19476 2011-07-14  Adam Barth  <abarth@webkit.org>
19477
19478         garden-o-matic should badge TIMEOUT failures
19479         https://bugs.webkit.org/show_bug.cgi?id=64435
19480
19481         Reviewed by Eric Seidel.
19482
19483         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
19484         (.regression .what a[draggable].TIMEOUT::before):
19485             - Add TIMEOUT.
19486
19487 2011-07-14  Adam Barth  <abarth@webkit.org>
19488
19489         Fix Python exception blocking the commit-queue.
19490
19491         * Scripts/webkitpy/common/checkout/scm/git.py:
19492
19493 2011-07-13  Eric Seidel  <eric@webkit.org>
19494
19495         Move webkitpy off of loose mocks
19496         https://bugs.webkit.org/show_bug.cgi?id=64508
19497
19498         Reviewed by Adam Barth.
19499
19500         Using Mock has caused us more pain than help.
19501         It's possible that there was a cleaner way to use it
19502         (maybe Mock(class) instead of inheriting from it?).
19503         But for now, I've removed all uses of Mock from mocktool.py.
19504
19505         I also moved run_command into the only 3 files which call it
19506         instead of leaving the deprecated method in executive.py.
19507
19508         I changed various direct calls to os.* to use filesystem instead.
19509
19510         * Scripts/webkitpy/common/checkout/checkout.py:
19511         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
19512         * Scripts/webkitpy/common/checkout/scm/git.py:
19513         * Scripts/webkitpy/common/checkout/scm/scm.py:
19514         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
19515         * Scripts/webkitpy/common/checkout/scm/svn.py:
19516         * Scripts/webkitpy/common/system/executive.py:
19517         * Scripts/webkitpy/common/system/executive_unittest.py:
19518         * Scripts/webkitpy/tool/mocktool.py:
19519         * Scripts/webkitpy/tool/steps/cleanworkingdirectory_unittest.py:
19520         * Scripts/webkitpy/tool/steps/ensurelocalcommitifneeded.py:
19521
19522 2011-07-13  Eric Seidel  <eric@webkit.org>
19523
19524         NRWT doesn't store the svn revision in full_results.json on chromium-win
19525         https://bugs.webkit.org/show_bug.cgi?id=64492
19526
19527         Reviewed by Adam Barth.
19528
19529         Add a warning message in the revision='' case to help us diagnose
19530         what's going wrong on the bot.
19531
19532         * Scripts/webkitpy/layout_tests/controllers/manager.py:
19533
19534 2011-07-13  Dirk Pranke  <dpranke@chromium.org>
19535
19536         new-run-webkit-tests: print baseline search path as part of config output
19537         https://bugs.webkit.org/show_bug.cgi?id=64499
19538
19539         Reviewed by Eric Seidel.
19540
19541         * Scripts/webkitpy/layout_tests/controllers/manager.py:
19542         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
19543
19544 2011-07-13  Adam Barth  <abarth@webkit.org>
19545
19546         gardening server should have an API for parsing changelogs
19547         https://bugs.webkit.org/show_bug.cgi?id=64495
19548
19549         Reviewed by Eric Seidel.
19550
19551         This patch exposes much of the same information from CommitInfo in a
19552         dictionary form, which is easier to send over-the-wire as JSON to the
19553         frontend.
19554
19555         * Scripts/webkitpy/common/checkout/checkout.py:
19556         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
19557         * Scripts/webkitpy/tool/mocktool.py:
19558         * Scripts/webkitpy/tool/servers/gardeningserver.py:
19559         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
19560         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
19561
19562 2011-07-13  Eric Seidel  <eric@webkit.org>
19563
19564         REGRESSION: GitTestWithMock.test_create_patch fails
19565         https://bugs.webkit.org/show_bug.cgi?id=62945
19566
19567         Reviewed by Daniel Bates.
19568
19569         I was not able to reproduce the exact failure seen in the bug,
19570         however this test was failing on my machine for other reasons.
19571
19572         I went through and did an audit of our run_command usage, it's
19573         entirely in scm classes after this change.  (Not surprising given
19574         that scm.py was the second file ever created in webkit.py.)
19575
19576         The real bug I'm fixing here is that we were setting executive.should_log
19577         when the value had changed to executive._should_log.  Now we set the right one
19578         and the test works again.
19579
19580         * Scripts/webkitpy/common/checkout/checkout.py:
19581         * Scripts/webkitpy/common/checkout/scm/git.py:
19582         * Scripts/webkitpy/common/checkout/scm/scm.py:
19583         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
19584         * Scripts/webkitpy/common/checkout/scm/svn.py:
19585
19586 2011-07-13  Ilya Sherman  <isherman@chromium.org>
19587
19588         Fix WTF header guard style check
19589         https://bugs.webkit.org/show_bug.cgi?id=64488
19590
19591         Reviewed by David Levin.
19592
19593         * Scripts/webkitpy/style/checkers/cpp.py:
19594         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Verify that we don't *always* suggest "WTF_" as a prefix
19595
19596 2011-07-13  Ojan Vafai  <ojan@chromium.org>
19597
19598         bring flakiness_dashboard.html closer to webkit style
19599         https://bugs.webkit.org/show_bug.cgi?id=64477
19600
19601         Reviewed by Adam Barth.
19602
19603         * TestResultServer/static-dashboards/dashboard_base.js:
19604         Fix bad variable name that would hit only when running the tests.
19605         * TestResultServer/static-dashboards/flakiness_dashboard.html:
19606         No code/logic changes. All moving/removing brackets, indents and moving things
19607         to one line.
19608
19609 2011-07-13  Xan Lopez  <xlopez@igalia.com>
19610
19611         [GTK] Fix distcheck
19612
19613         Reviewed by Martin Robinson.
19614
19615         * WebKitTestRunner/GNUmakefile.am: mark built sources as nodist.
19616
19617 2011-07-13  Joseph Pecoraro  <joepeck@webkit.org>
19618
19619         Unreviewed fix. Add a "\n" to a printf that somehow got lost.
19620
19621         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
19622         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
19623
19624 2011-07-13  Joseph Pecoraro  <joepeck@webkit.org>
19625
19626         ApplicationCache Quota Output is Flakey
19627         https://bugs.webkit.org/show_bug.cgi?id=64410
19628
19629         Reviewed by Alexey Proskuryakov.
19630
19631         Unify the delegate logging for reaching application cache
19632         quotas. Truncate the space needed to the nearest 10000
19633         for less flakey test output. This also better supports
19634         printing NSUInteger on non-64 bit systems by casting to
19635         an unsigned long.
19636
19637         * DumpRenderTree/mac/UIDelegate.mm:
19638         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:totalSpaceNeeded:]):
19639         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
19640         (WebCore::DumpRenderTree::dumpApplicationCacheQuota):
19641
19642 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
19643
19644         Remove unused TestExpectations._overrides.
19645         https://bugs.webkit.org/show_bug.cgi?id=64470
19646
19647         Reviewed by Adam Barth.
19648
19649         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Removed unused member.
19650
19651 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
19652
19653         Introduce TestExpectationSerializer.list_to_string.
19654         https://bugs.webkit.org/show_bug.cgi?id=64462
19655
19656         Reviewed by Adam Barth.
19657
19658         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Added list_to_string and change the relevant callsite to use it.
19659         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Added tests for it.
19660
19661 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
19662
19663         Consolidate expectations parsing code.
19664         https://bugs.webkit.org/show_bug.cgi?id=64460
19665
19666         Reviewed by Adam Barth.
19667
19668         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Folded TestExpectationParser._split_expectation_string into TestExpectationParser.parse.
19669
19670 2011-07-12  Brent Fulgham  <bfulgham@webkit.org>
19671
19672         Standardize WinCairo conditionalized code under PLATFORM macro.
19673         https://bugs.webkit.org/show_bug.cgi?id=64377
19674
19675         Reviewed by Maciej Stachowiak.
19676
19677         Update compile-time conditions that had previously used #ifdef
19678         of WIN_CAIRO to consistently use PLATFORM(WIN_CAIRO).
19679
19680         * DumpRenderTree/config.h: Switch to PLATFORM(WIN_CAIRO)
19681         * DumpRenderTree/win/DumpRenderTreeCairo.vsprops: Remove redundant
19682           define of WIN_CAIRO=1.
19683         * DumpRenderTree/win/DumpRenderTreeDebugCairoCFLite.vsprops: Add
19684           WinCairo.vsprops to set (to match Release build).
19685
19686 2011-07-13  Dimitri Glazkov  <dglazkov@chromium.org>
19687
19688         Eliminate TestExpectationsFile.
19689         https://bugs.webkit.org/show_bug.cgi?id=64458
19690
19691         Turns out, we can just use a Python list.
19692
19693         Reviewed by Adam Barth.
19694
19695         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Folded TestExpectationsFile.append into TestExpectationParser.parse_list,
19696             removed TestExpectationsFile.
19697         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Moved tests to better reflect new names, removed iterator test,
19698             since there's no more custom iterator machinery.
19699
19700 2011-07-13  Philippe Normand  <pnormand@igalia.com>
19701
19702         Unreviewed, added my other email addresses.
19703
19704         * Scripts/webkitpy/common/config/committers.py:
19705
19706 2011-07-13  Adam Barth  <abarth@webkit.org>
19707
19708         Folks should only be listed once.
19709
19710         * Scripts/webkitpy/common/config/committers.py:
19711
19712 2011-07-13  Philippe Normand  <pnormand@igalia.com>
19713
19714         Unreviewed, adding myself as Reviewer.
19715
19716         * Scripts/webkitpy/common/config/committers.py:
19717
19718 2011-07-13  Adam Barth  <abarth@webkit.org>
19719
19720         Fix flickering bug introduced by my previous patch.  I forgot to change
19721         the name of the class everywhere.
19722
19723         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
19724
19725 2011-07-12  Philippe Normand  <pnormand@igalia.com>
19726
19727         [GTK] media/media-blocked-by-willsendrequest.html fails
19728         https://bugs.webkit.org/show_bug.cgi?id=63699
19729
19730         Reviewed by Martin Robinson.
19731
19732         * DumpRenderTree/gtk/DumpRenderTree.cpp:
19733         (willSendRequestCallback): Abort the request if explicitely asked
19734         by the LayoutTestController's willSendRequestReturnsNull() function.
19735
19736 2011-07-13  Adam Barth  <abarth@webkit.org>
19737
19738         Remove "Dismiss" button from garden-o-matic butterbar
19739         https://bugs.webkit.org/show_bug.cgi?id=64443
19740
19741         Reviewed by Dimitri Glazkov.
19742
19743         We don't have any persistent butter bar messages yet, so the dismiss
19744         button is premature (and fairly heavy, visually).
19745
19746         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
19747         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
19748
19749 2011-07-13  Adam Barth  <abarth@webkit.org>
19750
19751         Refine garden-o-matic status pane
19752         https://bugs.webkit.org/show_bug.cgi?id=64442
19753
19754         Reviewed by Dimitri Glazkov.
19755
19756         This patch makes two improvements to the status pane:
19757
19758         1) We only query the server for failure types that we're expecting.
19759            This dramatically reduces the number of HTTP requests, making
19760            loading the status pane faster.
19761
19762         2) The status pane now displays which test and which builder it is
19763            showing results for.
19764
19765         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
19766         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
19767         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
19768         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
19769         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
19770         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
19771         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
19772
19773 2011-07-13  Adam Barth  <abarth@webkit.org>
19774
19775         garden-o-matic should show test results
19776         https://bugs.webkit.org/show_bug.cgi?id=64440
19777
19778         Reviewed by Adam Roben.
19779
19780         This patch introduces the results detail pane, which appears at the
19781         bottom of the window and contains test results from the bots.
19782         Currently, you can activate the pane by mousing over one of the builder
19783         names associated with a failing test.
19784
19785         This is just a first iteration of the UI.  There's no way to resize or
19786         zoom in on elements of the details pane, and images likely aren't sized
19787         correctly, but it's a place to start.
19788
19789         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js:
19790             - Add constants for our data attributes so we don't typo them!
19791         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
19792             - Add DOM structure for the details pane.
19793         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
19794         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
19795             - Bind events for showing and hiding the details pane.
19796         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
19797             - Reduce the number of result types that we query for to avoid
19798               overfilling the details pane.  We'll probably need another
19799               solution here in the long-term.  For example, we could use tabs
19800               to pack more results into the pane.
19801         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
19802         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
19803             - Now that we're using the config package during testing, we need
19804               to include it in the testing HTML.
19805         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
19806         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
19807             - Change the DOM structure of the results detail to make it fit
19808               nicely in the details pane (rather than flowing freely in the
19809               body, as it did before).
19810
19811 2011-07-13  Adam Roben  <aroben@apple.com>
19812
19813         Make TestFailures show every time a possibly-flaky test failed, but hide it by default
19814
19815         It's useful to be able to see every time a flaky test failed to see whether it failed the
19816         same way every time. But doing so takes a lot of space, so the list of failures is now
19817         collapsed by default and can be revealed using a disclosure triangle.
19818
19819         Fixes <http://webkit.org/b/64455> TestFailures page doesn't show as much information for
19820         flaky tests as I would like, even though the page is already so long
19821
19822         Reviewed by Dimitri Glazkov.
19823
19824         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
19825         (FlakyLayoutTestDetector.prototype.allFailures): Replaced flakinessExamples with this
19826         function. Now returns all failures for the given test.
19827
19828         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector_unittests.js:
19829         Added. This just contains some simple tests of the FlakyLayoutTestDetector class. We'll add
19830         more over time.
19831
19832         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
19833         (LayoutTestHistoryAnalyzer.prototype.start): Updated the documentation comment to reflect
19834         that we no longer return passing builds for possibly-flaky tests.
19835
19836         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
19837         (.existing-bugs-list, .suspect-revisions-list, .flakiness-examples-list): Make the list of
19838         flakiness examples small, too, since it can get quite long.
19839
19840         (.disclosure-triangle):
19841         (.expanded > .disclosure-triangle):
19842         Simple styles for the disclosure triangle.
19843
19844         (.flakiness-examples-list): Collapse the list by default.
19845
19846         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
19847         (ViewController.prototype._displayBuilder): Pass the total number of builds analyzed to
19848         _domForPossiblyFlakyTests.
19849         (ViewController.prototype._domForPossiblyFlakyTests): Put a disclosure triangle to the left
19850         of each test name, and the number of failures to the right. When the disclosure triangle is
19851         clicked for the first time, we build up the list of failures and expand the element. After
19852         that we just collapse or expand the element on subsequent clicks.
19853
19854         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
19855         Pulled in new tests.
19856
19857 2011-07-13  Adam Roben  <aroben@apple.com>
19858
19859         Teach TestFailures to understand NRWT's output when it exits early due to too many failures
19860
19861         Fixes <http://webkit.org/b/64456> TestFailures page reports way too many failures when NRWT
19862         exits early
19863
19864         Reviewed by Dimitri Glazkov.
19865
19866         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
19867         (Builder.prototype.getNumberOfFailingTests): Relaxed the "Exiting early" test not to require
19868         it to be at the beginning of the line, since NRWT prints a bunch of junk earlier in the
19869         line. Tightened up the regex that's used to parse the number of failing tests to require the
19870         leading number to be followed by whitespace so that we won't parse the "2011" in
19871         "2011-07-13" as a number of failures.
19872
19873         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
19874         Added a test for the above.
19875
19876 2011-07-12  Dimitri Glazkov  <dglazkov@chromium.org>
19877
19878         Extract model-like TestExpectationLine and TestExpectationFile from TestExpectations.
19879         https://bugs.webkit.org/show_bug.cgi?id=64386
19880
19881         This is the first step in converting TestExpectations to a real model.
19882         * TestExpectationsLine represents a line in the test_expectations.txt file, and
19883         * TestExpectationsFile represents the file, which is a collection of lines.
19884
19885         Reviewed by Adam Barth.
19886
19887         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
19888         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py:
19889         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
19890
19891 2011-07-13  Xan Lopez  <xlopez@igalia.com>
19892
19893         [GTK] Do not grab focus too early in DRT.
19894
19895         Reviewed by Gustavo Noronha.
19896
19897         It causes a layout to happen and a progress signal to be emitted
19898         since r90900, but at this point we don't have a
19899         LayoutTestController object and we'll eventually crash. Since we
19900         already grab focus at the beginning of runTest() this is
19901         redundant, so get rid of it to fix the crash.
19902
19903         * DumpRenderTree/gtk/DumpRenderTree.cpp:
19904         (main): remove call to grab_focus
19905
19906 2011-07-13  Sheriff Bot  <webkit.review.bot@gmail.com>
19907
19908         Unreviewed, rolling out r90893 and r90894.
19909         http://trac.webkit.org/changeset/90893
19910         http://trac.webkit.org/changeset/90894
19911         https://bugs.webkit.org/show_bug.cgi?id=64441
19912
19913         NRWT still doesn't work on qt-mac platform (Requested by Ossy
19914         on #webkit).
19915
19916         * Scripts/run-webkit-tests:
19917         (useNewRunWebKitTests):
19918
19919 2011-07-13  Csaba Osztrogonác  <ossy@webkit.org>
19920
19921         [Qt] NRWT should pick up the right httpd config file
19922         https://bugs.webkit.org/show_bug.cgi?id=64086
19923
19924         * Scripts/run-webkit-tests: Enable NRWT on qt-mac platform after r90810.
19925         (useNewRunWebKitTests):
19926
19927 2011-07-12  Mark Rowe  <mrowe@apple.com>
19928
19929         Fix the 32-bit build.
19930
19931         * DumpRenderTree/mac/UIDelegate.mm:
19932         (-[UIDelegate webView:exceededApplicationCacheOriginQuotaForSecurityOrigin:totalSpaceNeeded:]):
19933         Cast the NSUInteger value to unsigned long to match the format specifier.
19934
19935 2011-07-12  Adam Barth  <abarth@webkit.org>
19936
19937         Improve garden-o-matic UI when the bots fail to report revision numbers
19938         https://bugs.webkit.org/show_bug.cgi?id=64427
19939
19940         Reviewed by Dimitri Glazkov.
19941
19942         At least Win (dbg)(2) seems to fail to report the SVN revision number
19943         in full_results.json.  This patch makes garden-o-matic more robust to
19944         missing revision numbers.
19945
19946         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
19947         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
19948         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
19949         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
19950         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
19951
19952 2011-07-12  Adam Barth  <abarth@webkit.org>
19953
19954         Adjust garden-o-matic layout to use fewer lines
19955         https://bugs.webkit.org/show_bug.cgi?id=64422
19956
19957         Reviewed by Ojan Vafai.
19958
19959         This patch effectively merges the test name line with the list of
19960         builders on which the test fails.  Each test failure now occupies two
19961         lines instead of three.
19962
19963         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
19964         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
19965         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
19966
19967 2011-07-12  Adam Barth  <abarth@webkit.org>
19968
19969         garden-o-matic should label tests failures of new tests as such
19970         https://bugs.webkit.org/show_bug.cgi?id=64421
19971
19972         Reviewed by Ojan Vafai.
19973
19974         Also, we shouldn't dim failures of new tests, even if we've only seen
19975         them once because they're likely to be real problems that need
19976         attention.
19977
19978         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
19979         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout.js: Added.
19980         * Scripts/webkitpy/tool/servers/data/gardeningserver/checkout_unittests.js: Added.
19981         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
19982         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
19983         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
19984         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
19985         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
19986         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
19987         * Scripts/webkitpy/tool/servers/gardeningserver.py:
19988
19989 2011-07-12  Adam Barth  <abarth@webkit.org>
19990
19991         garden-o-matic should display how many times we've seen a failure
19992         https://bugs.webkit.org/show_bug.cgi?id=64417
19993
19994         Reviewed by Ojan Vafai.
19995
19996         This patch adds some UI to display how many times we've seen a given
19997         failure, which can be helpful for determining whether that failure is a
19998         real failure or a flaky test.
19999
20000         When a failure has only been seen once (i.e., only a single run on a
20001         single bot), we set the opacity of to 50% to avoid distracting the
20002         gardener.
20003
20004         This patch also refactors the failure walker to have a simpler API
20005         internally by moving from an object-oriented paradigm to a functional
20006         paradigm.
20007
20008         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
20009         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
20010         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
20011         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
20012         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
20013         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
20014
20015 2011-07-12  Chris Rogers  <crogers@google.com>
20016
20017         Enable Web Audio for chromium DRT
20018         https://bugs.webkit.org/show_bug.cgi?id=64409
20019
20020         Reviewed by James Robinson.
20021
20022         * DumpRenderTree/chromium/TestShell.cpp:
20023         (TestShell::TestShell):
20024
20025 2011-07-12  Adam Barth  <abarth@webkit.org>
20026
20027         Tweak some UI in garden-o-matic now that the tree actually has a
20028         failure and I can see what this all looks like.
20029
20030         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css:
20031         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
20032         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
20033
20034 2011-07-12  Joseph Pecoraro  <joepeck@webkit.org>
20035
20036         Unreviewed. Skipping a few tests which fail due to differing output
20037            Unit tests shouldn't be hitting the disk anyway.  It's possible Config should move off of Port and onto Tool/Host directly.
20038         * Scripts/webkitpy/layout_tests/port/factory.py:
20039         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
20040         * Scripts/webkitpy/layout_tests/port/gtk.py:
20041         * Scripts/webkitpy/layout_tests/port/gtk_unittest.py: Added.
20042         * Scripts/webkitpy/layout_tests/port/mac.py:
20043         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
20044         * Scripts/webkitpy/layout_tests/port/qt.py:
20045         * Scripts/webkitpy/layout_tests/port/qt_unittest.py:
20046         * Scripts/webkitpy/layout_tests/port/webkit.py:
20047
20048 2011-07-12  Adam Roben  <aroben@apple.com>
20049
20050         Teach TestFailures to recognize when run-webkit-tests gets killed by buildbot
20051
20052         Fixes <http://webkit.org/b/64358> TestFailures page thinks all tests passed in
20053         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14672
20054
20055         Reviewed by Daniel Bates.
20056
20057         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
20058         (Builder.prototype.getNumberOfFailingTests): If run-webkit-tests exited with a non-zero
20059         exit status but we didn't find any failure counts, assume that there was some error that
20060         caused run-webkit-tests to die early (like being killed by buildbot due to a timeout).
20061
20062         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
20063         Added a new test that shows that we get a failingTestCount of -1 when run-webkit-tests dies
20064         early.
20065         (runGetNumberOfFailingTestsTest): Moved most code here from the only pre-existing test in this
20066         file.
20067
20068         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
20069         (LayoutTestResultsLoader.prototype.start): Bump the cache number so old cached data that was
20070         tainted by the bug fixed in this patch will be evicted.
20071
20072 2011-07-12  Adam Barth  <abarth@webkit.org>
20073
20074         cr-linux-ews complains about tests that aren't actually failing
20075         https://bugs.webkit.org/show_bug.cgi?id=64367
20076
20077         The underlying problem here is that full_results.json doesn't have
20078         enough information to determine whether a given test result was
20079         expected because whether an actual result is expected depends on
20080         whether full_results.json was generated duing a test run that included
20081         pixel tests.
20082
20083         The right long-term solution is to make full_results.json a complete
20084         description of what happened durning a test run.  In this patch
20085         however, to stop the spam, this patch makes the jsonresultsparser
20086         ignore pixel failures.
20087
20088         (I'm landing this patch unreviewed in order to stop the bots from
20089         spamming.  I'll happily address any review comments in a follow-up
20090         patch.)
20091
20092         * Scripts/webkitpy/common/net/resultsjsonparser.py:
20093         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
20094
20095 2011-07-12  Adam Barth  <abarth@webkit.org>
20096
20097         Re-work garden-o-matic UI to begin to look like the mocks
20098         https://bugs.webkit.org/show_bug.cgi?id=64334
20099
20100         Reviewed by Dimitri Glazkov.
20101
20102         This patch changes how we display failures to be more like the mock the
20103         dglazkov made. We now can display the same information much more
20104         compactly.
20105
20106         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
20107         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js:
20108         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
20109         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.css: Added.
20110         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
20111         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
20112         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
20113         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
20114         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
20115         * Scripts/webkitpy/tool/servers/gardeningserver.py:
20116
20117 2011-07-11  Adam Roben  <aroben@apple.com>
20118
20119         Extract TestFailures's bug-filing code into two new classes
20120
20121         Fixes <http://webkit.org/b/64300> TestFailures page's new-bug-filing code is a mess!
20122
20123         Reviewed by Darin Adler and Adam Barth.
20124
20125         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js:
20126         (Buildbot.prototype.resultsDirectoryURL): Changed to return a URI-encoded URL. Otherwise the
20127         URL isn't valid (and it's harder to mock this function).
20128
20129         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot_unittests.js:
20130         Added. Just tests the above change (for now).
20131
20132         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
20133         Added a license header, enclosed everything in a closure to avoid polluting the global
20134         namespace, changed the test name to actually describe the passing condition, and replaced
20135         uses of equals() with equal(). (The latter is the actual name of the function, and matches
20136         deepEqual, etc.)
20137
20138         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm.js: Added.
20139         (NewBugForm): This class knows how to construct a <form> used for filing new bugs in
20140         Bugzilla based on some parameters.
20141         (NewBugForm.prototype.domElement): Creates and returns the <form> element. Code came from
20142         ViewController.prototype._domForNewAndExistingBugs.
20143
20144         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NewBugForm_unittests.js:
20145         Added. Tests the above code.
20146
20147         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm.js: Added.
20148         (TestFailureBugForm): This class knows how to construct a <form> element used for filing new
20149         bugs specifically about test failures. Code came from
20150         ViewController.prototype._domForNewAndExistingBugs.
20151         (TestFailureBugForm.prototype.domElement): Slightly customizes the <form> element returned
20152         by NewBugForm.
20153         (TestFailureBugForm.prototype._computeOperatingSystem):
20154         (TestFailureBugForm.prototype._computePlatform):
20155         (TestFailureBugForm.prototype._createBugTitle):
20156         (TestFailureBugForm.prototype._failingResultsHTMLURL):
20157         (TestFailureBugForm.prototype._failingRevision):
20158         (TestFailureBugForm.prototype._passingRevision):
20159         (TestFailureBugForm.prototype._regressionRangeString):
20160         Code came from ViewController.prototype._domForNewAndExistingBugs. I broke it out into
20161         separate functions to break up the rat's nest a bit.
20162
20163         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailureBugForm_unittests.js:
20164         Added. Tests the above code.
20165
20166         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
20167         (ViewController.prototype._domForNewAndExistingBugs): Moved code from here to
20168         TestFailureBugForm/NewBugForm, and changed this code to use a TestFailureBugForm.
20169
20170         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
20171         NewBugForm/TestFailureBugForm.js.
20172
20173         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
20174         Added new tests and required files.
20175
20176 2011-07-12  Adam Roben  <aroben@apple.com>
20177
20178         Test that no intermediate WTF::Strings are created when concatenating with string literals
20179
20180         Test for <http://webkit.org/b/63330> Concatenating string literals and WTF::Strings using
20181         operator+ is suboptimal
20182
20183         Reviewed by Darin Adler.
20184
20185         * TestWebKitAPI/Tests/WTF/StringOperators.cpp: Added.
20186         (TestWebKitAPI::TEST): Test that a bunch of different string concatenation expressions don't
20187         create any intermediate WTF::Strings while they're being evaluated.
20188
20189         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
20190         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
20191         Added new file.
20192
20193 2011-07-12  Eric Seidel  <eric@webkit.org>
20194
20195         [Qt] NRWT should pick up the right httpd config file
20196         https://bugs.webkit.org/show_bug.cgi?id=64086
20197
20198         Reviewed by Adam Barth.
20199
20200         This is more fall-out from the Port class inappropriately encapsulating
20201         both platform and port knowledge.  We need to split out some of this
20202         platform knowledge into a separate class which can be better shared
20203         between ports.
20204
20205         The fix was to move all the _path_to_apache_config_file logic down
20206         into the WebKitPort baseclass so that all of the ports can find the
20207         right apache config file, regardless of what platform they may be running on.
20208
20209         I did not move this down into base.Port (even though I considered it).
20210         Chromium duplicates some of this logic, but since they have separate
20211         subclasses for each port-OS combination (e.g. ChromiumMac, ChromiumWin)
20212         they wouldn't notice the change.  Eventually we'll move this logic
20213         out of Port entirely, and then it will be shared by all ports.
20214
20215         I also cleaned up the http-config logic for ORWT while I was there,
20216         although since we're killing that code, I'm happy to revert that part
20217         of the change if changing it is viewed as needlessly risky.
20218
20219         During this cleanup, I noticed that no ports use the "default" httpd.conf
20220         which is left over from Mac OS X Tiger (and old linux distros) which used
20221         Apache 1.3.  I've removed httpd.conf (and associated support in ORWT) since
20222         we no longer support any ports which use this httpd.conf.
20223
20224         * Scripts/webkitperl/httpd.pm:
20225         (getHTTPDConfigPathForTestDirectory):
20226         * Scripts/webkitpy/layout_tests/port/base.py:
20227         * Scripts/webkitpy/layout_tests/port/gtk.py:
20228         * Scripts/webkitpy/layout_tests/port/mac.py:
20229         * Scripts/webkitpy/layout_tests/port/qt.py:
20230         * Scripts/webkitpy/layout_tests/port/webkit.py:
20231         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
20232         * Scripts/webkitpy/layout_tests/port/win.py:
20233         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
20234
20235 2011-07-11  Hironori Bono  <hbono@chromium.org>
20236
20237         Reviewed by Adam Roben.
20238
20239         Implement layoutTestController.setTextDirection for WebKit2 and Windows.
20240         https://bugs.webkit.org/show_bug.cgi?id=61931
20241
20242         This change implements layoutTextController.setTextDirection for WebKit2
20243         and Windows so we can run a layout test added by r87770 on them. Each
20244         implementation adds a wrapper function for Editor::setBaseWritingDirection()
20245         so LayoutTestController can call it.
20246
20247         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
20248         (LayoutTestController::setTextDirection): Implemented the binding function
20249         for layoutTestController.setTextDirection (Windows).
20250         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
20251         Added a binding function for layoutTestController.setTextDirection (WebKit2).
20252         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp: ditto.
20253         (WTR::LayoutTestController::setTextDirection):
20254         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: ditto.
20255
20256 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
20257
20258         [EFL] ImageDiff: Make sure gEcoreEvas is destroyed at the right time.
20259         https://bugs.webkit.org/show_bug.cgi?id=64293
20260
20261         Reviewed by Kent Tamura.
20262
20263         gEcoreEvas was being destroyed after shutdownEfl() was called, so the
20264         call to ecore_evas_free() failed.
20265
20266         We now shut it down manually before shutdownEfl() is called.
20267
20268         * DumpRenderTree/efl/ImageDiff.cpp:
20269         (main): Destroy gEcoreEvas before shutting down the EFL.
20270
20271 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
20272
20273         Make Chromium Mac and Linux tests bot core builders
20274         https://bugs.webkit.org/show_bug.cgi?id=63196
20275
20276         Reviewed by Adam Barth.
20277
20278         Added Chromium Mac and Linux Release Tests bots core builders.
20279
20280         Also rearranged the order of bots so that test bots show up right next to
20281         respective builders.
20282
20283         * BuildSlaveSupport/build.webkit.org-config/config.json:
20284         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
20285         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
20286
20287 2011-07-11  Raphael Kubo da Costa  <kubo@profusion.mobi>
20288
20289         [EFL] DRT: Add code related to font management.
20290         https://bugs.webkit.org/show_bug.cgi?id=63989
20291
20292         Reviewed by Kent Tamura.
20293
20294         The added files are responsible for managing fontconfig and adding the
20295         fonts required by some layout tests to fontconfig's path.
20296
20297         * DumpRenderTree/efl/FontManagement.cpp: Added.
20298         (getFontDirectories):
20299         (getFontFiles):
20300         (addFontDirectories):
20301         (addFontFiles):
20302         (addFontsToEnvironment):
20303         * DumpRenderTree/efl/FontManagement.h: Added.
20304
20305 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
20306
20307         nrwt: linting fixes
20308         https://bugs.webkit.org/show_bug.cgi?id=64225
20309
20310         Reviewed by Eric Siedel.
20311
20312         Miscellaneous linting fixes. The most notable change is that
20313         we add public attributes for user, executive, filesystem, and
20314         options on the Port object, so we don't have to refer to the
20315         "protected" versions all over the place".
20316
20317         * Scripts/webkitpy/layout_tests/controllers/manager.py:
20318         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
20319         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
20320         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py:
20321         * Scripts/webkitpy/layout_tests/controllers/worker.py:
20322         * Scripts/webkitpy/layout_tests/port/base.py:
20323         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
20324
20325 2011-07-11  Ryosuke Niwa  <rniwa@webkit.org>
20326
20327         webkit-patch roll-chromium-deps no longer works
20328         https://bugs.webkit.org/show_bug.cgi?id=64324
20329
20330         Reviewed by James Robinson.
20331
20332         Lower the minimum commit message length to 10.
20333
20334         * Scripts/webkitpy/tool/steps/commit.py:
20335
20336 2011-07-11  Alice Boxhall  <aboxhall@chromium.org>
20337
20338         Convert json_results_generator.py to output version 4 JSON.
20339         https://bugs.webkit.org/show_bug.cgi?id=60869
20340
20341         Reviewed by Ojan Vafai.
20342
20343         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
20344         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
20345         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
20346
20347 2011-07-11  Martin Robinson  <mrobinson@igalia.com>
20348
20349         [GTK] [NRWT] Pixel tests do not work
20350         https://bugs.webkit.org/show_bug.cgi?id=64091
20351
20352         Reviewed by Eric Seidel.
20353
20354         Fix pixel test runs for GTK+ new-run-webkit-tests by exposing the
20355         path to the ImageDiff binary.
20356
20357         * Scripts/webkitpy/layout_tests/port/gtk.py: Implement _path_to_image_diff.
20358
20359 2011-07-11  Dirk Pranke  <dpranke@chromium.org>
20360
20361         Fix hang on win32 in a multiprocessing test that shouldn't have been running.
20362
20363         Unreviewed, build fix.
20364
20365         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
20366
20367 2011-07-11  Adam Roben  <aroben@apple.com>
20368
20369         Don't count leaks as test failures on TestFailures's front page
20370
20371         As a bonus, this patch adds our first unit test.
20372
20373         Fixes <http://webkit.org/b/64303> REGRESSION (r90489): TestFailures page says far more tests
20374         are failing on the Leaks bot than actually are
20375
20376         Reviewed by Adam Barth.
20377
20378         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
20379         (Builder.prototype.getNumberOfFailingTests): Exclude lines containing "leak" when summing
20380         failure counts.
20381
20382         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder_unittests.js:
20383         Added. Contains a simple test to show that leaks aren't counted as test failures.
20384
20385         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/run-unittests.html:
20386         Added. This is the test harness.
20387
20388 2011-06-14  Mike Stegeman  <mrstegeman@gmail.com>
20389
20390         [GTK] Add API to allow setting local storage database path
20391         https://bugs.webkit.org/show_bug.cgi?id=62091
20392
20393         Reviewed by Martin Robinson and Gustavo Noronha.
20394
20395         Expose the path of the localStorage databases through a setting
20396         to allow HTML5 localStorage to be persistent. New setting is:
20397         html5-local-storage-database-path
20398
20399         * DumpRenderTree/gtk/DumpRenderTree.cpp:
20400         (resetDefaultsToConsistentValues):
20401
20402 2011-07-11  Adam Barth  <abarth@webkit.org>
20403
20404         garden-o-matic should be able to rebaseline tests
20405         https://bugs.webkit.org/show_bug.cgi?id=64186
20406
20407         Reviewed by Eric Seidel.
20408
20409         This patch contains a simple HTTP server binding for the
20410         rebaseline-test command.  The frontend will appear in another patch.
20411
20412         * Scripts/webkitpy/tool/servers/gardeningserver.py:
20413         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
20414         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
20415
20416 2011-07-11  Adam Barth  <abarth@webkit.org>
20417
20418         Add some basic UI for showing regression ranges
20419         https://bugs.webkit.org/show_bug.cgi?id=64243
20420
20421         Reviewed by Eric Seidel.
20422
20423         This UI is all temporary.  It exists mostly just to have some buttons
20424         to click to exercise the code.
20425
20426         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
20427         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
20428         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
20429         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
20430         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
20431
20432 2011-07-11  Adam Barth  <abarth@webkit.org>
20433
20434         garden-o-matic should be able to determine which revisions caused a given failure
20435         https://bugs.webkit.org/show_bug.cgi?id=64189
20436
20437         Reviewed by Adam Roben.
20438
20439         Walking the failure history looking for failures turns out to be
20440         slightly tricky because the network requests are asynchronous.
20441         Currently we do all the fetches serially and our cache is unbounded.
20442         We'll probably optimize both those parameters eventually.
20443
20444         This patch also generalizes some functionality in the unit testing
20445         framework to make testing this sort of code easier.
20446
20447         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
20448         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
20449         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
20450
20451 2011-07-11  Adam Barth  <abarth@webkit.org>
20452
20453         Add a webkit-patch command for rebaselining an individual test
20454         https://bugs.webkit.org/show_bug.cgi?id=64246
20455
20456         Reviewed by Eric Seidel.
20457
20458         This patch introduces a command that's able to rebaseline a single
20459         test.  Currently, the command works only with the build.chromium.org
20460         buildbots, but extending it to work with the build.webkit.org bots
20461         shouldn't be that hard.
20462
20463         A complete rebaseling tool should also include an "optimize baselines"
20464         command (which moves/deletes baselines in order to reduce the number of
20465         expected results files), but that will come in a future patch.
20466
20467         Really BuilderToPort should be merged into builders.py, but I'm going
20468         to save that for a future patch as well.  (We need to stop shaving yaks
20469         at some point.)
20470
20471         * Scripts/webkitpy/tool/commands/rebaseline.py:
20472
20473 2011-07-11  Adam Barth  <abarth@webkit.org>
20474
20475         garden-o-matic should be able to roll out patches
20476         https://bugs.webkit.org/show_bug.cgi?id=64185
20477
20478         Reviewed by Eric Seidel.
20479
20480         This gardening server API is simply a binding to the rollout machinery
20481         already present in webkit-patch.
20482
20483         * Scripts/webkitpy/tool/commands/gardenomatic.py:
20484         * Scripts/webkitpy/tool/servers/gardeningserver.py:
20485         * Scripts/webkitpy/tool/servers/gardeningserver_unittest.py: Added.
20486
20487 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
20488
20489         [Qt][Mac] Unreviewed fix after r90746.
20490
20491         * Scripts/run-webkit-tests:
20492         (useNewRunWebKitTests): Really disable NRWT for qt-mac platform.
20493
20494 2011-07-11  Adam Roben  <aroben@apple.com>
20495
20496         Teach TestFailures that ORWT's results.html file might be missing due to all tests passing
20497
20498         There are three reasons why we might fail to fetch ORWT's results.html:
20499           - All tests passed, so no results.html was generated
20500           - Some error during the test run caused results.html not to be generated (e.g., ORWT
20501             timed out)
20502           - Some network error occurred when fetching results.html
20503
20504         We were failing to account for the first possibility in some cases. For test runs before
20505         r89610, we first check build.webkit.org/json to determine how many tests failed and whether
20506         ORWT exited early due to too many failures; if all tests passed then we don't fetch
20507         results.html at all. r89610 changed ORWT to put information in results.html about exiting
20508         early due to too many failures, so we no longer needed to check build.webkit.org/json to get
20509         that information, and in r89619 I changed TestFailures to do just that. But I forgot that we
20510         still needed to check build.webkit.org/json to find out if all tests passed!
20511
20512         Now, for test runs after r89610, we check results.html first, and then check
20513         build.webkit.org/json if we fail to fetch results.html. This lets us distinguish between all
20514         tests passing and the error cases.
20515
20516         Fixes <http://webkit.org/b/64280> TestFailures page can't pinpoint that r90699 caused 13
20517         tests to fail on Windows 7 Release (WebKit2 Tests)
20518
20519         Reviewed by David Kilzer.
20520
20521         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
20522         (LayoutTestResultsLoader.prototype.start): Bumped the cache version so that old, buggy
20523         cached data will get evicted. We were marking builds where all tests passed as errors!
20524         (LayoutTestResultsLoader.prototype._fetchAndParseORWTResults): Added success/error callback
20525         parameters to the fetchAndParseResultsHTML helper function, and added a similar
20526         fetchNumberOfFailingTests function that fetches data from build.webkit.org/json (code came
20527         from later in the function). For test runs before r89610, we first check
20528         build.webkit.org/json then check results.html. For builds after r89610, we first check
20529         results.html and then check build.webkit.org/json if we couldn't fetch results.html.
20530
20531 2011-07-11  Csaba Osztrogonác  <ossy@webkit.org>
20532
20533         new-run-webkit-tests does not support qt-arm or qt-4.8 results
20534         https://bugs.webkit.org/show_bug.cgi?id=64071
20535
20536         [Qt] NRWT should pick up the right httpd config file
20537         https://bugs.webkit.org/show_bug.cgi?id=64086
20538
20539         Reviewed by Zoltán Herczeg.
20540
20541         * Scripts/run-webkit-tests: Disable NRWT for qt-mac, qt-arm and qt-4.8 platforms until fix.
20542         (useNewRunWebKitTests):
20543
20544 2011-07-11  Kenichi Ishibashi  <bashi@chromium.org>
20545
20546         Add TestNetscapePlugIn/Tests/FormValue.cpp to DRT build files
20547         https://bugs.webkit.org/show_bug.cgi?id=64248
20548
20549         Reviewed by Kent Tamura.
20550
20551         * DumpRenderTree/DumpRenderTree.gypi: Added FormValue.cpp.
20552         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Ditto.
20553         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Ditto.
20554         * GNUmakefile.am: Ditto.
20555
20556 2011-07-10  Kenichi Ishibashi  <bashi@chromium.org>
20557
20558         Let plugins participate in form submission
20559         https://bugs.webkit.org/show_bug.cgi?id=13061
20560
20561         Adds NPPVformValue support to the testing plugin.
20562
20563         Reviewed by Anders Carlsson.
20564
20565         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added FormValue.cpp
20566         * DumpRenderTree/TestNetScapePlugIn/Tests/FormValue.cpp: Added.
20567         (FormValue::FormValue): Ditto.
20568         (FormValue::NPP_GetValue): Ditto.
20569
20570 2011-07-11  Roland Steiner  <rolandsteiner@chromium.org>
20571
20572         Reviewed by Tony Chang.
20573
20574         Complete functions in filesystem.py
20575         https://bugs.webkit.org/show_bug.cgi?id=63528
20576
20577         - missing text file functions added
20578         - functions sorted
20579         - removed 'append' optional parameters (were unused)
20580         - adapted filesystem_mock in the same way
20581
20582         * Scripts/webkitpy/common/system/filesystem.py:
20583         * Scripts/webkitpy/common/system/filesystem_mock.py:
20584         * Scripts/webkitpy/common/system/filesystem_unittest.py:
20585
20586 2011-07-10  Adam Barth  <abarth@webkit.org>
20587
20588         Clean up style in fallback path calculation
20589         https://bugs.webkit.org/show_bug.cgi?id=64239
20590
20591         Reviewed by Daniel Bates.
20592
20593         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
20594         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
20595         * Scripts/webkitpy/layout_tests/port/mac.py:
20596         * Scripts/webkitpy/layout_tests/port/win.py:
20597
20598 2011-07-10  Balazs Kelemen  <kbalazs@webkit.org>
20599
20600         [Qt][WK2] Implement the rest of WTR::PlatformWebView
20601         https://bugs.webkit.org/show_bug.cgi?id=63630
20602
20603         Reviewed by Kenneth Rohde Christiansen.
20604
20605         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
20606         (WTR::PlatformWebView::windowFrame):
20607         (WTR::PlatformWebView::setWindowFrame):
20608
20609 2011-07-10  Mark Rowe  <mrowe@apple.com>
20610
20611         Teach build-webkit how to find the latest WebKitSystemInterface binary.
20612
20613         * Scripts/build-webkit:
20614
20615 2011-07-09  Adam Roben  <aroben@apple.com>
20616
20617         Teach TestFailures to abbreviate the examples of test flakiness
20618
20619         These lists can get quite long, and it's not really helpful in most cases to have soooooo
20620         many examples of flakiness.
20621
20622         Fixes <http://webkit.org/b/64203> Lists of flaky revisions on TestFailures page can get so
20623         long they're hard to navigate
20624
20625         Reviewed by Dan Bates.
20626
20627         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
20628         (FlakyLayoutTestDetector.prototype.flakinessExamples): If we have more than a certain number
20629         of examples, replace the middle items with a separator.
20630
20631         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
20632         (.flakiness-example-separator): Added styles for the separator.
20633
20634         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
20635         (ViewController.prototype._domForPossiblyFlakyTests): Use a vertical ellipsis to represent
20636         the separator.
20637
20638 2011-07-09  Dirk Pranke  <dpranke@chromium.org>
20639
20640         nrwt: stack traces from worker-side exceptions aren't very useful inside test-webkitpy
20641         https://bugs.webkit.org/show_bug.cgi?id=64218
20642
20643         Reviewed by Eric Seidel.
20644
20645         Exceptions aren't picklable and can't be sent across the
20646         manager/worker message queue without losing information. NRWT
20647         handles this by turning the stack trace into a set of strings,
20648         and logging the strings when we receive an exception from the
20649         worker. However, when you are running tests and something
20650         crashes on the worker side, test-webkitpy prints the
20651         manager-side stack trace, which is just confusing and useless.
20652
20653         This patch changes the logic so that exceptions are passed
20654         through as-is when the worker and manager are in the same
20655         process (the --worker-model=inline option). This increases the
20656         code paths slightly but makes crashes much more useful.
20657
20658         * Scripts/webkitpy/layout_tests/controllers/manager.py:
20659         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
20660         * Scripts/webkitpy/layout_tests/controllers/message_broker.py:
20661         * Scripts/webkitpy/layout_tests/controllers/worker.py:
20662         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
20663
20664 2011-07-09  Darin Fisher  <darin@chromium.org>
20665
20666         Eliminate bad dependency on gfx::Point.
20667         https://bugs.webkit.org/show_bug.cgi?id=64228
20668
20669         Reviewed by Kent Tamura.
20670
20671         * DumpRenderTree/chromium/EventSender.cpp:
20672         (initMouseEvent):
20673
20674 2011-07-08  Ojan Vafai  <ojan@chromium.org>
20675
20676         pull static dashboard files into the appengine server from the chromium repository
20677         https://bugs.webkit.org/show_bug.cgi?id=64208
20678
20679         Reviewed by Dirk Pranke.
20680
20681         These files belong in the WebKit repo since they are tied primarily to the webkit tests.
20682         They have some extra bits to support chromium's gtests, but that seems fine.
20683         Mainly, this will allow other WebKit hackers to hack on the dashboards.
20684
20685         As a nice side-effect, we can now push the dashboard files when we do appengine pushes
20686         instead of the weird thing we used to do of pulling them from the Chromium repository
20687         and storing them in the appengine datastore. This allows for cleaning up a lot of code
20688         and will likely make the dashboards load a bit faster.
20689
20690         The new JS files don't fully match WebKit style, but I'd like to clean that up in a
20691         followup patch if possible to maintain my sanity with this patch.
20692
20693         * TestResultServer/app.yaml:
20694         * TestResultServer/handlers/dashboardhandler.py: Removed.
20695         * TestResultServer/handlers/menu.py:
20696         * TestResultServer/main.py:
20697         * TestResultServer/model/dashboardfile.py: Removed.
20698         * TestResultServer/static-dashboards/LICENSE.dygraph.txt: Added.
20699         * TestResultServer/static-dashboards/README.dygraph.txt: Added.
20700         * TestResultServer/static-dashboards/README.webtreemap.txt: Added.
20701         * TestResultServer/static-dashboards/aggregate_results.html: Added.
20702         * TestResultServer/static-dashboards/builders.js: Added.
20703         * TestResultServer/static-dashboards/dashboard_base.js: Added.
20704         * TestResultServer/static-dashboards/dygraph-combined.js: Added.
20705         * TestResultServer/static-dashboards/flakiness_dashboard.html: Added.
20706         * TestResultServer/static-dashboards/flakiness_dashboard_tests.js: Added.
20707         * TestResultServer/static-dashboards/timeline_explorer.html: Added.
20708         * TestResultServer/static-dashboards/treemap.html: Added.
20709         * TestResultServer/static-dashboards/webtreemap.css: Added.
20710         * TestResultServer/static-dashboards/webtreemap.js: Added.
20711         * TestResultServer/stylesheets/dashboardfile.css: Removed.
20712         * TestResultServer/stylesheets/menu.css:
20713         * TestResultServer/templates/dashboardfilelist.html: Removed.
20714         * TestResultServer/templates/menu.html:
20715
20716 2011-07-08  Adam Roben  <aroben@apple.com>
20717
20718         Make TestFailures's list of flaky tests look more like the list of non-flaky tests
20719
20720         Fixes <http://webkit.org/b/64204> TestFailures page's flaky tests list is ugly!
20721
20722         Reviewed by Daniel Bates.
20723
20724         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
20725         (#failure-history, #possibly-flaky-tests): Expanded this rule to apply to the list of flaky
20726         tests.
20727         (#failure-history > li, #possibly-flaky-tests > li): Ditto, but moved the 50px left padding
20728         from here...
20729         (#failure-history > li): ...to here.
20730
20731         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
20732         (ViewController.prototype._domForPossiblyFlakyTests): Give the list an id attribute for
20733         styling purposes.
20734
20735 2011-07-08  Jeffrey Pfau  <jpfau@apple.com>
20736
20737         Unreviewed, add myself as committer.
20738
20739         * Scripts/webkitpy/common/config/committers.py:
20740
20741 2011-07-08  Adam Barth  <abarth@webkit.org>
20742
20743         TestResultsServer should keep old test results
20744         https://bugs.webkit.org/show_bug.cgi?id=64199
20745
20746         Reviewed by Ojan Vafai.
20747
20748         Having historical data will help us do failure archeology.
20749
20750         * TestResultServer/handlers/testfilehandler.py:
20751         * TestResultServer/model/testfile.py:
20752         * TestResultServer/templates/showfilelist.jsonp: Added.
20753
20754 2011-07-08  Adam Roben  <aroben@apple.com>
20755
20756         Remove commit-log-editor's dependency on Module::Load::Conditional
20757
20758         This module isn't available in Perl 5.8.8 (the version used on Leopard).
20759
20760         Fixes <http://webkit.org/b/64198> REGRESSION (r90583):
20761         webkitpy.common.checkout.checkout_unittest failing on Leopard
20762
20763         Reviewed by Daniel Bates.
20764
20765         * Scripts/commit-log-editor: Use the new loadTermReadKey() function instead of
20766         Module::Load::Conditional::can_load.
20767         (loadTermReadKey): Added. Tries to load Term::ReadKey and returns true if it's successful.
20768
20769 2011-07-08  Adam Roben  <aroben@apple.com>
20770
20771         Make checkout_unittest more robust against files moving around
20772
20773         Fixes <http://webkit.org/b/64197> checkout_unittest contains ugly, fragile code to find the
20774         Scripts directory
20775
20776         Reviewed by Adam Barth.
20777
20778         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
20779         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Instantiate a real SCM
20780         object and use it to get the path to the Scripts directory, rather than hard-coding the
20781         relative path from this file to Scripts.
20782
20783 2011-07-08  Adam Roben  <aroben@apple.com>
20784
20785         Make TestFailures remember that run-webkit-tests timed out, even across reloads
20786
20787         Fixes <http://webkit.org/b/64193> TestFailures page incorrectly thinks all tests passed in
20788         http://build.webkit.org/builders/Windows%207%20Release%20(Tests)/builds/14589 after a reload
20789
20790         Reviewed by Daniel Bates.
20791
20792         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
20793         (LayoutTestResultsLoader.prototype.start): Store an "error" attribute in the cached data.
20794         When true, it indicates that there was an error retrieving the results for this build and
20795         that the errorCallback should be called.
20796
20797 2011-07-08  Adam Barth  <abarth@webkit.org>
20798
20799         Teach garden-o-matic how to display test results
20800         https://bugs.webkit.org/show_bug.cgi?id=64141
20801
20802         Reviewed by Ojan Vafai.
20803
20804         This patch includes basic infrastructure for probing build.chromium.org
20805         for test results.  We only handle text and image tests, not anything
20806         complicated like reftests.  Also, we're using the revision/build
20807         independent results store on the server, so we're avoiding that
20808         complication for now.
20809
20810         It's slightly hacky that we need to probe the server to see what kinds
20811         of results exist.  A better solution would be to add CORS support to
20812         the server or to use the local server to help.
20813
20814         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
20815         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
20816         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
20817         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
20818         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js:
20819         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js:
20820
20821 2011-07-08  Dirk Pranke  <dpranke@chromium.org>
20822
20823         REGRESSION(90419) NRWT's httpd locking is broken for --child-processes=1
20824         https://bugs.webkit.org/show_bug.cgi?id=64092
20825
20826         Reviewed by Tony Chang.
20827
20828         The code was incorrectly creating empty shards in the
20829         shard_in_two case.
20830
20831         * Scripts/webkitpy/layout_tests/controllers/manager.py:
20832         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
20833
20834 2011-07-08  Adam Roben  <aroben@apple.com>
20835
20836         Teach buildbot to figure out how many webkitpy/webkitperl tests failed
20837
20838         Fixes <http://webkit.org/b/64192> It's hard to tell how many test-webkitpy/test-webkitperl
20839         tests failed when looking at build.webkit.org
20840
20841         Reviewed by Eric Seidel.
20842
20843         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
20844         (TestWithFailureCount): New class that represents a test build step which has an associated
20845         failure count. Eventually we should move more of our test classes to deriving from this.
20846         (TestWithFailureCount.countFailures): Method for subclasses to override to say how many
20847         failures occurred.
20848
20849         (TestWithFailureCount.commandComplete):
20850         (TestWithFailureCount.evaluateCommand):
20851         (TestWithFailureCount.getText):
20852         (TestWithFailureCount.getText2):
20853         These were all based on RunGtkAPITests.
20854
20855         (RunPythonTests): Changed to inherit from TestWithFailureCount.
20856         (RunPythonTests.countFailures): Parses the test-webkitpy output looking for the count of
20857         failures.
20858         (RunPerlTests): Changed to inherit from TestWithFailureCount.
20859         (RunPerlTests.countFailures): Parses the test-webkitperl output looking for the count of
20860         failures.
20861
20862 2011-07-08  Adam Roben  <aroben@apple.com>
20863
20864         Ensure $CHANGE_LOG_EMAIL_ADDRESS is set when testing webkitpy's commit-log-editor integration
20865
20866         Fixes <http://webkit.org/b/64180> REGRESSION (r90564): test-webkitpy failing on multiple
20867         bots due to commit-log-editor errors
20868
20869         Reviewed by Adam Barth.
20870
20871         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
20872         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Set
20873         $CHANGE_LOG_EMAIL_ADDRESS to the patch author's email address. This ensures that
20874         commit-log-editor can find a value for the committer's email, and that the committer and
20875         author email addresses match, which will prevent commit-log-editor from inserting a "Patch
20876         by" line in the commit message.
20877
20878 2011-07-08  Andreas Kling  <kling@webkit.org>
20879
20880         [Qt][WK2] Views should expose QActions for basic navigation.
20881         https://bugs.webkit.org/show_bug.cgi?id=64174
20882
20883         Reviewed by Benjamin Poulain.
20884
20885         Bring the toolbar in Qt's MiniBrowser back to life.
20886
20887         * MiniBrowser/qt/BrowserView.cpp:
20888         (BrowserView::navigationAction):
20889         * MiniBrowser/qt/BrowserView.h:
20890         * MiniBrowser/qt/BrowserWindow.cpp:
20891         (BrowserWindow::BrowserWindow):
20892
20893 2011-07-08  Adam Roben  <aroben@apple.com>
20894
20895         Teach TestFailures how to find test names in commit-log-editor-style commit messages
20896
20897         TestFailures was relying on Trac turning the list of modified files in our commit messages
20898         into an HTML list. But Trac only does this when the list of modified files is indented.
20899         commit-log-editor doesn't indent the file list, so the list wasn't being turned into an HTML
20900         list, which was confusing TestFailures.
20901
20902         TestFailures now does much simpler parsing of the commit message (i.e., just a substring
20903         search) without relying at all on its structure.
20904
20905         Fixes <http://webkit.org/b/64173> TestFailures page fails to blame r90608 for breaking
20906         fast/dom/HTMLProgressElement/progress-element-markup.html on Windows
20907
20908         Reviewed by David Kilzer.
20909
20910         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
20911         (Trac.prototype.getCommitDataForRevisionRange): Instead of trying to parse the commit
20912         message, just return its text.
20913
20914         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
20915         (ViewController.prototype._domForRegressionRange): Instead of searching for test names in
20916         each commit's list of modified files, just search for test names anywhere in the commit's
20917         message.
20918
20919 2011-07-08  Adam Barth  <abarth@webkit.org>
20920
20921         sheriffbot is too spammy in IRC
20922         https://bugs.webkit.org/show_bug.cgi?id=64153
20923
20924         Reviewed by Eric Seidel.
20925
20926         Reporting failures in IRC worked well when the tree was greener than it
20927         is today.  Nowadays, this feature mostly just results in spam about
20928         false positives.  If we reach a greener state, we can bring this code
20929         back.
20930
20931         (Another possibility is to restrict this feature to builder bots,
20932         rather than including testers, as we did before this patch.)
20933
20934         * Scripts/webkitpy/tool/commands/sheriffbot.py:
20935         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
20936
20937 2011-07-08  Adam Roben  <aroben@apple.com>
20938
20939         Ensure commit-log-editor adds a "Patch by" line when the author and committer are different
20940
20941         Previously we were only adding a "Patch by" line when the ChangeLog contained a "Reviewed
20942         by" line. But some patches (like rollout patches) don't contain that line. Now we always add
20943         "Patch by" to the commit log regardless of the ChangeLog's contents.
20944
20945         Fixes <http://webkit.org/b/64127> Committer for r90588 is commit-queue@webkit.org, but
20946         should have been abarth@webkit.org
20947
20948         Reviewed by Anders Carlsson.
20949
20950         * Scripts/commit-log-editor:
20951         (createCommitMessage): Try to put the "Patch by" line just above the "Reviewed by" line, as
20952         before. If there is no "Reviewed by" line, try to put it just above the first modified file.
20953         If all else fails, put it at the end of the commit message.
20954         (patchAuthorshipString): Added. Code came from createCommitMessage.
20955
20956 2011-07-08  Herczeg Zoltan  <zherczeg@webkit.org>
20957
20958         Reviewed by Andreas Kling.
20959
20960         Adding myself as a reviewer.
20961
20962         * Scripts/webkitpy/common/config/committers.py:
20963
20964 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
20965
20966         [wx] Unreviewed build fix, remove old files from the tree and update paths to new ones.
20967
20968         * wx/build: Removed.
20969         * wx/build/build_utils.py: Removed.
20970         * wx/build/settings.py: Removed.
20971         * wx/build/waf_extensions.py: Removed.
20972         * wx/build/wxpresets.py: Removed.
20973         * wx/packaging/build-mac-installer.py:
20974
20975 2011-07-07  Kevin Ollivier  <kevino@theolliviers.com>
20976
20977         [wx] Unreviewed build fix, more fixes for install name issues.
20978
20979         * wx/packaging/build-mac-installer.py:
20980
20981 2011-07-07  Dirk Pranke  <dpranke@chromium.org>
20982
20983         test-webkitpy fails on chromium win
20984         https://bugs.webkit.org/show_bug.cgi?id=64137
20985
20986         Reviewed by Eric Seidel.
20987
20988         test-webkitpy doesn't play nicely with the multiprocessing
20989         module on win32. An earlier change actually reenabled the tests
20990         on win32 by mistake.
20991
20992         This patch also fixes a few cases where path names will trip
20993         things up on win32 (e.g., by testing unix paths on windows).
20994         We do not lose any real coverage here with those fixes.
20995
20996         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
20997         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
20998         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
20999         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
21000
21001 2011-07-07  Leandro Pereira  <leandro@profusion.mobi>
21002
21003         [ImageDiff] Calculate/print difference right after reading baseline image.
21004         https://bugs.webkit.org/show_bug.cgi?id=64117
21005
21006         Reviewed by Kent Tamura.
21007
21008         Otherwise, ImageDiff will block on fgets() until the universe ends or you ^C
21009         it.  Whichever happens first.
21010
21011         * DumpRenderTree/efl/ImageDiff.cpp:
21012         (main):
21013
21014 2011-07-07  Raphael Kubo da Costa  <kubo@profusion.mobi>
21015
21016         [EFL] DRT: Leak cairo_t to make the pixel tests stop crashing
21017         https://bugs.webkit.org/show_bug.cgi?id=64107
21018
21019         Reviewed by Kent Tamura.
21020
21021         The pointer is later managed by BitmapContext, but it was being
21022         de-refed and deleted earlier by our RefPtr, causing crashes in all
21023         pixel tests.
21024
21025         We now call leakRef() to make sure it is not removed when our
21026         createBitmapContextFromWebView goes out of context.
21027
21028         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp:
21029         (createBitmapContextFromWebView):
21030
21031 2011-07-07  Adam Roben  <aroben@apple.com>
21032
21033         Update TestFailures's title and header immediately upon navigation
21034
21035         Fixes <http://webkit.org/b/64125> TestFailures page seems slow to react on link clicks
21036
21037         Reviewed by Daniel Bates.
21038
21039         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
21040         (ViewController.prototype.loaded): Create and store an h1 element for displaying the page's
21041         title.
21042         (ViewController.prototype._displayBuilder): Set the page's title and clear out any old
21043         content immediately instead of waiting for the first set of results for the history
21044         analyzer. As we receive new results from the analyzer we'll just clear out the main content
21045         area and replace it with the new info.
21046         (ViewController.prototype._displayTesters): Set the page's title and clear out any old
21047         content immediately instead of waiting for the list of testers from the buildbot. Sprinkled
21048         in a little use of our removeAllChildren helper, too.
21049         (ViewController.prototype._setTitle): Added. Set's the page's title and the header text.
21050
21051 2011-07-07  Adam Roben  <aroben@apple.com>
21052
21053         Make Checkout use SCM's Executive instead of conjuring up its own
21054
21055         This will improve integration with the rest of webkitpy, particularly when invoked via
21056         webkit-patch.
21057
21058         Fixes <http://webkit.org/b/64115> REGRESSION (r90564): webkitpy's Checkout class uses
21059         Executive inappropriately
21060
21061         Reviewed by Adam Barth.
21062
21063         * Scripts/webkitpy/common/checkout/checkout.py:
21064         (Checkout.commit_message_for_this_commit): Use SCM.run instead of creating an Executive for
21065         our own use. SCM might have some super-special Executive that it uses under the covers, and
21066         we want to use it, too!
21067
21068         * Scripts/webkitpy/common/checkout/checkout_unittest.py:
21069         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Mock the SCM.run
21070         method to call through to Executive.run_command.
21071
21072 2011-07-07  Adam Roben  <aroben@apple.com>
21073
21074         Make Term/ReadKey.pm an optional dependency of commit-log-editor
21075
21076         When Term/ReadKey.pm isn't present, the --regenerate-log switch will be non-functional.
21077
21078         Fixes <http://webkit.org/b/64113> REGRESSION (r90564):
21079         webkitpy.common.checkout.checkout_unittest failing on GTK bots due to missing
21080         Term/ReadKey.pm module
21081
21082         Reviewed by Xan Lopez.
21083
21084         * Scripts/commit-log-editor: Use Module::Load::Conditional::can_load to only load
21085         Term/ReadKey.pm if it's present. If it isn't present, ignore the --regenerate-log switch so
21086         that existing commit messages will be preserved. (Users can manually remove existing commit
21087         messages using git-reset.)
21088
21089 2011-07-07  Adam Roben  <aroben@apple.com>
21090
21091         Keep testing that commit messages containing Unicode are handled properly by webkitpy
21092
21093         Fixes <http://webkit.org/b/64109> REGRESSION (r90571): test-webkitpy no longer tests that we
21094         correctly handle Unicode in commit messages
21095
21096         Reviewed by Eric Seidel.
21097
21098         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Put some Unicode characters in
21099         places in the ChangeLog that will end up being included in the commit message.
21100
21101 2011-07-07  Andreas Kling  <kling@webkit.org>
21102
21103         [Qt][WK2] Remove Symbian code.
21104         https://bugs.webkit.org/show_bug.cgi?id=64101
21105
21106         Reviewed by Benjamin Poulain.
21107
21108         * MiniBrowser/qt/BrowserWindow.cpp:
21109         (BrowserWindow::BrowserWindow):
21110         (BrowserWindow::screenshot):
21111         * MiniBrowser/qt/MiniBrowser.pro:
21112         * MiniBrowser/qt/MiniBrowserApplication.h:
21113         (WindowOptions::WindowOptions):
21114
21115 2011-07-07  Andreas Kling  <kling@webkit.org>
21116
21117         [Qt][WK2] Don't support ridiculous matrix of QT_NO_FEATURE combinations.
21118         https://bugs.webkit.org/show_bug.cgi?id=64099
21119
21120         Reviewed by Benjamin Poulain.
21121
21122         * MiniBrowser/qt/BrowserWindow.cpp:
21123         (BrowserWindow::openFile):
21124         (BrowserWindow::screenshot):
21125         (BrowserWindow::loadURLListFromFile):
21126
21127 2011-07-07  Adam Roben  <aroben@apple.com>
21128
21129         Completely ignore too-many-failures builds in TestFailures in most circumstances
21130
21131         Because a semi-arbitrary subset of tests are run in a too-many-failures build, we can't
21132         really use them to perform regression analysis. The only time we want to pay attention to
21133         too-many-failures builds is when we're trying to explain when the current bout of
21134         too-many-failures started.
21135
21136         Fixes <http://webkit.org/b/64106> TestFailures page sometimes claims a test started failing
21137         in a build that didn't even run it (because it exited early due to too many failues)
21138
21139         Reviewed by David Kilzer.
21140
21141         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
21142         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Removed old, broken
21143         too-many-failures handling that would cause us to blame builds that didn't even run a given
21144         test for breaking it. Instead, skip over all too-many-failures builds unless the most recent
21145         build was itself a too-many-failures build.
21146
21147 2011-07-07  Adam Roben  <aroben@apple.com>
21148
21149         Teach webkitpy's Checkout class to use commit-log-editor to create commit messages
21150
21151         Fixes <http://webkit.org/b/26755> webkit-patch's commit messages are less readable than
21152         commit-log-editor's
21153
21154         Reviewed by David Kilzer.
21155
21156         * Scripts/webkitpy/common/checkout/checkout.py:
21157         (Checkout.commit_message_for_this_commit): Run commit-log-editor, passing it the paths of
21158         the modified ChangeLogs, to generate the commit message, rather than trying to generate one
21159         ourselves.
21160
21161         * Scripts/webkitpy/common/checkout/checkout_unittest.py: Updated the expected commit message
21162         to match commit-log-editor's format.
21163         (CommitMessageForThisCommitTest.setUp): Write the ChangeLogs into Tools and LayoutTests
21164         directories so we can see how the various entries get labeled in the commit message.
21165         (CommitMessageForThisCommitTest.test_commit_message_for_this_commit): Create a mock SCM
21166         instance that knows how to find commit-log-editor and pass it to our Checkout instance.
21167         Don't bother capturing output, since there shouldn't be any.
21168
21169 2011-07-07  Adam Roben  <aroben@apple.com>
21170
21171         Add a --print-log option to commit-log-editor
21172
21173         When specified, commit-log-editor takes a set of ChangeLog paths on the command line,
21174         generates a commit message from those ChangeLogs, prints it to stdout, and exits.
21175
21176         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
21177         readable than commit-log-editor's
21178
21179         Reviewed by David Kilzer.
21180
21181         * Scripts/commit-log-editor: Changed to use Getopt::Long to parse options. Added --print-log
21182         option, which calls through to createCommitMessage, prints the result, and exits.
21183         (printUsageAndExit): Renamed from "usage" for clarity. Beefed up the usage statement to
21184         explain commit-log-editor's 2.5 (normal, --print-log, and --help) modes of operation.
21185
21186 2011-07-07  Adam Roben  <aroben@apple.com>
21187
21188         Extract commit-log-editor's commit-message-generation code into a separate function
21189
21190         As a bonus, we now skip a bunch of work in the case where we are generating a commit message
21191         for staged git changes.
21192
21193         Prep work for fixing <http://webkit.org/b/26755> webkit-patch's commit messages are less
21194         readable than commit-log-editor's
21195
21196         Reviewed by David Kilzer.
21197
21198         * Scripts/commit-log-editor: Moved the default definition of $endl up toward the top of the
21199         file so that it will be defined even if we aren't operating on an existing log message.
21200         Moved a bunch of top-level code to generate the commit message from here...
21201         (createCommitMessage): ...to here.
21202
21203 2011-07-07  Carlos Garcia Campos  <cgarcia@igalia.com>
21204
21205         [GTK] Open links in a new window when clicking with the middle button in MiniBrowser
21206         https://bugs.webkit.org/show_bug.cgi?id=63988
21207
21208         Reviewed by Martin Robinson.
21209
21210         * MiniBrowser/gtk/BrowserWindow.c:
21211         (browserWindowConstructed):
21212         (decidePolicyForNavigationAction): Ignore the action if a link is
21213         clicked with the middle buttonm and open the link in a new window.
21214         (browserWindowPolicyClientInit):
21215
21216 2011-07-07  Adam Barth  <abarth@webkit.org>
21217
21218         Remove style_references.py
21219         https://bugs.webkit.org/show_bug.cgi?id=64070
21220
21221         Reviewed by Eric Seidel.
21222
21223         We never really adopted the _references design, and this file is a bit
21224         of an orphan at this point.
21225
21226         * Scripts/check-webkit-style:
21227         * Scripts/webkitpy/style/checker.py:
21228         * Scripts/webkitpy/style/checker_unittest.py:
21229         * Scripts/webkitpy/style/checkers/python.py:
21230         * Scripts/webkitpy/style/checkers/test_expectations.py:
21231         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
21232         * Scripts/webkitpy/style/main_unittest.py:
21233         * Scripts/webkitpy/style_references.py: Removed.
21234
21235 2011-07-07  Eric Seidel  <eric@webkit.org>
21236
21237         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
21238         https://bugs.webkit.org/show_bug.cgi?id=64056
21239
21240         Unreviewed.
21241
21242         I had written this unittest prior to landing my change, but
21243         had named it qt_unitest.py, so it wasn't running (nor was
21244         it added to my git repo).
21245
21246         Once I fixe the name of the unittest file, it was very easy
21247         to fix the error in webkit.py (which already had a FIXME).
21248
21249         * Scripts/webkitpy/layout_tests/port/qt_unittest.py: Added.
21250         * Scripts/webkitpy/layout_tests/port/webkit.py:
21251
21252 2011-07-07  Eric Seidel  <eric@webkit.org>
21253
21254         Unreviewed.  Add more debug logging about which test expectations we're using.
21255
21256         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
21257         * Scripts/webkitpy/layout_tests/port/webkit.py:
21258
21259 2011-07-06  Adam Barth  <abarth@webkit.org>
21260
21261         Remove unexpected_results.json
21262         https://bugs.webkit.org/show_bug.cgi?id=64065
21263
21264         Reviewed by Eric Seidel.
21265
21266         No one should be using this file any more.  This patch removes it from
21267         the face of the earth.  This patch also removes two NRWT-specific
21268         command line options that were implemented using
21269         unexpected_results.json.
21270
21271         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
21272         * Scripts/webkitpy/layout_tests/controllers/manager.py:
21273         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
21274         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
21275         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
21276         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
21277         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
21278
21279 2011-07-07  Adam Barth  <abarth@webkit.org>
21280
21281         REGRESSION(90520): chromium GPU tests produce output in the LayoutTests directory
21282         https://bugs.webkit.org/show_bug.cgi?id=64067
21283
21284         Reviewed by Eric Seidel.
21285
21286         The author of r90520 forgot to change the chromium_gpu.py implemenation
21287         of this method as well.
21288
21289         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
21290
21291 2011-07-06  Eric Seidel  <eric@webkit.org>
21292
21293         Fix WebKit2 expected results search paths for Mac and Qt under new-run-webkit-tests
21294         https://bugs.webkit.org/show_bug.cgi?id=64056
21295
21296         Reviewed by Adam Barth.
21297
21298         This fixes the MacPort's version of baseline_search_path to include
21299         wk2 fallback like WebKitPort's does.  This also re-works the QtPort
21300         to set self._name (as other ports do) so that WebKitPort's default
21301         implementations will understand Qt operating system flavors correctly
21302         (including when running wk2 tests).
21303
21304         In trying to test this, I realized that nearly all of our port-tests
21305         use the default FileSystem(), User() and Executive() objects, which
21306         is really bad, because they expose details of your system in the unittest results!
21307
21308         Once I fixed many of the Port() calls to pass MockFileSystem(), then
21309         MacPort.test_tests_for_other_platforms started failing, and I had to
21310         teach MockFileSystem.glob how to handle directories to make it work again.
21311
21312         I removed the useless PortTestCase.test_baseline_search_path and replaced
21313         it with more useful test_baseline_search_path tests in MacPortTest and QtPortTest.
21314
21315         * Scripts/webkitpy/common/config/ports.py:
21316         * Scripts/webkitpy/common/system/filesystem_mock.py:
21317         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
21318         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
21319         * Scripts/webkitpy/layout_tests/port/mac.py:
21320         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
21321         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
21322         * Scripts/webkitpy/layout_tests/port/qt.py:
21323         * Scripts/webkitpy/layout_tests/port/test_files.py:
21324         * Scripts/webkitpy/layout_tests/port/win.py:
21325         * Scripts/webkitpy/to_be_moved/deduplicate_tests.py:
21326         * Scripts/webkitpy/tool/mocktool.py:
21327
21328 2011-07-06  Adam Barth  <abarth@webkit.org>
21329
21330         Wean rebaseline-server off unexpected_results.json
21331         https://bugs.webkit.org/show_bug.cgi?id=64063
21332
21333         Reviewed by Eric Seidel.
21334
21335         In the process of changing this code to use full_results.json, I
21336         noticed that the code was broken (because it wasn't tested).  This
21337         patch also adds test coverage for the broken code.
21338
21339         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
21340         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
21341         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
21342
21343 2011-07-06  Adam Barth  <abarth@webkit.org>
21344
21345         Wean resultsjsonparser off unexpected_results.json
21346         https://bugs.webkit.org/show_bug.cgi?id=64061
21347
21348         Reviewed by Eric Seidel.
21349
21350         We're about to delete unexpected_results.json, so we need to remove all
21351         the users.
21352
21353         * Scripts/webkitpy/common/config/ports.py:
21354         * Scripts/webkitpy/common/net/resultsjsonparser.py:
21355         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
21356         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
21357
21358 2011-07-06  Adam Barth  <abarth@webkit.org>
21359
21360         Cleanup result_summary.py
21361         https://bugs.webkit.org/show_bug.cgi?id=64057
21362
21363         Reviewed by Eric Seidel.
21364
21365         This class had a bunch of out-of-date docstrings that no longer make
21366         any sense.
21367
21368         * Scripts/webkitpy/layout_tests/models/result_summary.py:
21369
21370 2011-07-06  Adam Barth  <abarth@webkit.org>
21371
21372         Tweak the ChangeLog for DEPS rolls to say "Unreviewed" so that the
21373         patches can be landed by the commit-queue.
21374
21375         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
21376
21377 2011-07-06  Adam Barth  <abarth@webkit.org>
21378
21379         Normalize callers of logging.getLogger to use __name__
21380         https://bugs.webkit.org/show_bug.cgi?id=64054
21381
21382         Reviewed by Eric Seidel.
21383
21384         Most callers were already using __name__.  This patch makes everyone
21385         consistent.
21386
21387         * Scripts/webkitpy/common/checkout/diff_parser.py:
21388         * Scripts/webkitpy/common/checkout/scm/svn.py:
21389         * Scripts/webkitpy/common/net/statusserver.py:
21390         * Scripts/webkitpy/common/system/executive.py:
21391         * Scripts/webkitpy/common/system/file_lock.py:
21392         * Scripts/webkitpy/common/system/user.py:
21393         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
21394         * Scripts/webkitpy/layout_tests/models/result_summary.py:
21395         * Scripts/webkitpy/layout_tests/models/test_expectations.py:
21396         * Scripts/webkitpy/layout_tests/port/chromium.py:
21397         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
21398         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
21399         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
21400         * Scripts/webkitpy/layout_tests/port/gtk.py:
21401         * Scripts/webkitpy/layout_tests/port/http_lock.py:
21402         * Scripts/webkitpy/layout_tests/port/mac.py:
21403         * Scripts/webkitpy/layout_tests/port/qt.py:
21404         * Scripts/webkitpy/layout_tests/port/server_process.py:
21405         * Scripts/webkitpy/layout_tests/port/webkit.py:
21406         * Scripts/webkitpy/layout_tests/port/win.py:
21407         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
21408         * Scripts/webkitpy/layout_tests/servers/http_server.py:
21409         * Scripts/webkitpy/layout_tests/servers/http_server_base.py:
21410         * Scripts/webkitpy/layout_tests/servers/websocket_server.py:
21411         * Scripts/webkitpy/layout_tests/views/printing.py:
21412         * Scripts/webkitpy/python24/versioning.py:
21413         * Scripts/webkitpy/style/checker.py:
21414         * Scripts/webkitpy/style/checkers/test_expectations.py:
21415         * Scripts/webkitpy/style/patchreader.py:
21416         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py:
21417
21418 2011-07-06  Adam Barth  <abarth@webkit.org>
21419
21420         Move model classes out of layout_package and into models
21421         https://bugs.webkit.org/show_bug.cgi?id=64053
21422
21423         Reviewed by Eric Seidel.
21424
21425         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
21426         * Scripts/webkitpy/common/net/layouttestresults.py:
21427         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
21428         * Scripts/webkitpy/common/net/resultsjsonparser.py:
21429         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
21430         * Scripts/webkitpy/layout_tests/controllers/manager.py:
21431         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
21432         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
21433         * Scripts/webkitpy/layout_tests/controllers/worker.py:
21434         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
21435         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
21436         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
21437         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py: Removed.
21438         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py: Removed.
21439         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py: Removed.
21440         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Removed.
21441         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py: Removed.
21442         * Scripts/webkitpy/layout_tests/layout_package/test_input.py: Removed.
21443         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
21444         * Scripts/webkitpy/layout_tests/layout_package/test_results.py: Removed.
21445         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py: Removed.
21446         * Scripts/webkitpy/layout_tests/models: Added.
21447         * Scripts/webkitpy/layout_tests/models/__init__.py: Added.
21448         * Scripts/webkitpy/layout_tests/models/result_summary.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/result_summary.py.
21449         * Scripts/webkitpy/layout_tests/models/test_expectations.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations.py.
21450         * Scripts/webkitpy/layout_tests/models/test_expectations_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py.
21451         * Scripts/webkitpy/layout_tests/models/test_failures.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures.py.
21452         * Scripts/webkitpy/layout_tests/models/test_failures_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py.
21453         * Scripts/webkitpy/layout_tests/models/test_input.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_input.py.
21454         * Scripts/webkitpy/layout_tests/models/test_results.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results.py.
21455         * Scripts/webkitpy/layout_tests/models/test_results_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py.
21456         * Scripts/webkitpy/layout_tests/port/chromium.py:
21457         * Scripts/webkitpy/layout_tests/views/printing.py:
21458         * Scripts/webkitpy/layout_tests/views/printing_unittest.py:
21459         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
21460         * Scripts/webkitpy/style_references.py:
21461         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
21462         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
21463         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
21464         * Scripts/webkitpy/tool/commands/queues_unittest.py:
21465         * Scripts/webkitpy/tool/commands/rebaseline.py:
21466
21467 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
21468
21469         nrwt: fix typo preventing http startup on windows
21470         https://bugs.webkit.org/show_bug.cgi?id=64050
21471
21472         Reviewed by Eric Siedel.
21473
21474         NRWT checks whether it needs to start the servers by looking
21475         for "/http/" in the test names. After r90520, the leading slash
21476         is not present, so the check isn't quite right.
21477
21478         * Scripts/webkitpy/layout_tests/controllers/manager.py:
21479         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
21480
21481 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
21482
21483         nrwt: remove --use-apache from the command line
21484         https://bugs.webkit.org/show_bug.cgi?id=63358
21485
21486         Reviewed by Adam Barth.
21487
21488         Re-land the fix in bug 63358 - the prior fix was flawed in that
21489         we were trying to start the http server during check_sys_deps()
21490         but not actually setting up the environment properly. This
21491         broken prior to this change, but exposed by it, since the patch
21492         changes the chromium port to call the base class's checks.
21493
21494         * Scripts/webkitpy/common/system/executive.py:
21495         * Scripts/webkitpy/layout_tests/port/base.py:
21496         * Scripts/webkitpy/layout_tests/port/chromium.py:
21497         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
21498         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
21499         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
21500         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
21501         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
21502         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
21503
21504 2011-07-06  Adam Barth  <abarth@webkit.org>
21505
21506         Move view-related code out of layout_package and into views
21507         https://bugs.webkit.org/show_bug.cgi?id=64051
21508
21509         Reviewed by Eric Seidel.
21510
21511         * Scripts/webkitpy/layout_tests/controllers/manager.py:
21512         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py:
21513         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
21514         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
21515         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py: Removed.
21516         * Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py: Removed.
21517         * Scripts/webkitpy/layout_tests/layout_package/printing.py: Removed.
21518         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py: Removed.
21519         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
21520         * Scripts/webkitpy/layout_tests/views: Added.
21521         * Scripts/webkitpy/layout_tests/views/__init__.py: Added.
21522         * Scripts/webkitpy/layout_tests/views/metered_stream.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream.py.
21523         * Scripts/webkitpy/layout_tests/views/metered_stream_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py.
21524         * Scripts/webkitpy/layout_tests/views/printing.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing.py.
21525         * Scripts/webkitpy/layout_tests/views/printing_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py.
21526
21527 2011-07-06  Adam Barth  <abarth@webkit.org>
21528
21529         Add goofy comments to explain why we have these files.
21530
21531         * Scripts/webkitpy/layout_tests/controllers/__init__.py:
21532         * Scripts/webkitpy/layout_tests/layout_package/__init__.py:
21533
21534 2011-07-06  Adam Barth  <abarth@webkit.org>
21535
21536         Move a bunch of classes out of layout_tests.layout_package and into layout_tests.controllers
21537         https://bugs.webkit.org/show_bug.cgi?id=64048
21538
21539         Reviewed by Eric Seidel.
21540
21541         * Scripts/webkitpy/layout_tests/controllers: Added.
21542         * Scripts/webkitpy/layout_tests/controllers/__init__.py: Added.
21543         * Scripts/webkitpy/layout_tests/controllers/manager.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager.py.
21544         * Scripts/webkitpy/layout_tests/controllers/manager_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py.
21545         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py.
21546         * Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py.
21547         * Scripts/webkitpy/layout_tests/controllers/message_broker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker.py.
21548         * Scripts/webkitpy/layout_tests/controllers/message_broker_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py.
21549         * Scripts/webkitpy/layout_tests/controllers/single_test_runner.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py.
21550         * Scripts/webkitpy/layout_tests/controllers/worker.py: Copied from Tools/Scripts/webkitpy/layout_tests/layout_package/worker.py.
21551         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Removed.
21552         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py: Removed.
21553         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py: Removed.
21554         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py: Removed.
21555         * Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Removed.
21556         * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: Removed.
21557         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
21558         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: Removed.
21559         * Scripts/webkitpy/layout_tests/layout_package/worker.py: Removed.
21560         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
21561
21562 2011-07-06  Adam Barth  <abarth@webkit.org>
21563
21564         Attempt to fix unittest on Leopard.  Let's enumerate the keys in sorted
21565         order so the test results are predictable.
21566
21567         * Scripts/webkitpy/tool/bot/irc_command.py:
21568         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
21569
21570 2011-07-06  Dirk Pranke  <dpranke@chromium.org>
21571
21572         Reviewed by Eric Siedel.
21573
21574         nrwt: make the code be consistent about using test names instead of filenames or paths
21575
21576         https://bugs.webkit.org/show_bug.cgi?id=63597
21577
21578         It can be hard to figure out where in the code we're assuming
21579         test names are listed as unix-style relative filenames and where
21580         they are either absolute paths or relative paths following the
21581         host filesystem convention.
21582
21583         This patch changes things so that everything outside of the
21584         Port object uses (and must assume) unix-style relative
21585         filenames (with one exception, which is specifying host-local
21586         filenames as a list of test arguments on the command line).
21587
21588         This should make things clearer and more consistent, and also
21589         removes a lot of path conversion calls.
21590
21591         The changes in this patch outside of port/base.py are largely
21592         cosmetic changes of "filename" to "test". We add a few routines
21593         to the base Port implementation to manipulate test names to
21594         split them, figure out if we're referring to a directory of
21595         tests, etc.
21596
21597         * Scripts/webkitpy/common/net/layouttestresults.py:
21598         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
21599         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
21600         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
21601         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
21602         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
21603         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
21604         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
21605         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
21606         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
21607         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
21608         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
21609         * Scripts/webkitpy/layout_tests/layout_package/test_input.py:
21610         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
21611         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
21612         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
21613         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
21614         * Scripts/webkitpy/layout_tests/port/base.py:
21615         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
21616         * Scripts/webkitpy/layout_tests/port/chromium.py:
21617         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
21618         * Scripts/webkitpy/layout_tests/port/dryrun.py:
21619         * Scripts/webkitpy/layout_tests/port/test.py:
21620         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
21621         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
21622         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
21623         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
21624         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
21625         * Scripts/webkitpy/tool/commands/rebaseline.py:
21626         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
21627
21628 2011-07-06  Chris Guillory  <ctguil@chromium.org>
21629
21630         No review necessary.
21631
21632         Adding myself to committers list in committers.py.
21633         https://bugs.webkit.org/show_bug.cgi?id=63478
21634
21635         * Scripts/webkitpy/common/config/committers.py:
21636
21637 2011-07-06  Johnny Ding  <jnd@chromium.org>
21638
21639         https://bugs.webkit.org/show_bug.cgi?id=63857
21640         Implement createPopupMenu for Chromium DRT.
21641
21642         Reviewed by Tony Chang.
21643
21644         * DumpRenderTree/chromium/WebViewHost.cpp:
21645         (WebViewHost::createPopupMenu):
21646         (WebViewHost::~WebViewHost):
21647         * DumpRenderTree/chromium/WebViewHost.h:
21648
21649 2011-07-06  Mihai Parparita  <mihaip@chromium.org>
21650
21651         Unreviewed, rolling out r90503.
21652         http://trac.webkit.org/changeset/90503
21653         https://bugs.webkit.org/show_bug.cgi?id=63358
21654
21655         Makes Chromium Windows canary bots hang.
21656
21657         * Scripts/webkitpy/layout_tests/port/base.py:
21658         * Scripts/webkitpy/layout_tests/port/chromium.py:
21659         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
21660         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
21661         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
21662         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
21663         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
21664         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
21665
21666 2011-07-06  Adam Barth  <abarth@webkit.org>
21667
21668         Move buildbot URL in webkitpy to config.urls
21669         https://bugs.webkit.org/show_bug.cgi?id=64041
21670
21671         Reviewed by Eric Seidel.
21672
21673         This patch prepares for using the buildbot class to talk to the
21674         Chromium buildbots.
21675
21676         * Scripts/webkitpy/common/config/urls.py:
21677         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
21678         * Scripts/webkitpy/tool/commands/download_unittest.py:
21679         * Scripts/webkitpy/tool/commands/queries.py:
21680         * Scripts/webkitpy/tool/mocktool.py:
21681         * Scripts/webkitpy/tool/steps/ensurebuildersaregreen.py:
21682
21683 2011-07-06  Kevin Ollivier  <kevino@theolliviers.com>
21684
21685         [wx] Unreviewed build fix, add the correct path to the waf build settings.
21686
21687         * wx/packaging/build-win-installer.py:
21688
21689 2011-07-06  Adam Barth  <abarth@webkit.org>
21690
21691         Add missing include.
21692
21693         * Scripts/webkitpy/tool/bot/irc_command.py:
21694
21695 2011-07-06  Adam Barth  <abarth@webkit.org>
21696
21697         Add roll-chromium-deps command to sheriff-bot
21698         https://bugs.webkit.org/show_bug.cgi?id=64037
21699
21700         Reviewed by Eric Seidel.
21701
21702         Lo, and the DEPS were rolled on command.
21703
21704         * Scripts/webkitpy/tool/bot/irc_command.py:
21705         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
21706         * Scripts/webkitpy/tool/bot/sheriff.py:
21707         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
21708         * Scripts/webkitpy/tool/commands/roll.py:
21709         * Scripts/webkitpy/tool/commands/roll_unittest.py:
21710         * Scripts/webkitpy/tool/steps/updatechromiumdeps.py:
21711
21712 2011-07-06  Eric Seidel  <eric@webkit.org>
21713
21714         Split Driver, DriverOutput and DriverInput out into a new driver.py file
21715         https://bugs.webkit.org/show_bug.cgi?id=64031
21716
21717         Reviewed by Adam Barth.
21718
21719         This ended up kinda large and I'm happy to break it up if desired.
21720
21721         As part of cleaning up the Port module, I split Driver, DriverInput and DriverOutput
21722         out into its own driver.py file.
21723
21724         I also cleaned up how we import Port objects, by making port/__init__.py
21725         expose Port, Driver, DriverInput, and DriverOutput.  I removed exposure of
21726         some various test objects as they were causing circular dependencies.
21727         base.py is now an implementation detail of port.  No longer is "base" referenced
21728         anywhere in the code.
21729
21730         I also made Driver.__init__ non-virtual so that sub-classers could all share
21731         the data storage on Driver.  This more closely matches how other subclasses
21732         work throughout our python code.
21733
21734         Standard line-unwrapping and doc-string removal as I went through the code.
21735
21736         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
21737         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
21738         * Scripts/webkitpy/layout_tests/port/__init__.py:
21739         * Scripts/webkitpy/layout_tests/port/base.py:
21740         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
21741         * Scripts/webkitpy/layout_tests/port/chromium.py:
21742         * Scripts/webkitpy/layout_tests/port/driver.py: Added.
21743         * Scripts/webkitpy/layout_tests/port/dryrun.py:
21744         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
21745         * Scripts/webkitpy/layout_tests/port/test.py:
21746         * Scripts/webkitpy/layout_tests/port/webkit.py:
21747         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
21748         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py:
21749         * Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
21750
21751 2011-07-06  Adam Roben  <aroben@apple.com>
21752
21753         Clear TestFailures's getNumberOfFailingTests cache
21754
21755         Now that we support NRWT, we need to clear out old zero values for the NRWT bots that might
21756         have been cached by old versions of TestFailures.
21757
21758         Fixes <http://webkit.org/b/64029> TestFailures page still doesn't show NRWT bots if you've
21759         used it recently
21760
21761         Reviewed by Adam Barth.
21762
21763         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
21764         (Builder.prototype.getNumberOfFailingTests): Added a .version property to the cache data
21765         that must match the current version. Otherwise we'll ignore the cached data.
21766
21767 2011-06-24  Dirk Pranke  <dpranke@chromium.org>
21768
21769         Reviewed by Adam Barth.
21770
21771         nrwt: remove --use-apache from the command line
21772         https://bugs.webkit.org/show_bug.cgi?id=63358
21773
21774         This change removes the --use-apache command line argument. It
21775         was initially put in when we were trying to get the cygwin
21776         apache instance to work with Chromium win, but that code has
21777         bitrotted and doesn't work at all now.
21778
21779         Arguably we should remove all of the code to allow for a choice
21780         of web servers, but since we may still want to switch off of
21781         LigHTTPd at some point on Windows, I'll leave the rest in for
21782         now (or at least yank it in a different change.)
21783
21784         * Scripts/webkitpy/layout_tests/port/base.py:
21785         * Scripts/webkitpy/layout_tests/port/chromium.py:
21786         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
21787         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
21788         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
21789         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
21790
21791 2011-07-06  Dimitri Glazkov  <dglazkov@chromium.org>
21792
21793         garden-o-matic should use a favicon to indicate current state.
21794         https://bugs.webkit.org/show_bug.cgi?id=64027
21795
21796         Reviewed by Adam Barth.
21797
21798         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-green.png: Added.
21799         * Scripts/webkitpy/tool/servers/data/gardeningserver/favicon-red.png: Added.
21800         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Added blurb about icons.
21801         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Added setting icon state.
21802         * Scripts/webkitpy/tool/servers/gardeningserver.py: Allowed the use of favicons.
21803
21804 2011-07-06  Daniel Bates  <dbates@rim.com>
21805
21806         Fix webkitpy error after changeset 90483 <http://trac.webkit.org/changeset/90483>; Add missing
21807         parameter to _fetch_file_from_results lamdba function.
21808
21809         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
21810
21811 2011-07-06  Kevin Ollivier  <kevino@theolliviers.com>
21812
21813         [wx] Unreviewed build fix. Generate the proper install_name for dylibs on Mac.
21814
21815         * wx/packaging/build-mac-installer.py:
21816
21817 2011-07-06  Xan Lopez  <xlopez@igalia.com>
21818
21819         [GTK] Only set env. variables in start when strictly needed
21820         https://bugs.webkit.org/show_bug.cgi?id=64026
21821
21822         Reviewed by Eric Seidel.
21823
21824         * Scripts/webkitpy/layout_tests/port/gtk.py: add a
21825         setup_environ_for_server method in GtkPort and set as many
21826         env. variables as possible there.
21827
21828 2011-07-06  Adam Roben  <aroben@apple.com>
21829
21830         Teach TestFailures how to load, parse, and interpret NRWT test results
21831
21832         Fixes <http://webkit.org/b/61877> TestFailures page doesn't show testers that use
21833         new-run-webkit-tests
21834
21835         Reviewed by Adam Barth.
21836
21837         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
21838         (Builder.prototype.failureDiagnosisTextAndURL): Added support for the new 'flaky' failure
21839         type. For now we don't account for image-only flakes (but TestFailures doesn't deal with
21840         pixel tests at all currently).
21841         (Builder.prototype.getNumberOfFailingTests): Relaxed the regex used for parsing the number
21842         of failing tests from the buildbot output. Make sure not to count "new passes" (a new
21843         category introduced by NRWT) as failures.
21844
21845         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
21846         (LayoutTestResultsLoader.prototype.start): Moved a bunch of code from here to
21847         _fetchAndParseORWTResults. This function now attempts to load NRWT results, then falls back
21848         to loading ORWT results.
21849         (LayoutTestResultsLoader.prototype._fetchAndParseNRWTResults): Added. Tries to load and
21850         parse the full_results.json file for the given build.
21851         (LayoutTestResultsLoader.prototype._fetchAndParseORWTResults): Added. Code came from start.
21852         Fixed a bug along the way where we were sometimes calling the error callback instead of the
21853         success callback when all tests passed.
21854
21855         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/NRWTResultsParser.js: Added.
21856         (NRWTResultsParser): Do-nothing constructor.
21857         (NRWTResultsParser.prototype.parse): Uses eval() (eww!) to get the test results out of the
21858         JS string, then iterates over all the tests in the results data and builds up a data
21859         structure matching what ORWTResultsParser returns.
21860
21861         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
21862         (Array.prototype.contains):
21863         (String.prototype.contains):
21864         Added these simple helper functions.
21865
21866         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
21867         NRWTResultsParser.js.
21868
21869 2011-07-06  Adam Roben  <aroben@apple.com>
21870
21871         Extract code to parse ORWT's results.html file into its own class
21872
21873         Prep work for fixing <http://webkit.org/b/61877> TestFailures page doesn't show testers that
21874         use new-run-webkit-tests
21875
21876         Reviewed by Adam Barth.
21877
21878         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
21879         (LayoutTestResultsLoader.prototype.start): Moved a bunch of code from here...
21880
21881         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ORWTResultsParser.js: Added.
21882         (ORWTResultsParser):
21883         (ORWTResultsParser.prototype.parse):
21884         ...to here.
21885
21886         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
21887         ORWTResultsParser.js.
21888
21889 2011-07-06  Eric Seidel  <eric@webkit.org>
21890
21891         Unreviewed.  Fix a unittest from hitting the network. :)
21892
21893         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
21894
21895 2011-07-06  Eric Seidel  <eric@webkit.org>
21896
21897         Unreviewed.  Fixing an exception in test-webkitpy, caused by bug 64006.
21898
21899         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
21900
21901 2011-07-06  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
21902
21903         Run an Xvfb instance to run the API tests on, like we do for the
21904         layout tests.
21905
21906         Reviewed by Xan Lopez.
21907
21908         * Scripts/run-gtk-tests:
21909
21910 2011-07-06  Daniel Bates  <dbates@webkit.org>
21911
21912         Make SCM unit tests faster
21913         https://bugs.webkit.org/show_bug.cgi?id=63883
21914
21915         Speeds up the SCM unit tests by a factor of 4. Currently, we create a
21916         mock SVN repo for each test_ method in SVNTest and GitTest and creating
21917         this repo is expensive.
21918
21919         Instead, it is significantly faster to create the mock SVN repo once
21920         and then perform a filesystem copy of it for each test_ method.
21921
21922         Note, Python 2.7's unittest module implements support for per class and
21923         per module setup and tear down methods which could be used to implement
21924         similar functionality. At the time of writing, test-webkitpy is designed
21925         to support Python 2.5. So, we can't take advantage of these Python 2.7
21926         features :(
21927
21928         Reviewed by Eric Seidel.
21929
21930         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
21931
21932 2011-07-06  Adam Barth  <abarth@webkit.org>
21933
21934         Remove unused parameter from _add_test in test_expectations.py
21935         https://bugs.webkit.org/show_bug.cgi?id=64008
21936
21937         Reviewed by Eric Seidel.
21938
21939         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
21940
21941 2011-07-06  Adam Barth  <abarth@webkit.org>
21942
21943         Use old-run-webkit-tests when --leaks is present
21944         https://bugs.webkit.org/show_bug.cgi?id=64012
21945
21946         Reviewed by Adam Roben.
21947
21948         We'll remove this branch once support of --leaks is added to
21949         new-run-webkit-tests.
21950
21951         * Scripts/run-webkit-tests:
21952
21953 2011-07-06  Benjamin Poulain  <benjamin@webkit.org>
21954
21955         [Qt] [WK2] Add touch mocking to Qt's MiniBrowser
21956         https://bugs.webkit.org/show_bug.cgi?id=63995
21957
21958         Reviewed by Andreas Kling.
21959
21960         The mouse events are intercepted through QCoreApplication::notify(),
21961         and fake touch events are generated for mouse events when necessary.
21962
21963         If touch events are received from the platform, we assume the current
21964         hardware has a touch screen and we stop generating fake events.
21965
21966         * MiniBrowser/qt/MiniBrowserApplication.cpp:
21967         (isTouchEvent):
21968         (isMouseEvent):
21969         (MiniBrowserApplication::MiniBrowserApplication):
21970         (MiniBrowserApplication::notify):
21971         * MiniBrowser/qt/MiniBrowserApplication.h:
21972
21973 2011-07-06  Xan Lopez  <xlopez@igalia.com>
21974
21975         Reviewed by Gustavo Noronha.
21976
21977         * Scripts/webkitpy/layout_tests/port/gtk.py: use Popen correctly.
21978
21979 2011-07-06  Eric Seidel  <eric@webkit.org>
21980
21981         webkit-patch failure-reason does not understand NRWT results
21982         https://bugs.webkit.org/show_bug.cgi?id=64006
21983
21984         Reviewed by Adam Barth.
21985
21986         Adam Barth tells me failure-reason should be deleted soon,
21987         but aroben's fancy new TestFailures/ page doesn't work for
21988         NRWT yet, and I needed to know when
21989         third-party-cookie-relaxing started failing.
21990
21991         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
21992         * Scripts/webkitpy/tool/commands/queries.py:
21993
21994 2011-07-06  Xan Lopez  <xlopez@igalia.com>
21995
21996         [GTK] Add missing environment variables in NWRT
21997         https://bugs.webkit.org/show_bug.cgi?id=64004
21998
21999         Reviewed by Gustavo Noronha Silva.
22000
22001         Add missing environment variables needed by the GTK+ port.
22002
22003         * Scripts/webkitpy/layout_tests/port/gtk.py: ditto.
22004
22005 2011-07-06  Adam Roben  <aroben@apple.com>
22006
22007         Make run-api-tests output intelligible on the bots
22008
22009         run-api-tests's output on the bots was very hard to read for two reasons:
22010           1) It was mixing its own output with gtest's
22011           2) It was using ANSI escape sequences to print colored text, but the bots don't support
22012              that
22013
22014         Now, the --verbose flag turns off almost all of run-api-tests's own output so that gtest's
22015         output will not be obscured. We still print "Timeout" messages even in verbose mode, since
22016         gtest doesn't have any native support for timeouts. Also, when our output is being
22017         redirected to a file, we don't print ANSI escape sequences.
22018
22019         Fixes <http://webkit.org/b/63996> It's very hard to read run-api-tests output on the bots
22020
22021         Reviewed by Sam Weinig.
22022
22023         * Scripts/run-api-tests: Don't use the :constants interface of Term::ANSIColor. We want to
22024         use the colored function instead.
22025         (runAllTestsInSuite): Don't print out suite names in verbose mode; gtest will give us enough
22026         context that they aren't needed.
22027         (runTest): Don't print out test names or pass/fail messages in verbose mode; gtest will do
22028         that for us. Also, use the new possiblyColored function instead of always coloring output.
22029         (possiblyColored): Added. When printing to a tty, returns the string with the appropriate
22030         ANSI color escape sequences added. Otherwise just returns the string unmodified.
22031
22032 2011-07-06  Xan Lopez  <xlopez@igalia.com>
22033
22034         Reviewed by Gustavo Noronha.
22035
22036         Launch Xvfb (mostly) the same way we were launching it in the
22037         bots, for maximum compatibility.
22038
22039         * Scripts/webkitpy/layout_tests/port/gtk.py:
22040
22041 2011-07-06  Andreas Kling  <kling@webkit.org>
22042
22043         [Qt][WK2] Split Qt API into two different web views (touch and desktop)
22044         https://bugs.webkit.org/show_bug.cgi?id=63950
22045
22046         Co-authored with Benjamin Poulain <benjamin@webkit.org>.
22047
22048         Reviewed by Kenneth Rohde Christiansen.
22049
22050         Port WTR/Qt to use QDesktopWebView and force a 800x600 viewport.
22051
22052         The Qt MiniBrowser now defaults to QDesktopWebView but can be started
22053         in QTouchWebView mode by passing -touch on the command line.
22054
22055         A lot of functionality has been removed and/or disabled from
22056         MiniBrowser until we reintroduce the necessary API's.
22057
22058         * MiniBrowser/qt/BrowserView.cpp:
22059         (BrowserView::BrowserView):
22060         (BrowserView::~BrowserView):
22061         (BrowserView::load):
22062         (BrowserView::view):
22063         (BrowserView::touchWebView):
22064         (BrowserView::desktopWebView):
22065         * MiniBrowser/qt/BrowserView.h:
22066         * MiniBrowser/qt/BrowserWindow.cpp:
22067         (newPageFunction):
22068         (BrowserWindow::BrowserWindow):
22069         (BrowserWindow::webView):
22070         (BrowserWindow::newWindow):
22071         (BrowserWindow::showUserAgentDialog):
22072         (BrowserWindow::updateUserAgentList):
22073         * MiniBrowser/qt/BrowserWindow.h:
22074         * MiniBrowser/qt/MiniBrowserApplication.cpp:
22075         (MiniBrowserApplication::handleUserOptions):
22076         * MiniBrowser/qt/MiniBrowserApplication.h:
22077         (WindowOptions::WindowOptions):
22078         * MiniBrowser/qt/UrlLoader.cpp:
22079         (UrlLoader::UrlLoader):
22080         * MiniBrowser/qt/main.cpp:
22081         (main):
22082         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
22083         (WTR::WebView::wkView):
22084         (WTR::WebView::pageRef):
22085         (WTR::WebView::WebView):
22086         (WTR::PlatformWebView::page):
22087         (WTR::PlatformWebView::windowFrame):
22088         (WTR::PlatformWebView::setWindowFrame):
22089
22090 2011-07-06  Eric Seidel  <eric@webkit.org>
22091
22092         Unreviewed..
22093
22094         * Scripts/run-webkit-tests:
22095         (useNewRunWebKitTests):
22096          - Fix the pass-thru of --gtk and --qt for old-run-webkit-tests,
22097            it was previously broken by the existence of isGtk/isQt checks which
22098            removed the --gtk/--qt args from @ARGV.
22099
22100 2011-07-06  Eric Seidel  <eric@webkit.org>
22101
22102         new-run-webkit-tests uses a 35s timeout for Mac and 6s for all other ports
22103         https://bugs.webkit.org/show_bug.cgi?id=63983
22104
22105         Reviewed by Xan Lopez.
22106
22107         All WebKit ports want to use a 35s timeout.  Only chromium currently
22108         uses a 6s timeout.  Eventually we'll move all WebKit ports to 6s too
22109         to further speed up new-run-webkit-tests runs, but that's going to be a while.
22110
22111         * Scripts/webkitpy/layout_tests/port/mac.py:
22112         * Scripts/webkitpy/layout_tests/port/webkit.py:
22113         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
22114
22115 2011-07-06  Xan Lopez  <xlopez@igalia.com>
22116
22117         Unreviewed.
22118
22119         NWRT actually seems to work locally, so it should work in the bots
22120         as long as we take care of the TIMEOUT issue (which is being fixed
22121         as I write this). Let's give this another shot and fix whatever
22122         breaks.
22123
22124         * Scripts/run-webkit-tests:
22125         (useNewRunWebKitTests):
22126
22127 2011-07-06  Xan Lopez  <xlopez@igalia.com>
22128
22129         Unreviewed.
22130
22131         Disable NWRT on GTK. At the very least it needs an updated exected
22132         results file since NWRT is more sensitive to timeouts, not sure
22133         what the exact problem is.
22134
22135         * Scripts/run-webkit-tests:
22136         (useNewRunWebKitTests):
22137
22138 2011-07-06  Adam Barth  <abarth@webkit.org>
22139
22140         Fix global variable leak detected by noglobals.
22141
22142         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
22143
22144 2011-07-06  Adam Barth  <abarth@webkit.org>
22145
22146         garden-o-matic shouldn't show flaky tests by default
22147         https://bugs.webkit.org/show_bug.cgi?id=63979
22148
22149         Reviewed by Ojan Vafai.
22150
22151         If any of the actual test results are PASS, the test is flaky and we
22152         can ignore it (at least for now).  If we get awesome at dealing with
22153         failing tests, we might want to tackle flaky tests by refering the user
22154         to the flakiness dashboard, which is way more awesome at dealing with
22155         flaky tests that this tool ever will be.
22156
22157         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
22158         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
22159
22160 2011-07-05  Adam Barth  <abarth@webkit.org>
22161
22162         garden-o-matic should list the failing tests
22163         https://bugs.webkit.org/show_bug.cgi?id=63978
22164
22165         Reviewed by Eric Seidel.
22166
22167         * Scripts/webkitpy/tool/servers/data/gardeningserver/config.js: Added.
22168         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
22169         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js:
22170         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
22171         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js:
22172         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
22173         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui.js: Added.
22174         * Scripts/webkitpy/tool/servers/data/gardeningserver/ui_unittests.js: Added.
22175         * Scripts/webkitpy/tool/servers/gardeningserver.py:
22176
22177
22178 2011-07-05  Eric Seidel  <eric@webkit.org>
22179
22180         Re-enable NRWT on Leopard now that bug 63973 is fixed.
22181
22182         Reviewed by Adam Barth.
22183
22184         I also made the should-enable expression a black-list instead
22185         of a white-list since every bot on build.webkit.org is
22186         now using NRWT except windows and WebKit2 bots.
22187
22188         * Scripts/run-webkit-tests:
22189         (useNewRunWebKitTests):
22190
22191 2011-07-05  Eric Seidel  <eric@webkit.org>
22192
22193         Perl string concat failure on Leopard (causes NRWT to fail)
22194         https://bugs.webkit.org/show_bug.cgi?id=63973
22195
22196         Reviewed by Daniel Bates.
22197
22198         Leopard's "xcodebuild -version" has a different format from Snow Leopard/Lion
22199         thus the version parsing logic in webkitdirs.pm seems to fail.
22200
22201         Default to "3.0" instead of undef when we can't parse the version number.
22202         This should send us down the correct path for Leopard.
22203
22204         * Scripts/webkitdirs.pm:
22205         (determineBaseProductDir):
22206
22207 2011-07-05  Kevin Ollivier  <kevino@theolliviers.com>
22208
22209         [wx] Unreviewed build fix after addition of Source/WebCore/xml/parser dir.
22210
22211         * waf/build/settings.py:
22212
22213 2011-07-05  Eric Seidel  <eric@webkit.org>
22214
22215         Disable NRWT for Leopard until https://bugs.webkit.org/show_bug.cgi?id=63973 can be fixed.
22216
22217         Reviewed by Adam Barth.
22218
22219         * Scripts/run-webkit-tests:
22220         (useNewRunWebKitTests):
22221
22222 2011-07-05  Eric Seidel  <eric@webkit.org>
22223
22224         Make run-webkit-tests pass --verbose to new-run-webkit-tests on qt/gtk bots.
22225
22226         Reviewed by Adam Barth.
22227
22228         * Scripts/run-webkit-tests:
22229         (runningOnBuildBot):
22230
22231 2011-07-05  Adam Barth  <abarth@webkit.org>
22232
22233         Add results_unittests.js to garden-o-matic
22234         https://bugs.webkit.org/show_bug.cgi?id=63971
22235
22236         Reviewed by Eric Seidel.
22237
22238         Adding these unit tests forced me to commit to an API for the results
22239         module.  I'm not sure if this is the API we'll end up with, but it's
22240         something not entirely unreasonable.
22241
22242         Also, I've changed results.js over to WebKit style, as discussed previously.
22243
22244         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
22245         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
22246         * Scripts/webkitpy/tool/servers/data/gardeningserver/results_unittests.js: Added.
22247         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html:
22248
22249 2011-07-05  Eric Seidel  <eric@webkit.org>
22250
22251         Teach run-webkit-tests how to translate --qt and --gtk for new-run-webkit-tests.
22252
22253         Reviewed by Adam Barth.
22254
22255         * Scripts/run-webkit-tests:
22256
22257 2011-07-05  Adam Barth  <abarth@webkit.org>
22258
22259         Add first unit tests for garden-o-matic
22260         https://bugs.webkit.org/show_bug.cgi?id=63969
22261
22262         Reviewed by Eric Seidel.
22263
22264         I've also reformated base.js to match WebKit style, as discussed in the
22265         previous bug.
22266
22267         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js:
22268         * Scripts/webkitpy/tool/servers/data/gardeningserver/base_unittests.js: Added.
22269         * Scripts/webkitpy/tool/servers/data/gardeningserver/run-unittests.html: Added.
22270
22271 2011-07-05  Eric Seidel  <eric@webkit.org>
22272
22273         Move Qt, Gtk and Leopard to NRWT.
22274
22275         We'll roll out any of these if we see problems.
22276
22277         Reviewed by Adam Barth.
22278
22279         * Scripts/run-webkit-tests:
22280         (useNewRunWebKitTests):
22281
22282 2011-07-05  Adam Barth  <abarth@webkit.org>
22283
22284         garden-o-matic should understand which tests have unexpected failures
22285         https://bugs.webkit.org/show_bug.cgi?id=63965
22286
22287         Reviewed by Eric Seidel.
22288
22289         Currently, this code just logs the list of unexpected failures to the
22290         console.  A future patch will do something useful with this
22291         information.  A testing harness will also come in the next patch.
22292
22293         * Scripts/webkitpy/tool/servers/data/gardeningserver/base.js: Added.
22294         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
22295         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js:
22296         * Scripts/webkitpy/tool/servers/gardeningserver.py:
22297
22298 2011-07-05  Adam Barth  <abarth@webkit.org>
22299
22300         garden-o-matic should know how to fetch test results from the (NRWT) bots
22301         https://bugs.webkit.org/show_bug.cgi?id=63959
22302
22303         Reviewed by Eric Seidel.
22304
22305         Currently, this code just fetches the results and expectations JSON
22306         blob for Linux from the server and logs it to the console.
22307
22308         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
22309         * Scripts/webkitpy/tool/servers/data/gardeningserver/results.js: Added.
22310         * Scripts/webkitpy/tool/servers/gardeningserver.py:
22311
22312 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
22313
22314         Reviewed by Eric Seidel.
22315
22316         Python tests are failing on leopard
22317         https://bugs.webkit.org/show_bug.cgi?id=63842
22318
22319         Fix the missing flush() call that appears to be needed by the
22320         logging package in python 2.5. Also fix the regressions
22321         introduced in the run_webkit_test tests when we switched from
22322         thread to processes (since processes aren't available on 2.5).
22323
22324         Also fix a missing "from __future__ import with_statement' in
22325         testfilehandler (unreviewed).
22326
22327         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py:
22328         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
22329         * TestResultServer/handlers/testfilehandler.py:
22330
22331 2011-07-05  Eric Seidel  <eric@webkit.org>
22332
22333         Reviewed by Adam Barth.
22334
22335         re-enabling new-run-webkit-tests on the Snow Leopard WebKit1 bots
22336         now that we believe the stray http-server issue to be solved.
22337
22338         * Scripts/run-webkit-tests:
22339         (useNewRunWebKitTests):
22340
22341 2011-07-05  Adam Barth  <abarth@webkit.org>
22342
22343         Turns out we need to support numerals too because that's what jQuery
22344         uses by default.
22345
22346         * TestResultServer/handlers/testfilehandler.py:
22347
22348 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
22349
22350         nrwt: allow for multiple http shards
22351         https://bugs.webkit.org/show_bug.cgi?id=63116
22352
22353         Reviewed by Tony Chang.
22354
22355         This modifies the sharding logic to support multiple http
22356         shards, but for now we clamp to one shard until we can test
22357         perf impact and flakiness impact.
22358
22359         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
22360         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
22361
22362 2011-07-05  Dirk Pranke  <dpranke@chromium.org>
22363
22364         Re-land nrwt: make sharding tests needing locks less hard-coded
22365         https://bugs.webkit.org/show_bug.cgi?id=63112
22366
22367         Reviewed by Ojan Vafai.
22368
22369         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
22370         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
22371
22372 2011-07-05  Adam Barth  <abarth@webkit.org>
22373
22374         TestResultServer should support callback parameter for JSON
22375         https://bugs.webkit.org/show_bug.cgi?id=63961
22376
22377         Reviewed by Ojan Vafai.
22378
22379         This makes using jQuery.ajax so much more pleasant.
22380
22381         * TestResultServer/handlers/testfilehandler.py:
22382
22383 2011-07-05  Eric Seidel  <eric@webkit.org>
22384
22385         buildbot needs to understand whether NRWT exited early after having too many failures
22386         https://bugs.webkit.org/show_bug.cgi?id=63839
22387
22388         Reviewed by Adam Barth.
22389
22390         Fix-up the exited early messages printed by NRWT so that
22391         the buildbot can parse them as expected.
22392         It looks for lines using "if line.find('Exiting early') >= 0:"
22393
22394         I also plumbed the "early exit" status through to results.json
22395         in the form of an "interrupted" bool.  It was unclear to me
22396         if results.json already had enough information to compute this bool
22397         itself.  It's possible Ojan could come up with a better fix.
22398
22399         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
22400         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
22401         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
22402         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
22403
22404 2011-07-05  Eric Seidel  <eric@webkit.org>
22405
22406         new-run-webkit-tests fails to start http server if one is already running
22407         https://bugs.webkit.org/show_bug.cgi?id=63956
22408
22409         Reviewed by Adam Barth.
22410
22411         * BuildSlaveSupport/kill-old-processes:
22412          - Removed ^M line endings and sorted the process names.
22413          - There are no functional changes to this file.
22414            (except that now that its using unix line endings it can be executed directly!)
22415         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
22416          - Change the name to 'httpd' to match old-run-webkit-test paths.
22417
22418 2011-07-05  Adam Barth  <abarth@webkit.org>
22419
22420         Add basic ajax support to garden-o-matic
22421         https://bugs.webkit.org/show_bug.cgi?id=63874
22422
22423         Reviewed by Eric Seidel.
22424
22425         This patch adds some basic infrastructure to garden-o-matic.
22426         Currently, the infrastructure is just used to make the "quit" command
22427         use Ajax, but in the future, this infrastructure will be used to do
22428         more sophistocated remote proceedure calls.
22429
22430         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html:
22431         * Scripts/webkitpy/tool/servers/data/gardeningserver/main.js: Added.
22432         * Scripts/webkitpy/tool/servers/gardeningserver.py:
22433         * Scripts/webkitpy/tool/servers/reflectionhandler.py:
22434
22435 2011-07-05  Adam Barth  <abarth@webkit.org>
22436
22437         Add trivial garden-o-matic command (with server)
22438         https://bugs.webkit.org/show_bug.cgi?id=63872
22439
22440         Reviewed by Eric Seidel.
22441
22442         This patch adds a trivial "Hello, world" garden-o-matic command,
22443         complete with an HTTP server.  This command re-uses a bunch of code
22444         from the existing rebaseline-server command.  Over time, this command
22445         will grow to be a tool useful for gardening the WebKit tree.
22446
22447         * Scripts/webkitpy/tool/commands/__init__.py:
22448         * Scripts/webkitpy/tool/commands/abstractservercommand.py: Added.
22449         * Scripts/webkitpy/tool/commands/gardenomatic.py: Added.
22450         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
22451         * Scripts/webkitpy/tool/servers/data/gardeningserver/index.html: Added.
22452         * Scripts/webkitpy/tool/servers/gardeningserver.py: Added.
22453         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
22454
22455 2011-07-05  Adam Roben  <aroben@apple.com>
22456
22457         Make prepare-ChangeLog include modified Perl functions in its ChangeLog template
22458
22459         This is a very simple first cut. Functions must start with a line that starts with "sub "
22460         and end with a line that starts with a closing brace. No leading whitespace is allowed.
22461         Package names aren't parsed at all.
22462
22463         Fixes <http://webkit.org/b/21591> prepare-ChangeLog should know how to find functions in
22464         Perl files
22465
22466         Reviewed by David Kilzer.
22467
22468         * Scripts/prepare-ChangeLog: Removed redundant code that ignored certain files when
22469         searching for function line ranges. This is already done inside the get_function_line_ranges
22470         function.
22471         (get_function_line_ranges): Cleaned up coding style a little bit. Call
22472         get_function_line_ranges_for_perl for files with .pl and .pm extensions. For files with an
22473         unknown extension or no extension, read the shebang line to try to determine the script
22474         interpreter. Call get_function_line_ranges_for_perl if the interpreter seems to be Perl.
22475         (get_function_line_ranges_for_perl): Added. Does extremely basic parsing of the file to find
22476         lines starting with "sub " or "}".
22477
22478 2011-07-05  Adam Roben  <aroben@apple.com>
22479
22480         Clean up run-api-tests output on Windows
22481
22482         We were mixing run-api-tests output with gtest output, and the result was a mess.
22483
22484         Fixes <http://webkit.org/b/63954> run-api-tests output is very confusing on Windows
22485
22486         Reviewed by Dan Bates.
22487
22488         * Scripts/run-api-tests:
22489         (runTest):
22490         (populateTests):
22491         Made Windows use the formerly-Mac-only codepaths which correctly suppress gtest output
22492         except in verbose mode. Most changes are due to indentation. Use of the arch utility is now
22493         the only Mac-specific thing in these functions. Some other platform-specific code was moved
22494         from here...
22495
22496         (prepareEnvironmentForRunningTestTool):
22497         (testToolPath):
22498         ...to here.
22499
22500 2011-07-04  Gavin Barraclough  <barraclough@apple.com>
22501
22502         Reviewed by Sam Weinig.
22503
22504         https://bugs.webkit.org/show_bug.cgi?id=16652
22505         Firefox and JavaScriptCore differ in Number.toString(integer)
22506
22507         Added forwarding header.
22508
22509         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Added.
22510
22511 2011-07-04  Anders Carlsson  <andersca@apple.com>
22512
22513         Try to fix Qt build.
22514
22515         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
22516
22517 2011-07-04  Anders Carlsson  <andersca@apple.com>
22518
22519         NP_RemoveProperty is not called back by Safari when delete npObject.prop is encountered in JavaScript
22520         https://bugs.webkit.org/show_bug.cgi?id=63915
22521         <rdar://problem/7124300>
22522
22523         Reviewed by Sam Weinig.
22524
22525         Extend the NPRuntimeRemoveProperty to handle delete object.property from JavaScript.
22526
22527         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
22528         (PluginTest::NPN_IdentifierIsString):
22529         (PluginTest::NPN_UTF8FromIdentifier):
22530         (PluginTest::NPN_IntFromIdentifier):
22531         (PluginTest::NPN_RetainObject):
22532         (PluginTest::NPN_ReleaseObject):
22533         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
22534         (PluginTest::Object::removeProperty):
22535         (PluginTest::Object::identifierIs):
22536         (PluginTest::Object::NP_RemoveProperty):
22537         (PluginTest::Object::npClass):
22538         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
22539         (NPRuntimeRemoveProperty::TestObject::TestObject):
22540         (NPRuntimeRemoveProperty::TestObject::hasProperty):
22541         (NPRuntimeRemoveProperty::TestObject::getProperty):
22542         (NPRuntimeRemoveProperty::TestObject::removeProperty):
22543         (NPRuntimeRemoveProperty::PluginObject::PluginObject):
22544         (NPRuntimeRemoveProperty::PluginObject::~PluginObject):
22545         (NPRuntimeRemoveProperty::PluginObject::hasMethod):
22546         (NPRuntimeRemoveProperty::PluginObject::invoke):
22547         (NPRuntimeRemoveProperty::PluginObject::hasProperty):
22548         (NPRuntimeRemoveProperty::PluginObject::getProperty):
22549         (NPRuntimeRemoveProperty::NPP_GetValue):
22550
22551 2011-07-04  Leandro Pereira  <leandro@profusion.mobi>
22552
22553         Reviewed by Kent Tamura.
22554
22555         [EFL] Add LayoutTestControllerEfl.
22556         https://bugs.webkit.org/show_bug.cgi?id=61974
22557
22558         Only the core features are implemented. Most functions are actually
22559         stubs and will be implemented as soon as its features are needed.
22560         Build system changes will be made as soon as DRT is completely
22561         upstreamed.
22562
22563         * DumpRenderTree/efl/LayoutTestControllerEfl.cpp: Added.
22564         (LayoutTestController::~LayoutTestController):
22565         (LayoutTestController::addDisallowedURL):
22566         (LayoutTestController::clearBackForwardList):
22567         (LayoutTestController::copyDecodedHostName):
22568         (LayoutTestController::copyEncodedHostName):
22569         (LayoutTestController::dispatchPendingLoadRequests):
22570         (LayoutTestController::display):
22571         (LayoutTestController::counterValueForElementById):
22572         (LayoutTestController::keepWebHistory):
22573         (LayoutTestController::computedStyleIncludingVisitedInfo):
22574         (LayoutTestController::nodesFromRect):
22575         (LayoutTestController::layerTreeAsText):
22576         (LayoutTestController::pageNumberForElementById):
22577         (LayoutTestController::numberOfPages):
22578         (LayoutTestController::pageProperty):
22579         (LayoutTestController::isPageBoxVisible):
22580         (LayoutTestController::pageSizeAndMarginsInPixels):
22581         (LayoutTestController::webHistoryItemCount):
22582         (LayoutTestController::workerThreadCount):
22583         (LayoutTestController::notifyDone):
22584         (LayoutTestController::pathToLocalResource):
22585         (LayoutTestController::queueLoad):
22586         (LayoutTestController::setAcceptsEditing):
22587         (LayoutTestController::setAlwaysAcceptCookies):
22588         (LayoutTestController::setCustomPolicyDelegate):
22589         (LayoutTestController::waitForPolicyDelegate):
22590         (LayoutTestController::setScrollbarPolicy):
22591         (LayoutTestController::addOriginAccessWhitelistEntry):
22592         (LayoutTestController::removeOriginAccessWhitelistEntry):
22593         (LayoutTestController::setMainFrameIsFirstResponder):
22594         (LayoutTestController::setTabKeyCyclesThroughElements):
22595         (LayoutTestController::setUseDashboardCompatibilityMode):
22596         (LayoutTestController::setUserStyleSheetEnabled):
22597         (LayoutTestController::setUserStyleSheetLocation):
22598         (LayoutTestController::setValueForUser):
22599         (LayoutTestController::setViewModeMediaFeature):
22600         (LayoutTestController::setWindowIsKey):
22601         (LayoutTestController::setSmartInsertDeleteEnabled):
22602         (waitToDumpWatchdogFired):
22603         (LayoutTestController::setWaitToDump):
22604         (LayoutTestController::windowCount):
22605         (LayoutTestController::setPrivateBrowsingEnabled):
22606         (LayoutTestController::setJavaScriptCanAccessClipboard):
22607         (LayoutTestController::setXSSAuditorEnabled):
22608         (LayoutTestController::setFrameFlatteningEnabled):
22609         (LayoutTestController::setSpatialNavigationEnabled):
22610         (LayoutTestController::setAllowUniversalAccessFromFileURLs):
22611         (LayoutTestController::setAllowFileAccessFromFileURLs):
22612         (LayoutTestController::setAuthorAndUserStylesEnabled):
22613         (LayoutTestController::setAutofilled):
22614         (LayoutTestController::disableImageLoading):
22615         (LayoutTestController::setMockDeviceOrientation):
22616         (LayoutTestController::setMockGeolocationPosition):
22617         (LayoutTestController::setMockGeolocationError):
22618         (LayoutTestController::setGeolocationPermission):
22619         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
22620         (LayoutTestController::addMockSpeechInputResult):
22621         (LayoutTestController::setIconDatabaseEnabled):
22622         (LayoutTestController::setJavaScriptProfilingEnabled):
22623         (LayoutTestController::setSelectTrailingWhitespaceEnabled):
22624         (LayoutTestController::setPopupBlockingEnabled):
22625         (LayoutTestController::setPluginsEnabled):
22626         (LayoutTestController::elementDoesAutoCompleteForElementWithId):
22627         (LayoutTestController::execCommand):
22628         (LayoutTestController::findString):
22629         (LayoutTestController::isCommandEnabled):
22630         (LayoutTestController::setCacheModel):
22631         (LayoutTestController::setPersistentUserStyleSheetLocation):
22632         (LayoutTestController::clearPersistentUserStyleSheet):
22633         (LayoutTestController::clearAllApplicationCaches):
22634         (LayoutTestController::setApplicationCacheOriginQuota):
22635         (LayoutTestController::clearApplicationCacheForOrigin):
22636         (LayoutTestController::localStorageDiskUsageForOrigin):
22637         (LayoutTestController::originsWithApplicationCache):
22638         (LayoutTestController::applicationCacheDiskUsageForOrigin):
22639         (LayoutTestController::clearAllDatabases):
22640         (LayoutTestController::setDatabaseQuota):
22641         (LayoutTestController::originsWithLocalStorage):
22642         (LayoutTestController::deleteAllLocalStorage):
22643         (LayoutTestController::deleteLocalStorageForOrigin):
22644         (LayoutTestController::observeStorageTrackerNotifications):
22645         (LayoutTestController::syncLocalStorage):
22646         (LayoutTestController::setDomainRelaxationForbiddenForURLScheme):
22647         (LayoutTestController::goBack):
22648         (LayoutTestController::setDefersLoading):
22649         (LayoutTestController::setAppCacheMaximumSize):
22650         (LayoutTestController::pauseAnimationAtTimeOnElementWithId):
22651         (LayoutTestController::pauseTransitionAtTimeOnElementWithId):
22652         (LayoutTestController::sampleSVGAnimationForElementAtTime):
22653         (LayoutTestController::numberOfActiveAnimations):
22654         (LayoutTestController::suspendAnimations):
22655         (LayoutTestController::resumeAnimations):
22656         (LayoutTestController::overridePreference):
22657         (LayoutTestController::addUserScript):
22658         (LayoutTestController::addUserStyleSheet):
22659         (LayoutTestController::setDeveloperExtrasEnabled):
22660         (LayoutTestController::setAsynchronousSpellCheckingEnabled):
22661         (LayoutTestController::showWebInspector):
22662         (LayoutTestController::closeWebInspector):
22663         (LayoutTestController::evaluateInWebInspector):
22664         (LayoutTestController::evaluateScriptInIsolatedWorld):
22665         (LayoutTestController::removeAllVisitedLinks):
22666         (LayoutTestController::callShouldCloseOnWebView):
22667         (LayoutTestController::apiTestNewWindowDataLoadBaseURL):
22668         (LayoutTestController::apiTestGoToCurrentBackForwardItem):
22669         (LayoutTestController::setWebViewEditable):
22670         (LayoutTestController::markerTextForListItem):
22671         (LayoutTestController::authenticateSession):
22672         (LayoutTestController::setEditingBehavior):
22673         (LayoutTestController::abortModal):
22674         (LayoutTestController::hasSpellingMarker):
22675         (LayoutTestController::hasGrammarMarker):
22676         (LayoutTestController::dumpConfigurationForViewport):
22677         (LayoutTestController::setSerializeHTTPLoads):
22678         (LayoutTestController::setMinimumTimerInterval):
22679         (LayoutTestController::setTextDirection):
22680         (LayoutTestController::allowRoundingHacks):
22681
22682 2011-06-24  Vsevolod Vlasov  <vsevik@chromium.org>
22683
22684         Unreviewed, add myself as committer.
22685
22686         * Scripts/webkitpy/common/config/committers.py:
22687
22688 2011-07-03  Sheriff Bot  <webkit.review.bot@gmail.com>
22689
22690         Unreviewed, rolling out r90347.
22691         http://trac.webkit.org/changeset/90347
22692         https://bugs.webkit.org/show_bug.cgi?id=63886
22693
22694         Build breaks on Leopard, Chromium-win, WinCairo, and WinCE.
22695         (Requested by tkent on #webkit).
22696
22697         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Removed.
22698
22699 2011-07-03  Daniel Bates  <dbates@webkit.org>
22700
22701         Import Checkout in scm_unittests.py and logging in svn.py to fix
22702         Python missing global variable errors
22703
22704         Rubber-stamped by Adam Barth.
22705
22706         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py: Import Checkout.
22707         * Scripts/webkitpy/common/checkout/scm/svn.py: Import logging.
22708
22709 2011-07-02  Gavin Barraclough  <barraclough@apple.com>
22710
22711         Reviewed by Sam Weinig.
22712
22713         https://bugs.webkit.org/show_bug.cgi?id=16652
22714         Firefox and JavaScriptCore differ in Number.toString(integer)
22715
22716         Added forwarding header.
22717
22718         * DumpRenderTree/ForwardingHeaders/wtf/StdLibExtras.h: Added.
22719
22720 2011-04-02  Robert Hogan  <robert@webkit.org>
22721
22722         Reviewed by Benjamin Poulain.
22723
22724         [Qt] Support third-party cookie policy for Qt clients
22725         https://bugs.webkit.org/show_bug.cgi?id=45455
22726
22727         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
22728         (LayoutTestController::reset):
22729         (LayoutTestController::setAlwaysAcceptCookies):
22730         (LayoutTestController::setAlwaysBlockCookies):
22731         * DumpRenderTree/qt/LayoutTestControllerQt.h:
22732
22733 2011-07-03  Adam Barth  <abarth@webkit.org>
22734
22735         Factor ReflectionHandler out of the rebaseline server.  Just code motion.
22736
22737         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
22738         * Scripts/webkitpy/tool/servers/rebaselineserver.py:
22739         * Scripts/webkitpy/tool/servers/reflectionhandler.py: Added.
22740
22741 2011-07-03  Adam Barth  <abarth@webkit.org>
22742
22743         Move the rebaseline server out of the commands package and into the
22744         (new) servers package.  This patch prepares for refactoring this code
22745         to share server infrastructure with the (forthcoming) garden-o-matic
22746         command.
22747
22748         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
22749         * Scripts/webkitpy/tool/servers/__init__.py: Added.
22750         * Scripts/webkitpy/tool/servers/data/rebaselineserver/loupe.js:
22751         * Scripts/webkitpy/tool/servers/data/rebaselineserver/main.css:
22752         * Scripts/webkitpy/tool/servers/data/rebaselineserver/queue.js:
22753         * Scripts/webkitpy/tool/servers/rebaselineserver.py: Added.
22754
22755 2011-07-01  Adam Barth  <abarth@webkit.org>
22756
22757         Fix typo in master.cfg, hoping to trigger a master restart.
22758
22759         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
22760
22761 2011-07-01  Adam Barth  <abarth@webkit.org>
22762
22763         Turn off new-run-webkit-tests and remove some debug logging.
22764
22765         * Scripts/run-webkit-tests:
22766         * Scripts/webkitpy/layout_tests/port/http_lock.py:
22767
22768 2011-07-01  Adam Barth  <abarth@webkit.org>
22769
22770         Add some more debug logging to try to understand why we're getting
22771         stuck acquring the HTTP lock.
22772
22773         * Scripts/webkitpy/layout_tests/port/http_lock.py:
22774
22775 2011-07-01  Adam Barth  <abarth@webkit.org>
22776
22777         Refactor NRWT parsing logic in master.cfg to understand that run-webkit-tests might run NRWT
22778         https://bugs.webkit.org/show_bug.cgi?id=63854
22779
22780         Reviewed by Dirk Pranke.
22781
22782         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
22783
22784 2011-07-01  Eric Seidel  <eric@webkit.org>
22785
22786         new-run-webkit-tests results does not understand that mac uses test_expectations files
22787         https://bugs.webkit.org/show_bug.cgi?id=63838
22788
22789         Reviewed by Adam Barth.
22790
22791         We've decided to "enable" test_expectations for the Mac port for now.
22792         It makes the results.html results much more confusing to read, but at least they're
22793         no longer lying to us.
22794
22795         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
22796         * Scripts/webkitpy/layout_tests/port/base.py:
22797         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
22798
22799 2011-07-01  Adam Barth  <abarth@webkit.org>
22800
22801         Temporarily disable waiting for ReportCrash to see how this code
22802         affects performance on the bots.
22803
22804         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
22805         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
22806
22807 2011-07-01  Dmitry Lomov  <dslomov@google.com>
22808
22809         Reviewed by Adam Roben.
22810
22811         https://bugs.webkit.org/show_bug.cgi?id=63818
22812         Test bots shouldn't attempt to build TestWebKitAPI on Windows.
22813         TestWebKitAPI builds as a part of general webkit build and is a
22814         part of build products archive.
22815
22816         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
22817         * Scripts/run-api-tests:
22818
22819 2011-07-01  Adam Barth  <abarth@webkit.org>
22820
22821         We shouldn't wait for ReportCrash while the timeout is running
22822         https://bugs.webkit.org/show_bug.cgi?id=63837
22823
22824         Reviewed by Eric Seidel.
22825
22826         Instead, we should move this check to worker.py.
22827
22828         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
22829         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
22830
22831 2011-07-01  Adam Roben  <aroben@apple.com>
22832
22833         Stop setting unsafe headers on POSTs via XHR from TestFailures page
22834
22835         Fixes <http://webkit.org/b/63833> TestFailures page causes a bunch of spew in the
22836         Inspector's console about unsafe headers
22837
22838         Reviewed by Dan Bates.
22839
22840         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
22841         (fetchResource): Removed unnecessary code to set Content-Length and Connection headers.
22842
22843 2011-07-01  Eric Seidel  <eric@webkit.org>
22844
22845         Reviewed by Adam Barth.
22846
22847         Switch Snow Leopard over to new-run-webkit-tests (except for wk2)
22848         https://bugs.webkit.org/show_bug.cgi?id=63786
22849
22850         We're down to one bug blocking the "move to NRWT" bug:
22851         https://bugs.webkit.org/show_bug.cgi?id=34984
22852         so we're trying transitioning Snow Leopard to NRWT as an experiment.
22853
22854         If we run into any blocking issues we'll of course roll this back out
22855         (or anyone reading this should feel encouraged to).
22856
22857         * Scripts/run-webkit-tests:
22858
22859 2011-07-01  Adam Barth  <abarth@webkit.org>
22860
22861         Add a compat shim for --leaks.  Support for --leaks will be added in
22862         https://bugs.webkit.org/show_bug.cgi?id=63832.
22863
22864         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
22865
22866 2011-07-01  Adam Barth  <abarth@webkit.org>
22867
22868         Reviewed by Darin Adler.
22869
22870         ReportCrash destabilizes new-run-webkit-tests
22871         https://bugs.webkit.org/show_bug.cgi?id=63795
22872
22873         ReportCrash chews up too many system resources that it destablizes
22874         tests run concurrently.  This patch causes us to spin down all the
22875         worker processes when ReportCrash is running.
22876
22877         Also, this patch causes the master process to tell the user we're
22878         waiting for the crash reporter rather than having that be the worker's
22879         job.  This stops the user from getting spammed with too many "waiting
22880         for crash reporter" messages and also cleans up some minor UI issues
22881         with the meter.
22882
22883         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
22884         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
22885         * Scripts/webkitpy/layout_tests/port/base.py:
22886         * Scripts/webkitpy/layout_tests/port/server_process.py:
22887
22888 2011-07-01  Adam Roben  <aroben@apple.com>
22889
22890         Teach webkitpy about the new format of our ChangeLog template
22891
22892         r90229 moved the "Need a short description and bug URL" line to the top of the ChangeLog
22893         template. But webkitpy didn't know this, so the various rollout-related commands were
22894         leaving that line in the ChangeLog.
22895
22896         Fixes <http://webkit.org/b/63815> REGRESSION (r90229): webkit-patch rollout and sheriffbot
22897         rollout are broken
22898
22899         Reviewed by Dan Bates.
22900
22901         * Scripts/webkitpy/common/checkout/changelog.py:
22902         (ChangeLog.update_with_unreviewed_message): Changed to look for the "Need a short
22903         description and bug URL" line instead of the "Reviewed by" line as the first line of the
22904         ChangeLog template.
22905
22906         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
22907         (ChangeLogTest): Updated _new_entry_boilerplate to match what prepare-ChangeLog now
22908         generates.
22909
22910 2011-07-01  Adam Roben  <aroben@apple.com>
22911
22912         Make TestFailures stop loading old builds once we've explained all current failures
22913
22914         We used to do this, but some refactoring in r90165 broke this logic.
22915
22916         Fixes <http://webkit.org/b/63808> REGRESSION (r90165): TestFailures page always loads all
22917         builds back to the beginning of time
22918
22919         Reviewed by Dan Bates.
22920
22921         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
22922         (LayoutTestHistoryAnalyzer.prototype.start): Stop fetching old builds once we're told that
22923         all current failures have been explained.
22924
22925 2011-07-01  Adam Roben  <aroben@apple.com>
22926
22927         commit-log-editor reorders ChangeLog entries in unexpected ways
22928         https://bugs.webkit.org/show_bug.cgi?id=63804
22929
22930         Reviewed by Darin Adler.
22931
22932         commit-log-editor was changed in r46899 to move the bug title and URL to the top of the
22933         commit message in order to make git-based tools, which expect the first line of the commit
22934         message to be a summary of the change, to work better. But this reordering of the ChangeLog
22935         entry was unpredictable and unexpected.
22936
22937         I changed prepare-ChangeLog to put the bug title and URL at the top of the ChangeLog entry
22938         so that commit-log-editor doesn't have to move them. Then I changed commit-log-editor not to
22939         try to move them.
22940
22941         * Scripts/commit-log-editor: Partially reverted r46899. We no longer save the "Reviewed by"
22942         line and try to move it down below the bug title and URL. The order of the text in the
22943         ChangeLog is now preserved, except that we insert a "Patch by" line just above the "Reviewed
22944         by" line in cases where the committer and author are not the same person.
22945
22946         * Scripts/prepare-ChangeLog: Moved the "Reviewed by" line down below the bug title and URL.
22947
22948 2011-07-01  Mihnea Ovidenie  <mihnea@adobe.com>
22949
22950         Reviewed by Adam Roben.
22951
22952         [CSSRegions] Add build slave
22953         https://bugs.webkit.org/show_bug.cgi?id=62372
22954
22955         * BuildSlaveSupport/build.webkit.org-config/config.json:
22956         Add an entry for a release build of Mac.
22957
22958 2011-07-01  Sheriff Bot  <webkit.review.bot@gmail.com>
22959
22960         Unreviewed, rolling out r90213.
22961         http://trac.webkit.org/changeset/90213
22962         https://bugs.webkit.org/show_bug.cgi?id=63792
22963
22964         The SL bot has an errant HTTP server holding port 8080
22965         (Requested by abarth on #webkit).
22966
22967         * Scripts/run-webkit-tests:
22968
22969 2011-07-01  Carlos Garcia Campos  <cgarcia@igalia.com>
22970
22971         Unreviewed. Fix WebKit2 GTK build after r90163.
22972
22973         * MiniBrowser/gtk/WebBundle/WebBundleMain.c:
22974         (didCreatePage):
22975
22976 2011-07-01  Adam Barth  <abarth@webkit.org>
22977
22978         Exit faster when we get a keyboard interrupt.  Previously, we'd open
22979         Safari if you ^C NRWT after it found a failure.
22980
22981         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
22982
22983 2011-07-01  Eric Seidel  <eric@webkit.org>
22984
22985         Reviewed by Adam Barth.
22986
22987         Switch Snow Leopard over to new-run-webkit-tests (except for wk2)
22988         https://bugs.webkit.org/show_bug.cgi?id=63786
22989
22990         We're down to one bug blocking the "move to NRWT" bug:
22991         https://bugs.webkit.org/show_bug.cgi?id=34984
22992         so we're trying transitioning Snow Leopard to NRWT as an experiment.
22993
22994         If we run into any blocking issues we'll of course roll this back out
22995         (or anyone reading this should feel encouraged to).
22996
22997         * Scripts/run-webkit-tests:
22998
22999 2011-07-01  Sheriff Bot  <webkit.review.bot@gmail.com>
23000
23001         Unreviewed, rolling out r90192.
23002         http://trac.webkit.org/changeset/90192
23003         https://bugs.webkit.org/show_bug.cgi?id=63788
23004
23005         Appears to have caused NRWT on Chromium WebKit Vista to hang
23006         (Requested by abarth on #webkit).
23007
23008         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
23009         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
23010
23011 2011-06-30  Adam Barth  <abarth@webkit.org>
23012
23013         Reviewed by Eric Seidel.
23014
23015         Remove the concept of "being wedged" from new-run-webkit-tests
23016         https://bugs.webkit.org/show_bug.cgi?id=63767
23017
23018         Worker processes shouldn't ever become wedged.  My understanding is
23019         that this code was originally motivated by the old threading-based
23020         design but no longer servers any purpose.
23021
23022         Note: If we actually have a problem with the test harness getting
23023         stuck, buildbot will kill us.
23024
23025         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
23026         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
23027         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
23028
23029 2011-06-30  Adam Barth  <abarth@webkit.org>
23030
23031         The win32 implementation of fork is sad about forking test-webkitpy.
23032         Hopefully this will heal the bot.
23033
23034         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
23035
23036 2011-06-30  Eric Seidel  <eric@webkit.org>
23037
23038         Reviewed by Adam Barth.
23039
23040         Update new-run-webkit-tests now that the wk2 shared skipped list has moved
23041         https://bugs.webkit.org/show_bug.cgi?id=63780
23042
23043         The shared list was moved from mac-wk2 to wk2 as part of
23044         http://trac.webkit.org/changeset/90125.
23045
23046         * Scripts/webkitpy/layout_tests/port/webkit.py:
23047         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
23048
23049 2011-06-30  Kenichi Ishibashi  <bashi@chromium.org>
23050
23051         Reviewed by Darin Adler.
23052
23053         Style Checker should flag removal of "developmentRegion = English;" from project.pbxproj.
23054         https://bugs.webkit.org/show_bug.cgi?id=62022
23055
23056         Adds XcodeProjectFileChecker to check removal of "developmentRegion".
23057
23058         * Scripts/webkitpy/style/checker.py: Added XcodeProjectFileChecker as a checker for *.pbxproj.
23059         * Scripts/webkitpy/style/checkers/xcodeproj.py: Added.
23060         * Scripts/webkitpy/style/checkers/xcodeproj_unittest.py: Added.
23061
23062 2011-06-30  Dirk Pranke  <dpranke@chromium.org>
23063
23064         Reviewed by Ojan Vafai.
23065
23066         nrwt: make sharding tests needing locks less hard-coded
23067         https://bugs.webkit.org/show_bug.cgi?id=63112
23068
23069         This change also changes the manager logic so that it will
23070         drop the server lock as soon as all of the shards requiring
23071         the lock have completed.
23072
23073         This change includes some minor namespace/import changes in the
23074         unit tests, and also makes the Manager a new-style object, which
23075         it should've been all along.
23076
23077         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
23078         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
23079
23080 2011-06-30  Adam Barth  <abarth@webkit.org>
23081
23082         Reviewed by Eric Seidel.
23083
23084         Remove threaded mode from new-run-webkit-tests
23085         https://bugs.webkit.org/show_bug.cgi?id=63771
23086
23087         This mode is not used and is buggy.  Rather than carry around a bunch
23088         of unused buggy code, we should rip it out and focus on the
23089         multiprocess implementation.
23090
23091         * Scripts/webkitpy/common/system/executive.py:
23092         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
23093         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
23094         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
23095         * Scripts/webkitpy/layout_tests/port/base.py:
23096         * Scripts/webkitpy/layout_tests/port/chromium.py:
23097         * Scripts/webkitpy/layout_tests/port/mac.py:
23098         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
23099         * Scripts/webkitpy/layout_tests/port/server_process.py:
23100         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
23101         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
23102
23103 2011-06-30  Adam Barth  <abarth@webkit.org>
23104
23105         Reviewed by Darin Adler.
23106
23107         Clean up output from new-run-webkit-tests
23108         https://bugs.webkit.org/show_bug.cgi?id=63759
23109
23110         Printing messages from the child process looks super ugly because of
23111         the way the pretty-printer works.  Printing a blank line first is a
23112         hack, but it makes things at least partially sane.
23113
23114         Also, handle the case where calling sample throws an exception.
23115
23116         * Scripts/webkitpy/layout_tests/port/server_process.py:
23117         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
23118
23119 2011-06-27  Diego Gonzalez  <diegohcg@webkit.org>
23120
23121         Reviewed by Antonio Gomes.
23122
23123         [Qt] DRT: Provide LayoutTestController::setDefersLoading and ::goBack support
23124         https://bugs.webkit.org/show_bug.cgi?id=63489
23125
23126         Unskipped:
23127             loader/navigation-while-deferring-loads.html
23128
23129         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
23130         (LayoutTestController::goBack):
23131         (LayoutTestController::setDefersLoading):
23132         * DumpRenderTree/qt/LayoutTestControllerQt.h:
23133
23134 2011-06-30  Adam Barth  <abarth@webkit.org>
23135
23136         Reviewed by Dirk Pranke.
23137
23138         new-run-webkit-tests should stop when I ^C it
23139         https://bugs.webkit.org/show_bug.cgi?id=63763
23140
23141         It still doesn't stop fast enough, but at least with this patch it
23142         doesn't run more tests.
23143
23144         Also, I tweaked the UI that happens when there's a keyboard interrupt
23145         to not be amazingly ugly.  There's still a garbage "Te" left on the
23146         console, but that's a problem for another day.
23147
23148         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
23149
23150 2011-06-30  Adam Roben  <aroben@apple.com>
23151
23152         Make TestFailures never treat a too-many-failures build as a last-passing build for a test
23153
23154         Fixes <http://webkit.org/b/63758> TestFailures sometimes incorrectly says that a build was
23155         the last passing build for a test, when in fact the test wasn't run due to too many failures
23156         causing old-run-webkit-tests to exit early
23157
23158         Reviewed by Anders Carlsson.
23159
23160         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
23161         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): If there were too many
23162         failures in this build, assume that all the tests failed in the last build would have failed
23163         in this build, too, had they been run. Introduced some new local variables to reduce typing,
23164         and used them throughout.
23165
23166         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
23167         (LayoutTestResultsLoader.prototype.start): Fix a regression from r89841 where we wouldn't
23168         correctly remember whether old-run-webkit-tests exited early due to too many failures for
23169         builds older than r89610. Also removed an unused parameter while I was at it.
23170
23171 2011-06-30  Adam Roben  <aroben@apple.com>
23172
23173         Make TestFaiulres only load old-results directories as needed
23174
23175         Fixes <http://webkit.org/b/63752> Tester pages on TestFailures page load very slowly
23176
23177         Reviewed by Anders Carlsson.
23178
23179         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
23180         (Builder.prototype.getBuildNames): Moved up into the API section of the class. Now just
23181         calls through to _getBuildNamesFromResultsDirectory.
23182         (Builder.prototype.getOldBuildNames): Added. Just calls through to
23183         _getBuildNamesFromResultsDirectory.
23184         (Builder.prototype._getBuildNamesFromResultsDirectory): Renamed from getBuildNames. Now
23185         takes the directory URL as an argument and only fetches that single URL.
23186
23187         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
23188         (LayoutTestHistoryAnalyzer.prototype.start): Moved most logic to _analyzeBuilds. First
23189         analyzes builds from Builder.getBuildNames, then from Builder.getOldBuildNames if needed.
23190         (LayoutTestHistoryAnalyzer.prototype._analyzeBuilds): Moved logic here from start. (Most
23191         changes are just indentation.) Now takes a callback to call when we've finished analyzing
23192         all builds in buildNames so that we can try to fetch more build names if needed.
23193
23194 2011-06-30  Mark Rowe  <mrowe@apple.com>
23195
23196         Reviewed by Anders Carlsson.
23197
23198         Update MiniBrowser and WebKitTestRunner to use the client version constants.
23199
23200         * MiniBrowser/gtk/BrowserWindow.c:
23201         (browserWindowLoaderClientInit):
23202         (browserWindowUIClientInit):
23203         * MiniBrowser/gtk/WebBundle/WebBundleMain.c:
23204         (didCreatePage):
23205         (WKBundleInitialize):
23206         * MiniBrowser/gtk/main.c:
23207         (createWKContextWithInjectedBundle):
23208         * MiniBrowser/mac/AppDelegate.m:
23209         (-[BrowserAppDelegate init]):
23210         * MiniBrowser/mac/BrowserWindowController.m:
23211         (-[BrowserWindowController awakeFromNib]):
23212         * MiniBrowser/mac/WebBundle/WebBundleMain.m:
23213         (WKBundleInitialize):
23214         * MiniBrowser/win/BrowserView.cpp:
23215         (BrowserView::create):
23216         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
23217         (WTR::InjectedBundle::initialize):
23218         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
23219         (WTR::InjectedBundlePage::InjectedBundlePage):
23220         * WebKitTestRunner/TestController.cpp:
23221         (WTR::TestController::createOtherPage):
23222         (WTR::TestController::initialize):
23223
23224 2011-06-30  Adam Roben  <aroben@apple.com>
23225
23226         Identify revisions that modified failing tests on TestFailures page
23227
23228         Fixes <http://webkit.org/b/63716> TestFailures page should try to infer which commit caused
23229         a failure by looking at commit logs
23230
23231         Reviewed by David Kilzer.
23232
23233         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
23234         (.existing-bugs-list, .suspect-revisions-list): Made this rule apply to the suspect
23235         revisions list, too.
23236
23237         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
23238         (Trac): Initialize our memory cache.
23239         (Trac.prototype.getCommitDataForRevisionRange): Added. Fetches the commit log in RSS form,
23240         then parses out the revision number, Trac's idea of the commit title, and the files modified
23241         by the commit. (To get the modified files, we rely on the commit message including a
23242         prepare-ChangeLog-style file list.)
23243         (Trac.prototype.logURL): Added a new formatAsRSS parameter.
23244
23245         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
23246         (removePathExtension): Added. Returns a new string with the last dot and everything after it
23247         removed.
23248         (sorted): Added sortFunction parameter.
23249         (Node.prototype.removeAllChildren): Added. Does what it says.
23250
23251         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
23252         (ViewController.prototype._displayBuilder): Updated for change to _domForRegressionRange.
23253         (ViewController.prototype._domForRegressionRange): Changed argument order to match
23254         _domForNewAndExistingBugs. Now takes the list of failing tests and uses it to figure out
23255         which revisions modified the failing tests and displays the suspect revisions in a list.
23256
23257 2011-06-30  Eric Seidel  <eric@webkit.org>
23258
23259         Reviewed by Adam Barth.
23260
23261         Rename should_kill_dump_render_tree to driver_needs_restart in preparation for supporting WebKitTestRunner
23262         https://bugs.webkit.org/show_bug.cgi?id=63743
23263
23264         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
23265         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
23266
23267 2011-06-30  Martin Robinson  <mrobinson@igalia.com>
23268
23269         Reviewed by Anders Carlsson.
23270
23271         [GTK] Crash observed with nspluginwrapper and flash
23272         https://bugs.webkit.org/show_bug.cgi?id=62249
23273
23274         Added a TestNetscapePlugin test which verifies that WebKit properly
23275         handles situations where InvalidateRect is called with a null instance.
23276
23277         * DumpRenderTree/TestNetscapePlugIn/Tests/x11/CallInvalidateRectWithNullNPPArgument.cpp: Added.
23278         (CallInvalidateRectWithNullNPPArgument::CallInvalidateRectWithNullNPPArgument):
23279         (CallInvalidateRectWithNullNPPArgument::NPP_New):
23280         * GNUmakefile.am: Add the new file to sources list.
23281
23282 2011-06-30  Eric Seidel  <eric@webkit.org>
23283
23284         Reviewed by Adam Barth.
23285
23286         Remove redundant tests in TestFailuresTest
23287         https://bugs.webkit.org/show_bug.cgi?id=63742
23288
23289         These tests are all already covered by TestFailuresTest.test_loads.
23290         I also added the missing "T" from TestFailuresest. :)
23291
23292         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
23293
23294 2011-06-30  Adam Barth  <abarth@webkit.org>
23295
23296         Reviewed by Eric Seidel.
23297
23298         REGRESSION(r84294): new-run-webkit-tests results.html generate links to diffs.html or diff.png that don't exist
23299         https://bugs.webkit.org/show_bug.cgi?id=63494
23300
23301         This patch does two things:
23302
23303         1) Refactor diff_image to return the image diff instead of writing the
23304         diff to a file.  This fixes a bunch of hacks and disentangles a bunch
23305         of code.
23306
23307         2) When there's a checksum mismatch but not image diff, we no longer
23308         report an IMAGE failure to results.html.  That fixes the bug in
23309         question because results.html won't try to link to a non-existent image
23310         diff.
23311
23312         * Scripts/webkitpy/common/system/executive_mock.py:
23313         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
23314         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
23315         * Scripts/webkitpy/layout_tests/port/base.py:
23316         * Scripts/webkitpy/layout_tests/port/chromium.py:
23317         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
23318         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
23319         * Scripts/webkitpy/layout_tests/port/test.py:
23320         * Scripts/webkitpy/layout_tests/port/webkit.py:
23321         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
23322         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
23323
23324 2011-06-28  Hans Wennborg  <hans@chromium.org>
23325
23326         Reviewed by Tony Chang.
23327
23328         IndexedDB: Prepare for running layout tests with LevelDB
23329         https://bugs.webkit.org/show_bug.cgi?id=63593
23330
23331         The TestShell must always provide a temporary folder for LevelDB,
23332         not just when the backing store type is overridden.
23333
23334         * DumpRenderTree/chromium/LayoutTestController.cpp:
23335         (LayoutTestController::setOverrideIndexedDBBackingStore):
23336         (LayoutTestController::clearAllDatabases):
23337         * DumpRenderTree/chromium/LayoutTestController.h:
23338         * DumpRenderTree/chromium/TestShell.cpp:
23339         (TestShell::TestShell):
23340         * DumpRenderTree/chromium/TestShell.h:
23341
23342 2011-06-30  Zsolt Fehér  <h490267@stud.u-szeged.hu>
23343
23344         Reviewed by Csaba Osztrogonác.
23345
23346         [WK2] Create a general wk2 specific LayoutTests/platform/wk2 platform
23347         qt-wk2, win-wk2, mac-wk2 use this common Skipped list too
23348         https://bugs.webkit.org/show_bug.cgi?id=61785
23349
23350         * Scripts/old-run-webkit-tests:
23351
23352 2011-06-30  Adam Roben  <aroben@apple.com>
23353
23354         Use a colon to separate test names from failure info on TestFailures
23355
23356         This matches how the "Possibly Flaky Tests" list works, and is much prettier when the
23357         failure info itself includes parentheses.
23358
23359         Fixes <http://webkit.org/b/63722> Crashing symbols on TestFailures page look ugly due to
23360         double parentheses
23361
23362         Reviewed by Darin Adler.
23363
23364         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
23365         (ViewController.prototype._domForFailedTest): Use a colon instead of surrounding the failure
23366         info in parentheses.
23367
23368 2011-06-30  Adam Roben  <aroben@apple.com>
23369
23370         Robustify PersistentCache against cached data that contains the string ': '
23371
23372         Fixes <http://webkit.org/b/63721> Exceptions thrown while loading TestFailures page due to
23373         passing a bad string to JSON.parse
23374
23375         Reviewed by Darin Adler.
23376
23377         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/PersistentCache.js:
23378         (PersistentCache._parseDateAndJSONFromString): Manually split the string on the separator
23379         using indexOf and substring. (String.prototype.split always splits on all instances of the
23380         separator.)
23381
23382 2011-06-30  Adam Roben  <aroben@apple.com>
23383
23384         Include the crashing symbol in crash logs links on TestFailures
23385
23386         Fixes <http://webkit.org/b/63465> Links to crash logs on TestFailures page should include
23387         the crashing symbol (like links in results.html do)
23388
23389         Reviewed by David Kilzer.
23390
23391         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
23392         (LayoutTestResultsLoader.prototype.start): Bumped the cache version because we now store
23393         crashing symbols for crashing tests. Renamed testsForResultTable to parseResultTable because
23394         it now returns more than just the test names. Specifically, it now looks for crash log links
23395         and extracts the crashing symbol name from them. Updated callers of parseResultTable to
23396         match its new behavior. Changed to store the crashing symbol along with the failure type in
23397         the data we pass to the callback for tests which crashed.
23398
23399         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
23400         (code): Make <code> elements a little smaller because their contents can be quite long.
23401
23402         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
23403         (ViewController.prototype._domForFailureDiagnosis): Include the crashing symbol inside a
23404         <code> element in the link, if there is a crashing symbol.
23405
23406 2011-06-30  Adam Roben  <aroben@apple.com>
23407
23408         Use objects instead of strings to represent a test result in TestFailures code
23409
23410         This will eventually allow us to store more than just the type of failure for each test.
23411         (E.g., we can store the name of the crashing symbol for tests which crashed.)
23412
23413         Prep work for <http://webkit.org/b/63465> Links to crash logs on TestFailures page should
23414         include the crashing symbol (like links in results.html do)
23415
23416         Reviewed by David Kilzer.
23417
23418         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
23419         (Builder.prototype.failureDiagnosisTextAndURL): Changed to expect a testResult object
23420         instead of just a failureType string.
23421
23422         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
23423         (FlakyLayoutTestDetector.prototype.incorporateTestResults): Changed to store a
23424         testResult-like object for passing tests.
23425         (FlakyLayoutTestDetector.prototype.flakinessExamples): Changed to expect testResult-like
23426         objects.
23427
23428         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
23429         (LayoutTestResultsLoader.prototype.start): Store a version number along with the cached data
23430         so we can throw away cached data that's in an old format. Store a testResult object for each
23431         test instead of just its failure type.
23432
23433         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
23434         (ViewController.prototype._domForFailedTest):
23435         (ViewController.prototype._domForFailureDiagnosis):
23436         Changed to expect testResult objects instead of failureType strings.
23437
23438 2011-06-30  Adam Roben  <aroben@apple.com>
23439
23440         Show full commit logs when visiting Trac from TestFailures page
23441
23442         Fixes <http://webkit.org/b/63715> Links to Trac on TestFailures page
23443         should show full commit logs
23444
23445         Reviewed by David Kilzer.
23446
23447         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
23448         (Trac.prototype.logURL): Added new 'showFullCommitLogs' parameter. When specified, we add
23449         verbose=on to the log URL to tell Trac to show full commit logs.
23450
23451         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
23452         (ViewController.prototype._domForRegressionRange): Specify true for showFullCommitLogs.
23453
23454 2011-06-30  Adam Roben  <aroben@apple.com>
23455
23456         Add a link to file a bug about TestFailures to the TestFailures page
23457
23458         Fixes <http://webkit.org/b/63466> TestFailures page should include a link to file a bug
23459         about TestFailures itself
23460
23461         Reviewed by Anders Carlsson.
23462
23463         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
23464         (aside): Added some styles for the new link. It is positioned in the upper-right and
23465         hopefully not to glaringly obvious.
23466
23467         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
23468         (ViewController.prototype.loaded): Set up our main content element and auxiliary UI.
23469
23470         (ViewController.prototype._displayBuilder):
23471         (ViewController.prototype._displayTesters):
23472         Modify the main content element instead of the entire body, as there are other things in the
23473         body that we want to preserve.
23474
23475         (ViewController.prototype._domForAuxiliaryUIElements): Create and return an <aside> element
23476         that contains a link to file a bug about the TestFailures page.
23477
23478 2011-06-30  Adam Roben  <aroben@apple.com>
23479
23480         Ignore callbacks that correspond to old navigations of the TestFailures page
23481
23482         Fixes <http://webkit.org/b/63711> TestFailures page sometimes ignores back/forward
23483         navigations
23484
23485         Reviewed by Anders Carlsson.
23486
23487         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
23488         (LayoutTestHistoryAnalyzer.prototype.start): Check the return value of the callback. If the
23489         callback returns false, stop fetching more data.
23490
23491         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
23492         (ViewController): Initialize new _navigationID member. This member is used to assign each
23493         navigation by the user a unique ID.
23494         (ViewController.prototype.parseHash): Increment the navigation ID.
23495
23496         (ViewController.prototype._displayBuilder):
23497         (ViewController.prototype._displayTesters):
23498         At the start of these functions, record the current navigation ID. Bail out from callbacks
23499         if the navigation ID has changed, indicating that the user has navigated elsewhere.
23500
23501 2011-06-30  Carlos Garcia Campos  <cgarcia@igalia.com>
23502
23503         Reviewed by Martin Robinson.
23504
23505         [GTK] Add back/forward menu to MiniBrowser toolbar
23506         https://bugs.webkit.org/show_bug.cgi?id=63445
23507
23508         * MiniBrowser/gtk/BrowserWindow.c:
23509         (browser_window_init): Use a GtkMenuToolButton for navigation
23510         widgets.
23511         (browserWindowConstructed): Get the page BackForwardList.
23512         (browserWindowHistoryItemActivated): Go to activated history item.
23513         (browserWindowHistoryItemSelected): Show the url of currently
23514         selected history item in the statusbar.
23515         (createGtkActionFromBackForwardItem): Create a GtkAction from a
23516         WKBackForwardListItemRef.
23517         (browserWindowCreateMenuItemFromBackForwardItem): Create a
23518         GtkMenuItem from a WKBackForwardListItemRef.
23519         (browserWindowCreateBackForwardMenu): Create a GtkMenu with the
23520         given history items.
23521         (browserWindowUpdateNavigationActions): Update navigation widgets
23522         sensitivity and history menus.
23523         (didChangeBackForwardList): Call
23524         browserWindowUpdateNavigationActions().
23525         (browserWindowLoaderClientInit): Add didChangeBackForwardList
23526         callback.
23527
23528 2011-06-30  Dmitry Lomov  <dslomov@google.com>
23529
23530         Reviewed by Adam Roben.
23531
23532         https://bugs.webkit.org/show_bug.cgi?id=63660
23533         This enables unit-tests step on Windows and MacOS SnowLeopard.
23534
23535         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
23536
23537 2011-06-30  Balazs Kelemen  <kbalazs@webkit.org>
23538
23539         Reviewed by Kenneth Rohde Christiansen.
23540
23541         [Qt][WK2] Implement the rest of WTR::PlatformWebView
23542         https://bugs.webkit.org/show_bug.cgi?id=63630
23543
23544         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
23545         (WTR::PlatformWebView::windowFrame):
23546         (WTR::PlatformWebView::setWindowFrame):
23547
23548 2011-06-29  Dirk Pranke  <dpranke@chromium.org>
23549
23550         Reviewed by Adam Barth.
23551
23552         test-webkitpy is spamming lots of messages about updating test expectations
23553         https://bugs.webkit.org/show_bug.cgi?id=63680
23554
23555         This patch changes the passing_run() function to capture the
23556         output to stdout and stderr and assert that it is empty, and
23557         also specifies --no-new-test-results by default to avoid getting
23558         "generating baseline" messages for the expected-missing results.
23559
23560         Also check that we are logging messages when we generate new
23561         baselines.
23562
23563         Also fix a minor bug in the way test_tolerance() was
23564         checking for whether the tests were passing or not.
23565
23566         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
23567
23568 2011-06-29  Yuta Kitamura  <yutak@chromium.org>
23569
23570         Reviewed by Kent Tamura.
23571
23572         DRT: Make Hixie76WebSocketProtocolEnabled preference flag configurable from LayoutTestController
23573         https://bugs.webkit.org/show_bug.cgi?id=63532
23574
23575         LayoutTestController::overridePreference() will be used to change the value of
23576         the Hixie76WebSocketProtocolEnabled preference flag from layout tests. In this way,
23577         we can avoid writing a lot of boilerplate code in each port's DumpRenderTree, but
23578         at least we need to reset preference values in DumpRenderTree every time we start
23579         a new test because preference values changed in some test should not be leaked
23580         to the next test.
23581
23582         * DumpRenderTree/chromium/LayoutTestController.cpp:
23583         (LayoutTestController::overridePreference):
23584         The name of preference key is defined in WebKit/{mac/WebView,win}/WebPreferenceKeysPrivate.h.
23585         * DumpRenderTree/chromium/WebPreferences.cpp:
23586         (WebPreferences::reset):
23587         Set the default value. This will be called from TestShell::resetTestController().
23588         (WebPreferences::applyTo):
23589         * DumpRenderTree/chromium/WebPreferences.h:
23590         * DumpRenderTree/mac/DumpRenderTree.mm:
23591         (resetDefaultsToConsistentValues): Set the default value.
23592         * DumpRenderTree/win/DumpRenderTree.cpp:
23593         (resetDefaultsToConsistentValues): Ditto.
23594
23595 2011-06-29  Kenichi Ishibashi  <bashi@chromium.org>
23596
23597         Reviewed by Adam Barth.
23598
23599         check-webkit-style fails to detect https://webkit.org/b/XXXXX as a bug number
23600         https://webkit.org/b/57360
23601
23602         Modify the regular expression to match https scheme.
23603
23604         * Scripts/webkitpy/common/checkout/changelog_unittest.py: Added a test case for https://webkit.org/b/XXXXX.
23605         * Scripts/webkitpy/common/config/urls.py: Modified bug_url_short to match https scheme.
23606
23607 2011-06-29  Adam Barth  <abarth@webkit.org>
23608
23609         Reviewed by Eric Seidel.
23610
23611         new-run-webkit-tests: images page switch actual and expected images before they're loaded
23612         https://bugs.webkit.org/show_bug.cgi?id=63199
23613
23614         We now preload the images and keep the DOM nodes in memory instead of
23615         going back to the MemoryCache all the time.  (Also, remove some nutty
23616         </img> HTML.)
23617
23618         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
23619
23620 2011-06-29  Adam Barth  <abarth@webkit.org>
23621
23622         Reviewed by Eric Seidel.
23623
23624         new-run-webkit-tests should tell you when it creates new expectations
23625         https://bugs.webkit.org/show_bug.cgi?id=63675
23626
23627         Unforunately, this patch creates a bunch of test-webkitpy spam because
23628         we don't understand how Python logging works well enough to stop the
23629         spam.  :(
23630
23631         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
23632
23633 2011-06-29  Eric Seidel  <eric@webkit.org>
23634
23635         Reviewed by Adam Barth.
23636
23637         Refactor WebKitDriver.run_test into smaller pieces
23638         https://bugs.webkit.org/show_bug.cgi?id=63673
23639
23640         There is no functional change here.  Just moving code around.
23641         This is in preparation for adding support for #CRASHED and #CRASHED - WebProcess
23642
23643         * Scripts/webkitpy/layout_tests/port/webkit.py:
23644
23645 2011-06-29  Eric Seidel  <eric@webkit.org>
23646
23647         Reviewed by Adam Barth.
23648
23649         Simplify skipped file finding in preparation for adding wk2 skipped list fallback
23650         https://bugs.webkit.org/show_bug.cgi?id=63501
23651
23652         Re-landing after fixing a bug in WebKitPort.baseline_search_path which broke
23653         the rebaselineserver unit tests.  Also fixed nits brought up by Dirk after previous landing.
23654
23655         The goal was to add support for wk2 skipped lists.
23656         However, I found that our skipped list computation was a
23657         manual hack (which only worked for the mac port).
23658
23659         So I fixed a FIXME to move the skipped list finding
23660         into WebKitPort instead of MacPort.
23661         Doing so required the concept of a "port_name", since previously
23662         the only name accessible from a port object was name()
23663         which includes many things beyond "mac" or "chromium", etc.
23664
23665         Eventually I believe we'll want to re-think the way that we pass
23666         in a port_name argument to Port subclasses and expect them to parse
23667         it.  But for now I just added a cls.port_name variable which contains
23668         the static information needed to compute wk2 names as well as
23669         compute Skipped list fallback which works for Mac/Win/Qt and Gtk.
23670
23671         In order to test my new _skipped_file_search_paths method, I
23672         fixed another FIXME by making it return relative paths.
23673
23674         I also fixed the test_expectations_path code in WebKitPort to use port_name.
23675         It was using name() which would return PORT-VERSION so MacPort was overriding
23676         it to use just PORT.  After fixing test_expectations_path to use port_name
23677         (and making it aware of webkit2) I was able to remove the MacPort implementation.
23678
23679         * Scripts/webkitpy/layout_tests/port/base.py:
23680          - Add port_name() to access "mac", since name() returns "mac-leopard" etc.
23681          - Document that real_name() seems to have no purpose.
23682         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
23683          - Add _parse_port_name(), eventually we might call this from WebKitPort directly.
23684         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
23685          - Add _parse_port_name.
23686         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
23687          - Add _parse_port_name.
23688         * Scripts/webkitpy/layout_tests/port/gtk.py:
23689         * Scripts/webkitpy/layout_tests/port/mac.py:
23690          - Move Skipped-file finding code down to WebKitPort
23691         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
23692         * Scripts/webkitpy/layout_tests/port/qt.py:
23693         * Scripts/webkitpy/layout_tests/port/webkit.py:
23694         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
23695
23696 2011-06-29  Sheriff Bot  <webkit.review.bot@gmail.com>
23697
23698         Unreviewed, rolling out r90070.
23699         http://trac.webkit.org/changeset/90070
23700         https://bugs.webkit.org/show_bug.cgi?id=63676
23701
23702         Broke test-webkitpy rebaseline tests (Requested by eseidel2 on
23703         #webkit).
23704
23705         * Scripts/webkitpy/layout_tests/port/base.py:
23706         * Scripts/webkitpy/layout_tests/port/chromium.py:
23707         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
23708         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
23709         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
23710         * Scripts/webkitpy/layout_tests/port/gtk.py:
23711         * Scripts/webkitpy/layout_tests/port/mac.py:
23712         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
23713         * Scripts/webkitpy/layout_tests/port/qt.py:
23714         * Scripts/webkitpy/layout_tests/port/webkit.py:
23715         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
23716         * Scripts/webkitpy/layout_tests/port/win.py:
23717         * Scripts/webkitpy/tool/commands/rebaselineserver_unittest.py:
23718
23719 2011-06-29  Adam Barth  <abarth@webkit.org>
23720
23721         Reviewed by Eric Seidel.
23722
23723         new-run-webkit-tests doesn't support sample-on-timeout
23724         https://bugs.webkit.org/show_bug.cgi?id=56731
23725
23726         * Scripts/webkitpy/layout_tests/port/server_process.py:
23727         * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
23728         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
23729
23730 2011-06-29  Eric Seidel  <eric@webkit.org>
23731
23732         Reviewed by Adam Barth.
23733
23734         Simplify skipped file finding in preparation for adding wk2 skipped list fallback
23735         https://bugs.webkit.org/show_bug.cgi?id=63501
23736
23737         The goal was to add support for wk2 skipped lists.
23738         However, I found that our skipped list computation was a
23739         manual hack (which only worked for the mac port).
23740
23741         So I fixed a FIXME to move the skipped list finding
23742         into WebKitPort instead of MacPort.
23743         Doing so required the concept of a "port_name", since previously
23744         the only name accessible from a port object was name()
23745         which includes many things beyond "mac" or "chromium", etc.
23746
23747         Eventually I believe we'll want to re-think the way that we pass
23748         in a port_name argument to Port subclasses and expect them to parse
23749         it.  But for now I just added a cls.port_name variable which contains
23750         the static information needed to compute wk2 names as well as
23751         compute Skipped list fallback which works for Mac/Win/Qt and Gtk.
23752
23753         In order to test my new _skipped_file_search_paths method, I
23754         fixed another FIXME by making it return relative paths.
23755
23756         I also fixed the test_expectations_path code in WebKitPort to use port_name.
23757         It was using name() which would return PORT-VERSION so MacPort was overriding
23758         it to use just PORT.  After fixing test_expectations_path to use port_name
23759         (and making it aware of webkit2) I was able to remove the MacPort implementation.
23760
23761         * Scripts/webkitpy/layout_tests/port/base.py:
23762          - Add port_name() to access "mac", since name() returns "mac-leopard" etc.
23763          - Document that real_name() seems to have no purpose.
23764         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
23765          - Add _parse_port_name(), eventually we might call this from WebKitPort directly.
23766         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
23767          - Add _parse_port_name.
23768         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
23769          - Add _parse_port_name.
23770         * Scripts/webkitpy/layout_tests/port/gtk.py:
23771         * Scripts/webkitpy/layout_tests/port/mac.py:
23772          - Move Skipped-file finding code down to WebKitPort
23773         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
23774         * Scripts/webkitpy/layout_tests/port/qt.py:
23775         * Scripts/webkitpy/layout_tests/port/webkit.py:
23776         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
23777
23778 2011-06-29  Adam Barth  <abarth@webkit.org>
23779
23780         Reviewed by Dirk Pranke.
23781
23782         new-run-webkit-tests complains about missing pixel results instead of plopping down new expectations
23783         https://bugs.webkit.org/show_bug.cgi?id=38063
23784
23785         This patch changes new-run-webkit-tests to match old-run-webkit-tests
23786         in generating new expected results when the expected results are
23787         missing.
23788
23789         There are still a couple details that are different:
23790
23791         1) Image baselines aren't generated unless you run with -p.
23792         2) Render tree dumps are places in the cross-platform directory instead
23793         of the platform-specific directory.
23794
23795         I'm inclined to deal with both of these issues in follow-up patches.
23796
23797         * Scripts/webkitpy/common/net/layouttestresults.py:
23798         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
23799         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
23800         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
23801         * Scripts/webkitpy/layout_tests/port/test.py:
23802         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
23803         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
23804
23805 2011-06-29  Adam Roben  <aroben@apple.com>
23806
23807         Teach TestFailures to detect possibly flaky tests and list them separately
23808
23809         Fixes <http://webkit.org/b/61061> <rdar://problem/9452796> TestFailures page blames
23810         arbitrary revisions for breaking flaky tests
23811
23812         Reviewed by Dan Bates.
23813
23814         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js: Added.
23815         (FlakyLayoutTestDetector): This class identifies flaky tests when given the test results
23816         from various builds (in reverse-chronological order).
23817         (FlakyLayoutTestDetector.prototype.incorporateTestResults): Detects flaky tests. Tests move
23818         monotonically through three states: LastSeenFailing, LastSeenPassing, and PossiblyFlaky.
23819         (FlakyLayoutTestDetector.prototype.flakinessExamples): Finds examples of flakiness for the
23820         given test. Essentially, finds all the transitions from passing to failing (or vice-versa)
23821         and puts them in an array in reverse-chronological order.
23822         (FlakyLayoutTestDetector.prototype.get possiblyFlakyTests): Returns all tests we've detected
23823         to be possibly flaky.
23824
23825         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
23826         (LayoutTestHistoryAnalyzer): Initialize new members.
23827         (LayoutTestHistoryAnalyzer.prototype.start): Now passes the callback an object with two
23828         properties: history and possiblyFlaky. history holds the data this function used to pass to
23829         the callback, while possiblyFlaky lists all tests that might be flaky and examples of their
23830         flakiness. Updated documentation comment to match.
23831         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Now uses a
23832         FlakyLayoutTestDetector to identify possibly flaky tests. Any possibly flaky tests are
23833         removed from the failure history, since when they started failing is no longer meaningful.
23834         We tell our caller to keep calling until all current failures have been explained and we've
23835         gone through 5 builds without any new flaky tests being identified.
23836
23837         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
23838         (sorted): New helper function to return a sorted copy of an array.
23839         (Array.prototype.findLast): New helper function. Like findFirst, but finds the last item
23840         that matches the predicate.
23841
23842         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
23843         (ViewController.prototype._displayBuilder): Updated for change in the object passed to us by
23844         the analyzer. Now puts the list of possibly flaky tests after the failure history.
23845         (ViewController.prototype._domForFailedTest): Moved some code from here...
23846         (ViewController.prototype._domForFailureDiagnosis): ...to here.
23847         (ViewController.prototype._domForPossiblyFlakyTests): New function, builds up a list of
23848         possibly flaky tests and examples of their flakiness and returns it.
23849
23850         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
23851         FlakyLayoutTestDetector.js.
23852
23853 2011-06-29  Eric Seidel  <eric@webkit.org>
23854
23855         Adam says cowboys don't review (or unit test).
23856
23857         * Scripts/webkitpy/tool/commands/download.py:
23858
23859 2011-06-29  Eric Seidel  <eric@webkit.org>
23860
23861         Reviewed by Adam Barth.
23862
23863         Make port/config.py mockable for easier unit testing
23864         https://bugs.webkit.org/show_bug.cgi?id=63661
23865
23866         Fix an exception from the unit tests.
23867
23868         * Scripts/webkitpy/layout_tests/port/config.py:
23869
23870 2011-06-29  Brent Fulgham  <bfulgham@webkit.org>
23871
23872         Unreviewed build correction.
23873
23874         * DumpRenderTree/win/ImageDiffWinCairoCommon.vsprops:
23875         Link against zlib1.dll, not the static zlib.lib.
23876
23877 2011-06-29  Eric Seidel  <eric@webkit.org>
23878
23879         Reviewed by Adam Barth.
23880
23881         Make port/config.py mockable for easier unit testing
23882         https://bugs.webkit.org/show_bug.cgi?id=63661
23883
23884         Right now config.py leaks real system paths for
23885         webkit_base_dir due to using __file__.  This patch
23886         adds a filsystem.py wrapper around __file__ so that
23887         we don't have to jump through hoops in webkit_unittest.py
23888         to avoid leaking local paths to the unit test results.
23889
23890         There are probably many more places where we should now
23891         use filesystem.path_for_module, but starting with just this one.
23892
23893         * Scripts/webkitpy/common/system/filesystem.py:
23894         * Scripts/webkitpy/common/system/filesystem_mock.py:
23895         * Scripts/webkitpy/layout_tests/port/config.py:
23896         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
23897         * Scripts/webkitpy/tool/commands/rebaselineserver_unittest.py:
23898
23899 2011-06-29  Eric Seidel  <eric@webkit.org>
23900
23901         Reviewed by Adam Barth.
23902
23903         Remove duplicate methods in filesystem.py
23904         https://bugs.webkit.org/show_bug.cgi?id=63658
23905
23906         Looks like there was a bad merge at some point.
23907
23908         I also removed a bunch of redundant docstrings.
23909
23910         * Scripts/webkitpy/common/system/filesystem.py:
23911
23912 2011-06-29  Adam Barth  <abarth@webkit.org>
23913
23914         Reviewed by Eric Seidel.
23915
23916         Pass --no-new-test-results to NRWT on build.webkit.org
23917         https://bugs.webkit.org/show_bug.cgi?id=63653
23918
23919         This is in preparation for https://bugs.webkit.org/show_bug.cgi?id=38063.
23920
23921         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
23922
23923 2011-06-29  Xan Lopez  <xlopez@igalia.com>
23924
23925         Reviewed by Adam Barth.
23926
23927         [GTK] overlapping drag&drop tests fail on NRWT
23928         https://bugs.webkit.org/show_bug.cgi?id=57640
23929
23930         Make a GTK test driver that spawns one Xvfb instance per
23931         thread. This avoids bad interactions in DnD tests between threads.
23932
23933         * Scripts/webkitpy/layout_tests/port/gtk.py:
23934
23935 2011-06-29  Sheriff Bot  <webkit.review.bot@gmail.com>
23936
23937         Unreviewed, rolling out r89888.
23938         http://trac.webkit.org/changeset/89888
23939         https://bugs.webkit.org/show_bug.cgi?id=63609
23940
23941          "Seems to be causing merging issues on the dashboard, e.g.
23942         http://test-
23943         results.appspot.com/dashboards/flakiness_dashboard.html#group=%40ToT%20-%20chromium.org&tests=fast%2Fcanvas"
23944         (Requested by aboxhall on #webkit).
23945
23946         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
23947         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
23948         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
23949
23950 2011-06-29  Philippe Normand  <pnormand@igalia.com>
23951
23952         Reviewed by Martin Robinson.
23953
23954         [webkitpy] Executive.running_pids() sometimes doesn't correctly process "ps" output
23955         https://bugs.webkit.org/show_bug.cgi?id=63624
23956
23957         * Scripts/webkitpy/common/system/executive.py: In some cases the
23958         ps output lines can contain one or more leading white-spaces, so
23959         strip it before splitting the pid and command name.
23960
23961 2011-06-29  Noel Gordon  <noel.gordon@gmail.com>
23962
23963         Reviewed by Kent Tamura.
23964
23965         [Chromium] [DRT] ImageDiff should compile in VS2010
23966         https://bugs.webkit.org/show_bug.cgi?id=63614
23967
23968         * DumpRenderTree/chromium/ImageDiff.cpp: Include <iterator> for back_inserter.
23969
23970 2011-06-27  Diego Gonzalez  <diegohcg@webkit.org>
23971
23972         Reviewed by Kenneth Rohde Christiansen.
23973
23974         [Qt] DRT support for setInteractiveFormValidationEnabled
23975         https://bugs.webkit.org/show_bug.cgi?id=63496
23976
23977         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
23978         (WebCore::DumpRenderTree::DumpRenderTree):
23979
23980 2011-06-28  Eric Seidel  <eric@webkit.org>
23981
23982         Reviewed by Adam Barth.
23983
23984         Make new-run-webkit-tests -2 actually use WebKitTestRunner
23985         https://bugs.webkit.org/show_bug.cgi?id=63572
23986
23987         new-run-webkit-tests -2 works with this change.  However a zillion
23988         tests fail since it can't find the right skipped lists for wk2.
23989         The Skipped-list fixing is covered in bug 63501.
23990
23991         * Scripts/webkitpy/layout_tests/port/gtk.py:
23992         * Scripts/webkitpy/layout_tests/port/qt.py:
23993         * Scripts/webkitpy/layout_tests/port/webkit.py:
23994
23995 2011-06-28  Ryosuke Niwa  <rniwa@webkit.org>
23996
23997         Reviewed by Adam Roben.
23998
23999         Add more machines to Chromium Mac Release and Chromium Mac Release (Tests)
24000         https://bugs.webkit.org/show_bug.cgi?id=63500
24001
24002         Added google-mac-3 and google-mac-4.
24003
24004         * BuildSlaveSupport/build.webkit.org-config/config.json:
24005
24006 2011-06-28  Eric Seidel  <eric@webkit.org>
24007
24008         Reviewed by Dirk Pranke.
24009
24010         Remove webkitpy's tiger support now that WebKit no longer supports tiger
24011         https://bugs.webkit.org/show_bug.cgi?id=63568
24012
24013         I don't think our support actually fully worked anyway.  Regardless
24014         tiger is dead now (thankfully)!
24015
24016         I also unwrapped some short lines and added a couple FIXMEs while
24017         touring the code.
24018
24019         No new tests, only removing old ones which tested our Tiger support.
24020
24021         * Scripts/webkitpy/layout_tests/port/base.py:
24022         * Scripts/webkitpy/layout_tests/port/builders.py:
24023         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
24024         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
24025         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
24026         * Scripts/webkitpy/layout_tests/port/mac.py:
24027         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
24028         * Scripts/webkitpy/tool/commands/rebaseline.py:
24029         * Scripts/webkitpy/tool/commands/rebaselineserver_unittest.py:
24030
24031 2011-06-28  Eric Seidel  <eric@webkit.org>
24032
24033         Unreviewed, just fixing a typo.
24034
24035         Fix misspellings of "thead" to "thread".
24036
24037         * Scripts/webkitpy/common/checkout/scm/git.py:
24038         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
24039
24040 2011-06-28  Eric Seidel  <eric@webkit.org>
24041
24042         Reviewed by Tony Chang.
24043
24044         new-run-webkit-tests needs a --webkit-test-runner option
24045         https://bugs.webkit.org/show_bug.cgi?id=63439
24046
24047         This was accidentally fixed in bug 63501 as well.
24048         The root problem here was that the webkit unit tests were not using
24049         enough mocks.  I've fixed them here to use more mocks
24050         which should unbreak windows.
24051
24052         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
24053
24054 2011-06-28  Eric Seidel  <eric@webkit.org>
24055
24056         Reviewed by Tony Chang.
24057
24058         new-run-webkit-tests should not run mthml tests when MHTML support is disabled
24059         https://bugs.webkit.org/show_bug.cgi?id=63549
24060
24061         This way mhtml tests won't show up as "missing" results on all non-chromium platforms.
24062         I happen to go looking for what code was deciding that .mht was a supported
24063         extension, found test_files.py and then removed some redundant doc-strings and
24064         if branching.
24065
24066         * Scripts/webkitpy/layout_tests/port/test_files.py:
24067         * Scripts/webkitpy/layout_tests/port/webkit.py:
24068         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
24069
24070 2011-06-28  Roland Steiner  <rolandsteiner@chromium.org>
24071
24072         Reviewed by Tony Chang.
24073
24074         Fix filesystem_unittest.FileSystemTest test_read_and_write_file()
24075         https://bugs.webkit.org/show_bug.cgi?id=63514
24076
24077         - change write_text_file to write_binary_file
24078         - check that text_path and binary_path are actually existant files
24079         - change cleanup block from 'except' to 'finally'
24080
24081         * Scripts/webkitpy/common/system/filesystem_unittest.py:
24082
24083 2011-06-28  Adam Roben  <aroben@apple.com>
24084
24085         Roll out r89924 and r89925
24086
24087         I didn't mean to commit them!
24088
24089         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
24090         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
24091         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
24092
24093 2011-06-27  Adam Barth  <abarth@webkit.org>
24094
24095         Reviewed by Eric Seidel.
24096
24097         NRWT should wait for ReportCrash
24098         https://bugs.webkit.org/show_bug.cgi?id=63495
24099
24100         When a test crashes on Mac, ReportCrash confuses us into thinking the
24101         test times out.
24102
24103         I'm not sure how to test this in a meaningful way in our current test
24104         harness...
24105
24106         * Scripts/webkitpy/common/system/crashlogs.py:
24107         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
24108         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
24109         * Scripts/webkitpy/layout_tests/port/base.py:
24110         * Scripts/webkitpy/layout_tests/port/mac.py:
24111         * Scripts/webkitpy/layout_tests/port/server_process.py:
24112         * Scripts/webkitpy/tool/commands/queries.py:
24113
24114 2011-06-27  Sheriff Bot  <webkit.review.bot@gmail.com>
24115
24116         Unreviewed, rolling out r89881.
24117         http://trac.webkit.org/changeset/89881
24118         https://bugs.webkit.org/show_bug.cgi?id=63513
24119
24120         Makes Qt massively slow (Requested by abarth on #webkit).
24121
24122         * Scripts/webkitpy/common/system/crashlogs.py:
24123         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
24124         * Scripts/webkitpy/common/system/executive.py:
24125         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
24126         * Scripts/webkitpy/layout_tests/port/base.py:
24127         * Scripts/webkitpy/layout_tests/port/mac.py:
24128         * Scripts/webkitpy/layout_tests/port/server_process.py:
24129         * Scripts/webkitpy/tool/commands/queries.py:
24130
24131 2011-06-27  Alice Boxhall  <aboxhall@chromium.org>
24132
24133         Reviewed by Ojan Vafai.
24134
24135         Convert json_results_generator.py to output version 4 JSON.
24136         https://bugs.webkit.org/show_bug.cgi?id=60869
24137
24138         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
24139         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
24140         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
24141
24142 2011-06-27  Adam Barth  <abarth@webkit.org>
24143
24144         Reviewed by Eric Seidel.
24145
24146         NRWT should wait for ReportCrash
24147         https://bugs.webkit.org/show_bug.cgi?id=63495
24148
24149         When a test crashes on Mac, ReportCrash confuses us into thinking the
24150         test times out.
24151
24152         I'm not sure how to test this in a meaningful way in our current test
24153         harness...
24154
24155         * Scripts/webkitpy/common/system/crashlogs.py:
24156         * Scripts/webkitpy/common/system/crashlogs_unittest.py:
24157         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
24158         * Scripts/webkitpy/layout_tests/port/base.py:
24159         * Scripts/webkitpy/layout_tests/port/mac.py:
24160         * Scripts/webkitpy/layout_tests/port/server_process.py:
24161         * Scripts/webkitpy/tool/commands/queries.py:
24162
24163 2011-06-27  Adam Barth  <abarth@webkit.org>
24164
24165         Reviewed by Eric Seidel.
24166
24167         new-run-webkit-tests should upload crash logs
24168         https://bugs.webkit.org/show_bug.cgi?id=55907
24169
24170         NRWT used to have a different name for crash logs.  I've changed the
24171         name to patch ORWT.
24172
24173         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
24174         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
24175
24176 2011-06-27  Eric Seidel  <eric@webkit.org>
24177
24178         Reviewed by Adam Barth.
24179
24180         new-run-webkit-tests needs a --webkit-test-runner option
24181         https://bugs.webkit.org/show_bug.cgi?id=63439
24182
24183         NRWT doesn't actually know how to run with the WebKitTestRunner yet
24184         but it does have a --webkit-test-runner option and will build WebKitTestRunner correctly.
24185
24186         There is a bunch of other little cleanup in this patch which I added as I
24187         took a tour through all of our hard-coded DumpRenderTree strings.
24188
24189         * Scripts/webkitpy/common/net/layouttestresults.py: Added FIXME.
24190         * Scripts/webkitpy/common/system/outputcapture.py: Fixed spacing to pass PEP8.
24191         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Made the FIXME slightly stronger (bad bug).
24192         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: test_shell is dead.
24193         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Unwrapped silly wrapping.
24194         * Scripts/webkitpy/layout_tests/port/base.py: Removed two dead methods.
24195         * Scripts/webkitpy/layout_tests/port/config.py:
24196          - build_dumprendertree had no business in this class, removed it.
24197          - Exposed _FLAGS_FROM_CONFIGURATIONS through flag_for_configuration()
24198          - Exposed _script_path as script_path() (this probably belongs elsewhere).
24199         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
24200          - Moved these tests to webkit_unittests.
24201         * Scripts/webkitpy/layout_tests/port/webkit.py:
24202          - Now respects the --webkit-test-runner option.
24203          - setup_test_run is empty in base.py too, no need to override it.
24204         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
24205          - Test the new hotness.
24206          - Bad, bad, bad!  The old code was using a real Executive during unit-testing!
24207            I think this code is still hitting disk during the unit tests. :(
24208         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
24209          - Add the option.
24210         * Scripts/webkitpy/tool/bot/queueengine_unittest.py:
24211          - Remove optional args option.
24212         * Scripts/webkitpy/tool/commands/commandtest.py:
24213          - Remove optional args option.
24214         * Scripts/webkitpy/tool/commands/queues_unittest.py:
24215          - Remove optional args option.
24216
24217 2011-06-27  Adam Barth  <abarth@webkit.org>
24218
24219         Reviewed by Dirk Pranke.
24220
24221         new-run-webkit-tests does not report stderr output
24222         https://bugs.webkit.org/show_bug.cgi?id=37739
24223
24224         Almost all the infrastructure appears to be in place to report stderr
24225         in results.html.  We just need to wire up a bool to tell results.html
24226         to create the hyperlink.
24227
24228         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
24229         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
24230         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
24231         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
24232         * Scripts/webkitpy/layout_tests/port/base.py:
24233
24234 2011-06-27  Adam Barth  <abarth@webkit.org>
24235
24236         Reviewed by Dirk Pranke.
24237
24238         webkitpy unit tests should have more descriptive names than just "Test"
24239         https://bugs.webkit.org/show_bug.cgi?id=63479
24240
24241         Luckily we kept these classes in separate namespaces in the harness so
24242         we were actually running them all even though they had the same name.
24243
24244         * Scripts/webkitpy/common/system/stack_utils_unittest.py:
24245         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
24246         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
24247
24248 2011-06-27  Greg Simon  <gregsimon@chromium.org>
24249
24250         No review necessary.
24251
24252         Add email addr to contributor list
24253         https://bugs.webkit.org/show_bug.cgi?id=63478
24254
24255         * Scripts/webkitpy/common/config/committers.py:
24256
24257 2011-06-27  Sheriff Bot  <webkit.review.bot@gmail.com>
24258
24259         Unreviewed, rolling out r89834.
24260         http://trac.webkit.org/changeset/89834
24261         https://bugs.webkit.org/show_bug.cgi?id=63476
24262
24263         Broke test-webkitpy (Requested by abarth on #webkit).
24264
24265         * Scripts/webkitpy/common/net/layouttestresults.py:
24266         * Scripts/webkitpy/common/system/outputcapture.py:
24267         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
24268         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
24269         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
24270         * Scripts/webkitpy/layout_tests/port/base.py:
24271         * Scripts/webkitpy/layout_tests/port/config.py:
24272         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
24273         * Scripts/webkitpy/layout_tests/port/webkit.py:
24274         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
24275         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
24276         * Scripts/webkitpy/tool/bot/queueengine_unittest.py:
24277         * Scripts/webkitpy/tool/commands/commandtest.py:
24278         * Scripts/webkitpy/tool/commands/queues_unittest.py:
24279
24280 2011-06-27  Jeff Miller  <jeffm@apple.com>
24281
24282         Reviewed by Adam Roben.
24283
24284         Stop copying obsolete WebKit libraries on Windows
24285         https://bugs.webkit.org/show_bug.cgi?id=63474
24286
24287         Don't copy obsolete WebKit library directories and files on Windows, they no longer exist and just generate extraneous warnings. These include the CharacterSets directory, dnssd.dll, and icudt40.dll.
24288
24289         * DumpRenderTree/win/DumpRenderTreePostBuild.cmd:
24290         * MiniBrowser/MiniBrowserPostBuild.cmd:
24291         * WebKitTestRunner/win/WebKitTestRunnerPostBuild.cmd:
24292
24293 2011-06-27  Ryosuke Niwa  <rniwa@webkit.org>
24294
24295         Add Wyatt Carss to the list of non-committer contributors.
24296
24297         * Scripts/webkitpy/common/config/committers.py:
24298
24299 2011-06-27  Adam Barth  <abarth@webkit.org>
24300
24301         Reviewed by Eric Seidel.
24302
24303         webkitpy should understand crash logs
24304         https://bugs.webkit.org/show_bug.cgi?id=63468
24305
24306         We're planning to use this functionality to upload crash logs along
24307         with test results for new-run-webkit-tests.
24308
24309         * Scripts/webkitpy/common/system/crashlog.py: Added.
24310         * Scripts/webkitpy/common/system/crashlog_unittest.py: Added.
24311         * Scripts/webkitpy/common/system/executive.py:
24312         * Scripts/webkitpy/common/system/executive_unittest.py:
24313         * Scripts/webkitpy/common/system/filesystem.py:
24314         * Scripts/webkitpy/common/system/filesystem_mock.py:
24315         * Scripts/webkitpy/tool/commands/queries.py:
24316
24317 2011-06-27  Adam Roben  <aroben@apple.com>
24318
24319         Make LayoutTestResultsLoader cache whether old-run-webkit-tests exited early due to too many
24320         failures
24321
24322         Fixes <http://webkit.org/b/63470> TestFailures page for a particular builder forgets
24323         old-run-webkit-tests exited early after reload
24324
24325         Reviewed by Anders Carlsson.
24326
24327         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
24328         (LayoutTestResultsLoader.prototype.start): Store both the set of failing tests and whether
24329         old-run-webkit-tests exited early due to too many failures in PersistentCache.
24330
24331 2011-06-27  Adam Roben  <aroben@apple.com>
24332
24333         Make TestFailures load build names from build.webkit.org/old-results too
24334
24335         Build results are periodically moved from results to old-results. This change makes those
24336         builds still visible to TestFailures.
24337
24338         Fixes <http://webkit.org/b/63453> TestFailures page doesn't show information for builds that
24339         have been moved to build.webkit.org/old-results
24340
24341         Reviewed by Anders Carlsson.
24342
24343         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
24344         (Builder.prototype.getBuildNames): Extracted code to parse build names from a
24345         build.webkit.org directory listing into a separate function. Instead of omitting .zip files,
24346         we now only include directory entries whose names are parseable as build names (since
24347         old-results sometimes contains other random files/directories from who knows what). We now
24348         fetch both results and old-results (with a FIXME about loading old-results on demand),
24349         extract build names from each, and concatenate the two sets of names.
24350
24351         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js:
24352         (WebKitBuildbot.prototype.parseBuildName): Changed to return null when the build name isn't
24353         parseable, rather than throwing an exception.
24354
24355 2011-06-26  Adam Roben  <aroben@apple.com>
24356
24357         Store analyzed history directly in LayoutTestHistoryAnalyzer instead of in a generic cache
24358
24359         It seems silly to have a whole generic cache just to store a single object.
24360
24361         Fixes <http://webkit.org/b/63407> LayoutTestHistoryAnalyzer has a generic cache but doesn't
24362         need one
24363
24364         Reviewed by Anders Carlsson.
24365
24366         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
24367         (LayoutTestHistoryAnalyzer): Replaced this._cache with this._history.
24368         (LayoutTestHistoryAnalyzer.prototype.start): Ditto.
24369         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): Ditto, and removed the
24370         no-longer-needed history argument.
24371
24372 2011-06-26  Adam Roben  <aroben@apple.com>
24373
24374         Move some inappropriate code out of Builder
24375
24376         Fixes <http://webkit.org/b/63406> TestFailure page's Builder class has a bunch of code that
24377         isn't really about a builder
24378
24379         Reviewed by Anders Carlsson.
24380
24381         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
24382         (Builder.prototype.getBuildNames): Renamed from _getBuildNames, since it now needs to be
24383         called by other objects.
24384
24385         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
24386         Added. Code came from Builder.
24387         (LayoutTestHistoryAnalyzer):
24388         (LayoutTestHistoryAnalyzer.prototype.start):
24389         (LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory):
24390
24391         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
24392         Added. Code came from Builder.
24393         (LayoutTestResultsLoader):
24394         (LayoutTestResultsLoader.prototype.start):
24395
24396         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
24397         (ViewController.prototype._displayBuilder): Changed to use LayoutTestHistoryAnalyzer.
24398
24399         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in new
24400         scripts.
24401
24402 2011-05-17  Nat Duca  <nduca@chromium.org>
24403
24404         Reviewed by James Robinson.
24405
24406         [chromium] Implement CCLayerTreeHost and CCLayerTreeHostImpl portions of threaded compositor
24407         https://bugs.webkit.org/show_bug.cgi?id=58408
24408
24409         * DumpRenderTree/chromium/WebViewHost.cpp:
24410         (WebViewHost::paintInvalidatedRegion):
24411
24412 2011-06-27  Eric Seidel  <eric@webkit.org>
24413
24414         Reviewed by Adam Barth.
24415
24416         new-run-webkit-tests needs a --webkit-test-runner option
24417         https://bugs.webkit.org/show_bug.cgi?id=63439
24418
24419         NRWT doesn't actually know how to run with the WebKitTestRunner yet
24420         but it does have a --webkit-test-runner option and will build WebKitTestRunner correctly.
24421
24422         There is a bunch of other little cleanup in this patch which I added as I
24423         took a tour through all of our hard-coded DumpRenderTree strings.
24424
24425         * Scripts/webkitpy/common/net/layouttestresults.py: Added FIXME.
24426         * Scripts/webkitpy/common/system/outputcapture.py: Fixed spacing to pass PEP8.
24427         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Made the FIXME slightly stronger (bad bug).
24428         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: test_shell is dead.
24429         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py: Unwrapped silly wrapping.
24430         * Scripts/webkitpy/layout_tests/port/base.py: Removed two dead methods.
24431         * Scripts/webkitpy/layout_tests/port/config.py:
24432          - build_dumprendertree had no business in this class, removed it.
24433          - Exposed _FLAGS_FROM_CONFIGURATIONS through flag_for_configuration()
24434          - Exposed _script_path as script_path() (this probably belongs elsewhere).
24435         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
24436          - Moved these tests to webkit_unittests.
24437         * Scripts/webkitpy/layout_tests/port/webkit.py:
24438          - Now respects the --webkit-test-runner option.
24439          - setup_test_run is empty in base.py too, no need to override it.
24440         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
24441          - Test the new hotness.
24442          - Bad, bad, bad!  The old code was using a real Executive during unit-testing!
24443            I think this code is still hitting disk during the unit tests. :(
24444         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
24445          - Add the option.
24446         * Scripts/webkitpy/tool/bot/queueengine_unittest.py:
24447          - Remove optional args option.
24448         * Scripts/webkitpy/tool/commands/commandtest.py:
24449          - Remove optional args option.
24450         * Scripts/webkitpy/tool/commands/queues_unittest.py:
24451          - Remove optional args option.
24452
24453 2011-06-27  Chang Shu  <cshu@webkit.org>
24454
24455         Reviewed by Darin Adler.
24456
24457         [WK2] WebKitTestRunner should reset FrameFlattening flag to disable before each test
24458         https://bugs.webkit.org/show_bug.cgi?id=63457
24459
24460         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
24461         (WTR::InjectedBundle::beginTesting):
24462
24463 2011-06-27  Raphael Kubo da Costa  <kubo@profusion.mobi>
24464
24465         Reviewed by Kent Tamura.
24466
24467         [EFL] DRT: Add an ewk_view specialization for EFL's DumpRenderTree.
24468         It will be plugged into the build system as soon as the rest of the
24469         DRT code is fully upstreamed.
24470         https://bugs.webkit.org/show_bug.cgi?id=63086
24471
24472         * DumpRenderTree/efl/DumpRenderTreeView.cpp: Added.
24473         (onConsoleMessage):
24474         (onJavaScriptAlert):
24475         (onJavaScriptConfirm):
24476         (onJavaScriptPrompt):
24477         (drtViewTiledAdd):
24478         * DumpRenderTree/efl/DumpRenderTreeView.h: Added.
24479
24480 2011-06-27  Leandro Pereira  <leandro@profusion.mobi>
24481
24482         Reviewed by Kent Tamura.
24483
24484         [EFL] Add an EventSender implementation to EFL's DumpRenderTree along
24485         with support files.
24486         Build system changes will be made as soon as the whole DumpRenderTree
24487         is upstreamed.
24488         https://bugs.webkit.org/show_bug.cgi?id=61962
24489
24490         * DumpRenderTree/efl/EventSender.cpp: Added.
24491         (setEvasModifiers):
24492         (translateMouseButtonNumber):
24493         (sendMouseEvent):
24494         (sendClick):
24495         (scheduleAsynchronousClickCallback):
24496         (updateClickCount):
24497         (modifierFromJSValue):
24498         (modifiersFromJSValue):
24499         (mouseDownCallback):
24500         (mouseUpCallback):
24501         (mouseMoveToCallback):
24502         (evasMouseEventFromHorizontalAndVerticalOffsets):
24503         (mouseScrollByCallback):
24504         (continuousMouseScrollByCallback):
24505         (keyPadNameFromJSValue):
24506         (keyNameFromJSValue):
24507         (keyDownCallback):
24508         (scalePageByCallback):
24509         (getClass):
24510         (makeEventSender):
24511         * DumpRenderTree/efl/EventSender.h: Added.
24512
24513 2011-06-27  Eric Seidel  <eric@webkit.org>
24514
24515         Reviewed by Adam Barth.
24516
24517         Remove evil uses of hasattr
24518         https://bugs.webkit.org/show_bug.cgi?id=63430
24519
24520         For some reason these classes believe that they may be called with
24521         various flavors of "option" elements and so carefully check to make
24522         sure that the options element has their option before checking it.
24523
24524         We had a set_option_default method which was never called, so I made it
24525         do what callsites seemed to want it to do and replaced 3 callers
24526         who previously used hasattr manually to use set_option_default instead.
24527
24528         * Scripts/webkitpy/layout_tests/port/base.py:
24529         * Scripts/webkitpy/layout_tests/port/webkit.py:
24530
24531 2011-06-27  Kent Tamura  <tkent@chromium.org>
24532
24533         Reviewed by Hajime Morita.
24534
24535         [Chromium] WebViewHost should be destructed before LayoutTestController, etc.
24536         https://bugs.webkit.org/show_bug.cgi?id=63427
24537
24538         ~WebViewHost() navigates to about:blank, so it calls some
24539         callbacks and LayoutTestController functions.
24540
24541         * DumpRenderTree/chromium/TestShell.h:
24542          Move m_printer upper and move m_webViewHost lower.
24543
24544 2011-06-27  Adam Barth  <abarth@webkit.org>
24545
24546         Reviewed by Eric Seidel.
24547
24548         Move server config files to webkitpy.layout_tests.servers
24549         https://bugs.webkit.org/show_bug.cgi?id=63436
24550
24551         The config files belong with the servers.
24552
24553         * Scripts/webkitpy/layout_tests/port/httpd2.pem: Removed.
24554         * Scripts/webkitpy/layout_tests/port/lighttpd.conf: Removed.
24555         * Scripts/webkitpy/layout_tests/servers/http_server.py:
24556         * Scripts/webkitpy/layout_tests/servers/httpd2.pem: Copied from Tools/Scripts/webkitpy/layout_tests/port/httpd2.pem.
24557         * Scripts/webkitpy/layout_tests/servers/lighttpd.conf: Copied from Tools/Scripts/webkitpy/layout_tests/port/lighttpd.conf.
24558
24559 2011-06-27  Adam Barth  <abarth@webkit.org>
24560
24561         Reviewed by Eric Seidel.
24562
24563         Move websocket_server and http_server to webkitpy.layout_package.servers
24564         https://bugs.webkit.org/show_bug.cgi?id=63434
24565
24566         These classes do not represent ports.  They represent servers.
24567
24568         * Scripts/new-run-webkit-httpd:
24569         * Scripts/new-run-webkit-websocketserver:
24570         * Scripts/webkitpy/layout_tests/port/base.py:
24571         * Scripts/webkitpy/layout_tests/port/chromium.py:
24572         * Scripts/webkitpy/layout_tests/port/http_server.py: Removed.
24573         * Scripts/webkitpy/layout_tests/port/http_server_base.py: Removed.
24574         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Removed.
24575         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
24576         * Scripts/webkitpy/layout_tests/port/websocket_server.py: Removed.
24577         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py:
24578         * Scripts/webkitpy/layout_tests/servers/http_server.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/http_server.py.
24579         * Scripts/webkitpy/layout_tests/servers/http_server_base.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/http_server_base.py.
24580         * Scripts/webkitpy/layout_tests/servers/http_server_integrationtest.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py.
24581         * Scripts/webkitpy/layout_tests/servers/websocket_server.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/websocket_server.py.
24582
24583 2011-06-27  Eric Seidel  <eric@webkit.org>
24584
24585         Reviewed by Adam Barth.
24586
24587         Remove Chromium and ChromiumMac implementations of driver_name now that everything uses DumpRenderTree
24588         https://bugs.webkit.org/show_bug.cgi?id=63428
24589
24590         * Scripts/webkitpy/layout_tests/port/chromium.py:
24591         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
24592
24593 2011-06-27  Adam Barth  <abarth@webkit.org>
24594
24595         Reviewed by Eric Seidel.
24596
24597         Move apache_http_server.py to webkitpy.layout_package.servers
24598         https://bugs.webkit.org/show_bug.cgi?id=63433
24599
24600         The apache_http_server.py is a server, not a port.
24601
24602         * Scripts/webkitpy/layout_tests/port/apache_http_server.py: Removed.
24603         * Scripts/webkitpy/layout_tests/port/base.py:
24604         * Scripts/webkitpy/layout_tests/servers: Added.
24605         * Scripts/webkitpy/layout_tests/servers/__init__.py: Added.
24606         * Scripts/webkitpy/layout_tests/servers/apache_http_server.py: Copied from Tools/Scripts/webkitpy/layout_tests/port/apache_http_server.py.
24607
24608 2011-06-27  Adam Barth  <abarth@webkit.org>
24609
24610         Reviewed by Eric Seidel.
24611
24612         Move rebaseline-chromium-webkit-tests.py to webkitpy.to_be_moved
24613         https://bugs.webkit.org/show_bug.cgi?id=63432
24614
24615         This code isn't in the dependency cone of new-run-webkit-tests. It's
24616         unclear what it's final home should be, but this patch moves it out of
24617         layout_tests to simply the package.
24618
24619         * Scripts/rebaseline-chromium-webkit-tests:
24620         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py: Removed.
24621         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py: Removed.
24622         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py: Copied from Tools/Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py.
24623         * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py.
24624
24625 2011-06-27  Adam Barth  <abarth@webkit.org>
24626
24627         Reviewed by Eric Seidel.
24628
24629         Move update_webgl_conformance_tests.py to webkitpy.to_be_moved
24630         https://bugs.webkit.org/show_bug.cgi?id=63429
24631
24632         This package as nothing to do with the layout_tests package.
24633
24634         * Scripts/update-webgl-conformance-tests:
24635         * Scripts/webkitpy/layout_tests/update_webgl_conformance_tests.py: Removed.
24636         * Scripts/webkitpy/layout_tests/update_webgl_conformance_tests_unittest.py: Removed.
24637         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py: Copied from Tools/Scripts/webkitpy/layout_tests/update_webgl_conformance_tests.py.
24638         * Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/update_webgl_conformance_tests_unittest.py.
24639
24640 2011-06-27  Adam Barth  <abarth@webkit.org>
24641
24642         Reviewed by Eric Seidel.
24643
24644         Move deduplicate_tests.py into webkitpy.to_be_moved
24645         https://bugs.webkit.org/show_bug.cgi?id=63426
24646
24647         deduplicate_tests.py doesn't haven anything to do with layout tests.
24648         It's really a separate script.  This patch moves it to the (new)
24649         to_be_moved package so that it doesn't distract us when working on NRWT.
24650
24651         * Scripts/deduplicate-tests:
24652         * Scripts/webkitpy/layout_tests/deduplicate_tests.py: Removed.
24653         * Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py: Removed.
24654         * Scripts/webkitpy/test/main.py:
24655         * Scripts/webkitpy/to_be_moved: Added.
24656         * Scripts/webkitpy/to_be_moved/__init__.py: Added.
24657         * Scripts/webkitpy/to_be_moved/deduplicate_tests.py: Copied from Tools/Scripts/webkitpy/layout_tests/deduplicate_tests.py.
24658         * Scripts/webkitpy/to_be_moved/deduplicate_tests_unittest.py: Copied from Tools/Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py.
24659
24660 2011-06-26  Adam Barth  <abarth@webkit.org>
24661
24662         Rubber-stamped by Eric Seidel.
24663
24664         Remove empty directory.
24665
24666         * Scripts/webkitpy/layout_tests/test_types: Removed.
24667
24668 2011-06-25  Naoki Takano  <honten@chromium.org>
24669
24670         Reviewed by Eric Seidel.
24671
24672         Change my entry from contributors to committers.
24673         https://bugs.webkit.org/show_bug.cgi?id=63043
24674
24675         * Scripts/webkitpy/common/config/committers.py: Change my e-mail from takano.naoki@gmail.com to honten@chromium.org. And move the entry to commiters from contributors.
24676
24677 2011-06-25  Adam Barth  <abarth@webkit.org>
24678
24679         Reviewed by Eric Seidel.
24680
24681         [Chromium] Update WebKit for WEBKIT_FRAME_TO_DOCUMENT_API_MOVE
24682         https://bugs.webkit.org/show_bug.cgi?id=63394
24683
24684         * DumpRenderTree/chromium/LayoutTestController.cpp:
24685         (LayoutTestController::queueLoad):
24686
24687 2011-06-25  Brent Fulgham  <bfulgham@webkit.org>
24688
24689         Unreviewed build correction after r89426.
24690
24691         * WebKitTestRunner/Configurations/WebKitTestRunnerCFLite.vsprops:
24692           Link with Cairo DLL for new functionality.
24693         * WebKitTestRunner/cairo/TestInvocationCairo.cpp: Include compatibility
24694           code for snprintf.
24695
24696 2011-06-25  Dimitri Glazkov  <dglazkov@chromium.org>
24697
24698         Reviewed by Adam Barth.
24699
24700         run-bindings-tests does not work on Chromium Windows.
24701         https://bugs.webkit.org/show_bug.cgi?id=63382
24702
24703         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Removed run-bindings-tests as a build step for cr-win.
24704
24705 2011-06-25  Dan Bernstein  <mitz@apple.com>
24706
24707         Try to fix the Qt and Chromium builds.
24708
24709         * DumpRenderTree/chromium/LayoutTestController.cpp:
24710         (LayoutTestController::setEncodedAudioData):
24711         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
24712
24713 2011-06-24  Dan Bernstein  <mitz@apple.com>
24714
24715         Reviewed by Anders Carlsson.
24716
24717         Add an option to enable legacy rounding hacks
24718         https://bugs.webkit.org/show_bug.cgi?id=63363
24719
24720         Added layoutTestController.allowRoundingHacks() and implemented it
24721         on OS X.
24722
24723         * DumpRenderTree/LayoutTestController.cpp:
24724         (allowRoundingHacksCallback):
24725         (LayoutTestController::staticFunctions):
24726         * DumpRenderTree/LayoutTestController.h:
24727         * DumpRenderTree/chromium/LayoutTestController.cpp:
24728         (LayoutTestController::allowRoundingHacks):
24729         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
24730         (LayoutTestController::allowRoundingHacks):
24731         * DumpRenderTree/mac/DumpRenderTree.mm:
24732         (resetWebViewToConsistentStateBeforeTesting):
24733         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
24734         (LayoutTestController::allowRoundingHacks):
24735         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
24736         (LayoutTestController::allowRoundingHacks):
24737         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
24738         (LayoutTestController::allowRoundingHacks):
24739         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
24740         (LayoutTestController::allowRoundingHacks):
24741
24742 2011-06-24  Adam Roben  <aroben@apple.com>
24743
24744         Include the directory containing the failing tests in titles of bugs filed from TestFailures
24745         when all the tests won't fit
24746
24747         Fixes <http://webkit.org/b/63350> Short bug titles from TestFailures page give no indication
24748         which tests are failing
24749
24750         Reviewed by David Kilzer.
24751
24752         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
24753         (longestCommonPathPrefix): Added. Given a set of paths, returns the longest common prefix
24754         that ends in a path separator.
24755
24756         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
24757         (ViewController.prototype._domForNewAndExistingBugs): When we can't fit all the test names
24758         in the title, first try to include the longest common prefix of the test names, then, if
24759         that's still too long or there's no common prefix, fall back to not including any
24760         information about which tests are failing.
24761
24762 2011-06-24  Adam Roben  <aroben@apple.com>
24763
24764         Include links to Trac in bugs filed from TestFailures
24765
24766         Fixes <http://webkit.org/b/63348> Bugs filed from TestFailures page should include links to
24767         Trac for suspect revisions
24768
24769         Reviewed by David Kilzer.
24770
24771         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
24772         (Trac.prototype.changesetURL): New function, returns the URL for a particular revision in
24773         Trac.
24774
24775         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
24776         (ViewController.prototype._domForNewAndExistingBugs): If there's only one suspect revision,
24777         include a link to Trac for it in the first sentence of the description. Otherwise include a
24778         link to the suspect range after the list of failing tests.
24779
24780 2011-06-24  Nat Duca  <nduca@chromium.org>
24781
24782         Unreviewed, rolling out r89694.
24783         http://trac.webkit.org/changeset/89694
24784         https://bugs.webkit.org/show_bug.cgi?id=58408
24785
24786         Test shell still not ready for animate changes.
24787
24788         * DumpRenderTree/chromium/WebViewHost.cpp:
24789         (WebViewHost::paintInvalidatedRegion):
24790
24791 2011-06-24  Adam Roben  <aroben@apple.com>
24792
24793         Fix typos introduced in r89696
24794
24795         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
24796         (Builder.prototype.getNumberOfFailingTests): Pass the result data's constituent parts to the
24797         callback, since that's what it expects.
24798
24799 2011-06-24  Adam Roben  <aroben@apple.com>
24800
24801         Make TestFailures correctly remember whether old-run-webkit-tests exited early
24802
24803         Fixes <http://webkit.org/b/63342> TestFailures page incorrectly claims test run ran to
24804         completion after reload
24805
24806         Reviewed by David Kilzer.
24807
24808         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
24809         (Builder.prototype.getNumberOfFailingTests): Instead of just storing the number of failures
24810         in the PeristentCache, store an object that contains both the number of failures and whether
24811         old-run-webkit-tests exited early.
24812
24813 2011-06-24  Adam Roben  <aroben@apple.com>
24814
24815         Add links to regression ranges in Trac to the TestFailures page
24816
24817         Fixes <http://webkit.org/b/61060> <rdar://problem/9452153> TestFailures page should provide
24818         links to regression ranges in Trac
24819
24820         Reviewed by David Kilzer.
24821
24822         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js: Added.
24823         (Trac): This new class represents a particular instance of Trac for a single project.
24824         (Trac.prototype.logURL): Returns the URL that can be used to see the log of the given
24825         repository path between the two specified revisions
24826         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
24827         (ViewController): Added new trac argument.
24828         (ViewController.prototype._displayBuilder): Moved code to create the DOM for the
24829         passed/failed builds from here...
24830         (ViewController.prototype._domForRegressionRange): ...to here. Now also includes a link to
24831         Trac if there are multiple suspect revisions.
24832
24833         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
24834         Trac.js and pass a Trac instance for trac.webkit.org to the ViewController.
24835
24836 2011-05-17  Nat Duca  <nduca@chromium.org>
24837
24838         Reviewed by James Robinson.
24839
24840         [chromium] Implement CCLayerTreeHost and CCLayerTreeHostImpl portions of threaded compositor
24841         https://bugs.webkit.org/show_bug.cgi?id=58408
24842
24843         * DumpRenderTree/chromium/WebViewHost.cpp:
24844         (WebViewHost::paintInvalidatedRegion):
24845
24846 2011-06-24  Dominic Cooney   <dominicc@chromium.org>
24847
24848         Reviewed by Dimitri Glazkov.
24849
24850         Convert shadow DOM-related tests to use window.internals
24851         https://bugs.webkit.org/show_bug.cgi?id=61671
24852
24853         Remove ensureShadowRoot, shadowRoot, shadowPseudoId and
24854         removeShadowRoot from layoutTestController; these have been
24855         replaced by equivalents in window.internals instead.
24856
24857         * DumpRenderTree/LayoutTestController.cpp:
24858         (LayoutTestController::staticFunctions):
24859         * DumpRenderTree/LayoutTestController.h:
24860         * DumpRenderTree/chromium/LayoutTestController.cpp:
24861         (LayoutTestController::LayoutTestController):
24862         * DumpRenderTree/chromium/LayoutTestController.h:
24863         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
24864         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
24865         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
24866         * DumpRenderTree/qt/LayoutTestControllerQt.h:
24867         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
24868         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
24869         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
24870         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
24871         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
24872
24873 2011-06-24  Csaba Osztrogonác  <ossy@webkit.org>
24874
24875         Reviewed by Dirk Schulze.
24876
24877         Make run-bindings-tests remove its temporary directories.
24878         https://bugs.webkit.org/show_bug.cgi?id=63320
24879
24880         * Scripts/run-bindings-tests:
24881
24882 2011-06-23  MORITA Hajime  <morrita@google.com>
24883
24884         Reviewed by Kent Tamura.
24885
24886         [Chromium] Invalid write inside WebKit::FrameLoaderClientImpl::dispatchDidClearWindowObjectInWorld
24887         https://bugs.webkit.org/show_bug.cgi?id=63216
24888
24889         Deleted m_webViewHost which looks leaked.
24890
24891         * DumpRenderTree/chromium/TestShell.h:
24892         * DumpRenderTree/chromium/TestShell.cpp:
24893         (TestShell::~TestShell):
24894
24895 2011-06-24  Carlos Garcia Campos  <cgarcia@igalia.com>
24896
24897         Unreviewed. Fix WebKit2 GTK build.
24898
24899         * WebKitTestRunner/GNUmakefile.am:
24900
24901 2011-06-23  Ryuan Choi  <ryuan.choi@samsung.com>
24902
24903         Unreviewed, add myself as committer.
24904
24905         * Scripts/webkitpy/common/config/committers.py:
24906
24907 2011-06-23  Nat Duca  <nduca@chromium.org>
24908
24909         Unreviewed, add myself as committer.
24910
24911         * Scripts/webkitpy/common/config/committers.py:
24912
24913 2011-06-23  Dirk Pranke  <dpranke@chromium.org>
24914
24915         Reviewed by Tony Chang.
24916
24917         nrwt: move http locking code into manager
24918         https://bugs.webkit.org/show_bug.cgi?id=63103
24919
24920         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
24921         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
24922
24923 2011-06-23  Alok Priyadarshi  <alokp@chromium.org>
24924
24925         Reviewed by James Robinson.
24926
24927         [chromium] Add support for running layout tests with accelerated rendering
24928         https://bugs.webkit.org/show_bug.cgi?id=63274
24929
24930         * Scripts/webkitpy/layout_tests/port/chromium.py:
24931         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
24932
24933 2011-06-23  Adam Roben  <aroben@apple.com>
24934
24935         Avoid fetching JSON data when possible on TestFailures page to determine if
24936         old-run-webkit-tests exited early
24937
24938         As of r89610, results.html includes information about whether old-run-webkit-tests exited
24939         early, so we don't need to look at JSON data to get this information for builds newer than
24940         that.
24941
24942         Fixes <http://webkit.org/b/63281> TestFailures page fetches build JSON for every build,
24943         which is slow and limits history to the last build master restart
24944
24945         Reviewed by David Kilzer.
24946
24947         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
24948         (Builder.prototype._getFailingTests): Moved code dealing with results.html into a new
24949         function. If we're dealing with a build that has early-exit information in results.html,
24950         don't bother fetching JSON data and go straight to parsing results.html. Otherwise first use
24951         the JSON data to determine if old-run-webkit-tests exited early, and then parse results.html
24952         (as before).
24953
24954 2011-06-23  Eric Seidel  <eric@webkit.org>
24955
24956         Reviewed by Adam Barth.
24957
24958         sheriffbot can no longer restart
24959         https://bugs.webkit.org/show_bug.cgi?id=63221
24960
24961         Fixed and unit tested.
24962
24963         * Scripts/webkitpy/tool/bot/queueengine.py:
24964         * Scripts/webkitpy/tool/bot/sheriffircbot.py:
24965         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
24966
24967 2011-06-23  Martin Robinson  <mrobinson@igalia.com>
24968
24969         Reviewed by Xan Lopez.
24970
24971         [GTK] [WebKit2] WebKitTestRunner does not properly load TestNetscapePlugin
24972         https://bugs.webkit.org/show_bug.cgi?id=63287
24973
24974         Send the path to the TestNetscapePlugin to the WKTR via an environment variable
24975         so that the plugin loads properly.
24976
24977         * Scripts/old-run-webkit-tests: Pass the appropriate environment variable to WKTR.
24978         * WebKitTestRunner/gtk/TestControllerGtk.cpp:
24979         (WTR::getEnvironmentVariableAsUTF8String): Added this helper which abstracts away some
24980         of the work for getting an environment variable and converting it to UTF-8.
24981         (WTR::TestController::initializeInjectedBundlePath): Use the helper.
24982         (WTR::TestController::initializeTestPluginDirectory): Get the path from an environment variable now.
24983
24984 2011-06-23  Adam Roben  <aroben@apple.com>
24985
24986         Record whether old-run-webkit-tests aborted early in results.html
24987
24988         Fixes <http://webkit.org/b/63280> results.html gives no indication of whether
24989         old-run-webkit-tests stopped running tests early
24990
24991         Reviewed by David Kilzer.
24992
24993         * Scripts/old-run-webkit-tests:
24994         (top level): Switched to heredoc syntax for printing the start of results.html for clarity.
24995         Added a <style> element with some styles for the new message. If we stopped running early,
24996         put the reason why at the top of results.html.
24997         (stopRunningTestsEarlyIfNeeded): Store the message we print in the new
24998         $stoppedRunningEarlyMessage global.
24999
25000 2011-06-23  Adam Roben  <aroben@apple.com>
25001
25002         Cache some of the TestFailures page's data in localStorage
25003
25004         This makes reloading TestFailures much faster. Right now we only store the number of failing
25005         tests and the list of failing tests for each build. We may choose to store more later, but
25006         it's easy to run up against quota limits.
25007
25008         Fixes <http://webkit.org/b/61520> TestFailures page should take advantage of LocalStorage
25009         APIs (or similar) to improve loading performance
25010
25011         Reviewed by David Kilzer.
25012
25013         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
25014         (Builder.prototype.getNumberOfFailingTests): Changed to use the new PersistentCache object.
25015         (Builder.prototype.startFetchingBuildHistory): Changed to pass whether or not we're still
25016         fetching data to the callback.
25017         (Builder.prototype._getFailingTests): Changed to use the new PersistentCache object. We now
25018         store the tests in the cache just before calling the callback. (The previous code in this
25019         function relied on being able to modify the tests object after storing it in the cache and
25020         having the cached version be updated. This worked while it was a non-serialized cache, but
25021         PersistentCache uses serialization.)
25022
25023         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/PersistentCache.js: Added.
25024         (PersistentCache): This object wraps localStorage. It uses JSON to serialize/deserialize
25025         values, and stores the date that each value was initially stored along with it. This is
25026         later used for pruning the cache.
25027         (PersistentCache.contains): Checks whether the key exists in localStorage.
25028         (PersistentCache.get): Fetch the string we stored in localStorage and extract the original
25029         value out of it.
25030         (PersistentCache.set): Serialize the value, add the date to it, and store it in
25031         localStorage. If this fails due to quota limits, empty the whole cache and try again.
25032         (PersistentCache.prune): Delete any cached data that is deemed old enough.
25033         (PersistentCache._addDateToJSONString): Prepend the current date to the string.
25034         (PersistentCache._emptyCache): Delete everything from localStorage.
25035         (PersistentCache._parseDateAndJSONFromString): Split apart the date and the JSON string and
25036         return them.
25037
25038         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
25039         (ViewController.prototype._displayBuilder): Updated for change to callback signature. When
25040         we finish fetching data, prune the PersistentCache. While I was here I also fixed a bug
25041         where we'd never show the new bug link for tests for which we couldn't determine a passing
25042         revision.
25043
25044         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pull in
25045         PersistentCache.js.
25046
25047 2011-06-23  Adam Roben  <aroben@apple.com>
25048
25049         Make finding existing bugs and filing new bugs work on TestFailures even when lots of tests are failing
25050
25051         One bug this fixes is <http://webkit.org/b/61660> New bug links on TestFailures page often
25052         contain titles that are so long they are rejected by Bugzilla
25053
25054         Reviewed by David Kilzer.
25055
25056         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js:
25057         (Bugzilla.prototype.quickSearch): Use fetchResource to POST the search query rather than
25058         using a query string on the URL. If the search query is very long, using a query string can
25059         cause the request to be rejected due to the URL being too long. POSTing avoids this issue.
25060         (Bugzilla.maximumBugTitleLength): Added this constant based on Bugzilla's implementation.
25061
25062         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
25063         (.new-bug-form): Hide the form that we secretly use to file a new bug.
25064
25065         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
25066         (fetchResource): Added. Code came from getResource. If we're using a GET request, add the
25067         query parameters to the URL. Otherwise, send them as the body of the request along with the
25068         appropriate headers.
25069         (getResource): Now just calls through to fetchResource.
25070         (urlEncodedQueryParameters): Added. Moved some code here...
25071         (addQueryParametersToURL): ...from here.
25072
25073         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
25074         (ViewController.prototype._domForNewAndExistingBugs): Shorten the title to just mention the
25075         number of failing tests if mentioning all the tests would make the title too long. Improved
25076         the description for large numbers of failures by listing the tests one-per-line instead of
25077         just having them be comma-separated, which was hard to read. Use a form to file the new bug
25078         instead of a link so that we can POST the form data. (Using a URL with a query string can
25079         result in the URL being too long.) The new bug link now just submits the form.
25080
25081 2011-06-23  Adam Roben  <aroben@apple.com>
25082
25083         Show closed bugs on the TestFailures page in addition to open ones
25084
25085         Fixes <http://webkit.org/b/63194> TestFailures page should show closed bugs, too
25086
25087         Reviewed by David Kilzer.
25088
25089         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js:
25090         (Bugzilla.prototype.quickSearch): Added code to extract the bug's status and store it in the
25091         returned data.
25092         (Bugzilla.isOpenStatus): New function, returns true if the passed-in status indicates that
25093         the associated bug is still open.
25094
25095         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
25096         (addQueryParametersToURL): Add a missing semicolon.
25097         (Node.prototype.appendChildren): New function, like appendChild but takes an array-like
25098         object and appends each of the values stored within.
25099
25100         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
25101         (ViewController.prototype._domForNewAndExistingBugs): Prepend 'ALL' to the query so closed
25102         bugs will be included in the results. Split the returned bugs into two sets: those which are
25103         open, and those which are closed. Put the open bugs at the top level of the list, and the
25104         closed bugs in a second level.
25105
25106 2011-06-23  Adam Roben  <aroben@apple.com>
25107
25108         Don't count new tests as failures on the TestFailures page
25109
25110         Fixes <http://webkit.org/b/63254> TestFailures page calls new tests "failures", even though
25111         they aren't
25112
25113         Reviewed by Anders Carlsson.
25114
25115         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
25116         (Builder.prototype.getNumberOfFailingTests): Don't add the number of new tests to the number
25117         of failures.
25118
25119 2011-06-23  Dirk Pranke  <dpranke@chromium.org>
25120
25121         Reviewed by Eric Seidel.
25122
25123         nrwt: make TestInput objects printable
25124         https://bugs.webkit.org/show_bug.cgi?id=63225
25125
25126         This makes debugging slighlty easier.
25127
25128         * Scripts/webkitpy/layout_tests/layout_package/test_input.py:
25129
25130 2011-06-22  Daniel Bates  <dbates@webkit.org>
25131
25132         Reviewed by Adam Barth.
25133
25134         test-webkitpy --all errors out because scm_unittest.py can't find module checkout
25135         https://bugs.webkit.org/show_bug.cgi?id=62943
25136
25137         Remove "from .checkout import Checkout" from scm_unittest.py since it's included
25138         by Scripts/webkitpy/common/checkout/__init__.py.
25139
25140         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
25141
25142 2011-06-22  Dominic Cooney  <dominicc@chromium.org>
25143
25144         Reviewed by Mark Rowe.
25145
25146         Add window.internals to WebKit2's WebKitTestRunner.
25147         https://bugs.webkit.org/show_bug.cgi?id=61073
25148
25149         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pxbproj:
25150         * WebKitTestRunner/Configurations/Base.xcconfig:
25151         * WebKitTestRunner/Configurations/InjectedBundleCFLite.vsprops:
25152         * WebKitTestRunner/Configurations/InjectedBundleCommon.vsprops:
25153         * WebKitTestRunner/Configurations/InjectedBundleCoreFoundation.vsprops:
25154         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
25155         (WTR::InjectedBundlePage::didClearWindowForFrame):
25156         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
25157         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
25158
25159 2011-06-22  Eric Seidel  <eric@webkit.org>
25160
25161         Reviewed by Adam Barth.
25162
25163         Make sheriff-bot rollout messages a little nicer
25164         https://bugs.webkit.org/show_bug.cgi?id=63107
25165
25166         itertools.chain.from_iterable is new in 2.6,
25167         use itertools.chain(*list) for 2.5 compat.
25168
25169         * Scripts/webkitpy/tool/bot/irc_command.py:
25170
25171 2011-06-22  Eric Seidel  <eric@webkit.org>
25172
25173         Reviewed by Adam Barth.
25174
25175         Make sheriff-bot robust against exceptions from commands
25176         https://bugs.webkit.org/show_bug.cgi?id=63211
25177
25178         sheriff-bot was acting strange this afternoon.
25179         We don't know if this will fix the cause, but
25180         at least it adds some unit tests and catches
25181         one possible cause.
25182
25183         * Scripts/webkitpy/tool/bot/sheriffircbot.py:
25184         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
25185
25186 2011-06-22  Mark Rowe  <mrowe@apple.com>
25187
25188         Reviewed by David Levin.
25189
25190         <http://webkit.org/b/63212> TestWebKitAPI Xcode project has bogus settings since r86287
25191
25192         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Fix the reference to libgtest.a to
25193         be relative to the built products directory rather than using a hard-coded path. Remove the
25194         explicit settings of FRAMEWORK_SEARCH_PATHS, HEADER_SEARCH_PATHS and LIBRARY_SEARCH_PATHS
25195         as they're all unnecessary. If they were necessary they should be set in the .xcconfig file
25196         rather than in the .xcodeproj directly.
25197
25198 2011-06-22  Dirk Pranke  <dpranke@chromium.org>
25199
25200         Reviewed by Ojan Vafai.
25201
25202         new-run-webkit-tests: remove obsolete port.shut_down_http_server method
25203         https://bugs.webkit.org/show_bug.cgi?id=59993
25204
25205         * Scripts/webkitpy/layout_tests/port/base.py:
25206         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
25207         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
25208         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
25209         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
25210         * Scripts/webkitpy/layout_tests/port/gtk.py:
25211         * Scripts/webkitpy/layout_tests/port/mac.py:
25212         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
25213         * Scripts/webkitpy/layout_tests/port/qt.py:
25214         * Scripts/webkitpy/layout_tests/port/win.py:
25215
25216 2011-06-22  Luke Macpherson   <macpherson@chromium.org>
25217
25218         Reviewed by Andreas Kling.
25219
25220         Move macpherson from contributor list to committer list.
25221         https://bugs.webkit.org/show_bug.cgi?id=63179
25222
25223         * Scripts/webkitpy/common/config/committers.py:
25224         Move macpherson from contributor list to committer list.
25225
25226 2011-06-22  David Levin  <levin@chromium.org>
25227
25228         Reviewed by Adam Barth.
25229
25230         check-webkit-style should detect returning (Own|Ref)Ptr instead of the Pass*Ptr version.
25231         https://bugs.webkit.org/show_bug.cgi?id=63204
25232
25233         * Scripts/webkitpy/style/checkers/cpp.py: Added a check for the return value and combined
25234           with similar code for the parameter checking.
25235         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Removed pass_ptr checks from
25236           those done for single lines since they don't make sense in that case (variable decls look like function decls).
25237           Removed some redundant comments (one of which was slightly wrong).
25238           Added checks for the new functionality and minor other test changes.
25239
25240 2011-06-22  Nate Chapin  <japhet@chromium.org>
25241
25242         Reviewed by Adam Barth.
25243
25244         Add libsoup to libWebCoreInternals build.
25245         https://bugs.webkit.org/show_bug.cgi?id=62066
25246
25247         * GNUmakefile.am:
25248
25249 2011-06-22  Dirk Pranke  <dpranke@chromium.org>
25250
25251         Reviewed by Tony Chang.
25252
25253         nrwt: handle missing httpd cleanly
25254         https://bugs.webkit.org/show_bug.cgi?id=62027
25255
25256         We had reverted the change in r89414, so this adds it back in
25257         (modifying check_sys_deps() in port/base.py to check for an
25258         installed web server).
25259
25260         This change then fixes the change in r89414 to stub out the
25261         check_sys_deps() in the test port so that the unit tests run correctly.
25262
25263         * Scripts/webkitpy/layout_tests/port/base.py:
25264         * Scripts/webkitpy/layout_tests/port/test.py:
25265
25266 2011-06-22  David Levin  <levin@chromium.org>
25267
25268         Reviewed by Adam Barth.
25269
25270         check-webkit-style should check for invalid uses of RefPtr/OwnPtr as parameters.
25271         https://bugs.webkit.org/show_bug.cgi?id=63188
25272
25273         * Scripts/webkitpy/style/checkers/cpp.py: Added the check.
25274         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests.
25275
25276 2011-06-22  Tony Chang  <tony@chromium.org>
25277
25278         Reviewed by Ojan Vafai.
25279
25280         Add a buildbot for testing ENABLE(CSS3_FLEXBOX)
25281         https://bugs.webkit.org/show_bug.cgi?id=62891
25282
25283         * BuildSlaveSupport/build.webkit.org-config/config.json:
25284
25285 2011-06-22  Dirk Pranke  <dpranke@chromium.org>
25286
25287         Reviewed by Ojan Vafai.
25288
25289         nrwt: don't look for http lock when running the test port
25290         https://bugs.webkit.org/show_bug.cgi?id=63158
25291
25292         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
25293         * Scripts/webkitpy/layout_tests/port/test.py:
25294
25295 2011-06-22  Balazs Kelemen  <kbalazs@webkit.org>
25296
25297         Reviewed by Andreas Kling.
25298
25299         [Qt][WK2] Set up plugin tests
25300         https://bugs.webkit.org/show_bug.cgi?id=63066
25301
25302         * WebKitTestRunner/qt/TestControllerQt.cpp:
25303         (WTR::TestController::initializeTestPluginDirectory):
25304         Use the QTWEBKIT_PLUGIN_PATH environment variable as the path
25305         of the test plugin like DRT does.
25306
25307 2011-06-22  Eric Seidel  <eric@webkit.org>
25308
25309         Reviewed by Ojan Vafai.
25310
25311         Make sheriff-bot rollout messages a little nicer
25312         https://bugs.webkit.org/show_bug.cgi?id=63107
25313
25314         It annoyed me this afternoon that I had to convert sheriff-bots "r12345" revisions
25315         into urls myself.  So I have now fixed its "preparing" message to include a url.
25316
25317         I also figured that I should make the messages mention all of the responsible parties
25318         so that rollouts are never surprises.  If you're in the channel and were involved
25319         in a patch, you will see if someone is using sheriff-bot to rollout a patch.
25320
25321         As part of doing this I also changed (and tested) _parse_args to fail-fast
25322         when given invalid args.
25323
25324         * Scripts/webkitpy/tool/bot/irc_command.py:
25325         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
25326         * Scripts/webkitpy/tool/bot/sheriff.py:
25327         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
25328
25329 2011-06-22  Dirk Pranke  <dpranke@chromium.org>
25330
25331         Reviewed by Tony Chang.
25332
25333         run-webkit-websocketserver fails to stop websocket server
25334         https://bugs.webkit.org/show_bug.cgi?id=63123
25335
25336         There's a ten-character change in new-run-webkit-websocketserver
25337         that fixes the actual bug, and then we update the code in
25338         new-run-webkit-httpd to do the actual stop call, and add more tests
25339         (refactoring the existing test code so that it is shared).
25340
25341         * Scripts/new-run-webkit-httpd:
25342         * Scripts/new-run-webkit-websocketserver:
25343         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py:
25344
25345 2011-06-22  Adam Roben  <aroben@apple.com>
25346
25347         Add links to existing bugs related to failing tests on TestFailures page
25348
25349         I changed the layout of the page a little to make it easier to read with all the new
25350         information. Passing/failing revisions have been moved down below the list of tests to be
25351         closer to the existing bugs and the new bug link. And each set of tests and its relevant
25352         information is in a light gray box.
25353
25354         Fixes <http://webkit.org/b/61665> TestFailures page should link to existing bugs when
25355         possible
25356
25357         Reviewed by Darin Adler.
25358
25359         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Bugzilla.js: Added.
25360         (Bugzilla): This new class represents a single Bugzilla instance.
25361         (Bugzilla.prototype.quickSearch): Searches Bugzilla using its Quick Search functionality,
25362         passing the resulting bug titles and URLs to the callback when complete. If called multiple
25363         times with the same query before the query returns, caches the callbacks so that only one
25364         query is sent over the wire. When the query completes, all pending callbacks are called.
25365
25366         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
25367         (#failure-history): Reduce the margin/padding on the top-level list a bit.
25368         (#failure-history > li): Put each set of tests in a gray box, and indent most information
25369         inside the box.
25370         (.test-list): Unindent the list of failing tests so it is visually at the top level.
25371         (.new-and-existing-bugs): Reduce the space at the bottom of this area so that the bottom of
25372         each box isn't a big empty space.
25373         (.existing-bugs-list): Use a smaller text size for existing bugs, since their titles can be
25374         quite long.
25375
25376         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
25377         (addQueryParametersToURL): New function extracted from
25378         ViewController.prototype._domForNewAndExistingBugs.
25379
25380         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
25381         (ViewController): Take and store an optional Bugzilla instance.
25382         (ViewController.prototype._displayBuilder): Give the top-level list an id for styling
25383         purposes and move the list of failing tests above all other information. Only show bug
25384         information once we've determined the most-recent passing revision for a set of tests. It's
25385         not that useful to file a new bug before this information has been determined, and searching
25386         for existing bugs before we've figured out which tests started failing at the same time
25387         would end up giving you information about a bunch of unrelated tests.
25388         (ViewController.prototype._domForNewAndExistingBugs): Renamed from _domForNewBugLink. Now
25389         returns a DocumentFragment instead of an HTMLParagraphElement. If we don't have a Bugzilla
25390         instance, just returns an empty DocumentFragment. Starts a search for bugs related to the
25391         failing tests, and adds links to the bugs when the search completes.
25392
25393         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Pass a
25394         Bugzilla instance for bugs.webkit.org to the ViewController.
25395
25396 2011-06-22  Carlos Garcia Campos  <cgarcia@igalia.com>
25397
25398         Reviewed by Martin Robinson.
25399
25400         [GTK] Apply current view settings to newly created views in GtkLauncher
25401         https://bugs.webkit.org/show_bug.cgi?id=63142
25402
25403         * GtkLauncher/main.c:
25404         (createWebViewCb):
25405
25406 2011-06-22  Carlos Garcia Campos  <cgarcia@igalia.com>
25407
25408         Reviewed by Martin Robinson.
25409
25410         [GTK] Use GOption to parse main arguments in GtkLauncher
25411         https://bugs.webkit.org/show_bug.cgi?id=63060
25412
25413         Instead of getting the settings from the view, create a new
25414         settings object not associated with any view so that we can parse
25415         the command line arguments before creating the view. It will also
25416         allows to use the same settings for more than one view if we
25417         eventually support passing more than one uri from the command
25418         line.
25419
25420         * GtkLauncher/main.c:
25421         (parseOptionEntryCallback): Make this function static. Also use
25422         g_ascii_strtoll/g_ascii_strtod to parse int/float values instead
25423         of atoi/atof to provide better error messages. And always fill the
25424         error struct when the function fails.
25425         (getOptionEntriesFromWebKitWebSettings): Use an array of
25426         GOptionEntry structs instead of a GArray, since that is what
25427         g_option_group_add_entries() expects and the array size doesn't
25428         change.
25429         (addWebSettingsGroupToContext): Helper function to add the
25430         websettings group to the option context.
25431         (main): Create the option context to parse command line arguments.
25432
25433 2011-06-22  Martin Robinson  <mrobinson@igalia.com>
25434
25435         Reviewed by Adam Roben.
25436
25437         [GTK] Implement pixel dump support for WebKitTestRunner
25438         https://bugs.webkit.org/show_bug.cgi?id=58242
25439
25440         * DumpRenderTree/gtk/GNUmakefile.ImageDiff.am: Added.
25441         * GNUmakefile.am: Separate out the ImageDiff bits of the GNUmakefile.
25442         * Scripts/old-run-webkit-tests: Update to add support for the GTK+ port.
25443         * WebKitTestRunner/GNUmakefile.am: Added Cairo specific files to the build.
25444         * WebKitTestRunner/TestInvocation.cpp: Moved this code to the platform-independent
25445         file so that we can share it across platforms.
25446         (WTR::TestInvocation::compareActualHashToExpectedAndDumpResults):
25447         * WebKitTestRunner/TestInvocation.h: Updated method list.
25448         * WebKitTestRunner/cairo/TestInvocationCairo.cpp: Added Cairo implementation.
25449         * WebKitTestRunner/cg/TestInvocationCG.cpp:
25450         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected): Updated to use new helper.
25451
25452 2011-06-22  Sheriff Bot  <webkit.review.bot@gmail.com>
25453
25454         Unreviewed, rolling out r89414.
25455         http://trac.webkit.org/changeset/89414
25456         https://bugs.webkit.org/show_bug.cgi?id=63130
25457
25458         It broke webkitpy tests (Requested by Ossy on #webkit).
25459
25460         * Scripts/webkitpy/layout_tests/port/base.py:
25461
25462 2011-06-22  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
25463
25464         Reviewed by Dirk Pranke.
25465
25466         [NRWT] Print error message when there is no httpd present on the system
25467         https://bugs.webkit.org/show_bug.cgi?id=62027
25468
25469         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
25470
25471 2011-06-21  Dirk Pranke  <dpranke@chromium.org>
25472
25473         Unreviewed, build fix.
25474         Fix crashes in new-run-webkit-tests resulting from the
25475         change to the http_server logic in r89400. Python 2.5
25476         on Mac 10.5 has some weird error paths.
25477
25478         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
25479         * Scripts/webkitpy/common/system/executive.py:
25480
25481 2011-06-21  MORITA Hajime  <morrita@google.com>
25482
25483         Unreviewed, rolling out r89401 and r89403.
25484         http://trac.webkit.org/changeset/89401
25485         http://trac.webkit.org/changeset/89403
25486         https://bugs.webkit.org/show_bug.cgi?id=62970
25487
25488         Breaks mac build and mistakenly enables the spellcheck API
25489
25490         * Scripts/build-webkit:
25491         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
25492
25493 2011-06-21  Kent Tamura  <tkent@chromium.org>
25494
25495         [Mac] Sort Xcode project files.
25496
25497         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
25498
25499 2011-06-20  MORITA Hajime  <morrita@google.com>
25500
25501         Reviewed by Kent Tamura.
25502
25503         Spellcheck API should be build-able.
25504         https://bugs.webkit.org/show_bug.cgi?id=62970
25505
25506         * Scripts/build-webkit:
25507
25508 2011-06-21  Dirk Pranke  <dpranke@chromium.org>
25509
25510         Reviewed by Tony Chang.
25511
25512         nrwt: attempt #3 at fixing server startup/shutdown
25513         https://bugs.webkit.org/show_bug.cgi?id=62829
25514
25515         Attempt yet again to make starting and stopping servers
25516         reliable. It turns out that apache has races between when the
25517         ctl process exists and when it writes/removes the pid file.
25518         This change accounts for the races.
25519
25520         * Scripts/new-run-webkit-httpd:
25521         * Scripts/webkitpy/common/system/executive.py:
25522         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
25523         * Scripts/webkitpy/layout_tests/port/base.py:
25524         * Scripts/webkitpy/layout_tests/port/http_server.py:
25525         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
25526         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Added.
25527         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
25528         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
25529
25530 2011-06-21  Dmitry Lomov  <dslomov@google.com>
25531
25532         Reviewed by Adam Roben.
25533
25534         https://bugs.webkit.org/show_bug.cgi?id=63080
25535         Unit-tests step on test-only bot is broken.
25536         Removing the step until run-api-tests is fixed.
25537
25538         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
25539
25540 2011-06-20  Dmitry Lomov  <dslomov@google.com>
25541
25542         Reviewed by David Levin.
25543
25544         https://bugs.webkit.org/show_bug.cgi?id=63017
25545         "Unit tests" step on bots should provide more information in case of failure.
25546
25547         * BuildSlaveSupport/build.webkit.org-config/master.cfg: added --verbose flag to UnitTests step.
25548
25549 2011-06-20  Dirk Pranke  <dpranke@chromium.org>
25550
25551         Reviewed by Ojan Vafai.
25552
25553         new-run-webkit-tests: order shards by name, not number of tests
25554         https://bugs.webkit.org/show_bug.cgi?id=62753
25555
25556         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
25557
25558 2011-06-20  Dirk Pranke  <dpranke@chromium.org>
25559
25560         Reviewed by Ojan Vafai.
25561
25562         we should only log the test names to the tests_run.txt files during layout_tests
25563         https://bugs.webkit.org/show_bug.cgi?id=62751
25564
25565         * Scripts/old-run-webkit-tests:
25566         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
25567
25568 2011-06-20  Eric Seidel  <eric@webkit.org>
25569
25570         Reviewed by Adam Barth.
25571
25572         commit-queue/ews-bots should run run-bindings-tests for every patch
25573         https://bugs.webkit.org/show_bug.cgi?id=63010
25574
25575         * Scripts/webkitpy/common/config/ports.py:
25576         * Scripts/webkitpy/tool/commands/download_unittest.py:
25577         * Scripts/webkitpy/tool/steps/runtests.py:
25578         * Scripts/webkitpy/tool/steps/runtests_unittest.py:
25579         * Scripts/webkitpy/tool/steps/steps_unittest.py:
25580
25581 2011-06-20  Chang Shu  <cshu@webkit.org>
25582
25583         Reviewed by Adam Barth.
25584
25585         fast/js/sputnik/Conformance/10_Execution_Contexts/10.2_Entering_An_Execution_Context/10.2.2_Eval_Code tests fail when showModalDialog isn't supported
25586         https://bugs.webkit.org/show_bug.cgi?id=53676
25587
25588         * WebKitTestRunner/TestController.cpp:
25589         (WTR::TestController::initialize):
25590
25591 2011-06-20  Eric Seidel  <eric@webkit.org>
25592
25593         Reviewed by Adam Barth.
25594
25595         buildbots should run run-bindings-tests
25596         https://bugs.webkit.org/show_bug.cgi?id=63001
25597
25598         run-bindings-tests takes 2.4 seconds on my machine, so
25599         I expect this will not be noticeable on the bots.
25600
25601         I also removed tiger-specific checks since Tiger is no
25602         longer a supported platform for webkit.
25603
25604         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
25605
25606 2011-06-20  Carlos Garcia Campos  <cgarcia@igalia.com>
25607
25608         Unreviewed. Fix WebKit2 GTK build after r89249.
25609
25610         * GtkLauncher/main.c:
25611         (main):
25612
25613 2011-06-20  Pavel Feldman  <pfeldman@chromium.org>
25614
25615         Reviewed by Yury Semikhatsky.
25616
25617         Web Inspector: remove LayoutTestController::setTimelineProfilingEnabled.
25618         https://bugs.webkit.org/show_bug.cgi?id=62994
25619
25620         * DumpRenderTree/LayoutTestController.cpp:
25621         (closeWebInspectorCallback):
25622         (LayoutTestController::staticFunctions):
25623         * DumpRenderTree/LayoutTestController.h:
25624         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
25625         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
25626         * DumpRenderTree/chromium/LayoutTestController.cpp:
25627         (LayoutTestController::LayoutTestController):
25628         * DumpRenderTree/chromium/LayoutTestController.h:
25629         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
25630         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
25631         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
25632         * DumpRenderTree/qt/LayoutTestControllerQt.h:
25633         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
25634         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
25635         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
25636         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
25637         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
25638
25639 2011-06-20  Lukasz Slachciak  <l.slachciak@samsung.com>
25640
25641         Reviewed by Martin Robinson.
25642
25643         [GTK] General mechanism for adjusting WebKitWebSettings in GtkLauncher.
25644         https://bugs.webkit.org/show_bug.cgi?id=55308
25645
25646         * GtkLauncher/main.c: Added general mechanism for adjusting WebKitWebSettings.
25647         (parseOptionEntryCallback): Callback for parsing option entry.
25648         (getOptionEntriesFromWebKitWebSettings): Basing on the WebKitWebSettings properties prepare
25649         list of option entries.
25650         (transformStringToBoolean): Transform function for boolean convertion used by parseOptionEntryCallback.
25651         (transformStringToInt): Transform function for int convertion used by parseOptionEntryCallback.
25652         (transformStringToFloat): Transform function for float convertion used by parseOptionEntryCallback.
25653         (parseAdditionalOptions): Help function parsing additional commandline options.
25654         (main): Added call to parseAdditionalOptions.
25655
25656 2011-06-20  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
25657
25658         Reviewed by Andreas Kling.
25659
25660         [Qt] [WK2] Compile error in ActivateFontsQt.cpp on Mac
25661         https://bugs.webkit.org/show_bug.cgi?id=62869
25662
25663         Add QFont include to fix "incomlete type QFont" error.
25664
25665         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
25666
25667 2011-06-19  Adam Bergkvist  <adam.bergkvist@ericsson.com>
25668
25669         Reviewed by Martin Robinson.
25670
25671         [GTK] Enable Media Stream feature and make it default on
25672         https://bugs.webkit.org/show_bug.cgi?id=60394
25673
25674         * Scripts/build-webkit:
25675
25676 2011-06-19  Sheriff Bot  <webkit.review.bot@gmail.com>
25677
25678         Unreviewed, rolling out r89198.
25679         http://trac.webkit.org/changeset/89198
25680         https://bugs.webkit.org/show_bug.cgi?id=62953
25681
25682         Chromium breakage on "Linux Tests" (Requested by senorblanco
25683         on #webkit).
25684
25685         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
25686         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
25687         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
25688
25689 2011-06-18  Daniel Bates  <dbates@webkit.org>
25690
25691         Reviewed by Eric Seidel.
25692
25693         Git.push_local_commits_to_server() should use passed username and
25694         password when we don't have cached credentials
25695         https://bugs.webkit.org/show_bug.cgi?id=62941
25696
25697         Fixes a issue where Git.push_local_commits_to_server() would always look
25698         for cached credentials regardless of whether a username and password
25699         were passed as arguments. Instead, if given both a username and password
25700         then we shouldn't look for cached credentials (since we were explicitly given
25701         credentials).
25702
25703         * Scripts/webkitpy/common/checkout/scm/git.py:
25704         * Scripts/webkitpy/common/checkout/scm/scm_unittest.py: Added unit tests.
25705
25706 2011-06-18  Daniel Bates  <dbates@rim.com>
25707
25708         Rubber-stamped by Eric Seidel.
25709
25710         Remove run-javascriptcore-tests --skip-build option; Instead add --build/no-build options
25711         https://bugs.webkit.org/show_bug.cgi?id=62767
25712
25713         Part 3 of 3
25714
25715         Remove the command-line option --skip-build now that the master BuildBot has been restarted
25716         following <http://trac.webkit.org/changeset/89138>.
25717
25718         Instead, specify --no-build to get analogous functionality.
25719
25720         * Scripts/run-javascriptcore-tests:
25721
25722 2011-06-18  Dirk Pranke  <dpranke@chromium.org>
25723
25724         Reviewed by Tony Chang.
25725
25726         new-run-webkit-tests: fix test sorting
25727         https://bugs.webkit.org/show_bug.cgi?id=58691
25728
25729         This changes the sorting to use a "natural" sort order such that
25730         foo2.html sorts before foo13.html, and sorts the tests so
25731         that all the tests in a directory are run before any test in a
25732         subdirectory.
25733
25734         This change also adds a split() method to the filesystem since
25735         that is needed to cluster files into a directory otherwise.
25736
25737         * Scripts/webkitpy/common/system/filesystem.py:
25738         * Scripts/webkitpy/common/system/filesystem_mock.py:
25739         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
25740         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py:
25741
25742 2011-06-18  Evan Martin  <evan@chromium.org>
25743
25744         Reviewed by Darin Adler.
25745
25746         Compiler warning in double->int conversion in TestNetscapePlugIn
25747         https://bugs.webkit.org/show_bug.cgi?id=51554
25748
25749         This code is providing a JS API for removing a property by numeric ID.
25750         We get it from JS as a double, and must convert it to an int before using it.
25751
25752         Test: LayoutTests/plugins/npruntime/remove-property.html
25753
25754         * DumpRenderTree/TestNetscapePlugIn/Tests/NPRuntimeRemoveProperty.cpp:
25755         (NPRuntimeRemoveProperty::TestObject::invoke):
25756
25757 2011-06-18  Patrick Gansterer  <paroga@webkit.org>
25758
25759         Reviewed by Eric Seidel.
25760
25761         Don't exit webkit-patch when running on Win32 python
25762         https://bugs.webkit.org/show_bug.cgi?id=55811
25763
25764         Most of the basic actions work in the meantime,
25765         so let people play with in on Win32 now.
25766
25767         * Scripts/webkit-patch:
25768
25769 2011-06-18  Alice Boxhall  <aboxhall@chromium.org>
25770
25771         Reviewed by Ojan Vafai.
25772
25773         Convert json_results_generator.py to output version 4 JSON.
25774         https://bugs.webkit.org/show_bug.cgi?id=60869
25775
25776         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
25777         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
25778         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
25779
25780 2011-06-18  Berend-Jan Wever  <skylined@chromium.org>
25781
25782         Reviewed by Eric Seidel.
25783
25784         Some names with spaces are incorrectly reported as not containing spaces
25785         while running prepare-ChangeLog
25786         https://bugs.webkit.org/show_bug.cgi?id=55572
25787
25788         * Scripts/VCSUtils.pm: Updated CHANGE_LOG_NAME check for spaces to allow
25789           non-alphanumeric characters in names.
25790
25791 2011-06-18  Michael Saboff  <msaboff@apple.com>
25792
25793         Reviewed by Eric Seidel.
25794
25795         Enhancement: New Script to Format malloc_history output as tree
25796         https://bugs.webkit.org/show_bug.cgi?id=56378
25797
25798         New script to process malloc_history output to tree format.
25799
25800         * Scripts/malloc-tree: Added.
25801
25802 2011-06-17  Dmitry Lomov  <dslomov@google.com>
25803
25804         Reviewed by Daniel Bates.
25805
25806         https://bugs.webkit.org/show_bug.cgi?id=62918
25807         Pass configuration to UnitTests step in master.cfg
25808
25809         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
25810
25811 2011-06-17  Leandro Pereira  <leandro@profusion.mobi>
25812
25813         Reviewed by Eric Seidel.
25814
25815         [EFL] Add Evas-based ImageDiff implementation.
25816         https://bugs.webkit.org/show_bug.cgi?id=61942
25817
25818         Build system changes will be performed as soon as the complete
25819         DumpRenderTree implementation is upstreamed.
25820
25821         * DumpRenderTree/efl/ImageDiff.cpp: Added.
25822         (pixelFromImageData):
25823         (differenceImageFromDifferenceBuffer):
25824         (computeDistanceBetweenPixelComponents):
25825         (calculatePixelDifference):
25826         (calculateDifference):
25827         (getTemporaryFile):
25828         (printImage):
25829         (printImageDifferences):
25830         (resizeEcoreEvasIfNeeded):
25831         (readImageFromStdin):
25832         (parseCommandLineOptions):
25833         (shutdownEfl):
25834         (abortWithErrorMessage):
25835         (initEfl):
25836         (main):
25837
25838 2011-06-17  Tony Chang  <tony@chromium.org>
25839
25840         Reviewed by Dirk Pranke.
25841
25842         [chromium] force rebaseline-chromium-webkit-tests to use the chromium port
25843         https://bugs.webkit.org/show_bug.cgi?id=62907
25844
25845         This fixes a bug where the ImageDiff on the user's machine doesn't
25846         match the logic used on the Chromium buildbots.  This has the downside
25847         of requiring that the Chromium ImageDiff is built, but it turns out
25848         that even with --tolerance 0, the CG ImageDiff will allow small pixel
25849         differences through.
25850
25851         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
25852
25853 2011-06-17  Brent Fulgham  <bfulgham@webkit.org>
25854
25855         [WinCairo] Unreviewed build correction.
25856
25857         * DumpRenderTree/win/DumpRenderTreeCairo.vsprops:
25858         Update WinCairo property sheet to link against the Cairo DLL,
25859         now that we are using a more recent Cairo build.
25860
25861 2011-06-17  Brent Fulgham  <bfulgham@webkit.org>
25862
25863         Reviewed by Brian Weinstein.
25864
25865         Create a simple example of using the COM-based C++ listeners.
25866         A new onclick event is attached (via C++) to the WebKit logo
25867         in the default test pattern displayed on startup.  Clicking on
25868         the logo causes a message box to be displayed.
25869         https://bugs.webkit.org/show_bug.cgi?id=61885
25870
25871         * WinLauncher/DOMDefaultImpl.h: Added.  Stub implementation of
25872           the WebScriptObject and DOMEventListener.
25873         * WinLauncher/WinLauncher.cpp:
25874         (SimpleEventListener::SimpleEventListener): Example implementation
25875           of a simple DOM event listener.
25876         (SimpleEventListener::handleEvent):
25877         (WinLauncherWebHost::didFinishLoadForFrame): Added implementation
25878          to bind a C++ method to the 'onclick' event for the WebKit logo.
25879         (_tWinMain):
25880         * WinLauncher/WinLauncher.h:
25881         * WinLauncher/WinLauncher.vcproj: Add new DOMDefaultImpl.h file.
25882
25883 2011-06-17  Chang Shu  <cshu@webkit.org>
25884
25885         Reviewed by Andreas Kling.
25886
25887         [Qt] [WK2] Support undo/redo in MiniBrowser
25888         https://bugs.webkit.org/show_bug.cgi?id=62809
25889
25890         * MiniBrowser/qt/BrowserWindow.cpp:
25891         (BrowserWindow::BrowserWindow):
25892
25893 2011-06-17  Daniel Bates  <dbates@rim.com>
25894
25895         Rubber-stamped by Eric Seidel.
25896
25897         Remove run-javascriptcore-tests --skip-build option; Instead add --build/no-build options
25898         https://bugs.webkit.org/show_bug.cgi?id=62767
25899
25900         Part 2 of 3
25901
25902         Update the BuildBot master.cfg to use the new --no-build run-javascriptcore-tests
25903         command line option instead of --skip-build.
25904
25905         We'll remove --skip-build from run-javascriptcore-test in a follow up patch once the
25906         master buildbot picks up the change.
25907
25908         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
25909
25910 2011-06-17  Daniel Bates  <dbates@rim.com>
25911
25912         Rubber-stamped by Eric Seidel.
25913
25914         Remove run-javascriptcore-tests --skip-build option; Instead add --build/no-build options
25915         https://bugs.webkit.org/show_bug.cgi?id=62767
25916
25917         Part 1 of 3
25918
25919         Towards renaming the option --skip-build to --no-build. Add --build/--no-build options to
25920         run-javascriptcore-tests. Also, add a remark about these options to the usage message
25921         printed by run-javascriptcore-tests.
25922
25923         We'll remove --skip-build in a follow up patch once the master buildbot picks up the changes.
25924
25925         * Scripts/run-javascriptcore-tests: Additionally, remove unnecessary initialization
25926           of $showHelp.
25927
25928 2011-06-17  Joone Hur  <joone.hur@collabora.co.uk>
25929
25930         Reviewed by Martin Robinson.
25931
25932         [GTK] Replace GdkRectangle by cairo_rectangle_int_t
25933         https://bugs.webkit.org/show_bug.cgi?id=60687
25934
25935         Replace GdkRectangle by cairo_rectangle_int_t.
25936
25937         * DumpRenderTree/gtk/PixelDumpSupportGtk.cpp: Replace GdkRectangle by cairo_rectangle_int_t.
25938         (createBitmapContextFromWebView):
25939         * DumpRenderTree/gtk/TextInputController.cpp: Ditto.
25940         (firstRectForCharacterRangeCallback):
25941
25942 2011-06-16  Dmitry Lomov  <dslomov@google.com>
25943
25944         Reviewed by David Levin.
25945
25946         https://bugs.webkit.org/show_bug.cgi?id=62603
25947         Run run-api-tests on bots in debug mode.
25948         Release mode is disabled due to https://bugs.webkit.org/show_bug.cgi?id=61812.
25949
25950         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
25951
25952 2011-06-16  Leandro Pereira  <leandro@profusion.mobi>
25953
25954         Reviewed by Eric Seidel.
25955
25956         [EFL] DumpRenderTree: Add GCController, PixelDumpSupport and
25957         WorkQueueItem.
25958         They will be plugged into the build system once the EFL's DRT is
25959         fully upstreamed.
25960         https://bugs.webkit.org/show_bug.cgi?id=61958
25961
25962         * DumpRenderTree/efl: Added.
25963         * DumpRenderTree/efl/GCControllerEfl.cpp: Added.
25964         (GCController::collect):
25965         (GCController::collectOnAlternateThread):
25966         (GCController::getJSObjectCount):
25967         * DumpRenderTree/efl/PixelDumpSupportEfl.cpp: Added.
25968         (createBitmapContextFromWebView):
25969         * DumpRenderTree/efl/WorkQueueItemEfl.cpp: Added.
25970         (LoadItem::invoke):
25971         (LoadHTMLStringItem::invoke):
25972         (ReloadItem::invoke):
25973         (ScriptItem::invoke):
25974         (BackForwardItem::invoke):
25975         * DumpRenderTree/efl/WorkQueueItemEfl.h: Added.
25976         (WorkQueueItem::~WorkQueueItem):
25977         (LoadItem::LoadItem):
25978         (LoadHTMLStringItem::LoadHTMLStringItem):
25979         (ScriptItem::ScriptItem):
25980         (LoadingScriptItem::LoadingScriptItem):
25981         (LoadingScriptItem::invoke):
25982         (NonLoadingScriptItem::NonLoadingScriptItem):
25983         (NonLoadingScriptItem::invoke):
25984         (BackForwardItem::BackForwardItem):
25985         (BackItem::BackItem):
25986         (ForwardItem::ForwardItem):
25987
25988 2011-06-16  Kevin Ollivier  <kevino@theolliviers.com>
25989
25990         [wx] Unreviewed build fix, make sure the waf build checks the webKitBranchBuild
25991         preference like the perl scripts do.
25992
25993         * waf/build/build_utils.py:
25994         * waf/build/settings.py:
25995
25996 2011-06-16  Tommy Widenflycht  <tommyw@google.com>
25997
25998         Reviewed by Tony Gentilcore.
25999
26000         MediaStream API: Runtime enabled features
26001         https://bugs.webkit.org/show_bug.cgi?id=62798
26002
26003         Add the WebRuntimeFeatures API for Media Streams and enable the feature by default in DumpRenderTree.
26004
26005         * DumpRenderTree/chromium/TestShell.cpp:
26006         (TestShell::TestShell):
26007
26008 2011-06-16  Torsten Kurbad  <gentoo@tk-webart.de>
26009
26010         Reviewed by Eric Seidel.
26011
26012         Add support for sys.platform == 'linux3' by relaxing
26013         conditional statements and/or extending lookup dicts.
26014         This way, Linux kernels >=3.0-rc1 get supported in the
26015         build process.
26016         https://bugs.webkit.org/show_bug.cgi?id=62730
26017
26018         * Scripts/webkitpy/common/system/executive.py:
26019         * Scripts/webkitpy/common/system/file_lock.py:
26020         * Scripts/webkitpy/common/system/path_unittest.py:
26021         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
26022         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
26023         * Scripts/webkitpy/layout_tests/port/chromium_gpu_unittest.py:
26024         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
26025         * Scripts/webkitpy/layout_tests/port/factory.py:
26026         * Scripts/webkitpy/layout_tests/port/factory_unittest.py:
26027         * Scripts/webkitpy/layout_tests/port/qt.py:
26028         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
26029
26030 2011-06-16  Sheriff Bot  <webkit.review.bot@gmail.com>
26031
26032         Unreviewed, rolling out r88796.
26033         http://trac.webkit.org/changeset/88796
26034         https://bugs.webkit.org/show_bug.cgi?id=62790
26035
26036         It made fast/dom/nodesFromRect-basic.html time out on Qt,
26037         64-bit, debug mode (Requested by Ossy on #webkit).
26038
26039         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
26040         (LayoutTestController::nodesFromRect):
26041         * DumpRenderTree/qt/LayoutTestControllerQt.h:
26042         * DumpRenderTree/qt/PlainTextControllerQt.cpp:
26043         (PlainTextController::plainText):
26044         * DumpRenderTree/qt/PlainTextControllerQt.h:
26045
26046 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
26047
26048         Unreviewed, build fix.
26049
26050         Revert the patches in r88995 and r88999; the http servers don't
26051         seem to be shutting down reliably.
26052
26053         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
26054         * Scripts/webkitpy/layout_tests/port/base.py:
26055         * Scripts/webkitpy/layout_tests/port/http_server.py:
26056         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
26057         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Removed.
26058         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
26059         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
26060
26061 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
26062
26063         Unreviewed, build fix.
26064
26065         Add some debugging messages to triage problems introduced in
26066         r88995.
26067
26068         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
26069
26070 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
26071
26072         Reviewed by Tony Chang.
26073
26074         nrwt: should clean up stale server processes from a previous run
26075         https://bugs.webkit.org/show_bug.cgi?id=62615
26076
26077         This change fixes cases where NRWT might start a server and then
26078         exit, leaving the server still running. Now NRWT should detect
26079         this case and clean up on the next run, without shutting down
26080         any processes it didn't start.
26081
26082         This change also fixes a bug in executive.kill_process on UNIX
26083         where zombies weren't being handled properly.
26084
26085         * Scripts/webkitpy/common/system/executive.py:
26086         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
26087         * Scripts/webkitpy/layout_tests/port/base.py:
26088         * Scripts/webkitpy/layout_tests/port/http_server.py:
26089         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
26090         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
26091         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
26092
26093 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
26094
26095         Reviewed by Tony Chang.
26096
26097         nrwt: fix http, websocket server startup, shutdown
26098         https://bugs.webkit.org/show_bug.cgi?id=62180
26099
26100         This change refactors all of the server-related code to
26101         have consistent semantics. It pushes the start()/stop()
26102         logic into the base class and adds four much smaller routines
26103         for specialization:
26104         _prepare_config() - for optionally writing config files etc.
26105             to disk prior to starting the server
26106         _remove_stale_log_files()
26107         _spawn_process() - the actual launch of the subprocess
26108         _cleanup_after_stop() - and custom cleanup code
26109
26110         Prior to this change the three server implementations were
26111         fairly inconsistent, leading to lots of weirdness on the bots
26112         with servers being left around, etc.
26113
26114         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
26115         * Scripts/webkitpy/layout_tests/port/http_server.py:
26116         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
26117         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
26118
26119 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
26120
26121         Reviewed by Tony Chang.
26122
26123         webkitpy: clean up code prior to functional changes for server startup/shutdown
26124         https://bugs.webkit.org/show_bug.cgi?id=62256
26125
26126         This patch does a bunch of minor cleanup of the code, but there
26127         should be no functional changes except that all of the modules
26128         will now use the same exception type. Other changes include
26129         removing unused code paths and functions, pushing 'name' and
26130         'pid_file' into the base class of the server objects, and using
26131         the Filesystem object instead of the codecs, tempfile, and
26132         shutils modules.
26133
26134         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
26135         * Scripts/webkitpy/layout_tests/port/http_server.py:
26136         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
26137         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
26138
26139 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
26140
26141         Reviewed by Tony Chang.
26142
26143         webkitpy: add integration tests for new-run-webkit-httpd, stop calling shut_down_http_server
26144         https://bugs.webkit.org/show_bug.cgi?id=62251
26145
26146         shut_down_http_server() was a total hack that was only used by
26147         new-run-webkit-httpd, so I've moved the code there and switched
26148         to using executive.kill_process() for the common case. The
26149         method itself will be removed in the patch on bug 59993.
26150
26151         * Scripts/new-run-webkit-httpd:
26152         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
26153         * Scripts/webkitpy/layout_tests/port/http_server.py:
26154         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Added.
26155         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
26156
26157 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
26158
26159         Unreviewed, build fix.
26160
26161         Fix test-webkitpy failure introduced in r88986.
26162
26163         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
26164
26165 2011-06-15  Dirk Pranke  <dpranke@chromium.org>
26166
26167         Reviewed by Eric Seidel.
26168
26169         NRWT: doesn't support webarchives, has stale skipping logic.
26170         https://bugs.webkit.org/show_bug.cgi?id=58293
26171
26172         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
26173         * Scripts/webkitpy/layout_tests/port/base.py:
26174         * Scripts/webkitpy/layout_tests/port/webkit.py:
26175
26176 2011-06-15  Jeremy Orlow  <jorlow@webkit.org>
26177
26178         Change my email to @webkit.org
26179
26180         * Scripts/webkitpy/common/config/committers.py:
26181
26182 2011-06-15  Nikolas Zimmermann  <nzimmermann@rim.com>
26183
26184         Reviewed by David Hyatt.
26185
26186         CSS 2.1: Support replaced elements with relative intrinsic sizes
26187         https://bugs.webkit.org/show_bug.cgi?id=15849
26188
26189         CSS 2.1 failure: html4/replaced-intrinsic-ratio-001
26190         https://bugs.webkit.org/show_bug.cgi?id=53099
26191
26192         Recognize *.htm tests in the old & new DRT implementation, needed for the vanilla css2.1 html4 tests.
26193
26194         * Scripts/old-run-webkit-tests:
26195         * Scripts/webkitpy/layout_tests/port/test_files.py:
26196
26197 2011-06-15  Carlos Garcia Campos  <cgarcia@igalia.com>
26198
26199         Reviewed by Martin Robinson.
26200
26201         [GTK] Add a statusbar to MiniBrowser
26202         https://bugs.webkit.org/show_bug.cgi?id=62634
26203
26204         It shows the url of the current hovered link.
26205
26206         * MiniBrowser/gtk/BrowserWindow.c:
26207         (browserWindowConstructed): Create the GtkSatusbar.
26208         (mouseDidMoveOverElement): Update statusbar text.
26209         (browserWindowUIClientInit): Add implementation for
26210         mouseDidMoveOverElement().
26211         * MiniBrowser/gtk/GNUmakefile.am: Add new files to compilation.
26212         * MiniBrowser/gtk/WebBundle/WebBundleMain.c: Added.
26213         (mouseDidMoveOverElement): Pass the url of the current hovered
26214         link to the UI process.
26215         (didCreatePage): Set the UI client adding an implementation for
26216         mouseDidMoveOverElement().
26217         (WKBundleInitialize):
26218         * MiniBrowser/gtk/main.c:
26219         (createWKContextWithInjectedBundle): Create a global context with
26220         the injected bundle.
26221         (loadURI): Use the global context.
26222         (main):
26223
26224 2011-06-14  Lucas Forschler  <lforschler@apple.com>
26225
26226         Reviewed by Stephanie Lewis.
26227
26228         https://bugs.webkit.org/show_bug.cgi?id=62495
26229         Bug 62495 - combine windows and mac kill-old-processes script
26230         Add a platform agnostic script.  Remove the recently added mac flavor.
26231         Note: Leaving the windows one alone, since it is currently used by the WebKit bots.
26232
26233         * BuildSlaveSupport/kill-old-processes: Added.
26234         * BuildSlaveSupport/mac: Removed.
26235         * BuildSlaveSupport/mac/kill-old-processes: Removed.
26236
26237 2011-06-14  Dirk Pranke  <dpranke@chromium.org>
26238
26239         Reviewed by Tony Chang.
26240
26241         nrwt: handle worker exceptions cleanly
26242         https://bugs.webkit.org/show_bug.cgi?id=62614
26243
26244         This change modifiers new-run-webkit-tests to handle exceptions
26245         raised by worker threads better, by:
26246         - capturing the worker's stack and logging it in the manager
26247         - propagating the worker's exception in the caller correctly
26248         - attempting to cancel the workers and clean up even when
26249           we get an unexpected exception
26250
26251         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
26252         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
26253         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
26254         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
26255
26256 2011-06-14  Qi Zhang  <qi.2.zhang@nokia.com>
26257
26258         Reviewed by Laszlo Gombos.
26259
26260         [Qt] fast/dom/HTMLScriptElement/nested-execution.html failed
26261         https://bugs.webkit.org/show_bug.cgi?id=62227
26262
26263         QtWebkit does not yet support different CacheModels. This change will
26264         expose setCacheModel() with a stub implementation, which is enough to pass the LayoutTest.
26265
26266         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
26267         (LayoutTestController::setCacheModel):
26268         * DumpRenderTree/qt/LayoutTestControllerQt.h:
26269
26270 2011-06-14  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
26271
26272         Reviewed by Andreas Kling.
26273
26274         [Qt] tst_QWebFrame::overloadedSlots() fails
26275         https://bugs.webkit.org/show_bug.cgi?id=37319
26276
26277         Since we don't implictly convert 'document' object to QWebElement
26278         in metacalls anymore, change the controller to expect QVariantMap instead.
26279
26280         The method plainText() was updated to use QVariantMap as well to let the bridge
26281         do the conversion directly for us.
26282
26283         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
26284         (LayoutTestController::nodesFromRect):
26285         * DumpRenderTree/qt/LayoutTestControllerQt.h:
26286         * DumpRenderTree/qt/PlainTextControllerQt.cpp:
26287         (PlainTextController::plainText):
26288         * DumpRenderTree/qt/PlainTextControllerQt.h:
26289
26290 2011-06-14  Andras Becsi  <abecsi@webkit.org>
26291
26292         Reviewed by Csaba Osztrogonác.
26293
26294         Remove dead code in DumpRenderTree/TestNetscapePlugIn/main.cpp
26295         https://bugs.webkit.org/show_bug.cgi?id=62630
26296
26297         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
26298         (NPP_New): remove dead variable 'forceCarbon'
26299
26300 2011-06-13  Hayato Ito  <hayato@chromium.org>
26301
26302         Reviewed by Ojan Vafai.
26303
26304         [NRWT] Fix a WebKitDriver so that it returns an actual hash value if DRT returns that.
26305         https://bugs.webkit.org/show_bug.cgi?id=59188
26306
26307         A current WebKitDriver doesn't try to return an actual hash value if DRT does not return an image.
26308         WebKitDriver should always return an actual hash value if DRT returns that.
26309
26310         No tests since I'll add sample reftests to make sure mismatch reftests work later,
26311         which should be in a separate patch.
26312
26313         * Scripts/webkitpy/layout_tests/port/webkit.py:
26314
26315 2011-06-13  Kent Tamura  <tkent@chromium.org>
26316
26317         [Chromium-win] Fix two wrong assertions exposed by r88757.
26318
26319         * DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
26320         (WebThemeEngineDRTWin::paintTrackbar):
26321
26322 2011-06-13  Alexey Proskuryakov  <ap@apple.com>
26323
26324         Added an updated e-mail address for Matt Lilek, to make Bugzilla autocomplete
26325         and commit queue happy.
26326
26327         * Scripts/webkitpy/common/config/committers.py:
26328
26329 2011-06-13  Chris Rogers  <crogers@google.com>
26330
26331         Unreviewed build fix.
26332
26333         Build fix to chromium DRT until we can rebaseline some tests
26334         https://bugs.webkit.org/show_bug.cgi?id=62608
26335
26336         * DumpRenderTree/chromium/TestShell.cpp:
26337         (TestShell::TestShell):
26338
26339 2011-06-13  Chris Rogers  <crogers@google.com>
26340
26341         Reviewed by Dirk Pranke.
26342
26343         Add Web Audio support to chromium DRT
26344         https://bugs.webkit.org/show_bug.cgi?id=62306
26345
26346         * DumpRenderTree/chromium/LayoutTestController.cpp:
26347         (LayoutTestController::LayoutTestController):
26348         (LayoutTestController::reset):
26349         (LayoutTestController::setEncodedAudioData):
26350         * DumpRenderTree/chromium/LayoutTestController.h:
26351         (LayoutTestController::encodedAudioData):
26352         (LayoutTestController::shouldDumpAsAudio):
26353         (LayoutTestController::setShouldDumpAsAudio):
26354         * DumpRenderTree/chromium/TestEventPrinter.cpp:
26355         (DRTPrinter::handleAudioHeader):
26356         (TestShellPrinter::handleAudioHeader):
26357         * DumpRenderTree/chromium/TestEventPrinter.h:
26358         * DumpRenderTree/chromium/TestShell.cpp:
26359         (TestShell::TestShell):
26360         (TestShell::dump):
26361
26362 2011-06-13  Jaehun Lim  <ljaehun.lim@samsung.com>
26363
26364         Reviewed by Eric Seidel.
26365
26366         [EFL] Add load error handler to EWebLauncher
26367         https://bugs.webkit.org/show_bug.cgi?id=62347
26368
26369         Add a handler function of "load,error" to EWebLauncher.
26370         EWebLauncher now displays a simple error page on load errors.
26371
26372         * EWebLauncher/main.c:
26373         (on_load_error):
26374         (browserCreate):
26375
26376 2011-06-13  Martin Robinson  <mrobinson@igalia.com>
26377
26378         Reviewed by Eric Seidel.
26379
26380         [GTK] [Qt] Eliminate duplicate TestNetscapePlugin implementation
26381         https://bugs.webkit.org/show_bug.cgi?id=62385
26382
26383         Remove duplicate TestNetscapePlugIn implementation. Having two copies
26384         of this code makes keeping it in sync much more difficult. The files are
26385         almost identical and this change ports the X11 changes to main.cpp.
26386
26387         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
26388         (NPP_New): Do not set the plugin as windowless always. Doing this breaks
26389         some results on Qt and this seems to be unnecessary for the moment.
26390         (keyEventToChar): Added this helper which converts a X11 keycode into a char.
26391         (handleEventX11): Use the handler to properly convert the keycode. Do not print
26392         adjustCursorEvent output. The tests currently do not trigger this for Cocoa and thus
26393         it should not be in the results.
26394         (NPP_GetValue): Properly handle NPPVpluginNeedsXEmbed and make it clearer why the
26395         XP_UNIX block is first by guarding against using a null instance.
26396         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Update source list add
26397         the XP_UNIX define for Linux.
26398         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp: Removed.
26399         * GNUmakefile.am: Update source list.
26400
26401 2011-06-13  Sheriff Bot  <webkit.review.bot@gmail.com>
26402
26403         Unreviewed, rolling out r88671.
26404         http://trac.webkit.org/changeset/88671
26405         https://bugs.webkit.org/show_bug.cgi?id=62591
26406
26407         "didn't work" (Requested by dpranke on #webkit).
26408
26409         * Scripts/new-run-webkit-httpd:
26410         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
26411         * Scripts/webkitpy/layout_tests/port/http_server.py:
26412         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
26413         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Removed.
26414         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
26415         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
26416
26417 2011-06-13  Nate Chapin  <japhet@chromium.org>
26418
26419         Reviewed by Darin Fisher.
26420
26421         New test method on TestNetscapePlugin for https://bugs.webkit.org/show_bug.cgi?id=61482.
26422
26423         * DumpRenderTree/TestNetscapePlugIn/PluginObject.cpp:
26424
26425 2011-06-13  Dirk Pranke  <dpranke@chromium.org>
26426
26427         Reviewed by Tony Chang.
26428
26429         webkitpy: add integration tests for new-run-webkit-httpd, stop calling shut_down_http_server
26430         https://bugs.webkit.org/show_bug.cgi?id=62251
26431
26432         shut_down_http_server() was a total hack that was only used by
26433         new-run-webkit-httpd, so I've moved the code there and switched
26434         to using executive.kill_process() for the common case. The
26435         method itself will be removed in the patch on bug 59993.
26436
26437         * Scripts/new-run-webkit-httpd:
26438         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
26439         * Scripts/webkitpy/layout_tests/port/http_server.py:
26440         * Scripts/webkitpy/layout_tests/port/http_server_integrationtest.py: Added.
26441         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
26442
26443 2011-06-13  Tony Chang  <tony@chromium.org>
26444
26445         Reviewed by Dimitri Glazkov.
26446
26447         rename ENABLE_NEW_FLEXBOX to ENABLE_CSS3_FLEXBOX
26448         https://bugs.webkit.org/show_bug.cgi?id=62578
26449
26450         * Scripts/build-webkit:
26451
26452 2011-06-13  Tony Chang  <tony@chromium.org>
26453
26454         Reviewed by Adam Barth.
26455
26456         rename ENABLE_FLEXBOX to ENABLE_NEW_FLEXBOX
26457         https://bugs.webkit.org/show_bug.cgi?id=62545
26458
26459         * Scripts/build-webkit:
26460
26461 2011-06-13  Carlos Garcia Campos  <cgarcia@igalia.com>
26462
26463         Reviewed by Martin Robinson.
26464
26465         [GTK] Export an API similar to WebKit1
26466         https://bugs.webkit.org/show_bug.cgi?id=57820
26467
26468         Build GtkLauncher2 for WebKit2. It's actually the same GtkLauncher
26469         code, but linking to libwebkit2gtk.
26470
26471         * GNUmakefile.am:
26472         * GtkLauncher/main.c:
26473         (createBrowser):
26474         (main):
26475
26476 2011-06-12  MORITA Hajime  <morrita@google.com>
26477
26478         Unreviewed, rolling out r88625.
26479         http://trac.webkit.org/changeset/88625
26480         https://bugs.webkit.org/show_bug.cgi?id=61073
26481
26482         Breaks SL Webkit2 Tests
26483
26484         * WebKitTestRunner/Configurations/InjectedBundleCFLite.vsprops:
26485         * WebKitTestRunner/Configurations/InjectedBundleCommon.vsprops:
26486         * WebKitTestRunner/Configurations/InjectedBundleCoreFoundation.vsprops:
26487         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
26488         (WTR::InjectedBundlePage::didClearWindowForFrame):
26489         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
26490
26491 2011-06-12  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
26492
26493         Reviewed by Antonio Gomes.
26494
26495         [Qt] LayoutTestController needs setTextDirection implementation
26496         https://bugs.webkit.org/show_bug.cgi?id=62442
26497
26498         Implements LayoutController.setTextDirection.
26499
26500         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
26501         (LayoutTestController::setTextDirection):
26502         * DumpRenderTree/qt/LayoutTestControllerQt.h:
26503
26504 2011-06-12  Dominic Cooney  <dominicc@chromium.org>
26505
26506         Reviewed by Hajime Morita.
26507
26508         Add window.internals to WebKit2's WebKitTestRunner.
26509         https://bugs.webkit.org/show_bug.cgi?id=61073
26510
26511         Test: fast/harness/internals-object.html
26512
26513         * WebKitTestRunner/Configurations/InjectedBundleCFLite.vsprops:
26514         * WebKitTestRunner/Configurations/InjectedBundleCommon.vsprops:
26515         * WebKitTestRunner/Configurations/InjectedBundleCoreFoundation.vsprops:
26516         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
26517         (WTR::InjectedBundlePage::didClearWindowForFrame): initialize window.internals
26518         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
26519
26520 2011-06-12  Sheriff Bot  <webkit.review.bot@gmail.com>
26521
26522         Unreviewed, rolling out r88616.
26523         http://trac.webkit.org/changeset/88616
26524         https://bugs.webkit.org/show_bug.cgi?id=62517
26525
26526         It broke editing/text-iterator/findString.html (Requested by
26527         tonikitoo on #webkit).
26528
26529         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
26530         * DumpRenderTree/qt/LayoutTestControllerQt.h:
26531
26532 2011-06-12  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
26533
26534         Reviewed by Antonio Gomes.
26535
26536         [Qt] LayoutTestController needs setTextDirection implementation
26537         https://bugs.webkit.org/show_bug.cgi?id=62442
26538
26539         Implements LayoutController.setTextDirection.
26540
26541         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
26542         (LayoutTestController::setTextDirection):
26543         * DumpRenderTree/qt/LayoutTestControllerQt.h:
26544
26545 2011-06-10  Ryosuke Niwa  <rniwa@webkit.org>
26546
26547         Reviewed by Ojan Vafai.
26548
26549         new-run-webkit-tests: 'images' links on results.html don't work
26550         https://bugs.webkit.org/show_bug.cgi?id=62194
26551
26552         Use relative path when generating images-diffs.html
26553
26554         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
26555         Added _output_testname.
26556
26557 2011-06-10  Lucas Forschler  <lforschler@apple.com>
26558
26559         Reviewed by Stephanie Lewis.
26560
26561         Script to kill old processes on Mac.
26562         This matches the windows version.
26563         https://bugs.webkit.org/show_bug.cgi?id=62479
26564
26565         * BuildSlaveSupport/mac: Added.
26566         * BuildSlaveSupport/mac/kill-old-processes: Added.
26567
26568 2011-06-10  Adam Barth  <abarth@webkit.org>
26569
26570         Reviewed by James Robinson.
26571
26572         When sheriffbot reports failures on IRC instead of saying "(and more..)" it should say the number of tests that are failing
26573         https://bugs.webkit.org/show_bug.cgi?id=62489
26574
26575         * Scripts/webkitpy/tool/bot/sheriff.py:
26576         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
26577
26578 2011-06-10  Tony Chang  <tony@chromium.org>
26579
26580         Reviewed by Ojan Vafai.
26581
26582         add a compile guard ENABLE(FLEXBOX)
26583         https://bugs.webkit.org/show_bug.cgi?id=62049
26584
26585         * Scripts/build-webkit:
26586
26587 2011-06-10  Mark Rowe  <mrowe@apple.com>
26588
26589         Reviewed by Dimitri Glazkov.
26590
26591         <rdar://problem/9562114> Fix DumpRenderTree build for production configuration.
26592
26593         Ensure that DumpRenderTree can find libWebCoreTestSupport.dylib and the associated
26594         header file.
26595
26596         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
26597         * DumpRenderTree/mac/Configurations/Base.xcconfig:
26598
26599 2011-06-09  Csaba Osztrogonác  <ossy@webkit.org>
26600
26601         Unreviewed rolling out r88471, because it broke plugin tests on Qt.
26602
26603         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
26604         (handleEventX11):
26605         (NPP_GetValue):
26606         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
26607         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp: Added.
26608         (webkit_test_plugin_new_instance):
26609         (webkit_test_plugin_destroy_instance):
26610         (webkit_test_plugin_set_window):
26611         (executeScript):
26612         (webkit_test_plugin_new_stream):
26613         (webkit_test_plugin_destroy_stream):
26614         (webkit_test_plugin_stream_as_file):
26615         (webkit_test_plugin_write_ready):
26616         (webkit_test_plugin_write):
26617         (webkit_test_plugin_print):
26618         (keyEventToChar):
26619         (webkit_test_plugin_handle_event):
26620         (webkit_test_plugin_url_notify):
26621         (webkit_test_plugin_get_value):
26622         (webkit_test_plugin_set_value):
26623         (NP_GetMIMEDescription):
26624         (NP_Initialize):
26625         (NP_Shutdown):
26626         (NP_GetValue):
26627         * GNUmakefile.am:
26628
26629 2011-06-09  Martin Robinson  <mrobinson@igalia.com>
26630
26631         Reviewed by Eric Seidel.
26632
26633         [GTK] Implement the next bit of the TextInputController (hasMarkedText and markedRange)
26634         https://bugs.webkit.org/show_bug.cgi?id=55603
26635
26636         Add callbacks to support hasMarkedText and markedRange and also clean up existing
26637         TextInputController support.
26638
26639         * DumpRenderTree/gtk/TextInputController.cpp:
26640         (setMarkedTextCallback): Accept parameters in terms of start and length, instead
26641         of start and end. This matches the other ports. Do not use GLib style early returns.
26642         (hasMarkedTextCallback): Added.
26643         (markedRangeCallback): Added.
26644         (insertTextCallback): Cleanup as described above.
26645         (unmarkTextCallback): Cleanup as described above.
26646         (firstRectForCharacterRangeCallback): Cleanup as described above.
26647         (selectedRangeCallback): Cleanup as described above.
26648
26649 2011-06-09  Dirk Pranke  <dpranke@chromium.org>
26650
26651         Reviewed by Tony Chang.
26652
26653         nrwt: support webaudio in chromium driver
26654         https://bugs.webkit.org/show_bug.cgi?id=62226
26655
26656         * Scripts/webkitpy/layout_tests/port/chromium.py:
26657
26658 2011-06-09  Dan Bernstein  <mitz@apple.com>
26659
26660         Reviewed by Anders Carlsson.
26661
26662         Added a test for Vector::reverse()
26663         https://bugs.webkit.org/show_bug.cgi?id=62393
26664
26665         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
26666         * TestWebKitAPI/Tests/WTF/VectorReverse.cpp: Added.
26667         (TestWebKitAPI::TEST):
26668         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
26669
26670 2011-06-09  Martin Robinson  <mrobinson@igalia.com>
26671
26672         Reviewed by Andreas Kling.
26673
26674         [GTK] [Qt] Eliminate duplicate TestNetscapePlugin implementation
26675         https://bugs.webkit.org/show_bug.cgi?id=62385
26676
26677         Remove duplicate TestNetscapePlugIn implementation. Having two copies
26678         of this code makes keeping it in sync much more difficult. The files are
26679         almost identical and this change ports the X11 changes to main.cpp.
26680
26681         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
26682         (keyEventToChar): Added this helper which converts a X11 keycode into a char.
26683         (handleEventX11): Use the handler to properly convert the keycode. Do not print
26684         adjustCursorEvent output. The tests currently do not trigger this for Cocoa and thus
26685         it should not be in the results.
26686         (NPP_GetValue): Properly handle NPPVpluginNeedsXEmbed.
26687         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro: Update source list.
26688         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp: Removed.
26689         * GNUmakefile.am: Update source list.
26690
26691 2011-06-09  Noel Gordon  <noel.gordon@gmail.com>
26692
26693         Reviewed by Tony Chang.
26694
26695         Test plugin should support event logging on the windows port.
26696         https://bugs.webkit.org/show_bug.cgi?id=61721
26697
26698         The WebKit win port interprets a 0 return from plugin NPP_HandleEvent()
26699         calls to mean the plugin handled the event.
26700
26701         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
26702         (handleEventWin): Return 0 for each event on windows.
26703
26704 2011-06-09  Chang Shu  <cshu@webkit.org>
26705
26706         Reviewed by Andreas Kling.
26707
26708         [Qt] [WK2] Many editing tests failed on missing shouldBeginEditingInDOMRange:range printout
26709         https://bugs.webkit.org/show_bug.cgi?id=62381
26710
26711         Activate QGraphicsScene and setFocus on QGraphicsItem to make sure the view has the focus.
26712         QWKPagePrivate::isViewFocused() will check it later.
26713
26714         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
26715         (WTR::PlatformWebView::PlatformWebView):
26716
26717 2011-06-09  Robert Hogan  <robert@webkit.org>
26718
26719         Reviewed by Andreas Kling.
26720
26721         Teach Qt about window.internals
26722         https://bugs.webkit.org/show_bug.cgi?id=61074
26723
26724         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
26725         (WebCore::DumpRenderTree::initJSObjects):
26726
26727 2011-06-09  Eric Seidel  <eric@webkit.org>
26728
26729         Reviewed by Adam Barth.
26730
26731         Disable running layout tests on Mac EWS bots (and fix cr-linux bots to run them again)
26732         https://bugs.webkit.org/show_bug.cgi?id=62362
26733
26734         The mac bots are failing inspector tests.  We don't know why yet.
26735         So for now, turning run-webkit-tests off for the mac-ews.
26736
26737         Also noticed that due to the wrong variable name the cr-linux
26738         bots weren't running tests!
26739
26740         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
26741
26742 2011-06-09  Zoltan Herczeg  <zherczeg@inf.u-szeged.hu>
26743
26744         Reviewed by Csaba Osztrogonac.
26745
26746         [Qt][WK2] Canvas tests make css2.1 ones failing
26747         https://bugs.webkit.org/show_bug.cgi?id=53427
26748
26749         Adding a workaround for this issue by mimicing the behaviour of DumpRenderTree.
26750         Since the Qt port should eventually go to QRawFont, this issue will be resolved
26751         in the future. Both clearMemoryCaches() and removeAllApplicationFonts() are
26752         needed for the workaround.
26753
26754         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
26755         (WTR::LayoutTestController::platformInitialize):
26756
26757 2011-06-09  Mario Sanchez Prada  <msanchez@igalia.com>
26758
26759         Reviewed by Chris Fleizach.
26760
26761         [GTK] Provide a way in DRT to check the platform name
26762         https://bugs.webkit.org/show_bug.cgi?id=56855
26763
26764         Patch for adding 'platformName' static property to DRT's
26765         LayoutController and provide platform specific implementations.
26766
26767         In principle, this feature will be only of interest for writing
26768         accessibility tests, so that's why it's only implemented in the
26769         Mac, Gtk and Win platforms.
26770
26771         * DumpRenderTree/LayoutTestController.h: New method platformName().
26772         * DumpRenderTree/LayoutTestController.cpp:
26773         (getPlatformNameCallback): New callback for 'platformName'.
26774         (LayoutTestController::staticValues): New static property.
26775
26776         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
26777         (LayoutTestController::platformName): Implemented.
26778
26779         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
26780         (LayoutTestController::platformName): Implemented.
26781
26782         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
26783         (LayoutTestController::platformName): Implemented.
26784
26785 2011-06-08  Mike Lawther  <mikelawther@chromium.org>
26786
26787         Reviewed by James Robinson.
26788
26789         Add mikelawther to committers.py
26790         https://bugs.webkit.org/show_bug.cgi?id=62354
26791
26792         * Scripts/webkitpy/common/config/committers.py:
26793
26794 2011-06-08  Kent Tamura  <tkent@chromium.org>
26795
26796         [Chromium] Fix a typo.
26797
26798         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
26799         (WebThemeControlDRTWin::draw):
26800
26801 2011-06-08  Kent Tamura  <tkent@chromium.org>
26802
26803         Reviewed by Hajime Morita.
26804
26805         [Chromium] Support vertical sliders in Windows DRT
26806         https://bugs.webkit.org/show_bug.cgi?id=62352
26807
26808         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
26809         (WebThemeControlDRTWin::draw):
26810           Paint VerticalSliderThumbType and VerticalSliderTrackType.
26811         * DumpRenderTree/chromium/WebThemeControlDRTWin.h:
26812           Add VerticalSliderTrackType and VerticalSliderThumbType.
26813         * DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
26814         (WebThemeEngineDRTWin::paintTrackbar):
26815           Convert Windows identifiers to the internal identifiers for vertical sliders.
26816
26817 2011-06-08  Ojan Vafai  <ojan@chromium.org>
26818
26819         Reviewed by Tony Chang.
26820
26821         add a short-flag for --experimental-fully-parallel
26822         https://bugs.webkit.org/show_bug.cgi?id=62321
26823
26824         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
26825
26826 2011-06-08  Ryosuke Niwa  <rniwa@webkit.org>
26827
26828         Reviewed by Tony Chang.
26829
26830         new-run-webkit-tests: don't turn the tree red when there are new passes
26831         https://bugs.webkit.org/show_bug.cgi?id=62311
26832
26833         Look for "new passes" instead of "passed unexpectedly" in evaluateCommand
26834         to match commandComplete.
26835
26836         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
26837
26838 2011-06-08  Tom Hudson  <tomhudson@google.com>
26839
26840         Reviewed by Mihai Parparita.
26841
26842         Allow sorting in RebaselineServer based on 'metric' field in unexpected_results.json
26843         https://bugs.webkit.org/show_bug.cgi?id=60964
26844
26845         * Scripts/webkitpy/tool/commands/data/rebaselineserver/index.html:
26846         Add 'Sort tests by metric' link.
26847         * Scripts/webkitpy/tool/commands/data/rebaselineserver/main.js:
26848         (disableSorting): Activate 'Sort tests by metric' link.
26849         (enableSorting): Deactivate 'Sort tests by metric' link.
26850         (selectDirectory): Call enableSorting()/disableSorting() depending
26851         on currently selected failure type, and sort tests if requested.
26852         * Scripts/webkitpy/tool/commands/data/rebaselineserver/main.css:
26853         New .disabled-control class for deactivated links.
26854
26855 2011-06-08  Tom Sepez  <tsepez@chromium.org>
26856
26857         Reviewed by Adam Barth.
26858
26859         Add web permissions callbacks for mixed content blocking.
26860         https://bugs.webkit.org/show_bug.cgi?id=61946
26861
26862         * DumpRenderTree/chromium/LayoutTestController.cpp:
26863         (LayoutTestController::setAllowDisplayOfInsecureContent):
26864         (LayoutTestController::setAllowRunningOfInsecureContent):
26865         (LayoutTestController::overridePreference):
26866         * DumpRenderTree/chromium/WebPermissions.cpp:
26867         (WebPermissions::allowDisplayingInsecureContent):
26868         (WebPermissions::allowRunningInsecureContent):
26869         (WebPermissions::setDisplayingInsecureContentAllowed):
26870         (WebPermissions::setRunningInsecureContentAllowed):
26871         (WebPermissions::reset):
26872         * DumpRenderTree/chromium/WebPermissions.h:
26873
26874 2011-06-08  Dirk Pranke  <dpranke@chromium.org>
26875
26876         Reviewed by Tony Chang.
26877
26878         webkitpy: fix typo causing us to skip the port unit tests in chromium_linux
26879         https://bugs.webkit.org/show_bug.cgi?id=62259
26880
26881         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
26882
26883 2011-06-08  Ryosuke Niwa  <rniwa@webkit.org>
26884
26885         Reviewed by Tony Chang.
26886
26887         new-run-webkit-tests: master.cfg errors when nrwt successfully runs test
26888         https://bugs.webkit.org/show_bug.cgi?id=62303
26889
26890         Access self.incorrectLayoutLines instead of incorrectLayoutLines in evaluateCommand.
26891
26892         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
26893
26894 2011-06-08  Greg Simon  <gregsimon@chromium.org>
26895
26896         Reviewed by Dimitri Glazkov.
26897
26898         Control Indexeddb backends from LayoutTestController
26899         https://bugs.webkit.org/show_bug.cgi?id=61000
26900
26901         * DumpRenderTree/chromium/LayoutTestController.cpp:
26902         (LayoutTestController::LayoutTestController):
26903         (LayoutTestController::setOverrideIndexedDBBackingStore):
26904         (LayoutTestController::clearAllDatabases):
26905         * DumpRenderTree/chromium/LayoutTestController.h:
26906
26907 2011-06-08  Andreas Kling  <kling@webkit.org>
26908
26909         Add Pierre Rossi to contributors list for EZ-CC.
26910
26911         * Scripts/webkitpy/common/config/committers.py:
26912
26913 2011-06-08  Ilya Sherman  <isherman@chromium.org>
26914
26915         Reviewed by Andreas Kling.
26916
26917         Add Ilya Sherman to list of committers
26918         https://bugs.webkit.org/show_bug.cgi?id=62270
26919
26920         * Scripts/webkitpy/common/config/committers.py:
26921
26922 2011-06-07  Ryosuke Niwa  <rniwa@webkit.org>
26923
26924         Reviewed by Tony Chang.
26925
26926         new-run-webkit-tests: Bot master should print useful information on waterfall/console for nrwt
26927         https://bugs.webkit.org/show_bug.cgi?id=62178
26928
26929         Added commandComplete and evaluateCommand for NewRunWebKitTests class
26930         to parse new-run-webkit-tests' results.
26931
26932         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
26933
26934 2011-06-07  Dirk Pranke  <dpranke@chromium.org>
26935
26936         Reviewed by Eric Seidel.
26937
26938         webkitpy: add a popen() call to executive
26939         https://bugs.webkit.org/show_bug.cgi?id=62179
26940
26941         This change adds a popen() wrapper call to the Executive object.
26942         This will allow other webkitpy scripts that currently call
26943         subprocess directly to use Executive, instead.
26944
26945         No additional tests are necessary (the existing tests cover the code).
26946
26947         * Scripts/webkitpy/common/system/executive.py:
26948
26949 2011-06-07  Sam Weinig  <sam@webkit.org>
26950
26951         Reviewed by Anders Carlsson.
26952
26953         Remove repeated IPC traffic from SetRenderTreeSize message
26954         https://bugs.webkit.org/show_bug.cgi?id=62244
26955
26956         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
26957         (WTR::InjectedBundlePage::InjectedBundlePage):
26958         Add stub for didLayoutForFrame.
26959
26960 2011-06-07  Dirk Pranke  <dpranke@chromium.org>
26961
26962         Reviewed by Tony Chang.
26963
26964         rebaseline-chromium-webkit-tests: fix baselining order for linux, linux_x86_64
26965         https://bugs.webkit.org/show_bug.cgi?id=62241
26966
26967         * Scripts/webkitpy/layout_tests/port/chromium.py:
26968
26969 2011-06-07  Adam Barth  <abarth@webkit.org>
26970
26971         Rubber-stamped by Eric Seidel.
26972
26973         Enable tests on the mac-ews
26974
26975         When we bring the mac-ews back up, we're going to try running tests!
26976
26977         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
26978
26979 2011-06-07  Dmitry Lomov  <dslomov@google.com>
26980
26981         Reviewed by David Levin.
26982
26983         https://bugs.webkit.org/show_bug.cgi?id=62215
26984         Allow comparisons with 0 in LIKELY and UNLIKELY macros.
26985
26986         * Scripts/webkitpy/style/checkers/cpp.py:
26987         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
26988
26989 2011-06-07  Nico Weber  <thakis@chromium.org>
26990
26991         Reviewed by Dimitri Glazkov.
26992
26993         [chromium] -Wdelete-non-virtual-dtor pass for DumpRenderTree
26994         https://bugs.webkit.org/show_bug.cgi?id=62210
26995
26996         The change to TestEventPrinter fixes a latent bug, because
26997         objects are deleted through the TestEventPrinter type, but none
26998         of the subclasses have destructors or non-POD members.
26999
27000         The changes to NotificationPresenter and WebViewHost do _not_ to fix a
27001         real bug, they just make clang's -Wdelete-non-virtual-dtor happy. As
27002         discussed at http://codereview.chromium.org/7094005/, we prefer making
27003         leaf class destructors virtual over making the leaf classes final.
27004
27005         * DumpRenderTree/chromium/NotificationPresenter.cpp:
27006         (NotificationPresenter::~NotificationPresenter):
27007         * DumpRenderTree/chromium/NotificationPresenter.h:
27008         * DumpRenderTree/chromium/TestEventPrinter.cpp:
27009         (TestEventPrinter::~TestEventPrinter):
27010         * DumpRenderTree/chromium/TestEventPrinter.h:
27011         * DumpRenderTree/chromium/WebViewHost.h:
27012
27013 2011-06-06  Ryosuke Niwa  <rniwa@webkit.org>
27014
27015         Reviewed by Dirk Pranke.
27016
27017         [chromium] Chromium bots on build.webkit.org don't upload results.html
27018         https://bugs.webkit.org/show_bug.cgi?id=61961
27019
27020         Don't fix the path on chromium port erroneously.
27021
27022         * BuildSlaveSupport/test-result-archive:
27023
27024 2011-06-06  Martin Robinson  <mrobinson@igalia.com>
27025
27026         Fix the GTK+ build by ensuring that the autogenerated sources
27027         necessary for libWebCoreInternals are recorded in a predeclared
27028         variable before assigning to BUILT_SOURCES.
27029
27030         * GNUmakefile.am: Predeclare libwebcoreinternals_built_sources.
27031
27032 2011-06-06  Yong Li  <yoli@rim.com>
27033
27034         Reviewed by Eric Seidel.
27035
27036         https://bugs.webkit.org/show_bug.cgi?id=62097
27037         Fix the problem that integral bitfield check incorrectly barked at ?: expressions.
27038
27039         * Scripts/webkitpy/style/checkers/cpp.py:
27040         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
27041
27042 2011-06-06  Xan Lopez  <xlopez@igalia.com>
27043
27044         Reviewed by Martin Robinson.
27045
27046         [GTK] Enable feature/symbol detection in NWRT/GTK
27047         https://bugs.webkit.org/show_bug.cgi?id=62136
27048
27049         Disable feature detection through DRT, since we don't support it,
27050         and set the libwebcore patch correctly so that symbol detection
27051         through 'nm' works.
27052
27053         * Scripts/webkitpy/layout_tests/port/gtk.py:
27054
27055 2011-06-06  Shishir Agrawal  <shishir@chromium.org>
27056
27057         Reviewed by Tony Gentilcore.
27058
27059         Add a "prerender" state for page visibility.
27060         https://bugs.webkit.org/show_bug.cgi?id=62062
27061
27062         * DumpRenderTree/chromium/LayoutTestController.cpp:
27063         (LayoutTestController::setPageVisibility):
27064
27065 2011-06-06  Csaba Osztrogonác  <ossy@webkit.org>
27066
27067         [Qt] Implement download feature for QtTestBrowser
27068         https://bugs.webkit.org/show_bug.cgi?id=61865
27069
27070         Buildfix for --minimal and CONFIG+=qt_minimal build after r88161.
27071
27072         * QtTestBrowser/launcherwindow.cpp:
27073         (LauncherWindow::LauncherWindow):
27074         * QtTestBrowser/launcherwindow.h:
27075
27076 2011-06-06  Barát Tibor  <Barat.Tibor@stud.u-szeged.hu>
27077
27078         Reviewed by Csaba Osztrogonác.
27079
27080         [Qt] Implement download feature for QtTestBrowser
27081         https://bugs.webkit.org/show_bug.cgi?id=61865
27082
27083         * QtTestBrowser/launcherwindow.cpp:
27084         (LauncherWindow::LauncherWindow):
27085         (LauncherWindow::downloadRequest):
27086         (LauncherWindow::fileDownloadFinished):
27087         * QtTestBrowser/launcherwindow.h:
27088
27089 2011-06-04  Dominic Cooney  <dominicc@chromium.org>
27090
27091         Reviewed by Dimitri Glazkov.
27092
27093         Make DumpRenderTree initialize window.internals.
27094         https://bugs.webkit.org/show_bug.cgi?id=61076
27095
27096         * DumpRenderTree/win/DumpRenderTreeCommon.vsprops: link WebCoreTestSupport.
27097         * DumpRenderTree/win/FrameLoadDelegate.cpp:
27098         (FrameLoadDelegate::didClearWindowObjectForFrameInStandardWorld):
27099
27100 2011-06-06  Alexandru Chiculita  <achicu@adobe.com>
27101
27102         Reviewed by Kent Tamura.
27103
27104         Add ENABLE_CSS_EXCLUSIONS support for build-webkit script
27105         https://bugs.webkit.org/show_bug.cgi?id=61628
27106
27107         * Scripts/build-webkit:
27108
27109 2011-06-06  Mihnea Ovidenie  <mihnea@adobe.com>
27110
27111         Reviewed by Kent Tamura.
27112
27113         Add ENABLE(CSS_REGIONS) guard for CSS Regions support
27114         https://bugs.webkit.org/show_bug.cgi?id=61631
27115
27116         * Scripts/build-webkit:
27117
27118 2011-06-06  Carlos Garcia Campos  <cgarcia@igalia.com>
27119
27120         Unreviewed. Fix the GTK+ build.
27121
27122         * GNUmakefile.am: Use javascriptcore_cflags when building libWebCoreInternals.
27123
27124 2011-06-05  Alexey Proskuryakov  <ap@apple.com>
27125
27126         Add an e-mail address for Cameron McCormack that matches Bugzilla (so that auto-completion
27127         works properly).
27128
27129         * Scripts/webkitpy/common/config/committers.py:
27130
27131 2011-06-04  Kevin Ollivier  <kevino@theolliviers.com>
27132
27133         [wx] Unreviewed build fix. Update git branch build dir handling to match what other
27134         WebKit scripts expect, so other scripts work properly with wx.
27135
27136         * waf/build/build_utils.py:
27137         * waf/build/settings.py:
27138
27139 2011-06-04  Kevin Ollivier  <kevino@theolliviers.com>
27140
27141         [wx] Unreviewed build fix. Restore the PPC build and allow users to specify architectures
27142         to build on Mac.
27143
27144         * waf/build/settings.py:
27145         * wx/install-unix-extras:
27146
27147 2011-06-04  Kevin Ollivier  <kevino@theolliviers.com>
27148
27149         [wx] Unreviewed build fixes for recent trunk changes.
27150
27151         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
27152         (LayoutTestController::setTextDirection):
27153         * waf/build/settings.py:
27154
27155 2011-06-04  Martin Robinson  <mrobinson@igalia.com>
27156
27157         Fix the GTK+ build.
27158
27159         * GNUmakefile.am: Add some missing newline escape.
27160
27161 2011-06-03  Martin Robinson  <mrobinson@igalia.com>
27162
27163         Try to fix the GTK+ bulid.
27164
27165         * GNUmakefile.am: Add JSDOMWrapper.cpp to the libWebCoreInternals
27166         source list.
27167
27168 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
27169
27170         Reviewed by Adam Barth.
27171
27172         nrwt: should skip chromium test expectation overrides on upstream bots
27173         https://bugs.webkit.org/show_bug.cgi?id=61888
27174
27175         * Scripts/webkitpy/layout_tests/port/chromium.py:
27176         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
27177
27178 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
27179
27180         Reviewed by Ojan Vafai.
27181
27182         new-run-webkit-tests: minor logging, code cleanup
27183         https://bugs.webkit.org/show_bug.cgi?id=62055
27184
27185         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
27186         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
27187
27188 2011-06-03  Martin Robinson  <mrobinson@igalia.com>
27189
27190         Try to fix the build failure on the GTK+ 32-bit Release bot.
27191
27192         * GNUmakefile.am: Add JSInternal.{cpp,h} to the BUILT_SOURCES list.
27193
27194 2011-05-31  Martin Robinson  <mrobinson@igalia.com>
27195
27196         Reviewed by Ryosuke Niwa.
27197
27198         [GTK] Support smart replace for the pasteboard
27199         https://bugs.webkit.org/show_bug.cgi?id=61734
27200
27201         Call into the DumpRenderTreeSupportGtk hooks to toggle smart replace.
27202         This is required to produce consistent test results.
27203
27204         * DumpRenderTree/gtk/DumpRenderTree.cpp:
27205         (resetDefaultsToConsistentValues): Set the smart replace setting to on by default.
27206         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
27207         (LayoutTestController::setSmartInsertDeleteEnabled): Added an implementation of this
27208         methods which calls into DumpRenderTreeSupportGtk.
27209
27210 2011-06-03  Martin Robinson  <mrobinson@igalia.com>
27211
27212         Reviewed by Dimitri Glazkov.
27213
27214         Teach Gtk build about window.internals
27215         https://bugs.webkit.org/show_bug.cgi?id=61071
27216
27217         * DumpRenderTree/gtk/DumpRenderTree.cpp:
27218         (webViewWindowObjectCleared): Inject the new internals object when the
27219         window object clears.
27220         * GNUmakefile.am: Add new source files and includes necessary for window.internals.
27221         Also do a little bit of cleanup.
27222
27223 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
27224
27225         Reviewed by Tony Chang.
27226
27227         Worker may not be stopped after LT.
27228         https://bugs.webkit.org/show_bug.cgi?id=61847
27229
27230         Re-submit a variant of the first fix for this bug. It's possible
27231         that actually calling cleanup() in the worker threads was taking
27232         too long and we triggered the assertions in the manager thread.
27233
27234         I've bumped up the timeout in the manager and added better
27235         diagnostics. I also cleaned up the cleanup() code a bit because
27236         there was some unnecessary if-checks.
27237
27238         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
27239         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
27240
27241 2011-06-03  Dirk Pranke  <dpranke@chromium.org>
27242
27243         Reviewed by Tony Chang.
27244
27245         nrwt: fix duplicate logging regression
27246         https://bugs.webkit.org/show_bug.cgi?id=61981
27247
27248         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
27249
27250 2011-06-03  Robert Kroeger  <rjkroege@chromium.org>
27251
27252         Reviewed by Adam Barth.
27253
27254         Modified addTouchPoint to correctly re-use touchPoint id values starting
27255         from the lowest not-un-use touchPoint id as is done by the embedder
27256         APIs. (Fixes BUGCR36415)
27257
27258         https://bugs.webkit.org/show_bug.cgi?id=62016
27259
27260         * DumpRenderTree/chromium/EventSender.cpp:
27261         (EventSender::addTouchPoint):
27262
27263 2011-06-03  Cary Clark  <caryclark@google.com>
27264
27265         Reviewed by Eric Seidel.
27266
27267         Draw DRT scroll bar thumb with CG when WebKit renders with Skia
27268         https://bugs.webkit.org/show_bug.cgi?id=61534
27269
27270         Skia on Mac uses Skia to render WebKit, and CG
27271         to render UI elements. The CG elements need a
27272         transcribed graphics context that preserves the
27273         canvas matrix, and the canvas clip.
27274
27275         The SkiaBitLocker utility class sets up a CGContext
27276         from the SkCanvas, locks the bitmap's bits, and
27277         releases the lock when the class goes out of scope.
27278
27279         The SkiaBitLocker implementation is described by
27280         this patch:
27281         http://codereview.chromium.org/7031006/
27282
27283         When WebKit uses Skia for rendering, the WebCanvas
27284         is an SkCanvas. Create an equivalent CGContext to
27285         draw the scrollbar thumb.
27286
27287         As the Chrome Mac port does not yet use Skia, this
27288         patch has no effect and requires no tests.
27289
27290         * DumpRenderTree/chromium/WebThemeEngineDRTMac.mm:
27291         (WebThemeEngineDRTMac::paintHIThemeScrollbarThumb):
27292
27293 2011-06-03  Andras Becsi  <abecsi@webkit.org>
27294
27295         Unreviewed build fix.
27296
27297         Fix typo.
27298
27299         * Scripts/old-run-webkit-tests:
27300
27301 2011-06-03  Andras Becsi  <abecsi@webkit.org>
27302
27303         Reviewed by Csaba Osztrogonác.
27304
27305         Print error message when there is no httpd present on the system
27306         https://bugs.webkit.org/show_bug.cgi?id=61939
27307
27308         * Scripts/old-run-webkit-tests: Propose to run with --no-httpd.
27309         * Scripts/webkitperl/httpd.pm:
27310
27311 2011-06-03  Mikhail Naganov  <mnaganov@chromium.org>
27312
27313         Reviewed by Yury Semikhatsky.
27314
27315         Web Inspector: [Chromium] cpu-profiler-profiling layout test is flaky on Linux Debug.
27316         https://bugs.webkit.org/show_bug.cgi?id=61533
27317
27318         Skip cpu-profiler-profiling in debug, implement a fast headless alternative.
27319
27320         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
27321         (DRTDevToolsAgent::setJavaScriptProfilingEnabled):
27322         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
27323         * DumpRenderTree/chromium/LayoutTestController.cpp:
27324         (LayoutTestController::LayoutTestController):
27325         (LayoutTestController::setJavaScriptProfilingEnabled):
27326         * DumpRenderTree/chromium/LayoutTestController.h:
27327         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
27328         (LayoutTestController::setJavaScriptProfilingEnabled):
27329         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
27330         (WTR::LayoutTestController::setJavaScriptProfilingEnabled):
27331         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
27332
27333 2011-06-02  Dirk Pranke  <dpranke@chromium.org>
27334
27335         Reviewed by Tony Chang.
27336
27337         NRWT needs a way to log to a file without including backspaces without needing --verbose
27338         https://bugs.webkit.org/show_bug.cgi?id=60328
27339
27340         This change modifies NRWT so that if it can tell if it is
27341         writing to a terminal, it will overwrite any messages printed
27342         in the 'one-line-progress' and 'updates' print options, and if
27343         not (or if --verbose was specified) the messages will be printed
27344         without backspaces.
27345
27346         If the overwriting updates are disabled, one-line-progress will
27347         only print every 10 seconds instead of on every update. This
27348         ensures that we still get updates once in a while while the
27349         tests are running, even if we're not in --verbose mode. Note
27350         that one-line-progress will now still be printed in --verbose,
27351         since it's useful as long as you're not flooded with the
27352         updates.
27353
27354         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
27355         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
27356         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
27357
27358 2011-06-02  Martin Robinson  <mrobinson@igalia.com>
27359
27360         Reviewed by Gustavo Noronha Silva.
27361
27362         [GTK] autogen.sh is run twice for each buld on the bots
27363         https://bugs.webkit.org/show_bug.cgi?id=61951
27364
27365         * Scripts/webkitdirs.pm: Only check the previous autotools arguments
27366         for the WebKit project. This prevents build-jsc runs from forcing subsequent
27367         build-webkit runs to re-rerun autogen.sh.
27368
27369 2011-06-02  Sheriff Bot  <webkit.review.bot@gmail.com>
27370
27371         Unreviewed, rolling out r87946.
27372         http://trac.webkit.org/changeset/87946
27373         https://bugs.webkit.org/show_bug.cgi?id=61965
27374
27375         Cause NRWT to break (Requested by abarth|gardener on #webkit).
27376
27377         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
27378
27379 2011-06-02  Dimitri Glazkov  <dglazkov@chromium.org>
27380
27381         Reviewed by Darin Adler.
27382
27383         Add build logistics and plumbing for window.internals object.
27384         https://bugs.webkit.org/show_bug.cgi?id=60313
27385
27386         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
27387         * DumpRenderTree/chromium/TestShell.cpp:
27388         (TestShell::bindJSObjectsToWindow): Added injection code.
27389         * DumpRenderTree/mac/FrameLoadDelegate.mm:
27390         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
27391
27392 2011-06-02  Hao Zheng  <zhenghao@chromium.org>
27393
27394         Reviewed by Dirk Pranke.
27395
27396         Call cleanup() explicitly at the end of worker.run().
27397         https://bugs.webkit.org/show_bug.cgi?id=61847
27398
27399         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
27400
27401 2011-06-02  Dimitri Glazkov  <dglazkov@chromium.org>
27402
27403         Unreviewed, rolling out r87926.
27404         http://trac.webkit.org/changeset/87926
27405         https://bugs.webkit.org/show_bug.cgi?id=60313
27406
27407         Fails to find WebCoreTestSupport.dylib on bots.
27408
27409         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
27410         * DumpRenderTree/chromium/TestShell.cpp:
27411         (TestShell::bindJSObjectsToWindow):
27412         * DumpRenderTree/mac/FrameLoadDelegate.mm:
27413         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]):
27414
27415 2011-05-21  Dimitri Glazkov  <dglazkov@chromium.org>
27416
27417         Reviewed by Darin Adler.
27418
27419         Add build logistics and plumbing for window.internals object.
27420         https://bugs.webkit.org/show_bug.cgi?id=60313
27421
27422         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
27423         * DumpRenderTree/chromium/TestShell.cpp:
27424         (TestShell::bindJSObjectsToWindow): Added injection code.
27425         * DumpRenderTree/mac/FrameLoadDelegate.mm:
27426         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
27427
27428 2011-06-02  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
27429
27430         Reviewed by Eric Seidel.
27431
27432         Adjust efl coding style to more efl directories
27433         https://bugs.webkit.org/show_bug.cgi?id=61903
27434
27435         * Scripts/webkitpy/style/checker.py: Adjust efl coding style to efl directories.
27436
27437 2011-06-01  Dirk Pranke  <dpranke@chromium.org>
27438
27439         Reviewed by Tony Chang.
27440
27441         old-run-webkit-tests: add support for audio files
27442         https://bugs.webkit.org/show_bug.cgi?id=57992
27443
27444         * Scripts/old-run-webkit-tests:
27445
27446 2011-06-01  Jochen Eisinger  <jochen@chromium.org>
27447
27448         Reviewed by Adam Barth.
27449
27450         Implement layoutTestController.setImagesAllowed for controlling the respective behavior of the WebPermissionClient
27451         https://bugs.webkit.org/show_bug.cgi?id=34314
27452
27453         * DumpRenderTree/DumpRenderTree.gypi:
27454         * DumpRenderTree/chromium/LayoutTestController.cpp:
27455         (LayoutTestController::LayoutTestController):
27456         (LayoutTestController::setImagesAllowed):
27457         (LayoutTestController::setStorageAllowed):
27458         * DumpRenderTree/chromium/LayoutTestController.h:
27459         * DumpRenderTree/chromium/WebPermissions.cpp: Copied from Tools/DumpRenderTree/chromium/WebPermissions.h.
27460         (WebPermissions::WebPermissions):
27461         (WebPermissions::~WebPermissions):
27462         (WebPermissions::allowImages):
27463         (WebPermissions::allowStorage):
27464         (WebPermissions::setImagesAllowed):
27465         (WebPermissions::setStorageAllowed):
27466         (WebPermissions::reset):
27467         * DumpRenderTree/chromium/WebPermissions.h:
27468
27469 2011-06-01  Noel Gordon  <noel.gordon@gmail.com>
27470
27471         Reviewed by Tony Chang.
27472
27473         Test plugin should support event logging on the windows port.
27474         https://bugs.webkit.org/show_bug.cgi?id=61721
27475
27476         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
27477         (handleEventWin): Well, handle them.
27478
27479 2011-06-01  Daniel Cheng  <dcheng@chromium.org>
27480
27481         Reviewed by Tony Chang.
27482
27483         [chromium] Use correct file name for dragging out images.
27484         https://bugs.webkit.org/show_bug.cgi?id=24887
27485
27486         * DumpRenderTree/chromium/EventSender.cpp:
27487         (EventSender::EventSender):
27488         (EventSender::dumpFilenameBeingDragged):
27489         * DumpRenderTree/chromium/EventSender.h:
27490
27491 2011-06-01  Dirk Pranke  <dpranke@chromium.org>
27492
27493         Reviewed by Tony Chang.
27494
27495         nrwt: make webaudio work w/o content-length header
27496         https://bugs.webkit.org/show_bug.cgi?id=61819
27497
27498         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
27499         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
27500         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
27501         * Scripts/webkitpy/layout_tests/port/webkit.py:
27502
27503 2011-06-01  Viatcheslav Ostapenko  <ostapenko.viatcheslav@nokia.com>
27504
27505         Reviewed by Andreas Kling.
27506
27507         [Qt][WK2] Add QGLWidget viewport support to MiniBrowser
27508         https://bugs.webkit.org/show_bug.cgi?id=61802
27509
27510         Add command line parameter and menu item to MiniBrowser application
27511         enabling use of QGLWidget for browser viewport.
27512
27513         * MiniBrowser/qt/BrowserWindow.cpp:
27514         (BrowserWindow::BrowserWindow):
27515         (BrowserWindow::toggleGLViewport):
27516         * MiniBrowser/qt/BrowserWindow.h:
27517         * MiniBrowser/qt/MiniBrowserApplication.cpp:
27518         (MiniBrowserApplication::handleUserOptions):
27519         * MiniBrowser/qt/MiniBrowserApplication.h:
27520         (WindowOptions::WindowOptions):
27521
27522 2011-06-01  Martin Robinson  <mrobinson@igalia.com>
27523
27524         Reviewed by Xan Lopez.
27525
27526         Print out autogen arguments when they differ from the last set, to help
27527         with debugging build issues.
27528
27529         * Scripts/webkitdirs.pm: Print out previous and current build arguments
27530         when they differ.
27531
27532 2011-06-01  Carlos Garcia Campos  <cgarcia@igalia.com>
27533
27534         Reviewed by Martin Robinson.
27535
27536         [GTK] run-launcher should run MiniBrowser in WebKit2
27537         https://bugs.webkit.org/show_bug.cgi?id=61858
27538
27539         * Scripts/run-launcher: Check whether -2 command line option has
27540         been passed and run MiniBrowser instead of GtkLauncher.
27541
27542 2011-05-31  Hao Zheng  <zhenghao@chromium.org>
27543
27544         Reviewed by Dirk Pranke.
27545
27546         Fix exception in verbose printing mode.
27547         https://bugs.webkit.org/show_bug.cgi?id=61723
27548
27549         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
27550
27551 2011-05-31  Yong Li  <yoli@rim.com>
27552
27553         Reviewed by Eric Seidel.
27554
27555         https://bugs.webkit.org/show_bug.cgi?id=54807
27556         We have been assuming plain bitfields (like "int a : 31") are always signed integers.
27557         However some compilers can treat them as unsigned. For example, RVCT 4.0 states plain
27558         bitfields (declared without either signed or unsigned qualifiers) are treats as unsigned.
27559         http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.dui0348c/Babjddhe.html
27560         Although we can use "--signed-bitfields" flag to make RVCT 4.0 behave as most other compilers,
27561         always using "signed"/"unsigned" qualifier to declare integral type bitfields is still a good
27562         rule we should have in order to make our code independent from compilers and compiler flags.
27563
27564         * Scripts/webkitpy/style/checkers/cpp.py:
27565
27566 2011-05-31  Hironori Bono  <hbono@chromium.org>
27567
27568         Reviewed by Eric Seidel.
27569
27570         Add a JavaScript function layoutTestController.setTextDirection().
27571         https://bugs.webkit.org/show_bug.cgi?id=50952
27572
27573         This change adds a new JavaScript function setTextDirection() to the
27574         LayoutTestController class of Safari and Chrome to write layout tests
27575         that changes the text direction of an editable element.
27576
27577         * DumpRenderTree/LayoutTestController.cpp:
27578         (setTextDirectionCallback):
27579         (LayoutTestController::staticFunctions):
27580         * DumpRenderTree/LayoutTestController.h:
27581         * DumpRenderTree/chromium/LayoutTestController.cpp:
27582         (LayoutTestController::LayoutTestController):
27583         (LayoutTestController::setTextDirection):
27584         * DumpRenderTree/chromium/LayoutTestController.h:
27585         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
27586         (LayoutTestController::setTextDirection):
27587         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
27588         (LayoutTestController::setTextDirection):
27589         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
27590         (LayoutTestController::setTextDirection):
27591
27592 2011-05-31  Ademar de Souza Reis Jr  <ademar.reis@openbossa.org>
27593
27594         Reviewed by Simon Hausmann.
27595
27596         [Qt] User-agent list of Qt test browsers is very outdated
27597         https://bugs.webkit.org/show_bug.cgi?id=61531
27598
27599         The user can edit the user-agent manually, so there's no need to have
27600         a super complete and up to date list, but it's nice to have it
27601         minimally updated.
27602
27603         * MiniBrowser/qt/useragentlist.txt:
27604         * QtTestBrowser/useragentlist.txt:
27605
27606 2011-03-30  Martin Robinson  <mrobinson@igalia.com>
27607
27608         Reviewed by Adam Roben.
27609
27610         [GTK] [WebKit2] Implement a basic WebKitTestRunner
27611         https://bugs.webkit.org/show_bug.cgi?id=57068
27612
27613         Add an implementation of WebKitTestRunner for GTK+.
27614
27615         * Scripts/build-webkittestrunner: Added knowledge of GTK+ TestRunner.
27616         * Scripts/old-run-webkit-tests: Ditto.
27617         * Scripts/run-launcher: Ditto.
27618         * Scripts/webkitdirs.pm: Ditto.
27619         * WebKitTestRunner/GNUmakefile.am: Added.
27620         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
27621         * WebKitTestRunner/InjectedBundle/gtk/ActivateFontsGtk.cpp: Added.
27622         * WebKitTestRunner/InjectedBundle/gtk/InjectedBundleGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
27623         * WebKitTestRunner/InjectedBundle/gtk/LayoutTestControllerGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
27624         * WebKitTestRunner/PlatformWebView.h:
27625         * WebKitTestRunner/gtk/PlatformWebViewGtk.cpp: Added.
27626         * WebKitTestRunner/gtk/TestControllerGtk.cpp: Added.
27627         * WebKitTestRunner/gtk/TestInvocationGtk.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
27628         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
27629         * WebKitTestRunner/gtk/main.cpp: Copied from Source/WebKit2/WebProcess/InjectedBundle/gtk/InjectedBundleGtk.cpp.
27630         (main): Added.
27631
27632 2011-05-31  Dirk Pranke  <dpranke@chromium.org>
27633
27634         Reviewed by Tony Chang.
27635
27636         new-run-webkit-tests: doesn't wait for children if it gets a ctrl-c
27637         https://bugs.webkit.org/show_bug.cgi?id=60241
27638
27639         * Scripts/new-run-webkit-tests:
27640
27641 2011-05-31  Eric Seidel  <eric@webkit.org>
27642
27643         Reviewed by David Kilzer.
27644
27645         webkitpy should create zips with zip -9
27646         https://bugs.webkit.org/show_bug.cgi?id=61789
27647
27648         Dave Kilzer suggested we add this to make zips uploaded to bugzilla smaller.
27649
27650         * Scripts/webkitpy/common/system/workspace.py:
27651         * Scripts/webkitpy/common/system/workspace_unittest.py:
27652
27653 2011-05-31  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
27654
27655         Reviewed by Csaba Osztrogonác.
27656
27657         [Qt] Implement find feature for QtTestBrowser
27658         https://bugs.webkit.org/show_bug.cgi?id=61425
27659
27660         Buildfix for --minimal and CONFIG+=qt_minimal build.
27661
27662         * QtTestBrowser/launcherwindow.cpp:
27663         (LauncherWindow::LauncherWindow):
27664         (LauncherWindow::createChrome):
27665         * QtTestBrowser/launcherwindow.h:
27666
27667 2011-05-31  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
27668
27669         Reviewed by Csaba Osztrogonác.
27670
27671         [Qt] Implement find feature for QtTestBrowser
27672         https://bugs.webkit.org/show_bug.cgi?id=61425
27673
27674         * QtTestBrowser/launcherwindow.cpp:
27675         (LauncherWindow::LauncherWindow):
27676         (LauncherWindow::createChrome):
27677         (LauncherWindow::showFindBar):
27678         (LauncherWindow::find):
27679         * QtTestBrowser/launcherwindow.h:
27680
27681 2011-05-30  Carlos Garcia Campos  <cgarcia@igalia.com>
27682
27683         Reviewed by Martin Robinson.
27684
27685         [GTK] Implement JavaScript dialogs in MiniBrowser
27686         https://bugs.webkit.org/show_bug.cgi?id=61725
27687
27688         * MiniBrowser/gtk/BrowserWindow.c:
27689         (WKStringGetCString): Helper function to get a UTF-8 c string from
27690         a WKStringRef.
27691         (WKURLGetCString): Helper function to get a UTF-8 c string from a
27692         WKURLRef.
27693         (browserWindowUpdateURL): Update to use WKURLGetCString.
27694         (didReceiveTitleForFrame): Update to use WKStringGetCString.
27695         (createMessageDialog): Create a GtkMessageDialog.
27696         (runJavaScriptAlert):
27697         (runJavaScriptConfirm):
27698         (runJavaScriptPrompt):
27699         (browserWindowUIClientInit):
27700
27701 2011-05-29  Yuta Kitamura  <yutak@chromium.org>
27702
27703         Reviewed by Kent Tamura.
27704
27705         WebSocket closing handshake
27706         https://bugs.webkit.org/show_bug.cgi?id=35721
27707
27708         * Scripts/webkitpy/thirdparty/__init__.py:
27709         Pull in pywebsocket 0.6b1. We need to update pywebsocket
27710         to get the right behavior of closing handshake.
27711
27712 2011-05-29  Daniel Bates  <dbates@rim.com>
27713
27714         Reviewed by David Kilzer.
27715
27716         REGRESSION (r86515): svn-apply ignores diffs that omit line count in chunk range
27717         https://bugs.webkit.org/show_bug.cgi?id=61162
27718
27719         Fixes an issue where svn-apply may ignore a diff that contains a chunk range line
27720         that omits a line count. In particular, the chunk range regular expression does
27721         not match a chunk range line that omits a line count. GNU diff(1) will omit the
27722         line count in the chunk range if the line count is exactly 1. For example, appending
27723         a new line to the end of an existing file F that contains exactly one line of text will
27724         be represented in a diff with a chunk range line that omits the line count for F.
27725
27726         * Scripts/VCSUtils.pm:
27727           (parseChunkRange): Added.
27728         * Scripts/webkitperl/VCSUtils_unittest/parseChunkRange.pl: Added.
27729         * Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
27730           - Added unit test "Git: Append new line to the end of an existing file".
27731
27732 2011-05-28  Adam Barth  <abarth@webkit.org>
27733
27734         Reviewed by Eric Seidel.
27735
27736         EWS builds patches that fail to build twice, which seems useless and slows down the bots
27737         https://bugs.webkit.org/show_bug.cgi?id=55585
27738
27739         This patch switches all the early warning system bots over to the new
27740         PatchAnalysisTask-based infrastructure.  This patch makes these bots
27741         more efficient (in the case where patches fail to build) and paves the
27742         way for running tests on these bots!
27743
27744         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
27745         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py:
27746         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
27747         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
27748         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
27749         * Scripts/webkitpy/tool/commands/queues.py:
27750
27751 2011-05-28  Kenichi Ishibashi  <bashi@chromium.org>
27752
27753         Reviewed by Kent Tamura.
27754
27755         Add bashi as a committer.
27756
27757         https://bugs.webkit.org/show_bug.cgi?id=61685
27758
27759         * Scripts/webkitpy/common/config/committers.py:
27760
27761 2011-05-27  Adam Barth  <abarth@webkit.org>
27762
27763         Reviewed by Eric Seidel.
27764
27765         When checking whether the tree is red, the EWS posts a link to the wrong log
27766         https://bugs.webkit.org/show_bug.cgi?id=61072
27767
27768         We need to cache the original script error because that contains the
27769         failure log we want to upload.  If we don't cache that script error,
27770         self._script_error will get overwritten when we sanity check the clean
27771         tree (and it also has test failures).
27772
27773         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
27774             - This testing approach is slightly goofy.  We'd like to use
27775               assertRaisesRegexp, but that's not available until Python 2.7.
27776         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
27777
27778 2011-05-27  Jochen Eisinger  <jochen@chromium.org>
27779
27780         Reviewed by Adam Barth.
27781
27782         Add layoutTestController.setStorageAllowed() to control whether access
27783         to the localStorage API is enabled via the WebPermissionClient
27784         https://bugs.webkit.org/show_bug.cgi?id=61581
27785
27786         * DumpRenderTree/chromium/LayoutTestController.cpp:
27787         (LayoutTestController::LayoutTestController):
27788         (LayoutTestController::setStorageAllowed):
27789         * DumpRenderTree/chromium/LayoutTestController.h:
27790         * DumpRenderTree/chromium/TestShell.cpp:
27791         (TestShell::TestShell):
27792         (TestShell::createNewWindow):
27793         * DumpRenderTree/chromium/TestShell.h:
27794         * DumpRenderTree/chromium/WebPermissions.h: Added.
27795         (WebPermissions::WebPermissions):
27796         (WebPermissions::allowStorage):
27797         (WebPermissions::setStorageAllowed):
27798         (WebPermissions::reset):
27799
27800 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
27801
27802         Reviewed by Eric Seidel.
27803
27804         NRWT: clean up metered_stream code in preparation for 'nooverwriting' patch
27805         https://bugs.webkit.org/show_bug.cgi?id=60326
27806
27807         This patch removes a lot of the complexity from the
27808         metered_stream implementation that was unnecessary since there
27809         was only one caller and the logic could be coordinated better.
27810
27811         There should be no functional changes in this patch, just code
27812         getting deleted and cleaned up.
27813
27814         * Scripts/webkitpy/layout_tests/layout_package/metered_stream.py:
27815         * Scripts/webkitpy/layout_tests/layout_package/metered_stream_unittest.py:
27816         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
27817         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
27818
27819 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
27820
27821         Reviewed by Eric Seidel.
27822
27823         NRWT: minor cleanup in printing module
27824         https://bugs.webkit.org/show_bug.cgi?id=60329
27825
27826         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
27827
27828 2011-05-27  Steve Lacey  <sjl@chromium.org>
27829
27830         Reviewed by David Levin.
27831
27832         Add sjl@chromium.org as a committer.
27833
27834         https://bugs.webkit.org/show_bug.cgi?id=61672
27835
27836         * Scripts/webkitpy/common/config/committers.py:
27837
27838 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
27839
27840         Reviewed by Adam Barth.
27841
27842         NRWT: remove --print detailed-progress
27843         https://bugs.webkit.org/show_bug.cgi?id=60324
27844
27845         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
27846         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
27847         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
27848         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
27849
27850 2011-05-27  Dirk Pranke  <dpranke@chromium.org>
27851
27852         Reviewed by Ojan Vafai.
27853
27854         NRWT: debug messages from the workers are being logged twice
27855         https://bugs.webkit.org/show_bug.cgi?id=60428
27856
27857         It looks like when the workers are run in separate processes
27858         we end up getting two copies of every log message they print.
27859         This has to do with the multiprocessing module on UNIX cloning
27860         the log configuration in a way I wasn't expecting, and so two
27861         log handlers end up getting registered.
27862
27863         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
27864         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
27865         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
27866         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
27867
27868 2011-05-27  Adam Roben  <aroben@apple.com>
27869
27870         Always decode tester names from location.hash
27871
27872         When navigating directly to a tester-specific URL (e.g.,
27873         <http://build.webkit.org/TestFailures/#/Windows%207%20Release%20(Tests)>), you could end up
27874         with a URL-encoded tester name in the page, and then the encoded name could make its way
27875         into Bugzilla bugs. Bad!
27876
27877         Reviewed by David Kilzer.
27878
27879         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
27880         (ViewController.prototype.parseHash): Pass the tester name we got from location.hash through
27881         decodeURIComponent before trying to use it.
27882
27883 2011-05-27  Adam Roben  <aroben@apple.com>
27884
27885         Remove testing code that snuck into r87528
27886
27887         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
27888         (ViewController.prototype._domForNewBugLink): Remove a line I was using for some testing.
27889
27890 2011-05-27  Adam Roben  <aroben@apple.com>
27891
27892         Add links to TestFailures to easily file bugs about failing tests
27893
27894         Fixes <http://webkit.org/b/61636> <rdar://problem/9514886> TestFailures page should have
27895         links to file bugs for failing tests
27896
27897         Reviewed by David Kilzer.
27898
27899         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
27900         (Builder.prototype.resultsPageURL): New helper function to return the URL to results.html
27901         for a particular build.
27902         (Builder.prototype._getFailingTests): Use resultsPageURL.
27903
27904         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
27905         (ViewController.prototype._displayBuilder): Add links to file bugs about each set of
27906         failures.
27907         (ViewController.prototype._domForBuildName): Use resultsPageURL.
27908         (ViewController.prototype._domForNewBugLink): Added. Creates an element that represents the
27909         new bug link. The bug has a title, description, and URL already filled in for you!
27910
27911 2011-05-27  Sheriff Bot  <webkit.review.bot@gmail.com>
27912
27913         Unreviewed, rolling out r87464.
27914         http://trac.webkit.org/changeset/87464
27915         https://bugs.webkit.org/show_bug.cgi?id=61643
27916
27917         client-close/server-close tests are flaky on Chromium Mac and
27918         Windows (Requested by aklein on #webkit).
27919
27920         * Scripts/webkitpy/thirdparty/__init__.py:
27921
27922 2011-05-27  Alexey Proskuryakov  <ap@apple.com>
27923
27924         Added Douglas Davidson to contributors list for easier CC'ing.
27925
27926         * Scripts/webkitpy/common/config/committers.py:
27927
27928 2011-05-26  Mihai Parparita  <mihaip@chromium.org>
27929
27930         Reviewed by Adam Barth.
27931
27932         Fix worldID and destinationDomain argument names
27933         https://bugs.webkit.org/show_bug.cgi?id=61571
27934
27935         As part of working on r87423, I noticed a couple of inconsistencies in
27936         argument names:
27937         - We would use worldId in .h files but worldID in .cpp files.
27938           Standardize on the latter
27939         - SecurityOrigin::addOriginAccessWhitelistEntry would take a
27940           destinationDomains argument, even though the actual parameter was for
27941           a single domain (renamed to be singular).
27942
27943         * DumpRenderTree/LayoutTestController.h:
27944         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
27945         (LayoutTestController::evaluateScriptInIsolatedWorld):
27946
27947 2011-05-26  Yuta Kitamura  <yutak@chromium.org>
27948
27949         Reviewed by Kent Tamura.
27950
27951         WebSocket closing handshake
27952         https://bugs.webkit.org/show_bug.cgi?id=35721
27953
27954         * Scripts/webkitpy/thirdparty/__init__.py:
27955         Pull in pywebsocket 0.6b1. We need to update pywebsocket
27956         to get the right behavior of closing handshake.
27957
27958 2011-05-26  Qi Zhang  <qi.2.zhang@nokia.com>
27959
27960         Reviewed by Andreas Kling.
27961
27962         WebKitTestRunner needs layoutTestController.setAuthorAndUserStylesEnabled
27963         https://bugs.webkit.org/show_bug.cgi?id=42676
27964
27965         Implemented layoutTestController.setAuthorAndUserStylesEnabled for WebKitTestRunner.
27966
27967         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
27968         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
27969         (WTR::InjectedBundle::beginTesting):
27970         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
27971         (WTR::LayoutTestController::setAuthorAndUserStylesEnabled):
27972         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
27973
27974 2011-05-26  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
27975
27976         Reviewed by Andreas Kling.
27977
27978         [Qt] QtTestBrowser "Switch User Agent" dialog is broken
27979         https://bugs.webkit.org/show_bug.cgi?id=61527
27980
27981         The setMaximumHeight() call is unecessary and breaks the dialog at
27982         least on my setup.
27983
27984         * QtTestBrowser/launcherwindow.cpp:
27985         (LauncherWindow::showUserAgentDialog):
27986
27987 2011-05-02  Robert Hogan  <robert@webkit.org>
27988
27989         Reviewed by Adam Roben.
27990
27991         [Qt] Allow popup windows from plugins when initiated by a user gesture
27992
27993         https://bugs.webkit.org/show_bug.cgi?id=41292
27994
27995         Support parameters introduced in plugin-initiate-popup-window.html
27996
27997         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
27998         (webkit_test_plugin_new_instance):
27999         (webkit_test_plugin_handle_event):
28000
28001 2011-05-26  Mihai Parparita  <mihaip@chromium.org>
28002
28003         Reviewed by Adam Barth.
28004
28005         Support cross-origin XMLHttpRequest in isolated worlds
28006         https://bugs.webkit.org/show_bug.cgi?id=59843
28007
28008         Expose setIsolatedWorldSecurityOrigin in LayoutTestController.
28009
28010         * DumpRenderTree/chromium/LayoutTestController.cpp:
28011         (LayoutTestController::LayoutTestController):
28012         (LayoutTestController::setIsolatedWorldSecurityOrigin):
28013         * DumpRenderTree/chromium/LayoutTestController.h:
28014
28015 2011-05-25  Brian Weinstein  <bweinstein@apple.com>
28016
28017         Reviewed by Adam Roben.
28018
28019         WebKit2: Status bar, toolbar, and menu bar checks should be in the injected bundle
28020         https://bugs.webkit.org/show_bug.cgi?id=61474
28021         <rdar://problem/9468337>
28022
28023         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
28024         (WTR::InjectedBundlePage::InjectedBundlePage): Add empty entries in the WKBundlePageUIClient.
28025
28026 2011-05-26  Andreas Kling  <kling@webkit.org>
28027
28028         Unreviewed, correct address to webkit-committers mailing list.
28029
28030         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
28031
28032 2011-05-25  Adam Roben  <aroben@apple.com>
28033
28034         Don't assume that builds finish in the order they were started
28035
28036         Fixes <http://webkit.org/b/61464> TestFailures page shows failures from a not-most-recent
28037         build when one slave is taking a long time to finish a build
28038
28039         Reviewed by Antti Koivisto.
28040
28041         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
28042         (Builder.prototype.getMostRecentCompletedBuildNumber): Don't assume that all builds that are
28043         newer than the oldest in-progress build are also themselves in progress. Just check whether
28044         each build is actually in progress (as signified by its presence in the currentBuilds set).
28045
28046 2011-05-25  Adam Roben  <aroben@apple.com>
28047
28048         Correctly detect cases where only a single test failed/crashed/timed out
28049
28050         Fixes <http://webkit.org/b/61463> TestFailures page doesn't show testers with only a single
28051         failing test
28052
28053         Reviewed by Darin Adler.
28054
28055         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
28056         (Builder.prototype.getNumberOfFailingTests): Don't require "test cases" to be plural. It
28057         will be singular when only a single test fails/crashes/times out.
28058
28059         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
28060         (ViewController.prototype._displayTesters): Only pluralize "test" when more than one test
28061         failed/crashed/timed out.
28062
28063 2011-05-26  Tom Hudson  <tomhudson@google.com>
28064
28065         Reviewed by Tony Chang.
28066
28067         Add flags to Chromium ImageDiff to write image comparison metrics on
28068         stdout.
28069         https://bugs.webkit.org/show_bug.cgi?id=60569
28070
28071         * DumpRenderTree/chromium/ImageDiff.cpp:
28072         (Image::createFromStdin): Fix spelling.
28073         (maxOf3): New function to do 3-way maximum.
28074         (getRedComponent): Extract red component from ABGR packed encoding.
28075         (getGreenComponent): Extract green component from ABGR packed encoding.
28076         (getBlueComponent): Extract blue component from ABGR packed encoding.
28077         (weightedPercentageDifferent): Determine % of pixels different in two
28078         images multiplied by intensity difference of each pixel.
28079         (printHelp): Document new arguments.
28080         (compareImages): Parameterize.
28081         (untestedCompareImages): Parameterize.
28082         (diffImages): Parameterize; write results on stdout if requested.
28083         (main): Parse new arguments.
28084
28085 2011-05-25  Qi Zhang  <qi.2.zhang@nokia.com>
28086
28087         Reviewed by Simon Hausmann.
28088
28089         WebKitTestRunner needs layoutTestController.setPrivateBrowsingEnabled
28090         https://bugs.webkit.org/show_bug.cgi?id=42697
28091
28092         Implemented layoutTestController.setPrivateBrowsingEnabled for WebKitTestRunner.
28093
28094         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
28095         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
28096         (WTR::InjectedBundle::beginTesting):
28097         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
28098         (WTR::LayoutTestController::setPrivateBrowsingEnabled):
28099         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
28100
28101 2011-05-25  Jon Honeycutt  <jhoneycutt@apple.com>
28102
28103         REGRESSION (WebKit2): Crash in Flash on USA Today photo gallery
28104         https://bugs.webkit.org/show_bug.cgi?id=61428
28105         <rdar://problem/9457006>
28106
28107         Reviewed by Adam Roben.
28108
28109         The crash occurs when Flash posts a message to a window that it
28110         creates, and in processing the message, it calls NPN_Evaluate to
28111         evaluate JavaScript that removes the plug-in from the page. Flash then
28112         crashes when we return to Flash code.
28113
28114         This test emulates that behavior.
28115
28116         * DumpRenderTree/TestNetscapePlugIn/win/CallJSThatDestroysPlugin.cpp: Added.
28117         (CallJSThatDestroysPlugin::CallJSThatDestroysPlugin):
28118         Initialize member vars.
28119         (CallJSThatDestroysPlugin::~CallJSThatDestroysPlugin):
28120         Remove our custom property from the message window, and destroy it.
28121         (CallJSThatDestroysPlugin::NPP_Destroy):
28122         Set m_isDestroyed, log that the plug-in was destroyed, and notify the
28123         layout test controller that we're done.
28124         (wndProc):
28125         Get the PluginTest object, and call its runTest() function.
28126         (CallJSThatDestroysPlugin::NPP_New):
28127         Setup the test: register a class for the message-only window, create
28128         it, and post a message to it to run the test.
28129         (CallJSThatDestroysPlugin::runTest):
28130         Execute JS that removes the plug-in from the page, and if we're not
28131         destroyed, log a success message.
28132
28133         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
28134         Add new test to project.
28135
28136 2011-05-25  Tony Chang  <tony@chromium.org>
28137
28138         Reviewed by Adam Barth.
28139
28140         webkitpy's NetworkTransaction shouldn't use mechanize
28141         https://bugs.webkit.org/show_bug.cgi?id=61450
28142
28143         mechanize.HTTPError is just an alias for urllib2.HTTPError, so
28144         use urllib2 directly.  Re-add NetworkTransaction to
28145         test_results_uploader.py (reverting r87124).
28146
28147         * Scripts/webkitpy/common/net/networktransaction.py:
28148         * Scripts/webkitpy/layout_tests/layout_package/test_results_uploader.py:
28149
28150 2011-05-25  Kevin Ollivier  <kevino@theolliviers.com>
28151
28152         [Chromium] Unreviewed build fix. Do not define WTF/JS_EXPORT_PRIVATE to JS_EXPORTDATA
28153         until the port begins using those export macros.
28154
28155         * DumpRenderTree/chromium/config.h:
28156
28157 2011-05-25  Sheriff Bot  <webkit.review.bot@gmail.com>
28158
28159         Unreviewed, rolling out r87257.
28160         http://trac.webkit.org/changeset/87257
28161         https://bugs.webkit.org/show_bug.cgi?id=61457
28162
28163         This broke layout tests, see bug 61431. (Requested by
28164         hwennborg on #webkit).
28165
28166         * DumpRenderTree/chromium/LayoutTestController.cpp:
28167         (LayoutTestController::LayoutTestController):
28168         (LayoutTestController::clearAllDatabases):
28169         * DumpRenderTree/chromium/LayoutTestController.h:
28170
28171 2011-05-25  Qi Zhang  <qi.2.zhang@nokia.com>
28172
28173         Reviewed by Andreas Kling.
28174
28175         [Qt] fast/css/disabled-author-styles.html failed
28176         https://bugs.webkit.org/show_bug.cgi?id=61438
28177
28178         Implemented API setAuthorAndUserStylesEnabled in LayoutTestControllerQt.
28179
28180         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
28181         (WebCore::WebPage::resetSettings):
28182         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
28183         (LayoutTestController::setAuthorAndUserStylesEnabled):
28184         * DumpRenderTree/qt/LayoutTestControllerQt.h:
28185
28186 2011-05-25  Adam Roben  <aroben@apple.com>
28187
28188         Identify, rather than skip, builds where run-webkit-tests exited early due to too many failures
28189
28190         Fixes <http://webkit.org/b/61441> TestFailures page should show when run-webkit-tests
28191         started exiting early due to too many crashes, timeouts, or failures
28192
28193         Reviewed by David Kilzer.
28194
28195         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
28196         (Builder.prototype.getNumberOfFailingTests): Rather than returning -1 ("unknown") when
28197         run-webkit-tests exits early due to too many failures, pass that information in another
28198         argument to the callback.
28199         (Builder.prototype.startFetchingBuildHistory): Updated the documentation comment to reflect
28200         the new structure of the object passed to the callback.
28201         (Builder.prototype._getFailingTests): Updated to expect a tooManyFailures boolean from
28202         getNumberOfFailingTests and to pass that along to our own callbacks.
28203         (Builder.prototype._incorporateBuildHistory): Updated to expect a tooManyFailures boolean
28204         from _getFailingTests and to store that value in the history object.
28205
28206         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
28207         (.info): Added.
28208
28209         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
28210         (ViewController.prototype._displayBuilder): Updated for change in structure to the history
28211         object and to add a note when run-webkit-tests exited early due to too many failures.
28212         (ViewController.prototype._displayTesters): Renamed testersAndFailureCounts to
28213         latestBuildInfos since each entry now contains more than just the tester and failure count.
28214         Now displays a message for testers where the latest build exited early due to too many
28215         failures. Updated to expect a tooManyFailures boolean from getNumberOfFailingTests and to
28216         store that value in latestBuildInfos.
28217
28218 2011-05-24  Keishi Hattori  <keishi@webkit.org>
28219
28220         Reviewed by Kent Tamura.
28221
28222         Disable textfield implementation of <input type=color>. Add INPUT_COLOR feature flag. Add input color sanitizer.
28223         https://bugs.webkit.org/show_bug.cgi?id=61273
28224
28225         * Scripts/build-webkit: Added INPUT_COLOR feature flag.
28226
28227 2011-05-24  Greg Simon  <gregsimon@chromium.org>
28228
28229         Reviewed by Dimitri Glazkov.
28230
28231         Control Indexeddb backends from LayoutTestController
28232         https://bugs.webkit.org/show_bug.cgi?id=61000
28233
28234         * DumpRenderTree/chromium/LayoutTestController.cpp:
28235         (LayoutTestController::LayoutTestController):
28236         (LayoutTestController::setOverrideIndexedDBBackingStore):
28237         (LayoutTestController::clearAllDatabases):
28238         * DumpRenderTree/chromium/LayoutTestController.h:
28239
28240 2011-05-24  Robert Hogan  <robert@webkit.org>
28241
28242         Reviewed by Ryosuke Niwa.
28243
28244         [Qt] Fix fast/events/selectstart-by-single-click-with-shift.html
28245         https://bugs.webkit.org/show_bug.cgi?id=61393
28246
28247         * DumpRenderTree/qt/EventSenderQt.h:
28248         * DumpRenderTree/qt/EventSenderQt.cpp:
28249         (getModifiers): static function for interpreting modifiers
28250         (EventSender::mouseDown): Respect keyboard modifiers
28251         (EventSender::keyDown): use new static function
28252         (EventSender::sendOrQueueEvent): whitespace fix
28253
28254 2011-05-24  Jay Civelli  <jcivelli@chromium.org>
28255
28256         Reviewed by Adam Barth.
28257
28258         Adding MHTML reading support.
28259         https://bugs.webkit.org/show_bug.cgi?id=7168
28260
28261         * Scripts/build-webkit:
28262         * Scripts/old-run-webkit-tests:
28263         * Scripts/webkitperl/features.pm:
28264         * Scripts/webkitpy/layout_tests/port/test_files.py:
28265         * Scripts/webkitpy/layout_tests/port/webkit.py:
28266
28267 2011-05-24  Stephen White  <senorblanco@chromium.org>
28268
28269         Reviewed by Kenneth Russell.
28270
28271         Remove most style checks on Source/ThirdParty/glu, since it doesn't
28272         follow WebKit style.
28273         https://bugs.webkit.org/show_bug.cgi?id=61365
28274
28275         * Scripts/webkitpy/style/checker.py:
28276
28277 2011-05-24  Adam Roben  <aroben@apple.com>
28278
28279         Don't allow forcing builds through the build.webkit.org web interface
28280
28281         Someone has been forcing bogus builds.
28282
28283         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
28284
28285 2011-05-24  Qi Zhang  <qi.2.zhang@nokia.com>
28286
28287         Reviewed by Adam Roben.
28288
28289         WebKitTestRunner needs layoutTestController.setJavaScriptCanAccessClipboard
28290         https://bugs.webkit.org/show_bug.cgi?id=42669
28291
28292         Fixes the regression issue caused by patch 94265. The patch didn't intialize the setting before the
28293         layouttest starting, it will make the test cases after editing/execCommand/clipboard-access.html failed.
28294         Because that case changed the setting.
28295
28296         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
28297         (WTR::InjectedBundle::beginTesting):
28298
28299 2011-05-24  Csaba Osztrogonác  <ossy@webkit.org>
28300
28301         Reviewed by Kenneth Rohde Christiansen.
28302
28303         [ORWT]Make --skipped=only option work with parameters with LayoutTests/ prefix
28304         https://bugs.webkit.org/show_bug.cgi?id=61052
28305
28306         * Scripts/old-run-webkit-tests:
28307
28308 2011-05-24  Adam Roben  <aroben@apple.com>
28309
28310         Make TestFailures show how many tests are failing on each tester, and omit testers with no failures
28311
28312         Fixes <http://webkit.org/b/61063> <rdar://problem/9460533> TestFailures page shows testers
28313         that don't have any failing tests, which isn't useful
28314
28315         Reviewed by David Kilzer.
28316
28317         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js:
28318         (Buildbot.prototype.getTesters): Renamed from getTesterNames. Now returns Builder objects
28319         instead of name strings.
28320         (Buildbot.prototype._buildersForNames): Added. Helper function to convert an array of
28321         builder names into an array of builders.
28322
28323         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
28324         (Builder.prototype.getMostRecentCompletedBuildNumber): Added. Returns the build number of
28325         the most recently completed build, or -1 if there is no such build.
28326         (Builder.prototype.getNumberOfFailingTests): Added. Returns the number of tests that failed
28327         in the given build, or -1 if the number could not be determined. Some of this code came from
28328         _getFailingTests.
28329         (Builder.prototype._getBuildJSON): Added. Code came from _getFailingTests.
28330         (Builder.prototype._getFailingTests): Changed to use new _getBuildJSON and
28331         getNumberOfFailingTests functions.
28332
28333         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
28334         (ViewController.prototype._displayTesters): Get the current number of test failures for each
28335         tester and show it in the list. Omit testers that have no failures at all. We keep the
28336         testers and failure counts in an array and sort it before displaying the current data, as
28337         the order in which data will be fetched is unpredictable.
28338
28339 2011-05-23  Tony Chang  <tony@chromium.org>
28340
28341         Reviewed by Ojan Vafai.
28342
28343         [nrwt] don't use NetworkTranslation when uploading results
28344         https://bugs.webkit.org/show_bug.cgi?id=61315
28345
28346         This avoids autoinstalling mechanize.  It didn't matter anyway
28347         because test_results_uploader.py wasn't using mechanize to make the
28348         request so NetworkTransaction::run wasn't doing anything extra
28349         (it only matters if it catches mechanize.HTTPError).
28350
28351         * Scripts/webkitpy/layout_tests/layout_package/test_results_uploader.py:
28352
28353 2011-05-23  Tony Chang  <tony@chromium.org>
28354
28355         Reviewed by Dirk Pranke.
28356
28357         [chromium] fix the rebaseline tool to work with Linux GPU failures
28358         https://bugs.webkit.org/show_bug.cgi?id=61312
28359
28360         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
28361
28362 2011-05-23  Darin Adler  <darin@apple.com>
28363
28364         Ignore generated .pyc files in a directory full of python code.
28365
28366         * Scripts/webkitpy/common/checkout/scm: Added property svn:ignore.
28367
28368 2011-05-23  Adam Barth  <abarth@webkit.org>
28369
28370         Reviewed by Eric Seidel.
28371
28372         sheriffbot should give more details about the failures in IRC
28373         https://bugs.webkit.org/show_bug.cgi?id=61233
28374
28375         With this patch, sheriffbot will annonce the set of failing tests,
28376         which might help folks triage the problem.
28377
28378         * Scripts/webkitpy/tool/commands/sheriffbot.py:
28379         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
28380
28381 2011-05-23  Sheriff Bot  <webkit.review.bot@gmail.com>
28382
28383         Unreviewed, rolling out r87078.
28384         http://trac.webkit.org/changeset/87078
28385         https://bugs.webkit.org/show_bug.cgi?id=61294
28386
28387         Patch contains errors and seems likely to break a bunch of
28388         bots (Requested by abarth on #webkit).
28389
28390         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
28391         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
28392         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
28393
28394 2011-05-23  Alice Boxhall  <aboxhall@chromium.org>
28395
28396         Reviewed by Ojan Vafai.
28397
28398         Convert json_results_generator.py to output version 4 JSON.
28399         https://bugs.webkit.org/show_bug.cgi?id=60869
28400
28401         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
28402         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
28403         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
28404
28405 2011-05-23  Patrick Gansterer  <paroga@webkit.org>
28406
28407         Reviewed by Adam Barth.
28408
28409         Move generation of shell_command from port to executive
28410         https://bugs.webkit.org/show_bug.cgi?id=61251
28411
28412         After move it's possible to use the code in other places too.
28413
28414         * Scripts/webkitpy/common/config/ports.py:
28415         * Scripts/webkitpy/common/system/executive.py:
28416
28417 2011-05-23  Qi Zhang  <qi.2.zhang@nokia.com>
28418
28419         Reviewed by Darin Adler.
28420
28421         WebKitTestRunner needs layoutTestController.setJavaScriptCanAccessClipboard
28422         https://bugs.webkit.org/show_bug.cgi?id=42669
28423
28424         Implement setJavaScriptCanAccessClipboard in WebKitTestRunner.
28425
28426         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
28427         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
28428         (WTR::LayoutTestController::setJavaScriptCanAccessClipboard):
28429         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
28430
28431 2011-05-22  Robert Hogan  <robert@webkit.org>
28432
28433         Reviewed by Kenneth Rohde Christiansen.
28434
28435         Fix policyDelegate in Qt DRT
28436         https://bugs.webkit.org/show_bug.cgi?id=61247
28437
28438         Use the 'policy delegate' implemented in FrameLoaderClient
28439         for layout tests. The partial implementation in DumpRenderTreeQt
28440         is redundant, so remove it.
28441
28442         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
28443         (WebCore::WebPage::acceptNavigationRequest):
28444         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
28445         (LayoutTestController::waitForPolicyDelegate):
28446
28447 2011-05-22  Patrick Gansterer  <paroga@webkit.org>
28448
28449         Reviewed by Darin Adler.
28450
28451         Use double quotes for filename arguments
28452         https://bugs.webkit.org/show_bug.cgi?id=61250
28453
28454         Windows does not support single quotes on the command line,
28455         so replace them with double quotes.
28456
28457         * Scripts/prepare-ChangeLog:
28458
28459 2011-05-20  Simon Fraser  <simon.fraser@apple.com>
28460
28461         Reviewed by Sam Weinig.
28462
28463         WebKitTestRunner needs layoutTestController.pauseTransitionAtTimeOnElementWithId
28464         https://bugs.webkit.org/show_bug.cgi?id=42550
28465
28466         Implement pauseTransitionAtTimeOnElementWithId in WebKitTestRunner.
28467
28468         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
28469         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
28470         Copy the code for pausing animations, FIXME comments and all.
28471         (WTR::LayoutTestController::pauseTransitionAtTimeOnElementWithId):
28472         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
28473
28474 2011-05-20  Alok Priyadarshi  <alokp@chromium.org>
28475
28476         Reviewed by Stephen White.
28477
28478         Adding myself to the committers list.
28479
28480         * Scripts/webkitpy/common/config/committers.py:
28481
28482 2011-05-19  Adam Roben  <aroben@apple.com>
28483
28484         Test that the WebKit2 UI process doesn't crash when starting a download
28485
28486         Test for <http://webkit.org/b/61142> <rdar://problem/9471680> REGRESSION (r86812): Crash
28487         (preceded by assertion) in fastMalloc when downloading a file
28488
28489         Reviewed by Darin Adler.
28490
28491         * TestWebKitAPI/Tests/WebKit2/18-characters.html: Added.
28492
28493         * TestWebKitAPI/Tests/WebKit2/DownloadDecideDestinationCrash.cpp: Added.
28494         (TestWebKitAPI::decidePolicyForNavigationAction): Start a download.
28495         (TestWebKitAPI::decideDestinationWithSuggestedFilename): Record that the download was
28496         started, cancel the download, and return a bogus string.
28497
28498         (TestWebKitAPI::setContextDownloadClient):
28499         (TestWebKitAPI::setPagePolicyClient):
28500         Simple helper functions.
28501
28502         (TestWebKitAPI::TEST): Load 18-characters.html, which should trigger a download thanks to
28503         our policy client, and run until we know that the download was started. If we haven't
28504         crashed, we win!
28505
28506         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
28507         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
28508         * TestWebKitAPI/win/copy-resources.cmd:
28509         Added new files.
28510
28511 2011-05-20  Kent Tamura  <tkent@chromium.org>
28512
28513         Reviewed by Ryosuke Niwa.
28514
28515         Fix style errors in DumpRenderTree/chromium/.
28516         https://bugs.webkit.org/show_bug.cgi?id=61172
28517
28518         * DumpRenderTree/chromium/CppBoundClass.cpp:
28519         (CppBoundClass::getAsCppVariant):
28520         (CppBoundClass::bindToJavascript):
28521         * DumpRenderTree/chromium/CppBoundClass.h:
28522         (CppBoundClass::GetterCallback::~GetterCallback):
28523         (CppBoundClass::CppBoundClass):
28524         (CppBoundClass::Callback::~Callback):
28525         (CppBoundClass::MemberCallback::MemberCallback):
28526         (CppBoundClass::MemberCallback::~MemberCallback):
28527         (CppBoundClass::MemberGetterCallback::MemberGetterCallback):
28528         (CppBoundClass::MemberGetterCallback::~MemberGetterCallback):
28529         * DumpRenderTree/chromium/CppVariant.h:
28530         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
28531         (DRTDevToolsAgent::~DRTDevToolsAgent):
28532         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
28533         (DRTDevToolsClient::sendFrontendLoaded):
28534         * DumpRenderTree/chromium/DRTDevToolsClient.h:
28535         * DumpRenderTree/chromium/DumpRenderTree.cpp:
28536         (main):
28537         * DumpRenderTree/chromium/EventSender.cpp:
28538         (SavedEvent::SavedEvent):
28539         (EventSender::EventSender):
28540         (EventSender::keyDown):
28541         * DumpRenderTree/chromium/EventSender.h:
28542         * DumpRenderTree/chromium/ImageDiff.cpp:
28543         (Image::Image):
28544         * DumpRenderTree/chromium/LayoutTestController.cpp:
28545         (LayoutTestController::LayoutTestController):
28546         (LayoutTestController::WorkQueue::reset):
28547         (WorkItemBackForward::WorkItemBackForward):
28548         (WorkItemLoadingScript::WorkItemLoadingScript):
28549         (WorkItemNonLoadingScript::WorkItemNonLoadingScript):
28550         (WorkItemLoad::WorkItemLoad):
28551         (WorkItemLoadHTMLString::WorkItemLoadHTMLString):
28552         (LayoutTestController::pathToLocalResource):
28553         * DumpRenderTree/chromium/LayoutTestController.h:
28554         (LayoutTestController::WorkItem::~WorkItem):
28555         (LayoutTestController::WorkQueue::WorkQueue):
28556         * DumpRenderTree/chromium/MockSpellCheck.cpp:
28557         (MockSpellCheck::MockSpellCheck):
28558         (MockSpellCheck::~MockSpellCheck):
28559         * DumpRenderTree/chromium/MockSpellCheck.h:
28560         * DumpRenderTree/chromium/NotificationPresenter.h:
28561         (NotificationPresenter::NotificationPresenter):
28562         * DumpRenderTree/chromium/Task.cpp:
28563         (WebTask::WebTask):
28564         * DumpRenderTree/chromium/Task.h:
28565         (TaskList::TaskList):
28566         * DumpRenderTree/chromium/TestEventPrinter.cpp:
28567         (DRTPrinter::DRTPrinter):
28568         (TestShellPrinter::TestShellPrinter):
28569         * DumpRenderTree/chromium/TestNavigationController.cpp:
28570         (TestNavigationEntry::TestNavigationEntry):
28571         (TestNavigationEntry::~TestNavigationEntry):
28572         (TestNavigationController::TestNavigationController):
28573         (TestNavigationController::reload):
28574         (TestNavigationController::loadEntry):
28575         (TestNavigationController::didNavigateToEntry):
28576         * DumpRenderTree/chromium/TestNavigationController.h:
28577         (TestShellExtraData::TestShellExtraData):
28578         * DumpRenderTree/chromium/TestShell.cpp:
28579         (dumpHistoryItem):
28580         * DumpRenderTree/chromium/TestShell.h:
28581         (TestParams::TestParams):
28582         * DumpRenderTree/chromium/TestShellWin.cpp:
28583         (TestShell::waitTestFinished):
28584         * DumpRenderTree/chromium/TestWebWorker.h:
28585         (TestWebWorker::startWorkerContext):
28586         (TestWebWorker::terminateWorkerContext):
28587         (TestWebWorker::postMessageToWorkerContext):
28588         (TestWebWorker::clientDestroyed):
28589         (TestWebWorker::postMessageToWorkerObject):
28590         (TestWebWorker::postExceptionToWorkerObject):
28591         (TestWebWorker::postConsoleMessageToWorkerObject):
28592         (TestWebWorker::confirmMessageFromWorkerObject):
28593         (TestWebWorker::reportPendingActivity):
28594         (TestWebWorker::workerContextClosed):
28595         (TestWebWorker::~TestWebWorker):
28596         * DumpRenderTree/chromium/TextInputController.cpp:
28597         * DumpRenderTree/chromium/WebThemeControlDRTWin.h:
28598         * DumpRenderTree/chromium/WebThemeEngineDRTMac.h:
28599         * DumpRenderTree/chromium/WebThemeEngineDRTWin.cpp:
28600         (WebThemeEngineDRTWin::paintButton):
28601         (WebThemeEngineDRTWin::paintMenuList):
28602         (WebThemeEngineDRTWin::paintTrackbar):
28603         * DumpRenderTree/chromium/WebThemeEngineDRTWin.h:
28604         (WebThemeEngineDRTWin::WebThemeEngineDRTWin):
28605         * DumpRenderTree/chromium/WebViewHost.cpp:
28606         (WebViewHost::navigate):
28607         (WebViewHost::updateSessionHistory):
28608         (WebViewHost::paintInvalidatedRegion):
28609         * DumpRenderTree/chromium/WebViewHost.h:
28610
28611 2011-05-19  Dmitry Lomov  <dslomov@google.com>
28612
28613         Reviewed by Adam Roben.
28614
28615         Detect hangs in run-api-tests
28616         https://bugs.webkit.org/show_bug.cgi?id=48043
28617
28618         * Scripts/run-api-tests: Added test timeouts
28619
28620 2011-05-19  Dmitry Lomov  <dslomov@google.com>
28621
28622         Reviewed by Adam Roben.
28623
28624         run-api-tests should run one test per process
28625         https://bugs.webkit.org/show_bug.cgi?id=61088
28626
28627         * Scripts/run-api-tests: Resurrecting the previous revison of this file, with fixes to system call under Windows,
28628         return code, and parsing GTest output format.
28629
28630 2011-05-15  Robert Hogan  <robert@webkit.org>
28631
28632         Reviewed by Antonio Gomes.
28633
28634         [Qt][GTK] plugins/get-url-with-javascript-url.html fails
28635         https://bugs.webkit.org/show_bug.cgi?id=60834
28636         Fix unix test plugin for plugins/get-url-with-javascript-url.html
28637
28638         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
28639         (webkit_test_plugin_new_stream):
28640         (webkit_test_plugin_write_ready):
28641         (webkit_test_plugin_write):
28642
28643 2011-05-19  Sheriff Bot  <webkit.review.bot@gmail.com>
28644
28645         Unreviewed, rolling out r86869, r86873, r86875, and r86877.
28646         http://trac.webkit.org/changeset/86869
28647         http://trac.webkit.org/changeset/86873
28648         http://trac.webkit.org/changeset/86875
28649         http://trac.webkit.org/changeset/86877
28650         https://bugs.webkit.org/show_bug.cgi?id=61139
28651
28652         broke builds and debug DRT (Requested by rniwa on #webkit).
28653
28654         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
28655         * DumpRenderTree/chromium/TestShell.cpp:
28656         (TestShell::bindJSObjectsToWindow):
28657         * DumpRenderTree/mac/FrameLoadDelegate.mm:
28658         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]):
28659
28660 2011-05-18  Dimitri Glazkov  <dglazkov@chromium.org>
28661
28662         Reviewed by Darin Adler.
28663
28664         Add build logistics and plumbing for window.internals object.
28665         https://bugs.webkit.org/show_bug.cgi?id=60313
28666
28667         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj: Added linking new WebCoreTestSupport library.
28668         * DumpRenderTree/chromium/TestShell.cpp:
28669         (TestShell::bindJSObjectsToWindow): Added injection code.
28670         * DumpRenderTree/mac/FrameLoadDelegate.mm:
28671         (-[FrameLoadDelegate didClearWindowObjectInStandardWorldForFrame:]): Ditto.
28672
28673 2011-05-19  Zsolt Fehér  <h490267@stud.u-szeged.hu>
28674
28675         Reviewed by Csaba Osztrogonác.
28676
28677         [Qt] Implement eventSender.scalePageBy
28678         https://bugs.webkit.org/show_bug.cgi?id=60015
28679
28680         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
28681         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
28682         * DumpRenderTree/qt/EventSenderQt.cpp:
28683         (EventSender::scalePageBy):
28684         * DumpRenderTree/qt/EventSenderQt.h:
28685
28686 2011-05-19  Kenichi Ishibashi  <bashi@chromium.org>
28687
28688         Reviewed by Kent Tamura.
28689
28690         [Chromium] IME candidate window appears wrong position in an iframe
28691         https://bugs.webkit.org/show_bug.cgi?id=61023
28692
28693         Call the focused frame's firstRectForCharacterRange() instead of the
28694         main frame so that DRT behaves as the same as Chromium.
28695
28696         * DumpRenderTree/chromium/TextInputController.cpp:
28697         (TextInputController::firstRectForCharacterRange):
28698
28699 2011-05-18  Adam Roben  <aroben@apple.com>
28700
28701         Add a new page to build.webkit.org to help find when tests started failing
28702
28703         The page is accessible at <http://build.webkit.org/TestFailures/>. It is pretty minimalist
28704         right now, but already shows some useful information. It's somewhat similar to webkit-patch
28705         failure-reason and sheriffbot, and perhaps can be combined with them eventually. It's a
28706         little more convenient than either of them, though, because it's all done in the browser
28707         (and thus it's easy to go directly to the relevant test results).
28708
28709         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Buildbot.js: Added.
28710         (Buildbot): This class represents a Buildbot server.
28711         (Buildbot.prototype.buildURL): Returns the URL for the summary page for a particular build.
28712         (Buildbot.prototype.builderNamed): Returns a Builder with the given name.
28713         (Buildbot.prototype.getTesterNames): Fetches the names of all testers and passes them to the
28714         callback.
28715         (Buildbot.prototype.parseBuildName): Breaks up a build name into its constituent parts. Must
28716         be implemented by a derived class that understands this server's build naming scheme.
28717         (Buildbot.prototype.resultsDirectoryURL): Returns the URL for the results directory for a
28718         particular build.
28719
28720         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js: Added.
28721         (Builder): This class represents one builder on the buildbot.
28722         (Builder.prototype.buildURL): Returns the URL for the summary page for a particular build.
28723         (Builder.prototype.failureDiagnosisTextAndURL): Returns data that provides a little more
28724         information about a particular test failure.
28725         (Builder.prototype.startFetchingBuildHistory): Periodically calls the callback with
28726         information about when tests started failing.
28727         (Builder.prototype.resultsDirectoryURL): Returns the URL for the results directory for a
28728         particular build.
28729         (Builder.prototype._getBuildNames): Fetches the names of all builds and passes them to the
28730         callback.
28731         (Builder.prototype._getFailingTests): Fetches the results.html page for the given build and
28732         extracts all the failing tests listed in it, passing them to the callback.
28733         (Builder.prototype._incorporateBuildHistory): Gets the failing tests for the specified
28734         build, merges them into the build history, and calls the callback telling it whether the
28735         next build should be fetched to provide more information.
28736
28737         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
28738         Added. Just some simple styles.
28739
28740         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js: Added.
28741         (createDefinitionList): Takes an array of pairs and turns them into a DL element.
28742         (getResource): Wrapper around XMLHttpRequest.
28743         (Array.prototype.findFirst): Finds the first element matching the given predicate and
28744         returns it.
28745         (Array.prototype.last): Returns the last element of the array.
28746
28747         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js: Added.
28748         (ViewController): This class contains the main logic for displaying the page.
28749         (ViewController.loaded): Just calls through to parseHash.
28750         (ViewController.parseHash): Either starts analyzing failures on a particular builder, or
28751         shows the list of all testers so one can be chosen. This function is called when the page
28752         loads and whenever we get a hashchange event.
28753         (ViewController._displayBuilder): Asks the builder to fetch build history, and displays it
28754         as it is fetched. The display ends up grouping tests by when they started failing.
28755         (ViewController._displayTesters): Gets the list of testers and displays it.
28756
28757         (ViewController._domForBuildName):
28758         (ViewController._domForFailedTest):
28759         Helper functions to create descriptions and links for a particular build or failed test.
28760
28761         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/WebKitBuildbot.js: Added.
28762         (WebKitBuildbot): Calls up to the base class constructor with the correct base URL.
28763         (WebKitBuildbot.prototype.parseBuildName): Parses a build.webkit.org-style build name.
28764
28765         * BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/index.html: Added. Just
28766         loads all the files and sets up a ViewController, which does the rest.
28767
28768         * BuildSlaveSupport/build.webkit.org-config/templates/root.html: Added a link to the new
28769         page.
28770
28771 2011-05-18  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
28772
28773         Reviewed by Csaba Osztrogonác.
28774
28775         [Qt] Implement layoutTestController.setValueForUser()
28776         https://bugs.webkit.org/show_bug.cgi?id=60956
28777
28778         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
28779         (LayoutTestController::setValueForUser):
28780         * DumpRenderTree/qt/LayoutTestControllerQt.h:
28781
28782 2011-05-16  Robert Hogan  <robert@webkit.org>
28783
28784         Reviewed by Kenneth Rohde Christiansen.
28785
28786         plugins/invalidate_rect.html fails on linux ports
28787
28788         - Make the unix test netscape plugin recognize the onPaintEvent
28789           and windowedPlugin parameters.
28790
28791         https://bugs.webkit.org/show_bug.cgi?id=54051
28792
28793         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
28794         (handleEventX11):
28795         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
28796         (LayoutTestController::displayInvalidatedRegion):
28797         * DumpRenderTree/qt/LayoutTestControllerQt.h:
28798         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
28799         (webkit_test_plugin_new_instance):
28800         (webkit_test_plugin_handle_event):
28801
28802 2011-05-17  Tony Chang  <tony@chromium.org>
28803
28804         Reviewed by Ojan Vafai.
28805
28806         [chromium] move Lucid 64 bit results into LayoutTests/platform/chromium-linux
28807         https://bugs.webkit.org/show_bug.cgi?id=60895
28808
28809         Update the tools to handle the move.  Lucid 32 results now go in
28810         chromium-linux-x86 and the default platform on Linux is now x86_64.
28811
28812         * Scripts/webkitpy/layout_tests/deduplicate_tests.py: Default to x86_64
28813         * Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py:
28814         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py: Default to Lucid
28815         * Scripts/webkitpy/layout_tests/port/base.py: ditto
28816         * Scripts/webkitpy/layout_tests/port/chromium_linux.py: Update directory fallback and default to Lucid 64
28817         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
28818         * Scripts/webkitpy/layout_tests/port/test.py: Update tests to default to x86_64
28819         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py: Update bot names.
28820         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py: Whitespace cleanup.
28821
28822 2011-05-17  Adam Roben  <aroben@apple.com>
28823
28824         Make run-api-tests work on Windows when there are spaces in the path
28825
28826         Fixes <http://webkit.org/b/60954> REGRESSION (r86511): run-api-tests fails if there are
28827         spaces in the path to TestWebKitAPI.exe
28828
28829         Reviewed by David Levin.
28830
28831         * Scripts/run-api-tests:
28832         (runTestTool): Use the "direct object" form of system() to avoid having the path to
28833         TestWebKitAPI.exe be split by the shell.
28834
28835 2011-05-16  Ian Henderson  <ianh@apple.com>
28836
28837         Reviewed by Joseph Pecoraro.
28838
28839         Page::goToItem doesn't work while loading is deferred
28840         https://bugs.webkit.org/show_bug.cgi?id=60412
28841
28842         Add setDefersLoading and goBack methods to LayoutTestController.  We
28843         need to use goBack() instead of history.back() because the latter goes
28844         through NavigationScheduler, hence doesn't exhibit the bug.
28845
28846         * DumpRenderTree/LayoutTestController.cpp:
28847         (goBackCallback):
28848         (setDefersLoadingCallback):
28849         (LayoutTestController::staticFunctions):
28850         * DumpRenderTree/LayoutTestController.h:
28851         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
28852         (LayoutTestController::goBack):
28853         (LayoutTestController::setDefersLoading):
28854         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
28855         (LayoutTestController::goBack):
28856         (LayoutTestController::setDefersLoading):
28857         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
28858         (LayoutTestController::goBack):
28859         (LayoutTestController::setDefersLoading):
28860         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
28861         (LayoutTestController::goBack):
28862         (LayoutTestController::setDefersLoading):
28863
28864 2011-05-16  Sam Weinig  <sam@webkit.org>
28865
28866         Reviewed by Anders Carlsson.
28867
28868         TestWebKitAPI should build with clang if it can
28869         https://bugs.webkit.org/show_bug.cgi?id=60918
28870
28871         * TestWebKitAPI/Configurations/CompilerVersion.xcconfig:
28872         Update CompilerVersion.xcconfig to match others.
28873
28874         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
28875         Add CompilerVersion.xcconfig to the project.
28876
28877         * TestWebKitAPI/Tests/WebKit2/FailedLoad.cpp:
28878         (TestWebKitAPI::didFailProvisionalLoadWithErrorForFrame):
28879         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
28880         (TestWebKitAPI::didSameDocumentNavigationForFrame):
28881         Add some casts to quiet warnings from clang.
28882
28883 2011-05-16  Dirk Pranke  <dpranke@chromium.org>
28884
28885         Reviewed by David Levin.
28886
28887         add dpranke as a reviewer
28888         https://bugs.webkit.org/show_bug.cgi?id=60919
28889
28890         * Scripts/webkitpy/common/config/committers.py:
28891
28892 2011-05-16  Robert Hogan  <robert@webkit.org>
28893
28894         Reviewed by Anders Carlsson.
28895
28896         [Gtk] plugins/get-url-notify-with-url-that-fails-to-load.html crashes on buildbot
28897
28898         https://bugs.webkit.org/show_bug.cgi?id=60838
28899
28900         The unix test plugin needs to call the test's NPP_URLNotify
28901         rather than the browsers.
28902
28903         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
28904         (webkit_test_plugin_url_notify):
28905
28906 2011-05-16  Brent Fulgham  <bfulgham@webkit.org>
28907
28908         Rubber-stamped by Adam Roben.
28909
28910         Adding myself as a reviewer.
28911         * Scripts/webkitpy/common/config/committers.py:
28912
28913 2011-05-16  Sam Weinig  <sam@webkit.org>
28914
28915         Fix typo pointed out by Dave Levin.
28916
28917         * TestWebKitAPI/PlatformUtilities.h:
28918         (TestWebKitAPI::Util::assertWKStringEqual):
28919         Strig -> String.
28920
28921 2011-05-16  Sam Weinig  <sam@webkit.org>
28922
28923         Reviewed by David Levin.
28924
28925         Convert api tester over to using gtest expectations directly
28926         https://bugs.webkit.org/show_bug.cgi?id=60862
28927
28928         * TestWebKitAPI/PlatformUtilities.cpp:
28929         (TestWebKitAPI::Util::toSTD):
28930         * TestWebKitAPI/PlatformUtilities.h:
28931         (TestWebKitAPI::Util::assertWKStrigEqual):
28932         Add convenience macro to compare WK2 strings. Add some overloads of toSTD to make
28933         the implementation of the macro simpler.
28934
28935         * TestWebKitAPI/Test.h:
28936         Remove TEST_ASSERT forwarder.
28937
28938         [Test changes elided]
28939
28940 2011-05-16  David Kilzer  <ddkilzer@apple.com>
28941
28942         <http://webkit.org/b/60913> C++ exceptions should not be enabled when building with llvm-gcc-4.2
28943         <rdar://problem/9446430>
28944
28945         Reviewed by Mark Rowe.
28946
28947         * MiniBrowser/Configurations/Base.xcconfig: Fixed typo.
28948         * TestWebKitAPI/Configurations/Base.xcconfig: Ditto.
28949         * WebKitTestRunner/Configurations/Base.xcconfig: Ditto.
28950
28951 2011-05-16  Jon Lee  <jonlee@apple.com>
28952
28953         Reviewed by Simon Fraser.
28954
28955         Can't horizontally scroll iframes and overflow because wheel events are always accepted
28956         https://bugs.webkit.org/show_bug.cgi?id=60779
28957
28958         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
28959         (WTR::InjectedBundlePage::InjectedBundlePage): Set new default method to nil.
28960
28961 2011-05-16  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
28962
28963         Reviewed by Csaba Osztrogonác.
28964
28965         [Qt] Tools.pro misses include(common.pri)
28966         https://bugs.webkit.org/show_bug.cgi?id=60883
28967
28968         It includes features.pri, which depends on common.pri (the build is
28969         not broken today with the default options, but this is the right
28970         thing to do).
28971
28972         * Tools.pro:
28973
28974 2011-05-15  Dominic Cooney  <dominicc@chromium.org>
28975
28976         Unreviewed: adding myself to committers.py.
28977
28978         * Scripts/webkitpy/common/config/committers.py:
28979
28980 2011-05-15  Daniel Bates  <dbates@webkit.org>
28981
28982         Reviewed by Chris Jerdonek.
28983
28984         svn-apply can't handle Git diff that only changes executable bit
28985         https://bugs.webkit.org/show_bug.cgi?id=60848
28986
28987         Fixes an issue where svn-{apply, unapply} fail to handle a Git diff for a file
28988         that only has an executable bit change.
28989
28990         * Scripts/VCSUtils.pm:
28991           - Modified regex $chunkRangeRegEx to match chunk range with ending text, say: @@ -2,6 +2,18 @@ foo().
28992         (parseDiff): Modified to count the number of text chunks in the diff. This should also help towards
28993         fixing <https://bugs.webkit.org/show_bug.cgi?id=29684>.
28994         * Scripts/svn-apply:
28995         (patch): Only use standard patch tool if the patch is non-empty and for a text file (i.e. not a binary patch).
28996         * Scripts/svn-unapply:
28997         (patch): Ditto.
28998         * Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl:
28999           - Updated expected results for test cases to account for numTextChunks.
29000           - Fixed malformed chunk range in test cases: "Git: simple", and "Git: Git diff followed by SVN diff".
29001           - Added test case "Git: file that only has an executable bit change".
29002         * Scripts/webkitperl/VCSUtils_unittest/parseDiffWithMockFiles.pl:
29003           - Updated expected results for test cases to account for numTextChunks.
29004
29005 2011-05-15  Sam Weinig  <sam@webkit.org>
29006
29007         Fix windows build.
29008
29009         * TestWebKitAPI/win/main.cpp:
29010         (main):
29011
29012 2011-05-15  Sam Weinig  <sam@webkit.org>
29013
29014         Reviewed by David Levin.
29015
29016         Make API tester more gtest friendly
29017         https://bugs.webkit.org/show_bug.cgi?id=60859
29018
29019         * Scripts/run-api-tests:
29020         Numerous changes:
29021         - Remove logic to run tests one at a time.
29022         - Remove logic to build up a list of tests, instead just pass --gtest_list_tests to
29023           the test runner and allow gtest to print it for us.
29024         - Forward stdout from the test tool to stdout, even in non-verbose runs.
29025         - Run test tool from the current working directory instead of the root webkit
29026           directory.
29027
29028         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
29029         Move plist to be with other bundle sources.
29030
29031         * TestWebKitAPI/TestsController.cpp:
29032         (TestWebKitAPI::TestsController::TestsController):
29033         (TestWebKitAPI::TestsController::run):
29034         * TestWebKitAPI/TestsController.h:
29035         * TestWebKitAPI/mac/main.mm:
29036         Simplify TestController down to a single run function. We should consider removing the class
29037         entirely at some point.
29038
29039 2011-05-15  Sheriff Bot  <webkit.review.bot@gmail.com>
29040
29041         Unreviewed, rolling out r86504.
29042         http://trac.webkit.org/changeset/86504
29043         https://bugs.webkit.org/show_bug.cgi?id=60853
29044
29045         "Broke Qt EventSender in editing/selection tests" (Requested
29046         by mwenge on #webkit).
29047
29048         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
29049         (WebCore::WebPage::WebPage):
29050         (WebCore::DumpRenderTree::DumpRenderTree):
29051         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
29052         (WebCore::DumpRenderTree::initJSObjects):
29053         * DumpRenderTree/qt/DumpRenderTreeQt.h:
29054         (WebCore::DumpRenderTree::eventSender):
29055         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
29056         (webkit_test_plugin_new_instance):
29057         (webkit_test_plugin_set_window):
29058         (webkit_test_plugin_handle_event):
29059
29060 2011-05-12  Robert Hogan  <robert@webkit.org>
29061
29062         Reviewed by Benjamin Poulain.
29063
29064         [Qt] fix http/tests/plugins/plugin-document-has-focus.html
29065
29066         The support for this test added to the unix test plugin here
29067         may allow other platforms to pass it (nearly everyone
29068         skips it). On Qt it required a bit of trickiness with
29069         the page's EventSender object to get it working fully though,
29070         so only unskipping Qt here.
29071
29072         https://bugs.webkit.org/show_bug.cgi?id=60722
29073
29074         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
29075         (WebCore::WebPage::WebPage):
29076         (WebCore::WebPage::eventSender):
29077         (WebCore::DumpRenderTree::DumpRenderTree):
29078         (WebCore::DumpRenderTree::initJSObjects):
29079         (WebCore::DumpRenderTree::createWindow):
29080         * DumpRenderTree/qt/DumpRenderTreeQt.h:
29081         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
29082         (webkit_test_plugin_new_instance):
29083         (webkit_test_plugin_set_window):
29084         (webkit_test_plugin_handle_event):
29085
29086 2011-05-14  Arno Renevier  <arno@renevier.net>
29087
29088         Reviewed by Martin Robinson.
29089
29090         make run-gtk-tests executable
29091         https://bugs.webkit.org/show_bug.cgi?id=59278
29092
29093         * Scripts/run-gtk-tests:
29094
29095 2011-05-14  Anders Carlsson  <andersca@apple.com>
29096
29097         Reviewed by Kevin Decker.
29098
29099         NPP_URLNotify is not called if a URL passed to NPN_GetURLNotify fails to load
29100         https://bugs.webkit.org/show_bug.cgi?id=60823
29101         <rdar://problem/9430386>
29102
29103         Add test that calls NPN_GetURLNotify on a URL that can't be loaded, and checks if NPP_URLNotify is called.
29104
29105         * DumpRenderTree/DumpRenderTree.gypi:
29106         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
29107         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Added.
29108         (GetURLNotifyWithURLThatFailsToLoad::GetURLNotifyWithURLThatFailsToLoad):
29109         (GetURLNotifyWithURLThatFailsToLoad::NPP_New):
29110         (GetURLNotifyWithURLThatFailsToLoad::NPP_URLNotify):
29111         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
29112         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
29113         * GNUmakefile.am:
29114
29115 2011-05-14  Sheriff Bot  <webkit.review.bot@gmail.com>
29116
29117         Unreviewed, rolling out r86476.
29118         http://trac.webkit.org/changeset/86476
29119         https://bugs.webkit.org/show_bug.cgi?id=60832
29120
29121         It made plugins/get-url-notify-with-url-that-fails-to-
29122         load.html crash on GTK (Requested by Ossy on #webkit).
29123
29124         * DumpRenderTree/DumpRenderTree.gypi:
29125         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
29126         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Removed.
29127         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
29128         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
29129         * GNUmakefile.am:
29130
29131 2011-05-13  Anders Carlsson  <andersca@apple.com>
29132
29133         Reviewed by Sam Weinig.
29134
29135         NPP_URLNotify is not called if a URL passed to NPN_GetURLNotify fails to load
29136         https://bugs.webkit.org/show_bug.cgi?id=60823
29137         <rdar://problem/9430386>
29138
29139         Add test that calls NPN_GetURLNotify on a URL that can't be loaded, and checks if NPP_URLNotify is called.
29140
29141         * DumpRenderTree/DumpRenderTree.gypi:
29142         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
29143         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLNotifyWithURLThatFailsToLoad.cpp: Added.
29144         (GetURLNotifyWithURLThatFailsToLoad::GetURLNotifyWithURLThatFailsToLoad):
29145         (GetURLNotifyWithURLThatFailsToLoad::NPP_New):
29146         (GetURLNotifyWithURLThatFailsToLoad::NPP_URLNotify):
29147         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
29148         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
29149         * GNUmakefile.am:
29150
29151 2011-05-13  Anders Carlsson  <andersca@apple.com>
29152
29153         Build fix.
29154
29155         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURL.cpp:
29156
29157 2011-05-13  Anders Carlsson  <andersca@apple.com>
29158
29159         Reviewed by Sam Weinig.
29160
29161         NPN_GetURL with a javascript: URL doesn't send back the result in the stream
29162         https://bugs.webkit.org/show_bug.cgi?id=60810
29163
29164         Add a test that calls NPN_GetURL on a javascript: URL and checks that it gets back the expected result.
29165
29166         * DumpRenderTree/DumpRenderTree.gypi:
29167         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
29168         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
29169         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
29170         * GNUmakefile.am:
29171         Add files.
29172
29173         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
29174         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
29175         Add new NPN and NPP wrappers and sort them according to the order they appear in npfunctions.h
29176
29177         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURL.cpp: Added.
29178         (GetURLWithJavaScriptURL::GetURLWithJavaScriptURL):
29179         (GetURLWithJavaScriptURL::NPP_New):
29180         (GetURLWithJavaScriptURL::NPP_NewStream):
29181         (GetURLWithJavaScriptURL::NPP_DestroyStream):
29182         (GetURLWithJavaScriptURL::NPP_WriteReady):
29183         (GetURLWithJavaScriptURL::NPP_Write):
29184         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
29185         (NPP_NewStream):
29186         (NPP_WriteReady):
29187         (NPP_Write):
29188         (NPP_URLNotify):
29189         call the PluginTest NPP wrappers.
29190
29191 2011-05-13  Sheriff Bot  <webkit.review.bot@gmail.com>
29192
29193         Unreviewed, rolling out r86447.
29194         http://trac.webkit.org/changeset/86447
29195         https://bugs.webkit.org/show_bug.cgi?id=60809
29196
29197         "Broke some uses of EventSender object on Qt" (Requested by
29198         mwenge on #webkit).
29199
29200         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
29201         (WebCore::WebPage::WebPage):
29202         (WebCore::DumpRenderTree::DumpRenderTree):
29203         (WebCore::DumpRenderTree::initJSObjects):
29204         (WebCore::DumpRenderTree::createWindow):
29205         * DumpRenderTree/qt/DumpRenderTreeQt.h:
29206         (WebCore::DumpRenderTree::eventSender):
29207         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
29208         (webkit_test_plugin_new_instance):
29209         (webkit_test_plugin_set_window):
29210         (webkit_test_plugin_handle_event):
29211
29212 2011-05-13  Robert Hogan  <robert@webkit.org>
29213
29214         Reviewed by Benjamin Poulain.
29215
29216         [Qt] fix http/tests/plugins/plugin-document-has-focus.html
29217
29218         The support for this test added to the unix test plugin here
29219         may allow other platforms to pass it (nearly everyone
29220         skips it). On Qt it required a bit of trickiness with
29221         the page's EventSender object to get it working fully though,
29222         so only unskipping Qt here.
29223
29224         https://bugs.webkit.org/show_bug.cgi?id=60722
29225
29226         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
29227         (WebCore::WebPage::WebPage):
29228         (WebCore::WebPage::eventSender):
29229         (WebCore::DumpRenderTree::DumpRenderTree):
29230         (WebCore::DumpRenderTree::initJSObjects):
29231         (WebCore::DumpRenderTree::createWindow):
29232         * DumpRenderTree/qt/DumpRenderTreeQt.h:
29233         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
29234         (webkit_test_plugin_new_instance):
29235         (webkit_test_plugin_set_window):
29236         (webkit_test_plugin_handle_event):
29237
29238 2011-05-13  Adam Roben  <aroben@apple.com>
29239
29240         Exclude the Leopard NRWT bot from the core builder set
29241
29242         Fixes <http://webkit.org/b/60392> Leopard Intel Release (NRWT) shouldn't be a core builder
29243
29244         Reviewed by Darin Adler.
29245
29246         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
29247         (BuildBot.__init__): Tighten up the Leopard regexp to only allow the non-NRWT Leopard bots
29248         in the core builder set.
29249
29250         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
29251         (BuildBotTest.test_builder_name_regexps): Updated to match the current set of builders and
29252         regexps.
29253
29254 2011-05-13  Adam Roben  <aroben@apple.com>
29255
29256         Group all the NRWT bots together
29257
29258         This also gets the new NRWT bots out of the middle of Apple's bots. Since these new NRWT
29259         bots are experimental, it seems best not to clutter up the non-experimental bots with
29260         experimental ones.
29261
29262         Fixes <http://webkit.org/b/60769> New "(NRWT)" bots are right in the middle of Apple's bots
29263         on build.webkit.org
29264
29265         Reviewed by Dan Bernstein.
29266
29267         * BuildSlaveSupport/build.webkit.org-config/config.json: Moved the "(NRWT)" bots next to the
29268         Chromium testers, which also use new-run-webkit-tests.
29269
29270 2011-05-13  Adam Roben  <aroben@apple.com>
29271
29272         Add a plugin test that dumps the plugin window's rect
29273
29274         Part of the test for <http://webkit.org/b/60194> <rdar://problem/9383760> REGRESSION
29275         (r79167): Windowed plugins in Google Reader don't move when the article list is scrolled
29276
29277         Reviewed by Dan Bernstein.
29278
29279         * DumpRenderTree/TestNetscapePlugIn/Tests/win/DumpWindowRect.cpp: Added.
29280         (DumpWindowRect::DumpWindowRect): Just call up to the base class.
29281         (DumpWindowRect::performWindowGeometryTest): Find our window rect relative to the test
29282         harness window and log it.
29283
29284         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added DumpWindowRect.
29285
29286 2011-05-13  Adam Roben  <aroben@apple.com>
29287
29288         Move code required for most tests of plugin window geometry into a base class
29289
29290         WebKit2's asynchronous plugin window positioning makes testing window geometry tricky. By
29291         having all the trickiness in a shared base class, writing new window geometry tests will be
29292         much easier.
29293
29294         Fixes <http://webkit.org/b/60776> It's hard to write tests that check plugin window geometry
29295
29296         Reviewed by Anders Carlsson.
29297
29298         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowRegionIsSetToClipRect.cpp:
29299         (WindowRegionIsSetToClipRect::WindowRegionIsSetToClipRect): Changed to inherit from the new
29300         WindowGeometryTest base class. Most of the code from this class was moved to the new base
29301         class.
29302         (WindowRegionIsSetToClipRect::performWindowGeometryTest): Renamed from checkWindowRegion.
29303         This is called by the base class when it is safe to check the window's geometry. Note that
29304         the m_didCheckWindowRegion logic was removed completely; it has never been needed and was
29305         just vestigial code from an earlier version of this test.
29306
29307         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added new
29308         WindowGeometryTest class.
29309
29310         * DumpRenderTree/TestNetscapePlugIn/win/WindowGeometryTest.cpp: Added. Code came from
29311         WindowRegionIsSetToClipRect.
29312         (WindowGeometryTest::finishTest): This is the only change from WindowRegionIsSetToClipRect.
29313         We now call the new virtual performWindowGeometryTest function instead of checkWindowRegion.
29314         Subclasses can implement that function to perform their own tests.
29315
29316         * DumpRenderTree/TestNetscapePlugIn/win/WindowGeometryTest.h: Added.
29317
29318 2011-05-13  Zoltan Herczeg  <zherczeg@webkit.org>
29319
29320         Reviewed by Csaba Osztrogonác.
29321
29322         Missing 2nd #EOF if WebProcess is killed.
29323         https://bugs.webkit.org/show_bug.cgi?id=60761
29324
29325         If the WebProcess is killed, only one #EOF is yielded
29326         by WebKitTestRunner. However, run-webkit-test wait for
29327         a second #EOF, which signals the finishing of pixel dump.
29328         Since this never happens, run-webkit-test waits forever.
29329
29330         * WebKitTestRunner/TestInvocation.cpp:
29331         (WTR::TestInvocation::dump):
29332         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
29333         * WebKitTestRunner/TestInvocation.h:
29334
29335 2011-05-13  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
29336
29337         Reviewed by Csaba Osztrogonác.
29338
29339         [NRWT] chromium-linux tests shouldn't run on other platforms
29340         https://bugs.webkit.org/show_bug.cgi?id=60162
29341
29342         Remove _tests_for_other_platforms defs because r82753
29343         implemented it properly in the baseclass in webkit.py.
29344
29345         * Scripts/webkitpy/layout_tests/port/gtk.py:
29346         * Scripts/webkitpy/layout_tests/port/qt.py:
29347         * Scripts/webkitpy/layout_tests/port/win.py:
29348
29349 2011-05-12  Ryuan Choi  <ryuan.choi@samsung.com>
29350
29351         Reviewed by Antonio Gomes.
29352
29353         [EFL] Remove unused option of EWeblauncher
29354         https://bugs.webkit.org/show_bug.cgi?id=60747
29355
29356         Remove sudo-workaround option.
29357         This option is not used and make other option wrong.
29358
29359         * EWebLauncher/main.c:
29360
29361 2011-05-12  Lucas Forschler  <lforschler@apple.com>
29362
29363         Reviewed by Stephanie Lewis.
29364
29365         Fix config.json to address missing WK2 builder/tester config.
29366         Fix naming of machines (xserve not macpro)
29367
29368         * BuildSlaveSupport/build.webkit.org-config/config.json:
29369
29370 2011-05-12  Lucas Forschler  <lforschler@apple.com>
29371
29372         Reviewed by Stephanie Lewis.
29373
29374         Add SnowLeopard Debug builder and Test bot.
29375         apple-xserve-8 = SL Debug Tester.
29376         apple-xserve-9 = SL Debug Builder.
29377
29378         * BuildSlaveSupport/build.webkit.org-config/config.json:
29379
29380 2011-05-12  Adam Barth  <abarth@webkit.org>
29381
29382         This patch disables sheriffbot's bug postings.  He's just been way too
29383         spammy recently due to test flakiness.
29384
29385         * Scripts/webkitpy/tool/commands/sheriffbot.py:
29386         * Scripts/webkitpy/tool/commands/sheriffbot_unittest.py:
29387
29388 2011-05-11  Brian Weinstein  <bweinstein@apple.com>
29389
29390         Reviewed by Anders Carlsson.
29391
29392         After relaunching an unresponsive web process, web page stops responding to mouse movement
29393         https://bugs.webkit.org/show_bug.cgi?id=60669
29394         <rdar://problem/9409934>
29395
29396         This test follows the steps needed to reproduce this bug:
29397         1) Hang the web process
29398         2) While the web process is hung, move the mouse (so WebPageProxy thinks it is processing a mouse move event)
29399         3) Kill and restart the WebProcess
29400         4) Move the mouse
29401
29402         We are testing that we receive mouse move events after the WebProcess has restarted.
29403
29404         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash.cpp: Added.
29405         (TestWebKitAPI::didFinishLoadForFrame): Set a flag that we finished loading.
29406         (TestWebKitAPI::processDidBecomeUnresponsive): Kill the web process once we become unresponsive.
29407         (TestWebKitAPI::processDidCrash): Reload the page when the process crashes, which restarts the web process.
29408         (TestWebKitAPI::setPageLoaderClient):
29409         (TestWebKitAPI::TEST):
29410         * TestWebKitAPI/Tests/WebKit2/MouseMoveAfterCrash_Bundle.cpp: Added.
29411         (TestWebKitAPI::MouseMoveAfterCrashTest::MouseMoveAfterCrashTest):
29412         (TestWebKitAPI::MouseMoveAfterCrashTest::didReceiveMessage):
29413         * TestWebKitAPI/Tests/WebKit2/mouse-move-listener.html: Added.
29414
29415         * TestWebKitAPI/PlatformWebView.h:
29416         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
29417         (TestWebKitAPI::PlatformWebView::simulateMouseMove): Sends the messages that simulate a mouse move.
29418         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
29419         (TestWebKitAPI::PlatformWebView::simulateMouseMove): Ditto.
29420
29421         Add new files.
29422         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
29423         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
29424         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
29425         * TestWebKitAPI/win/copy-resources.cmd:
29426
29427 2011-05-12  Tomasz Morawski  <t.morawski@samsung.com>
29428
29429         Reviewed by Antonio Gomes.
29430
29431         [EFL] Added initialization/shutdown for ecore_file library
29432         https://bugs.webkit.org/show_bug.cgi?id=60457
29433
29434         Added initialization and shuthown for ecore_file library and replaced
29435         realpath function to ecore_file_realpath.
29436
29437         * EWebLauncher/main.c:
29438         (quit):
29439         (findThemePath):
29440         (main):
29441
29442 2011-05-12  Young Han Lee  <joybro@company100.net>
29443
29444         Reviewed by Kenneth Rohde Christiansen.
29445
29446         [Qt] Implement layoutTestController.layerTreeAsText()
29447         https://bugs.webkit.org/show_bug.cgi?id=60367
29448
29449         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
29450         (LayoutTestController::layerTreeAsText):
29451         * DumpRenderTree/qt/LayoutTestControllerQt.h:
29452
29453 2011-05-11  Eric Seidel  <eric@webkit.org>
29454
29455         Reviewed by David Levin.
29456
29457         Teach validate-committer-lists how to validate committers bugzilla_email()
29458         https://bugs.webkit.org/show_bug.cgi?id=60651
29459
29460         It turns out we currently have 9 committers with invalid values for their
29461         bugzilla_email field (meaning we don't know what email is their bugzilla account).
29462         "Alexander Kellett" <lypanov@mac.com>
29463         "Andre Boule" <aboule@apple.com>
29464         "Cameron McCormack" <cam@webkit.org>
29465         "Chris Petersen" <cpetersen@apple.com>
29466         "Eli Fidler" <eli@staikos.net>
29467         "Matt Lilek" <webkit@mattlilek.com>
29468         "Pierre d'Herbemont" <pdherbemont@free.fr>
29469         "Scott Violet" <sky@chromium.org>
29470         "Adam Treat" <treat@kde.org>
29471
29472         Most of those haven't committed in a very long time, so it's not suprising
29473         that we haven't noticed that their bugzilla emails are bogus.
29474
29475         * Scripts/validate-committer-lists:
29476         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
29477         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
29478
29479 2011-05-11  Dmitry Lomov  <dslomov@google.com>
29480
29481         Reviewed by David Levin.
29482
29483         https://bugs.webkit.org/show_bug.cgi?id=59561
29484         Switch TestWebKitAPI to GTest
29485
29486         Fixes Makefiles and WinCairo build as well
29487
29488         * Makefile:
29489         * Scripts/build-api-tests:
29490         * Scripts/run-api-tests:
29491         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
29492         * TestWebKitAPI/Test.h:
29493         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
29494         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
29495         (TestWebKitAPI::TEST):
29496         * TestWebKitAPI/TestsController.cpp:
29497         (TestWebKitAPI::TestsController::TestsController):
29498         (TestWebKitAPI::TestsController::dumpTestNames):
29499         (TestWebKitAPI::TestsController::runTestNamed):
29500         (TestWebKitAPI::TestsController::runAllTests):
29501         * TestWebKitAPI/TestsController.h:
29502         * TestWebKitAPI/mac/main.mm:
29503         (main):
29504
29505 2011-05-11  Eric Seidel  <eric@webkit.org>
29506
29507         Reviewed by Alexey Proskuryakov.
29508
29509         Add Luke Macpherson to committers.py as a Contributor
29510         https://bugs.webkit.org/show_bug.cgi?id=60639
29511
29512         * Scripts/webkitpy/common/config/committers.py:
29513
29514 2011-05-11  Noam Rosenthal  <noam.rosenthal@nokia.com>
29515
29516         Reviewed by Kenneth Rohde Christiansen.
29517
29518         [Texmap][Qt] Upstream texture-mapper changes from Qt's WebKit2 branch
29519         https://bugs.webkit.org/show_bug.cgi?id=60439
29520
29521         Patch 4/12: Update the build script to enable 3D rendering by default for Qt.
29522
29523         * Scripts/build-webkit:
29524
29525 2011-05-11  Brian Weinstein  <bweinstein@apple.com>
29526
29527         Reviewed by Adam Roben.
29528
29529         REGRESSION (r86161): WKPreferences API test failing on Windows
29530         https://bugs.webkit.org/show_bug.cgi?id=60658
29531
29532         * TestWebKitAPI/Tests/WebKit2/WKPreferences.cpp:
29533         (TestWebKitAPI::TEST):
29534
29535 2011-05-11  Alice Boxhall  <aboxhall@chromium.org>
29536
29537         Reviewed by Ojan Vafai.
29538
29539         Modify jsonresults_unittest.py to use a dict format for its test data, and modify jsonresults.py to flatten hierarchical directory structures in input JSON.
29540         https://bugs.webkit.org/show_bug.cgi?id=60521
29541
29542         First steps towards making the results JSON all hierarchical.
29543
29544         * TestResultServer/model/jsonresults.py:
29545         * TestResultServer/model/jsonresults_unittest.py:
29546
29547 2011-05-11  Kevin Ollivier  <kevino@theolliviers.com>
29548
29549         [wx] Build fix, adding new LayoutTestController method stubs.
29550
29551         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
29552         (LayoutTestController::localStorageDiskUsageForOrigin):
29553         (LayoutTestController::applicationCacheDiskUsageForOrigin):
29554
29555 2011-05-11  Jessie Berlin  <jberlin@apple.com>
29556
29557         Reviewed by Darin Adler.
29558
29559         [WebKit2] exceededDatabaseQuota function on WKPageUIClient needs to provide information
29560         about the usage for both the origin overall and the database in question.
29561         https://bugs.webkit.org/show_bug.cgi?id=60573
29562
29563         * WebKitTestRunner/TestController.cpp:
29564         (WTR::exceededDatabaseQuota):
29565
29566 2011-05-10  Stephen White  <senorblanco@chromium.org>
29567
29568         Reviewed by David Levin.
29569
29570         Adding myself as a reviewer.
29571         https://bugs.webkit.org/show_bug.cgi?id=60587
29572
29573         * Scripts/webkitpy/common/config/committers.py:
29574
29575 2011-05-11  Eric Seidel  <eric@webkit.org>
29576
29577         Reviewed by Csaba Osztrogonác.
29578
29579         queues.webkit.org should show 1000+ instead of 1000 when we hit the query limit
29580         https://bugs.webkit.org/show_bug.cgi?id=60614
29581
29582         AppEngine has a result limit of 1000.  We hit that when trying
29583         to find out how many patches the style-queue (and other queues)
29584         have passed in the last 30 days.  So lets display 1000+ instead
29585         of 1000 to avoid confusion. (Ossy was asking about this in #webkit tonight.)
29586
29587         I also made the status pages show "never" instead of just "ago" when we
29588         haven't seen a pass or boot message in the last 1000 messages.
29589
29590         * QueueStatusServer/handlers/queuestatus.py:
29591         * QueueStatusServer/templates/queuestatus.html:
29592
29593 2011-05-11  Adam Barth  <abarth@webkit.org>
29594
29595         Reviewed by Eric Seidel.
29596
29597         sheriffbot shouldn't spam bugs when it can't figure out who to blame
29598         https://bugs.webkit.org/show_bug.cgi?id=60613
29599
29600         This patch limits the number of bugs sheriffbot will spam to three.
29601         Ideally, we'd come up with some smarter huerstics, but hopefully this
29602         will cut down on the bugmail spam for slow bots.
29603
29604         * Scripts/webkitpy/tool/commands/sheriffbot.py:
29605
29606 2011-05-10  Eric Seidel  <eric@webkit.org>
29607
29608         Reviewed by Adam Barth.
29609
29610         EWS bots should be robust against test-webkitpy hangs
29611         https://bugs.webkit.org/show_bug.cgi?id=60579
29612
29613         Added a killAfterTimeout subroutine, which takes a child_pid
29614         and timeout, and will setup a watchdog process to kill that child
29615         after the timeout.  killAfterTimeout will waitpid on the child.
29616
29617         * EWSTools/start-queue.sh:
29618
29619 2011-05-10  MORITA Hajime  <morrita@google.com>
29620
29621         Unreviewed, moved myself to the reviewers list.
29622
29623         * Scripts/webkitpy/common/config/committers.py:
29624
29625 2011-05-10  Anton D'Auria  <adauria@apple.com>
29626
29627         Reviewed by David Levin.
29628
29629         StorageTracker should report actual local storage usage on disk
29630         https://bugs.webkit.org/show_bug.cgi?id=60558
29631
29632         * DumpRenderTree/LayoutTestController.cpp:
29633         (localStorageDiskUsageForOriginCallback):
29634         (LayoutTestController::staticFunctions):
29635         * DumpRenderTree/LayoutTestController.h:
29636         * DumpRenderTree/chromium/LayoutTestController.cpp:
29637         (LayoutTestController::LayoutTestController):
29638         (LayoutTestController::localStorageDiskUsageForOrigin):
29639         * DumpRenderTree/chromium/LayoutTestController.h:
29640         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
29641         (LayoutTestController::localStorageDiskUsageForOrigin):
29642         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
29643         (LayoutTestController::localStorageDiskUsageForOrigin):
29644         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
29645         (LayoutTestController::localStorageDiskUsageForOrigin):
29646         * DumpRenderTree/qt/LayoutTestControllerQt.h:
29647         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
29648         (LayoutTestController::localStorageDiskUsageForOrigin):
29649
29650 2011-05-10  Anton D'Auria  <adauria@apple.com>
29651
29652         Reviewed by Alexey Proskuryakov.
29653
29654         Add ApplicationCache API to get disk usage per origin
29655         https://bugs.webkit.org/show_bug.cgi?id=57127
29656
29657         Added test that loads two frames with ~15Kb image, checks usage,
29658         deletes app cache, verifies usage is 0.
29659
29660         * DumpRenderTree/LayoutTestController.cpp:
29661         (applicationCacheDiskUsageForOriginCallback):
29662         (LayoutTestController::staticFunctions):
29663         * DumpRenderTree/LayoutTestController.h:
29664         * DumpRenderTree/chromium/LayoutTestController.cpp: Added stub.
29665         (LayoutTestController::LayoutTestController):
29666         (LayoutTestController::applicationCacheDiskUsageForOrigin):
29667         * DumpRenderTree/chromium/LayoutTestController.h:
29668         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Added stub.
29669         (LayoutTestController::applicationCacheDiskUsageForOrigin):
29670         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
29671         (LayoutTestController::applicationCacheDiskUsageForOrigin):
29672         * DumpRenderTree/qt/LayoutTestControllerQt.cpp: Added stub.
29673         (LayoutTestController::applicationCacheDiskUsageForOrigin):
29674         * DumpRenderTree/qt/LayoutTestControllerQt.h:
29675         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Added stub.
29676         (LayoutTestController::applicationCacheDiskUsageForOrigin):
29677
29678 2011-05-10  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
29679
29680         Unreviewed, adding Rafael Antognolli as a committer.
29681
29682         * Scripts/webkitpy/common/config/committers.py:
29683
29684 2011-05-10  Chris Evans  <cevans@chromium.org>
29685
29686         Reviewed by Adam Barth.
29687
29688         Add WebCore::Setting to block displaying and/or running insecure content on secure pages
29689         https://bugs.webkit.org/show_bug.cgi?id=58378
29690
29691         * DumpRenderTree/chromium/LayoutTestController.cpp:
29692         (LayoutTestController::LayoutTestController):
29693         (LayoutTestController::setAllowDisplayOfInsecureContent):
29694         (LayoutTestController::setAllowRunningOfInsecureContent):
29695         * DumpRenderTree/chromium/LayoutTestController.h:
29696         * DumpRenderTree/chromium/WebPreferences.cpp:
29697         (WebPreferences::reset):
29698         (WebPreferences::applyTo):
29699         * DumpRenderTree/chromium/WebPreferences.h:
29700         Add plumbing for new settings to Chromium DRT.
29701
29702 2011-05-10  Enrica Casucci  <enrica@apple.com>
29703
29704         Reviewed by Beth Dakin.
29705
29706         Added my name to the reviewers list and removed my name from
29707         to committer-only list.
29708
29709         * Scripts/webkitpy/common/config/committers.py:
29710
29711 2011-05-10  David Levin  <levin@chromium.org>
29712
29713         Reviewed by Shinichiro Hamaji.
29714
29715         check-webkit-style shouldn't filter the errors using the modified lines for the xml and test expectations checkers.
29716         https://bugs.webkit.org/show_bug.cgi?id=60466
29717
29718         * Scripts/webkitpy/style/checker.py: (unrelated) style fix.
29719         * Scripts/webkitpy/style/checker_unittest.py: Remove the file_path check since
29720           that property was removed.
29721         * Scripts/webkitpy/style/checkers/test_expectations.py: Turn off the line
29722           filtering for errors.
29723         * Scripts/webkitpy/style/checkers/test_expectations_unittest.py: Add testing to
29724           verify that the line filtering is turned off.
29725         * Scripts/webkitpy/style/checkers/xml.py: Turn off the line filtering for errors
29726           and remove the unused file_path parameter.
29727         * Scripts/webkitpy/style/checkers/xml_unittest.py: Add testing to verify that the
29728           line filtering is turned off (and fix test_no_error).
29729         * Scripts/webkitpy/style/error_handlers.py: Added a way to turn off the filtering
29730           of errors based on the lines changed in the patch.
29731         * Scripts/webkitpy/style/error_handlers_unittest.py: Added a test for turning off
29732           the filtering.
29733
29734 2011-05-10  Adam Roben  <aroben@apple.com>
29735
29736         Update some plugin tests to work with WebKit2's asynchronous windowed plugin positioning on Windows
29737
29738         Fixes <http://webkit.org/b/60442> <rdar://problem/9403200> REGRESSION (r85795):
29739         platform/win/plugins/window-geometry-initialized-before-set-window.html and
29740         platform/win/plugins/window-region-is-set-to-clip-rect.html failing on Windows 7 Release
29741         (WebKit2 Tests)
29742
29743         Reviewed by Anders Carlsson.
29744
29745         * DumpRenderTree/TestNetscapePlugIn/Tests/win/NPNInvalidateRectInvalidatesWindow.cpp:
29746         (NPNInvalidateRectInvalidatesWindow::NPP_SetWindow): Moved some code from here to the new
29747         WindowedPluginTest::testHarnessWindow function.
29748
29749         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowGeometryInitializedBeforeSetWindow.cpp:
29750         (WindowGeometryInitializedBeforeSetWindow::NPP_SetWindow): Only check the window's size, not
29751         its position. In WebKit2, the position won't have been set yet, so we can't test it. But it
29752         turns out only the size mattered for <http://webkit.org/b/46716>, which this test was
29753         created for.
29754
29755         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowRegionIsSetToClipRect.cpp:
29756         (WindowRegionIsSetToClipRect::WindowRegionIsSetToClipRect): Converted to be a
29757         WindowedPluginTest, and initialized new members.
29758         (WindowRegionIsSetToClipRect::startTest): Make our plugin window visible, and set things up
29759         to ensure we get a WM_WINDOWPOSCHANGED message about it becoming visible.
29760         (WindowRegionIsSetToClipRect::finishTest): Check that the window region is what we expect,
29761         then clean up and tell the test harness we're done.
29762         (WindowRegionIsSetToClipRect::checkWindowRegion): Moved code to check the window region here
29763         from NPP_SetWindow. Changed to use the fancy new log() function instead of pluginLog().
29764         (WindowRegionIsSetToClipRect::showTestHarnessWindowIfNeeded): Show the test harness's window
29765         if it is hidden. This is needed to ensure we get a WM_WINDOWPOSCHANGED message when our own
29766         window becomes visible.
29767         (WindowRegionIsSetToClipRect::hideTestHarnessWindowIfNeeded): Hide the test harness's window
29768         if it was hidden at the start of the test.
29769         (WindowRegionIsSetToClipRect::wndProc): When our timer fires, tell the UI process to paint.
29770         This should cause our window to become visible and our window region to get set. When we get
29771         a WM_WINDOWPOSCHANGED message, we can finally finish the test.
29772         (WindowRegionIsSetToClipRect::NPP_GetValue): Return our script object.
29773         (WindowRegionIsSetToClipRect::ScriptObject::hasMethod): We only have one method:
29774         "startTest".
29775         (WindowRegionIsSetToClipRect::ScriptObject::invoke): Call through to the PluginTest object.
29776
29777         * DumpRenderTree/TestNetscapePlugIn/win/WindowedPluginTest.cpp:
29778         (WindowedPluginTest::testHarnessWindow):
29779         * DumpRenderTree/TestNetscapePlugIn/win/WindowedPluginTest.h:
29780         Added this helper function. Code came from NPNInvalidateRectInvalidatesWindow.
29781
29782 2011-05-10  Csaba Osztrogonác  <ossy@webkit.org>
29783
29784         Reviewed by Kenneth Rohde Christiansen.
29785
29786         [Qt] Add new qt-4.8 platform for layout testing
29787         https://bugs.webkit.org/show_bug.cgi?id=60544
29788
29789         * Scripts/old-run-webkit-tests:
29790         * Scripts/webkitdirs.pm:
29791
29792 2011-05-09  Sheriff Bot  <webkit.review.bot@gmail.com>
29793
29794         Unreviewed, rolling out r86108.
29795         http://trac.webkit.org/changeset/86108
29796         https://bugs.webkit.org/show_bug.cgi?id=60533
29797
29798         Broke makefile based builds and Cairo (Requested by weinig on
29799         #webkit).
29800
29801         * Scripts/build-api-tests:
29802         * Scripts/run-api-tests:
29803         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
29804         * TestWebKitAPI/Test.h:
29805         (TestWebKitAPI::Test::~Test):
29806         (TestWebKitAPI::Test::name):
29807         (TestWebKitAPI::Test::Register::Register):
29808         (TestWebKitAPI::Test::Register::create):
29809         (TestWebKitAPI::Test::Test):
29810         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
29811         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
29812         (TestWebKitAPI::TEST):
29813         * TestWebKitAPI/TestsController.cpp:
29814         (TestWebKitAPI::TestsController::TestsController):
29815         (TestWebKitAPI::TestsController::dumpTestNames):
29816         (TestWebKitAPI::TestsController::runTestNamed):
29817         (TestWebKitAPI::TestsController::testFailed):
29818         (TestWebKitAPI::TestsController::registerCreateTestFunction):
29819         * TestWebKitAPI/TestsController.h:
29820         * TestWebKitAPI/mac/main.mm:
29821         (main):
29822
29823 2011-05-09  Hayato Ito  <hayato@chromium.org>
29824
29825         Reviewed by Ojan Vafai.
29826
29827         Skip reftests if pixel tests are disabled.
29828
29829         https://bugs.webkit.org/show_bug.cgi?id=59078
29830
29831         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
29832         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
29833         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
29834         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
29835
29836 2011-05-09  Chang Shu  <cshu@webkit.org>
29837
29838         Reviewed by Kenneth Rohde Christiansen.
29839
29840         [Qt] DumpRenderTreeQt needs an implementation of unmarkText
29841         https://bugs.webkit.org/show_bug.cgi?id=60499
29842
29843         The implementation is similar to gtk.
29844
29845         * DumpRenderTree/qt/TextInputControllerQt.cpp:
29846         (TextInputController::unmarkText):
29847         * DumpRenderTree/qt/TextInputControllerQt.h:
29848
29849 2011-05-09  James Robinson  <jamesr@chromium.org>
29850
29851         Reviewed by Eric Seidel.
29852
29853         [chromium] Teach new-run-webkit-tests to follow symlinks when trying to determine the arch of DumpRenderTree on linux
29854         https://bugs.webkit.org/show_bug.cgi?id=60520
29855
29856         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
29857
29858 2011-05-09  Dmitry Lomov  <dslomov@google.com>
29859
29860         Reviewed by Sam Weinig.
29861
29862         Switch TestWebKitAPI to GTest
29863         https://bugs.webkit.org/show_bug.cgi?id=59561
29864
29865         * Scripts/build-api-tests:
29866         * Scripts/run-api-tests:
29867         * TestWebKitAPI/Configurations/TestWebKitAPICommon.vsprops:
29868         * TestWebKitAPI/Test.h:
29869         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
29870         * TestWebKitAPI/Tests/WTF/VectorBasic.cpp:
29871         (TestWebKitAPI::TEST):
29872         * TestWebKitAPI/TestsController.cpp:
29873         (TestWebKitAPI::TestsController::TestsController):
29874         (TestWebKitAPI::TestsController::dumpTestNames):
29875         (TestWebKitAPI::TestsController::runTestNamed):
29876         (TestWebKitAPI::TestsController::runAllTests):
29877         * TestWebKitAPI/TestsController.h:
29878         * TestWebKitAPI/mac/main.mm:
29879         (main):
29880
29881 2011-05-09  Igor Oliveira  <igor.oliveira@openbossa.org>
29882
29883         Reviewed by Andreas Kling.
29884
29885         [Qt] DRT eventSender.addTouchPoint method is not setting the start position
29886         https://bugs.webkit.org/show_bug.cgi?id=60487
29887
29888         When adding a new touch point, the start position of this point needs to be set.
29889
29890         * DumpRenderTree/qt/EventSenderQt.cpp:
29891         (EventSender::addTouchPoint):
29892
29893 2011-05-09  Robert Hogan  <robert@webkit.org>
29894
29895         Reviewed by Adam Roben.
29896
29897         https://bugs.webkit.org/show_bug.cgi?id=57137
29898
29899         Unskip plugins/embed-prefers-plugins-for-images.html
29900
29901         Add image/png mime-type to test netscape plugin on Unix.
29902
29903         * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
29904         (NP_GetMIMEDescription):
29905
29906 2011-05-09  Alice Boxhall  <aboxhall@chromium.org>
29907
29908         Reviewed by Ojan Vafai.
29909
29910         Pull jsonresults_unittest test cases out into separate test methods.
29911         https://bugs.webkit.org/show_bug.cgi?id=60418
29912
29913         * TestResultServer/model/jsonresults_unittest.py:
29914
29915 2011-05-09  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
29916
29917         Reviewed by Andreas Kling.
29918
29919         [Qt] fix fast/dom/Geolocation/window-close-crash.html
29920         https://bugs.webkit.org/show_bug.cgi?id=60458
29921
29922         Update all pages with Mock position/error/permission.
29923
29924         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
29925         (LayoutTestController::setGeolocationPermission):
29926         (LayoutTestController::setMockGeolocationError):
29927         (LayoutTestController::setMockGeolocationPosition):
29928
29929 2011-05-09  Shishir Agrawal  <shishir@chromium.org>
29930
29931         Reviewed by Tony Gentilcore.
29932
29933         Implement Page Visibility API.
29934         https://bugs.webkit.org/show_bug.cgi?id=54181
29935
29936         * DumpRenderTree/LayoutTestController.cpp:
29937         (setPageVisibilityCallback):
29938         (resetPageVisibilityCallback):
29939         (LayoutTestController::staticFunctions):
29940         * DumpRenderTree/LayoutTestController.h:
29941         (LayoutTestController::setPageVisibility):
29942         (LayoutTestController::resetPageVisibility):
29943         * DumpRenderTree/chromium/LayoutTestController.cpp:
29944         (LayoutTestController::LayoutTestController):
29945         (LayoutTestController::resetPageVisibility):
29946         (LayoutTestController::setPageVisibility):
29947         * DumpRenderTree/chromium/LayoutTestController.h:
29948
29949 2011-05-07  Adam Barth  <abarth@webkit.org>
29950
29951         Reviewed by Eric Seidel.
29952
29953         Testing EWS spins on patches with a large number of failures
29954         https://bugs.webkit.org/show_bug.cgi?id=60441
29955
29956         In cases where the tree has a small number of persistent failures but a
29957         patch has a large number of failures, we weren't believing any results.
29958         That lead to us retrying the patch forever while we waited for
29959         trustworthy results that would never come.
29960
29961         This patch loosens the semantics of unexpected_failures to return a
29962         subset (rather than the exact set) of unexpected failures.  In the case
29963         where the tree has a bounded number of failures and the patch has an
29964         unbounded number of failures, we can accurately compute such a subset
29965         and reject the patch.
29966
29967         * Scripts/webkitpy/tool/bot/expectedfailures.py:
29968         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py:
29969         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
29970         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
29971         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
29972
29973 2011-05-07  Joe Mason  <jmason@rim.com>
29974
29975         Reviewed by Daniel Bates.
29976
29977         FrameLoader::isProcessingUserGesture is wrong in dispatchWillPerformClientRedirect
29978         https://bugs.webkit.org/show_bug.cgi?id=52211
29979
29980         Dump isUserProcessingGesture in willPerformClientRedirect so that the layout tests can be extended.
29981         (For the Mac and Chromium ports - other ports don't support dumping user gestures in DRT.)
29982
29983         * DumpRenderTree/chromium/WebViewHost.cpp:
29984         (WebViewHost::willPerformClientRedirect):
29985         (WebViewHost::didStartProvisionalLoad):
29986         * DumpRenderTree/mac/FrameLoadDelegate.mm:
29987         (-[FrameLoadDelegate webView:willPerformClientRedirectToURL:delay:fireDate:forFrame:]):
29988
29989 2011-05-07  Daniel Bates  <dbates@webkit.org>
29990
29991         Attempt to fix the build; remove some more references to isTiger(), which was removed in changeset 86007
29992         <http://trac.webkit.org/changeset/86007> (bug #60438).
29993
29994         * Scripts/old-run-webkit-tests:
29995           - Removed reference to isTiger() from openDumpTool(), captureSavedCrashLog(), and
29996             file local variable.
29997
29998 2011-05-07  Pratik Solanki  <psolanki@apple.com>
29999
30000         Reviewed by Darin Adler.
30001
30002         Remove references to Tiger in scripts
30003         https://bugs.webkit.org/show_bug.cgi?id=60438
30004
30005         Remove isTiger() and all of the code that depends on it.
30006
30007         * Scripts/build-webkit:
30008         * Scripts/ensure-valid-python:
30009         * Scripts/gdb-safari:
30010         * Scripts/webkitdirs.pm:
30011
30012 2011-05-06  David Grogan  <dgrogan@chromium.org>
30013
30014         Reviewed by Eric Seidel.
30015
30016         update link to message about mac 64bit git binaries
30017         https://bugs.webkit.org/show_bug.cgi?id=60411
30018
30019         The original message implied (correctly, at-the-time) that
30020         http://code.google.com/p/git-osx-installer/ doesn't provide 64-bit
30021         binaries. The new message notes that they do. This might prevent someone
30022         else from spending a few minutes confused.
30023
30024         * Scripts/webkitpy/common/checkout/scm/git.py:
30025
30026 2011-05-06  Dirk Pranke  <dpranke@chromium.org>
30027
30028         Reviewed by Ojan Vafai.
30029
30030         new-run-webkit-tests: crashes in race with crashing DRT
30031         https://bugs.webkit.org/show_bug.cgi?id=60238
30032
30033         This fixes a race where if we wrote the URL to DRT and
30034         DRT crashed immediately, we could get a broken pipe
30035         error back during the write() and not handle it correctly,
30036         causing NRWT to also crash.
30037
30038         * Scripts/webkitpy/layout_tests/port/server_process.py:
30039         * Scripts/webkitpy/layout_tests/port/webkit.py:
30040
30041 2011-05-06  Adam Barth  <abarth@webkit.org>
30042
30043         Reviewed by Eric Seidel.
30044
30045         Add a --skip-integrationtests mode to test-webkitpy
30046         https://bugs.webkit.org/show_bug.cgi?id=60354
30047
30048         These two test suites account for 2/3 of test-webkitpy's running time
30049         because they're not actually unit tests.  They're integration tests.
30050         This patch adds the ability to skip integration tests by passing a
30051         command line argument to test-webkitpy.  The buildbots will continue to
30052         run these integration tests to make sure they don't regress.  This
30053         option is for use by developers when working on parts of the code that
30054         are well-covered by unit tests.
30055
30056         * Scripts/webkitpy/common/system/file_lock_integrationtest.py: Copied from Tools/Scripts/webkitpy/common/system/file_lock_unittest.py.
30057         * Scripts/webkitpy/common/system/file_lock_unittest.py: Removed.
30058         * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py: Copied from Tools/Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py.
30059         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py: Removed.
30060         * Scripts/webkitpy/test/main.py:
30061
30062 2011-05-06  Adam Barth  <abarth@webkit.org>
30063
30064         Reviewed by Eric Seidel.
30065
30066         The results.zip generated by the Chromium EWS are too big
30067         https://bugs.webkit.org/show_bug.cgi?id=60352
30068
30069         Currently results.zip created by the Chromium EWS bot contain roughly
30070         1000 tests that we expect to fail.  These test results are not helpful
30071         to contributors.
30072
30073         There are several approaches to solving this problem:
30074
30075         1) We could be more selective about which files to include in the zip.
30076         2) We could ask new-run-webkit-tests not to generate results for
30077            expected failures.
30078         3) We could ask new-run-webkit-tests not to run tests that we expect to
30079            fail.
30080
30081         This patch takes approach (3).  Currently, we're ignoring the outcome
30082         of failing tests anyway, so there's not much point in actually running
30083         them.  As a side benefit, this will speed up the test run by about 6%.
30084         In the future, we might revisit this decision and choose another
30085         approach.
30086
30087         * Scripts/webkitpy/common/config/ports.py:
30088         * Scripts/webkitpy/common/config/ports_unittest.py:
30089
30090 2011-05-06  Gabor Rapcsanyi  <rgabor@webkit.org>
30091
30092         Reviewed by Csaba Osztrogonác.
30093
30094         [Qt] Use qt-arm platform instead of qt-linux on ARM platform.
30095
30096         * Scripts/old-run-webkit-tests:
30097
30098 2011-05-06  Adam Barth  <abarth@webkit.org>
30099
30100         Reviewed by Ojan Vafai.
30101
30102         Add an option to NRWT to skip failing tests
30103         https://bugs.webkit.org/show_bug.cgi?id=60295
30104
30105         This option is similar to the way old-run-webkit-tests works.  We're
30106         planning to use this option on the testing EWS bots because they don't
30107         use information from failing tests yet.
30108
30109         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
30110         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
30111         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
30112
30113 2011-05-04  Philippe Normand  <pnormand@igalia.com>
30114
30115         Reviewed by Dimitri Glazkov.
30116
30117         Implement LayoutTestController::pseudoShadowId()
30118         https://bugs.webkit.org/show_bug.cgi?id=60034
30119
30120         shadowPseudoId support in the various LayoutTestController
30121         implementations.
30122
30123         * DumpRenderTree/LayoutTestController.cpp:
30124         (shadowPseudoIdCallback):
30125         (LayoutTestController::staticFunctions):
30126         * DumpRenderTree/LayoutTestController.h:
30127         * DumpRenderTree/chromium/LayoutTestController.cpp:
30128         (LayoutTestController::LayoutTestController):
30129         (LayoutTestController::shadowPseudoId):
30130         * DumpRenderTree/chromium/LayoutTestController.h:
30131         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
30132         (LayoutTestController::shadowPseudoId):
30133         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
30134         (LayoutTestController::shadowPseudoId):
30135         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
30136         (LayoutTestController::shadowPseudoId):
30137         * DumpRenderTree/qt/LayoutTestControllerQt.h:
30138         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
30139         (LayoutTestController::shadowPseudoId):
30140         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
30141         (LayoutTestController::shadowPseudoId):
30142         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
30143         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
30144         (WTR::LayoutTestController::shadowPseudoId):
30145         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
30146
30147 2011-05-06  Adam Barth  <abarth@webkit.org>
30148
30149         EWS can't reject patches
30150         https://bugs.webkit.org/show_bug.cgi?id=60345
30151
30152         We're supposed to use patch.id() for the attachment ID, silly.
30153
30154         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
30155         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
30156
30157 2011-05-05  Steve Lacey  <sjl@chromium.org>
30158
30159         Reviewed by Ojan Vafai.
30160
30161         Fix generation of results html file by passing relative path to scm
30162         methods.
30163         Generate diff images for results html file.
30164
30165         https://bugs.webkit.org/show_bug.cgi?id=60086
30166
30167         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
30168         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
30169
30170 2011-05-05  Andy Estes  <aestes@apple.com>
30171
30172         Reviewed by Daniel Bates.
30173
30174         update-webkit should handle Internal using a different VCS than OpenSource.
30175         https://bugs.webkit.org/show_bug.cgi?id=60310
30176
30177         isGit() and isSVN() cache their results, so Internal will always be
30178         updated using the VCS commands used for OpenSource. If Internal uses a
30179         different VCS, this will fail. Fix this by calling isGitDirectory() and
30180         isSVNDirectory() instead, which do not cache their results.
30181
30182         * Scripts/update-webkit:
30183
30184 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
30185
30186         Reviewed by Ojan Vafai.
30187
30188         nrwt: revamp integration tests incl. tests for servers
30189         https://bugs.webkit.org/show_bug.cgi?id=59985
30190
30191         This change adds the ability to run "integration tests" that
30192         depend on (and affect) the local environment on your machine.
30193         These tests can be run manually by invoking python on the
30194         test file in question and passing the '-i' flag, but are
30195         not run automatically as part of test-webkitpy.
30196
30197         Integration tests are identified by being methods named as
30198         "integration_test_*" instead of "test_*", and we use a
30199         custom TestLoader instead of the default one provided by
30200         unittest.
30201
30202         This change also adds a protected "_default_to_apache" method
30203         to the port object so we can do the right thing when starting
30204         and stopping web servers with less mechanism.
30205
30206         * Scripts/webkitpy/layout_tests/port/base.py:
30207         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
30208         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
30209         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
30210         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
30211         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
30212
30213 2011-05-05  Adam Roben  <aroben@apple.com>
30214
30215         Robustify DRT's pixel-dumping code against semi-random alpha channel values on Windows
30216
30217         DRT generates a pixel dump on Windows using the WM_PRINT message. This message (just like
30218         most of GDI) doesn't write anything meaningful into the alpha channel. In some cases, the
30219         alpha channel will even get filled with zeros (like when ::GradientFill is used). To
30220         safeguard against this, we fill the alpha channel with constant full opacity after the
30221         WM_PRINT message is handled to ensure we end up with an opaque image. (Note that we can't
30222         just ignore the alpha channel when creating the PNG; Mac's pixel results have an alpha
30223         channel, and we want to be able to compare Mac and Windows pixel results, so Windows pixel
30224         results need an alpha channel, too.)
30225
30226         Fixes <http://webkit.org/b/60300> draws-gradient.html's pixel results can't be correctly
30227         generated due to alpha channel weirdness
30228
30229         Reviewed by Sam Weinig.
30230
30231         * DumpRenderTree/win/PixelDumpSupportWin.cpp:
30232         (makeAlphaChannelOpaque): Added this helper function to fill a bitmap's alpha channel with
30233         constant full opacity.
30234         (createBitmapContextFromWebView): Make the bitmap's alpha channel opaque before wrapping it
30235         in a CGContext or cairo_surface.
30236
30237 2011-05-05  Adam Roben  <aroben@apple.com>
30238
30239         Add support for layoutTestController.dumpAsText(true) to WebKitTestRunner
30240
30241         Passing true tells the test harness to dump pixels even though we're not dumping a render
30242         tree.
30243
30244         Fixes <http://webkit.org/b/60298> Calling layoutTestController.dumpAsText(true) in
30245         WebKitTestRunner doesn't result in pixel results being generated
30246
30247         Reviewed by Sam Weinig.
30248
30249         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Added dumpPixels
30250         argument to dumpAsText.
30251
30252         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
30253         (WTR::InjectedBundlePage::dump): Updated for dumpAsText change.
30254
30255         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
30256         (WTR::LayoutTestController::dumpAsText): Store the dumpPixels argument.
30257
30258         * WebKitTestRunner/InjectedBundle/LayoutTestController.h: Updated for IDL change.
30259
30260         * WebKitTestRunner/win/InjectedBundle.vcproj: Added .idl files as a convenience for Windows
30261         developers.
30262
30263 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
30264
30265         Reviewed by Ojan Vafai.
30266
30267         new-run-webkit-tests: merge os, version fields in test config
30268         https://bugs.webkit.org/show_bug.cgi?id=60004
30269
30270         This change removes the 'o/s' field from the TestConfiguration
30271         object, and changes the 'MAC', 'WIN', and 'LINUX' modifiers in
30272         the expectations file to macros that expand to the version names
30273         for each platform.
30274
30275         This means that you can once again put any subset of the
30276         operating system versions on a single line, e.g.:
30277           BUGX LEOPARD XP LINUX : fast/html/foo.html = FAIL
30278
30279         This change will break any existing test_expectations.txt
30280         file that uses the "OS-VERSION" macros, which are no longer
30281         needed. It turns out they were never needed, as just VERSION
30282         by itself was sufficient. The test_expectations.txt file should
30283         be updated accordingly prior to landing this change.
30284
30285         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
30286         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
30287         * Scripts/webkitpy/layout_tests/port/base.py:
30288
30289 2011-05-05  Tony Chang  <tony@chromium.org>
30290
30291         Reviewed by Adam Barth.
30292
30293         [chromium] switch from Lucida to free fonts for Hindi and Thai
30294         https://bugs.webkit.org/show_bug.cgi?id=60287
30295
30296         The fonts are provided by the ttf-thai-tlwg and ttf-indic-fonts
30297         packages on Lucid. There are actually multiple Thai fonts in the
30298         package, so I just picked one.
30299
30300         * DumpRenderTree/chromium/TestShellGtk.cpp:
30301         (setupFontconfig):
30302
30303 2011-05-05  Tony Chang  <tony@chromium.org>
30304
30305         Reviewed by Ojan Vafai.
30306
30307         [chromium] google-chrome-64 builds should now fallback to chromium-linux-x86_64 results
30308         https://bugs.webkit.org/show_bug.cgi?id=60290
30309
30310         It used to be that google-chrome-32 was hardy32, google-chrome-64 was
30311         hardy64, chromium-linux was hardy32 and chromium-linux-x86_64 was
30312         lucid64.  Because of that, google-chrome-64 used to skip the
30313         chromium-linux-x86_64 results since they were more incorrect.
30314
30315         Now that we're switching to lucid for google-chrome, we want
30316         google-chrome-64 to fall back to chromium-linux-x86_64.
30317
30318         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
30319
30320 2011-05-05  Dirk Pranke  <dpranke@chromium.org>
30321
30322         Reviewed by Ojan Vafai.
30323
30324         new-run-webkit-tests: merge TestExpectations, TestExpectationsFile classes
30325         https://bugs.webkit.org/show_bug.cgi?id=60002
30326
30327         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
30328         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
30329         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
30330         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
30331         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
30332         * Scripts/webkitpy/style/checkers/test_expectations.py:
30333
30334 2011-05-05  Eric Seidel  <eric@webkit.org>
30335
30336         Reviewed by Adam Barth.
30337
30338         commit-queue should reject patches which fail ewses
30339         https://bugs.webkit.org/show_bug.cgi?id=47534
30340
30341         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
30342         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
30343         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
30344
30345 2011-05-05  Chang Shu  <cshu@webkit.org>
30346
30347         Reviewed by Andreas Kling.
30348
30349         [Qt][WK2] Support PageGroup in QGraphicsWKView/QWKPage constructors
30350         https://bugs.webkit.org/show_bug.cgi?id=59638
30351
30352         1. Call QGraphicsWKView constructor with pageGroupRef.
30353         2. Disable setFontFamily calls as it breaks Qt test fonts.
30354
30355         * WebKitTestRunner/TestController.cpp:
30356         (WTR::TestController::resetStateToConsistentValues):
30357         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
30358         (WTR::WebView::WebView):
30359         (WTR::PlatformWebView::PlatformWebView):
30360
30361 2011-05-05  Joseph Pecoraro  <joepeck@webkit.org>
30362
30363         Reviewed by David Kilzer.
30364
30365         Add --only-webkit option to build-webkit
30366         https://bugs.webkit.org/show_bug.cgi?id=60245
30367
30368         * Scripts/build-webkit: If --only-webkit was specified
30369         just build the WebKit project, and not its dependencies.
30370
30371 2011-05-05  Jeff Miller  <jeffm@apple.com>
30372
30373         Reviewed by Darin Adler.
30374
30375         Add didNotHandleWheelEvent callback to WKPageUIClient
30376         https://bugs.webkit.org/show_bug.cgi?id=60203
30377
30378         * MiniBrowser/gtk/BrowserWindow.c:
30379         (browserWindowUIClientInit): Stub out didNotHandleWheelEvent callback.
30380         * MiniBrowser/mac/BrowserWindowController.m:
30381         (-[BrowserWindowController awakeFromNib]): Stub out didNotHandleWheelEvent callback.
30382         * MiniBrowser/win/BrowserView.cpp:
30383         (BrowserView::create): Stub out didNotHandleWheelEvent callback.
30384         * WebKitTestRunner/TestController.cpp:
30385         (WTR::TestController::createOtherPage): Stub out didNotHandleWheelEvent callback.
30386         (WTR::TestController::initialize): Stub out didNotHandleWheelEvent callback.
30387
30388 2011-05-04  Eric Seidel  <eric@webkit.org>
30389
30390         Reviewed by Adam Barth.
30391
30392         Add more missing imports from scm refactor
30393         https://bugs.webkit.org/show_bug.cgi?id=60239
30394
30395         * Scripts/webkitpy/common/checkout/scm/git.py:
30396         * Scripts/webkitpy/common/checkout/scm/svn.py:
30397
30398 2011-05-04  Eric Seidel  <eric@webkit.org>
30399
30400         Reviewed by Adam Barth.
30401
30402         sheriff-bot is having more exception trouble
30403         https://bugs.webkit.org/show_bug.cgi?id=60233
30404
30405         A webpage-forced build was causing SB to throw an exception.
30406         We've long had this bug, but we just didn't notice it until now.
30407
30408         In debugging this I found we fetch an absurd amount of data now
30409         that we use buildbot's json (instead of xmlrpc), so I added
30410         the filter=1 parameter to reduce it a little.  That required
30411         me to handle the case where filter=1 would strip
30412         result=0 in some cases (it's unclear why it does that).
30413
30414         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
30415         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
30416
30417 2011-05-04  Cris Neckar  <cdn@chromium.org>
30418
30419         Unreviewed, Adding Cris Neckar as a committer
30420
30421         * Scripts/webkitpy/common/config/committers.py:
30422
30423 2011-05-04  Siddharth Mathur  <siddharth.mathur@nokia.com>
30424
30425         Reviewed by Laszlo Gombos.
30426
30427         [Qt][Wk2][Symbian] Capabilities for Qt MiniBrowser and QtWebProcess on Symbian
30428         https://bugs.webkit.org/show_bug.cgi?id=59462
30429
30430         * MiniBrowser/qt/MiniBrowser.pro: Add Symbian OS capabilities to executable. Also enable data paging.
30431
30432 2011-05-04  Eric Seidel  <eric@webkit.org>
30433
30434         Reviewed by Adam Barth.
30435
30436         The testing EWS bots should upload layout-test-results.zip to bugs after failures
30437         https://bugs.webkit.org/show_bug.cgi?id=60223
30438
30439         This required sharing a bit of code between the commit-queue
30440         (which already knew how to do this) and the new EWS testing bots.
30441
30442         In the process I also cleaned up EWS testing a little and
30443         removed some dead code from the commit-queue.
30444
30445         * Scripts/webkitpy/common/config/committervalidator.py:
30446         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
30447         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
30448         * Scripts/webkitpy/tool/commands/queues.py:
30449         * Scripts/webkitpy/tool/commands/queues_unittest.py:
30450
30451 2011-05-04  James Kozianski  <koz@chromium.org>
30452
30453         Reviewed by Eric Seidel.
30454
30455         Add method to make a Port able to retrieve TestOutputSets from its builders.
30456         https://bugs.webkit.org/show_bug.cgi?id=58211
30457
30458         This will be used in the new rebaseline script - TestOutputs are
30459         potential new baselines.
30460
30461         * Scripts/webkitpy/common/config/urls.py:
30462         * Scripts/webkitpy/layout_tests/port/base.py:
30463         * Scripts/webkitpy/layout_tests/port/chromium.py:
30464         * Scripts/webkitpy/layout_tests/port/factory.py:
30465         * Scripts/webkitpy/layout_tests/port/webkit.py:
30466
30467 2011-05-04  Eric Seidel  <eric@webkit.org>
30468
30469         Reviewed by Ojan Vafai.
30470
30471         sheriff-bot can't parse results.html
30472         https://bugs.webkit.org/show_bug.cgi?id=60196
30473
30474         The problem was that the NRWT bots are currently core
30475         and so sheriff-bot was trying to diagnose failures on
30476         them.  It was trying to parse their results.html file
30477         and BeautifulSoup was throwing an exception.
30478
30479         It seems our version of BeautifulSoup was quite old
30480         and unable to handle <script> tags correctly.
30481         This moves us to 3.2 which doesn't seem to have this bug.
30482
30483         * Scripts/webkitpy/thirdparty/BeautifulSoup.py:
30484
30485 2011-05-04  Adam Barth  <abarth@webkit.org>
30486
30487         Reviewed by Eric Seidel.
30488
30489         Make the EWS test logging less verbose
30490         https://bugs.webkit.org/show_bug.cgi?id=60201
30491
30492         This should help us generate less than a hundred MB of logs per hour.
30493
30494         * Scripts/webkitpy/common/config/ports.py:
30495         * Scripts/webkitpy/common/config/ports_unittest.py:
30496
30497 2011-05-04  Tao Bai  <michaelbai@chromium.org>
30498
30499         Reviewed by David Kilzer.
30500
30501         Populate touch-icon url to FrameLoaderClient
30502         https://bugs.webkit.org/show_bug.cgi?id=59143
30503
30504         Disable touch icon by default.
30505
30506         * Scripts/build-webkit:
30507
30508 2011-05-04  Justin Novosad  <junov@chromium.org>
30509
30510         Reviewed by Kenneth Russell.
30511
30512         [Chromium] Add gpu acceleration flags to DumpRenderTree
30513         https://bugs.webkit.org/show_bug.cgi?id=60168
30514         Adding support for command line flags
30515         enable-accelerated-2d-canvas an
30516         enable-accelerated-drawing in DumpRenderTree.
30517
30518         * DumpRenderTree/chromium/DumpRenderTree.cpp:
30519         (main):
30520         * DumpRenderTree/chromium/TestShell.cpp:
30521         (TestShell::TestShell):
30522         (TestShell::resetWebSettings):
30523         * DumpRenderTree/chromium/TestShell.h:
30524         (TestShell::setLegacyAccelerated2dCanvasEnabled):
30525         (TestShell::setAcceleratedDrawingEnabled):
30526         * DumpRenderTree/chromium/WebPreferences.cpp:
30527         (WebPreferences::reset):
30528         (WebPreferences::applyTo):
30529         * DumpRenderTree/chromium/WebPreferences.h:
30530
30531 2011-05-04  Ademar de Souza Reis Jr.  <ademar.reis@openbossa.org>
30532
30533         Reviewed by Csaba Osztrogonác.
30534
30535         [Qt] run-qtwebkit-tests misses exec bit
30536         https://bugs.webkit.org/show_bug.cgi?id=60188
30537
30538         Also removing an unecessary import (mostly to let this bugfix goes
30539         through the system, as this is just a chmod).
30540
30541         * Scripts/run-qtwebkit-tests:
30542
30543 2011-05-04  Adam Barth  <abarth@webkit.org>
30544
30545         Fix import of ircproxy module to actually import the module.  I don't
30546         really understand how this code could have worked before.
30547
30548         * Scripts/webkitpy/common/host.py:
30549
30550 2011-05-04  Siddharth Mathur  <siddharth.mathur@nokia.com>
30551
30552         Reviewed by Andreas Kling.
30553
30554         [Qt][Wk2][Symbian] Minor UI facelift for MiniBrowser
30555         https://bugs.webkit.org/show_bug.cgi?id=60172
30556
30557         * MiniBrowser/qt/BrowserWindow.cpp: Bring in UI layout changes from QtTestBrowser
30558         (BrowserWindow::BrowserWindow):
30559
30560 2011-05-04  Mihai Parparita  <mihaip@chromium.org>
30561
30562         Reviewed by Ojan Vafai.
30563
30564         [NRWT] --new-baseline should put baselines in the directory specified by --additional-platform-directory
30565         https://bugs.webkit.org/show_bug.cgi?id=60088
30566
30567         Makes --new-baseline put new baselines into the first entry in the
30568         baseline search path, including any additional directories passed in via
30569         --additional-platform-directory.
30570
30571         Simplifies baseline_path() by making it return the first entry in the
30572         baseline search path instead of requiring each port to have an
30573         implementation (should not result in any behavior change).
30574
30575         * Scripts/webkitpy/layout_tests/port/base.py:
30576         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
30577         * Scripts/webkitpy/layout_tests/port/chromium.py:
30578         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
30579         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
30580         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
30581         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
30582         * Scripts/webkitpy/layout_tests/port/mac.py:
30583         * Scripts/webkitpy/layout_tests/port/test.py:
30584         * Scripts/webkitpy/layout_tests/port/webkit.py:
30585
30586 2011-05-04  Adam Barth  <abarth@webkit.org>
30587
30588         Reviewed by Eric Seidel.
30589
30590         Fix circular dependency in webkitpy
30591         https://bugs.webkit.org/show_bug.cgi?id=60075
30592
30593         These functions don't belong in bugzilla.py.  They only exist there
30594         because they are old.  Really, these functions shouldn't be free
30595         functions at all, but that's a patch for another day.
30596
30597         * Scripts/webkitpy/common/checkout/changelog.py:
30598         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
30599         * Scripts/webkitpy/common/checkout/checkout.py:
30600         * Scripts/webkitpy/common/config/urls.py:
30601         * Scripts/webkitpy/common/net/bugzilla/__init__.py:
30602         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
30603         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
30604         * Scripts/webkitpy/style/checkers/changelog.py:
30605         * Scripts/webkitpy/tool/bot/irc_command.py:
30606         * Scripts/webkitpy/tool/bot/sheriff.py:
30607         * Scripts/webkitpy/tool/commands/upload.py:
30608
30609 2011-05-03  Pratik Solanki  <psolanki@apple.com>
30610
30611         Reviewed by Dan Bernstein.
30612
30613         https://bugs.webkit.org/show_bug.cgi?id=60141
30614         Remove reference to libWebKitSystemInterfaceTiger.a from build-webkit
30615
30616         * Scripts/build-webkit:
30617
30618 2011-05-03  Mihai Parparita  <mihaip@chromium.org>
30619
30620         Reviewed by Ojan Vafai.
30621
30622         Update rebaseline queue server to handle hierarchical test results
30623         https://bugs.webkit.org/show_bug.cgi?id=60063
30624
30625         Update test result parsing done by the rebaseline queue server to handle
30626         the hierarchical test result output that is generated as of r85254.
30627
30628         * RebaselineQueueServer/templates/builder-queue-edit.html:
30629
30630 2011-05-03  Eric Seidel  <eric@webkit.org>
30631
30632         Reviewed by Adam Roben.
30633
30634         webkit-patch post-commits HEAD^.. throws an exception
30635         https://bugs.webkit.org/show_bug.cgi?id=60027
30636
30637         * Scripts/webkitpy/common/checkout/scm/git.py:
30638
30639 2011-05-03  Luiz Agostini  <luiz.agostini@openbossa.org>
30640
30641         Reviewed by Kenneth Rohde Christiansen.
30642
30643         [Qt] Layout tests may timeout if loadFinished receives false
30644         https://bugs.webkit.org/show_bug.cgi?id=60064
30645
30646         When using layoutTestController.queueLoad(), if QWebPage::loadFinished(bool) receives false
30647         then the layout test may timout. It happens that QWebPage::loadFinished() receives false when
30648         a response is ignored, like it should be for 204 responses. It is a problem for creating
30649         tests for 204 responses for example.
30650
30651         This change has been tested and it causes no regressions in layout tests.
30652
30653         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
30654         (LayoutTestController::maybeDump):
30655
30656 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
30657
30658         Unreviewed, rolling out r85659.
30659         http://trac.webkit.org/changeset/85659
30660         https://bugs.webkit.org/show_bug.cgi?id=59977
30661
30662         restarting isn't working correctly
30663
30664         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
30665         * Scripts/webkitpy/layout_tests/port/http_server.py:
30666         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
30667         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
30668
30669 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
30670
30671         Reviewed by Ojan Vafai.
30672
30673         new-run-webkit-tests: fix http server startup/shutdown
30674
30675         Previous versions of the code had three problems that made startup
30676         and shutdown flaky. The first is that it would throw exceptions
30677         if it couldn't delete stale log files, which was overly
30678         paranoid. The second is that some of the exceptions weren't
30679         defined properly. The third, and most important, is that it was
30680         using urllib to check if ports were available, which was leaving
30681         sockets in a half-closed state, and keeping ports from being
30682         reused. By switching to raw sockets, we are able to now reliably
30683         restart.
30684
30685         This change also switches the code to using Executives to stop
30686         processes, which will let us delete a bunch of code in the
30687         port/* implementations and fix a weird layering problem in a
30688         subsequent patch.
30689
30690         https://bugs.webkit.org/show_bug.cgi?id=59977
30691
30692         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
30693         * Scripts/webkitpy/layout_tests/port/http_server.py:
30694         * Scripts/webkitpy/layout_tests/port/http_server_base.py:
30695         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
30696
30697 2011-05-03  Dirk Pranke  <dpranke@chromium.org>
30698
30699         Reviewed by Adam Barth.
30700
30701         fix failing printing_unittest on win32
30702         https://bugs.webkit.org/show_bug.cgi?id=59963
30703
30704         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
30705
30706 2011-05-03  Adam Roben  <aroben@apple.com>
30707
30708         Fix most strict PassOwnPtr violations on Windows
30709
30710         Fixes <http://webkit.org/b/60025> Windows should (almost) build with strict PassOwnPtr
30711         enabled
30712
30713         Reviewed by Anders Carlsson.
30714
30715         * WebKitAPITest/Test.h: Use adoptPtr.
30716
30717 2011-05-03  Caio Marcelo de Oliveira Filho  <caio.oliveira@openbossa.org>
30718
30719         Reviewed by Andreas Kling.
30720
30721         Add myself to the committers list.
30722
30723         * Scripts/webkitpy/common/config/committers.py:
30724
30725 2011-05-03  James Robinson  <jamesr@chromium.org>
30726
30727         Reviewed by Kenneth Russell.
30728
30729         [chromium] Layout test animations/3d/replace-filling-transform.html failing
30730         https://bugs.webkit.org/show_bug.cgi?id=59839
30731
30732         Add animations/3d to the list of tests run in the chromium-gpu configuration since these tests exercise the compositor.
30733
30734         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
30735
30736 2011-05-03  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
30737
30738         Unreviewed, rolling out r85471.
30739         http://trac.webkit.org/changeset/85471
30740         https://bugs.webkit.org/show_bug.cgi?id=59922
30741
30742         The rule of thumb in EFL is that you init in all the places
30743         you use it. So, if in your application you call ecore_evas_*
30744         functions, you have to call ecore_evas_init() first.
30745
30746         * EWebLauncher/main.c: always call *_init() functions when you use
30747         that namespace.
30748         (quit):
30749         (main):
30750
30751 2011-05-02  Ojan Vafai  <ojan@chromium.org>
30752
30753         Reviewed by Tony Chang.
30754
30755         remove no longer used method
30756         https://bugs.webkit.org/show_bug.cgi?id=59981
30757
30758         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
30759         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
30760
30761 2011-05-02  Ojan Vafai  <ojan@chromium.org>
30762
30763         Reviewed by Tony Chang.
30764
30765         add image toggling behavior from old-run-webkit-tests
30766         https://bugs.webkit.org/show_bug.cgi?id=59978
30767
30768         Output -diffs.html files. This matches old-run-webkit-tests.
30769         In theory, we could use data urls in the results.html file, but
30770         data urls don't currently have access to local files.
30771
30772         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
30773
30774 2011-05-02  Tony Chang  <tony@chromium.org>
30775
30776         Reviewed by Ojan Vafai.
30777
30778         [nrwt] remove some obsolete references to .checksum files
30779         https://bugs.webkit.org/show_bug.cgi?id=59971
30780
30781         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
30782         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
30783         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
30784         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
30785         * Scripts/webkitpy/layout_tests/port/test.py:
30786         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
30787
30788 2011-05-02  Adam Roben  <aroben@apple.com>
30789
30790         Take advantage of implicit nullptr_t -> PassOwnPtr conversion
30791
30792         Fixes <http://webkit.org/b/59964> Implicit conversion from std::nullptr_t to PassOwnPtr
30793         doesn't work, but should
30794
30795         Reviewed by Adam Barth.
30796
30797         * DumpRenderTree/chromium/CppBoundClass.h:
30798         * DumpRenderTree/chromium/WebViewHost.cpp:
30799
30800 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
30801
30802         Reviewed by Xan Lopez.
30803
30804         [GTK] Compiler warning in Tools/MiniBrowser/main.c
30805         https://bugs.webkit.org/show_bug.cgi?id=59986
30806
30807         Remove a compiler warning.
30808
30809         * MiniBrowser/gtk/main.c: Use full array initializers to prevent some compilers
30810         from complaining.
30811
30812 2011-05-02  Adam Barth  <abarth@webkit.org>
30813
30814         Reviewed by Eric Seidel.
30815
30816         Add runtests_unittest.py
30817         https://bugs.webkit.org/show_bug.cgi?id=59955
30818
30819         I forgot to add this file in a previous patch.
30820
30821         * Scripts/webkitpy/tool/steps/runtests_unittest.py: Added.
30822
30823 2011-05-02  Adam Barth  <abarth@webkit.org>
30824
30825         Change the testing EWS to log failures explicitly to avoid an infinite loop.
30826
30827         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
30828
30829 2011-05-02  Adrienne Walker  <enne@google.com>
30830
30831         Unreviewed, fix misspelled contributor name.
30832
30833         * Scripts/webkitpy/common/config/committers.py:
30834
30835 2011-05-02  Adam Barth  <abarth@webkit.org>
30836
30837         Reviewed by Eric Seidel.
30838
30839         cr-linux-ews should run tests
30840         https://bugs.webkit.org/show_bug.cgi?id=59272
30841
30842         This patch turns on testing for the cr-linux-ews.  Yay!
30843
30844         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py:
30845         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
30846         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
30847         * Scripts/webkitpy/tool/steps/runtests.py:
30848         * Scripts/webkitpy/tool/steps/steps_unittest.py:
30849
30850 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
30851
30852         Incorporate a review comment that I mistakenly omitted from my last commit.
30853
30854         * DumpRenderTree/gtk/EventSender.cpp: Use unsigned here instead of unsigned int.
30855
30856 2011-05-02  Adam Barth  <abarth@webkit.org>
30857
30858         Reviewed by Eric Seidel.
30859
30860         EWS should report the list of unexpected failures
30861         https://bugs.webkit.org/show_bug.cgi?id=59923
30862
30863         We cache the results of the first layout test results run on the task
30864         so the EWS can grab them later.  Another option is to include them in
30865         the exception that's thrown, but this approach seemed slightly more
30866         aesthetic.
30867
30868         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
30869         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
30870
30871 2011-05-02  Adam Barth  <abarth@webkit.org>
30872
30873         Reviewed by Eric Seidel.
30874
30875         Add base case for a test-running EWS
30876         https://bugs.webkit.org/show_bug.cgi?id=59920
30877
30878         Later I will subclass this to run the Chromium Linux EWS.
30879
30880         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Added.
30881         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
30882         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
30883
30884 2011-05-02  Anders Carlsson  <andersca@apple.com>
30885
30886         Build fix.
30887
30888         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
30889         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
30890
30891 2011-05-02  Nico Weber  <thakis@chromium.org>
30892
30893         Reviewed by Mihai Parparita.
30894
30895         Add --no-webkit2 to the output of "build-webkit -h"
30896         https://bugs.webkit.org/show_bug.cgi?id=59942
30897
30898         * Scripts/build-webkit:
30899
30900 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
30901
30902         Reviewed by Xan Lopez.
30903
30904         [GTK] old-run-webkit-tests should disable the Ubuntu overlay scrollbar
30905         https://bugs.webkit.org/show_bug.cgi?id=59943
30906
30907         Disable the Ubuntu overlay scrollbar while testing, by injecting the
30908         LIBOVERLAY_SCROLLBAR=0 environment variable.
30909
30910         * Scripts/old-run-webkit-tests:
30911
30912 2011-05-02  Martin Robinson  <mrobinson@igalia.com>
30913
30914         Reviewed by Eric Seidel.
30915
30916         [GTK] editing/selection/select-out-of-editable.html failing after r83967
30917         https://bugs.webkit.org/show_bug.cgi?id=58667
30918
30919         Do not trigger mouse down events when the EventSender knows that a mouse button is
30920         already down. These type of situations can confuse Xvfb and cause dragging failures.
30921
30922         * DumpRenderTree/gtk/EventSender.cpp:
30923         (mouseDownCallback): Do not trigger another mouse button down event if the mouse button
30924         is already down. This can interfere with drags in Xvfb.
30925
30926 2011-05-02  Zsolt Fehér  <h490267@stud.u-szeged.hu>
30927
30928         Reviewed by Csaba Osztrogonác.
30929
30930         DRT sideeffect revealed by r71268.
30931         https://bugs.webkit.org/show_bug.cgi?id=48958
30932
30933         Reset QWebSettings::ZoomTextOnly attribute before all tests.
30934
30935         Call setTextSizeMultiplier(1.0) to reset TextZoomFactor and
30936         PageZoomFactor too. It should be done before resetsettings()
30937         to guarantee resetting QWebSettings::ZoomTextOnly correctly.
30938
30939         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
30940         (WebCore::WebPage::resetSettings):
30941         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
30942
30943 2011-05-02  Sheriff Bot  <webkit.review.bot@gmail.com>
30944
30945         Unreviewed, rolling out r85469.
30946         http://trac.webkit.org/changeset/85469
30947         https://bugs.webkit.org/show_bug.cgi?id=59928
30948
30949         Broke  webkitpy-tests on Windows 7, GTK and Qt. (Requested by
30950         bbandix on #webkit).
30951
30952         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Removed.
30953         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
30954         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
30955
30956 2011-05-02  Tomasz Morawski  <t.morawski@samsung.com>
30957
30958         Reviewed by Kenneth Rohde Christiansen.
30959
30960         Cleanup EFL initialization/shutdown in the EWebLauncher
30961         https://bugs.webkit.org/show_bug.cgi?id=59922
30962
30963         Removed ecore_evas_init and edje_init form main function due to they
30964         are also called in the ewk_init. Moved ewk_shutdown to quit function.
30965
30966         * EWebLauncher/main.c:
30967         (quit):
30968         (main):
30969
30970 2011-05-02  Adam Barth  <abarth@webkit.org>
30971
30972         Reviewed by Eric Seidel.
30973
30974         Add base case for a test-running EWS
30975         https://bugs.webkit.org/show_bug.cgi?id=59920
30976
30977         Later I will subclass this to run the Chromium Linux EWS.
30978
30979         * Scripts/webkitpy/tool/bot/earlywarningsystemtask.py: Added.
30980         * Scripts/webkitpy/tool/commands/earlywarningsystem.py:
30981         * Scripts/webkitpy/tool/commands/earlywarningsystem_unittest.py:
30982
30983 2011-05-02  Adam Barth  <abarth@webkit.org>
30984
30985         Reviewed by Eric Seidel.
30986
30987         Add the ability to skip python and perl unit tests on EC2
30988         https://bugs.webkit.org/show_bug.cgi?id=59918
30989
30990         * Scripts/webkitpy/common/config/ports.py:
30991             - This is slightly ugly, but it's useful...
30992         * Scripts/webkitpy/common/config/ports_unittest.py:
30993         * Scripts/webkitpy/tool/mocktool.py:
30994         * Scripts/webkitpy/tool/steps/runtests.py:
30995
30996 2011-05-02  Adam Barth  <abarth@webkit.org>
30997
30998         Reviewed by Eric Seidel.
30999
31000         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
31001
31002 2011-04-29  Kevin Ollivier  <kevino@theolliviers.com>
31003
31004         [wx] Unreviewed build fix, add new directory to the build.
31005
31006         * waf/build/settings.py:
31007
31008 2011-04-29  Carlos Garcia Campos  <cgarcia@igalia.com>
31009
31010         Reviewed by Martin Robinson.
31011
31012         [GTK] Implement UI client in minibrowser
31013         https://bugs.webkit.org/show_bug.cgi?id=58416
31014
31015         It also allows to open more than one minibrowser window from the
31016         command line.
31017
31018         * MiniBrowser/gtk/BrowserWindow.c:
31019         (browserWindowFinalize):
31020         (browser_window_init):
31021         (browserWindowConstructed):
31022         (createNewPage):
31023         (showPage):
31024         (closePage):
31025         (runJavaScriptAlert):
31026         (browserWindowUIClientInit):
31027         * MiniBrowser/gtk/main.c:
31028         (loadURI):
31029         (main):
31030
31031 2011-04-29  Carlos Garcia Campos  <cgarcia@igalia.com>
31032
31033         Reviewed by Martin Robinson.
31034
31035         [GTK] Implement page loader client in MiniBrowser
31036         https://bugs.webkit.org/show_bug.cgi?id=58024
31037
31038         A new class BrowserWindow has been added to implement the page
31039         loader client. It makes supporting multiple windows easier. Window
31040         title and url entry are now updated during the loading process.
31041
31042         * MiniBrowser/gtk/BrowserWindow.c: Added.
31043         (activateUriEntryCallback):
31044         (goBackCallback):
31045         (goForwardCallback):
31046         (browserWindowFinalize):
31047         (browserWindowGetProperty):
31048         (browserWindowSetProperty):
31049         (browser_window_init):
31050         (browserWindowConstructed):
31051         (browser_window_class_init):
31052         (browserWindowUpdateTitle):
31053         (browserWindowSetTitle):
31054         (browserWindowSetLoadProgress):
31055         (browserWindowUpdateURL):
31056         (didStartProvisionalLoadForFrame):
31057         (didReceiveServerRedirectForProvisionalLoadForFrame):
31058         (didFailProvisionalLoadWithErrorForFrame):
31059         (didCommitLoadForFrame):
31060         (didFinishDocumentLoadForFrame):
31061         (didFinishLoadForFrame):
31062         (didFailLoadWithErrorForFrame):
31063         (didReceiveTitleForFrame):
31064         (didFirstLayoutForFrame):
31065         (didFirstVisuallyNonEmptyLayoutForFrame):
31066         (didRemoveFrameFromHierarchy):
31067         (didStartProgress):
31068         (didChangeProgress):
31069         (didFinishProgress):
31070         (didBecomeUnresponsive):
31071         (didBecomeResponsive):
31072         (browserWindowLoaderClientInit):
31073         (browser_window_new):
31074         (browser_window_get_view):
31075         * MiniBrowser/gtk/BrowserWindow.h: Added.
31076         * MiniBrowser/gtk/GNUmakefile.am:
31077         * MiniBrowser/gtk/main.c:
31078         (createWindow):
31079
31080 2011-04-28  Stephanie Lewis  <slewis@apple.com>
31081
31082         Reviewed by Mark Rowe.
31083
31084         http://bugs.webkit.org/show_bug.cgi?id=59743
31085         <rdar://problem/9040413> Add an install target for production builds.
31086
31087         * DumpRenderTree/mac/PerlSupport/Makefile:
31088
31089 2011-04-28  Adam Roben  <aroben@apple.com>
31090
31091         Stop and restart ATSServer before each test run on Leopard
31092
31093         This is to work around an ATSServer leak that eventually causes tests
31094         to fail.
31095
31096         Fixes <http://webkit.org/b/51807> <rdar://problem/9273894>
31097         fast/blockflow/broken-ideograph-[small-caps|font].html sometimes fails on Leopard buildbots
31098         (due to ATSServer leak)
31099
31100         Reviewed by Dan Bernstein.
31101
31102         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
31103         (StopATSServer):
31104         (StartATSServer):
31105         Added these new commands to stop and start ATSServer.
31106
31107         (TestFactory.__init__): Stop and start ATSServer on Leopard before running tests.
31108
31109 2011-04-28  Dirk Pranke  <dpranke@chromium.org>
31110
31111         Reviewed by Adam Roben.
31112
31113         add Leopard NRWT bot to the scheduler
31114         https://bugs.webkit.org/show_bug.cgi?id=59766
31115
31116         * BuildSlaveSupport/build.webkit.org-config/config.json:
31117
31118 2011-04-28  Sam Weinig  <sam@webkit.org>
31119
31120         Fix WebKitTestRunner build.
31121
31122         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
31123         The Derived Sources target needs a product name.
31124
31125 2011-04-28  Sam Weinig  <sam@webkit.org>
31126
31127         Reviewed by Mark Rowe.
31128
31129         Add install support for WebKitTestRunner.
31130         <rdar://problem/9349055>
31131
31132         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
31133         Add missing CompilerVersion.xcconfig.
31134
31135         * WebKitTestRunner/Configurations/InjectedBundle.xcconfig:
31136         * WebKitTestRunner/Configurations/WebKitTestRunner.xcconfig:
31137         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
31138         Add install path and tell it not to skip installation.
31139
31140         Add missing CompilerVersion.xcconfig.
31141
31142 2011-04-28  David Levin  <levin@chromium.org>
31143
31144         Reviewed by Adam Barth.
31145
31146         Remove IMAGE_RESIZER related code.
31147         https://bugs.webkit.org/show_bug.cgi?id=59735
31148
31149         * Scripts/build-webkit:
31150
31151 2011-04-28  Sam Weinig  <sam@webkit.org>
31152
31153         Reviewed by Adam Roben.
31154
31155         Make reading skipped files from additional platform
31156         directories work in WebKit2 mode.
31157
31158         * Scripts/old-run-webkit-tests:
31159         (buildPlatformTestHierarchy): Simplify logic to always use the first
31160         and last platforms in the build hierarchy as the test hierarchy, and to
31161         insert the WebKit2 platform in the middle if it exists. Yum.
31162
31163 2011-04-28  Adam Barth  <abarth@webkit.org>
31164
31165         Reviewed by Eric Seidel.
31166
31167         Remove WML
31168         https://bugs.webkit.org/show_bug.cgi?id=59678
31169
31170         Remove build system integration and test harness integration for WML.
31171
31172         * Scripts/build-webkit:
31173         * Scripts/old-run-webkit-tests:
31174         * Scripts/webkitperl/features.pm:
31175         * Scripts/webkitpy/layout_tests/port/webkit.py:
31176         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
31177
31178 2011-04-28  Ojan Vafai  <ojan@chromium.org>
31179
31180         Reviewed by Mihai Parparita.
31181
31182         use a hierarchical data structure to store results json
31183         https://bugs.webkit.org/show_bug.cgi?id=59736
31184
31185         Instead of a map from test name to results dict, we now store a hierarchical
31186         data structure.
31187
31188         Old: { foo/bar/baz.html: results_dict, foo/bar/baz2.html: results_dict }
31189         New:
31190         { foo: {
31191             bar: {
31192               baz.html: results_dict,
31193               baz2.html: results_dict } } }
31194
31195         This cut ~52% of the size of times_ms.json. I expect it will cut
31196         30-40% of full_results.json and unexpected_results.json.
31197
31198         * Scripts/webkitpy/common/net/resultsjsonparser.py:
31199         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
31200         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
31201         Using os.sep was just wrong since relative_test_filename canonicalizes separators.
31202         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
31203         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
31204         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
31205         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
31206         Instead of plumbing through the change to all of the rebaselineserver
31207         python and JS, just convert to the old format after reading in the json.
31208
31209 2011-04-28  Eric Seidel  <eric@webkit.org>
31210
31211         Reviewed by Adam Barth.
31212
31213         EWS Bubbles should show commit-queue position when patch is in the commit-queue
31214         https://bugs.webkit.org/show_bug.cgi?id=59744
31215
31216         Add target="_top" to the bubble links.  This was
31217         forgotten when I moved from using an onclick to a normal link.
31218
31219         * QueueStatusServer/templates/statusbubble.html:
31220
31221 2011-04-28  Eric Seidel  <eric@webkit.org>
31222
31223         Reviewed by Adam Barth.
31224
31225         EWS bubbles should show "submit to ews" button if they've not been processed
31226         https://bugs.webkit.org/show_bug.cgi?id=59751
31227
31228         * QueueStatusServer/handlers/statusbubble.py:
31229         * QueueStatusServer/handlers/submittoews.py:
31230         * QueueStatusServer/templates/statusbubble.html:
31231         * QueueStatusServer/templates/submittoews.html:
31232
31233 2011-04-28  Eric Seidel  <eric@webkit.org>
31234
31235         Reviewed by Adam Barth.
31236
31237         EWS Bubbles should show commit-queue position when patch is in the commit-queue
31238         https://bugs.webkit.org/show_bug.cgi?id=59744
31239
31240         This was actually just a 2 line change to statusbubble.py.
31241         I also cleaned up statusbubble.html a little while I was there
31242         (the links are now actual <a> tags and have hover).
31243
31244         * QueueStatusServer/handlers/statusbubble.py:
31245         * QueueStatusServer/templates/statusbubble.html:
31246
31247 2011-04-28  Dirk Pranke  <dpranke@chromium.org>
31248
31249         Reviewed by Eric Seidel.
31250
31251         add a mac leopard release NRWT bot
31252         https://bugs.webkit.org/show_bug.cgi?id=59733
31253
31254         * BuildSlaveSupport/build.webkit.org-config/config.json:
31255
31256 2011-04-28  Ojan Vafai  <ojan@chromium.org>
31257
31258         Avoid crashing when running new-run-webkit-tests outside of
31259         version control.
31260
31261         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
31262
31263 2011-04-28  Xan Lopez  <xlopez@igalia.com>
31264
31265         Unreviewed build fix.
31266
31267         * GNUmakefile.am:
31268
31269 2011-04-28  Ojan Vafai  <ojan@chromium.org>
31270
31271         Reviewed by Eric Seidel.
31272
31273         fix python printing test
31274         https://bugs.webkit.org/show_bug.cgi?id=59719
31275
31276         Now that we run scm commands in summarize_results, --verbose prints
31277         debug logging to stderr.
31278
31279         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
31280
31281 2011-04-28  Ojan Vafai  <ojan@chromium.org>
31282
31283         Reviewed by Tony Chang.
31284
31285         link to tests at a specific revision in the results.html file
31286         https://bugs.webkit.org/show_bug.cgi?id=59711
31287
31288         * Scripts/webkitpy/common/checkout/scm.py:
31289         Add a method to get the head svn revision.
31290         * Scripts/webkitpy/common/checkout/scm_unittest.py:
31291         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
31292         Include the svn revision in the results json.
31293
31294 2011-04-28  Carlos Garcia Campos  <cgarcia@igalia.com>
31295
31296         Reviewed by Sam Weinig.
31297
31298         [GTK] Remove WKURLCreateWithURL()
31299         https://bugs.webkit.org/show_bug.cgi?id=59691
31300
31301         Use WKURLCreateWithUTF8CString() instead of WKURLCreateWithURL().
31302
31303         * MiniBrowser/gtk/main.c:
31304         (activateUriEntryCallback):
31305         (main):
31306
31307 2011-04-28  Carlos Garcia Campos  <cgarcia@igalia.com>
31308
31309         Reviewed by Martin Robinson.
31310
31311         [GTK] WKView should be a GtkWidget
31312         https://bugs.webkit.org/show_bug.cgi?id=59603
31313
31314         Don't use WKViewGetWindow(), use WKView as a GtkWidget instead.
31315
31316         * MiniBrowser/gtk/main.c:
31317         (createWindow):
31318         (main):
31319
31320 2011-04-28  Gabor Rapcsanyi  <rgabor@webkit.org>
31321
31322         Reviewed by Csaba Osztrogonác.
31323
31324         [Qt] Add ARM specific Skipped list
31325         https://bugs.webkit.org/show_bug.cgi?id=59680
31326
31327         * Scripts/old-run-webkit-tests:
31328         * Scripts/webkitdirs.pm:
31329
31330 2011-04-27  Eric Seidel  <eric@webkit.org>
31331
31332         Reviewed by Adam Barth.
31333
31334         error_log.txt should include PHP errors
31335         https://bugs.webkit.org/show_bug.cgi?id=59643
31336
31337         Turns out this was really easy after I finally learned
31338         enough about PHP to do it...
31339
31340         I'm adding this in hopes of better tracking down
31341         our random http failures.
31342
31343         * Scripts/webkitperl/httpd.pm:
31344
31345 2011-04-27  David Levin  <levin@chromium.org>
31346
31347         Reviewed by Adam Barth.
31348
31349         check-webkit-style should flags spacing errors for braces on the same line.
31350         https://bugs.webkit.org/show_bug.cgi?id=59665
31351
31352         * Scripts/webkitpy/style/checkers/cpp.py: Added the check.
31353         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added some tests for the check
31354           and fixed some test cases to not trigger this error since it isn't what they are verifying.
31355
31356 2011-04-27  David Levin  <levin@chromium.org>
31357
31358         Reviewed by Adam Barth.
31359
31360         check-webkit-style error for spaces after periods in a comment has a bad error
31361         message and warns about end of line spaces.
31362         https://bugs.webkit.org/show_bug.cgi?id=59663
31363
31364         * Scripts/webkitpy/style/checkers/cpp.py: Fix both issues.
31365         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests for spaces at the end of line
31366         and changed the error messages.
31367
31368 2011-04-27  Ojan Vafai  <ojan@chromium.org>
31369
31370         remove assertion now that platform/mac-tiger no longer exists
31371         https://bugs.webkit.org/show_bug.cgi?id=59660
31372
31373         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
31374
31375 2011-04-27  Ojan Vafai  <ojan@chromium.org>
31376
31377         Reviewed by Eric Seidel.
31378
31379         remove test that didn't make sense
31380         https://bugs.webkit.org/show_bug.cgi?id=59658
31381
31382         It was artifically saying the test wasn't expected to pass,
31383         but the expectations string was empty. The case of unexpected
31384         passes with a real expectations string is tested later in the
31385         same test.
31386
31387         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
31388
31389 2011-04-27  Sam Weinig  <sam@webkit.org>
31390
31391         Reviewed by Adam Roben.
31392
31393         WebKit2 needs to notify the client when a plugin fails to initialize
31394         https://bugs.webkit.org/show_bug.cgi?id=59657
31395
31396         * MiniBrowser/mac/BrowserWindowController.m:
31397         (-[BrowserWindowController awakeFromNib]):
31398         * WebKitTestRunner/TestController.cpp:
31399         (WTR::TestController::initialize):
31400         Add stubs.
31401
31402 2011-04-26  Sam Weinig  <sam@webkit.org>
31403
31404         Reviewed by David Levin.
31405
31406         Add a few more missing options to new-run-webkit-tests
31407         https://bugs.webkit.org/show_bug.cgi?id=59567
31408
31409         Add support for:
31410             --threaded
31411             --gc-between-tests
31412             --complex-text
31413
31414         * Scripts/webkitpy/layout_tests/port/webkit.py:
31415         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
31416         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
31417
31418 2011-04-27  Mark Rowe  <mrowe@apple.com>
31419
31420         Build fix.
31421
31422         Fix an OwnPtr-related issue that was missed before the strictness was turned on.
31423
31424         * WebKitTestRunner/TestController.cpp:
31425         (WTR::TestController::runTest):
31426
31427 2011-04-27  Ojan Vafai  <ojan@chromium.org>
31428
31429         Forgot to upload the new times_ms.json file after splitting
31430         it out from full_results.json.
31431
31432         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
31433
31434 2011-04-27  Ojan Vafai  <ojan@chromium.org>
31435
31436         Fix snafu in generating the JSON for the chromium tests.
31437
31438         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
31439
31440 2011-04-27  Ojan Vafai  <ojan@chromium.org>
31441
31442         Fix bad merge. This broke the chromium tests generating JSON files.
31443         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
31444
31445 2011-04-27  Mark Rowe  <mrowe@apple.com>
31446
31447         Rubber-stamped by Geoff Garen.
31448
31449         Don't look for results in a location that no longer exists.
31450
31451         * Scripts/old-run-webkit-tests:
31452
31453 2011-04-27  Adrienne Walker  <enne@google.com>
31454
31455         Unreviewed, add Chrome GPU folks to contributors that I'm tired of manually CCing
31456
31457         * Scripts/webkitpy/common/config/committers.py:
31458
31459 2011-04-22  Ojan Vafai  <ojan@chromium.org>
31460
31461         Reviewed by Tony Chang.
31462
31463         only include failures in full_results.json
31464         https://bugs.webkit.org/show_bug.cgi?id=59256
31465
31466         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
31467
31468 2011-04-22  Ojan Vafai  <ojan@chromium.org>
31469
31470         Reviewed by Adam Roben.
31471
31472         move times into their own json file
31473         https://bugs.webkit.org/show_bug.cgi?id=59239
31474
31475         The test runtimes are the only case where we need an entry for
31476         every test that was run, including passing tests. Move them into
31477         their own json file so we can prune the other json files to only
31478         include failing tests.
31479
31480         Use a trie-like data structure to hold the tests to avoid including
31481         full test paths. This cuts 52% of the filesize. The end result is ~700k
31482         pre-gzip.
31483
31484         * Scripts/webkitpy/common/net/resultsjsonparser.py:
31485         Exclude times from the parsed TestResults.
31486         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py:
31487         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
31488         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
31489         * Scripts/webkitpy/layout_tests/layout_package/manager.py:
31490
31491 2011-04-27  Anders Carlsson  <andersca@apple.com>
31492
31493         More casts and more clang++ build fixes.
31494
31495         * DumpRenderTree/mac/DumpRenderTree.mm:
31496         (createWebViewAndOffscreenWindow):
31497         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
31498         (TestWebKitAPI::PlatformWebView::PlatformWebView):
31499         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
31500         (WTR::PlatformWebView::PlatformWebView):
31501
31502 2011-04-27  Tony Chang  <tony@chromium.org>
31503
31504         Reviewed by Ojan Vafai.
31505
31506         lazily autoinstall thirdparty python libraries
31507         https://bugs.webkit.org/show_bug.cgi?id=55723
31508
31509         This helps by only having pywebsocket installed when the bots run.  It
31510         doesn't help much for devs since webkit-patch pulls in almost everything.
31511
31512         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
31513         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
31514         * Scripts/webkitpy/common/net/networktransaction.py:
31515         * Scripts/webkitpy/common/net/networktransaction_unittest.py:
31516         * Scripts/webkitpy/common/net/statusserver.py:
31517         * Scripts/webkitpy/layout_tests/port/websocket_server.py:
31518         * Scripts/webkitpy/thirdparty/__init__.py: Add an import hook so we can install based on
31519             the import command rather than just installing everything.
31520         * Scripts/webkitpy/thirdparty/__init___unittest.py: Added.
31521
31522 2011-04-27  Girish Ramakrishnan  <girish@forwardbias.in>
31523
31524         Add my IRC nick.
31525
31526         * Scripts/webkitpy/common/config/committers.py:
31527
31528 2011-04-27  Eric Seidel  <eric@webkit.org>
31529
31530         Reviewed by Mihai Parparita.
31531
31532         sherrifbot create-bug shouldn't assign bugs to webkit.review.bot
31533         https://bugs.webkit.org/show_bug.cgi?id=59545
31534
31535         To do this, I needed a way to look up contributors by irc-name
31536         (since anyone in #webkit who might use this command may not be a committer).
31537         To lookup contributors, I had to make Contributor a real object.
31538         Which led me to redesign parts of committers.py...
31539         and finally fix one spot in changelog.py where we wanted to be
31540         looking up contributors and not committers.
31541
31542         Overall a pretty simple fix, once you wade through the yak-hair.
31543
31544         This may not prevent *all* possible ways that bugs would get assigned
31545         to webkit.review.bot.  If we don't recognize the requester we will
31546         go through the previous code path (which shouldn't change the assignee
31547         on the bug from the default as far as I can tell).
31548
31549         * Scripts/webkitpy/common/checkout/changelog.py:
31550         * Scripts/webkitpy/common/config/committers.py:
31551         * Scripts/webkitpy/common/config/committers_unittest.py:
31552         * Scripts/webkitpy/tool/bot/irc_command.py:
31553         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
31554
31555 2011-04-27  Yi Shen  <yi.4.shen@nokia.com>
31556
31557         Reviewed by Kenneth Rohde Christiansen.
31558
31559         [Qt] Upstream Symbian platform plugin
31560         https://bugs.webkit.org/show_bug.cgi?id=58435
31561
31562         Exempting directory WebKit/qt/symbian/platformplugin from style guide.
31563
31564         * Scripts/webkitpy/style/checker.py:
31565         * Scripts/webkitpy/style/checker_unittest.py:
31566
31567 2011-04-27  Jeff Miller  <jeffm@apple.com>
31568
31569         Correct my IRC nickname.
31570
31571         * Scripts/webkitpy/common/config/committers.py:
31572
31573 2011-04-27  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
31574
31575         Reviewed by Darin Adler.
31576
31577         Change the email used on bugzilla
31578         https://bugs.webkit.org/show_bug.cgi?id=59582
31579
31580         * Scripts/webkitpy/common/config/committers.py:
31581
31582 2011-04-27  James Kozianski  <koz@chromium.org>
31583
31584         Reviewed by Ojan Vafai.
31585
31586         Extract map from port to builder name.
31587         https://bugs.webkit.org/show_bug.cgi?id=58301
31588
31589         * Scripts/webkitpy/layout_tests/port/__init__.py:
31590         * Scripts/webkitpy/layout_tests/port/builders.py: Added.
31591         * Scripts/webkitpy/layout_tests/port/builders_unittest.py: Added.
31592         * Scripts/webkitpy/layout_tests/port/factory.py:
31593
31594 2011-04-27  Ryuan Choi  <ryuan.choi@samsung.com>
31595
31596         Reviewed by Martin Robinson.
31597
31598         [GTK] Add proxy support to GtkLauncher
31599         https://bugs.webkit.org/show_bug.cgi?id=58852
31600
31601         Add SOUP_TYPE_PROXY_RESOLVER_DEFAULT feature or set SOUP_SESSION_PROXY_URI for developers behind proxy.
31602
31603         * GNUmakefile.am:
31604         * GtkLauncher/main.c:
31605         (main):
31606
31607 2011-04-27  Annie Sullivan  <sullivan@chromium.org>
31608
31609         Reviewed by David Levin.
31610
31611         Adding myself as a Chromium contributor to committers.py.
31612
31613         https://bugs.webkit.org/show_bug.cgi?id=59576
31614
31615         * Scripts/webkitpy/common/config/committers.py:
31616
31617 2011-04-26  David Grogan  <dgrogan@chromium.org>
31618
31619         Reviewed by David Levin.
31620
31621         Adding myself to Chromium contributors list in committers.py
31622         https://bugs.webkit.org/show_bug.cgi?id=59565
31623
31624         * Scripts/webkitpy/common/config/committers.py:
31625
31626 2011-04-26  Chang Shu  <cshu@webkit.org>
31627
31628         Reviewed by Laszlo Gombos.
31629
31630         WebKitTestRunner needs layoutTestController.addOriginAccessWhitelistEntry
31631         https://bugs.webkit.org/show_bug.cgi?id=42544
31632
31633         Also implemented removeOriginAccessWhitelistEntry and resetOriginAccessWhitelists.
31634
31635         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
31636         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
31637         (WTR::InjectedBundle::beginTesting):
31638         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
31639         (WTR::LayoutTestController::addOriginAccessWhitelistEntry):
31640         (WTR::LayoutTestController::removeOriginAccessWhitelistEntry):
31641         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
31642
31643 2011-04-26  Chris Fleizach  <cfleizach@apple.com>
31644
31645         Reviewed by Adele Peterson.
31646
31647         AX: Add more text marker routines to DRT
31648         https://bugs.webkit.org/show_bug.cgi?id=58949
31649
31650         Add a few more methods to help test the text marker system, including nextMarker, previousMarker and
31651         stringForMarkerRange.
31652
31653         * DumpRenderTree/AccessibilityUIElement.cpp:
31654         (nextTextMarkerCallback):
31655         (previousTextMarkerCallback):
31656         (stringForTextMarkerRangeCallback):
31657         (AccessibilityUIElement::previousTextMarker):
31658         (AccessibilityUIElement::nextTextMarker):
31659         (AccessibilityUIElement::stringForTextMarkerRange):
31660         (AccessibilityUIElement::getJSClass):
31661         * DumpRenderTree/AccessibilityUIElement.h:
31662         * DumpRenderTree/mac/AccessibilityUIElementMac.mm:
31663         (AccessibilityUIElement::previousTextMarker):
31664         (AccessibilityUIElement::nextTextMarker):
31665         (AccessibilityUIElement::stringForTextMarkerRange):
31666
31667 2011-04-26  Dan Bernstein  <mitz@apple.com>
31668
31669         Build fix.
31670
31671         * DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig:
31672         * MiniBrowser/Configurations/CompilerVersion.xcconfig:
31673         * WebKitTestRunner/Configurations/CompilerVersion.xcconfig:
31674
31675 2011-04-26  Kevin Ollivier  <kevino@theolliviers.com>
31676
31677         Rubberstamped by Eric Seidel.
31678
31679         Enable waf to be used to build other ports
31680         https://bugs.webkit.org/show_bug.cgi?id=58213
31681
31682         * Scripts/build-webkit:
31683         * Scripts/webkitdirs.pm:
31684         * waf: Added.
31685         * waf/build: Added.
31686         * waf/build/build_utils.py: Added.
31687         * waf/build/settings.py: Added.
31688         * waf/build/waf_extensions.py: Added.
31689         * waf/build/wxpresets.py: Added.
31690
31691 2011-04-26  Sam Weinig  <sam@webkit.org>
31692
31693         Reviewed by David Hyatt.
31694
31695         Remove Datagrid from the tree
31696         https://bugs.webkit.org/show_bug.cgi?id=59543
31697
31698         * Scripts/build-webkit:
31699         Remove feature.
31700
31701         * iExploder/iexploder-1.3.2/htdocs/htmltags.in:
31702         * iExploder/iexploder-1.7.2/src/html-tags/webkit:
31703         Remove from fuzzer.
31704
31705 2011-04-26  Brent Fulgham  <bfulgham@webkit.org>
31706
31707         Reviewed by Adam Roben.
31708
31709         Update WinLauncher with command-line arguments to allow the
31710         transparency to be tested.  Use --transparent to create a
31711         transparent view.  Use --desktop to have the view fill the
31712         desktop area of the screen.
31713         https://bugs.webkit.org/show_bug.cgi?id=58300
31714
31715         * WinLauncher/WinLauncher.cpp:
31716         (usesLayeredWebView): New method to indicate that the
31717         program is running with layered windows (alpha-blended
31718         transparent windows).
31719         (shouldUseFullDesktop): New method to indicate that the
31720         program is running across the full desktop (less any
31721         task bars.)
31722         (resizeSubViews): Layered window mode doesn't have any
31723         winapi child windows, so this routine is exited early
31724         for that case.
31725         (subclassForLayeredWindow): Subclass the window so we can
31726         override its defaut message loop.
31727         (computeFullDesktopFrame): Convenience function to get the
31728         desktop region and update the frame dimensions appropriately.
31729         (_tWinMain): Updated to accept the new '--transparent' and
31730         '--desktop' arguments.
31731         (InitInstance): Modified to exit early for the transparent
31732         background case.  WebKit will construct our primary window (with
31733         appropriate settings) so we don't need to create one ourselves.
31734         (WndProc):
31735         (TransparentWndProc): New message loop for transparent windows,
31736         since they need special handling to support dragging.
31737
31738 2011-04-26  Alexandre MAzari  <amazari@igalia.com>
31739
31740         Reviewed by Xan Lopez.
31741
31742         [Gtk] Implement LayoutTestController::setPluginsEnabled
31743         https://bugs.webkit.org/show_bug.cgi?id=59184
31744
31745         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
31746         (LayoutTestController::setPluginsEnabled):
31747
31748 2011-04-26  Chang Shu  <cshu@webkit.org>
31749
31750         Reviewed by Laszlo Gombos.
31751
31752         WebKitTestRunner needs layoutTestController.setAllowFileAccessFromFileURLs
31753         https://bugs.webkit.org/show_bug.cgi?id=57572
31754
31755         Remove the unnecessary code that updates preference.
31756
31757         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
31758         (WTR::InjectedBundle::beginTesting):
31759         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
31760         (WTR::LayoutTestController::setAllowUniversalAccessFromFileURLs):
31761
31762 2011-04-26  Mario Sanchez Prada  <msanchez@igalia.com>
31763
31764         Reviewed by Eric Seidel.
31765
31766         [GTK] accessibility/aria-combobox-expected.txt
31767         https://bugs.webkit.org/show_bug.cgi?id=59399
31768
31769         Implement isExpanded(), as it's used in the layout test.
31770
31771         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
31772         (AccessibilityUIElement::isExpanded): Implemented.
31773
31774 2011-04-26  Adrienne Walker  <enne@google.com>
31775
31776         Reviewed by Kent Tamura.
31777
31778         [chromium] Fix memory leak in DRT due to OwnPtr<T*>
31779         https://bugs.webkit.org/show_bug.cgi?id=59489
31780
31781         * DumpRenderTree/chromium/WebViewHost.h:
31782
31783 2011-04-26  Mario Sanchez Prada  <msanchez@igalia.com>
31784
31785         Reviewed by Eric Seidel.
31786
31787         [GTK] Implement isChecked in DRT
31788         https://bugs.webkit.org/show_bug.cgi?id=59385
31789
31790         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
31791         (AccessibilityUIElement::isChecked): Implemented.
31792
31793 2011-04-26  Alejandro G. Castro  <alex@igalia.com>
31794
31795         Unreviewed, rolling out r84949.
31796         http://trac.webkit.org/changeset/84949
31797         https://bugs.webkit.org/show_bug.cgi?id=58852
31798
31799         Broke GTK+ compilation
31800
31801         * GtkLauncher/main.c:
31802         (main):
31803
31804 2011-04-26  Eric Seidel  <eric@webkit.org>
31805
31806         Reviewed by Mihai Parparita.
31807
31808         sheriff-bot should have a create-bug command
31809         https://bugs.webkit.org/show_bug.cgi?id=59491
31810
31811         This patch has some stderr unittest spew, but I will
31812         fix it after my live demo.
31813
31814         * Scripts/webkitpy/tool/bot/irc_command.py:
31815         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
31816         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
31817
31818 2011-04-26  Ryuan Choi  <ryuan.choi@samsung.com>
31819
31820         Reviewed by Martin Robinson.
31821
31822         [GTK] Add proxy support to GtkLauncher
31823         https://bugs.webkit.org/show_bug.cgi?id=58852
31824
31825         Add SOUP_TYPE_PROXY_RESOLVER_DEFAULT feature or set SOUP_SESSION_PROXY_URI for developers behind proxy.
31826
31827         * GtkLauncher/main.c:
31828         (main):
31829
31830 2011-04-26  Dan Bernstein  <mitz@apple.com>
31831
31832         Reviewed by Mark Rowe.
31833
31834         Choose the compiler based on the Xcode version for Snow Leopard debug builds.
31835
31836         * DumpRenderTree/mac/Configurations/Base.xcconfig:
31837         * DumpRenderTree/mac/Configurations/CompilerVersion.xcconfig: Added.
31838         * MiniBrowser/Configurations/Base.xcconfig:
31839         * MiniBrowser/Configurations/CompilerVersion.xcconfig: Added.
31840         * TestWebKitAPI/Configurations/Base.xcconfig:
31841         * TestWebKitAPI/Configurations/CompilerVersion.xcconfig: Added.
31842         * WebKitTestRunner/Configurations/Base.xcconfig:
31843         * WebKitTestRunner/Configurations/CompilerVersion.xcconfig: Added.
31844
31845 2011-04-26  Finnur Thorarinsson  <finnur.webkit@gmail.com>
31846
31847         Reviewed by David Levin.
31848
31849         Adding myself as a Chromium contributors to committers.py.
31850
31851         https://bugs.webkit.org/show_bug.cgi?id=59444
31852
31853         * Scripts/webkitpy/common/config/committers.py:
31854
31855 2011-04-26  Adam Barth  <abarth@webkit.org>
31856
31857         Reviewed by Eric Seidel.
31858
31859         DumpRenderTree/chromium should play nice with strict OwnPtrs
31860         https://bugs.webkit.org/show_bug.cgi?id=59458
31861
31862         All these changes are pretty straight forward.
31863
31864         * DumpRenderTree/chromium/CppBoundClass.cpp:
31865         (GetterPropertyCallback::GetterPropertyCallback):
31866         (CppBoundClass::bindGetterCallback):
31867         * DumpRenderTree/chromium/CppBoundClass.h:
31868         (CppBoundClass::bindProperty):
31869         (CppBoundClass::bindFallbackCallback):
31870         (CppBoundClass::bindFallbackMethod):
31871         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
31872         (DRTDevToolsClient::DRTDevToolsClient):
31873         * DumpRenderTree/chromium/TestEventPrinter.cpp:
31874         (TestEventPrinter::createDRTPrinter):
31875         (TestEventPrinter::createTestShellPrinter):
31876         * DumpRenderTree/chromium/TestEventPrinter.h:
31877         * DumpRenderTree/chromium/TestShell.cpp:
31878         (TestShell::TestShell):
31879         (TestShell::createMainWindow):
31880         (TestShell::createDRTDevToolsClient):
31881         * DumpRenderTree/chromium/WebViewHost.cpp:
31882         (WebViewHost::geolocationClientMock):
31883         (WebViewHost::speechInputController):
31884         (WebViewHost::deviceOrientationClientMock):
31885         (WebViewHost::reset):
31886         (WebViewHost::navigate):
31887         (WebViewHost::setPendingExtraData):
31888         (WebViewHost::canvas):
31889         * DumpRenderTree/chromium/WebViewHost.h:
31890
31891 2011-04-26  Adam Roben  <aroben@apple.com>
31892
31893         Make DRT/win log loading errors the same way DRT/mac does
31894
31895         Fixes <http://webkit.org/b/59303> [Windows WebKit1 Tests]
31896         http/tests/misc/will-send-request-returns-null-on-redirect.html failing since r84742
31897
31898         Reviewed by Alexey Proskuryakov.
31899
31900         * DumpRenderTree/win/ResourceLoadDelegate.cpp:
31901         (ResourceLoadDelegate::descriptionSuitableForTestResult): Removed the workaround for
31902         <rdar://problem/5064234>, which has long since been fixed, and then made this function match
31903         -[NSError(DRTExtras) _drt_descriptionSuitableForTestResult] more closely.
31904
31905 2011-04-26  Dan Bernstein  <mitz@apple.com>
31906
31907         REGRESSION (84876): build-webkit produces an error message
31908         https://bugs.webkit.org/show_bug.cgi?id=59460
31909
31910         * Scripts/webkitdirs.pm:
31911
31912 2011-04-26  Mihai Parparita  <mihaip@chromium.org>
31913
31914         Reviewed by Eric Seidel.
31915
31916         Fix timestamp comparison in check-for-weak-vtables-and-externals
31917         https://bugs.webkit.org/show_bug.cgi?id=59416
31918
31919         We should only re-run the check-for-weak-vtables-and-externals script
31920         if the executable is more recent than the last run.
31921
31922         * Scripts/check-for-weak-vtables-and-externals:
31923
31924 2011-04-26  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
31925
31926         Reviewed by Csaba Osztrogonác.
31927
31928         [Qt] Implement LayoutTestController::setAutofilled
31929         https://bugs.webkit.org/show_bug.cgi?id=59439
31930
31931         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
31932         (LayoutTestController::setAutofilled):
31933         * DumpRenderTree/qt/LayoutTestControllerQt.h:
31934
31935 2011-04-25  Dan Bernstein  <mitz@apple.com>
31936
31937         Reviewed by Mark Rowe.
31938
31939         Teach build-webkit where newer Xcode versions store the application-wide build settings user default.
31940
31941         * Scripts/webkitdirs.pm:
31942
31943 2011-04-25  Sam Weinig  <sam@webkit.org>
31944
31945         Reviewed by Adam Roben.
31946
31947         Leaks Viewer throws "Cannot post cyclic structures" in Safari 5
31948         https://bugs.webkit.org/show_bug.cgi?id=56090
31949
31950         Make LeaksViewer work with versions of browsers that don't support message
31951         passing of cyclic structures.
31952
31953         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParser.js:
31954         (LeaksParser.this._worker.onmessage):
31955         (LeaksParser):
31956         (LeaksParser.prototype.addLeaksFile):
31957         Use LeaksParserImpl directly if we don't support passing cyclic structures.
31958
31959         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserImpl.js: Copied from BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js.
31960         (LeaksParserImpl):
31961         (LeaksParserImpl.prototype.addLeaksFile):
31962         (LeaksParserImpl.prototype._incorporateLeaks):
31963         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
31964         (onmessage):
31965         Factor out LeaksParser implementation into LeaksParserImpl.js.
31966
31967         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js:
31968         (workersSupportCyclicStructures):
31969         Add function to detect if workers support passing cyclic structures.
31970
31971         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html:
31972         Add LeaksParserImpl.js include.
31973
31974 2011-04-25  Jeff Miller  <jeffm@apple.com>
31975
31976         Reviewed by Timothy Hatcher.
31977
31978         Treat strings defined with UI_STRING_LOCALIZE_LATER_KEY() as debugging strings in extract-localizable-strings
31979         https://bugs.webkit.org/show_bug.cgi?id=59373
31980
31981         * Scripts/extract-localizable-strings: Add UI_STRING_LOCALIZE_LATER_KEY() to isDebugMacro var.
31982
31983 2011-04-25  Andrew Scherkus  <scherkus@chromium.org>
31984
31985         Reviewed by Dimitri Glazkov.
31986
31987         Adding a few Chromium contributors to committers.py.
31988
31989         https://bugs.webkit.org/show_bug.cgi?id=59327
31990
31991         * Scripts/webkitpy/common/config/committers.py:
31992
31993 2011-04-23  Dominic Cooney  <dominicc@chromium.org>
31994
31995         Reviewed by Dimitri Glazkov.
31996
31997         [V8] Nodes in shadow DOM should not be GCed while their hosts are alive
31998         https://bugs.webkit.org/show_bug.cgi?id=59284
31999
32000         Chromium DRT over-refcounted the NPObject wrappers it used to
32001         marshal shadowRoot nodes back to tests (WebBindings::makeNode
32002         births NPObjects with referenceCount == 1). This masked the GC bug
32003         59284 in tests by making shadowRoot nodes live forever.
32004
32005         * DumpRenderTree/chromium/LayoutTestController.cpp:
32006         (LayoutTestController::shadowRoot):
32007         (LayoutTestController::ensureShadowRoot):
32008
32009 2011-04-23  Kevin Ollivier  <kevino@theolliviers.com>
32010
32011         [wx] Unreviewed build fixes. Add new methods to wx LayoutTestController,
32012         and update the URL for downloading waf.
32013
32014         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
32015         (LayoutTestController::ensureShadowRoot):
32016         (LayoutTestController::removeShadowRoot):
32017         * Scripts/webkitdirs.pm:
32018
32019 2011-04-22  Sheriff Bot  <webkit.review.bot@gmail.com>
32020
32021         Unreviewed, rolling out r84627.
32022         http://trac.webkit.org/changeset/84627
32023         https://bugs.webkit.org/show_bug.cgi?id=59271
32024
32025         It broke Symbian build (Requested by Ossy on #webkit).
32026
32027         * Scripts/webkitpy/style/checker.py:
32028         * Scripts/webkitpy/style/checker_unittest.py:
32029
32030 2011-04-22  Dirk Pranke  <dpranke@chromium.org>
32031
32032         Reviewed by Adam Barth.
32033
32034         rebaseline-chromium-webkit-tests: clean up output
32035         https://bugs.webkit.org/show_bug.cgi?id=59240
32036
32037         This change dramatically changes the output the tool provides
32038         by default. The previous output (more or less) is still there
32039         in -v / verbose / debug mode, but the default output is
32040         way terser and more readable. It mostly says which files are
32041         updated, which are duplicates, and which are not found in an
32042         archive. Feedback desired :).
32043
32044         This change also adds a scm.exists() method to the SCM object
32045         so that we can cleanly figure out what we're doing while
32046         rebaselining.
32047
32048         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
32049         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
32050         * Scripts/webkitpy/common/system/filesystem_mock.py:
32051         * Scripts/webkitpy/common/checkout/scm.py:
32052         * Scripts/webkitpy/common/checkout/scm_unittest.py:
32053
32054 2011-04-22  Ilya Sherman  <isherman@chromium.org>
32055
32056         Reviewed by Alexey Proskuryakov.
32057
32058         Add Ilya Sherman (isherman@chromium.org) to the list of contributors
32059         https://bugs.webkit.org/show_bug.cgi?id=59231
32060
32061         * Scripts/webkitpy/common/config/committers.py:
32062
32063 2011-04-22  Yuta Kitamura  <yutak@chromium.org>
32064
32065         Reviewed by Ojan Vafai.
32066
32067         new-run-webkit-tests: Exception raised when stderr contains binary data
32068         https://bugs.webkit.org/show_bug.cgi?id=59083
32069
32070         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
32071         Use write_binary_file() instead of write_text_file().
32072
32073 2011-04-22  Chang Shu  <cshu@webkit.org>
32074
32075         Reviewed by Kenneth Rohde Christiansen.
32076
32077         WebKitTestRunner needs layoutTestController.setFrameFlatteningEnabled
32078         https://bugs.webkit.org/show_bug.cgi?id=42536
32079
32080         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
32081         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
32082         (WTR::LayoutTestController::setFrameFlatteningEnabled):
32083         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
32084
32085 2011-04-22  Adam Barth  <abarth@webkit.org>
32086
32087         Reviewed by Eric Seidel.
32088
32089         CommitQueue should remember the expected failures from patch to patch
32090         https://bugs.webkit.org/show_bug.cgi?id=59253
32091
32092         Previously, we stored the set of expected failures on the task object,
32093         which meant the commit-queue forgot about them after each patch.  This
32094         patch moves them to the CommitQueue object so they will have a longer
32095         lifetime.
32096
32097         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
32098         * Scripts/webkitpy/tool/bot/patchanalysistask.py:
32099         * Scripts/webkitpy/tool/commands/queues.py:
32100
32101 2011-04-22  Sam Weinig  <sam@webkit.org>
32102
32103         Reviewed by Brian Weinstein.
32104
32105         WebKit2: Implement JavaScript unresponsiveness timer/callbacks
32106         https://bugs.webkit.org/show_bug.cgi?id=59248
32107
32108         Stub out shouldInterruptJavaScript callbacks.
32109
32110         * MiniBrowser/mac/BrowserWindowController.m:
32111         (-[BrowserWindowController awakeFromNib]):
32112         * MiniBrowser/win/BrowserView.cpp:
32113         (BrowserView::create):
32114         * WebKitTestRunner/TestController.cpp:
32115         (WTR::TestController::createOtherPage):
32116         (WTR::TestController::initialize):
32117
32118 2011-04-22  Daniel Bates  <dbates@webkit.org>
32119
32120         Reviewed by Csaba Osztrogonác.
32121
32122         webkit-patch land hangs if svn prompts for credentials
32123         https://bugs.webkit.org/show_bug.cgi?id=31500
32124
32125         Make has_authorization_for_realm() return True only if there exists a credentials file
32126         that contains either the word "password" or "passtype". We assume that these words don't
32127         coincide with the actual credential data (e.g. a person's username is "password").
32128
32129         * Scripts/webkitpy/common/checkout/scm.py:
32130         * Scripts/webkitpy/common/checkout/scm_unittest.py:
32131           - Added test cases:
32132             test_has_authorization_for_realm_using_credentials_with_passtype(),
32133             test_has_authorization_for_realm_using_credentials_with_password(),
32134             test_not_have_authorization_for_realm_with_credentials_missing_password_and_passtype()
32135           - Renamed test_not_have_authorization_for_realm() to test_not_have_authorization_for_realm_when_missing_credentials_file()
32136             to better describe what it's testing.
32137           - Repurposed test_has_authorization_for_realm() to take realm and credential data to use
32138             and return the result of calling has_authorization_for_realm() so that the caller can
32139             assert the result; Renamed to _test_has_authorization_for_realm_using_credentials() to
32140             better reflect its new purpose.
32141
32142 2011-04-22  Adam Barth  <abarth@webkit.org>
32143
32144         Remove double-negative.  (Sorry this was bugging me.)
32145
32146         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
32147
32148 2011-04-22  Adam Barth  <abarth@webkit.org>
32149
32150         Reviewed by Eric Seidel.
32151
32152         Factor LayoutTestResultsReader out of CommitQueue
32153         https://bugs.webkit.org/show_bug.cgi?id=59244
32154
32155         This code will be shared with the EWS when they start running tests.
32156
32157         * Scripts/webkitpy/tool/bot/layouttestresultsreader.py: Added.
32158         * Scripts/webkitpy/tool/bot/layouttestresultsreader_unittest.py: Added.
32159         * Scripts/webkitpy/tool/commands/queues.py:
32160         * Scripts/webkitpy/tool/commands/queues_unittest.py:
32161
32162 2011-04-22  Eric Seidel  <eric@webkit.org>
32163
32164         Reviewed by Dimitri Glazkov.
32165
32166         webkit-patch rebaseline is broken
32167         https://bugs.webkit.org/show_bug.cgi?id=59207
32168
32169         My recent changes to how LayoutTestResults.results_from_string
32170         uncovered a bug in our use of urllib2.
32171         We now are treating the return value from _fetch_results_html()
32172         (which returns urllib2.openurl()) as a string.  That's incorrect,
32173         the return value is a file-like object (which auto-converts to a
32174         string in most cases).
32175
32176         I've updated our urllib2-using code to correctly call .read() on the
32177         resulting object before treating it like a string.
32178         Unfortunately there is no easy way to test this mis-understanding
32179         of the openurl() semantics.
32180
32181         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
32182
32183 2011-04-22  Adam Barth  <abarth@webkit.org>
32184
32185         Reviewed by Eric Seidel.
32186
32187         Factor PatchAnalysis task out of CommitQueueTask
32188         https://bugs.webkit.org/show_bug.cgi?id=59220
32189
32190         This is a first step towards teaching the EWS to run tests.
32191
32192         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
32193         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
32194         * Scripts/webkitpy/tool/bot/patchanalysistask.py: Added.
32195
32196 2011-04-22  Philippe Normand  <pnormand@igalia.com>
32197
32198         Reviewed by Martin Robinson.
32199
32200         [GTK] DRT: implement eventSender.scalePageBy
32201         https://bugs.webkit.org/show_bug.cgi?id=59082
32202
32203         * DumpRenderTree/gtk/EventSender.cpp:
32204         (scalePageByCallback): scalePageBy() support.
32205
32206 2011-04-21  Eric Seidel  <eric@webkit.org>
32207
32208         Reviewed by Adam Barth.
32209
32210         LayoutTestResults should know how to handle NRWT json files
32211         https://bugs.webkit.org/show_bug.cgi?id=59168
32212
32213         LayoutTestResults was originally created to be a parallel
32214         to existing NRWT architecture for dealing with results files.
32215         But then it turns out that NRWT has no such architecture. :(
32216         So this patch writes the necessary code to handle reading
32217         full_results.json and unexpected_results.json files from
32218         NRWT layout-test-results directories.
32219
32220         LayoutTestResults has thus morphed from being ORWT-only to
32221         supporting both ORWT and NRWT.  It's possible at some future
32222         point that other pieces of the NRWT architecture will learn
32223         how to read JSON files in which case this can go away.
32224
32225         This is all done for making it possible for the commit-queue
32226         to run the tests using NWRT and be able to understand the results
32227         (for flaky test reporting, etc.)
32228
32229         * Scripts/webkitpy/common/net/layouttestresults.py:
32230         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
32231         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
32232         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
32233         * Scripts/webkitpy/common/net/resultsjsonparser.py: Added.
32234         * Scripts/webkitpy/common/net/resultsjsonparser_unittest.py: Added.
32235
32236 2011-04-22  Yi Shen  <yi.4.shen@nokia.com>
32237
32238         Reviewed by Kenneth Rohde Christiansen.
32239
32240         [Qt] Upstream Symbian platform plugin
32241         https://bugs.webkit.org/show_bug.cgi?id=58435
32242
32243         Exempting directory WebKit/qt/symbian/platformplugin from style guide.
32244
32245         * Scripts/webkitpy/style/checker.py:
32246         * Scripts/webkitpy/style/checker_unittest.py:
32247
32248 2011-04-22  Balazs Kelemen  <kbalazs@webkit.org>
32249
32250         Reviewed by Csaba Osztrogonác.
32251
32252         [Qt][WK2] Qt should use the mac-wk2 skipped list
32253         https://bugs.webkit.org/show_bug.cgi?id=58536
32254
32255         * Scripts/old-run-webkit-tests:
32256
32257 2011-04-21  Tony Chang  <tony@chromium.org>
32258
32259         Reviewed by Mihai Parparita.
32260
32261         Remove find-mismatched-layout-test-results since it's not needed in a
32262         checksum free world.
32263         https://bugs.webkit.org/show_bug.cgi?id=59156
32264
32265         * Scripts/find-mismatched-layout-test-results: Removed.
32266
32267 2011-04-21  Brent Fulgham  <bfulgham@webkit.org>
32268
32269         Reviewed by Martin Robinson.
32270
32271         Teach build-webkit and update-webkit that the WinCairo port does
32272         not need the non-redistributable WebKitSupportLibraries.zip file.
32273         Also append the _Cairo_CFLite label to the build target for the
32274         default (no argument) case when building with '--wincairo'
32275
32276         https://bugs.webkit.org/show_bug.cgi?id=58801
32277
32278         * Scripts/build-webkit: Don't die if WebKitSupportLibraries.zip
32279           doesn't exist.
32280         * Scripts/webkitdirs.pm: Append '_Cairo_CFLite' to the default
32281           'Release' build target when passed the '--wincairo' flag.
32282         * Scripts/update-webkit: Don't die if WebKitSupportLibraries.zip
32283           doesn't exist.
32284
32285 2011-04-21  Dirk Pranke  <dpranke@chromium.org>
32286
32287         Reviewed by Eric Seidel.
32288
32289         new-run-webkit-tests: obsolete old code, part 5: rename message_broker2 to message_broker
32290         https://bugs.webkit.org/show_bug.cgi?id=58758
32291
32292         Since worker_mixin is now only used by worker, there's no point
32293         in having it split across two files.
32294
32295         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
32296         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
32297         * Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Added
32298         * Scripts/webkitpy/layout_tests/layout_package/message_broker2.py: Removed.
32299         * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: * Added
32300         * Scripts/webkitpy/layout_tests/layout_package/message_broker2_unittest.py: Removed.
32301
32302 2011-04-21  Dirk Pranke  <dpranke@chromium.org>
32303
32304         Reviewed by Tony Chang.
32305
32306         new-run-webkit-tests: obsolete old code, part 4: merge worker_mixin into worker
32307         https://bugs.webkit.org/show_bug.cgi?id=58756
32308
32309         Since worker_mixin is now only used by worker, there's no point
32310         in having it split across two files.
32311
32312         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
32313         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py: Removed.
32314
32315 2011-04-21  Dirk Pranke  <dpranke@chromium.org>
32316
32317         Reviewed by Tony Chang.
32318
32319         new-run-webkit-tests: obsolete old code, part 3 - rename test_runner to manager
32320         https://bugs.webkit.org/show_bug.cgi?id=58754
32321
32322         test_runner hasn't been a great name for some time, since all of
32323         the test running is actually done in worker and
32324         single_test_runner. Rename to manager to match the terminology
32325         elsewhere.
32326
32327         * Scripts/webkitpy/layout_tests/layout_package/manager.py: Added.
32328         * Scripts/webkitpy/layout_tests/layout_package/manager_unittest.py: Added.
32329         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
32330         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py: Removed.
32331         * Scripts/webkitpy/layout_tests/layout_package/test_runner_unittest.py: Removed.
32332         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
32333
32334 2011-04-21  Dirk Pranke  <dpranke@chromium.org>
32335
32336         Reviewed by Tony Chang.
32337
32338         new-run-webkit-tests: obsolete old threading code, part 2
32339         https://bugs.webkit.org/show_bug.cgi?id=58753
32340
32341         This merges the test_runner2 code back into test_runner since
32342         there's only a single type of test runner now.
32343
32344         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
32345         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py: Removed.
32346         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
32347
32348 2011-04-21  Maciej Stachowiak  <mjs@apple.com>
32349
32350         Reviewed by Adam Roben.
32351
32352         Add a feature define to allow <details> and <summary> to be disabled
32353         https://bugs.webkit.org/show_bug.cgi?id=59118
32354         <rdar://problem/9257045>
32355
32356         * Scripts/build-webkit:
32357         * Tools.pro:
32358
32359 2011-04-21  Adam Barth  <abarth@webkit.org>
32360
32361         Reviewed by Eric Seidel.
32362
32363         webkit-patch upload should assign the bug to you
32364         https://bugs.webkit.org/show_bug.cgi?id=33699
32365
32366         This patch only re-assigns the bug if it is currently unassigned.  We
32367         can be more aggressive in the future if people want us to be more
32368         agressive.
32369
32370         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
32371         * Scripts/webkitpy/tool/commands/upload_unittest.py:
32372         * Scripts/webkitpy/tool/mocktool.py:
32373         * Scripts/webkitpy/tool/steps/postdiff.py:
32374
32375 2011-04-21  Ojan Vafai  <ojan@chromium.org>
32376
32377         Reviewed by Tony Chang.
32378
32379         move json_results.html over to LayoutTests/fast/harness
32380         https://bugs.webkit.org/show_bug.cgi?id=59154
32381
32382         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
32383
32384 2011-04-21  Adam Barth  <abarth@webkit.org>
32385
32386         Reviewed by Eric Seidel.
32387
32388         ValidateChangeLogs doesn't work on SVN
32389         https://bugs.webkit.org/show_bug.cgi?id=59115
32390
32391         svn-create-patch lies about the diff in the working copy by moving
32392         ChangeLog entries to the top of the diff.  That's fine on most cases,
32393         but causes problems for ValidateChangeLogs, which is trying validate
32394         the where the ChangeLog entry appears.
32395
32396         I haven't added a test for this change because I couldn't figure out
32397         how to write one.  The issue is more of an integration issue, which
32398         we're not really set up to test in our unit testing framework.  If this
32399         patch had worked around the output from svn-create-patch, then I could
32400         have tested that we behave correctly on sample svn-create-patch output,
32401         but, in this case, I've removed the dependency on svn-create-patch.  I
32402         could test that we behave correctly on "svn diff" output, but we
32403         already have those tests.
32404
32405         * Scripts/webkitpy/tool/commands/download_unittest.py:
32406         * Scripts/webkitpy/tool/mocktool.py:
32407         * Scripts/webkitpy/tool/steps/validatechangelogs.py:
32408
32409 2011-04-21  Tony Chang  <tony@chromium.org>
32410
32411         Reviewed by Ojan Vafai.
32412
32413         old-run-webkit-tests/new-run-webkit-tests shouldn't read or write .checksum files
32414         https://bugs.webkit.org/show_bug.cgi?id=58402
32415
32416         * Scripts/old-run-webkit-tests: Don't read or write .checksum files.
32417         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py: Don't write .checksum files.
32418         * Scripts/webkitpy/layout_tests/port/base.py: Don't read .checksum files.
32419         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
32420         * Scripts/webkitpy/layout_tests/port/test.py: Update passing pngs to include a checksum.
32421         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
32422
32423 2011-04-21  Ojan Vafai  <ojan@chromium.org>
32424
32425         Reviewed by Tony Chang.
32426
32427         final tests for results.html file
32428         https://bugs.webkit.org/show_bug.cgi?id=59150
32429
32430         This gets us close to 100% code coverage. Once this is
32431         checked in, I'll move this file over to LayoutTests/fast/harness
32432         and make it a proper layout test.
32433
32434         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32435         * Scripts/webkitpy/layout_tests/layout_package/json_results_test.js:
32436
32437 2011-04-21  Ojan Vafai  <ojan@chromium.org>
32438
32439         Reviewed by Tony Chang.
32440
32441         only expand visible rows
32442         https://bugs.webkit.org/show_bug.cgi?id=59130
32443
32444         For the chromium port, where there are many expected failures,
32445         expanding all of them is quite slow. Only expand the visible ones.
32446         This makes the default case. Also, make the expanding async to
32447         avoid totally killing the process.
32448
32449         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32450         * Scripts/webkitpy/layout_tests/layout_package/json_results_test.js:
32451
32452 2011-04-21  Martin Robinson  <mrobinson@igalia.com>
32453
32454         [GTK] DRT: implement eventSender.scalePageBy
32455         https://bugs.webkit.org/show_bug.cgi?id=59082
32456
32457         Rolling out r84507 since it broke many GTK+ tests.
32458
32459         * DumpRenderTree/gtk/EventSender.cpp:
32460
32461 2011-04-21  Sam Weinig  <sam@webkit.org>
32462
32463         Reviewed by Simon Fraser.
32464
32465         Add way to garbage collect between each test when running layout tests
32466         https://bugs.webkit.org/show_bug.cgi?id=59126
32467
32468         * DumpRenderTree/mac/DumpRenderTree.mm:
32469         (initializeGlobalsFromCommandLineOptions):
32470         (runTest):
32471         * Scripts/old-run-webkit-tests:
32472         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
32473         (WTR::InjectedBundle::didReceiveMessage):
32474         * WebKitTestRunner/TestController.cpp:
32475         (WTR::TestController::TestController):
32476         (WTR::TestController::initialize):
32477         (WTR::TestController::resetStateToConsistentValues):
32478         * WebKitTestRunner/TestController.h:
32479
32480 2011-04-21  Ojan Vafai  <ojan@chromium.org>
32481
32482         Reviewed by Mihai Parparita.
32483
32484         fix another innocuous JS error in expand/collapseAllExpectations
32485         https://bugs.webkit.org/show_bug.cgi?id=59122
32486
32487         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32488         * Scripts/webkitpy/layout_tests/layout_package/json_results_test.js:
32489
32490 2011-04-20  Ojan Vafai  <ojan@chromium.org>
32491
32492         Reviewed by Tony Chang.
32493
32494         put unexpected passes into their own table
32495         https://bugs.webkit.org/show_bug.cgi?id=59016
32496
32497         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32498
32499 2011-04-20  Ojan Vafai  <ojan@chromium.org>
32500
32501         Reviewed by Tony Chang.
32502
32503         initial set of tests for new results.html file
32504         https://bugs.webkit.org/show_bug.cgi?id=59031
32505
32506         Also, fix the bugs (mostly innocuous JS errors) the tests found.
32507
32508         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32509         * Scripts/webkitpy/layout_tests/layout_package/json_results_test.js: Added.
32510
32511 2011-04-21  Philippe Normand  <pnormand@igalia.com>
32512
32513         Reviewed by Martin Robinson.
32514
32515         [GTK] missing LayoutTestController::callShouldCloseOnWebView()
32516         https://bugs.webkit.org/show_bug.cgi?id=59086
32517
32518         callShouldCloseOnWebView() implementation.
32519
32520         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
32521         (LayoutTestController::callShouldCloseOnWebView):
32522
32523 2011-04-21  Philippe Normand  <pnormand@igalia.com>
32524
32525         Reviewed by Martin Robinson.
32526
32527         [GTK] DRT: implement eventSender.scalePageBy
32528         https://bugs.webkit.org/show_bug.cgi?id=59082
32529
32530         * DumpRenderTree/gtk/EventSender.cpp:
32531         (scalePageByCallback): scalePageBy() support.
32532
32533 2011-04-20  Dominic Cooney  <dominicc@chromium.org>
32534
32535         Reviewed by Dimitri Glazkov.
32536
32537         layoutTestController can create and destroy shadow DOM
32538         https://bugs.webkit.org/show_bug.cgi?id=59058
32539
32540         Chromium, Mac, GTK and Qt DRT; and WK2 test runner.
32541
32542         * DumpRenderTree/LayoutTestController.cpp:
32543         (ensureShadowRootCallback):
32544         (removeShadowRootCallback):
32545         (LayoutTestController::staticFunctions):
32546         * DumpRenderTree/LayoutTestController.h:
32547         * DumpRenderTree/chromium/LayoutTestController.cpp:
32548         (LayoutTestController::LayoutTestController):
32549         (LayoutTestController::ensureShadowRoot):
32550         (LayoutTestController::removeShadowRoot):
32551         * DumpRenderTree/chromium/LayoutTestController.h:
32552         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
32553         (LayoutTestController::ensureShadowRoot):
32554         (LayoutTestController::removeShadowRoot):
32555         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
32556         (LayoutTestController::shadowRoot):
32557         (LayoutTestController::ensureShadowRoot):
32558         (LayoutTestController::removeShadowRoot):
32559         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
32560         (LayoutTestController::ensureShadowRoot):
32561         (LayoutTestController::removeShadowRoot):
32562         * DumpRenderTree/qt/LayoutTestControllerQt.h:
32563         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
32564         (LayoutTestController::ensureShadowRoot):
32565         (LayoutTestController::removeShadowRoot):
32566         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
32567         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
32568         (WTR::LayoutTestController::shadowRoot):
32569         (WTR::LayoutTestController::ensureShadowRoot):
32570         (WTR::LayoutTestController::removeShadowRoot):
32571         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
32572
32573 2011-04-20  Dirk Pranke  <dpranke@chromium.org>
32574
32575         Reviewed by Eric Seidel.
32576
32577         new-run-webkit-tests: handle ctrl-c more cleanly
32578         https://bugs.webkit.org/show_bug.cgi?id=58849
32579
32580         * Scripts/new-run-webkit-tests:
32581         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
32582
32583 2011-04-20  Sheriff Bot  <webkit.review.bot@gmail.com>
32584
32585         Unreviewed, rolling out r84431.
32586         http://trac.webkit.org/changeset/84431
32587         https://bugs.webkit.org/show_bug.cgi?id=59061
32588
32589         Breaks for folks who don't have editbugs (Requested by abarth
32590         on #webkit).
32591
32592         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
32593         * Scripts/webkitpy/tool/commands/upload_unittest.py:
32594         * Scripts/webkitpy/tool/mocktool.py:
32595         * Scripts/webkitpy/tool/steps/postdiff.py:
32596
32597 2011-04-20  Leo Yang  <leo.yang@torchmobile.com.cn>
32598
32599         Unreviewed.
32600
32601         Adding myself to committers.py.
32602
32603         * Scripts/webkitpy/common/config/committers.py:
32604
32605 2011-04-20  Adam Barth  <abarth@webkit.org>
32606
32607         Reviewed by Eric Seidel.
32608
32609         webkit-patch upload should assign the bug to you
32610         https://bugs.webkit.org/show_bug.cgi?id=33699
32611
32612         This patch only re-assigns the bug if it is currently unassigned.  We
32613         can be more aggressive in the future if people want us to be more
32614         agressive.
32615
32616         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
32617         * Scripts/webkitpy/tool/commands/upload_unittest.py:
32618         * Scripts/webkitpy/tool/mocktool.py:
32619         * Scripts/webkitpy/tool/steps/postdiff.py:
32620
32621 2011-04-20 Peter Gal <galpeter@inf.u-szeged.hu>
32622
32623         Reviewed by Csaba Osztrogonác.
32624
32625         Modify builder.html template to allow easy access
32626         to list more builds on the builder's page.
32627
32628         https://bugs.webkit.org/show_bug.cgi?id=59008
32629
32630         * BuildSlaveSupport/build.webkit.org-config/templates/builder.html:
32631
32632 2011-04-20  Peter Gal  <galpeter@inf.u-szeged.hu>
32633
32634         Reviewed by Csaba Osztrogonác.
32635
32636         Add the default template for builder page.
32637
32638         https://bugs.webkit.org/show_bug.cgi?id=59008
32639
32640         * BuildSlaveSupport/build.webkit.org-config/templates/builder.html: Added.
32641
32642 2011-04-20  Ojan Vafai  <ojan@chromium.org>
32643
32644         Reviewed by Eric Seidel.
32645
32646         move page generation into a single function
32647         https://bugs.webkit.org/show_bug.cgi?id=59027
32648
32649         Now executing the script doesn't create any state. We should then
32650         be able to write tests for this by setting custom results data.
32651
32652         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32653
32654 2011-04-20  Ojan Vafai  <ojan@chromium.org>
32655
32656         Reviewed by Eric Seidel.
32657
32658         move global state into a shared object
32659         https://bugs.webkit.org/show_bug.cgi?id=59023
32660
32661         This makes the code more testable.
32662
32663         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32664
32665 2011-04-20  Ojan Vafai  <ojan@chromium.org>
32666
32667         Reviewed by Tony Chang.
32668
32669         move global state into a shared object
32670         https://bugs.webkit.org/show_bug.cgi?id=59023
32671
32672         This makes the code more testable.
32673
32674         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32675
32676 2011-04-20  Ojan Vafai  <ojan@chromium.org>
32677
32678         Reviewed by Eric Seidel.
32679
32680         add expandAll/collapseAll links
32681         https://bugs.webkit.org/show_bug.cgi?id=59011
32682
32683         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32684
32685 2011-04-20  Kevin Ollivier  <kevino@theolliviers.com>
32686
32687         [wx] Unreviewed build fixes for wx 2.9.1, Cocoa and 64-bit builds.
32688
32689         * Scripts/webkitdirs.pm:
32690         * wx/build/build_utils.py:
32691         * wx/build/settings.py:
32692         * wx/install-unix-extras:
32693         * wx/packaging/build-mac-installer.py:
32694         * wx/packaging/build-win-installer.py:
32695
32696 2011-04-20  Adam Roben  <aroben@apple.com>
32697
32698         Replace old-run-webkit-tests's win-7 platform with win-7sp0
32699
32700         It was an error that we had a win-7 platform to begin with. Windows 7 was the most recent
32701         version of Windows we cared about, so its results should just go in win. But now that
32702         Windows 7 SP1 exists (and has different results for some tests), we need a place to store
32703         results for Windows 7 without SP1 (and older versions of Windows).
32704
32705         Part of <http://webkit.org/b/56617> Several tests fail on Windows 7 SP1 due to font
32706         differences from Windows 7 "SP0"
32707
32708         Reviewed by David Kilzer.
32709
32710         * Scripts/old-run-webkit-tests: Replaced win-7 with win-7sp0. Newer versions of Windows 7
32711         will just use "win".
32712
32713         * Scripts/webkitdirs.pm:
32714         (isWindows7SP0): Added. Replaces isWindows7, because that function had no remaining callers.
32715
32716 2011-04-20  Adam Roben  <aroben@apple.com>
32717
32718         Teach webkitdirs::winVersion about build numbers
32719
32720         The build number isn't currently used, but soon will be.
32721
32722         As a bonus, winVersion now works on non-Cygwin variants of Perl!
32723
32724         Fixes <http://webkit.org/b/58991> Need a way to distinguish between Windows 7 and Windows 7
32725         SP1
32726
32727         Reviewed by Anders Carlsson.
32728
32729         * Scripts/webkitdirs.pm:
32730         (isAnyWindows): Added. Returns true if we're running on any Windows-ish Perl.
32731         (determineWinVersion): Use `cmd /c ver` to get the version number, including the build
32732         number. Parse the version number into a hash, similar to how osXVersion() works.
32733
32734         (isWindows7):
32735         (isWindowsVista):
32736         (isWindowsXP):
32737         Updated for changes to winVersion().
32738
32739 2011-04-20  W. James MacLean  <wjmaclean@chromium.org>
32740
32741         Reviewed by Kenneth Russell.
32742
32743         Add support for composite-to-texture to DumpRenderTree.
32744         https://bugs.webkit.org/show_bug.cgi?id=58675
32745
32746         * DumpRenderTree/chromium/DumpRenderTree.cpp:
32747         (main):
32748         * DumpRenderTree/chromium/TestShell.cpp:
32749         (TestShell::TestShell):
32750         (TestShell::resetWebSettings):
32751         * DumpRenderTree/chromium/TestShell.h:
32752         (TestShell::setCompositeToTexture):
32753         * DumpRenderTree/chromium/WebPreferences.cpp:
32754         (WebPreferences::reset):
32755         (WebPreferences::applyTo):
32756         * DumpRenderTree/chromium/WebPreferences.h:
32757
32758 2011-04-19  Eric Seidel  <eric@webkit.org>
32759
32760         Reviewed by Adam Barth.
32761
32762         The commit-queue is confused when non-layout tests fail
32763         https://bugs.webkit.org/show_bug.cgi?id=58955
32764
32765         As seen in http://queues.webkit.org/results/8474435
32766
32767         It fails to create an archive of the layout test results
32768         and then throws an exception.
32769
32770         I've fixed this by making it catch the exception as well
32771         as made it so it never calls that path in the common case.
32772
32773         I've updated MockFileSystem to understand rmtree's affect on
32774         directories as well as files.
32775
32776         * Scripts/webkitpy/common/system/filesystem_mock.py:
32777         * Scripts/webkitpy/common/system/workspace.py:
32778         * Scripts/webkitpy/tool/commands/queues.py:
32779         * Scripts/webkitpy/tool/commands/queues_unittest.py:
32780         * Scripts/webkitpy/tool/mocktool.py:
32781
32782 2011-04-19  Ojan Vafai  <ojan@chromium.org>
32783
32784         Fix check for whether a failure is expected to deal with
32785         flaky tests and FAIL expectations.
32786         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32787
32788 2011-04-19  Ojan Vafai  <ojan@chromium.org>
32789
32790         Reviewed by Eric Seidel.
32791
32792         don't show missing text results for tests that only dump image results
32793         https://bugs.webkit.org/show_bug.cgi?id=58931
32794
32795         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32796         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
32797
32798 2011-04-19  Ojan Vafai  <ojan@chromium.org>
32799
32800         Reviewed by Eric Seidel.
32801
32802         don't show failures table if there are only new tests
32803         https://bugs.webkit.org/show_bug.cgi?id=58929
32804
32805         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32806
32807 2011-04-19  Ojan Vafai  <ojan@chromium.org>
32808
32809         Reviewed by Eric Seidel.
32810
32811         don't show failures table if there are only new tests
32812         https://bugs.webkit.org/show_bug.cgi?id=58929
32813
32814         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32815
32816 2011-04-19  Ojan Vafai  <ojan@chromium.org>
32817
32818         Fix unittest failures from http://trac.webkit.org/changeset/84294.
32819
32820         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
32821         * Scripts/webkitpy/layout_tests/layout_package/test_runner_unittest.py:
32822
32823 2011-04-18  Ojan Vafai  <ojan@chromium.org>
32824
32825         Reviewed by Eric Seidel.
32826
32827         switch new-run-webkit-tests to using the new results file
32828         https://bugs.webkit.org/show_bug.cgi?id=58861
32829
32830         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
32831         * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:
32832         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
32833
32834 2011-04-19  Renata Hodovan  <reni@webkit.org>
32835
32836         Reviewed by Eric Seidel.
32837
32838         Move the alignment related macros from Vector.h to new Alignment.h.
32839         https://bugs.webkit.org/show_bug.cgi?id=56000
32840
32841         Adding the new file to the forwarding headers of Tools.
32842
32843         * DumpRenderTree/ForwardingHeaders/wtf/Alignment.h: Added.
32844
32845 2011-04-19  Kevin Ollivier  <kevino@theolliviers.com>
32846
32847         [wx] Unreviewed build fix for Win / wx 2.9 wxPython installation.
32848
32849         * wx/packaging/build-win-installer.py:
32850         * wx/packaging/wxWebKitInstaller.iss.in:
32851
32852 2011-04-19  Adam Roben  <aroben@apple.com>
32853
32854         Skip another accelerated-compositing-sensitive test when accelerated compositing is disabled
32855
32856         * Scripts/old-run-webkit-tests: Skip media/controls-without-preload.html on Windows when
32857         accelerated compositing is disabled.
32858
32859 2011-04-19  Ojan Vafai  <ojan@chromium.org>
32860
32861         Remove accidentally committed debug code.
32862         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
32863
32864 2011-04-19  Brent Fulgham  <bfulgham@webkit.org>
32865
32866         Unreviewed build correction.
32867
32868         Avoid duplicate stub implementation of test routine
32869         for WinCairo build.
32870
32871         * WebKitTestRunner/win/TestInvocationWin.cpp: Exclude
32872         WinCairo from using this stub.
32873
32874 2011-04-18  Ojan Vafai  <ojan@chromium.org>
32875
32876         Reviewed by Eric Seidel.
32877
32878         make results file work with audio and reftests
32879         https://bugs.webkit.org/show_bug.cgi?id=58860
32880
32881         Also fix bug with timeout tests and store a bit in the JSON
32882         for new image tests instead of loading the image result to check if it's there.
32883
32884         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
32885         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
32886
32887 2011-04-19  Jer Noble  <jer.noble@apple.com>
32888
32889         Reviewed by Adam Roben.
32890
32891         REGRESSION (r84206): 50 tests failing on SnowLeopard Intel Release (WebKit2 Tests) due to extra "supportsFullScreen() == true" output
32892         https://bugs.webkit.org/show_bug.cgi?id=58890
32893
32894         Add a layoutTestController flag which controls when full screen callbacks are written to stdout.
32895         Defaults to false.
32896
32897         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl: Added dumpFullScreenCallbacks().
32898         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
32899         (WTR::InjectedBundlePage::supportsFullScreen): Only write to stdout if shouldDumpFullScreenCallbacks is set.
32900         (WTR::InjectedBundlePage::enterFullScreenForElement): Ditto.
32901         (WTR::InjectedBundlePage::exitFullScreenForElement): Ditto.
32902         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
32903         (WTR::LayoutTestController::LayoutTestController): Initialize new ivar.
32904         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
32905         (WTR::LayoutTestController::dumpFullScreenCallbacks): Added.
32906         (WTR::LayoutTestController::shouldDumpFullScreenCallbacks): Added.
32907
32908 2011-04-19  Kristóf Kosztyó  <Kosztyo.Kristof@stud.u-szeged.hu>
32909
32910         Reviewed by Andreas Kling.
32911
32912         [Qt] fast/xmlhttprequest/xmlhttprequest-nonexistent-file.html fails (DRT sideeffect)
32913         https://bugs.webkit.org/show_bug.cgi?id=35086
32914
32915         fast/xmlhttprequest/xmlhttprequest-no-file-access.html caused
32916         a DRT sideeffect problem, because the resetSettings didn't set
32917         the default value of LocalContentCanAccessFileUrls.
32918
32919         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
32920         (WebCore::WebPage::resetSettings):
32921
32922 2011-04-18  Zan Dobersek  <zandobersek@gmail.com>
32923
32924         Reviewed by Martin Robinson.
32925
32926         [GTK] Need support for dumping focus rectangles in pixel results
32927         https://bugs.webkit.org/show_bug.cgi?id=53647
32928
32929         Dump the selection rect if it has been requested.
32930
32931         * DumpRenderTree/gtk/PixelDumpSupportGtk.cpp:
32932         (createBitmapContextFromWebView):
32933
32934 2011-04-18  Eric Seidel  <eric@webkit.org>
32935
32936         Reviewed by Adam Barth.
32937
32938         test-webkitpy should run the ActiveWorkItems tests
32939         https://bugs.webkit.org/show_bug.cgi?id=58859
32940
32941         Also make the unit tests pass. :)
32942
32943         * QueueStatusServer/model/activeworkitems.py:
32944         * QueueStatusServer/model/queue.py:
32945         * QueueStatusServer/model/workitems.py:
32946
32947 2011-04-18  Eric Seidel  <eric@webkit.org>
32948
32949         Reviewed by Adam Barth.
32950
32951         queues.webkit.org should show trailing-days pass counts for queues
32952         https://bugs.webkit.org/show_bug.cgi?id=58812
32953
32954         This lets us see if each bot is pulling its own weight or not.
32955         Before I wrote this patch our suspicious was that one of the
32956         cq bots was doing most of the work.  Turns out they're actually
32957         pretty even in their recent pass counts.
32958
32959         * QueueStatusServer/handlers/queuestatus.py:
32960         * QueueStatusServer/index.yaml:
32961         * QueueStatusServer/templates/queuestatus.html:
32962
32963 2011-04-18  Dirk Pranke  <dpranke@chromium.org>
32964
32965         Reviewed by Eric Seidel.
32966
32967         new-run-webkit-tests: obsolete old threading code, part 1
32968         https://bugs.webkit.org/show_bug.cgi?id=58752
32969
32970         This patch removes support for the 'old-inline' and
32971         'old-threads' worker models, and deletes the corresponding code.
32972         Now that the new code is stable, this stuff is no longer
32973         necessary.
32974
32975         * Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py: Removed.
32976         * Scripts/webkitpy/layout_tests/layout_package/message_broker.py: Removed.
32977         * Scripts/webkitpy/layout_tests/layout_package/message_broker_unittest.py: Removed.
32978         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
32979         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
32980         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
32981
32982 2011-04-18  Dirk Pranke  <dpranke@chromium.org>
32983
32984         Reviewed by Eric Seidel.
32985
32986         new-run-webkit-tests: add an --additional-drt-flag option
32987         https://bugs.webkit.org/show_bug.cgi?id=58680
32988
32989         NRWT has a bunch of command line flags that exist to pass
32990         custom flags to DRT, especially on chromium. It would be nice
32991         if there was a generic mechanism to pass through flags so we
32992         didn't have all the custom ones.
32993
32994         * Scripts/webkitpy/layout_tests/port/chromium.py:
32995         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
32996         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
32997         * Scripts/webkitpy/layout_tests/port/test.py:
32998         * Scripts/webkitpy/layout_tests/port/webkit.py:
32999         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
33000
33001 2011-04-05  Jer Noble  <jer.noble@apple.com>
33002
33003         Reviewed by Sam Weinig.
33004
33005         WebKit2: WKTR should support WebKit2 full screen APIs
33006         https://bugs.webkit.org/show_bug.cgi?id=56318
33007
33008         Respond to enterFullScreenForElement() and exitFullScreenForElement(), allowing
33009         WKTR to test the LayoutTest/fullscreen/ tests.
33010
33011         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
33012         (WTR::InjectedBundlePage::InjectedBundlePage): Add support for WKBundlePageFullScreenClient.
33013         (WTR::InjectedBundlePage::supportsFullScreen): Added.
33014         (WTR::InjectedBundlePage::enterFullScreenForElement): Added.
33015         (WTR::InjectedBundlePage::exitFullScreenForElement): Added.
33016         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
33017         * WebKitTestRunner/TestController.cpp:
33018         (WTR::TestController::resetStateToConsistentValues): Enable the full screen preference.
33019
33020 2011-04-18  Dirk Pranke  <dpranke@chromium.org>
33021
33022         Reviewed by Tony Chang.
33023
33024         new-run-webkit-tests: remove chromium-specific relpath hack for --results-directory
33025         https://bugs.webkit.org/show_bug.cgi?id=58388
33026
33027         * Scripts/webkitpy/layout_tests/port/chromium.py:
33028
33029 2011-04-18  Evan Martin  <evan@chromium.org>
33030
33031         Reviewed by Eric Seidel.
33032
33033         [chromium] expose title direction to webkit client
33034         https://bugs.webkit.org/show_bug.cgi?id=58823
33035
33036         * DumpRenderTree/chromium/LayoutTestController.cpp:
33037         (LayoutTestController::LayoutTestController):
33038         (LayoutTestController::reset):
33039         * DumpRenderTree/chromium/LayoutTestController.h:
33040         (LayoutTestController::setTitleTextDirection):
33041         Add member and setter to let us stash the title direction at load time.
33042
33043         * DumpRenderTree/chromium/WebViewHost.cpp:
33044         (WebViewHost::didReceiveTitle):
33045         * DumpRenderTree/chromium/WebViewHost.h:
33046         Adapt to new API; push the title into layoutTestController when we
33047         receive it.
33048
33049 2011-04-18  Dirk Pranke  <dpranke@chromium.org>
33050
33051         Unreviewed.
33052
33053         Re-land r84112. It was rolled out in r84127 but it didn't need
33054         to be. Only r84111 actually broke things.
33055
33056         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
33057         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
33058
33059 2011-04-18  Dirk Pranke  <dpranke@chromium.org>
33060
33061         Reviewed by Tony Chang.
33062
33063         new-run-webkit-tests: don't run more processes than necessary
33064         https://bugs.webkit.org/show_bug.cgi?id=58751
33065
33066         Previously NRWT would start up 1 process for each processor on
33067         the machine by default, which was a waste if we were only
33068         running a few shards of tests.
33069
33070         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
33071         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
33072         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
33073
33074 2011-04-18  Ojan Vafai  <ojan@chromium.org>
33075
33076         Reviewed by Eric Seidel.
33077
33078         include links to actual.png files for new pixel tests
33079         https://bugs.webkit.org/show_bug.cgi?id=58831
33080
33081         Load image elements to confirm whether the files exist.
33082         Image elements will load without being appended to the DOM.
33083
33084         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33085
33086 2011-04-18  Ojan Vafai  <ojan@chromium.org>
33087
33088         Reviewed by Eric Seidel.
33089
33090         add a zoomed view for pixel results to the new results html file
33091         https://bugs.webkit.org/show_bug.cgi?id=58827
33092
33093         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33094
33095 2011-04-18  Ojan Vafai  <ojan@chromium.org>
33096
33097         Reviewed by Eric Seidel.
33098
33099         add labels to results html page
33100         https://bugs.webkit.org/show_bug.cgi?id=58803
33101
33102         -fix iframe size to be the size of the png results
33103         -fix a bunch of aesthetic nits
33104         -fix some incorrectly styled code
33105
33106         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33107
33108 2011-04-18  Chang Shu  <cshu@webkit.org>
33109
33110         Reviewed by Eric Seidel.
33111
33112         WebKitTestRunner needs layoutTestController.pathToLocalResource
33113         https://bugs.webkit.org/show_bug.cgi?id=42541
33114
33115         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
33116         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
33117         * WebKitTestRunner/InjectedBundle/mac/LayoutTestControllerMac.mm:
33118         (WTR::LayoutTestController::pathToLocalResource):
33119         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
33120         (WTR::LayoutTestController::pathToLocalResource):
33121         * WebKitTestRunner/InjectedBundle/win/LayoutTestControllerWin.cpp:
33122         (WTR::LayoutTestController::pathToLocalResource):
33123
33124 2011-04-18  Timothy Hatcher  <timothy@apple.com>
33125
33126         Make update-webkit-localizable-strings put WebKit/win strings in WebCore
33127         now that all localized strings in WebKit/win use WEB_UI_STRING.
33128
33129         https://webkit.org/b/58747
33130
33131         Reviewed by Dan Bernstein.
33132
33133         * Scripts/update-webkit-localizable-strings: Remove the code that updated WebKit/win differently.
33134
33135 2011-04-18  Patrick Gansterer  <paroga@webkit.org>
33136
33137         Reviewed by Adam Roben.
33138
33139         [WIN] Use WCHAR instead of TCHAR
33140         https://bugs.webkit.org/show_bug.cgi?id=58755
33141
33142         We always use the UNICODE versions of windows functions, so
33143         the usage of TCHAR makes no sense and mixing them is bad style.
33144
33145         * WinLauncher/PrintWebUIDelegate.cpp:
33146           Also fix various style issues (including indentation and removal of ugly C-casts).
33147         * WinLauncher/WinLauncher.cpp:
33148
33149 2011-04-18  Dominic Cooney  <dominicc@chromium.org>
33150
33151         Reviewed by Andreas Kling.
33152
33153         Add layoutTestController.shadowRoot to Qt DRT.
33154         https://bugs.webkit.org/show_bug.cgi?id=58759
33155
33156         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
33157         (LayoutTestController::shadowRoot): Added.
33158         * DumpRenderTree/qt/LayoutTestControllerQt.h:
33159
33160 2011-04-18  Sheriff Bot  <webkit.review.bot@gmail.com>
33161
33162         Unreviewed, rolling out r84111.
33163         http://trac.webkit.org/changeset/84111
33164         https://bugs.webkit.org/show_bug.cgi?id=58771
33165
33166         Chromium linux layout tests are broken. (Requested by loislo2
33167         on #webkit).
33168
33169         * Scripts/webkitpy/layout_tests/port/chromium.py:
33170         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
33171         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
33172
33173 2011-04-18  Sheriff Bot  <webkit.review.bot@gmail.com>
33174
33175         Unreviewed, rolling out r84112.
33176         http://trac.webkit.org/changeset/84112
33177         https://bugs.webkit.org/show_bug.cgi?id=58770
33178
33179         Chromium linux layout tests are broken. (Requested by loislo2
33180         on #webkit).
33181
33182         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
33183         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
33184
33185 2011-04-17  Daniel Bates  <dbates@webkit.org>
33186
33187         Reviewed by Eric Seidel and Adam Barth.
33188
33189         webkit-patch land hangs if svn prompts for credentials
33190         https://bugs.webkit.org/show_bug.cgi?id=31500
33191
33192         Prompt for username and password when using git-svn and there aren't cached SVN credentials.
33193
33194         * Scripts/webkitpy/common/checkout/scm.py:
33195           - Added mixin class SVNRepository and made both class SVN and Git inherit from it.
33196           - Moved SVN.has_authorization_for_realm() to class SVNRepository and removed default value
33197             for argument realm; modified call sites as needed.
33198           - Modified AuthenticationError constructor to take optional prompt_for_password argument.
33199           - Modified {SCM, SVN, Git}.commit_with_message() to take optional password argument.
33200           - Modified Git._commit_on_branch() to take optional username and password argument.
33201           - Modified Git.push_local_commits_to_server() to take optional username and password
33202             argument and to call has_authorization_for_realm().
33203         * Scripts/webkitpy/common/checkout/scm_unittest.py:
33204           - Modified SVNTest.test_commit_without_authorization() to take dummy realm argument.
33205           - Modified SVNTest.test_not_have_authorization_for_realm() to pass realm argument to
33206             SVN.has_authorization_for_realm().
33207         * Scripts/webkitpy/common/net/credentials.py:
33208           - Modified Credentials.read_credentials() to call User.prompt_password() instead
33209             of using getpass.getpass() directly.
33210         * Scripts/webkitpy/common/system/user.py:
33211           - Added User.prompt_password().
33212         * Scripts/webkitpy/tool/steps/commit.py:
33213           - Modified Commit.run() to prompt for a password if needed.
33214
33215 2011-04-17  Dirk Pranke  <dpranke@chromium.org>
33216
33217         Reviewed by Ojan Vafai.
33218
33219         new-run-webkit-tests: save stderr into the layout-test-results dir
33220         https://bugs.webkit.org/show_bug.cgi?id=58690
33221
33222         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
33223         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
33224
33225 2011-04-17  Dirk Pranke  <dpranke@chromium.org>
33226
33227         Reviewed by Ojan Vafai.
33228
33229         new-run-webkit-tests: read stderr from chromium DRT separately
33230         https://bugs.webkit.org/show_bug.cgi?id=58708
33231
33232         * Scripts/webkitpy/layout_tests/port/chromium.py:
33233         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
33234         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
33235
33236 2011-04-17  Patrick Gansterer  <paroga@webkit.org>
33237
33238         Reviewed by Adam Barth.
33239
33240         Rename PLATFORM(CA) to USE(CA)
33241         https://bugs.webkit.org/show_bug.cgi?id=58742
33242
33243         * DumpRenderTree/mac/DumpRenderTree.mm:
33244         (resetDefaultsToConsistentValues):
33245
33246 2011-04-17  Patrick Gansterer  <paroga@webkit.org>
33247
33248         Reviewed by Adam Barth.
33249
33250         Rename PLATFORM(CG) to USE(CG)
33251         https://bugs.webkit.org/show_bug.cgi?id=58729
33252
33253         * DumpRenderTree/PixelDumpSupport.cpp:
33254         * DumpRenderTree/chromium/WebViewHost.cpp:
33255         (WebViewHost::paintRect):
33256         * DumpRenderTree/config.h:
33257         * DumpRenderTree/win/PixelDumpSupportWin.cpp:
33258         (createBitmapContextFromWebView):
33259         * TestWebKitAPI/PlatformWebView.h:
33260         * WebKitTestRunner/config.h:
33261         * WebKitTestRunner/win/TestInvocationWin.cpp:
33262
33263 2011-04-16  Patrick Gansterer  <paroga@webkit.org>
33264
33265         Reviewed by Eric Seidel.
33266
33267         Rename PLATFORM(CAIRO) to USE(CAIRO)
33268         https://bugs.webkit.org/show_bug.cgi?id=55192
33269
33270         * DumpRenderTree/PixelDumpSupport.cpp:
33271         * DumpRenderTree/config.h:
33272         * DumpRenderTree/win/PixelDumpSupportWin.cpp:
33273         (createBitmapContextFromWebView):
33274         * WebKitTestRunner/config.h:
33275
33276 2011-04-16  Ojan Vafai  <ojan@chromium.org>
33277
33278         Reviewed by Eric Seidel.
33279
33280         show diffs+test inline in the results page
33281         https://bugs.webkit.org/show_bug.cgi?id=58723
33282
33283         Put each row in it's own tbody and sort based on tbodies. That way
33284         we can append the results to the tbody inline as another table row.
33285
33286         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33287
33288 2011-04-16  Ojan Vafai  <ojan@chromium.org>
33289
33290         Reviewed by Eric Seidel.
33291
33292         handle new tests in the new results html file
33293         https://bugs.webkit.org/show_bug.cgi?id=58715
33294
33295         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33296
33297 2011-04-15  Dirk Pranke  <dpranke@chromium.org>
33298
33299         Reviewed by Ojan Vafai.
33300
33301         old-run-webkit-tests: save the list of tests actually run to a file
33302         https://bugs.webkit.org/show_bug.cgi?id=58692
33303
33304         * Scripts/old-run-webkit-tests:
33305
33306 2011-04-15  Dirk Pranke  <dpranke@chromium.org>
33307
33308         Reviewed by Ojan Vafai.
33309
33310         fix MockFileSystem.abspath() on windows
33311         https://bugs.webkit.org/show_bug.cgi?id=58713
33312
33313         The existing code was borked on Windows if a path was passed
33314         in using a drive letter (e.g., "d:\foo.txt"), and would go
33315         into an infinite recursive loop. While win paths shouldn't
33316         normally be passed into the mock filesystem, this just
33317         bulletproofs things a bit (note that "/foo.txt" is considered
33318         an absolute path on windows python as well.
33319
33320         * Scripts/webkitpy/common/system/filesystem_mock.py:
33321
33322 2011-04-15  Ojan Vafai  <ojan@chromium.org>
33323
33324         Reviewed by Eric Seidel.
33325
33326         show stderr files and httpd log files
33327         https://bugs.webkit.org/show_bug.cgi?id=58714
33328
33329         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33330         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
33331
33332 2011-04-15  Shishir Agrawal  <shishir@chromium.org>
33333
33334         Reviewed by James Robinson.
33335
33336         Add a flag to guard Page Visibility API changes.
33337         https://bugs.webkit.org/show_bug.cgi?id=58464
33338
33339         * Scripts/build-webkit:
33340
33341 2011-04-15  Chang Shu  <cshu@webkit.org>
33342
33343         Reviewed by Alexey Proskuryakov.
33344
33345         When a message with url embedded is added to console, the "file:" scheme
33346         and path should be stripped.
33347         https://bugs.webkit.org/show_bug.cgi?id=58665
33348
33349         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
33350         (WTR::lastFileURLPathComponent):
33351         (WTR::InjectedBundlePage::willAddMessageToConsole):
33352
33353 2011-04-15  Jeff Miller  <jeffm@apple.com>
33354
33355         Reviewed by Sam Weinig.
33356
33357         Add takeFocus callback to WKPageUIClient
33358         https://bugs.webkit.org/show_bug.cgi?id=58686
33359
33360         Add empty WKPageUIClient entries for the takeFocus callback in the MiniBrowser and WebKitTestRunner.
33361
33362         * MiniBrowser/mac/BrowserWindowController.m:
33363         (-[BrowserWindowController awakeFromNib]): Added empty takeFocus entry.
33364         * MiniBrowser/win/BrowserView.cpp:
33365         (BrowserView::create): Added empty takeFocus entry.
33366         * WebKitTestRunner/TestController.cpp:
33367         (WTR::TestController::createOtherPage): Added empty takeFocus entry.
33368         (WTR::TestController::initialize): Added empty takeFocus entry.
33369
33370 2011-04-15  Ojan Vafai  <ojan@chromium.org>
33371
33372         Reviewed by Eric Seidel.
33373
33374         only show pretty-diff/wdiff if they were generated
33375         https://bugs.webkit.org/show_bug.cgi?id=58682
33376
33377         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33378         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
33379         * Scripts/webkitpy/layout_tests/port/base.py:
33380
33381 2011-04-15  Andreas Kling  <kling@webkit.org>
33382
33383         Reviewed by Antonio Gomes.
33384
33385         [Qt] DRT: Default custom policy delegate decision should be Ignore.
33386
33387         If layoutTestController.setCustomPolicyDelegate() is called with only
33388         one argument, the second ("permissive") should default to false.
33389
33390         * DumpRenderTree/qt/LayoutTestControllerQt.h:
33391
33392 2011-04-15  Ojan Vafai  <ojan@chromium.org>
33393
33394         Reviewed by Eric Seidel.
33395
33396         fix path to layout test in new results file
33397         https://bugs.webkit.org/show_bug.cgi?id=58618
33398
33399         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33400         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
33401
33402 2011-04-15  Ojan Vafai  <ojan@chromium.org>
33403
33404         Reviewed by Eric Seidel.
33405
33406         fix sorting in new results file
33407         https://bugs.webkit.org/show_bug.cgi?id=58616
33408
33409         -Fix the custom sort to return -1, 0, 1 instead of true/false.
33410         -Secondary sort by test name when sort values are equal.
33411         -Sort by test name by default.
33412
33413         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33414
33415 2011-04-15  Dmitry Lomov  <dslomov@google.com>
33416
33417         Reviewed by David Levin.
33418
33419         check-webkit-style shouldn't complain about not including a primary header file
33420         if none exists
33421         https://bugs.webkit.org/show_bug.cgi?id=39514
33422
33423         * Scripts/webkitpy/style/checkers/cpp.py:
33424         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
33425
33426 2011-04-15  Eric Seidel  <eric@webkit.org>
33427
33428         Reviewed by Mihai Parparita.
33429
33430         queues.webkit.org should display when a bot last rebooted
33431         https://bugs.webkit.org/show_bug.cgi?id=58562
33432
33433         There is more repeated code here than I would like.  I fear
33434         my django-fu isn't quite up to snuff.
33435
33436         * QueueStatusServer/handlers/queuestatus.py:
33437         * QueueStatusServer/templates/queuestatus.html:
33438
33439 2011-04-14  Alok Priyadarshi  <alokp@chromium.org>
33440
33441         Reviewed by Adam Barth.
33442
33443         Remove dependency on chromium skia::PlatformCanvas
33444         https://bugs.webkit.org/show_bug.cgi?id=57563
33445
33446         Fixed compile error on Mac when using Core Graphics.
33447
33448         * DumpRenderTree/chromium/WebViewHost.cpp:
33449         (WebViewHost::paintRect):
33450
33451 2011-04-14  Alok Priyadarshi  <alokp@chromium.org>
33452
33453         Reviewed by James Robinson.
33454
33455         Remove dependency on chromium skia::PlatformCanvas
33456         https://bugs.webkit.org/show_bug.cgi?id=57563
33457
33458         * DumpRenderTree/chromium/TestShell.cpp:
33459         (makeCanvasOpaque):
33460         (TestShell::dumpImage):
33461         * DumpRenderTree/chromium/TestShell.h:
33462         * DumpRenderTree/chromium/WebThemeControlDRTWin.cpp:
33463         (WebThemeControlDRTWin::WebThemeControlDRTWin):
33464         (WebThemeControlDRTWin::draw):
33465         (WebThemeControlDRTWin::drawTextField):
33466         (WebThemeControlDRTWin::drawProgressBar):
33467         * DumpRenderTree/chromium/WebThemeControlDRTWin.h:
33468         * DumpRenderTree/chromium/WebViewHost.cpp:
33469         (WebViewHost::canvas):
33470         * DumpRenderTree/chromium/WebViewHost.h:
33471
33472 2011-04-14  Keith Kyzivat  <keith.kyzivat@nokia.com>
33473
33474         Reviewed by Laszlo Gombos.
33475
33476         [Qt] Add -maximize flag to QtTestBrowser and MiniBrowser
33477         https://bugs.webkit.org/show_bug.cgi?id=58007
33478
33479         Adds flag to start the test browsers maximized.
33480         Symbian started maximized on QtTestBrowser - made MiniBrowser follow
33481         suit.
33482         Also fixes some ToggleFullScreen issues so that prior window state is
33483         restored.
33484
33485         * MiniBrowser/qt/BrowserWindow.cpp:
33486         (BrowserWindow::BrowserWindow):
33487         (BrowserWindow::toggleFullScreenMode):
33488         * MiniBrowser/qt/MiniBrowserApplication.cpp:
33489         (MiniBrowserApplication::handleUserOptions):
33490         * MiniBrowser/qt/MiniBrowserApplication.h:
33491         (WindowOptions::WindowOptions):
33492         * QtTestBrowser/launcherwindow.cpp:
33493         (LauncherWindow::init):
33494         (LauncherWindow::toggleFullScreenMode):
33495         * QtTestBrowser/launcherwindow.h:
33496         (WindowOptions::WindowOptions):
33497         * QtTestBrowser/main.cpp:
33498         (LauncherApplication::handleUserOptions):
33499
33500 2011-04-14  Ojan Vafai  <ojan@chromium.org>
33501
33502         Reviewed by Eric Seidel.
33503
33504         add checkbox to only show unexpected results to new results file
33505         https://bugs.webkit.org/show_bug.cgi?id=58606
33506
33507         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33508
33509 2011-04-13  Brent Fulgham  <bfulgham@webkit.org>
33510
33511         Reviewed by Martin Robinson.
33512
33513         [WinCairo] Implement ImageDiff Logic.
33514         http://bugs.webkit.org/show_bug.cgi?id=58486
33515
33516         * DumpRenderTree/win/ImageDiffCairo.cpp: Added.
33517         (strtof):
33518         (readFromData):
33519         (createImageFromStdin):
33520         (releaseMallocBuffer):
33521         (createDifferenceImage):
33522         (imageHasAlpha):
33523         (writeToData):
33524         (main):
33525         * DumpRenderTree/win/ImageDiff.vcproj: Updated for new
33526           ImageDiffCairo.cpp (excluding from Apple builds).
33527         * DumpRenderTree/win/ImageDiffDebugCairoCFLite.vsprops: Use
33528           new ImageDiffWinCairoCommon.vsprops file.
33529         * DumpRenderTree/win/ImageDiffReleaseCairoCFLite.vsprops: Use
33530           new ImageDiffWinCairoCommon.vsprops file.
33531         * DumpRenderTree/win/ImageDiffWinCairoCommon.vsprops: Added.
33532         * Scripts/old-run-webkit-tests: Make sure proper executable
33533           is built and executed for WinCairo port.
33534
33535 2011-04-14  Ojan Vafai  <ojan@chromium.org>
33536
33537         Reviewed by Eric Seidel.
33538
33539         only show expected failure type for ports that use an expectations file
33540         https://bugs.webkit.org/show_bug.cgi?id=58588
33541
33542         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33543         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
33544
33545 2011-04-14  Eric Seidel  <eric@webkit.org>
33546
33547         Reviewed by Adam Barth.
33548
33549         commit-queue fails to catch IOError when results.html is missing
33550         https://bugs.webkit.org/show_bug.cgi?id=58589
33551
33552         * Scripts/webkitpy/tool/commands/queues.py:
33553         * Scripts/webkitpy/tool/commands/queues_unittest.py:
33554
33555 2011-04-14  Kevin Ollivier  <kevino@theolliviers.com>
33556
33557         [wx] Unreviewed build fix for Mac installation script.
33558
33559         * wx/packaging/build-mac-installer.py:
33560
33561 2011-04-14  Ojan Vafai  <ojan@chromium.org>
33562
33563         Reviewed by Tony Chang.
33564
33565         sort columns in the new json_results.html file
33566         https://bugs.webkit.org/show_bug.cgi?id=58581
33567
33568         Also fix padding on TDs and the path to layout tests.
33569
33570         * Scripts/webkitpy/layout_tests/layout_package/json_results.html:
33571
33572 2011-04-14  Eric Seidel  <eric@webkit.org>
33573
33574         Reviewed by Dimitri Glazkov.
33575
33576         queues.webkit.org should display when a bot last passed a patch
33577         https://bugs.webkit.org/show_bug.cgi?id=58546
33578
33579         I am suspicious that some of our commit-queue instances are never
33580         actually passing anything.
33581
33582         This will let us know if those bots are never passing patches.
33583
33584         * QueueStatusServer/handlers/queuestatus.py:
33585         * QueueStatusServer/index.yaml:
33586         * QueueStatusServer/templates/queuestatus.html:
33587
33588 2011-04-14  Dirk Pranke  <dpranke@chromium.org>
33589
33590         Reviewed by Adam Roben.
33591
33592         REGRESSION (r83764): webkitpy.layout_tests.port.chromium_win_unittest.ChromiumWinTest.test_default_worker_model fails on apple-windows-13
33593         https://bugs.webkit.org/show_bug.cgi?id=58545
33594
33595         Remove platform-specific overrides; the default behavior should
33596         be working everywhere now, I think.
33597
33598         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
33599         * Scripts/webkitpy/layout_tests/port/mac.py:
33600
33601 2011-04-14  Eric Seidel  <eric@webkit.org>
33602
33603         Reviewed by Adam Barth.
33604
33605         commit-queues are hitting an exception trying to land when the tree is red
33606         https://bugs.webkit.org/show_bug.cgi?id=58558
33607
33608         Just an un-tested (and thus broken) code path from yesterday's changes.
33609
33610         * Scripts/webkitpy/common/net/layouttestresults.py:
33611         * Scripts/webkitpy/common/net/layouttestresults_unittest.py:
33612         * Scripts/webkitpy/tool/commands/queues.py:
33613         * Scripts/webkitpy/tool/commands/queues_unittest.py:
33614
33615 2011-04-14  Kevin Ollivier  <kevino@theolliviers.com>
33616
33617         [wx] Unreviewed build fixes for wxMSW and wx 2.9.1.1.
33618
33619         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
33620         (LayoutTestController::shadowRoot):
33621         * wx/build/wxpresets.py:
33622
33623 2011-04-14  Victoria Kirst  <vrk@google.com>
33624
33625         Reviewed by James Robinson.
33626
33627         [chromium] Failing GPU video tests
33628         https://bugs.webkit.org/show_bug.cgi?id=57926
33629
33630         Added/updated GPU directory names for rebaseline tool.
33631
33632         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
33633
33634 2011-04-13  Daniel Bates  <dbates@rim.com>
33635
33636         Reviewed by Adam Barth.
33637
33638         Perl unit test removeEOL.pl should print test case name on failure
33639         https://bugs.webkit.org/show_bug.cgi?id=58513
33640
33641         * Scripts/webkitperl/VCSUtils_unittest/removeEOL.pl: Pass $title as
33642           second argument of Test::Simple::ok().
33643
33644 2011-04-13  Kent Tamura  <tkent@chromium.org>
33645
33646         Reviewed by Eric Seidel.
33647
33648         gdb-safari should pass arguments to safari.
33649         https://bugs.webkit.org/show_bug.cgi?id=58439
33650
33651         We can do:
33652           gdb-safari --debug LayoutTests/.../foo.html
33653
33654         * Scripts/gdb-safari: Passing @ARGV to Safari via gdb --arg.
33655
33656 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
33657
33658         Unreviewed, build fix.
33659         https://bugs.webkit.org/show_bug.cgi?id=52763
33660
33661         r83799 broke test-webkitpy because it assumed we always have
33662         access to a real filesystem (we don't during the unit tests).
33663         Stubbing out the copy for now but we might need a better
33664         solution at some point.
33665
33666         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
33667
33668 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
33669
33670         Reviewed by Tony Chang.
33671
33672         new-run-webkit-tests --run-singly option is busted
33673         https://bugs.webkit.org/show_bug.cgi?id=55909
33674
33675         Python's broken lexical scoping strikes again :(.
33676
33677         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
33678         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
33679
33680 2011-04-13  Ojan Vafai  <ojan@chromium.org>
33681
33682         Reviewed by Eric Seidel.
33683
33684         generated unexpected_results.html from unexpected_results.json
33685         https://bugs.webkit.org/show_bug.cgi?id=52763
33686
33687         Eventually, we'll merge this with results.html and have a single richer results page.
33688         For now, I just want to get something checked in that we can iterate on.
33689
33690         * Scripts/webkitpy/layout_tests/layout_package/json_results.html: Added.
33691         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
33692
33693 2011-04-13  Eric Seidel  <eric@webkit.org>
33694
33695         Reviewed by Adam Barth.
33696
33697         Up the failure limit to allow the commit-queue to actually land when the tree is red
33698         https://bugs.webkit.org/show_bug.cgi?id=58499
33699
33700         This change increases the --exit-after-N-failures limit used by the commit-queue
33701         from 1 to 10.  This will cause the code added in bug 58494 to actually get
33702         exercised and the queues should start being able to land when the trees are red.
33703
33704         When testing I found that test_runtests_leopard_commit_queue_hack_command was the
33705         only unit test to actually detect this change!  Since we don't actually
33706         run on leopard anymore, I removed the hack code and repurposed the test.
33707
33708         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
33709         * Scripts/webkitpy/tool/steps/runtests.py:
33710         * Scripts/webkitpy/tool/steps/steps_unittest.py:
33711
33712 2011-04-13  Eric Seidel  <eric@webkit.org>
33713
33714         Reviewed by Adam Barth.
33715
33716         commit-queue should be able to land when tree is red
33717         https://bugs.webkit.org/show_bug.cgi?id=58494
33718
33719         There is some yak hair on my hands, I will admit.
33720
33721         This change is mostly about adding an ExpectedFailures
33722         class to track when the bots are red and we should be
33723         ignoring failures when landing from the commit-queue.
33724
33725         However, to make intelligent decisions about patches we
33726         need to know whether the run hit the --exit-after-N-failures limit
33727         or not.  Right now that information is not saved off in results.html
33728         so we have to pull the information from RunTests.
33729
33730         I've plumbed the --exit-after-N-failures information into
33731         LayoutTestResults for now to make the ExpectedFailures code cleaner.
33732
33733         As a result of adding all these additional calls to delegate.layout_test_results()
33734         I broke some of our flaky test detection tests and had to re-write them
33735         to not depend on the number of layout_test_results code.
33736
33737         At the same time I updated the commit-queue to use the newer filesystem
33738         API (to allow us to use MockFileSystem) which required further changes
33739         to the layout tests.  Changes were required in either case, since
33740         we're now calling layout_test_results() in more cases, which previously
33741         would try and hit the disk (until I moved it to use tool.filesystem).
33742
33743         I should note that *all* of this code is disabled for now, since our
33744         --exit-after-N-failures limit is currently 1!  (Thus were always in the
33745         case where we can't actually tell if the layout test results are legit.)
33746         I will up that limit in a second patch (which may require a couple more unit test tweaks).
33747
33748         * Scripts/webkitpy/common/net/layouttestresults.py:
33749         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
33750         * Scripts/webkitpy/tool/bot/commitqueuetask_unittest.py:
33751         * Scripts/webkitpy/tool/bot/expectedfailures.py: Added.
33752         * Scripts/webkitpy/tool/bot/expectedfailures_unittest.py: Added.
33753         * Scripts/webkitpy/tool/commands/queues.py:
33754         * Scripts/webkitpy/tool/commands/queues_unittest.py:
33755         * Scripts/webkitpy/tool/commands/queuestest.py:
33756         * Scripts/webkitpy/tool/steps/runtests.py:
33757
33758 2011-04-13  Brent Fulgham  <bfulgham@webkit.org>
33759
33760         Unreviewed build correction.
33761
33762         * TestWebKitAPI/PlatformWebView.h: Limit include of
33763           <CoreGraphics/CGGeometry> to PLATFORM(CG) users.
33764
33765 2011-04-13  Mihai Parparita  <mihaip@chromium.org>
33766
33767         Reviewed by Eric Seidel.
33768
33769         [NRWT] Rename --baseline-search-patch to --additional-platform-directory to match ORWT
33770         https://bugs.webkit.org/show_bug.cgi?id=58489
33771
33772         r83743 added --additional-platform-directory to old-run-webkit-tests,
33773         rename the equivalent flag in new-run-webkit-tests to be consistent.
33774
33775         * Scripts/webkitpy/layout_tests/port/base.py:
33776         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
33777         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
33778         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
33779
33780 2011-04-13  Mihai Parparita  <mihaip@chromium.org>
33781
33782         Reviewed by Eric Seidel.
33783
33784         Rebaseline queue server
33785         https://bugs.webkit.org/show_bug.cgi?id=57891
33786
33787         Adds a simple rebaseline queue server (meant to run at
33788         http://rebaseline-queue.appspot.com/). It presents a list of all (NRWT)
33789         builders and their failing tests, and allows those tests to be added to
33790         a per-builer rebaseline queue.
33791
33792         A subsequent change will add a mode to NRWT (and/or a wrapper script)
33793         for pulling the tests to be rebaselined from the server and running them
33794         with --reset-results (the initial use-case is for the GPU hardware bots,
33795         which do not have checked-in baselines, therefore there is no need for a
33796         check-in step yet).
33797
33798         * RebaselineQueueServer/app.yaml: Added.
33799         * RebaselineQueueServer/handlers/__init__.py: Added.
33800         * RebaselineQueueServer/handlers/builderqueue.py: Added.
33801         * RebaselineQueueServer/handlers/pages.py: Added.
33802         * RebaselineQueueServer/index.yaml: Added.
33803         * RebaselineQueueServer/main.py: Added.
33804         * RebaselineQueueServer/model/__init__.py: Added.
33805         * RebaselineQueueServer/model/queueentry.py: Added.
33806         * RebaselineQueueServer/static/builder-frame-empty.html: Added.
33807         * RebaselineQueueServer/static/styles.css: Added.
33808         * RebaselineQueueServer/templates/builder-picker.html: Added.
33809         * RebaselineQueueServer/templates/builder-queue-edit.html: Added.
33810         * RebaselineQueueServer/templates/builder-queue-list.html: Added.
33811         * RebaselineQueueServer/templates/home.html: Added.
33812
33813 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
33814
33815         Reviewed by Mihai Parparita.
33816
33817         nrwt: enable multiple processes by default on Chromium Win.
33818         https://bugs.webkit.org/show_bug.cgi?id=55163
33819
33820         Re-land r79268; it should be more stable now and will hopefully
33821         work.
33822
33823         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
33824         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
33825
33826 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
33827
33828         Reviewed by Ojan Vafai.
33829
33830         new-run-webkit-tests: configure logging in child processes properly
33831         https://bugs.webkit.org/show_bug.cgi?id=58296
33832
33833         Logging propagated properly from manager to worker processes on
33834         Unix but not on Windows; this fixes that.
33835
33836         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker.py:
33837         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
33838
33839 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
33840
33841         Reviewed by Ojan Vafai.
33842
33843         new-run-webkit-tests: suppress extraneous pretty patch warnings
33844         https://bugs.webkit.org/show_bug.cgi?id=58295
33845
33846         NRWT wasn't being careful about when it logged messages from
33847         PrettyPatch not being available; it should only log during
33848         check_build(), and be silent otherwise. This was causing us to
33849         get multiple errors at runtime, which was confusing.
33850
33851         * Scripts/webkitpy/layout_tests/port/base.py:
33852         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
33853         * Scripts/webkitpy/layout_tests/port/chromium.py:
33854
33855 2011-04-13  Dirk Pranke  <dpranke@chromium.org>
33856
33857         Reviewed by Jian Li.
33858
33859         add chromium-linux-x86_64 to rebaseline-chromium-webkit-tests
33860         https://bugs.webkit.org/show_bug.cgi?id=58461
33861
33862         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
33863
33864 2011-04-13  Tony Chang  <tony@chromium.org>
33865
33866         Unreviewed, fix NRWT.  Stack:
33867
33868             return super(ChromiumPort, self).results_directory(self)
33869         TypeError: results_directory() takes exactly 1 argument (2 given)
33870
33871         * Scripts/webkitpy/layout_tests/port/chromium.py:
33872
33873 2011-04-13  Adam Roben  <aroben@apple.com>
33874
33875         Teach ORWT about extra platform-specific directories to be searched before the default ones
33876
33877         Fixes <http://webkit.org/b/58446> ORWT needs a way to look in a user-specified directory for
33878         platform-specific results
33879
33880         Reviewed by Anders Carlsson.
33881
33882         * Scripts/old-run-webkit-tests:
33883         (top-level): Added @additionalPlatformDirectories and hooked up
33884         --additional-platform-directories to populate it.
33885         (buildPlatformResultHierarchy): Prepend the additional platform directories to the default
33886         list (after filtering out any of the additional directories that don't exist).
33887
33888 2011-04-13  Lucas Forschler  <lforschler@apple.com>
33889
33890         Reviewed by Adam Roben.
33891
33892         Add the /T command to taskkill.exe.
33893         This will kill spawned child processes.
33894
33895         * BuildSlaveSupport/win/kill-old-processes:
33896
33897 2011-04-13  Lucas Forschler  <lforschler@apple.com>
33898
33899         Reviewed by Adam Roben.
33900
33901         Sort kill list alphabeticaly.
33902         Fixed capitalization on VcBuildHelper.
33903
33904         * BuildSlaveSupport/win/kill-old-processes:
33905
33906 2011-04-13  Lucas Forschler  <lforschler@apple.com>
33907
33908         Reviewed by Adam Roben.
33909
33910         https://bugs.webkit.org/show_bug.cgi?id=58384
33911         Update kill-old-processes to include a few new ones.
33912
33913         * BuildSlaveSupport/win/kill-old-processes:
33914
33915 2011-04-13  Gustavo Noronha Silva  <gns@gnome.org>
33916
33917         Reviewed by Martin Robinson.
33918
33919         [GTK] PluginPackage should check whether a plugin mixes GTK+ 2 and 3 symbols itself
33920         https://bugs.webkit.org/show_bug.cgi?id=58297
33921
33922         Do not disable flash specifically when built with GTK+ 3 -
33923         PluginPackage should do the right thing while loading plugins.
33924
33925         * GtkLauncher/main.c:
33926         (main):
33927
33928 2011-04-12  Kent Tamura  <tkent@chromium.org>
33929
33930         Reviewed by Dimitri Glazkov.
33931
33932         [Chromium] Add WebSettings::setValidationMessageTimerMagnification()
33933         https://bugs.webkit.org/show_bug.cgi?id=57426
33934
33935         * DumpRenderTree/chromium/WebPreferences.cpp:
33936         (WebPreferences::applyTo): Disable the validation message timer.
33937
33938 2011-04-12  James Kozianski  <koz@chromium.org>
33939
33940         Unreviewed.
33941
33942         Adding myself to committers.py.
33943
33944         * Scripts/webkitpy/common/config/committers.py:
33945
33946 2011-04-12  Brent Fulgham  <bfulgham@webkit.org>
33947
33948         Unreviewed build correction for WinCairo bot after r83639.
33949         https://bugs.webkit.org/show_bug.cgi?id=51790
33950
33951         Correct typo for the new "update-webkit-wincairo-libs" script.
33952
33953         * Scripts/build-webkit: Incorrectly calling the
33954           update-webkit-cairo-libs which does not exist.
33955
33956 2011-04-12  Alice Liu  <alice.liu@apple.com>
33957
33958         Reviewed by Jon Honeycutt.
33959
33960         A change that should have been part of http://trac.webkit.org/changeset/83628
33961
33962         * DumpRenderTree/win/DumpRenderTree.cpp:
33963         (resetDefaultsToConsistentValues): Setting should be off for most tests. Reset to false.
33964
33965 2011-04-12  Dirk Pranke  <dpranke@chromium.org>
33966
33967         Reviewed by Tony Chang.
33968
33969         new-run-webkit-tests: --results-directory is relative to builddir, not $PWD
33970         https://bugs.webkit.org/show_bug.cgi?id=58272
33971
33972         NRWT was interpreting the --results-directory cmd line arg as
33973         relative to the build directory, not the current working
33974         directory (ORWT uses the latter, which is much more intuitive).
33975
33976         This patch fixes the base case, but includes an override for
33977         Chromium that is needed until the bots can be updated.
33978
33979         * Scripts/webkitpy/layout_tests/port/base.py:
33980         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
33981         * Scripts/webkitpy/layout_tests/port/chromium.py:
33982         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
33983         * Scripts/webkitpy/layout_tests/port/test.py:
33984         * Scripts/webkitpy/layout_tests/port/webkit.py:
33985         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
33986         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
33987
33988 2011-04-12  Philippe Normand  <pnormand@igalia.com>
33989
33990         Unreviewed, roll out r83621 as it broke the GTK build.
33991
33992         [GTK] arguments passed to build-jsc and build-webkit scripts are not taken into account
33993         https://bugs.webkit.org/show_bug.cgi?id=58333
33994
33995         * Scripts/build-jsc:
33996         * Scripts/build-webkit:
33997
33998 2011-04-12  Carl Lobo  <carllobo@gmail.com> and Brent Fulgham <bfulgham@webkit.org>
33999
34000         Reviewed by Adam Roben.
34001
34002         Patch to download the WinCairo dependancies as part of build-webkit.
34003         https://bugs.webkit.org/show_bug.cgi?id=51790
34004
34005         This patch downloads the dependancies from
34006             http://idisk.mac.com/bfulgham-Public/WinCairoRequirements.zip
34007         checking the modification timestamp etc.
34008
34009         * Tools/Scripts/build-webkit
34010         * Tools/Scripts/update-webkit: Modified to understand the --wincairo
34011           flag, and to update WinCairo dependencies when used.
34012         * Tools/Scripts/update-webkit-cairo-libs: added
34013         * Tools/Scripts/update-webkit-dependency: added
34014         * Tools/Scripts/update-webkit-auxiliary-libs
34015
34016 2011-04-12  Jeff Miller  <jeffm@apple.com>
34017
34018         Reviewed by Adam Roben.
34019
34020         WebKit2: Pressing Tab in Web Inspector's console does not cycle through completion options
34021         https://bugs.webkit.org/show_bug.cgi?id=56020
34022
34023         Don't call TranslateMessage() in the MiniBrowser or TestWebKitAPI for key messages destined for a WebKit2 view,
34024         since WebKit will do this for us.  If we didn't do this, TranslateMessage() would be called twice,
34025         which would generate two characters for every keypress (for example).  I didn't bother doing this for
34026         WebKitTestRunner, since it doesn't get any WM_KEYDOWN events.
34027
34028         Add new WebKit2/TranslateMessageGeneratesWMChar test to test expected TranslateMessage() behavior.
34029
34030         * MiniBrowser/win/main.cpp:
34031         (shouldTranslateMessage): Added.
34032         (_tWinMain): Don't call TranslateMessage() unless shouldTranslateMessage() says to.
34033         * TestWebKitAPI/PlatformUtilities.h: Added shouldTranslateMessage() on Windows.
34034         * TestWebKitAPI/PlatformWebView.h: Added simulateAKeyDown().
34035         * TestWebKitAPI/Tests/WebKit2/win/TranslateMessageGeneratesWMChar.cpp: Added.
34036         (TestWebKitAPI::didNotHandleKeyEventCallback): Added.
34037         (TestWebKitAPI::runAndWatchForWMChar): Added.
34038         (TestWebKitAPI::TEST): Added.
34039         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
34040         (TestWebKitAPI::Util::run): Don't call TranslateMessage() unless shouldTranslateMessage() says to.
34041         (TestWebKitAPI::Util::shouldTranslateMessage): Added.
34042         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
34043         (TestWebKitAPI::PlatformWebView::simulateAKeyDown): Added.
34044         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Added TranslateMessageGeneratesWMChar.cpp.
34045
34046 2011-04-12  Dirk Pranke  <dpranke@chromium.org>
34047
34048         Reviewed by Tony Chang.
34049
34050         new-run-webkit-tests: update code to use port.results_directory() instead of options.results_directory
34051         https://bugs.webkit.org/show_bug.cgi?id=58290
34052
34053         This is some preliminary refactoring for bug 58272.
34054
34055         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
34056         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
34057         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
34058         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
34059         * Scripts/webkitpy/layout_tests/port/test.py:
34060         * Scripts/webkitpy/layout_tests/port/chromium.py:
34061         * Scripts/webkitpy/layout_tests/port/base.py:
34062         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
34063
34064 2011-04-12  Dirk Pranke  <dpranke@chromium.org>
34065
34066         Reviewed by Eric Seidel.
34067
34068         webkitpy: fix mock_filesystem abspath to handle relative paths
34069         and add filesystem.chdir() and filesystem.getcwd() to be able
34070         to test this and mock it out.
34071
34072         https://bugs.webkit.org/show_bug.cgi?id=58288
34073
34074         * Scripts/webkitpy/common/system/filesystem.py:
34075         * Scripts/webkitpy/common/system/filesystem_mock.py:
34076         * Scripts/webkitpy/common/system/filesystem_unittest.py:
34077
34078 2011-04-12  Alice Liu  <alice.liu@apple.com>
34079
34080         Reviewed by Sam Weinig.
34081
34082         https://bugs.webkit.org/show_bug.cgi?id=58292
34083         Provide new setting to allow site icon loading despite disabling automatic image loading in general.
34084
34085         * DumpRenderTree/mac/DumpRenderTree.mm:
34086         (resetDefaultsToConsistentValues): Setting should be off for most tests. Reset to false.
34087
34088 2011-04-12  Thouraya ANDOLSI  <thouraya.andolsi@st.com>
34089
34090         Reviewed by Eric Seidel.
34091
34092         [Qt] Enable JIT build for SH4 platforms.
34093         https://bugs.webkit.org/show_bug.cgi?id=58317
34094
34095
34096         * DumpRenderTree/qt/DumpRenderTree.pro:
34097         * QtTestBrowser/QtTestBrowser.pro:
34098         * Scripts/webkitdirs.pm:
34099
34100 2011-04-12  Tony Chang  <tony@chromium.org>
34101
34102         Reviewed by Andreas Kling.
34103
34104         [Qt] embed checksums in PNGs written by Qt-DRT
34105         https://bugs.webkit.org/show_bug.cgi?id=58173
34106
34107         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
34108         (WebCore::DumpRenderTree::dump):
34109
34110 2011-04-12  Eric Seidel  <eric@webkit.org>
34111
34112         Reviewed by Adam Barth.
34113
34114         commit-queue should upload failure diffs when tests fail
34115         https://bugs.webkit.org/show_bug.cgi?id=58348
34116
34117         This change was mostly just plumbing.  We were already saving
34118         this information for flaky test reporting.  I just made it possible
34119         for normal failures to report archives as well.
34120
34121         I did a little abstraction work to try and share some code between
34122         flakytestreporter.py and this new code.  There is still more we could do.
34123
34124         In making this change I also went through and updated the various
34125         places we have urls hard-coded in our python and pointed them at
34126         common.config.urls.
34127
34128         * Scripts/webkitpy/common/checkout/scm.py:
34129         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
34130         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
34131         * Scripts/webkitpy/common/net/statusserver.py:
34132         * Scripts/webkitpy/layout_tests/port/base.py:
34133         * Scripts/webkitpy/tool/bot/botinfo.py: Added.
34134         * Scripts/webkitpy/tool/bot/botinfo_unittest.py: Added.
34135         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
34136         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
34137         * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
34138         * Scripts/webkitpy/tool/commands/queues.py:
34139
34140 2011-04-12  Daniel Bates  <dbates@rim.com>
34141
34142         Rubber-stamped by Eric Seidel.
34143
34144         svn-apply and/or patch(1) has trouble applying patches that makes changes to files
34145         with Windows line endings
34146         https://bugs.webkit.org/show_bug.cgi?id=53625
34147
34148         Fixes an issue where a patch made on a SCM checkout on a Unix file system cannot be
34149         applied to an SCM checkout on a Windows file system and vice-versa.
34150
34151         This issue manifests itself due to custom SCM settings as well as differences in
34152         gitattributes support in older versions of Git (e.g. 1.7.0.3). We implement support
34153         into svn-apply/unapply to compensate for differences in line endings by converting
34154         the line endings in a patch to match the line endings in the target file (if it exists).
34155
34156         * Scripts/VCSUtils.pm:
34157           - Added parseFirstEOL(), which returns the first end-of-line character(s) that appear in the
34158             content read from the specified file handle. Obviously, this heuristic will not work
34159             for files with mixed line endings, but I don't envision such files to be the norm. If
34160             this turns out to be an issue then we can improve the heuristic.
34161           - Added firstEOLInFile().
34162           - Modified parsePatch() and parseDiff() to take an optional hash reference to an options hash.
34163             In particular, added the hash key shouldNotUseIndexPathEOL to control whether to use the line
34164             endings in the diff instead of the line endings in the target file.
34165           - Extracted local variable $chunkRangeRegEx from fixChangeLogPatch() so that it can be
34166             re-used in parseDiff().
34167         * Scripts/webkitperl/VCSUtils_unittest/parseDiff.pl: Modified to pass {shouldNotUseIndexPathEOL => 1}
34168           to parsePatch() since these unit tests don't create mock files. Instead, created unit tests that use
34169           mock files in file VCSUtils_unittest/parseDiffWithMockFiles.pl.
34170         * Scripts/webkitperl/VCSUtils_unittest/parseDiffWithMockFiles.pl: Added.
34171         * Scripts/webkitperl/VCSUtils_unittest/parseFirstEOL.pl: Added.
34172
34173 2011-04-12  Sam Weinig  <sam@webkit.org>
34174
34175         Reviewed by Adam Roben.
34176
34177         REGRESSION (r83550): Multiple http tests crashing in the web process in SQLite beneath CFNetwork on Windows 7 Release (WebKit2 Tests)
34178         https://bugs.webkit.org/show_bug.cgi?id=58336
34179
34180         * WebKitTestRunner/TestController.cpp:
34181         (WTR::TestController::resetStateToConsistentValues):
34182         Remove cache clearing as that is causing tests to crash on Windows 7.
34183
34184 2011-04-11  Sam Weinig  <sam@webkit.org>
34185
34186         Reviewed by Anders Carlsson.
34187
34188         REGRESSION(r77974): http/tests/security/aboutBlank/security-context-window-open.html is failing
34189         <rdar://problem/8981346>
34190         https://bugs.webkit.org/show_bug.cgi?id=54159
34191
34192         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
34193         (WTR::InjectedBundlePage::InjectedBundlePage):
34194         (WTR::InjectedBundlePage::decidePolicyForNavigationAction):
34195         (WTR::InjectedBundlePage::decidePolicyForNewWindowAction):
34196         (WTR::InjectedBundlePage::decidePolicyForResponse):
34197         (WTR::InjectedBundlePage::unableToImplementPolicy):
34198         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.h:
34199         Add short-circuited policy client which matches WebKit1.
34200
34201         * WebKitTestRunner/TestController.cpp:
34202         (WTR::TestController::resetStateToConsistentValues):
34203         Clear the cache between tests to ensure consistent results.
34204
34205 2011-04-11  George Guo  <George.Guo@nokia.com>
34206
34207         Reviewed by Eric Seidel.
34208
34209         prepare-ChangeLog support email input but -h did not show the option
34210         https://bugs.webkit.org/show_bug.cgi?id=58164
34211
34212         Add email option to -h to improve the usability
34213
34214         * Scripts/prepare-ChangeLog:
34215
34216 2011-04-11  Dimitri Glazkov  <dglazkov@chromium.org>
34217
34218         Reviewed by Eric Carlson.
34219
34220         Rename MediaControls to MediaControlRootElement.
34221         https://bugs.webkit.org/show_bug.cgi?id=58250
34222
34223         * Scripts/do-webcore-rename: Documented the renaming.
34224
34225 2011-04-11  Sam Weinig  <sam@webkit.org>
34226
34227         Fix Windows build.
34228
34229         * MiniBrowser/win/BrowserView.cpp:
34230         (BrowserView::create):
34231
34232 2011-04-11  Sam Weinig  <sam@webkit.org>
34233
34234         Reviewed by Anders Carlsson.
34235
34236         Move focus management out of WebKit via  the UIClient
34237         <rdar://problem/8784068>
34238         https://bugs.webkit.org/show_bug.cgi?id=58278
34239
34240         * MiniBrowser/mac/BrowserWindowController.m:
34241         (-[BrowserWindowController awakeFromNib]):
34242         * WebKitTestRunner/TestController.cpp:
34243         (WTR::TestController::createOtherPage):
34244         (WTR::TestController::initialize):
34245         Add stubs for new UIClient functions.
34246
34247 2011-02-03  Martin Robinson  <mrobinson@igalia.com>
34248
34249         Reviewed by Eric Seidel.
34250
34251         [GTK] The GTK+ DRT needs an implementation of the PlainTextController
34252         https://bugs.webkit.org/show_bug.cgi?id=53605
34253
34254         Add an implementation of the PlainTextController for the GTK+ DRT.
34255
34256         * DumpRenderTree/gtk/DumpRenderTree.cpp:
34257         (addControllerToWindow): Added this helper method.
34258         (webViewWindowObjectCleared): Use the new helper to add EventSender
34259         and the PlainTextController to the window object.
34260         * GNUmakefile.am: Added PlainTextController source files.
34261         * DumpRenderTree/gtk/PlainTextController.cpp: Added. This will rely
34262         on DumpRenderTreeSupportGtk to convert a JSValueRef into a WebKitDOMRange
34263         until that functionality exists somewhere in the WebKitGTK+ stack.
34264         * DumpRenderTree/gtk/PlainTextController.h: Added.
34265
34266 2011-04-11  Yael Aharon  <yael.aharon@nokia.com>
34267
34268         Unreviewed.
34269
34270         Adding my IRC nick.
34271
34272         * Scripts/webkitpy/common/config/committers.py:
34273
34274 2011-04-11  Dirk Pranke  <dpranke@chromium.org>
34275
34276         Reviewed by Adam Roben.
34277
34278         new-run-webkit-tests: fix crash when running under windows cmd.exe
34279         https://bugs.webkit.org/show_bug.cgi?id=58197
34280
34281         We would intermittently crash when running NRWT directly under
34282         cmd.exe on windows because we were sharing the stdin file
34283         descriptor between the python process and the http server.
34284         cmd.exe really didn't like that, and there was no reason to
34285         share the descriptor, so we now use a PIPE instead.
34286
34287         * Scripts/webkitpy/layout_tests/port/http_server.py:
34288
34289 2011-04-11  Dirk Pranke  <dpranke@chromium.org>
34290
34291         Reviewed by Ojan Vafai.
34292
34293         new-run-webkit-tests: add unit tests for Port.diff_image()
34294         https://bugs.webkit.org/show_bug.cgi?id=58196
34295
34296         Add the unit tests for the fix in bug 58195.
34297
34298         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
34299         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
34300         * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
34301
34302 2011-04-11  Dirk Pranke  <dpranke@chromium.org>
34303
34304         Reviewed by Tony Chang.
34305
34306         new-run-webkit-tests: implement support for audio, take two
34307         https://bugs.webkit.org/show_bug.cgi?id=58195
34308
34309         Attempt to re-land the fix for bug 58101 (which was initially
34310         landed in r83330, but rolled out). This is the same patch but
34311         fixes crashes in Port.diff_image() caused by the change for
34312         empty image files being None instead of ''.
34313
34314         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
34315         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
34316         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
34317         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
34318         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
34319         * Scripts/webkitpy/layout_tests/port/base.py:
34320         * Scripts/webkitpy/layout_tests/port/chromium.py:
34321         * Scripts/webkitpy/layout_tests/port/dryrun.py:
34322         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
34323         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
34324         * Scripts/webkitpy/layout_tests/port/test.py:
34325         * Scripts/webkitpy/layout_tests/port/webkit.py:
34326
34327 2011-04-11  Mario Sanchez Prada  <msanchez@igalia.com>
34328
34329         Reviewed by Xan Lopez.
34330
34331         [GTK] Implement AccessibilityUIElement::{row|column}IndexRange in DRT
34332         https://bugs.webkit.org/show_bug.cgi?id=57854
34333
34334         Implement missing features in GTK's DRT.
34335
34336         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
34337         (indexRangeInTable): New helper function to get the range string
34338         for the current cell inside the parent table, either from the
34339         point of view of rows or columns.
34340         (AccessibilityUIElement::rowIndexRange): Implemented by relying on
34341         the new helper function indexRangeInTable().
34342         (AccessibilityUIElement::columnIndexRange): Ditto.
34343
34344 2011-04-11  Mario Sanchez Prada  <msanchez@igalia.com>
34345
34346         Reviewed by Chris Fleizach.
34347
34348         [GTK] Implement AccessibilityUIElement::cellForColumnAndRow in DRT
34349         https://bugs.webkit.org/show_bug.cgi?id=57826
34350
34351         Implement missing feature in GTK's DRT.
34352
34353         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
34354         (AccessibilityUIElement::cellForColumnAndRow): Implemented, by
34355         relying on the analogous function from the AtkTable interface.
34356
34357 2011-04-05  Timothy Hatcher  <timothy@apple.com>
34358
34359         Fix the extract-localizable-strings script errors that started after
34360         WEB_UI_STRING was added to LocalizedStrings.h in WebCore.
34361
34362         * Scripts/extract-localizable-strings: Skip LocalizedStrings.h.
34363
34364 2011-04-11  Tony Chang  <tony@chromium.org>
34365
34366         Reviewed by Eric Seidel.
34367
34368         fix crc computation on 64 bit machines
34369         https://bugs.webkit.org/show_bug.cgi?id=58243
34370
34371         unsigned long is dependenct on CPU architecture, so use unsigned which is always 32 bits
34372
34373         * DumpRenderTree/CyclicRedundancyCheck.cpp:
34374         (makeCrcTable):
34375         (computeCrc):
34376         * DumpRenderTree/CyclicRedundancyCheck.h:
34377         * DumpRenderTree/PixelDumpSupport.cpp:
34378         (appendIntToVector):
34379         (convertChecksumToPNGComment):
34380
34381 2011-04-10  Kimmo Kinnunen  <kimmo.t.kinnunen@nokia.com>
34382
34383         Reviewed by Eric Seidel.
34384
34385         Require no undefined symbols during compilation.
34386
34387         [Qt] [WK2] WebKitTestRunner, QtWebProcess and WTRInjectBundle should fail to compile when there's undefined symbols
34388         https://bugs.webkit.org/show_bug.cgi?id=54896
34389
34390         Add -Wl,--no-undefined to catch missing symbols early.
34391
34392         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
34393         * WebKitTestRunner/qt/WebKitTestRunner.pro:
34394
34395 2011-04-10  Kevin Ollivier  <kevino@theolliviers.com>
34396
34397         [wx] Unreviewed build fix. Remove collector dirs from the list of build dirs.
34398
34399         * wx/build/settings.py:
34400
34401 2011-04-10  Kevin Ollivier  <kevino@theolliviers.com>
34402
34403         [wx] Unreviewed build fix. Add Source/JavaScriptCore/heap to the list of directories.
34404
34405         * wx/build/settings.py:
34406
34407 2011-04-10  David Levin  <levin@chromium.org>
34408
34409         Reviewed by Dimitri Glazkov.
34410
34411         REGRESSION(r83384): Change to chromium.py is resulting in failures of the test framework on linux.
34412         https://bugs.webkit.org/show_bug.cgi?id=58201
34413
34414         * Scripts/webkitpy/layout_tests/port/chromium.py: Commented out the assert for now,
34415         and filed bug 58202.
34416
34417 2011-04-09  David Levin  <levin@chromium.org>
34418
34419         Unreviewed, rolling out r83394.
34420         http://trac.webkit.org/changeset/83394
34421         https://bugs.webkit.org/show_bug.cgi?id=53625
34422
34423         Patch was incorrect as noted in the bug.
34424
34425         * Scripts/svn-apply:
34426
34427 2011-04-09  David Levin  <levin@chromium.org>
34428
34429         Reviewed by Adam Barth.
34430
34431         svn-apply and/or patch(1) has trouble applying patches that makes changes to files with Windows line endings
34432         https://bugs.webkit.org/show_bug.cgi?id=53625
34433
34434         * Scripts/svn-apply: Ensure that the portions of patches
34435         which are for vcproj/vsprops files has DOS line endings.
34436
34437 2011-04-09  Dirk Pranke  <dpranke@chromium.org>
34438
34439         Reviewed by Tony Chang.
34440
34441         It looks like NRWT has not been stopping DRT/TestShell
34442         instances properly on windows, probably for a long time.
34443         This would go a long way to explaining why we often have
34444         processes lying around :)
34445
34446         https://bugs.webkit.org/show_bug.cgi?id=57807
34447
34448         * Scripts/webkitpy/layout_tests/port/chromium.py:
34449         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
34450
34451 2011-04-09  Keith Kyzivat  <keith.kyzivat@nokia.com>
34452
34453         Reviewed by Laszlo Gombos.
34454
34455         [Qt] Don't link against fontconfig or X11 if embedded
34456         https://bugs.webkit.org/show_bug.cgi?id=58104
34457
34458         * DumpRenderTree/qt/DumpRenderTree.pro: fontconfig !included in embedded
34459         * QtTestBrowser/QtTestBrowser.pro: ditto
34460         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro: ditto
34461         * WebKitTestRunner/qt/WebKitTestRunner.pro: ditto
34462
34463 2011-04-08  Jon Honeycutt  <jhoneycutt@apple.com>
34464
34465         http/tests/loading/preload-append-scan.php is failing on Windows
34466         https://bugs.webkit.org/show_bug.cgi?id=58178
34467
34468         Reviewed by Brian Weinstein.
34469
34470         * DumpRenderTree/win/DumpRenderTree.cpp:
34471         (lastPathComponent):
34472         Use CFURL API to find the last path component, rather than
34473         PathFindFileName(), because the latter will include the query string.
34474
34475 2011-04-08  Jian Li  <jianli@chromium.org>
34476
34477         Unreviewed, rolling out r83327, r83330 since these patches are very
34478         likely to break chromium webkit mac10.6 builders.
34479
34480         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
34481         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
34482         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
34483         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
34484         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
34485         * Scripts/webkitpy/layout_tests/port/base.py:
34486         * Scripts/webkitpy/layout_tests/port/chromium.py:
34487         * Scripts/webkitpy/layout_tests/port/dryrun.py:
34488         * Scripts/webkitpy/layout_tests/port/mock_drt.py:
34489         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
34490         * Scripts/webkitpy/layout_tests/port/test.py:
34491         * Scripts/webkitpy/layout_tests/port/webkit.py:
34492
34493 2011-04-08  Vincent Scheib  <scheib@chromium.org>
34494
34495         Reviewed by Eric Seidel.
34496
34497         add chromium-gpu-linux-x86_64 to the recognized list of ports
34498         https://bugs.webkit.org/show_bug.cgi?id=58099
34499
34500         Add chromium-gpu-linux-x86_64 to fix assert.
34501
34502         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
34503
34504 2011-04-08  Alpha Lam  <hclam@chromium.org>
34505
34506         Unreviewed, rolling out r83335.
34507         http://trac.webkit.org/changeset/83335
34508         https://bugs.webkit.org/show_bug.cgi?id=53556
34509
34510         GTK and QT bots are broken
34511
34512         * Scripts/build-webkit:
34513
34514 2011-04-07  Anna Cavender  <annacc@chromium.org>
34515
34516         Reviewed by Eric Carlson.
34517
34518         Setup ENABLE(TRACK) feature define
34519         https://bugs.webkit.org/show_bug.cgi?id=53556
34520
34521         * Scripts/build-webkit:
34522
34523 2011-04-08  Dirk Pranke  <dpranke@chromium.org>
34524
34525         Reviewed by Adam Roben.
34526
34527         new-run-webkit-tests: configure a NRWT test bot for Mac SL
34528         https://bugs.webkit.org/show_bug.cgi?id=58114
34529
34530         This change removes the old new-run-webkit-tests entry and
34531         updates it for a new test-only bot.
34532
34533         * BuildSlaveSupport/build.webkit.org-config/config.json:
34534
34535 2011-04-08  Dominic Cooney  <dominicc@google.com>
34536
34537         Reviewed by Adam Roben.
34538
34539         Make layoutTestController.shadowRoot return null, not undefined,
34540         when its argument is invalid.
34541         https://bugs.webkit.org/show_bug.cgi?id=58121
34542
34543         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
34544         (LayoutTestController::shadowRoot):
34545         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
34546         (WTR::LayoutTestController::shadowRoot):
34547
34548 2011-04-08  Pere Martir  <pere.martir4@gmail.com>
34549
34550         Reviewed by Adam Roben.
34551
34552         Locate NSTD.EXE in 64-bit Windows
34553         https://bugs.webkit.org/show_bug.cgi?id=57847
34554
34555         * Scripts/old-run-webkit-tests:
34556
34557 2011-04-08  Adam Roben  <aroben@apple.com>
34558
34559         Qt build fix
34560
34561         * DumpRenderTree/TestNetscapePlugIn/PluginObject.h: Pull in stdarg.h for va_list.
34562
34563 2011-04-07  Adam Roben  <aroben@apple.com>
34564
34565         Test that NPP_SetWindow is passed a null window handle during plugin destruction on non-Mac platforms
34566
34567         Test for <http://webkit.org/b/47009> WebKit2 needs to call NPP_SetWindow when destroying a
34568         plugin
34569
34570         Reviewed by Anders Carlsson.
34571
34572         * DumpRenderTree/TestNetscapePlugIn/PluginObject.cpp:
34573         (pluginLogWithArguments): Moved code to format and log the message here...
34574         (pluginLog): ...from here.
34575
34576         * DumpRenderTree/TestNetscapePlugIn/PluginObject.h: Added pluginLogWithArguments.
34577
34578         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
34579         (PluginTest::log): Added. Calls through to pluginLogWithArguments.
34580
34581         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h: Added log.
34582
34583         * DumpRenderTree/TestNetscapePlugIn/Tests/NPPSetWindowCalledDuringDestruction.cpp: Added.
34584         (NPPSetWindowCalledDuringDestruction::setWillBeDestroyed): Records that destruction is about
34585         to begin.
34586         (NPPSetWindowCalledDuringDestruction::NPPSetWindowCalledDuringDestruction): Simple
34587         constructor.
34588         (NPPSetWindowCalledDuringDestruction::NPP_GetValue): Creates and returns a ScriptObject that
34589         can be used to invoke our setWillBeDestroyed function.
34590         (NPPSetWindowCalledDuringDestruction::NPP_SetWindow): Records what has happened (and logs if
34591         anything unexpected happens).
34592         (NPPSetWindowCalledDuringDestruction::NPP_Destroy): On Mac, logs a failure message if
34593         NPP_SetWindow was called during destruction. On other platforms, logs a failure message if
34594         NPP_SetWindow was *not* called during destruction.
34595         (NPPSetWindowCalledDuringDestruction::ScriptObject::hasMethod): Return true for our only
34596         method, setWillBeDestroyed.
34597         (NPPSetWindowCalledDuringDestruction::ScriptObject::invoke): Call through to the PluginTest
34598         object.
34599
34600         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
34601         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
34602         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
34603         * GNUmakefile.am:
34604         Added new test.
34605
34606         * Scripts/old-run-webkit-tests: Skip the new test when using out-of-process plugins with
34607         WebKit1 on Mac, since it can't work properly due to <http://webkit.org/b/58077>.
34608
34609 2011-04-08  Mario Sanchez Prada  <msanchez@igalia.com>
34610
34611         Reviewed by Martin Robinson.
34612
34613         [GTK] Implement increment() and decrement() functions in DRT's AccessibilityUIElement
34614         https://bugs.webkit.org/show_bug.cgi?id=58039
34615
34616         Implement missing functions in GTK's DRT.
34617
34618         * DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:
34619         (AccessibilityUIElement::increment): Implemented.
34620         (AccessibilityUIElement::decrement): Implemented.
34621
34622 2011-04-08  Dominic Cooney  <dominicc@google.com>
34623
34624         Reviewed by Kent Tamura.
34625
34626         Make WK2 layoutTestController.shadowRoot return undefined, not
34627         null, when its argument is not an element.
34628         https://bugs.webkit.org/show_bug.cgi?id=58121
34629
34630         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
34631         (WTR::LayoutTestController::shadowRoot):
34632
34633 2011-04-07  Dirk Pranke  <dpranke@chromium.org>
34634
34635         Reviewed by Mihai Parparita.
34636
34637         Change reference port for Mac GPU baselines from
34638         Leopard to SnowLeopard, XP to Win7, and Linux-x86
34639         to Linux x86-64.
34640
34641         https://bugs.webkit.org/show_bug.cgi?id=58099
34642
34643         * Scripts/webkitpy/layout_tests/port/chromium.py:
34644
34645 2011-04-07  Andrew Scherkus  <scherkus@chromium.org>
34646
34647         Revert ENABLE_TRACK patch due to compile failures.
34648
34649         * Scripts/build-webkit:
34650
34651 2011-04-07  Tony Chang  <tony@chromium.org>
34652
34653         Reviewed by Eric Seidel.
34654
34655         update DRT to embed checksums in png files
34656         https://bugs.webkit.org/show_bug.cgi?id=57871
34657
34658         We insert the bytes for the comment in printPNG rather than at encode
34659         time because each platform does its own PNG encoding (either using CG
34660         or cairo).  Putting this in pringPNG avoids having to duplicate this
34661         code, although it's not as clean as doing it at encoding time.
34662
34663         We insert the comment right after the IHDR chunk of the PNG.
34664
34665         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
34666         * DumpRenderTree/CyclicRedundancyCheck.cpp: Added.
34667         (makeCrcTable): Ported from LayoutTests/fast/canvas/webgl/resources/pnglib.js
34668         (computeCrc): Ported from LayoutTests/fast/canvas/webgl/resources/pnglib.js
34669         * DumpRenderTree/CyclicRedundancyCheck.h: Added.
34670         * DumpRenderTree/PixelDumpSupport.cpp:
34671         (dumpWebViewAsPixelsAndCompareWithExpected):
34672         (convertChecksumToPNGComment): Generate the bytes to insert.
34673         (printPNG): Insert the png comment before the first IDAT section.
34674         * DumpRenderTree/PixelDumpSupport.h:
34675         * DumpRenderTree/cairo/PixelDumpSupportCairo.cpp:
34676         (printPNG):
34677         (dumpBitmap):
34678         * DumpRenderTree/cg/PixelDumpSupportCG.cpp:
34679         (printPNG):
34680         (dumpBitmap):
34681         * DumpRenderTree/win/DumpRenderTree.vcproj:
34682         * GNUmakefile.am:
34683
34684 2011-04-07  Jeff Miller  <jeffm@apple.com>
34685
34686         Reviewed by Adam Roben.
34687
34688         Replace WKStringGetCharactersPtr() with WKStringGetCharacters()
34689         https://bugs.webkit.org/show_bug.cgi?id=58058
34690
34691         * TestWebKitAPI/Tests/WebKit2/WKString.cpp:
34692         (TestWebKitAPI::TEST): Add tests for  WKStringGetLength() and WKStringGetCharactersPtr().
34693
34694 2011-04-07  Chang Shu  <cshu@webkit.org>
34695
34696         Reviewed by Darin Adler.
34697
34698         WebKitTestRunner needs layoutTestController.isPageBoxVisible
34699         https://bugs.webkit.org/show_bug.cgi?id=42695
34700
34701         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
34702         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
34703         (WTR::LayoutTestController::isPageBoxVisible):
34704         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
34705
34706 2011-04-07  Adam Roben  <aroben@apple.com>
34707
34708         Build fix
34709
34710         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Added a missing import.
34711
34712 2011-04-07  Adam Roben  <aroben@apple.com>
34713
34714         Move ExtractTestResults[AndLeaks]'s summarizing code back from commandComplete() to finished()
34715
34716         commandComplete() is never called for MasterShellCommands like ExtractTestResults[AndLeaks].
34717         (Unfortunately the buildbot documentation does not make this clear.) finished() is the only
34718         hook we have, so we have to do our work there. I added a new addCustomURLs method which can
34719         be overridden by subclasses to provide extra URLs before we call up to the base class (after
34720         which adding more URLs is no longer possible).
34721
34722         Really hopefully fixes <http://webkit.org/b/56032> Leaks viewer should be linked from leaks
34723         bot results page
34724
34725         Reviewed by John Sullivan.
34726
34727         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
34728         (ExtractTestResults.addCustomURLs): Moved code to add the "view results" URL here...
34729         (ExtractTestResults.finished): ...from here. This replaces commandComplete, since that
34730         method is never called for MasterShellCommands.
34731         (ExtractTestResultsAndLeaks.addCustomURLs): Replaced commandComplete (which is never called)
34732         with this method (which is).
34733
34734 2011-04-07  Adam Roben  <aroben@apple.com>
34735
34736         Move ExtractTestResults[AndLeaks]'s summarizing code from finished() to commandComplete()
34737
34738         This matches how most of our other build steps work, so is good just for improving
34739         consistency between build steps. It should also make it possible for
34740         ExtractTestResultsAndLeaks to successfully add a URL to Leaks Viewer. (Previously we were
34741         trying to do this in finished() after we had called up to the base class, but that was
34742         apparently too late to add more URLs.)
34743
34744         Hopefully fixes <http://webkit.org/b/56032> Leaks viewer should be linked from leaks bot
34745         results page
34746
34747         Reviewed by John Sullivan.
34748
34749         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
34750         (ExtractTestResults): Added a descriptionDone property so that we don't have to manually
34751         call setText to get the right text to show up for this step. This matches how most of our
34752         other build steps work.
34753         (ExtractTestResults.commandComplete): Replaced our override of finished with this function.
34754         This is how most of our other build steps work.
34755         (ExtractTestResultsAndLeaks.commandComplete): Replaced our override of finished with this
34756         function, to match the base class.
34757
34758 2011-04-07  Adam Barth  <abarth@webkit.org>
34759
34760         Reviewed by Maciej Stachowiak.
34761
34762         webkit-patch shouldn't state the whole working copy when calling check-webkit-style
34763         https://bugs.webkit.org/show_bug.cgi?id=58022
34764
34765         Previously, we were ignoring the args variable!  This patch fixes the
34766         regression introduced in http://trac.webkit.org/changeset/82771.
34767
34768         * Scripts/webkitpy/tool/commands/download_unittest.py:
34769         * Scripts/webkitpy/tool/mocktool.py:
34770         * Scripts/webkitpy/tool/steps/checkstyle.py:
34771
34772 2011-04-07  Maciej Stachowiak  <mjs@apple.com>
34773
34774         Reviewed by Adam Barth.
34775
34776         REGRESSION: "webkit-patch land" doesn't work correctly in SVN subdirectories
34777         https://bugs.webkit.org/show_bug.cgi?id=58017
34778
34779         * Scripts/webkitpy/common/checkout/scm.py:
34780         * Scripts/webkitpy/common/checkout/scm_unittest.py:
34781
34782 2011-04-06  Chang Shu  <cshu@webkit.org>
34783
34784         Reviewed by Darin Adler.
34785
34786         WebKitTestRunner needs layoutTestController.pageSizeAndMarginsInPixels
34787         https://bugs.webkit.org/show_bug.cgi?id=57984
34788
34789         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
34790         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
34791         (WTR::LayoutTestController::pageSizeAndMarginsInPixels):
34792         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
34793
34794 2011-04-06  Chris Rogers  <crogers@google.com>
34795
34796         Reviewed by Tony Chang.
34797
34798         Add web audio support to DumpRenderTree (mac port)
34799         https://bugs.webkit.org/show_bug.cgi?id=57969
34800
34801         * DumpRenderTree/LayoutTestController.cpp:
34802         (LayoutTestController::LayoutTestController):
34803         (setEncodedAudioDataCallback):
34804         (LayoutTestController::staticFunctions):
34805         * DumpRenderTree/LayoutTestController.h:
34806         (LayoutTestController::dumpAsAudio):
34807         (LayoutTestController::setDumpAsAudio):
34808         (LayoutTestController::encodedAudioData):
34809         (LayoutTestController::setEncodedAudioData):
34810         * DumpRenderTree/mac/DumpRenderTree.mm:
34811         (dumpAudio):
34812         (dump):
34813
34814 2011-04-06  Benjamin Poulain  <benjamin.poulain@nokia.com>
34815
34816         Reviewed by Kenneth Rohde Christiansen.
34817
34818         [Qt] LayoutTestController needs to implement numberOfPendingGeolocationPermissionRequests
34819         https://bugs.webkit.org/show_bug.cgi?id=56086
34820
34821         Add the missing method.
34822
34823         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
34824         (WebCore::DumpRenderTree::getAllPages): Add a method to return all the pages allocated for the current test.
34825         * DumpRenderTree/qt/DumpRenderTreeQt.h:
34826         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
34827         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
34828
34829 2011-04-06  Dai Mikurube  <dmikurube@chromium.org>
34830
34831         Reviewed by David Levin.
34832
34833         Add QUOTA build flag for unified quota API
34834         https://bugs.webkit.org/show_bug.cgi?id=57918
34835
34836         * Scripts/build-webkit: Added QUOTA build flag
34837
34838 2011-04-06  Dirk Pranke  <dpranke@chromium.org>
34839
34840         Reviewed by Tony Chang.
34841
34842         Adjust the apple webkit port's default timeout to match
34843         old-run-webkit-tests at 35 seconds.
34844
34845         https://bugs.webkit.org/show_bug.cgi?id=37738
34846
34847         * Scripts/webkitpy/layout_tests/port/mac.py:
34848
34849 2011-04-06  Scott Cameron  <sccameron@rim.com>
34850
34851         Reviewed by Eric Seidel.
34852
34853         Remove global variable $httpdPath and replace with the return value of
34854         getHTTPDPath().  Also remove unnecessary calls to getHTTPDPath().
34855         https://bugs.webkit.org/show_bug.cgi?id=53499
34856
34857         * Scripts/run-iexploder-tests:
34858         * Scripts/webkitperl/httpd.pm:
34859
34860 2011-04-06  Kevin Ollivier  <kevino@theolliviers.com>
34861
34862         Reviewed by Darin Adler.
34863
34864         Make sure JS_EXPORT_PRIVATE is an empty define when we aren't using the export macros.
34865
34866         https://bugs.webkit.org/show_bug.cgi?id=27551
34867
34868         * DumpRenderTree/config.h:
34869         * WebKitAPITest/config.h:
34870         * WebKitTestRunner/config.h:
34871
34872 2011-04-06  Tony Chang  <tony@chromium.org>
34873
34874         Reviewed by Darin Adler.
34875
34876         teach run-webkit-tests to read checksums from png files
34877         https://bugs.webkit.org/show_bug.cgi?id=57993
34878
34879         * Scripts/old-run-webkit-tests: Read the first 2k of a .png if there's
34880           no .checksum and look for the checksum in there.
34881
34882 2011-04-06  David Dorwin  <ddorwin@chromium.org>
34883
34884         Reviewed by David Levin.
34885
34886         Enable fullscreen layout tests for Chromium
34887         https://bugs.webkit.org/show_bug.cgi?id=55726
34888
34889         Make DumpRenderTree always run with fullscreen enabled (equivalent of --enable-fullscreen).
34890
34891         * DumpRenderTree/chromium/WebPreferences.cpp:
34892         (WebPreferences::applyTo):
34893
34894 2011-04-06  Adam Roben  <aroben@apple.com>
34895
34896         Add a "view leaks" link to builds on SnowLeopard Intel Leaks
34897
34898         Fixes <http://webkit.org/b/56032> Leaks viewer should be linked from leaks bot results page
34899
34900         Reviewed by David Kilzer.
34901
34902         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
34903         (ExtractTestResults.resultDirectoryURL): Added. Moved code to calculate the URL for the
34904         build's results directory here...
34905         (ExtractTestResults.finished): ...from here.
34906         (ExtractTestResultsAndLeaks): New class that's used by the leaks builder
34907         (ExtractTestResultsAndLeaks.finished): Calls up to the base class, but also adds a "view
34908         leaks" link to point to Leaks Viewer for this build.
34909         (TestFactory): Added ExtractTestResultsClass abstraction. This isn't overridden anywhere,
34910         but it seemed good to add for consistency with BuildAndTestFactory.
34911         (BuildAndTestFactory): Added ExtractTestResultsClass abstraction.
34912         (BuildAndTestLeaksFactory): Use ExtractTestResultsAndLeaks as our ExtractTestResultsClass so
34913         that we'll get a "view leaks" link.
34914
34915 2011-04-06  Zan Dobersek  <zandobersek@gmail.com>
34916
34917         Reviewed by Eric Seidel.
34918
34919         [Gtk] plugins/set-status.html fails
34920         https://bugs.webkit.org/show_bug.cgi?id=57844
34921
34922         Allow an empty status text to be dumped.
34923
34924         * DumpRenderTree/gtk/DumpRenderTree.cpp:
34925         (webViewStatusBarTextChanged):
34926
34927 2011-04-06  Chang Shu  <cshu@webkit.org>
34928
34929         Reviewed by Darin Adler.
34930
34931         WebKitTestRunner needs layoutTestController.pageNumberForElementById
34932         https://bugs.webkit.org/show_bug.cgi?id=42329
34933
34934         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
34935         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
34936         (WTR::LayoutTestController::pageNumberForElementById):
34937         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
34938
34939 2011-04-06  Sergio Villar Senin  <svillar@igalia.com>
34940
34941         Reviewed by Martin Robinson.
34942
34943         [GTK] DumpRenderTree: do not try to free NULL SoupURIs
34944         https://bugs.webkit.org/show_bug.cgi?id=57932
34945
34946         Some Layout tests have invalid URIs that do not generate valid
34947         SoupURI instances. Do not try to free those NULL SoupURIs.
34948
34949         * DumpRenderTree/gtk/DumpRenderTree.cpp:
34950         (willSendRequestCallback):
34951
34952 2011-04-06  Csaba Osztrogonác  <ossy@webkit.org>
34953
34954         Reviewed by Andreas Kling.
34955
34956         ORWT shouldn't generate diff files for tests without expected files
34957         https://bugs.webkit.org/show_bug.cgi?id=57846
34958
34959         * Scripts/old-run-webkit-tests:
34960
34961 2011-04-05  Kevin Ollivier  <kevino@theolliviers.com>
34962
34963         [wx] Mac build fix. Handle cases where 10.4 SDK is not installed, and also add x86_64 arch
34964         to deps.
34965
34966         * wx/install-unix-extras:
34967
34968 2011-04-05  Sheriff Bot  <webkit.review.bot@gmail.com>
34969
34970         Unreviewed, rolling out r82978, r82999, and r83001.
34971         http://trac.webkit.org/changeset/82978
34972         http://trac.webkit.org/changeset/82999
34973         http://trac.webkit.org/changeset/83001
34974         https://bugs.webkit.org/show_bug.cgi?id=57913
34975
34976         Does not work in Python 2.5 (Requested by abarth on #webkit).
34977
34978         * Scripts/webkitpy/common/system/executive.py:
34979         * Scripts/webkitpy/common/system/executive_unittest.py:
34980
34981 2011-04-05  Chang Shu  <cshu@webkit.org>
34982
34983         Reviewed by Darin Adler.
34984
34985         WebKitTestRunner needs layoutTestController.numberOfPages
34986         https://bugs.webkit.org/show_bug.cgi?id=42694
34987
34988         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
34989         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
34990         (WTR::LayoutTestController::numberOfPages):
34991         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
34992
34993 2011-04-05  Tony Chang  <tony@chromium.org>
34994
34995         Reviewed by Mihai Parparita.
34996
34997         [chromium] stop putting results downloaded from WebKit Linux in chromium-linux-x86_64
34998         https://bugs.webkit.org/show_bug.cgi?id=57889
34999
35000         * Scripts/webkitpy/layout_tests/port/chromium.py:
35001         * Scripts/webkitpy/layout_tests/port/test.py:
35002         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
35003
35004 2011-04-05  Adam Barth  <abarth@webkit.org>
35005
35006         Silly with statement, from the future!
35007
35008         * Scripts/webkitpy/common/system/executive.py:
35009
35010 2011-04-05  Adam Barth  <abarth@webkit.org>
35011
35012         Reviewed by Tony Chang.
35013
35014         Don't use Exception.message because it's deprecated
35015         https://bugs.webkit.org/show_bug.cgi?id=57892
35016
35017         Suppress the warning for now.  When we move to Python 3, we might need
35018         to something more dramatic.
35019
35020         * Scripts/webkitpy/common/system/executive.py:
35021
35022 2011-04-05  Mihai Parparita  <mihaip@chromium.org>
35023
35024         Reviewed by Ojan Vafai.
35025
35026         Add builders.js to dashboard file list
35027         https://bugs.webkit.org/show_bug.cgi?id=57899
35028
35029         Add file added by http://crrev.com/80538 to dashboard file list.
35030
35031         * TestResultServer/handlers/dashboardhandler.py:
35032
35033 2011-04-05  MORITA Hajime  <morrita@google.com>
35034
35035         Reviewed by Adam Barth.
35036
35037         webkit-patch should print git's stderr when git svn dcommit fail
35038         http://webkit.org/b/57861
35039
35040         * Scripts/webkitpy/common/checkout/scm.py:
35041
35042 2011-04-05  Mihai Parparita  <mihaip@chromium.org>
35043
35044         Reviewed by Tony Chang.
35045
35046         Add --baseline-search-path to NRWT
35047         https://bugs.webkit.org/show_bug.cgi?id=56233
35048
35049         Add NRWT option to specify additional directories to look for baselines
35050         (will be used by hardware GPU bots which will have local per-bot
35051         expectations for some tests)
35052
35053         * Scripts/webkitpy/common/system/filesystem_mock.py:
35054         * Scripts/webkitpy/layout_tests/port/base.py:
35055         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
35056         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
35057
35058 2011-04-05  Adam Roben  <aroben@apple.com>
35059
35060         Strip off /results.html from results URLs before trying to load leaks files from them
35061
35062         r82734 changed build.webkit.org's "view results" URLs to point straight to the results.html
35063         files, rather than pointing to the directory that contains them. This is more convenient for
35064         people browsing build.webkit.org, but confused Leaks Viewer.
35065
35066         Fixes <http://webkit.org/b/57869> REGRESSION (r82734): Links in Leaks Viewer's Recent Builds
35067         list don't work
35068
35069         Reviewed by Joseph Pecoraro.
35070
35071         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
35072         (RecentBuildsLoader.prototype.start): Strip off "/results.html" from the results URL.
35073
35074 2011-04-05  Carol Szabo  <carol@webkit.org>
35075
35076         Unreviewed.
35077
35078         Updated my info in committers.py
35079
35080         Scripts\webkitpy\common\config\committers.py
35081
35082 2011-04-05  Csaba Osztrogonác  <ossy@webkit.org>
35083
35084         Reviewed by Laszlo Gombos.
35085
35086         [Qt] Make WebKitLibraries optional for building QtWebKit
35087         https://bugs.webkit.org/show_bug.cgi?id=57542
35088
35089         * Scripts/build-webkit:
35090
35091 2011-04-05  Chang Shu  <cshu@webkit.org>
35092
35093         Reviewed by Laszlo Gombos.
35094
35095         [Qt] MiniBrowser defaultUrl does not work
35096         https://bugs.webkit.org/show_bug.cgi?id=57021
35097
35098         Match the behavior and coding of MiniBrowser to QtTestBrowser.
35099         * MiniBrowser/qt/main.cpp:
35100         (main):
35101
35102 2011-04-05  Jade Han  <jade.han@nokia.com>
35103
35104         Reviewed by Laszlo Gombos.
35105
35106         [Qt] [Symbian] Disable WebKitTestRunner for Symbian
35107         https://bugs.webkit.org/show_bug.cgi?id=54977
35108
35109         This change is a preparation to enable building webkit2 for Symbian.
35110         Bug 57834 is filed to fix and enable WebKitTestRunner for Symbian.
35111
35112         * Tools.pro:
35113
35114 2011-04-05  Zoltan Horvath  <zoltan@webkit.org>
35115
35116         Reviewed by Andreas Kling.
35117
35118         [Qt] Fix timeoutTimer of MiniBrowser's UrlLoader
35119         https://bugs.webkit.org/show_bug.cgi?id=57832
35120
35121         Only QWKPage has loadFinished signal so connect to it instead of BrowserWindow.
35122
35123         * MiniBrowser/qt/UrlLoader.cpp:
35124         (UrlLoader::UrlLoader):
35125
35126 2011-04-05  Zoltan Horvath  <zoltan@webkit.org>
35127
35128         [Qt] Linux Release minimal build fix after r82919.
35129
35130         * MiniBrowser/qt/BrowserWindow.cpp:
35131         (BrowserWindow::loadURLListFromFile):
35132         * QtTestBrowser/launcherwindow.cpp:
35133         (LauncherWindow::loadURLListFromFile):
35134
35135 2011-04-05  Zoltan Horvath  <zoltan@webkit.org>
35136
35137         Reviewed by Andreas Kling.
35138
35139         [Qt] Make URL loader accessible from testbrowser's menus
35140         https://bugs.webkit.org/show_bug.cgi?id=57823
35141
35142         Add "Load URLs from file" to QtTestBrowser's and to MiniBrowser's menu.
35143
35144         * MiniBrowser/qt/BrowserWindow.cpp:
35145         (BrowserWindow::BrowserWindow):
35146         (BrowserWindow::loadURLListFromFile):
35147         (BrowserWindow::~BrowserWindow):
35148         * MiniBrowser/qt/BrowserWindow.h:
35149         * QtTestBrowser/launcherwindow.cpp:
35150         (LauncherWindow::LauncherWindow):
35151         (LauncherWindow::~LauncherWindow):
35152         (LauncherWindow::createChrome):
35153         (LauncherWindow::loadURLListFromFile):
35154         * QtTestBrowser/launcherwindow.h:
35155
35156 2011-04-04  Sam Weinig  <sam@webkit.org>
35157
35158         Reviewed by Brian Weinstein.
35159
35160         Remove duplicate API from WKContext
35161         <rdar://problem/8727879>
35162         https://bugs.webkit.org/show_bug.cgi?id=57815
35163
35164         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
35165         * TestWebKitAPI/Tests/WebKit2/SendingMessagesToTheWebProcessBeforeItIsValid.cpp: Removed.
35166         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
35167         Remove SendingMessagesToTheWebProcessBeforeItIsValid since the API it was testing is now removed.
35168
35169 2011-04-04  Tony Chang  <tony@chromium.org>
35170
35171         Reviewed by Ojan Vafai.
35172
35173         [chromium] don't write .checksum files if a fallback platform has an embedded checksum
35174         https://bugs.webkit.org/show_bug.cgi?id=57783
35175
35176         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
35177         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
35178
35179 2011-04-04  Kevin Ollivier  <kevino@theolliviers.com>
35180
35181         [wx] Unreviewed build fix, run uninstall when cleaning to remove built files
35182         from WebKitBuild.
35183
35184         * Scripts/webkitdirs.pm:
35185
35186 2011-04-04  Martin Robinson  <mrobinson@igalia.com>
35187
35188         Reviewed by Gustavo Noronha Silva.
35189
35190         [GTK] WebGL support
35191         https://bugs.webkit.org/show_bug.cgi?id=31517
35192
35193         Add support to the DRT for turning on WebGL when a layout tests requests it.
35194
35195         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
35196         (LayoutTestController::overridePreference): Allow turning on WebGL from tests.
35197
35198 2011-04-04  Chang Shu  <cshu@webkit.org>
35199
35200         Reviewed by Darin Adler.
35201
35202         WebKitTestRunner needs layoutTestController.setAllowFileAccessFromFileURLs
35203         https://bugs.webkit.org/show_bug.cgi?id=57572
35204
35205         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
35206         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
35207         (WTR::LayoutTestController::setAllowFileAccessFromFileURLs):
35208         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
35209
35210 2011-04-04  Keith Kyzivat  <keith.kyzivat@nokia.com>
35211
35212         Reviewed by Csaba Osztrogonác.
35213
35214         [Qt] DumpRenderTree breaks compilation in some uClibc environments
35215         https://bugs.webkit.org/show_bug.cgi?id=57602
35216
35217         * DumpRenderTree/qt/main.cpp:
35218         (get_backtrace):
35219
35220 2011-04-04  Kevin Ollivier  <kevino@theolliviers.com>
35221
35222         [wx] Unreviewed build fix, add new LayoutTestController method stub to wx.
35223
35224         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
35225         (LayoutTestController::shadowRoot):
35226
35227 2011-04-04  Pavel Podivilov  <podivilov@chromium.org>
35228
35229         Unreviewed, fix exception in rebaseline tool.
35230
35231         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
35232
35233 2011-04-02  Beth Dakin  <bdakin@apple.com>
35234
35235         Rubber-stamped by Geoff Garen.
35236
35237         For Dan!
35238         * DumpRenderTree/mac/DumpRenderTree.mm:
35239         (resetWebViewToConsistentStateBeforeTesting):
35240
35241 2011-04-02  Beth Dakin  <bdakin@apple.com>
35242
35243         Rubber-stamped by Geoff Garen.
35244
35245         Need to reset the scale, much like zoom.
35246         * DumpRenderTree/mac/DumpRenderTree.mm:
35247         (resetWebViewToConsistentStateBeforeTesting):
35248
35249 2011-04-02  Sam Weinig  <sam@webkit.org>
35250
35251         Reviewed by Beth Dakin.
35252
35253         https://bugs.webkit.org/show_bug.cgi?id=57605
35254         Frame::pageScaleFactor() should not affect getBoundingClientRect() or
35255         getClientRects()
35256         -and corresponding-
35257         <rdar://problem/9194541>
35258
35259         Add DRT support for the scaleWebView SPI.
35260         * DumpRenderTree/mac/EventSendingController.mm:
35261         (+[EventSendingController isSelectorExcludedFromWebScript:]):
35262         (+[EventSendingController webScriptNameForSelector:]):
35263         (-[EventSendingController scalePageBy:atX:andY:]):
35264         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
35265         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
35266         (WTR::EventSendingController::scalePageBy):
35267         * WebKitTestRunner/InjectedBundle/EventSendingController.h:
35268         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
35269         (WTR::InjectedBundlePage::reset):
35270
35271 2011-04-02  Dominic Cooney  <dominicc@google.com>
35272
35273         Reviewed by Martin Robinson.
35274
35275         Add layoutTestController.shadowRoot to GTK DumpRenderTree.
35276         https://bugs.webkit.org/show_bug.cgi?id=57551
35277
35278         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
35279         (LayoutTestController::shadowRoot):
35280
35281 2011-04-02  Patrick Gansterer  <paroga@webkit.org>
35282
35283         Reviewed by Eric Seidel.
35284
35285         Remove AbstractStep._run_script and move script names to ports.py
35286         https://bugs.webkit.org/show_bug.cgi?id=57704
35287
35288         Replace deprecated _run_script with _tool.executive.run_and_throw_if_fail.
35289
35290         * Scripts/webkitpy/common/config/ports.py:
35291         * Scripts/webkitpy/tool/commands/download_unittest.py:
35292         * Scripts/webkitpy/tool/commands/roll_unittest.py:
35293         * Scripts/webkitpy/tool/commands/upload_unittest.py:
35294         * Scripts/webkitpy/tool/steps/abstractstep.py:
35295         * Scripts/webkitpy/tool/steps/checkstyle.py:
35296         * Scripts/webkitpy/tool/steps/preparechangelog.py:
35297         * Scripts/webkitpy/tool/steps/preparechangelogfordepsroll.py:
35298         * Scripts/webkitpy/tool/steps/preparechangelogforrevert.py:
35299
35300 2011-04-02  Patrick Gansterer  <paroga@webkit.org>
35301
35302         Reviewed by Eric Seidel.
35303
35304         Emulate shebang on Win32
35305         https://bugs.webkit.org/show_bug.cgi?id=55927
35306
35307         Scripts on Windows work only if they are called with the explicit interpreter.
35308         Read the first line of scripts to detect the correct executable.
35309
35310         * Scripts/webkitpy/common/config/ports.py:
35311         * Scripts/webkitpy/common/system/executive.py: Added interpreter_for_script().
35312         * Scripts/webkitpy/common/system/executive_unittest.py:
35313
35314 2011-04-01  Adam Barth  <abarth@webkit.org>
35315
35316         Reviewed by Eric Seidel.
35317
35318         apos entities shouldn't show up in ChangeLogs when using webkit-patch
35319         https://bugs.webkit.org/show_bug.cgi?id=57692
35320
35321         Previously, we were using BeautifulSoup to process XML from
35322         bugs.webkit.org, but that's incorrect.  We should be using
35323         BeautifulStoneSoup to process the XML.  We were getting the &apos;
35324         entity wrong because &apos; is an XML entity but not an HTML entity.
35325
35326         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
35327         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
35328
35329 2011-04-01  Sheriff Bot  <webkit.review.bot@gmail.com>
35330
35331         Unreviewed, rolling out r82721.
35332         http://trac.webkit.org/changeset/82721
35333         https://bugs.webkit.org/show_bug.cgi?id=57687
35334
35335         This patch introduced assertion failures on the GTK+ bots.
35336         (Requested by mrobinson on #webkit).
35337
35338         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
35339         (LayoutTestController::shadowRoot):
35340
35341 2011-04-01  Csaba Osztrogonác  <ossy@webkit.org>
35342
35343         Unreviewed buildfix.
35344
35345         [Qt][WK2] Build Webkit2 using "-2" option on Qt
35346         https://bugs.webkit.org/show_bug.cgi?id=55074
35347
35348         * Scripts/build-webkit: Ensure that "-2" isn't passed to qmake.
35349         (The isWK2() function removes it from @ARGV, but not from @options.)
35350
35351 2011-04-01  Keith Kyzivat  <keith.kyzivat@nokia.com>
35352
35353         Reviewed by Csaba Osztrogonác.
35354
35355         [Qt] [WK2] MiniBrowser.qrc not found - regression from rev 82671
35356         https://bugs.webkit.org/show_bug.cgi?id=57666
35357
35358         * MiniBrowser/qt/MiniBrowser.qrc: Renamed from Tools/MiniBrowser/MiniBrowser.qrc.
35359
35360 2011-04-01  Dirk Pranke  <dpranke@chromium.org>
35361
35362         Reviewed by Adam Barth.
35363
35364         new-run-webkit-tests: fix feature detection, skipped platform lists on mac
35365
35366         We apparently never implemented the code to skip tests based on
35367         what was compiled into DRT. Also, change the logic used to skip
35368         platform directories to match what old-run-webkit-tests does:
35369         skip every test not in a directory in the baseline search path.
35370
35371         https://bugs.webkit.org/show_bug.cgi?id=57662
35372
35373         * Scripts/webkitpy/layout_tests/port/mac.py:
35374         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
35375         * Scripts/webkitpy/layout_tests/port/webkit.py:
35376
35377 2011-03-29  Mark Rowe  <mrowe@apple.com>
35378
35379         Reviewed by Jon Honeycutt.
35380
35381         <http://webkit.org/b/56730> new-run-webkit-tests fails on Lion seed
35382
35383         Teach new-run-webkit-tests about the concept of an unreleased version of Mac OS X.
35384
35385         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
35386         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
35387         * Scripts/webkitpy/layout_tests/port/mac.py:
35388         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
35389
35390 2011-04-01  Chang Shu  <cshu@webkit.org>
35391
35392         Reviewed by Csaba Osztrogonác.
35393
35394         [Qt][WK2] Build Webkit2 using "-2" option on Qt
35395         https://bugs.webkit.org/show_bug.cgi?id=55074
35396
35397         * Scripts/build-webkit:
35398
35399 2011-04-01  Dirk Pranke  <dpranke@chromium.org>
35400
35401         Reviewed by Tony Chang.
35402
35403         new-run-webkit-tests: remove spurious port version override in webkit.py
35404         base.py provides a default implementation so this is just
35405         breaking things.
35406
35407         https://bugs.webkit.org/show_bug.cgi?id=57667
35408
35409         * Scripts/webkitpy/layout_tests/port/webkit.py:
35410         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
35411
35412 2011-04-01  Csaba Osztrogonác  <ossy@webkit.org>
35413
35414         Reviewed by Adam Roben.
35415
35416         Make view results on waterfall direct link to results.html
35417         https://bugs.webkit.org/show_bug.cgi?id=57671
35418
35419         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Add "/results.html" to the URL.
35420         * Scripts/old-run-webkit-tests: Add links to httpd access and error logs.
35421
35422 2011-04-01  Sam Weinig  <sam@webkit.org>
35423
35424         Fix windows build.
35425
35426         * TestWebKitAPI/Tests/WebKit2/win/HideFindIndicator.cpp:
35427         (TestWebKitAPI::TEST):
35428         * TestWebKitAPI/Tests/WebKit2/win/ResizeViewWhileHidden.cpp:
35429         (TestWebKitAPI::flushMessages):
35430
35431 2011-04-01  Sam Weinig  <sam@webkit.org>
35432
35433         Reviewed by Adam Roben.
35434
35435         Add adoptWK to WKRetainPtr.h
35436         https://bugs.webkit.org/show_bug.cgi?id=57670
35437
35438         * TestWebKitAPI/PlatformUtilities.h:
35439         * TestWebKitAPI/Tests/WebKit2/AboutBlankLoad.cpp:
35440         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest.cpp:
35441         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest_Bundle.cpp:
35442         * TestWebKitAPI/Tests/WebKit2/CookieManager.cpp:
35443         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp:
35444         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
35445         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp:
35446         * TestWebKitAPI/Tests/WebKit2/RestoreSessionStateContainingFormData.cpp:
35447         (TestWebKitAPI::createSessionStateContainingFormData):
35448         * WebKitTestRunner/StringFunctions.h:
35449         Replace custom versions of adoptWK with the API on in WebKit2/WKRetainPtr.h.
35450
35451 2011-04-01  Sam Weinig  <sam@webkit.org>
35452
35453         Reviewed by Adam Roben.
35454
35455         Fix leak noticed by Adam Roben in LayoutTestController::shadowRoot.
35456
35457         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
35458         (WTR::LayoutTestController::shadowRoot):
35459         Make judicious use of adoptWK().
35460
35461 2011-04-01  Sam Weinig  <sam@webkit.org>
35462
35463         Reviewed by Timothy Hatcher.
35464
35465         Fix extract-localizable-strings for macro change from UI_STRING -> WEB_UI_STRING.
35466
35467         * Scripts/extract-localizable-strings:
35468
35469 2011-04-01  Anders Carlsson  <andersca@apple.com>
35470
35471         Reviewed by Sam Weinig.
35472
35473         Shockwave plug-in doesn't accept mouse events
35474         https://bugs.webkit.org/show_bug.cgi?id=57653
35475         <rdar://problem/8483273>
35476
35477         Add a plug-in test.
35478
35479         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
35480         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
35481         (PluginTest::indicateTestFailure):
35482         Move code from NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject here.
35483
35484         (PluginTest::NPN_ConvertPoint):
35485         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
35486         * DumpRenderTree/TestNetscapePlugIn/Tests/NPDeallocateCalledBeforeNPShutdown.cpp:
35487         (NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject):
35488         Call indicateTestFailure.
35489
35490         * DumpRenderTree/TestNetscapePlugIn/Tests/mac: Added.
35491         * DumpRenderTree/TestNetscapePlugIn/Tests/mac/ConvertPoint.cpp: Added.
35492         (ConvertPoint::ConvertPoint):
35493         (ConvertPoint::testConvert):
35494         (ConvertPoint::NPP_New):
35495
35496 2011-04-01  Dominic Cooney  <dominicc@google.com>
35497
35498         Reviewed by Martin Robinson.
35499
35500         Add layoutTestController.shadowRoot to GTK DumpRenderTree.
35501         https://bugs.webkit.org/show_bug.cgi?id=57551
35502
35503         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
35504         (LayoutTestController::shadowRoot):
35505
35506 2011-04-01  Sam Weinig  <sam@webkit.org>
35507
35508         Reviewed by Anders Carlsson.
35509
35510         WebKitTestRunner needs layoutTestController.shadowRoot
35511         https://bugs.webkit.org/show_bug.cgi?id=57661
35512
35513         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
35514         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
35515         (WTR::LayoutTestController::shadowRoot):
35516         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
35517         Implement layoutTestController.shadowRoot for WebKit2.
35518
35519 2011-04-01  Chang Shu  <cshu@webkit.org>
35520
35521         Reviewed by Darin Adler.
35522
35523         WebKitTestRunner needs layoutTestController.setDatabaseQuota
35524         https://bugs.webkit.org/show_bug.cgi?id=57568
35525
35526         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
35527         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
35528         (WTR::LayoutTestController::setDatabaseQuota):
35529         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
35530
35531 2011-04-01  Dirk Pranke  <dpranke@chromium.org>
35532
35533         Reviewed by Tony Chang.
35534
35535         This patch completes the cleanup of
35536         rebaseline-chromium-webkit-tests to work with all of the
35537         variants of a platform (we can now rebaseline gpu- and non-gpu
35538         files at the same time).
35539
35540         When the rebaselining is complete, any lines declared as
35541         REBASELINE in the expectations file that matches a test that was
35542         actually rebaselined will be deleted, even if only one of the
35543         variants was actually rebaselined. This may cause odd problems,
35544         but is better than where we're at today.
35545
35546         This change removes the -g flag and deprecates -w. The -g flag is gone
35547         because GPU baselines are handled just like any other variant.
35548         The -w flag is deprecated because this tool now only works
35549         against the canaries, since that's the only place we have a full
35550         set of bots. It will be trivial to change this to
35551         build.webkit.org if we decide that's where we want them to be.
35552
35553         Also, this patch deletes a lot of cruft that is no longer needed
35554         in the test_expectations code and the port-specific code.
35555
35556         https://bugs.webkit.org/show_bug.cgi?id=55191
35557
35558         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
35559         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
35560         * Scripts/webkitpy/layout_tests/port/base.py:
35561         * Scripts/webkitpy/layout_tests/port/base_unittest.py:
35562         * Scripts/webkitpy/layout_tests/port/chromium.py:
35563         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
35564         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
35565         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
35566         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
35567         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
35568         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
35569         * Scripts/webkitpy/layout_tests/port/mac.py:
35570         * Scripts/webkitpy/layout_tests/port/test.py:
35571         * Scripts/webkitpy/layout_tests/port/webkit.py:
35572         * Scripts/webkitpy/layout_tests/port/win.py:
35573         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
35574         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
35575
35576 2011-04-01  Dirk Pranke  <dpranke@chromium.org>
35577
35578         Reviewed by Adam Barth.
35579
35580         rebaseline-chromium-webkit-tests does not work correctly with
35581         version-specific baselines. This patch updates the tool to use
35582         all of the version-specific bots on the canaries, and will now
35583         attempt to rebaseline all of the versions by default, although
35584         it will not update both GPU and CPU versions.
35585
35586         Also, it will no longer modify the test_expectations.txt file
35587         *at all*. You will have to manually delete the REBASELINE lines
35588         after running the tool and determining that it did what you
35589         wanted it to do. This should be fixed in a separate bug - see
35590         webkit bug #55191.
35591
35592         https://bugs.webkit.org/show_bug.cgi?id=55608
35593
35594         * Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
35595         * Scripts/webkitpy/layout_tests/layout_package/test_expectations_unittest.py:
35596         * Scripts/webkitpy/layout_tests/port/chromium.py:
35597         * Scripts/webkitpy/layout_tests/port/test.py:
35598         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
35599         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
35600
35601 2011-03-31  Tony Chang  <tony@chromium.org>
35602
35603         Reviewed by Ojan Vafai.
35604
35605         [chromium] update the rebaseline tool to know about pngs with checksums
35606         https://bugs.webkit.org/show_bug.cgi?id=57481
35607
35608         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py: Skip over .checksum files if the checksum is already in the png
35609         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests_unittest.py:
35610
35611 2011-04-01  Adam Roben  <aroben@apple.com>
35612
35613         Retrieve revision numbers from the build's got_revision property in Leaks Viewer
35614
35615         Previously, we were getting the revision of the first revision that triggered a build.
35616         Choosing the last revision would have been more accurate. But got_revision is what is used
35617         everywhere else on build.webkit.org, and should work even when there were no changes that
35618         triggered a build (e.g., if someone clicked the Force Build button).
35619
35620         Fixes <http://webkit.org/b/57630> Leaks viewer gets some revision numbers wrong in the
35621         recent builds list
35622
35623         Reviewed by Anders Carlsson.
35624
35625         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
35626         (RecentBuildsLoader.prototype.start): Pull the revision number out of the got_revision
35627         property, rather than out of the first (i.e., earliest) change in the sourceStamp object.
35628
35629         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js:
35630         (Array.prototype.first): Added this helper function to return the first element in an array
35631         that matches the given predicate, or null if no such element exists.
35632
35633 2011-03-31  Adam Roben  <aroben@apple.com>
35634
35635         Prefer (but don't require) bug URLs to be on their own line when parsing bug numbers from ChangeLogs
35636
35637         Fixes <http://webkit.org/b/57579> webkit-patch is too strict about bug URL formatting
35638
35639         Reviewed by Darin Adler.
35640
35641         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
35642         (parse_bug_id_from_changelog): Fall back to parse_bug_id if we weren't able to find a bug
35643         URL on its own line.
35644
35645         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
35646         (BugzillaTest.test_parse_bug_id_from_changelog): Updated expected results for test
35647         progression, and added a new test that uses a short bug URL while I was at it.
35648
35649 2011-04-01  Adam Roben  <aroben@apple.com>
35650
35651         Mark .vcproj/.vsprops/.sln files as being Windows-only
35652
35653         Fixes <http://webkit.org/b/57489> Mac builders built 82512, but shouldn't have
35654
35655         Reviewed by Anders Carlsson.
35656
35657         * Scripts/webkitpy/common/config/build.py:
35658         (_should_file_trigger_build): Added patterns to mark .vcproj/.vsprops/.sln files and .vcproj
35659         directories as Windows-only.
35660
35661         * Scripts/webkitpy/common/config/build_unittest.py:
35662         (ShoulBuildTest): Added test cases for the above.
35663
35664 2011-04-01  Keith Kyzivat  <keith.kyzivat@nokia.com>
35665
35666         Reviewed by Laszlo Gombos.
35667
35668         [Qt] Build MiniBrowser for Symbian
35669         https://bugs.webkit.org/show_bug.cgi?id=56319
35670
35671         Have MiniBrowser reference it's own copy of useragentlist.txt instead
35672         of copying QtTestBrowser's.
35673         Remove Tools/MiniBrowser/DerivedSources.pro
35674         This reduces complexity in the Tools scripts due to Symbian limitations.
35675
35676         * DerivedSources.pro:
35677         * MiniBrowser/DerivedSources.pro: Removed.
35678         * MiniBrowser/MiniBrowser.qrc:
35679         * MiniBrowser/qt/MiniBrowser.pro:
35680         * Scripts/webkitdirs.pm:
35681
35682 2011-03-31  Chang Shu  <cshu@webkit.org>
35683
35684         Reviewed by Darin Adler.
35685
35686         WebKitTestRunner needs layoutTestController.clearAllDatabases
35687         https://bugs.webkit.org/show_bug.cgi?id=42540
35688
35689         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
35690         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
35691         (WTR::InjectedBundle::beginTesting):
35692         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
35693         (WTR::LayoutTestController::clearAllDatabases):
35694         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
35695
35696 2011-03-31  Chang Shu  <cshu@webkit.org>
35697
35698         Reviewed by Darin Adler.
35699
35700         WebKitTestRunner needs layoutTestController.setAllowUniversalAccessFromFileURLs
35701         https://bugs.webkit.org/show_bug.cgi?id=42692
35702
35703         * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
35704
35705 2011-03-31  Darin Adler  <darin@apple.com>
35706
35707         Reviewed by Sam Weinig.
35708
35709         Implement mouseDown, mouseUp, and mouseMoveTo in WebKitTestRunner
35710         https://bugs.webkit.org/show_bug.cgi?id=57573
35711
35712         * WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
35713         Added real definitions for mouseDown, mouseUp, mouseMoveTo and leapForward.
35714         Removed fake definitions of keyDown and contextClick.
35715
35716         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
35717         (WTR::operator==): Added. So we can compare two WKPoint structs.
35718         (WTR::parseModifier): Added.
35719         (WTR::parseModifierArray): Added.
35720         (WTR::EventSendingController::EventSendingController): Initialize the
35721         new data members.
35722         (WTR::EventSendingController::mouseDown): Added. Calls
35723         WKBundlePageSimulateMouseDown.
35724         (WTR::EventSendingController::mouseUp): Added. Calls
35725         WKBundlePageSimulateMouseUp.
35726         (WTR::EventSendingController::mouseMoveTo): Added. Calls
35727         WKBundlePageSimulateMouseMotion.
35728         (WTR::EventSendingController::leapForward): Added.
35729         (WTR::EventSendingController::updateClickCount): Added. Used by the
35730         mouseDown/Up functions to create a click count.
35731
35732         * WebKitTestRunner/InjectedBundle/EventSendingController.h: Updated
35733         for the changes above.
35734
35735 2011-03-31  Sergio Villar Senin  <svillar@igalia.com>
35736
35737         Reviewed by Martin Robinson.
35738
35739         [GTK] implement LayoutTestController::setWillSendRequestReturnsNull
35740         https://bugs.webkit.org/show_bug.cgi?id=57362
35741
35742         Do not generate DRT output if willSendRequestReturnsNull is set.
35743
35744         * DumpRenderTree/gtk/DumpRenderTree.cpp:
35745         (willSendRequestCallback):
35746
35747 2011-03-31  Vamshikrishna.Yellenki  <vamshi@motorola.com> and Alejandro G. Castro  <alex@igalia.com>
35748
35749         Reviewed by Martin Robinson.
35750
35751         Implement MiniBrowser for Gtk port.
35752         https://bugs.webkit.org/show_bug.cgi?id=48512
35753
35754         Initial implementation of the Gtk MiniBrowser.
35755
35756         * MiniBrowser/gtk/GNUmakefile.am: Added.
35757         * MiniBrowser/gtk/main.c: Added.
35758         (activateUriEntryCallback):
35759         (destroyCallback):
35760         (goBackCallback):
35761         (goForwardCallback):
35762         (createToolbar):
35763         (createWebView):
35764         (createWindow):
35765         (argumentToURL):
35766         (main):
35767
35768 2011-03-30  Dominic Cooney  <dominicc@google.com>
35769
35770         Reviewed by Dimitri Glazkov.
35771
35772         Adds layoutTestController.shadowRoot accessor to Mac DRT.
35773         https://bugs.webkit.org/show_bug.cgi?id=57415
35774
35775         * DumpRenderTree/LayoutTestController.cpp:
35776         (shadowRootCallback):
35777         (LayoutTestController::staticFunctions):
35778         * DumpRenderTree/LayoutTestController.h:
35779         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
35780         (LayoutTestController::shadowRoot):
35781         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
35782         (LayoutTestController::shadowRoot):
35783
35784 2011-03-30  Matthew Delaney  <mdelaney@apple.com>
35785
35786         Reviewed by Chris Marrin.
35787
35788         Update fast/canvas tests to avoid dumping the render tree when possible
35789         https://bugs.webkit.org/show_bug.cgi?id=57493
35790
35791         * DumpRenderTree/mac/DumpRenderTree.mm: Make DRT aware of new default values for
35792           accelerated drawing and accelerated drawing for canvas
35793
35794 2011-03-30  Adam Roben  <aroben@apple.com>
35795
35796         Stop ignoring leaks in CGGradientCreateWithColorStops
35797
35798         Fixes <rdar://problem/7888547>.
35799
35800         Rubber-stamped by John Sullivan.
35801
35802         * Scripts/old-run-webkit-tests:
35803         (countAndPrintLeaks): Removed some code to ignore those leaks.
35804
35805 2011-03-30  Timur Iskhodzhanov  <timurrrr@google.com>
35806
35807         Reviewed by Alexey Proskuryakov.
35808
35809         Add some dynamic annotations to JavaScriptCore/wtf
35810         https://bugs.webkit.org/show_bug.cgi?id=53747
35811
35812         By using these annotations we can improve the precision of finding
35813         WebKit errors using dynamic analysis tools like ThreadSanitizer and Valgrind.
35814         These annotations don't affect the compiled binaries unless USE(DYNAMIC_ANNOTATIONS) is "1".
35815
35816         These files don't add new functionality, so don't need extra tests.
35817
35818         * DumpRenderTree/ForwardingHeaders/wtf/DynamicAnnotations.h: Added.
35819
35820 2011-03-30  Steve Falkenburg  <sfalken@apple.com>
35821
35822         Reviewed by Adam Roben.
35823
35824         Share most vsprops between Release and Production builds in releaseproduction.vsprops
35825         https://bugs.webkit.org/show_bug.cgi?id=57508
35826
35827         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginProduction.vsprops:
35828         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginRelease.vsprops:
35829         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseCairoCFLite.vsprops:
35830         * DumpRenderTree/win/DumpRenderTreeProduction.vsprops:
35831         * DumpRenderTree/win/DumpRenderTreeRelease.vsprops:
35832         * DumpRenderTree/win/DumpRenderTreeReleaseCairoCFLite.vsprops:
35833         * DumpRenderTree/win/ImageDiffProduction.vsprops:
35834         * DumpRenderTree/win/ImageDiffRelease.vsprops:
35835         * DumpRenderTree/win/ImageDiffReleaseCairoCFLite.vsprops:
35836         * FindSafari/FindSafariProduction.vsprops:
35837         * FindSafari/FindSafariRelease.vsprops:
35838         * FindSafari/FindSafariReleaseCairoCFLite.vsprops:
35839         * FindSafari/FindSafariReleasePGO.vsprops:
35840         * MiniBrowser/Configurations/MiniBrowserProduction.vsprops:
35841         * MiniBrowser/Configurations/MiniBrowserRelease.vsprops:
35842         * MiniBrowser/Configurations/MiniBrowserReleaseCairoCFLite.vsprops:
35843         * TestWebKitAPI/Configurations/TestWebKitAPIRelease.vsprops:
35844         * TestWebKitAPI/Configurations/TestWebKitAPIReleaseCairoCFLite.vsprops:
35845         * WebKitAPITest/WebKitAPITestProduction.vsprops:
35846         * WebKitAPITest/WebKitAPITestRelease.vsprops:
35847         * WebKitAPITest/WebKitAPITestReleaseCairoCFLite.vsprops:
35848         * WebKitLauncherWin/WebKitLauncherWinProduction.vsprops:
35849         * WebKitLauncherWin/WebKitLauncherWinRelease.vsprops:
35850         * WebKitLauncherWin/WebKitLauncherWinReleaseCairoCFLite.vsprops:
35851         * WebKitTestRunner/win/InjectedBundleProduction.vsprops:
35852         * WebKitTestRunner/win/InjectedBundleRelease.vsprops:
35853         * WebKitTestRunner/win/InjectedBundleReleaseCairoCFLite.vsprops:
35854         * WebKitTestRunner/win/WebKitTestRunnerProduction.vsprops:
35855         * WebKitTestRunner/win/WebKitTestRunnerRelease.vsprops:
35856         * WebKitTestRunner/win/WebKitTestRunnerReleaseCairoCFLite.vsprops:
35857         * WinLauncher/WinLauncherProduction.vsprops:
35858         * WinLauncher/WinLauncherRelease.vsprops:
35859         * WinLauncher/WinLauncherReleaseCairoCFLite.vsprops:
35860         * record-memory-win/record-memory-winProduction.vsprops:
35861         * record-memory-win/record-memory-winRelease.vsprops:
35862         * record-memory-win/record-memory-winReleaseCairoCFLite.vsprops:
35863
35864 2011-03-30  Steve Falkenburg  <sfalken@apple.com>
35865
35866         Reviewed by Adam Roben.
35867
35868         Update Windows production build logic for new production configurations
35869         https://bugs.webkit.org/show_bug.cgi?id=57494
35870
35871         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginProduction.vsprops:
35872         * DumpRenderTree/win/DumpRenderTreeProduction.vsprops:
35873         * DumpRenderTree/win/ImageDiffProduction.vsprops:
35874         * FindSafari/FindSafariProduction.vsprops:
35875         * FindSafari/FindSafariReleasePGO.vsprops:
35876         * MiniBrowser/Configurations/MiniBrowserProduction.vsprops:
35877         * WebKitAPITest/WebKitAPITestProduction.vsprops:
35878         * WebKitLauncherWin/WebKitLauncherWinProduction.vsprops:
35879         * WebKitTestRunner/win/InjectedBundleProduction.vsprops:
35880         * WebKitTestRunner/win/WebKitTestRunnerProduction.vsprops:
35881         * WinLauncher/WinLauncherProduction.vsprops:
35882         * record-memory-win/record-memory-winProduction.vsprops:
35883
35884 2011-03-30  Robert Hogan  <robert@webkit.org>
35885
35886         Reviewed by Antonio Gomes.
35887
35888         [Qt] Fix LoadHTMLStringItem::invoke() after r75966
35889         Unskip http/tests/navigation/go-back-to-error-page.html
35890
35891         Also add the location of DumpRenderTreeSupportQt.h
35892         to DRT's include paths.
35893
35894         https://bugs.webkit.org/show_bug.cgi?id=52614
35895
35896         * DumpRenderTree/qt/DumpRenderTree.pro:
35897         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
35898         * DumpRenderTree/qt/DumpRenderTreeQt.h:
35899         * DumpRenderTree/qt/GCControllerQt.cpp:
35900         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
35901         (LayoutTestController::queueLoadHTMLString):
35902         * DumpRenderTree/qt/LayoutTestControllerQt.h:
35903         * DumpRenderTree/qt/PlainTextControllerQt.cpp:
35904         * DumpRenderTree/qt/TextInputControllerQt.cpp:
35905         * DumpRenderTree/qt/WorkQueueItemQt.cpp:
35906         (LoadAlternateHTMLStringItem::invoke):
35907         * DumpRenderTree/qt/WorkQueueItemQt.h:
35908         (LoadAlternateHTMLStringItem::LoadAlternateHTMLStringItem):
35909         * QtTestBrowser/QtTestBrowser.pro:
35910         * QtTestBrowser/launcherwindow.h:
35911
35912 2011-03-30  Adam Barth  <abarth@webkit.org>
35913
35914         Reviewed by Adam Roben.
35915
35916         SheriffBot rollouts take too long
35917         https://bugs.webkit.org/show_bug.cgi?id=57498
35918
35919         We used to build before landing rollouts via the commit-queue to
35920         prevent further breakage, but now that our individual commit-queue
35921         machines are slower, building takes too long.  I can't remember the
35922         last time a rollout broke compile.  It seems like just landing the
35923         patch is the better trade-off.
35924
35925         * Scripts/webkitpy/tool/bot/commitqueuetask.py:
35926         * Scripts/webkitpy/tool/commands/queues_unittest.py:
35927
35928 2011-03-30  Adam Barth  <abarth@webkit.org>
35929
35930         Remove platform/chromium-mac-snowleopard
35931         https://bugs.webkit.org/show_bug.cgi?id=57486
35932
35933         This directory no longer exists, so we can remove it from the fallback
35934         chain.
35935
35936         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
35937
35938 2011-03-30  Steve Falkenburg  <sfalken@apple.com>
35939
35940         Reviewed by Adam Roben.
35941
35942         Rename Windows configuration Release_LTCG to Production for clarity
35943         https://bugs.webkit.org/show_bug.cgi?id=57465
35944
35945         * DumpRenderTree/DumpRenderTree.sln:
35946         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
35947         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginProduction.vsprops: Copied from Tools/DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseLTCG.vsprops.
35948         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseLTCG.vsprops: Removed.
35949         * DumpRenderTree/win/DumpRenderTree.vcproj:
35950         * DumpRenderTree/win/DumpRenderTreeProduction.vsprops: Copied from Tools/DumpRenderTree/win/DumpRenderTreeReleaseLTCG.vsprops.
35951         * DumpRenderTree/win/DumpRenderTreeReleaseLTCG.vsprops: Removed.
35952         * DumpRenderTree/win/ImageDiff.vcproj:
35953         * DumpRenderTree/win/ImageDiffProduction.vsprops: Copied from Tools/DumpRenderTree/win/ImageDiffReleaseLTCG.vsprops.
35954         * DumpRenderTree/win/ImageDiffReleaseLTCG.vsprops: Removed.
35955         * FindSafari/FindSafari.vcproj:
35956         * FindSafari/FindSafariProduction.vsprops: Copied from Tools/FindSafari/FindSafariReleaseLTCG.vsprops.
35957         * FindSafari/FindSafariReleaseLTCG.vsprops: Removed.
35958         * MiniBrowser/Configurations/MiniBrowserProduction.vsprops: Copied from Tools/MiniBrowser/Configurations/MiniBrowserReleaseLTCG.vsprops.
35959         * MiniBrowser/Configurations/MiniBrowserReleaseLTCG.vsprops: Removed.
35960         * MiniBrowser/MiniBrowser.vcproj:
35961         * Scripts/webkitdirs.pm:
35962         * TestWebKitAPI/win/TestWebKitAPI.sln:
35963         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
35964         * TestWebKitAPI/win/TestWebKitAPIGenerated.vcproj:
35965         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
35966         * WebKitAPITest/WebKitAPITest.vcproj:
35967         * WebKitAPITest/WebKitAPITestProduction.vsprops: Copied from Tools/WebKitAPITest/WebKitAPITestReleaseLTCG.vsprops.
35968         * WebKitAPITest/WebKitAPITestReleaseLTCG.vsprops: Removed.
35969         * WebKitLauncherWin/WebKitLauncherWin.vcproj:
35970         * WebKitLauncherWin/WebKitLauncherWinProduction.vsprops: Copied from Tools/WebKitLauncherWin/WebKitLauncherWinReleaseLTCG.vsprops.
35971         * WebKitLauncherWin/WebKitLauncherWinReleaseLTCG.vsprops: Removed.
35972         * WebKitTestRunner/WebKitTestRunner.sln:
35973         * WebKitTestRunner/win/InjectedBundle.vcproj:
35974         * WebKitTestRunner/win/InjectedBundleGenerated.vcproj:
35975         * WebKitTestRunner/win/InjectedBundleProduction.vsprops: Copied from Tools/WebKitTestRunner/win/InjectedBundleReleaseLTCG.vsprops.
35976         * WebKitTestRunner/win/InjectedBundleReleaseLTCG.vsprops: Removed.
35977         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
35978         * WebKitTestRunner/win/WebKitTestRunnerProduction.vsprops: Copied from Tools/WebKitTestRunner/win/WebKitTestRunnerReleaseLTCG.vsprops.
35979         * WebKitTestRunner/win/WebKitTestRunnerReleaseLTCG.vsprops: Removed.
35980         * WinLauncher/WinLauncher.vcproj:
35981         * WinLauncher/WinLauncherProduction.vsprops: Copied from Tools/WinLauncher/WinLauncherReleaseLTCG.vsprops.
35982         * WinLauncher/WinLauncherReleaseLTCG.vsprops: Removed.
35983         * record-memory-win/record-memory-win.vcproj:
35984         * record-memory-win/record-memory-winProduction.vsprops: Copied from Tools/record-memory-win/record-memory-winReleaseLTCG.vsprops.
35985         * record-memory-win/record-memory-winReleaseLTCG.vsprops: Removed.
35986
35987 2011-03-30  MORITA Hajime  <morrita@google.com>
35988
35989         Reviewed by Dimitri Glazkov.
35990
35991         [Chromium] Expose the shadow DOM to DumpRenderTree JS tests.
35992         https://bugs.webkit.org/show_bug.cgi?id=56573
35993
35994         Added LayoutTestController.shadowRoot() to Chromium DRT.
35995
35996         * DumpRenderTree/chromium/LayoutTestController.cpp:
35997         (LayoutTestController::LayoutTestController):
35998         (LayoutTestController::shadowRoot):
35999         * DumpRenderTree/chromium/LayoutTestController.h:
36000
36001 2011-03-30  Yuta Kitamura  <yutak@chromium.org>
36002
36003         Reviewed by Kent Tamura.
36004
36005         [Chromium] DumpRenderTree: Implement LayoutTestController::setPluginsEnabled
36006         https://bugs.webkit.org/show_bug.cgi?id=57430
36007
36008         * DumpRenderTree/chromium/LayoutTestController.cpp:
36009         (LayoutTestController::LayoutTestController):
36010         (LayoutTestController::setPluginsEnabled):
36011         * DumpRenderTree/chromium/LayoutTestController.h:
36012
36013 2011-03-30  Maciej Stachowiak  <mjs@apple.com>
36014
36015         Reviewed by Adam Barth.
36016
36017         make webkit-patch upload respect -d
36018         https://bugs.webkit.org/show_bug.cgi?id=57425
36019
36020         * Scripts/webkitpy/common/checkout/scm.py:
36021         * Scripts/webkitpy/tool/steps/commit.py:
36022
36023 2011-03-30  Maciej Stachowiak  <mjs@apple.com>
36024
36025         Reviewed by Adam Barth.
36026
36027         Make "webkit-patch --dry-run --verbose land" log the SVN command it's going to use
36028         https://bugs.webkit.org/show_bug.cgi?id=57429
36029
36030         * Scripts/webkitpy/common/checkout/scm.py:
36031
36032 2011-03-29  Sheriff Bot  <webkit.review.bot@gmail.com>
36033
36034         Unreviewed, rolling out r82383.
36035         http://trac.webkit.org/changeset/82383
36036         https://bugs.webkit.org/show_bug.cgi?id=57417
36037
36038         "nrwt isn't shutting down cleanly" (Requested by dpranke on
36039         #webkit).
36040
36041         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
36042         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
36043         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
36044
36045 2011-03-29  Kent Tamura  <tkent@chromium.org>
36046
36047         Reviewed by Dimitri Glazkov.
36048
36049         Make validation message bubble testable
36050         https://bugs.webkit.org/show_bug.cgi?id=57290
36051
36052         * DumpRenderTree/mac/DumpRenderTree.mm:
36053         (createWebViewAndOffscreenWindow): Change a setting so that validation
36054           bubbles isn't hidden automatically.
36055
36056 2011-03-29  Dirk Pranke  <dpranke@chromium.org>
36057
36058         Reviewed by Ojan Vafai.
36059
36060         new-run-webkit-tests: use 'threads' on win instead of 'old-threads'
36061
36062         It's a bit unclear whether multiple threads or multiple
36063         processes will be a better model on windows. This change will
36064         test threads for now while we're still working out the bugs in
36065         multiple-processes.
36066
36067         https://bugs.webkit.org/show_bug.cgi?id=57410
36068
36069         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
36070         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
36071         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
36072
36073 2011-03-29  David Levin  <levin@chromium.org>
36074
36075         Reviewed by Shinichiro Hamaji.
36076
36077         check-webkit-style confused by two ChangeLog entries in a row from same user
36078         https://bugs.webkit.org/show_bug.cgi?id=57250
36079
36080         * Scripts/webkitpy/style/checker.py: Add the line should be checked function to ChangeLogChecker.
36081         * Scripts/webkitpy/style/checker_unittest.py: Fix test due to that new function.
36082         * Scripts/webkitpy/style/checkers/changelog.py: Made this code aware of what lines were being checked.
36083           It basically assumes only one ChangeLog entry is being processed because that is the standard case and
36084           checking more than that would be very messey.
36085         * Scripts/webkitpy/style/checkers/changelog_unittest.py: Add testing to catch the broken case.
36086         * Scripts/webkitpy/style/error_handlers.py: Added should_line_be_checked.
36087
36088 2011-03-29  Kent Tamura  <tkent@chromium.org>
36089
36090         Reviewed by Dimitri Glazkov.
36091
36092         [Mac] Enable interactive-validation tests on Mac DRT
36093         https://bugs.webkit.org/show_bug.cgi?id=57308
36094
36095         * DumpRenderTree/mac/DumpRenderTree.mm:
36096         (createWebViewAndOffscreenWindow): Enable the form interactive validation feature.
36097
36098 2011-03-29  Chang Shu  <cshu@webkit.org>
36099
36100         Reviewed by Kenneth Rohde Christiansen.
36101
36102         WebKitTestRunner needs layoutTestController.setAllowUniversalAccessFromFileURLs
36103         https://bugs.webkit.org/show_bug.cgi?id=42692
36104
36105         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
36106         (WTR::InjectedBundle::beginTesting):
36107         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
36108         (WTR::LayoutTestController::setAllowUniversalAccessFromFileURLs):
36109         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
36110
36111 2011-03-24  Martin Robinson  <mrobinson@igalia.com>
36112
36113         Reviewed by Xan Lopez.
36114
36115         [GTK] [DRT] GtkScrolledWindow adds three pixels of padding between the WebView and the scrollbar
36116         https://bugs.webkit.org/show_bug.cgi?id=57067
36117
36118         Eliminate extra spacing between the DRT GtkScrolledWindow and the DRT WebView. This
36119         will allow WebKit1GTK+ and WebKit2GTK+ to share test results. Followup commits will
36120         update DRT and pixel results as hundreds of results will need to be updated.
36121
36122         * DumpRenderTree/gtk/DumpRenderTree.cpp:
36123         (setDefaultsToConsistentStateValuesForTesting): Eliminate spacing on the GtkScrolledWindow
36124         by overriding the theme RC/CSS files.
36125
36126 2011-03-29  Steve Falkenburg  <sfalken@apple.com>
36127
36128         Reviewed by Adam Roben.
36129
36130         Use per-configuration vsprops in tools projects to avoid WebKitVSPropsRedirectionDir removal by MSVC IDE
36131         https://bugs.webkit.org/show_bug.cgi?id=57394
36132
36133         Visual Studio's IDE was removing instances of $(WebKitVSPropsRedirectionDir) from
36134         InheritedPropertySheet rules in our vcproj files when the vcproj was edited from within
36135         the IDE. To avoid this, add a separate vsprops file for each project configuration that
36136         contains the required inherited property sheets.
36137
36138         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
36139         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginDebug.vsprops: Added.
36140         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginDebugAll.vsprops: Added.
36141         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginDebugCairoCFLite.vsprops: Added.
36142         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginRelease.vsprops: Added.
36143         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseCairoCFLite.vsprops: Added.
36144         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePluginReleaseLTCG.vsprops: Added.
36145         * DumpRenderTree/win/DumpRenderTree.vcproj:
36146         * DumpRenderTree/win/DumpRenderTreeDebug.vsprops: Added.
36147         * DumpRenderTree/win/DumpRenderTreeDebugAll.vsprops: Added.
36148         * DumpRenderTree/win/DumpRenderTreeDebugCairoCFLite.vsprops: Added.
36149         * DumpRenderTree/win/DumpRenderTreeRelease.vsprops: Added.
36150         * DumpRenderTree/win/DumpRenderTreeReleaseCairoCFLite.vsprops: Added.
36151         * DumpRenderTree/win/DumpRenderTreeReleaseLTCG.vsprops: Added.
36152         * DumpRenderTree/win/ImageDiff.vcproj:
36153         * DumpRenderTree/win/ImageDiffDebug.vsprops: Added.
36154         * DumpRenderTree/win/ImageDiffDebugAll.vsprops: Added.
36155         * DumpRenderTree/win/ImageDiffDebugCairoCFLite.vsprops: Added.
36156         * DumpRenderTree/win/ImageDiffRelease.vsprops: Added.
36157         * DumpRenderTree/win/ImageDiffReleaseCairoCFLite.vsprops: Added.
36158         * DumpRenderTree/win/ImageDiffReleaseLTCG.vsprops: Added.
36159         * FindSafari/FindSafari.vcproj:
36160         * FindSafari/FindSafariDebug.vsprops: Added.
36161         * FindSafari/FindSafariDebugAll.vsprops: Added.
36162         * FindSafari/FindSafariDebugCairoCFLite.vsprops: Added.
36163         * FindSafari/FindSafariRelease.vsprops: Added.
36164         * FindSafari/FindSafariReleaseCairoCFLite.vsprops: Added.
36165         * FindSafari/FindSafariReleaseLTCG.vsprops: Added.
36166         * FindSafari/FindSafariReleasePGO.vsprops: Added.
36167         * MiniBrowser/Configurations/MiniBrowserDebug.vsprops: Added.
36168         * MiniBrowser/Configurations/MiniBrowserDebugAll.vsprops: Added.
36169         * MiniBrowser/Configurations/MiniBrowserDebugCairoCFLite.vsprops: Added.
36170         * MiniBrowser/Configurations/MiniBrowserRelease.vsprops: Added.
36171         * MiniBrowser/Configurations/MiniBrowserReleaseCairoCFLite.vsprops: Added.
36172         * MiniBrowser/Configurations/MiniBrowserReleaseLTCG.vsprops: Added.
36173         * MiniBrowser/MiniBrowser.vcproj:
36174         * WebKitAPITest/WebKitAPITest.vcproj:
36175         * WebKitAPITest/WebKitAPITestDebug.vsprops: Added.
36176         * WebKitAPITest/WebKitAPITestDebugAll.vsprops: Added.
36177         * WebKitAPITest/WebKitAPITestDebugCairoCFLite.vsprops: Added.
36178         * WebKitAPITest/WebKitAPITestRelease.vsprops: Added.
36179         * WebKitAPITest/WebKitAPITestReleaseCairoCFLite.vsprops: Added.
36180         * WebKitAPITest/WebKitAPITestReleaseLTCG.vsprops: Added.
36181         * WebKitLauncherWin/WebKitLauncherWin.vcproj:
36182         * WebKitLauncherWin/WebKitLauncherWinDebug.vsprops: Added.
36183         * WebKitLauncherWin/WebKitLauncherWinDebugAll.vsprops: Added.
36184         * WebKitLauncherWin/WebKitLauncherWinDebugCairoCFLite.vsprops: Added.
36185         * WebKitLauncherWin/WebKitLauncherWinRelease.vsprops: Added.
36186         * WebKitLauncherWin/WebKitLauncherWinReleaseCairoCFLite.vsprops: Added.
36187         * WebKitLauncherWin/WebKitLauncherWinReleaseLTCG.vsprops: Added.
36188         * WebKitTestRunner/win/InjectedBundle.vcproj:
36189         * WebKitTestRunner/win/InjectedBundleDebug.vsprops: Added.
36190         * WebKitTestRunner/win/InjectedBundleDebugAll.vsprops: Added.
36191         * WebKitTestRunner/win/InjectedBundleDebugCairoCFLite.vsprops: Added.
36192         * WebKitTestRunner/win/InjectedBundleRelease.vsprops: Added.
36193         * WebKitTestRunner/win/InjectedBundleReleaseCairoCFLite.vsprops: Added.
36194         * WebKitTestRunner/win/InjectedBundleReleaseLTCG.vsprops: Added.
36195         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
36196         * WebKitTestRunner/win/WebKitTestRunnerDebug.vsprops: Added.
36197         * WebKitTestRunner/win/WebKitTestRunnerDebugAll.vsprops: Added.
36198         * WebKitTestRunner/win/WebKitTestRunnerDebugCairoCFLite.vsprops: Added.
36199         * WebKitTestRunner/win/WebKitTestRunnerRelease.vsprops: Added.
36200         * WebKitTestRunner/win/WebKitTestRunnerReleaseCairoCFLite.vsprops: Added.
36201         * WebKitTestRunner/win/WebKitTestRunnerReleaseLTCG.vsprops: Added.
36202         * WinLauncher/WinLauncher.vcproj:
36203         * WinLauncher/WinLauncherDebug.vsprops: Added.
36204         * WinLauncher/WinLauncherDebugAll.vsprops: Added.
36205         * WinLauncher/WinLauncherDebugCairoCFLite.vsprops: Added.
36206         * WinLauncher/WinLauncherRelease.vsprops: Added.
36207         * WinLauncher/WinLauncherReleaseCairoCFLite.vsprops: Added.
36208         * WinLauncher/WinLauncherReleaseLTCG.vsprops: Added.
36209         * record-memory-win/record-memory-win.vcproj:
36210         * record-memory-win/record-memory-winDebug.vsprops: Added.
36211         * record-memory-win/record-memory-winDebugAll.vsprops: Added.
36212         * record-memory-win/record-memory-winDebugCairoCFLite.vsprops: Added.
36213         * record-memory-win/record-memory-winRelease.vsprops: Added.
36214         * record-memory-win/record-memory-winReleaseCairoCFLite.vsprops: Added.
36215         * record-memory-win/record-memory-winReleaseLTCG.vsprops: Added.
36216
36217 2011-03-29  Adam Barth  <abarth@webkit.org>
36218
36219         Reviewed by Darin Adler.
36220
36221         test_runner2.py can crash due to undefined variable
36222         https://bugs.webkit.org/show_bug.cgi?id=57356
36223
36224         Replace reference to undefined variable with a variable that's actually
36225         defined.  AFAICT, there's no way to test this code.
36226
36227         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
36228
36229 2011-03-29  Tony Chang  <tony@chromium.org>
36230
36231         Reviewed by Eric Seidel.
36232
36233         cleanup rebaseline-chromium-webkit-tests
36234         https://bugs.webkit.org/show_bug.cgi?id=57375
36235
36236         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
36237             - Split _extract_and_add_new_baselines into 3 functions
36238             - Reduce the amount of line wrapping (the file already has lines
36239               over 80 col, so may as well try to make the file consistent)
36240             - Remove unnecessary ()s
36241
36242 2011-03-29  Brent Fulgham  <bfulgham@webkit.org>
36243
36244         Reviewed by Adam Roben.
36245
36246         Make WinCairo a core builder.
36247         https://bugs.webkit.org/show_bug.cgi?id=57373
36248
36249         * Scripts/webkitpy/common/net/buildbot/buildbot.py: Add WinCairo.
36250         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py: Add WinCairo.
36251
36252 2011-03-29  Timothy Hatcher  <timothy@apple.com>
36253
36254         Update update-webkit-localizable-strings to understand that WebKit and
36255         WebKit2 strings go in WebCore.
36256
36257         Also make extract-localizable-strings optionally use an exclude file. No individual
36258         warnings are produced about unlocalized strings when there is no exclude file. This
36259         is needed for WebCore, since there are too many bare strings.
36260
36261         https://webkit.org/b/57354
36262
36263         Reviewed by Sam Weinig.
36264
36265         * Scripts/extract-localizable-strings: Support "-" to mean no exclude file.
36266         * Scripts/update-webkit-localizable-strings: Update paths to scan WebCore and WebKit2.
36267
36268 2011-03-29  Darin Adler  <darin@apple.com>
36269
36270         Reviewed by Adam Roben.
36271
36272         WebKit2 bundle page needs to offer generated file hooks
36273         https://bugs.webkit.org/show_bug.cgi?id=57279
36274
36275         (Land missing part of the original patch after r82289.)
36276
36277         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
36278         (WTR::InjectedBundlePage::InjectedBundlePage): Added new callbacks.
36279
36280 2011-03-29  Tony Chang  <tony@chromium.org>
36281
36282         Reviewed by Ojan Vafai.
36283
36284         [chromium] NRWT should be able to read checksums from png comments
36285         https://bugs.webkit.org/show_bug.cgi?id=57280
36286
36287         * Scripts/read-checksum-from-png: Added. Utility script to read the checksum
36288             from a file.
36289         * Scripts/webkitpy/common/system/filesystem.py: Add open_binary_file_for_reading
36290         * Scripts/webkitpy/common/system/filesystem_mock.py:
36291         * Scripts/webkitpy/layout_tests/port/base.py: When loading a checksum,
36292             if the -checksum.txt file doesn't exist, try looking in the png.
36293         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
36294         * Scripts/webkitpy/layout_tests/port/test.py:
36295         * Scripts/webkitpy/layout_tests/read_checksum_from_png.py: Added. Scan the first
36296             2k for a png comment with the checksum
36297         * Scripts/webkitpy/layout_tests/read_checksum_from_png_unittest.py: Added.
36298
36299 2011-03-29  Philippe Normand  <pnormand@igalia.com>
36300
36301         Rubber-stamped by Gustavo Noronha Silva.
36302
36303         [GTK] http/tests/uri/username-with-no-hostname.html fails
36304         https://bugs.webkit.org/show_bug.cgi?id=57325
36305
36306         * DumpRenderTree/gtk/DumpRenderTree.cpp:
36307         (willSendRequestCallback): URI host value check done case-insensitively.
36308
36309 2011-03-29  Philippe Normand  <pnormand@igalia.com>
36310
36311         Reviewed by Martin Robinson.
36312
36313         [GTK] http/tests/uri/username-with-no-hostname.html fails
36314         https://bugs.webkit.org/show_bug.cgi?id=57325
36315
36316         * DumpRenderTree/gtk/DumpRenderTree.cpp:
36317         (willSendRequestCallback): Check URI before submitting the request
36318         to WebCore.
36319
36320 2011-03-29  Noel Gordon  <noel.gordon@gmail.com>
36321
36322         Reviewed by Ojan Vafai.
36323
36324         [chromium] DRT EventSender: remove identity from dragTargetDragEnter() calls
36325         https://bugs.webkit.org/show_bug.cgi?id=57303
36326
36327         * DumpRenderTree/chromium/EventSender.cpp:
36328         (EventSender::doDragDrop):
36329         (EventSender::beginDragWithFiles):
36330
36331 2011-03-29  Zoltan Horvath  <zoltan@webkit.org>
36332
36333         Reviewed by Andreas Kling.
36334
36335         [Qt] Add -print-loaded-urls option to Qt's MiniBrowser and to QtTestBrowser
36336         https://bugs.webkit.org/show_bug.cgi?id=57314
36337
36338         Provide possibility to print loaded urls to the standard output.
36339
36340         * MiniBrowser/qt/BrowserWindow.cpp:
36341         (BrowserWindow::BrowserWindow):
36342         (BrowserWindow::printURL):
36343         * MiniBrowser/qt/BrowserWindow.h:
36344         * MiniBrowser/qt/MiniBrowserApplication.cpp:
36345         (MiniBrowserApplication::handleUserOptions):
36346         * MiniBrowser/qt/MiniBrowserApplication.h:
36347         (WindowOptions::WindowOptions):
36348         * QtTestBrowser/launcherwindow.cpp:
36349         (LauncherWindow::initializeView):
36350         (LauncherWindow::printURL):
36351         (LauncherWindow::cloneWindow):
36352         * QtTestBrowser/launcherwindow.h:
36353         (WindowOptions::WindowOptions):
36354         * QtTestBrowser/main.cpp:
36355         (LauncherApplication::handleUserOptions):
36356
36357 2011-03-29  Gabor Loki  <loki@webkit.org>
36358
36359         Rubber-stamped by Csaba Osztrogonác.
36360
36361         [Qt] Teach build-jsc how to build JavaScriptCore on Qt
36362         https://bugs.webkit.org/show_bug.cgi?id=56918
36363
36364         * Scripts/build-jsc: Remove duplicated --qt options to avoid passing them to qmake.
36365
36366 2011-03-28  Gabor Loki  <loki@webkit.org>
36367
36368         Reviewed by Csaba Osztrogonác.
36369
36370         [Qt] Teach build-jsc how to build JavaScriptCore on Qt
36371         https://bugs.webkit.org/show_bug.cgi?id=56918
36372
36373         * Scripts/build-jsc:
36374         * Scripts/webkitdirs.pm:
36375
36376 2011-03-28  Andrew Foster  <andrewf@chromium.org>
36377
36378         Reviewed by Ojan Vafai.
36379
36380         webkit-patch should be more intelligent about whether a bug applies to a patch
36381         Create a new function, parse_bug_id_from_changelog() which determines
36382         that a bug is related to a patch by parsing the output generated from
36383         prepare-ChangeLog, rather than arbitrarily matching a URL to a bug in
36384         the description of a patch.
36385         https://bugs.webkit.org/show_bug.cgi?id=56989
36386
36387         * Scripts/webkitpy/common/checkout/api.py:
36388         * Scripts/webkitpy/common/checkout/changelog.py:
36389         * Scripts/webkitpy/common/net/bugzilla/__init__.py:
36390         * Scripts/webkitpy/common/net/bugzilla/bugzilla.py:
36391         * Scripts/webkitpy/common/net/bugzilla/bugzilla_unittest.py:
36392         * Scripts/webkitpy/style/checkers/changelog.py:
36393         * Scripts/webkitpy/tool/commands/upload.py:
36394
36395 2011-03-28  Maciej Stachowiak  <mjs@apple.com>
36396
36397         Revert accidentally commited change.
36398
36399         * DumpRenderTree/mac/ResourceLoadDelegate.mm:
36400         (-[ResourceLoadDelegate webView:resource:willSendRequest:redirectResponse:fromDataSource:]):
36401
36402 2011-03-28  Tony Chang  <tony@chromium.org>
36403
36404         Reviewed by Adam Barth.
36405
36406         [chromium] have DRT write the png checksum into a png comment
36407         https://bugs.webkit.org/show_bug.cgi?id=57255
36408
36409         * DumpRenderTree/chromium/TestShell.cpp:
36410         (TestShell::dumpImage):
36411
36412 2011-03-28  Adele Peterson  <adele@apple.com>
36413
36414         Removing Qt stub, since it doesn't use the same header as the others.
36415
36416         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
36417
36418 2011-03-28  Adele Peterson  <adele@apple.com>
36419
36420         Added a few more stubs.
36421
36422         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
36423         (LayoutTestController::hasGrammarMarker):
36424         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
36425         (LayoutTestController::hasGrammarMarker):
36426         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
36427         (LayoutTestController::hasGrammarMarker):
36428
36429 2011-03-28  Adele Peterson  <adele@apple.com>
36430
36431         Build fix.
36432
36433         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
36434         (LayoutTestController::hasGrammarMarker):
36435
36436 2011-03-28  Adele Peterson  <adele@apple.com>
36437
36438         Reviewed by Eric Seidel.
36439
36440         Testing support for <rdar://problem/9112694> REGRESSION (r79411): "Check grammar with spelling" context menu doesn't check as you type
36441         https://bugs.webkit.org/show_bug.cgi?id=57173
36442
36443         * DumpRenderTree/LayoutTestController.cpp:
36444         (hasGrammarMarkerCallback): Added.
36445         (LayoutTestController::staticFunctions): Added case for hasGrammarMarker.
36446         * DumpRenderTree/LayoutTestController.h:
36447         * DumpRenderTree/mac/DumpRenderTree.mm: (createWebViewAndOffscreenWindow): Call setGrammarCheckingEnabled.
36448         * DumpRenderTree/mac/LayoutTestControllerMac.mm: (LayoutTestController::hasGrammarMarker): Added. Call new hasGrammarMarker method.
36449
36450 2011-03-28  Csaba Osztrogonác  <ossy@webkit.org>
36451
36452         Reviewed by Benjamin Poulain.
36453
36454         [Qt] QtWebKit will not compile with QT_ASCII_CAST_WARNINGS enabled
36455         https://bugs.webkit.org/show_bug.cgi?id=57087
36456
36457         Trivial typo fix after r82082.
36458
36459         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
36460         (WTR::activateFonts):
36461
36462 2011-03-28  Alexis Menard  <alexis.menard@openbossa.org>
36463
36464         Reviewed by Andreas Kling.
36465
36466         Add myself to the committers list.
36467
36468         * Scripts/webkitpy/common/config/committers.py:
36469
36470 2011-03-28  Simon Fraser  <simon.fraser@apple.com>
36471
36472         Reviewed by Sam Weinig.
36473
36474         WebKit2 unfortunately uses code from WebKit, so MiniBrowser needs
36475         to link against WebKit.framework for now.
36476
36477         * MiniBrowser/MiniBrowser.xcodeproj/project.pbxproj:
36478
36479 2011-03-28  Benjamin Poulain  <benjamin.poulain@nokia.com>
36480
36481         Reviewed by Andreas Kling.
36482
36483         [Qt] QtWebKit will not compile with QT_ASCII_CAST_WARNINGS enabled
36484         https://bugs.webkit.org/show_bug.cgi?id=57087
36485
36486         Disable QT_ASCII_CAST_WARNINGS for applications.
36487
36488         * DumpRenderTree/qt/DumpRenderTree.pro:
36489         * DumpRenderTree/qt/ImageDiff.pro:
36490         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
36491         * MiniBrowser/qt/MiniBrowser.pro:
36492         * QtTestBrowser/QtTestBrowser.pro:
36493         * WebKitTestRunner/qt/WebKitTestRunner.pro:
36494
36495 2011-03-28  Andreas Kling  <kling@webkit.org>
36496
36497         Reviewed by Kenneth Rohde Christiansen.
36498
36499         [Qt] QtTestBrowser doesn't reset its window title when opening about:blank
36500         https://bugs.webkit.org/show_bug.cgi?id=57224
36501
36502         * QtTestBrowser/mainwindow.cpp:
36503         (MainWindow::buildUI):
36504         (MainWindow::onTitleChanged):
36505         * QtTestBrowser/mainwindow.h:
36506
36507 2011-03-27  Jer Noble  <jer.noble@apple.com>
36508
36509         Reviewed by Maciej Stachowiak.
36510
36511         Full Screen: disable keyboard access by default
36512         https://bugs.webkit.org/show_bug.cgi?id=56684
36513
36514         Accept the withKeyboard parameter to supportsFullScreenForElement.
36515
36516         * DumpRenderTree/mac/UIDelegate.mm:
36517         (-[UIDelegate webView:supportsFullScreenForElement:withKeyboard:]):
36518
36519 2011-03-27  Patrick Gansterer  <paroga@webkit.org>
36520
36521         Reviewed by David Levin.
36522
36523         check-webkit-style should check ChangeLog for a valid bug number
36524         https://bugs.webkit.org/show_bug.cgi?id=57184
36525
36526         * Scripts/webkitpy/style/checker.py:
36527         * Scripts/webkitpy/style/checker_unittest.py:
36528         * Scripts/webkitpy/style/checkers/changelog.py: Added.
36529         * Scripts/webkitpy/style/checkers/changelog_unittest.py: Added.
36530
36531 2011-03-27  Benjamin Poulain  <benjamin.poulain@nokia.com>
36532
36533         Reviewed by Andreas Kling.
36534
36535         [Qt] QtWebKit will not compile with QT_ASCII_CAST_WARNINGS enabled
36536         https://bugs.webkit.org/show_bug.cgi?id=57087
36537
36538         Use explicit conversion for string to avoid depending on the default codec
36539         installed by the user code.
36540
36541         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
36542         (WTR::activateFonts):
36543
36544 2011-03-26  Maciej Stachowiak  <mjs@apple.com>
36545
36546         Revert inadvertantly committed changes.
36547
36548         * Scripts/old-run-webkit-tests:
36549         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
36550
36551 2011-03-26  Andreas Kling  <kling@webkit.org>
36552
36553         Reviewed by Kenneth Rohde Christiansen.
36554
36555         [Qt] Show page icons (favicons) in QtTestBrowser location bar.
36556         https://bugs.webkit.org/show_bug.cgi?id=57162
36557
36558         * QtTestBrowser/QtTestBrowser.qrc:
36559         * QtTestBrowser/favicon.png: Added.
36560         * QtTestBrowser/locationedit.cpp:
36561         (defaultPageIcon):
36562         (LocationEdit::LocationEdit):
36563         (LocationEdit::setPageIcon):
36564         (LocationEdit::resizeEvent):
36565         (LocationEdit::updateInternalGeometry):
36566         * QtTestBrowser/locationedit.h:
36567         * QtTestBrowser/mainwindow.cpp:
36568         (MainWindow::buildUI):
36569         (MainWindow::onIconChanged):
36570         (MainWindow::onLoadStarted):
36571         * QtTestBrowser/mainwindow.h:
36572
36573 2011-03-23  Martin Robinson  <mrobinson@igalia.com>
36574
36575         Reviewed by Xan Lopez.
36576
36577         [GTK] Force DumpRenderTree to use 96 DPI
36578         https://bugs.webkit.org/show_bug.cgi?id=56866
36579
36580         Make sure that GTK+ is set to use 96 DPI when running tests. This ensures
36581         broader compatibility with WebKit2 and makes setting font sizes more straightforward.
36582
36583         * DumpRenderTree/gtk/DumpRenderTree.cpp:
36584         (initializeGtkFontSettings): Set the screen resolution and GTK+ xft property to 96 DPI.
36585         (resetDefaultsToConsistentValues):Remove the call which repeatedly set the GDK screen resolution.
36586         (setDefaultsToConsistentStateValuesForTesting): Update the font sizes to reflect the new DPI.
36587
36588 2011-03-25  Andy Estes  <aestes@apple.com>
36589
36590         Reviewed by Adele Peterson.
36591
36592         REGRESSION (r70748): latest nightly builds kills AC_QuickTime.js
36593         https://bugs.webkit.org/show_bug.cgi?id=49016
36594
36595         TestNetscapePlugIn needs to register support for an image MIME type so
36596         we can test that the embed tag prefers plug-ins to render image MIME
36597         types.
36598
36599         * DumpRenderTree/TestNetscapePlugIn/mac/Info.plist: Register image/png.
36600         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
36601         (NP_GetMIMEDescription): Ditto.
36602         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.rc: Ditto.
36603
36604 2011-03-25  Jessie Berlin  <jberlin@apple.com>
36605
36606         Reviewed by Sam Weinig.
36607
36608         WebKit2: Need to be able to set and get the Cookie Storage Policy.
36609         https://bugs.webkit.org/show_bug.cgi?id=50780
36610
36611         Add a test for getting and setting the HTTP Cookie Accept Policy in WebKit2.
36612
36613         The test is named "CookieManager" so that we can eventually exercise more of the
36614         CookieManager functionality in the test.
36615
36616         * TestWebKitAPI/Tests/WebKit2/CookieManager.cpp: Added.
36617         (TestWebKitAPI::didGetTestHTTPCookieAcceptPolicy):
36618         Assert that the policy returned is equal to the policy set in
36619         didGetUserHTTPCookieAcceptPolicy, and then restore the user's policy.
36620         (TestWebKitAPI::didGetUserHTTPCookieAcceptPolicy):
36621         Set the policy to something different than the user's policy and get the policy again.
36622         (TestWebKitAPI::didFinishLoadForFrame):
36623         Get the user's policy so that it can be restored at the end of the test.
36624         (TestWebKitAPI::TEST):
36625         Load about:blank so that the Web Process gets instatiated (needed to get the
36626         CookieManager).
36627
36628         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
36629         Add CookieManager.cpp.
36630
36631 2011-03-24  Kevin Ollivier  <kevino@theolliviers.com>
36632
36633         [wx] Fix the build so that gcc-4.1 and up can be used to build on SnowLeopard.
36634
36635         * wx/build/settings.py:
36636
36637 2011-03-24  Ilya Sherman  <isherman@chromium.org>
36638
36639         Reviewed by Adam Roben.
36640
36641         Implement layoutTestController.setAutoFilled in DRT on Windows
36642         https://bugs.webkit.org/show_bug.cgi?id=56828
36643
36644         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
36645         (LayoutTestController::setAutofilled): Implemented.
36646
36647 2011-03-24  Brent Fulgham  <bfulgham@webkit.org>
36648
36649         Unreviewed build correction.
36650
36651         Correct vsprops file for Debug variant of the CFLite build.
36652         Disable a CFNETWORK-specific test when building for CFLite.
36653
36654         * TestWebKitAPI/Configurations/TestWebKitAPIDebugCairoCFLite.vsprops:
36655         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
36656
36657 2011-03-24  Dirk Pranke  <dpranke@chromium.org>
36658
36659         Reviewed by Adam Barth.
36660
36661         fix regressions introduced in 81908 - there were a couple of
36662         code paths that only fired under python 2.5 that I missed.
36663
36664         Also change a couple of 'python' references to sys.executable
36665         in order to handle runing test-webkitpy with a binary other than
36666         something called 'python'.
36667
36668         https://bugs.webkit.org/show_bug.cgi?id=57063
36669
36670         * Scripts/webkitpy/layout_tests/port/mac.py:
36671         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
36672         * Scripts/webkitpy/test/cat.py:
36673         * Scripts/webkitpy/test/echo.py:
36674
36675 2011-03-24  Brent Fulgham  <bfulgham@webkit.org>
36676
36677         Unreviewed build correction.
36678
36679         Add a stub implementation of the TestInvocation dumping logic.
36680         Update project files to reflect new file.
36681
36682         * WebKitTestRunner/cairo: Added.
36683         * WebKitTestRunner/cairo/TestInvocationCairo.cpp: Added.
36684         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
36685         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
36686
36687 2011-03-24  Dirk Pranke  <dpranke@chromium.org>
36688
36689         Reviewed by Tony Chang.
36690
36691         new-run-webkit-tests: clean up worker model defaults. This
36692         changes the default behavior to 'threads' instead of
36693         'old-threads', and clarifies that chromium-win-* and
36694         chromium-mac-leopard are exceptions.
36695
36696         https://bugs.webkit.org/show_bug.cgi?id=56971
36697
36698         * Scripts/webkitpy/layout_tests/port/base.py:
36699         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
36700         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
36701         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
36702         * Scripts/webkitpy/layout_tests/port/test.py:
36703         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
36704
36705 2011-03-23  Leandro Gracia Gil  <leandrogracia@chromium.org>
36706
36707         Reviewed by Darin Adler.
36708
36709         Media Stream API: fix the order of the options in build-webkit.
36710         https://bugs.webkit.org/show_bug.cgi?id=56949
36711
36712         Fix the order of the options in build-webkit.
36713
36714         * Scripts/build-webkit:
36715
36716 2011-03-23  Adam Klein  <adamk@chromium.org>
36717
36718         Reviewed by Mihai Parparita.
36719
36720         Add an option to new-run-webkit-httpd to allow overriding the location of LayoutTests directory
36721         https://bugs.webkit.org/show_bug.cgi?id=56884
36722
36723         This option will be utilized by Chromium's ui_tests to allow
36724         referencing js-test-resources from a layout test over HTTP.
36725
36726         * Scripts/new-run-webkit-httpd:
36727         Added --layout_tests_dir option.
36728         * Scripts/webkitpy/layout_tests/port/http_server.py:
36729         Plumbed through as layout_tests_dir param.
36730
36731 2011-03-23  Dirk Pranke  <dpranke@chromium.org>
36732
36733         Reviewed by Mihai Parparita.
36734
36735         new-run-webkit-tests: run tests in ascending alphabetical order per
36736         dir. This is a second attempt at the fix; the first attempt, in
36737         r81597, broke the behavior on the old-inline and old-threads
36738         worker models.
36739
36740         https://bugs.webkit.org/show_bug.cgi?id=56760
36741
36742         * Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py:
36743         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
36744         * Scripts/webkitpy/layout_tests/port/test.py:
36745         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
36746
36747 2011-03-23  Mario Sanchez Prada  <msanchez@igalia.com>
36748
36749         Reviewed by Martin Robinson.
36750
36751         [GTK] GTK's DRT to ouput detail's string for the 'property-change' signal
36752         https://bugs.webkit.org/show_bug.cgi?id=56953
36753
36754         Print the string for the detail of 'property-change' signal.
36755
36756         * DumpRenderTree/gtk/AccessibilityCallbacks.cpp:
36757         (axObjectEventListener): Print the detail string.
36758
36759 2011-03-23  Yury Semikhatsky  <yurys@chromium.org>
36760
36761         Reviewed by Pavel Feldman.
36762
36763         [V8] Web Inspector: compile DebuggerScript.js into DebuggerScriptSource.h
36764         https://bugs.webkit.org/show_bug.cgi?id=56843
36765
36766         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
36767         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
36768
36769 2011-03-22  Eric Seidel  <eric@webkit.org>
36770
36771         Reviewed by Adam Barth.
36772
36773         Teach update-webkit about gyp
36774         https://bugs.webkit.org/show_bug.cgi?id=56881
36775
36776         This makes it possible for someone who builds from Xcode
36777         to just run update-webkit --gyp and have the project files
36778         generated after update automatically.
36779
36780         I also added a stand-alone "generate-project-files" script
36781         wrapping Source/gyp/configure.  This makes running gyp simpler
36782         since most people will not have gyp or Source/gyp in their path.
36783
36784         * Scripts/build-webkit:
36785         * Scripts/generate-project-files: Added.
36786         * Scripts/update-webkit:
36787
36788 2011-03-22  Anton D'Auria  <adauria@apple.com>
36789
36790         Reviewed by Alexey Proskuryakov.
36791
36792         Add +[WebApplicationCache getOriginsWithCache]
36793         https://bugs.webkit.org/show_bug.cgi?id=56722
36794
36795         * DumpRenderTree/LayoutTestController.cpp:
36796         (originsWithApplicationCacheCallback):
36797         (LayoutTestController::staticFunctions):
36798         * DumpRenderTree/LayoutTestController.h:
36799         * DumpRenderTree/chromium/LayoutTestController.cpp: Added stub.
36800         (LayoutTestController::originsWithApplicationCache):
36801         * DumpRenderTree/chromium/LayoutTestController.h:
36802         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Added stub.
36803         (LayoutTestController::originsWithApplicationCache):
36804         * DumpRenderTree/mac/LayoutTestControllerMac.mm: Pass array of origin's database identifiers.
36805         (originsArrayToJS): Helper function for converting array of WebSecurityOrigins to a JS array of origin identifiers.
36806         (LayoutTestController::originsWithApplicationCache):
36807         (LayoutTestController::originsWithLocalStorage):
36808         * DumpRenderTree/qt/LayoutTestControllerQt.cpp: Added stub.
36809         (LayoutTestController::originsWithApplicationCache):
36810         * DumpRenderTree/qt/LayoutTestControllerQt.h:
36811         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Added stub.
36812         (LayoutTestController::originsWithApplicationCache):
36813         * DumpRenderTree/wx/LayoutTestControllerWx.cpp: Added stub.
36814         (LayoutTestController::originsWithApplicationCache):
36815
36816 2011-03-22  Eric Seidel  <eric@webkit.org>
36817
36818         Reviewed by Adam Barth.
36819
36820         Add support to build-webkit for building with gyp-generated project files
36821         https://bugs.webkit.org/show_bug.cgi?id=56877
36822
36823         build-webkit is a cesspool.  This change just hacks in --gyp support for Mac.
36824         Eventually we'll want to support gyp-generated projects in a more general manner.
36825
36826         * Scripts/build-webkit:
36827
36828 2011-03-22  David Kilzer  <ddkilzer@apple.com>
36829
36830         <http://webkit.org/b/56781> Add --dsym switch to enable dsym generation when building with Xcode
36831
36832         Reviewed by Joseph Pecoraro.
36833
36834         * Scripts/build-webkit: Updated usage statement to include
36835         --dsym.
36836         * Scripts/webkitdirs.pm: Added $generateDsym variable.
36837         (generateDsym): Added. Call determineGenerateDsym() and
36838         return $generateDsym.
36839         (determineGenerateDsym): Added.  Parse @ARGV for --dsym and set
36840         $generateDsym.
36841         (argumentsForXcode): Added.  Returns additional arguments for
36842         xcodebuild based on command-line switches.
36843         (XcodeOptions): Updated to include argumentsForXcode() in the
36844         array returned.
36845
36846 2011-03-22  Adam Roben  <aroben@apple.com>
36847
36848         Make Leopard Debug bots and Windows XP Debug (Tests) core builders
36849
36850         These bots are consistently green again.
36851
36852         Fixes <http://webkit.org/b/56830> Leopard Debug bots and Windows XP Debug (Tests) should be
36853         core builders
36854
36855         Rubber-stamped by Antti Koivisto.
36856
36857         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
36858         (BuildBot.__init__): Loosened the regexps for Leopard and Windows.
36859
36860         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
36861         (BuildBotTest.test_builder_name_regexps): Updated expectations.
36862
36863 2011-03-21  Sheriff Bot  <webkit.review.bot@gmail.com>
36864
36865         Unreviewed, rolling out r81597.
36866         http://trac.webkit.org/changeset/81597
36867         https://bugs.webkit.org/show_bug.cgi?id=56801
36868
36869         Change results in lots of unexpected flaky on the dashboard.
36870         (Requested by pfeldman on #webkit).
36871
36872         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
36873         * Scripts/webkitpy/layout_tests/port/test.py:
36874         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
36875
36876 2011-03-21  Sam Weinig  <sam@webkit.org>
36877
36878         Fix failing tests on the WebKit2 bots.
36879
36880         * WebKitTestRunner/TestController.cpp:
36881         (WTR::TestController::resetStateToConsistentValues):
36882
36883 2011-03-21  Daniel Sievers  <sievers@google.com>
36884
36885         Reviewed by Simon Fraser.
36886
36887         [Chromium] Make RenderAsTextBehavior and LayerTreeAsTextBehavior tweakable from the DumpRenderTree commandline
36888         https://bugs.webkit.org/show_bug.cgi?id=56139
36889
36890         * DumpRenderTree/chromium/DumpRenderTree.cpp:
36891         (main):
36892         * DumpRenderTree/chromium/LayoutTestController.cpp:
36893         (LayoutTestController::LayoutTestController):
36894         (LayoutTestController::layerTreeAsText):
36895         * DumpRenderTree/chromium/LayoutTestController.h:
36896         (LayoutTestController::setShowDebugLayerTree):
36897         * DumpRenderTree/chromium/TestShell.cpp:
36898         (TestShell::runFileTest):
36899         (TestShell::dump):
36900         * DumpRenderTree/chromium/TestShell.h:
36901         (TestParams::TestParams):
36902
36903 2011-03-21  Adam Roben  <aroben@apple.com>
36904
36905         Try again to get buildbot to show run-javascriptcore-tests's actual.html as an HTML file
36906
36907         Fixes <http://webkit.org/b/56746> build.webkit.org shows run-javascriptcore-tests's HTML
36908         output as plain text
36909
36910         Reviewed by David Kilzer.
36911
36912         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
36913         (RunJavaScriptCoreTests): List actual.html as one of our log files so it will get uploaded
36914         to the master.
36915         (RunJavaScriptCoreTests.commandComplete): Reading the actual.html file here won't work,
36916         since we're running on the build master, not the slave. Instead, turn the raw source of
36917         actual.html which the build slave uploaded into an HTML log.
36918
36919 2011-03-21  Dirk Pranke  <dpranke@chromium.org>
36920
36921         Reviewed by Mihai Parparita.
36922
36923         new-run-webkit-tests: run tests in ascending alphabetical order per dir
36924         https://bugs.webkit.org/show_bug.cgi?id=56760
36925
36926         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
36927         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
36928         * Scripts/webkitpy/layout_tests/port/test.py:
36929
36930 2011-03-21  Leandro Gracia Gil  <leandrogracia@chromium.org>
36931
36932         Reviewed by Steve Block.
36933
36934         Media Stream API patch 0: adding compilation guards.
36935         https://bugs.webkit.org/show_bug.cgi?id=56458
36936
36937         Adding options to enable the media stream API feature in Chromium.
36938
36939         * Scripts/build-webkit:
36940
36941 2011-03-21  Qi Zhang  <qi.2.zhang@nokia.com>
36942
36943         Reviewed by Laszlo Gombos.
36944
36945         [Qt] Add a command line option to capture stdout and stderr for DumpRenderTree
36946         https://bugs.webkit.org/show_bug.cgi?id=56323
36947
36948         Using freopen to redirect STDOUT and STDERR when DumpRenderTree command line provide
36949         "--stdout" or "--stderr" option.
36950
36951         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
36952         (WebCore::DumpRenderTree::~DumpRenderTree):
36953         (WebCore::DumpRenderTree::processArgsLine):
36954         (WebCore::DumpRenderTree::loadNextTestInStandAloneMode):
36955         * DumpRenderTree/qt/DumpRenderTreeQt.h:
36956         (WebCore::DumpRenderTree::setRedirectOutputFileName):
36957         (WebCore::DumpRenderTree::setRedirectErrorFileName):
36958         * DumpRenderTree/qt/main.cpp:
36959         (isOption):
36960         (takeOptionValue):
36961         (printUsage):
36962         (main):
36963
36964 2011-03-21  Adam Roben  <aroben@apple.com>
36965
36966         Fix exceptions on the buildbot due to my last change
36967
36968         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
36969         (RunJavaScriptCoreTests.commandComplete): Catch exceptions due to actual.html not existing.
36970
36971 2011-03-21  Adam Roben  <aroben@apple.com>
36972
36973         Teach buildbot to treat run-javascriptcore-tests's actual.html as an HTML file
36974
36975         Fixes <http://webkit.org/b/56746> build.webkit.org shows run-javascriptcore-tests's HTML
36976         output as plain text
36977
36978         Reviewed by Darin Adler.
36979
36980         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
36981         (RunJavaScriptCoreTests): Removed the logfiles variable, which can only handle plaintext
36982         logs.
36983         (RunJavaScriptCoreTests.commandComplete): Use the addHTMLLog method to upload actual.html to
36984         the build master. This will correctly treat it as HTML.
36985
36986 2011-03-20  Bill Budge  <bbudge@chromium.org>
36987
36988         Reviewed by Adam Barth.
36989
36990         Rename ThreadSafeShared to ThreadSafeRefCounted
36991         https://bugs.webkit.org/show_bug.cgi?id=56714
36992
36993         No new tests. Exposes no new functionality.
36994
36995         * DumpRenderTree/ForwardingHeaders/wtf/ThreadSafeRefCounted.h: Copied from DumpRenderTree/ForwardingHeaders/wtf/ThreadSafeShared.h.
36996         * DumpRenderTree/ForwardingHeaders/wtf/ThreadSafeShared.h: Removed.
36997         * Scripts/do-webcore-rename:
36998
36999 2011-03-19  Anton D'Auria  <adauria@apple.com>
37000
37001         Reviewed by Dan Bernstein.
37002
37003         Add stub to LayoutTestControllerGtk.cpp to fix GTK build
37004         https://bugs.webkit.org/show_bug.cgi?id=56719
37005
37006         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Adding stub.
37007         (LayoutTestController::clearApplicationCacheForOrigin):
37008
37009 2011-03-19  Anton D'Auria  <adauria@apple.com>
37010
37011         Reviewed by Alexey Proskuryakov.
37012
37013         ApplicationCacheGroup is not obsolete after being deleted via ApplicationCacheStorage::deleteEntriesForOrigin
37014         https://bugs.webkit.org/show_bug.cgi?id=56415
37015
37016         * DumpRenderTree/LayoutTestController.cpp:
37017         (clearApplicationCacheForOriginCallback):
37018         (LayoutTestController::staticFunctions):
37019         * DumpRenderTree/LayoutTestController.h:
37020         * DumpRenderTree/chromium/LayoutTestController.cpp: Adding stubs.
37021         (LayoutTestController::LayoutTestController):
37022         (LayoutTestController::clearAllApplicationCaches):
37023         (LayoutTestController::clearApplicationCacheForOrigin):
37024         * DumpRenderTree/chromium/LayoutTestController.h:
37025         * DumpRenderTree/mac/LayoutTestControllerMac.mm: Adding stub.
37026         (LayoutTestController::clearApplicationCacheForOrigin):
37027         * DumpRenderTree/qt/LayoutTestControllerQt.cpp: Adding stub.
37028         (LayoutTestController::clearApplicationCacheForOrigin):
37029         * DumpRenderTree/qt/LayoutTestControllerQt.h:
37030         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Adding stub.
37031         (LayoutTestController::clearApplicationCacheForOrigin):
37032         * DumpRenderTree/wx/LayoutTestControllerWx.cpp: Adding stub.
37033         (LayoutTestController::clearApplicationCacheForOrigin):
37034
37035 2011-03-18  Dan Bernstein  <mitz@apple.com>
37036
37037         Revised build fix for r81135.
37038
37039         * WebKitTestRunner/Configurations/InjectedBundle.xcconfig:
37040         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
37041
37042 2011-03-18  Adam Roben  <aroben@apple.com>
37043
37044         Windows build fix
37045
37046         * WebKitTestRunner/config.h: Add a missing #endif.
37047
37048 2011-03-18  Andreas Kling  <kling@webkit.org>
37049
37050         Unbreak Qt build after r81494.
37051
37052         * WebKitTestRunner/qt/TestInvocationQt.cpp:
37053         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
37054
37055 2011-03-18  Sam Weinig  <sam@webkit.org>
37056
37057         Fix windows build. This was really Adam Roben's fault. I swear!
37058
37059         * WebKitTestRunner/config.h:
37060
37061 2011-03-18  Sam Weinig  <sam@webkit.org>
37062
37063         Reviewed by Adam Roben.
37064
37065         Pixel tests don't work in WebKitTestRunner on Windows
37066         https://bugs.webkit.org/show_bug.cgi?id=56630
37067
37068         Cross-platformize (Mac and Windows CG) the pixel dumping code in WebKitTestRunner to use WebKit2
37069         built-in snapshotting instead of using the window server on the Mac.
37070
37071         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
37072         (WTR::InjectedBundle::InjectedBundle):
37073         (WTR::InjectedBundle::didReceiveMessage):
37074         (WTR::InjectedBundle::beginTesting):
37075         (WTR::InjectedBundle::done):
37076         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
37077         (WTR::InjectedBundle::setPixelResult):
37078         (WTR::InjectedBundle::shouldDumpPixels):
37079         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
37080         (WTR::InjectedBundlePage::dump):
37081         Take the snapshot in the bundle and send it to the UIProcess for dumping.
37082
37083         * WebKitTestRunner/TestInvocation.cpp:
37084         (WTR::TestInvocation::invoke):
37085         (WTR::TestInvocation::dump):
37086         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
37087         * WebKitTestRunner/TestInvocation.h:
37088         * WebKitTestRunner/cg: Added.
37089         * WebKitTestRunner/cg/TestInvocationCG.cpp: Copied from WebKitTestRunner/mac/TestInvocationMac.mm.
37090         (WTR::createCGContextFromImage):
37091         (WTR::computeMD5HashStringForContext):
37092         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
37093         * WebKitTestRunner/mac/TestInvocationMac.mm: Removed.
37094         Making pixel dumping code shared for CG ports. Also use the WTF/MD5 instead of platform specific code.
37095
37096         * WebKitTestRunner/win/TestInvocationWin.cpp:
37097         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
37098         Keep stub for Cairo build.
37099
37100         * WebKitTestRunner/Configurations/WebKitTestRunnerCoreGraphics.vsprops: Added.
37101         Add CG configuration.
37102
37103         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
37104         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
37105         Add new files.
37106
37107 2011-03-18  Adam Roben  <aroben@apple.com>
37108
37109         Don't allow Leaks Viewer's vertical scrollbar to cover up long function names
37110
37111         Rubber-stamped by Simon Fraser.
37112
37113         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
37114         (.data-grid .data-container): Use overflow-y:auto instead of the Inspector's default
37115         overflow-y:overlay so that the scrollbar doesn't overlay the content.
37116
37117 2011-03-18  Adam Roben  <aroben@apple.com>
37118
37119         Don't ellipsize long or indented function names in Leaks Viewer
37120
37121         Fixes <http://webkit.org/b/56037> Leaks viewer needs horizontal
37122         scrollbar
37123
37124         Reviewed by Simon Fraser.
37125
37126         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
37127         (.data-grid .data-container):
37128         (.data-grid td > div, .data-grid th > div):
37129         Allow the contents to extend to the right.
37130
37131 2011-03-18  Adam Roben  <aroben@apple.com>
37132
37133         Fix some leaks in DRT seen on the leaks bot
37134
37135         Fixes <http://webkit.org/b/56638> addURLToRedirectCallback and
37136         setWillSendRequestClearHeaderCallback in DRT are leaky
37137
37138         Reviewed by Alexey Proskuryakov.
37139
37140         * DumpRenderTree/ForwardingHeaders/wtf/OwnArrayPtr.h: Added.
37141
37142         * DumpRenderTree/LayoutTestController.cpp:
37143         (addURLToRedirectCallback):
37144         (setWillSendRequestClearHeaderCallback):
37145         Use OwnArrayPtr to cause the strings to be deleted.
37146
37147 2011-03-18  Adam Roben  <aroben@apple.com>
37148
37149         Fix a NSURLCredential leak seen on the leaks bot
37150
37151         Fixes <http://webkit.org/b/56637> -[SynchronousLoader
37152         connection:didReceiveAuthenticationChallenge:] is leaky
37153
37154         Reviewed by John Sullivan.
37155
37156         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
37157         (-[SynchronousLoader connection:didReceiveAuthenticationChallenge:]): Use a RetainPtr to
37158         cause the NSURLCredential to be released.
37159
37160 2011-03-18  John Knottenbelt  <jknotten@chromium.org>
37161
37162         Unreviewed.
37163
37164         Remove unnecessary brackets around my email address.
37165
37166         * Scripts/webkitpy/common/config/committers.py:
37167
37168 2011-03-18  Adam Roben  <aroben@apple.com>
37169
37170         Remove some files I accidentally added in r81454
37171
37172         * WebKitTestRunner/Configurations/WebKitTestRunnerCoreGraphics.vsprops: Removed.
37173         * WebKitTestRunner/cg/TestInvocationCG.cpp: Removed.
37174
37175 2011-03-18  Adam Roben  <aroben@apple.com>
37176
37177         Ignore an NSNumberFormatter leak on SnowLeopard
37178
37179         * Scripts/old-run-webkit-tests:
37180         (countAndPrintLeaks): Ignore leaks in -[NSNumberFormatter
37181         getObjectValue:forString:errorDescription:].
37182
37183 2011-03-18  Adam Roben  <aroben@apple.com>
37184
37185         Fix a leak beneath LayoutTestController::authenticateSession seen on the bots
37186
37187         Rubber-stamped by Mark Rowe.
37188
37189         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
37190         (LayoutTestController::authenticateSession): Use a RetainPtr to cause the NSURLRequest to be
37191         released.
37192
37193 2011-03-17  Adam Roben  <aroben@apple.com>
37194
37195         Address a review comment I missed in r81445
37196
37197         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
37198         (LeaksParserWorker.prototype._parseLeaks): Anchor the RegExp used to find the number of
37199         leaked bytes to speed up searching on long lines.
37200
37201 2011-03-17  Adam Roben  <aroben@apple.com>
37202
37203         Mark leaky builds orange, even if all regression tests passed
37204
37205         Fixes <http://webkit.org/b/56609> SnowLeopard Intel Leaks bot appears green on
37206         build.webkit.org when all tests pass even though there are still leaks
37207
37208         Reviewed by Dan Bernstein.
37209
37210         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
37211         (RunWebKitLeakTests): Set the warnOnWarnings property to True so that the entire build will
37212         be marked orange if this build step generates warnings.
37213
37214 2011-03-17  Adam Roben  <aroben@apple.com>
37215
37216         Allow values in Leaks Viewer to be shown as percentages
37217
37218         Double-clicking on a column will toggle between showing percentages and showing real values.
37219
37220         Fixes <http://webkit.org/b/56628> Would like to view values as percentages of the total
37221         number of leaked bytes in Leaks Viewer
37222
37223         Reviewed by Joe Pecoraro.
37224
37225         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
37226         (LeaksParserWorker.prototype._incorporateLeaks): Calculate the (never-shown) top-level
37227         node's totalTime, which is used in calculating percentages.
37228
37229         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/WebInspectorShims.js:
37230         (monkeyPatchInspectorObjects): Only format values as bytes if we aren't supposed to be
37231         showing them as percentages.
37232
37233 2011-03-17  Adam Roben  <aroben@apple.com>
37234
37235         Show the number of leaked bytes, not just leaked allocations, in Leaks Viewer
37236
37237         The Self and Total columns now report the number of leaked bytes. The new Calls column shows
37238         the number of leaked allocations. And the new Average column shows the average number of
37239         leaked bytes per call.
37240
37241         Fixes <http://webkit.org/b/56344> Leaks Viewer should show how many bytes were leaked, not
37242         just how many allocations
37243
37244         Reviewed by Joe Pecoraro.
37245
37246         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js:
37247         (LeaksParserWorker.prototype._parseLeaks): Parse the number of leaked bytes from the "Leak:"
37248         line and store it with the leak stack.
37249         (LeaksParserWorker.prototype._createNode): Initialize the averageTime and numberOfCalls
37250         properties.
37251         (LeaksParserWorker.prototype._incorporateLeaks): Updated to get the stack from the .stack
37252         property, and changed to store the number of leaked bytes in .selfTime/.totalTime.
37253         .numberOfCalls now holds the number of leaked allocations.
37254
37255         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/WebInspectorShims.js:
37256         (Preferences): Set samplingCPUProfiler to false so that the Average and Calls columns will
37257         appear.
37258         (monkeyPatchInspectorObjects): Format the values as bytes.
37259
37260 2011-03-17  Adam Roben  <aroben@apple.com>
37261
37262         Make Leaks Viewer able to count leaks in builds that didn't fail any regression tests
37263
37264         Fixes <http://webkit.org/b/56626> Leaks Viewer omits recent builds which didn't fail any
37265         regression tests
37266
37267         Reviewed by James Robinson.
37268
37269         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
37270         (RecentBuildsLoader.prototype.start): Get the number of leaks out of the "text" array, which
37271         is always populated, even when all regression tests pass.
37272
37273 2011-03-10  Hayato Ito  <hayato@chromium.org>
37274
37275         Reviewed by Tony Chang.
37276
37277         Add a check logic for reftests in case that a reftest is marked AS REBASELINE.
37278
37279         https://bugs.webkit.org/show_bug.cgi?id=56076
37280
37281         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
37282
37283 2011-03-17  Adam Roben  <aroben@apple.com>
37284
37285         Skip a compositing-sensitive test when accelerated compositing is disabled
37286
37287         Reviewed by Sam Weinig.
37288
37289         * Scripts/old-run-webkit-tests: Skip media/media-document-audio-repaint.html when
37290         accelerated compositing is disabled.
37291
37292 2011-03-17  Adam Roben  <aroben@apple.com>
37293
37294         Start using PlatformSpecificScheduler for the Mac builders
37295
37296         The current set of file/directory patterns should work for Mac builders. We may discover
37297         that there are more files and directories we can exclude from the Mac builds, but they can
37298         be added later.
37299
37300         Fixes <http://webkit.org/b/56593> Mac builders do lots of unnecessary builds
37301
37302         Reviewed by Mark Rowe.
37303
37304         * BuildSlaveSupport/build.webkit.org-config/config.json: Moved the Mac builders into two
37305         PlatformSpecificSchedulers: one for Leopard, and one for SnowLeopard.
37306
37307         * Scripts/webkitpy/common/config/build.py:
37308         (_should_file_trigger_build): Allow "mac-leopard" and "mac-snowleopard" platforms, and
37309         cleaned up patterns a little bit.
37310
37311         * Scripts/webkitpy/common/config/build_unittest.py:
37312         (ShouldBuildTest): Fixed some expected results.
37313         (ShouldBuildTest.test_should_build): Start testing "mac-leopard" and "mac-snowleopard"
37314         platforms.
37315
37316 2011-03-17  Jeff Miller  <jeffm@apple.com>
37317
37318         Use a consistent set of file patterns in the svn:ignore property for all .xcodeproj directories, specifically:
37319
37320         *.mode*
37321         *.pbxuser
37322         *.perspective*
37323         project.xcworkspace
37324         xcuserdata
37325
37326         * DumpRenderTree/DumpRenderTree.xcodeproj: Modified property svn:ignore.
37327         * MiniBrowser/MiniBrowser.xcodeproj: Modified property svn:ignore.
37328         * TestWebKitAPI/TestWebKitAPI.xcodeproj: Modified property svn:ignore.
37329         * WebKitLauncher/WebKitLauncher.xcodeproj: Modified property svn:ignore.
37330         * WebKitTestRunner/WebKitTestRunner.xcodeproj: Modified property svn:ignore.
37331
37332 2011-03-17  Adam Roben  <aroben@apple.com>
37333
37334         Make old-run-webkit-tests pick up mac-wk2 results when running in WebKit2 mode on Windows
37335
37336         Fixes <http://webkit.org/b/56563> Running WebKit2 tests on Windows should use results from
37337         mac-wk2
37338
37339         Reviewed by Anders Carlsson.
37340
37341         * Scripts/old-run-webkit-tests:
37342         (expectedDirectoryForTest): Clean up the code to search in extra Mac platform result
37343         directories on Windows. Add mac-wk2 to the set of extra platforms we look in on Windows when
37344         we're running in WebKit2 mode.
37345
37346 2011-03-17  Mikhail Naganov  <mnaganov@chromium.org>
37347
37348         Reviewed by Yury Semikhatsky.
37349
37350         Web Inspector: Clean up Inspector strings.
37351         https://bugs.webkit.org/show_bug.cgi?id=56557
37352
37353         Modify the script to deal correctly with "\xNN" escapes, and to
37354         check strings in Chromium's DevTools.js.
37355
37356         * Scripts/check-inspector-strings:
37357
37358 2011-03-17  Keith Kyzivat  <keith.kyzivat@nokia.com>
37359
37360         Reviewed by Benjamin Poulain.
37361
37362         Update autoinstall to install mechanize 0.2.4
37363         https://bugs.webkit.org/show_bug.cgi?id=56499
37364
37365         Newer versions of mechanize support HTTPS proxies, which is necessary
37366         for people using webkit-patch behind firewalls.
37367
37368         * Scripts/webkitpy/common/system/autoinstall.py:
37369         * Scripts/webkitpy/thirdparty/__init__.py:
37370
37371 2011-03-16  Sam Weinig  <sam@webkit.org>
37372
37373         Reviewed by Adam Roben.
37374
37375         REGRESSION (r81084): media/invalid-media-url-crash.html and editing/execCommand/indent-paragraphs.html failing on WebKit2
37376         https://bugs.webkit.org/show_bug.cgi?id=56541
37377
37378         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
37379         (WTR::InjectedBundlePage::willSendRequestForFrame):
37380         Check for an empty host in addition to a null host.  This bug was introduced
37381         when porting off of the KURL code and onto the WKURLRef version.
37382
37383 2011-03-16  Sam Weinig  <sam@webkit.org>
37384
37385         Reviewed by Adam Roben.
37386
37387         Add Lucas Forschler to the list of WebKit committers.
37388
37389         * Scripts/webkitpy/common/config/committers.py:
37390
37391 2011-03-16  Sam Weinig  <sam@webkit.org>
37392
37393         Reviewed upon and dictated through Adam Roben.
37394
37395         Escape crashing function names in results.html. This fixes an issue
37396         when a crash involves a type with templates.
37397
37398         * Scripts/old-run-webkit-tests:
37399
37400 2011-03-16  Daniel Sievers  <sievers@google.com>
37401
37402         Reviewed by James Robinson.
37403
37404         Add setting to always force compositing mode
37405         https://bugs.webkit.org/show_bug.cgi?id=56156
37406
37407         * DumpRenderTree/chromium/DumpRenderTree.cpp:
37408         (main):
37409         * DumpRenderTree/chromium/TestShell.cpp:
37410         (TestShell::TestShell):
37411         (TestShell::resetWebSettings):
37412         * DumpRenderTree/chromium/TestShell.h:
37413         (TestShell::setForceCompositingMode):
37414         * DumpRenderTree/chromium/WebPreferences.cpp:
37415         (WebPreferences::reset):
37416         (WebPreferences::applyTo):
37417         * DumpRenderTree/chromium/WebPreferences.h:
37418
37419 2011-03-15  Hayato Ito  <hayato@chromium.org>
37420
37421         Reviewed by Ojan Vafai.
37422
37423         Ignore files that will be used by reftests in old-run-webkit-tests.
37424
37425         https://bugs.webkit.org/show_bug.cgi?id=55936
37426
37427         * Scripts/old-run-webkit-tests:
37428
37429 2011-03-15  Felipe Oliveira da Silva Netto  <f.dachshund@gmail.com>
37430
37431         Reviewed by Adam Roben.
37432
37433         Function relativeScriptsDir doesn't return the correct relative
37434         Scripts directory.
37435
37436         * Scripts/webkitdirs.pm:
37437         (relativeScriptsDir): Ensure we use the directory that contains
37438         webkitdirs.pm, not the one that contains the original script that
37439         was invoked.
37440
37441 2011-03-15  Kevin Ollivier  <kevino@theolliviers.com>
37442
37443         [wx] Build fixes for long linker commands on Win and method return error, plus a Mac
37444         fix to link against the Security framework after recent WebKitSystemInterface changes.
37445
37446         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
37447         (LayoutTestController::originsWithLocalStorage):
37448         * wx/build/settings.py:
37449         * wx/build/waf_extensions.py:
37450
37451 2011-03-15  Sam Weinig  <sam@webkit.org>
37452
37453         Reviewed by Darin Adler.
37454
37455         Add script to convert a patch to a prettypatch and show in the default browser
37456         https://bugs.webkit.org/show_bug.cgi?id=56416
37457
37458         This can be used either like
37459             $ svn-create-patch | show-pretty-diff
37460         or
37461             $ show-pretty-diff patch.diff
37462
37463         * Scripts/show-pretty-diff: Added.
37464
37465 2011-03-15  Brian Weinstein  <bweinstein@apple.com>
37466
37467         Rubber-stamped by Adam Roben.
37468
37469         Adding myself as a reviewer.
37470
37471         * Scripts/webkitpy/common/config/committers.py:
37472
37473 2011-03-15  Anders Carlsson  <andersca@apple.com>
37474
37475         Try to fix the Windows build.
37476
37477         * DumpRenderTree/TestNetscapePlugIn/Tests/NPDeallocateCalledBeforeNPShutdown.cpp:
37478         (NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject):
37479
37480 2011-03-15  Anders Carlsson  <andersca@apple.com>
37481
37482         Reviewed by Sam Weinig.
37483
37484         Make sure that NP_Shutdown is always the last NPP function called
37485         https://bugs.webkit.org/show_bug.cgi?id=56391
37486
37487         Add a test that times out if an NP_Deallocate object is called after NP_Shutdown.
37488
37489         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
37490         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
37491         (PluginTest::PluginTest):
37492         (PluginTest::NP_Shutdown):
37493         (PluginTest::registerNPShutdownFunction):
37494         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
37495         * DumpRenderTree/TestNetscapePlugIn/Tests/NPDeallocateCalledBeforeNPShutdown.cpp: Added.
37496         (NPDeallocateCalledBeforeNPShutdown::NPDeallocateCalledBeforeNPShutdown):
37497         (NPDeallocateCalledBeforeNPShutdown::TestObject::~TestObject):
37498         (NPDeallocateCalledBeforeNPShutdown::ScriptableObject::hasProperty):
37499         (NPDeallocateCalledBeforeNPShutdown::ScriptableObject::getProperty):
37500         (NPDeallocateCalledBeforeNPShutdown::NPP_New):
37501         (NPDeallocateCalledBeforeNPShutdown::NPP_GetValue):
37502         (NPDeallocateCalledBeforeNPShutdown::shutdown):
37503         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
37504         (NP_Shutdown):
37505         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
37506         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
37507         * GNUmakefile.am:
37508
37509 2011-03-15  Ilya Sherman  <isherman@chromium.org>
37510
37511         Reviewed by Tony Chang.
37512
37513         Autofilled form elements are assigned fixed background color but not text color
37514         https://bugs.webkit.org/show_bug.cgi?id=48382
37515
37516         * DumpRenderTree/LayoutTestController.cpp:
37517         (setAutofilledCallback): Added.
37518         (LayoutTestController::staticFunctions):
37519         * DumpRenderTree/LayoutTestController.h:
37520         * DumpRenderTree/chromium/LayoutTestController.cpp:
37521         (LayoutTestController::LayoutTestController):
37522         (LayoutTestController::setAutofilled): Added.
37523         * DumpRenderTree/chromium/LayoutTestController.h:
37524         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
37525         (LayoutTestController::setAutofilled): Added.
37526         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
37527         (LayoutTestController::setAutofilled): Added.
37528         (LayoutTestController::setValueForUser): Style tweak.
37529         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
37530         (LayoutTestController::setAutofilled): Added (stub).
37531         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
37532         (LayoutTestController::setAutofilled): Added (stub).
37533
37534 2011-03-15  Dan Bernstein  <mitz@apple.com>
37535
37536         Build fix after r81135.
37537
37538         * TestWebKitAPI/Configurations/InjectedBundle.xcconfig:
37539
37540 2011-03-15  Dan Bernstein  <mitz@apple.com>
37541
37542         Build fix after r81135.
37543
37544         * WebKitTestRunner/Configurations/InjectedBundle.xcconfig:
37545         * WebKitTestRunner/mac/TestControllerMac.mm:
37546         * WebKitTestRunner/mac/TestInvocationMac.mm:
37547         * WebKitTestRunner/mac/main.mm:
37548
37549 2011-03-15  Balazs Kelemen  <kbalazs@webkit.org>
37550
37551         Reviewed by Adam Barth.
37552
37553         REGRESSION(80977): breaks Qt ARMv7 Linux Release bot
37554         https://bugs.webkit.org/show_bug.cgi?id=56349
37555
37556         * QtTestBrowser/QtTestBrowser.pro: Link against fontconfig
37557         on platforms where we use it. Copied from DumpRenderTree.pro.
37558
37559 2011-03-15  Dan Bernstein  <mitz@apple.com>
37560
37561         Build fix after r81135.
37562
37563         * DumpRenderTree/cf/WebArchiveDumpSupport.cpp:
37564         * DumpRenderTree/mac/WebArchiveDumpSupportMac.mm:
37565
37566 2011-03-15  Kevin Ollivier  <kevino@theolliviers.com>
37567
37568         Reviewed by Darin Adler.
37569
37570         Introduce WTF_USE_EXPORT_MACROS, which will allow us to put shared library import/export
37571         info into the headers rather than in export symbol definition files, but disable it on
37572         all platforms initially so we can deal with port build issues one port at a time.
37573
37574         https://bugs.webkit.org/show_bug.cgi?id=27551
37575
37576         * DumpRenderTree/cg/ImageDiffCG.cpp:
37577         * DumpRenderTree/chromium/config.h:
37578         * DumpRenderTree/config.h:
37579         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
37580         * TestWebKitAPI/TestWebKitAPIPrefix.h:
37581         * WebKitAPITest/HostWindow.cpp:
37582         * WebKitAPITest/TestsController.cpp:
37583         * WebKitAPITest/config.h: Added.
37584         * WebKitAPITest/main.cpp:
37585         * WebKitAPITest/tests/WebViewDestruction.cpp:
37586         * WebKitTestRunner/InjectedBundle/Bindings/JSWrapper.cpp:
37587         * WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
37588         * WebKitTestRunner/InjectedBundle/GCController.cpp:
37589         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
37590         * WebKitTestRunner/InjectedBundle/InjectedBundleMain.cpp:
37591         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
37592         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
37593         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
37594         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
37595         * WebKitTestRunner/InjectedBundle/qt/LayoutTestControllerQt.cpp:
37596         * WebKitTestRunner/InjectedBundle/win/ActivateFonts.cpp:
37597         * WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp:
37598         * WebKitTestRunner/InjectedBundle/win/LayoutTestControllerWin.cpp:
37599         * WebKitTestRunner/TestController.cpp:
37600         * WebKitTestRunner/TestInvocation.cpp:
37601         * WebKitTestRunner/config.h: Added.
37602         * WebKitTestRunner/qt/PlatformWebViewQt.cpp:
37603         * WebKitTestRunner/qt/TestControllerQt.cpp:
37604         * WebKitTestRunner/qt/TestInvocationQt.cpp:
37605         * WebKitTestRunner/qt/main.cpp:
37606         * WebKitTestRunner/win/PlatformWebViewWin.cpp:
37607         * WebKitTestRunner/win/TestControllerWin.cpp:
37608         * WebKitTestRunner/win/TestInvocationWin.cpp:
37609         * WebKitTestRunner/win/main.cpp:
37610         * wx/build/settings.py:
37611
37612 2011-03-15  Hayato Ito  <hayato@chromium.org>
37613
37614         Reviewed by Shinichiro Hamaji.
37615
37616         [NRWT] Add support for reftests to new-run-webkit-tests.
37617
37618         https://bugs.webkit.org/show_bug.cgi?id=55457
37619
37620         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
37621         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
37622         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
37623         * Scripts/webkitpy/layout_tests/port/base.py:
37624         * Scripts/webkitpy/layout_tests/port/dryrun.py:
37625         * Scripts/webkitpy/layout_tests/port/test.py:
37626         * Scripts/webkitpy/layout_tests/port/test_files.py:
37627         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
37628
37629 2011-03-15  Alejandro G. Castro  <alex@igalia.com>
37630
37631         Reviewed by Martin Robinson.
37632
37633         [GTK] Fix compilation warning after r80892
37634         https://bugs.webkit.org/show_bug.cgi?id=56302
37635
37636         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
37637         (LayoutTestController::originsWithLocalStorage):
37638
37639 2011-03-15  Robert Hogan  <robert@webkit.org>
37640
37641         Reviewed by Adam Barth.
37642
37643         Document the qmakearg argument, and hint how to build webkit2 on Qt
37644         https://bugs.webkit.org/show_bug.cgi?id=56326
37645
37646         * Scripts/build-webkit: add --qmakearg to build-webkit --help
37647
37648 2011-03-15  Adam Roben  <aroben@apple.com>
37649
37650         Make Windows 7 Release (Tests) a core builder
37651
37652         We can probably make Windows XP Debug (Tests) a core builder, too, but we should probably
37653         give it a few more days of greenness first.
37654
37655         Fixes <http://webkit.org/b/55665> Windows 7 Release (Tests) should be a core builder
37656
37657         Reviewed by Sam Weinig.
37658
37659         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
37660         (BuildBot): Added a regexp to match Windows 7 Release (Tests).
37661
37662         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
37663         (BuildBotTest): Updated the list of builders to match the current list on build.webkit.org.
37664         Updated the core regexps to match the buildbot module. Updated the expected results to match
37665         the current set of core builders.
37666
37667 2011-03-14  Adam Roben  <aroben@apple.com>
37668
37669         Add forgotten file from r81116
37670
37671         * TestWebKitAPI/Configurations/TestWebKitAPICFNetwork.vsprops: Added.
37672
37673 2011-03-14  Sam Weinig  <sam@webkit.org>
37674
37675         Reviewed by Adam Roben.
37676
37677         about:blank fake responses don't get serialized when sent the UIProcess
37678         <rdar://problem/9108119>
37679         https://bugs.webkit.org/show_bug.cgi?id=56357
37680
37681         * TestWebKitAPI/Tests/WebKit2/AboutBlankLoad.cpp: Added.
37682         (TestWebKitAPI::decidePolicyForResponse):
37683         (TestWebKitAPI::TEST):
37684         Add test that loads of about:blank have a response with a MIMEType of text/html.
37685
37686         * TestWebKitAPI/PlatformUtilities.h:
37687         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
37688         (TestWebKitAPI::Util::MIMETypeForWKURLResponse):
37689         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
37690         (TestWebKitAPI::Util::MIMETypeForWKURLResponse):
37691         Add helper to get the MIMEType from a WKURLResponse.
37692
37693         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
37694         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
37695         Add new files.
37696
37697 2011-03-14  Jarkko Sakkinen  <jarkko.j.sakkinen@gmail.com>
37698
37699         Reviewed by Kenneth Rohde Christiansen.
37700
37701         [Qt] Command-line options for QGLWidget and WebGL to QtTestBrowser
37702         https://bugs.webkit.org/show_bug.cgi?id=56006
37703
37704         * QtTestBrowser/launcherwindow.cpp:
37705         (LauncherWindow::initializeView):
37706         * QtTestBrowser/main.cpp:
37707         (LauncherApplication::handleUserOptions):
37708
37709 2011-03-14  James Kozianski  <koz@chromium.org>
37710
37711         Reviewed by Ojan Vafai.
37712
37713         Add method to make BuildBot return test outputs
37714         https://bugs.webkit.org/show_bug.cgi?id=54374
37715
37716         The Builder.latest_cached_build will be used by the new rebaseline
37717         script to determine the build to use baselines from.
37718
37719         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
37720         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
37721
37722 2011-03-14  Adam Roben  <aroben@apple.com>
37723
37724         Test that WKBundlePageCanHandleRequest returns true for empty document URLs
37725
37726         Reviewed by Sam Weinig.
37727
37728         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest.cpp: Added.
37729         (TestWebKitAPI::didReceiveMessageFromInjectedBundle): Store the result of the test.
37730         (TestWebKitAPI::setInjectedBundleClient): Hook up our callback.
37731         (TestWebKitAPI::TEST): Register "emptyscheme" as an empty document scheme, load a page to
37732         ensure the web process is initialized, then ask the bundle to run the test and assert that
37733         it succeeded.
37734
37735         * TestWebKitAPI/Tests/WebKit2/CanHandleRequest_Bundle.cpp: Added.
37736         (TestWebKitAPI::CanHandleRequestTest::CanHandleRequestTest): Just call up to the base class.
37737         (TestWebKitAPI::canHandleURL): Helper function to test whether WebKit2 claims to be able to
37738         handle a given URL.
37739         (TestWebKitAPI::runTest): Check that empty document URLs can be handled (and that an unknown
37740         URL type cannot be handled).
37741         (TestWebKitAPI::CanHandleRequestTest::didReceiveMessage): Run the test and return the
37742         result.
37743
37744         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
37745         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
37746         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
37747         Added new files to the project.
37748
37749 2011-03-14  Sheriff Bot  <webkit.review.bot@gmail.com>
37750
37751         Unreviewed, rolling out r81094.
37752         http://trac.webkit.org/changeset/81094
37753         https://bugs.webkit.org/show_bug.cgi?id=56355
37754
37755         Broke the chromium DRT related build. (Requested by dave_levin
37756         on #webkit).
37757
37758         * DumpRenderTree/chromium/DumpRenderTree.cpp:
37759         (main):
37760         * DumpRenderTree/chromium/LayoutTestController.cpp:
37761         (LayoutTestController::LayoutTestController):
37762         (LayoutTestController::layerTreeAsText):
37763         * DumpRenderTree/chromium/LayoutTestController.h:
37764         * DumpRenderTree/chromium/TestShell.cpp:
37765         (TestShell::runFileTest):
37766         (TestShell::dump):
37767         * DumpRenderTree/chromium/TestShell.h:
37768         (TestParams::TestParams):
37769
37770 2011-03-14  Daniel Sievers  <sievers@google.com>
37771
37772         Reviewed by Simon Fraser.
37773
37774         [Chromium] Make RenderAsTextBehavior and LayerTreeAsTextBehavior tweakable from the DumpRenderTree commandline
37775         https://bugs.webkit.org/show_bug.cgi?id=56139
37776
37777         * DumpRenderTree/chromium/DumpRenderTree.cpp:
37778         (main):
37779         * DumpRenderTree/chromium/LayoutTestController.cpp:
37780         (LayoutTestController::LayoutTestController):
37781         (LayoutTestController::layerTreeAsText):
37782         * DumpRenderTree/chromium/LayoutTestController.h:
37783         (LayoutTestController::setShowDebugLayerTree):
37784         * DumpRenderTree/chromium/TestShell.cpp:
37785         (TestShell::runFileTest):
37786         (TestShell::dump):
37787         * DumpRenderTree/chromium/TestShell.h:
37788         (TestParams::TestParams):
37789
37790 2011-03-14  Balazs Kelemen  <kbalazs@webkit.org>
37791
37792         Reviewed by Adam Roben.
37793
37794         [Qt][WK2]Unbreak InjectedBundle on Qt
37795         https://bugs.webkit.org/show_bug.cgi?id=54109
37796
37797         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
37798         (WTR::InjectedBundlePage::willSendRequestForFrame): Use new WebKit2
37799         API's instead of KURL to analyze the URL.
37800         * WebKitTestRunner/StringFunctions.h:
37801         (WTR::adoptWK): Moved from InjectedBundlePage.cpp to make it available globally
37802         in the WebKitTestRunner project.
37803         (WTR::toWK): Use adoptWK.
37804         (WTR::operator<<): Added helper for printing a WKURL.
37805         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj: Do not link
37806         against WebCore.
37807
37808 2011-03-14  Balazs Kelemen  <kbalazs@webkit.org>
37809
37810         Reviewed by Kenneth Rohde Christiansen.
37811
37812         check-webkit-style falsely should not complain about WebKit2 API headers in the Shared directory
37813         https://bugs.webkit.org/show_bug.cgi?id=56321
37814
37815         * Scripts/webkitpy/style/checker.py: Add the patch "Source/WebKit2/Shared/API/c"
37816         to the list of exceptions for WebKit2 C API headers.
37817
37818 2011-03-14  Adam Roben  <aroben@apple.com>
37819
37820         Show the list of recent leaky builds in reverse chronological order
37821
37822         Previously they were being shown in a deterministic but surprising order based on
37823         lexicographically sorting the build offsets (-1, -10, -2, -3...).
37824
37825         Fixes <http://webkit.org/b/56342> Leaks Viewer shows recent builds in a strange order
37826
37827         Reviewed by Darin Adler.
37828
37829         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js:
37830         (RecentBuildsLoader.prototype.start): Sort the builds by revision before passing them to our
37831         callback.
37832
37833 2011-03-14  Adam Roben  <aroben@apple.com>
37834
37835         Show a list of up to 10 recent leaky builds when Leaks Viewer loads
37836
37837         Fixes <http://webkit.org/b/56043> Leaks Viewer should present a list of recent builds to
37838         analyze
37839
37840         Reviewed by David Kilzer and Joe Pecoraro.
37841
37842         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
37843         (#url-prompt): Made this a bit taller to make room for the list of builds.
37844
37845         (#recent-builds-loading-indicator):
37846         (#recent-builds-list):
37847         Added styles for the build list and loading indicator.
37848
37849         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
37850         (LeaksViewer.loaded): Pass the URL from the query through decodeURIComponent before loading
37851         it. This will unescape it.
37852         (LeaksViewer.urlPromptButtonClicked): Moved code from here to _urlChosenFromPrompt.
37853         (LeaksViewer._didLoadRecentBuilds): Added. Stores the builds and updates the prompt.
37854         (LeaksViewer._displayURLPrompt): Start loading the recent leaky builds.
37855         (LeaksViewer._updateURLPrompt): Create a list of the recent leaky builds and add them to the
37856         prompt UI.
37857         (LeaksViewer._urlChosenFromPrompt): Added. Code came from urlPromptButtonClicked. This is
37858         now also used when a recent build is clicked.
37859
37860         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/RecentBuildsLoader.js: Added.
37861         (RecentBuildsLoader): Store the callback.
37862         (RecentBuildsLoader.prototype.start): Load the N most recent builds from the specified
37863         builder. Find the revision number, leak count, and results URL for each one. Pass the
37864         resulting info to our callback.
37865
37866         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js:
37867         (range): Added this helper function to return an array containing the integers [0, n).
37868
37869         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html: Added
37870         RecentBuildsLoader.js and added text to the prompt UI to show the recent builds.
37871
37872 2011-03-14  Mark Rowe  <mrowe@apple.com>
37873
37874         Reviewed by Adam Roben.
37875
37876         Fix two memory leaks within LayoutTestController's storage support.
37877
37878         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
37879         (LayoutTestController::originsWithLocalStorage): Don't leak the JSStringRef that is returned
37880         by JSStringCreateWithCFString.
37881         (LayoutTestController::deleteLocalStorageForOrigin): Don't leak the WebSecurityOrigin that we
37882         allocate.
37883
37884 2011-03-14  Mark Rowe  <mrowe@apple.com>
37885
37886         Reviewed by Oliver Hunt.
37887
37888         <http://webkit.org/b/56304> REGRESSION(r80892): 100,000+ leaks seen on the build bot
37889
37890         * DumpRenderTree/LayoutTestController.cpp:
37891         (LayoutTestController::staticFunctions): Remove duplicate entries.
37892
37893 2011-03-14  Alice Liu  <alice.liu@apple.com>
37894
37895         Reviewed by Dan Bernstein.
37896
37897         DumpRenderTree Perl Support fails to build with ARCHS="i386 x86_64"
37898
37899         * DumpRenderTree/mac/PerlSupport/Makefile:
37900         Add whitespace when joining "-arch" with the architectures passed in as arguments.
37901
37902 2011-03-14  Daniel Bates  <dbates@rim.com>
37903
37904         https://bugs.webkit.org/show_bug.cgi?id=55438
37905         Add double quotes around value of -DCMAKE_WINCE_SDK that I inadvertently did not add
37906         when landing changeset 81041 <http://trac.webkit.org/changeset/81041>.
37907
37908         * Scripts/build-webkit:
37909
37910 2011-03-14  Daniel Bates  <dbates@rim.com>
37911
37912         Reviewed by David Kilzer.
37913
37914         Cleanup: Separate port-specific implementation details from webkitdirs::buildCMakeProject()
37915         https://bugs.webkit.org/show_bug.cgi?id=55438
37916
37917         Separate out the EFL- and WinCE-specific logic from the port-independent logic in
37918         webkitdirs::buildCMakeProject(). Also, remove redundant code for generating the
37919         CMake arguments for feature defines.
37920
37921         * Scripts/build-webkit:
37922           - Added cMakeArgsFromFeatures() to convert the @features array to CMake arguments.
37923           - Modified EFL and WinCE-specific building logic to call buildCMakeProjectOrExit().
37924           - Sorted forward declarations.
37925           - Initialize $makeArgs with the empty string so as to simplify its use in string operations;
37926             Modified call sites as needed.
37927         * Scripts/webkitdirs.pm:
37928           - Added cleanCMakeGeneratedProject() to clean a CMake build.
37929           - Added buildCMakeGeneratedProject() to build using generated build system.
37930           - Added buildCMakeProjectOrExit() which is a facade for building a CMake project.
37931           - Added generateBuildSystemFromCMakeProject() to generate the build system from
37932             a CMake project.
37933           - Removed buildCMakeProject(). This functionality is in buildCMakeProjectOrExit().
37934
37935 2011-03-14  John Knottenbelt  <jknotten@chromium.org>
37936
37937         Reviewed by Steve Block.
37938
37939         Detach Geolocation from Frame when Page destroyed.
37940         https://bugs.webkit.org/show_bug.cgi?id=52877
37941
37942         Extend the layout test controller to expose the number of pending
37943         geolocation requests, so that we can test that the requests have
37944         been cancelled on page close.
37945
37946         * DumpRenderTree/LayoutTestController.cpp:
37947         (numberOfPendingGeolocationPermissionRequestsCallback):
37948         (LayoutTestController::staticFunctions):
37949         * DumpRenderTree/LayoutTestController.h:
37950         * DumpRenderTree/chromium/LayoutTestController.cpp:
37951         (LayoutTestController::LayoutTestController):
37952         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
37953         * DumpRenderTree/chromium/LayoutTestController.h:
37954         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
37955         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
37956         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
37957         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
37958         * DumpRenderTree/mac/UIDelegate.h:
37959         * DumpRenderTree/mac/UIDelegate.mm:
37960         (-[UIDelegate numberOfPendingGeolocationPermissionRequests]):
37961         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
37962         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
37963         * DumpRenderTree/qt/LayoutTestControllerQt.h:
37964         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
37965         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
37966         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
37967         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
37968
37969
37970 2011-03-14  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
37971
37972         Unreviewed. Modify my email address in committers.py.
37973
37974         * Scripts/webkitpy/common/config/committers.py:
37975
37976 2011-03-13  Joe Wild  <joseph.wild@nokia.com>
37977
37978         Reviewed by Kenneth Rohde Christiansen.
37979
37980         [Qt] QtLauncher does not load the same set of fonts as the DRT
37981         https://bugs.webkit.org/show_bug.cgi?id=34959
37982
37983         This patch adds the option "-use-test-fonts" to the QtTestBrowser.
37984         When this option is used the webkit fonts are loaded the same
37985         as they are in DumpRenderTree.  This option can be used on
37986         QtTestBrowser and run-launcher.  It can only be used
37987         on Linux systems with FcInit and is configured as such.
37988
37989         * QtTestBrowser/launcherwindow.h:
37990         (WindowOptions::WindowOptions):
37991         * QtTestBrowser/main.cpp:
37992         (initWebKitTestFonts):
37993         (launcherMain):
37994         (LauncherApplication::handleUserOptions):
37995
37996 2011-03-12  Dan Bernstein  <mitz@apple.com>
37997
37998         Reviewed by Mark Rowe.
37999
38000         <rdar://problem/8995822> Build DumpRenderTree perl support only for the architectures
38001         DumpRenderTree is being built for.
38002
38003         * DumpRenderTree/mac/PerlSupport/Makefile:
38004
38005 2011-03-12  Dan Bernstein  <mitz@apple.com>
38006
38007         Suggested by Mark Rowe.
38008
38009         * DumpRenderTree/mac/Configurations/Base.xcconfig: Leave C++ exceptions enabled
38010         as they were before r80915.
38011
38012 2011-03-12  Anders Carlsson  <andersca@apple.com>
38013
38014         Reviewed by Sam Weinig.
38015
38016         Should be able to download a PDF to the Downloads folder
38017         https://bugs.webkit.org/show_bug.cgi?id=56256
38018
38019         * MiniBrowser/mac/BrowserWindowController.m:
38020         (-[BrowserWindowController awakeFromNib]):
38021         * WebKitTestRunner/TestController.cpp:
38022         (WTR::TestController::createOtherPage):
38023         (WTR::TestController::initialize):
38024         Update for API changes.
38025
38026 2011-03-11  Anton D'Auria  <adauria@apple.com>
38027
38028         Reviewed and landed by Brady Eidson.
38029
38030         https://bugs.webkit.org/show_bug.cgi?id=51878
38031         Add WebKit1 API to view and delete local storage
38032
38033         Added tests that write to LocalStorage, delete one origin, get list of origins with local storage, delete all origins.
38034
38035         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
38036         * DumpRenderTree/LayoutTestController.cpp:
38037         (syncLocalStorageCallback):
38038         (observeStorageTrackerNotificationsCallback):
38039         (deleteAllLocalStorageCallback):
38040         (deleteLocalStorageForOriginCallback):
38041         (originsWithLocalStorageCallback):
38042         (LayoutTestController::staticFunctions):
38043         * DumpRenderTree/LayoutTestController.h:
38044         * DumpRenderTree/StorageTrackerDelegate.h: Added.
38045         * DumpRenderTree/StorageTrackerDelegate.mm: Added.
38046         (-[StorageTrackerDelegate init]):
38047         (-[StorageTrackerDelegate logNotifications:controller:]):
38048         (-[StorageTrackerDelegate originModified:]):
38049         (-[StorageTrackerDelegate dealloc]):
38050         (-[StorageTrackerDelegate setControllerToNotifyDone:]):
38051         * DumpRenderTree/chromium/LayoutTestController.cpp:
38052         (LayoutTestController::LayoutTestController):
38053         (LayoutTestController::deleteAllLocalStorage):
38054         (LayoutTestController::originsWithLocalStorage):
38055         (LayoutTestController::deleteLocalStorageForOrigin):
38056         (observeStorageTrackerNotifications):
38057         (syncLocalStorage):
38058         * DumpRenderTree/chromium/LayoutTestController.h:
38059         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
38060         (LayoutTestController::originsWithLocalStorage):
38061         (LayoutTestController::deleteAllLocalStorage):
38062         (LayoutTestController::deleteLocalStorageForOrigin):
38063         (LayoutTestController::observeStorageTrackerNotifications):
38064         (LayoutTestController::syncLocalStorage):
38065         * DumpRenderTree/mac/DumpRenderTree.mm:
38066         (resetDefaultsToConsistentValues):
38067         (allocateGlobalControllers):
38068         (releaseGlobalControllers):
38069         * DumpRenderTree/mac/DumpRenderTreeMac.h:
38070         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
38071         (LayoutTestController::syncLocalStorage):
38072         (LayoutTestController::observeStorageTrackerNotifications):
38073         (LayoutTestController::deleteAllLocalStorage):
38074         (LayoutTestController::originsWithLocalStorage):
38075         (LayoutTestController::deleteLocalStorageForOrigin):
38076         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
38077         (LayoutTestController::originsWithLocalStorage):
38078         (LayoutTestController::deleteAllLocalStorage):
38079         (LayoutTestController::deleteLocalStorageForOrigin):
38080         (LayoutTestController::observeStorageTrackerNotifications):
38081         (LayoutTestController::syncLocalStorage):
38082         * DumpRenderTree/qt/LayoutTestControllerQt.h:
38083         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
38084         (LayoutTestController::clearAllApplicationCaches):
38085         (LayoutTestController::syncLocalStorage):
38086         (LayoutTestController::observeStorageTrackerNotifications):
38087         (LayoutTestController::clearAllDatabases):
38088         (LayoutTestController::deleteAllLocalStorage):
38089         (LayoutTestController::originsWithLocalStorage):
38090         (LayoutTestController::deleteLocalStorageForOrigin):
38091         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
38092         (LayoutTestController::syncLocalStorage):
38093         (LayoutTestController::observeStorageTrackerNotifications):
38094         (LayoutTestController::clearAllDatabases):
38095         (LayoutTestController::deleteAllLocalStorage):
38096         (LayoutTestController::originsWithLocalStorage):
38097         (LayoutTestController::deleteLocalStorageForOrigin):
38098
38099 2011-03-11  Jessie Berlin  <jberlin@apple.com>
38100
38101         Reviewed by Adam Roben.
38102
38103         Crash calling WebContext::clearResourceCaches(InMemoryResourceCachesOnly) before the Web
38104         Process has finished launching.
38105         https://bugs.webkit.org/show_bug.cgi?id=56208
38106
38107         * TestWebKitAPI/Tests/WebKit2/SendingMessagesToTheWebProcessBeforeItIsValid.cpp: Added.
38108         (TestWebKitAPI::TEST):
38109         Test that calling WKContextClearResourceCaches works with both options when called before
38110         the Web Process is valid.
38111
38112         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
38113         Add the new test.
38114         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
38115         Ditto.
38116
38117 2011-03-11  Adam Roben  <aroben@apple.com>
38118
38119         Make it possible to view all leaks from a build at once in Leaks Viewer
38120
38121         You can now give Leaks Viewer the URL of a build results page (e.g.,
38122         <http://build.webkit.org/results/SnowLeopard%20Intel%20Leaks/r80847%20(15493)/>), and it
38123         will load all the leaks files from that build.
38124
38125         Fixes <http://webkit.org/b/56030> Leaks Viewer: Would like to be able to look at all leaks
38126         files from a particular build at once, rather than one at a time
38127
38128         Reviewed by David Kilzer.
38129
38130         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksLoader.js: Added.
38131         (LeaksLoader): Initialize our properties.
38132         (LeaksLoader.prototype.start): If the URL ends in .txt, assume it's a single leaks file and
38133         load it. Otherwise assume it's a build results page URL and start fetching the leaks files
38134         from that page.
38135         (LeaksLoader.prototype._loadLeaksFiles): Tell our client how many leaks files we're loading
38136         so it can provide feedback. Then load each one and pass it to the client.
38137         (LeaksLoader.prototype._loadLeaksFromResultsPage): Load the HTML of the results page, parse
38138         it into a DOM, pull out all the links to leaks files, and load each one.
38139
38140         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParser.js: Added.
38141         (LeaksParser): Initialize our worker.
38142         (LeaksParser.prototype.addLeaksFile): Tell the worker about the leaks file's contents.
38143
38144         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksParserWorker.js: Added.
38145         This code mostly came from Worker.js.
38146         (LeaksParserWorker): Initialize our profile.
38147         (LeaksParserWorker.prototype.addLeaksFile): Parse the leaks file and incorporate it into our
38148         profile.
38149         (LeaksParserWorker.prototype._parseLeaks):
38150         (LeaksParserWorker.prototype._createNode):
38151         (LeaksParserWorker.prototype._incorporateLeaks):
38152         Moved these functions here from LeaksViewer. _incorporateLeaks is essentially just a renamed
38153         version of createProfile which adds to this.profile instead of making a new profile each
38154         time.
38155         (onmessage): Parse the file and send back the new profile.
38156
38157         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
38158         (#loading-indicator):
38159         (#spinner):
38160         (#loading-indicator-label):
38161         Tweaked styles to accomodate a longer label.
38162
38163         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
38164         (LeaksViewer.loaded): Set up a loader and parser.
38165
38166         (LeaksViewer.get filesLeftToParse):
38167         (LeaksViewer.set filesLeftToParse):
38168         Added these simple accessors.
38169
38170         (LeaksViewer._didCountLeaksFiles): Callback from LeaksLoader. Stores the count.
38171         (LeaksViewer._didLoadLeaksFile): Callback from LeaksLoader. Passes the file contents off to
38172         the parser.
38173         (LeaksViewer._didParseLeaksFile): Callback from LeaksParser. If all files have been parsed,
38174         tell the ProfilerAgent and mark that we're done loading. (Code came from the old
38175         _loadLeaksFromURL function).
38176         (LeaksViewer._loadLeaksFromURL): Now just calls through to the loader.
38177         (LeaksViewer._loadingIndicatorText): Added. Returns the text that should show up in the
38178         loading indicator, including the number of files being loaded.
38179         (LeaksViewer._loadingStatusChanged): Update the loading indicator's label, too.
38180         (LeaksViewer._updateLoadingIndicatorLabel): Added. Just updates the label!
38181         (LeaksViewer._updateTitle): Moved code to compute the "Loading" text to
38182         _loadingIndicatorText.
38183
38184         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Utilities.js: Added.
38185         (getResource): Moved here from LeaksViewer.js.
38186
38187         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Worker.js: Subsumed by
38188         LeaksParserWorker.js.
38189
38190         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html: Added new JS
38191         files, added a #loading-indicator-label element, and tweaked the prompt wording.
38192
38193 2011-03-11  Adam Roben  <aroben@apple.com>
38194
38195         Don't trigger a build when build.webkit.org's HTML files are modified
38196
38197         Fixes <http://webkit.org/b/56190> Windows bots shouldn't have built r80848/r80851, but did
38198
38199         Reviewed by Dan Bernstein.
38200
38201         * Scripts/webkitpy/common/config/build.py:
38202         (_should_file_trigger_build): Added build.webkit.org's public_html directory to the list of
38203         directories that shouldn't trigger builds. Re-sorted the list.
38204
38205         * Scripts/webkitpy/common/config/build_unittest.py:
38206         (ShouldBuildTest): Added a test to cover the above change.
38207
38208 2011-03-11  Adam Roben  <aroben@apple.com>
38209
38210         Roll out r80848
38211
38212         It was accidentally committed.
38213
38214         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
38215         (LeaksViewer._loadLeaksFromURL):
38216
38217 2011-03-11  Tony Chang  <tony@chromium.org>
38218
38219         Reviewed by Ojan Vafai.
38220
38221         kill hung lighttpd processes on the chromium win bot
38222         https://bugs.webkit.org/show_bug.cgi?id=56152
38223
38224         Sometimes this process doesn't exit and it holds on to
38225         a log file that we can't delete or copy, causing the tests
38226         to go red.
38227
38228         * BuildSlaveSupport/win/kill-old-processes:
38229
38230 2011-03-11  Ilya Sherman  <isherman@chromium.org>
38231
38232         Reviewed by Kent Tamura.
38233
38234         [Chromium] missing HTMLInputElement::setValueForUser() to tests
38235         https://bugs.webkit.org/show_bug.cgi?id=55854
38236
38237         * DumpRenderTree/chromium/LayoutTestController.cpp:
38238         (LayoutTestController::LayoutTestController):
38239         (LayoutTestController::setValueForUser): Added.
38240         * DumpRenderTree/chromium/LayoutTestController.h:
38241
38242 2011-03-10  Peter Kasting  <pkasting@google.com>
38243
38244         Reviewed by Dimitri Glazkov.
38245
38246         Clean up some gross code in TestShellWin.cpp.  No functional change.
38247         https://bugs.webkit.org/show_bug.cgi?id=56048
38248
38249         * DumpRenderTree/chromium/TestShellWin.cpp:
38250         (checkLayoutTestSystemDependencies):
38251
38252 2011-03-10  Keith Kyzivat  <keith.kyzivat@nokia.com>
38253
38254         Reviewed by Laszlo Gombos.
38255
38256         [Qt] Buildfix for platforms where X11 is not available (e.g. Symbian)
38257
38258         * WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
38259         (WTR::activateFonts): Guard X11-specific code with Q_WS_X11.
38260
38261 2011-03-10  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
38262
38263         Reviewed by Kenneth Rohde Christiansen.
38264
38265         [EFL] Make single backing store the default for EWebLauncher
38266         https://bugs.webkit.org/show_bug.cgi?id=55753
38267
38268         Set single backing store the default for EWebLauncher.
38269
38270         * EWebLauncher/main.c:
38271         (main):
38272
38273 2011-03-10  Adam Roben  <aroben@apple.com>
38274
38275         Ignore errors from files being held open when creating a test result archive
38276
38277         Fixes <http://webkit.org/b/55581> Test results don't get uploaded if test-result-archive
38278         fails to delete the results directory
38279
38280         Reviewed by Mark Rowe.
38281
38282         * BuildSlaveSupport/test-result-archive:
38283         (archiveTestResults): Ignore errno 2 when deleting the results directory, since it can be
38284         thrown when a process is holding a file open.
38285
38286 2011-03-10  Sheriff Bot  <webkit.review.bot@gmail.com>
38287
38288         Unreviewed, rolling out r80701.
38289         http://trac.webkit.org/changeset/80701
38290         https://bugs.webkit.org/show_bug.cgi?id=56126
38291
38292         fast/dom/Geolocation/window-close-crash.html fails on Snow
38293         Leopard release builds (Requested by mihaip on #webkit).
38294
38295         * DumpRenderTree/LayoutTestController.cpp:
38296         (LayoutTestController::staticFunctions):
38297         * DumpRenderTree/LayoutTestController.h:
38298         * DumpRenderTree/chromium/LayoutTestController.cpp:
38299         (LayoutTestController::LayoutTestController):
38300         * DumpRenderTree/chromium/LayoutTestController.h:
38301         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
38302         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
38303         * DumpRenderTree/mac/UIDelegate.h:
38304         * DumpRenderTree/mac/UIDelegate.mm:
38305         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
38306         * DumpRenderTree/qt/LayoutTestControllerQt.h:
38307         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
38308         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
38309
38310 2011-03-10  Philippe Normand  <pnormand@igalia.com>
38311
38312         Reviewed by Gustavo Noronha Silva.
38313
38314         [GTK] run-gtk-tests doesn't fail if a test crashes
38315         https://bugs.webkit.org/show_bug.cgi?id=56089
38316
38317         * Scripts/run-gtk-tests: gtester -k exits with a 0 exit status
38318         even if a test crashes. So run the tests individually and if a
38319         test fails remember its exit code so run-gtk-test fails too, if
38320         required.
38321
38322 2011-01-26  John Knottenbelt  <jknotten@chromium.org>
38323
38324         Reviewed by Dmitry Titov.
38325
38326         Detach Geolocation from Frame when Page destroyed.
38327         https://bugs.webkit.org/show_bug.cgi?id=52877
38328
38329         Extend the layout test controller to expose the number of pending
38330         geolocation requests, so that we can test that the requests have
38331         been cancelled on page close.
38332
38333         * DumpRenderTree/LayoutTestController.cpp:
38334         (numberOfPendingGeolocationPermissionRequestsCallback):
38335         (LayoutTestController::staticFunctions):
38336         * DumpRenderTree/LayoutTestController.h:
38337         * DumpRenderTree/chromium/LayoutTestController.cpp:
38338         (LayoutTestController::LayoutTestController):
38339         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
38340         * DumpRenderTree/chromium/LayoutTestController.h:
38341         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
38342         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
38343         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
38344         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
38345         * DumpRenderTree/mac/UIDelegate.h:
38346         * DumpRenderTree/mac/UIDelegate.mm:
38347         (-[UIDelegate numberOfPendingGeolocationPermissionRequests]):
38348         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
38349         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
38350         * DumpRenderTree/qt/LayoutTestControllerQt.h:
38351         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
38352         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
38353         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
38354         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
38355
38356
38357 2011-03-09  Adam Roben  <aroben@apple.com>
38358
38359         Hide Leaks Viewer's URL prompt by default
38360
38361         We will show it as needed. I think I left it visible by default by accident after doing some
38362         testing.
38363
38364         Fixes <http://webkit.org/b/56031> Leaks Viewer: URL prompt overlay shouldn't appear when a
38365         URL is passed as a query parameter, but does
38366
38367         Reviewed by Simon Fraser.
38368
38369         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html:
38370
38371 2011-03-09  Adam Roben  <aroben@apple.com>
38372
38373         Don't use Element.classList or box-shadow in Leaks Viewer
38374
38375         They aren't supported by Safari 5. Luckily we have substitutes for them!
38376
38377         Fixes <http://webkit.org/b/56035> Leaks viewer doesn't work in Safari 5
38378
38379         Reviewed by Simon Fraser.
38380
38381         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css:
38382         (#url-prompt): Use -webkit-box-shadow instead of box-shadow.
38383
38384         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js:
38385         (LeaksViewer.urlPromptButtonClicked):
38386         (LeaksViewer._displayURLPrompt):
38387         (LeaksViewer._setLoadingIndicatorHidden):
38388         Use add/removeStyleClass (from utilities.js) instead of Element.classList.
38389
38390 2011-03-09  Adam Roben  <aroben@apple.com>
38391
38392         Add a link to the Leaks Viewer tool on the build.webkit.org homepage
38393
38394         Rubber-stamped by David Kilzer.
38395
38396         * BuildSlaveSupport/build.webkit.org-config/templates/root.html:
38397
38398 2011-03-09  Adam Roben  <aroben@apple.com>
38399
38400         Add a Leaks Viewer tool to build.webkit.org
38401
38402         This tool parses leaks from the files generated by the leaks bot and displays them in a
38403         view similar to the Web Inspector's Profiles pane (in fact, it uses the Web Inspector's
38404         code).
38405
38406         The URL of the leaks file can be entered in a prompt when the tools is first shown, or can
38407         be passed to the tool via the "url" query parameter.
38408
38409         Only one leaks file can be viewed at a time for now. Only the number of leaks, as opposed to
38410         the number of leaked bytes, is shown. And this tool has made apparent some bugs in the Web
38411         Inspector's Heavy profile view. But it's still pretty useful as-is.
38412
38413         Fixes <http://webkit.org/b/56028> Would like a Shark-like tool for viewing leaks from leak bots
38414
38415         Reviewed by Darin Adler.
38416
38417         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.css: Added.
38418         (#url-prompt-container): This is the dark overlay that is shown behind the URL prompt.
38419         (#url-prompt): This is the prompt itself.
38420
38421         (#loading-indicator):
38422         (#loading-indicator > img):
38423         These show a spinner in the bottom-right when we're loading a leaks file.
38424
38425         (.percent-time-status-bar-item): Hide the Percent button since we never want to show leaks
38426         as percentages.
38427
38428         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/LeaksViewer.js: Added.
38429         (LeaksViewer.loaded): Initialize ourselves and start loading the leaks file (or prompt for a
38430         URL if one wasn't given to us as a query parameter).
38431
38432         (LeaksViewer.get loading):
38433         (LeaksViewer.set loading):
38434         (LeaksViewer.get url):
38435         (LeaksViewer.set url):
38436         Pretty simple accessors.
38437
38438         (LeaksViewer.urlPromptButtonClicked): Start loading the URL from the prompt and hide the
38439         prompt.
38440         (LeaksViewer._displayURLPrompt): Show the prompt and focus the URL input.
38441         (LeaksViewer._loadLeaksFromURL): Start fetching the leaks file. When we've retrieved the
38442         file, hand it off to a Web Worker that will parse it into a fake "profile", then hand off
38443         the profile to the Web Inspector code once it's been created.
38444         (LeaksViewer._loadingStatusChanged): Update our loading indicator and title.
38445         (LeaksViewer._setLoadingIndicatorHidden): Simply hide or show the indicator.
38446         (LeaksViewer._updateTitle): Set our title based on our URL and loading status.
38447         (getResource): Helper function to fetch a resource via XHR and pass it to a callback.
38448
38449         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/WebInspectorShims.js: Added.
38450         (WebInspector.UIString): Just a simple implementation since this tool isn't localized.
38451         (ProfilerAgent.getProfile): Store the callback.
38452         (ProfilerAgent.profileReady): Call the callback.
38453         (monkeyPatchInspectorObjects): Fix up WebInspector.ProfileDataGridNode to show the self leak
38454         count and total leak counts as real values, not percentages or milliseconds.
38455
38456         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/Worker.js: Added.
38457         (onmessage): Create a profile based on the leaks file and pass it back to our caller.
38458         (parseLeaks): Extract call stacks out of a leaks file.
38459         (createNode): Helper function to create a new empty profile node.
38460         (createProfile): Take the leak stacks and turn them into a fake "profile" for consumption by
38461         WebInspector.ProfileView.
38462
38463         * BuildSlaveSupport/build.webkit.org-config/public_html/LeaksViewer/index.html: Added.
38464
38465 2011-03-09  Dan Bernstein  <mitz@apple.com>
38466
38467         Reviewed by Adam Roben.
38468
38469         Combine multiple --ignore-tests values
38470         https://bugs.webkit.org/show_bug.cgi?id=56016
38471
38472         * Scripts/old-run-webkit-tests:
38473
38474 2011-03-08  Kenji Imasaki  <imasaki@chromium.org>
38475
38476         Reviewed by David Levin.
38477
38478         Adding myself to the committer file.
38479
38480         * Scripts/webkitpy/common/config/committers.py:
38481
38482 2011-03-08  Jessie Berlin  <jberlin@apple.com>
38483
38484         Rubber-stamped by Adam Roben.
38485
38486         Fix two leaks I introduced into the API tests.
38487
38488         * TestWebKitAPI/Tests/WebKit2/win/DoNotCopyANullCFURLResponse.cpp:
38489         (TestWebKitAPI::TEST):
38490
38491 2011-03-08  Jessie Berlin  <jberlin@apple.com>
38492
38493         Reviewed by Anders Carlsson.
38494
38495         Crash in CFNetwork visiting google.com
38496         https://bugs.webkit.org/show_bug.cgi?id=55958
38497
38498         Add a test for not trying to copy a null CFURLResponse.
38499
38500         * TestWebKitAPI/Tests/WebKit2/win/DoNotCopyANullCFURLResponse.cpp: Added.
38501         (TestWebKitAPI::TEST):
38502         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
38503
38504 2011-03-08  Hayato Ito  <hayato@chromium.org>
38505
38506         Reviewed by Ojan Vafai.
38507
38508         [NRWT] Renamed a FailureWithType class to a ComparisonTestFailure class.
38509
38510         FailureWithType is a confusing naming because we have removed test_type/* classes in r79840.
38511
38512         https://bugs.webkit.org/show_bug.cgi?id=55450
38513
38514         * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
38515
38516 2011-03-08  Carlos Garcia Campos  <cgarcia@igalia.com>
38517
38518         Reviewed by Martin Robinson.
38519
38520         [GTK] Add layoutTestController.setValueForUser() after r80412
38521         https://bugs.webkit.org/show_bug.cgi?id=55862
38522
38523         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
38524         (LayoutTestController::setValueForUser): Implement setValueForUser()
38525
38526 2011-03-08  Patrick Gansterer  <paroga@webkit.org>
38527
38528         Reviewed by Adam Barth.
38529
38530         Add script_shell_command to WebKitPort
38531         https://bugs.webkit.org/show_bug.cgi?id=55925
38532
38533         Add a central place where we generate the shell command from the script name.
38534         This function will be used later to detect the correct script interpreter on Win32.
38535
38536         * Scripts/webkitpy/common/config/ports.py:
38537         * Scripts/webkitpy/layout_tests/port/base.py:
38538         * Scripts/webkitpy/tool/steps/abstractstep.py:
38539         * Scripts/webkitpy/tool/steps/preparechangelog.py:
38540
38541 2011-03-07  Ojan Vafai  <ojan@chromium.org>
38542
38543         Reviewed by Adam Barth.
38544
38545         add webkit-patch attach-to-bug for posting to bugzilla from the commandline
38546         https://bugs.webkit.org/show_bug.cgi?id=55749
38547
38548         * Scripts/webkitpy/tool/commands/upload.py:
38549         * Scripts/webkitpy/tool/commands/upload_unittest.py:
38550         * Scripts/webkitpy/tool/steps/__init__.py:
38551         * Scripts/webkitpy/tool/steps/attachtobug.py: Added.
38552         * Scripts/webkitpy/tool/steps/options.py:
38553
38554 2011-03-07  Daniel Cheng  <dcheng@chromium.org>
38555
38556         Reviewed by David Levin.
38557
38558         Add support for DataTransferItems
38559         https://bugs.webkit.org/show_bug.cgi?id=55115
38560
38561         Enable data transfer items in DRT.
38562
38563         * DumpRenderTree/chromium/TestShell.cpp:
38564         (TestShell::TestShell):
38565
38566 2011-03-06  Ojan Vafai  <ojan@chromium.org>
38567
38568         Reviewed by Tony Chang.
38569
38570         [test dashboard] pull treemap files
38571         https://bugs.webkit.org/show_bug.cgi?id=55852
38572
38573         * TestResultServer/handlers/dashboardhandler.py:
38574
38575 2011-03-03  Hayato Ito  <hayato@chromium.org>
38576
38577         Reviewed by Mihai Parparita.
38578
38579         [NRWT] Clear output image file used by DumpRenderTree to make sure the
38580         previous image should not be used in the ChromiumDriver.
38581
38582         Also make sure that output image from the ChromiumDriver should be ''
38583         (empty string) if a test crashes for consistency with the WebKitDriver
38584         implementation.
38585
38586         https://bugs.webkit.org/show_bug.cgi?id=55746
38587
38588         * Scripts/webkitpy/layout_tests/port/chromium.py:
38589
38590 2011-03-07  Zan Dobersek  <zandobersek@gmail.com>
38591
38592         Reviewed by Martin Robinson.
38593
38594         [Gtk] fast/js/navigator-language.html fails locally
38595         https://bugs.webkit.org/show_bug.cgi?id=55880
38596
38597         Add LANG variable to the clean environment when running
38598         old-run-webkit-tests script.
38599
38600         * Scripts/old-run-webkit-tests:
38601
38602 2011-03-07  Tony Chang  <tony@chromium.org>
38603
38604         Reviewed by Mihai Parparita.
38605
38606         [Chromium] Find .checksum files without .pngs and vice-versa
38607         https://bugs.webkit.org/show_bug.cgi?id=55236
38608
38609         * Scripts/find-mismatched-layout-test-results: Added. A script for finding
38610             mismatched results.
38611
38612 2011-03-07  Chris Fleizach  <cfleizach@apple.com>
38613
38614         Reviewed by Beth Dakin.
38615
38616         AX: WK1 needs to use the ScrollView attachment for AXScrollArea, WK2 should not use this element.
38617         https://bugs.webkit.org/show_bug.cgi?id=55706
38618
38619         The rootElement in DRT should point to the scroll view. In WK1, that will be the parent of the rootObject.
38620
38621         * DumpRenderTree/mac/AccessibilityControllerMac.mm:
38622         (AccessibilityController::focusedElement):
38623            Remove FIXME comment about caching the focusedElement. In fact, we shouldn't cache this
38624            because it could change at any time.
38625         (AccessibilityController::rootElement):
38626
38627 2011-03-07  Sam Weinig  <sam@webkit.org>
38628
38629         Reviewed by Anders Carlsson.
38630
38631         Add unableToImplementPolicy callback for WebKit2 policy client
38632         <rdar://problem/9071902>
38633         https://bugs.webkit.org/show_bug.cgi?id=55884
38634
38635         * MiniBrowser/mac/BrowserWindowController.m:
38636         (-[BrowserWindowController awakeFromNib]):
38637         Update policy client initialization struct for new member.
38638
38639 2011-03-07  Sam Weinig  <sam@webkit.org>
38640
38641         Reviewed by Anders Carlsson.
38642
38643         Replace WebKit2's decidePolicyForMIMEType with decidePolicyForResponse
38644         https://bugs.webkit.org/show_bug.cgi?id=55827
38645
38646         * MiniBrowser/mac/BrowserWindowController.m:
38647         (decidePolicyForResponse):
38648         (-[BrowserWindowController awakeFromNib]):
38649         * TestWebKitAPI/Tests/WebKit2/PageLoadBasic.cpp:
38650         (TestWebKitAPI::decidePolicyForResponse):
38651         (TestWebKitAPI::TEST):
38652         Update tools for new name.
38653
38654 2011-03-06  Daniel Bates  <dbates@rim.com>
38655
38656         Reviewed by Eric Seidel and David Kilzer.
38657
38658         Fix misspelled word in build-webkit's checkForJavaSDK() and sort forward declarations
38659         https://bugs.webkit.org/show_bug.cgi?id=55503
38660
38661         Fix misspelled word "Dowloads" [sic] in the message printed to standard output
38662         in checkForJavaSDK().
38663
38664         Also, remove extraneous space in function prototype for unlinkZeroFiles.
38665
38666         * Scripts/build-webkit:
38667
38668 2011-03-05  Brian Weinstein  <bweinstein@apple.com>
38669
38670         Reviewed by Sam Weinig.
38671
38672         Fix Web Process crashes caused by https://bugs.webkit.org/show_bug.cgi?id=53919.
38673
38674         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
38675         (WTR::InjectedBundlePage::willSendRequestForFrame): Retain the WKURLRequestRef returned
38676             by willSendRequestForFrame. The API expects a retained reference to the URL request.
38677
38678 2011-03-05  Ilya Sherman  <isherman@chromium.org>
38679
38680         Reviewed by Darin Adler.
38681
38682         HTMLInputElement::setValue() should schedule change event when the element is focused
38683         In service of https://code.google.com/p/chromium/issues/detail?id=42716
38684         https://bugs.webkit.org/show_bug.cgi?id=53160
38685
38686         * DumpRenderTree/LayoutTestController.cpp:
38687         (setValueForUserCallback):
38688         (LayoutTestController::staticFunctions):
38689         * DumpRenderTree/LayoutTestController.h:
38690         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
38691         (LayoutTestController::setValueForUser): Stubbed out; need real implementation.
38692         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
38693         (LayoutTestController::setValueForUser): Expose HTMLInputElement::setValueForUser() to tests
38694         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
38695         (LayoutTestController::setValueForUser): Stubbed out; need real implementation.
38696         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
38697         (LayoutTestController::setValueForUser): Stubbed out; need real implementation.
38698         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
38699         (LayoutTestController::setValueForUser): Stubbed out; need real implementation.
38700
38701 2011-03-04  Mark Rowe  <mrowe@apple.com>
38702
38703         Reviewed by Alexey Proskuryakov.
38704
38705         <rdar://problem/9055252> DumpRenderTree is crashing with GuardMalloc
38706
38707         GuardMalloc does not expect the memory containing its malloc_zone_t to ever be read-only.
38708
38709         * DumpRenderTree/mac/CheckedMalloc.cpp:
38710         (protectionOfRegion): Retrieve the protection flags for the VM region containing the given address.
38711         (makeLargeMallocFailSilently): Save the initial protection flags, temporarily make the region writeable,
38712         and then restore the initial protection flags when we're done.
38713
38714 2011-03-04  Evan Martin  <evan@chromium.org>
38715
38716         Reviewed by Tony Chang.
38717
38718         [gdb] pretty-print KURL in gdb
38719         https://bugs.webkit.org/show_bug.cgi?id=55788
38720
38721         * gdb/webkit.py: extend the pretty-printers to print WTF::CString
38722         and KURLGooglePrivate.  As always, "p/r" can be used to sidestep
38723         pretty-printing.
38724
38725 2011-03-03  Timothy Hatcher  <timothy@apple.com>
38726
38727         Update the tests that use WKPageRunJavaScriptInMainFrame use the WKSerializedScriptValueRef.
38728
38729         Reviewed by Darin Adler.
38730
38731         * TestWebKitAPI/JavaScriptTest.cpp:
38732         (TestWebKitAPI::javaScriptCallback):
38733         * TestWebKitAPI/Tests/WebKit2/EvaluateJavaScript.cpp:
38734         (TestWebKitAPI::didRunJavaScript):
38735         * TestWebKitAPI/Tests/WebKit2/PageLoadDidChangeLocationWithinPageForFrame.cpp:
38736         (TestWebKitAPI::nullJavaScriptCallback):
38737         * TestWebKitAPI/Tests/WebKit2/PreventEmptyUserAgent.cpp:
38738         (TestWebKitAPI::didRunJavaScript):
38739
38740 2011-03-03  Timothy Hatcher  <timothy@apple.com>
38741
38742         Make run-api-tests show failed if a test fails.
38743
38744         https://webkit.org/b/55725
38745
38746         Reviewed by Sam Weinig.
38747
38748         * Scripts/run-api-tests:
38749         (runTest): Remove "my" from the nested $result assignment so the
38750         check outside the condition sees the real result.
38751
38752 2011-03-02  MORITA Hajime  <morrita@google.com>
38753
38754         Reviewed by Kent Tamura.
38755
38756         [Chromium][DRT] WebTask should use webkit_support::TaskAdaptor
38757         https://bugs.webkit.org/show_bug.cgi?id=55434
38758
38759         Changed WebTask to be a subclass of webkit_support::TaskAdaptor
38760         and passed the object webkit_support::PostDelayedTask directly.
38761
38762         Note that the change keeps postTask() as is because it has a
38763         semantics taht webkit_support::PostDelayedTask() has; It jumps to
38764         the main thread.
38765
38766         * DumpRenderTree/chromium/Task.cpp:
38767         (postDelayedTask):
38768         * DumpRenderTree/chromium/Task.h:
38769         (WebTask::Run):
38770
38771 2011-03-03  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
38772
38773         Reviewed by Kenneth Rohde Christiansen.
38774
38775         [QT] Implement mock client-based geolocation for layout testing
38776         https://bugs.webkit.org/show_bug.cgi?id=54334
38777
38778         Implement testing for Client-Based geolocation for QtWebkit.
38779
38780         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
38781         (WebCore::DumpRenderTree::DumpRenderTree):
38782         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
38783         (LayoutTestController::reset):
38784         (LayoutTestController::setGeolocationPermission):
38785         (LayoutTestController::setMockGeolocationError):
38786         (LayoutTestController::setMockGeolocationPosition):
38787
38788 2011-03-03  Sheriff Bot  <webkit.review.bot@gmail.com>
38789
38790         Unreviewed, rolling out r80237.
38791         http://trac.webkit.org/changeset/80237
38792         https://bugs.webkit.org/show_bug.cgi?id=55714
38793
38794         Windows bots aren't ready to be core (Requested by aroben on
38795         #webkit).
38796
38797         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
38798         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
38799
38800 2011-03-03  Mihai Parparita  <mihaip@chromium.org>
38801
38802         Reviewed by Tony Chang.
38803
38804         NRWT: AttributeError: TestRunner2 instance has no attribute '_cancel_workers'
38805         https://bugs.webkit.org/show_bug.cgi?id=55694
38806
38807         Fixes for interruptions in NRWT:
38808         - Make TestRunInterruptedException be pickleable correctly (the base
38809           Exception class defines a __reduce__ that does not include the reason)
38810         - Fix ordering of returned arguments from TestRunner2._run_tests
38811           (interrupted and keyboard_interrupted were reversed)
38812         - Fix cancel_workers callsites (was using old name).
38813         - In handle_exception re-raise actual exception instance that was thrown
38814
38815         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
38816         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
38817         * Scripts/webkitpy/layout_tests/layout_package/worker.py:
38818         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
38819
38820 2011-03-03  Qi Zhang  <qi.2.zhang@nokia.com>
38821
38822         Unreviewed.
38823
38824         Adding myself as a committer.
38825
38826         * Scripts/webkitpy/common/config/committers.py:
38827
38828 2011-03-03  Tony Gentilcore  <tonyg@chromium.org>
38829
38830         Reviewed by Mihai Parparita.
38831
38832         Teach sheriffbot to answer whois commands
38833         https://bugs.webkit.org/show_bug.cgi?id=55687
38834
38835         * Scripts/webkitpy/tool/bot/irc_command.py:
38836         * Scripts/webkitpy/tool/bot/irc_command_unittest.py:
38837         * Scripts/webkitpy/tool/bot/sheriffircbot_unittest.py:
38838
38839 2011-03-03  Victoria Kirst  <vrk@google.com>
38840
38841         Reviewed by Eric Carlson.
38842
38843         Enable media tests in chromium-gpu/test_expectations
38844         https://bugs.webkit.org/show_bug.cgi?id=53608
38845
38846         Add media folder to the list of GPU test paths.
38847
38848         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
38849
38850 2011-03-03  Adam Roben  <aroben@apple.com>
38851
38852         Work around a bug in KURL's parsing of Windows-style absolute file: URLs
38853
38854         Fixes <http://webkit.org/b/55674> Many tests fail in WebKit2 mode on Windows due to
38855         incorrect parsing of absolute Windows-style file: URLs
38856
38857         Reviewed by Tim Hatcher.
38858
38859         * WebKitTestRunner/TestInvocation.cpp:
38860         (WTR::createWKURL): Add a "localhost" host to file: URLs we create to work around
38861         <http://webkit.org/b/55683>.
38862
38863 2011-03-03  Tony Chang  <tony@chromium.org>
38864
38865         Reviewed by Dimitri Glazkov.
38866
38867         [chromium] Run kill-old-processes on the chromium-win bot too
38868         https://bugs.webkit.org/show_bug.cgi?id=55630
38869
38870         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
38871         * BuildSlaveSupport/win/kill-old-processes: Add wdiff.exe to the kill list.
38872
38873 2011-03-03  Carlos Garcia Campos  <cgarcia@igalia.com>
38874
38875         Reviewed by Martin Robinson.
38876
38877         [GTK] DRT needs implementation of EventSender.scheduleAsynchronousClick
38878         https://bugs.webkit.org/show_bug.cgi?id=53960
38879
38880         * DumpRenderTree/gtk/EventSender.cpp:
38881         (sendClick):
38882         (scheduleAsynchronousClickCallback):
38883
38884 2011-03-03  Adam Roben  <aroben@apple.com>
38885
38886         Look for Windows-style paths when checking whether a test is an Inspector test
38887
38888         WTR part of <http://webkit.org/b/55672> <rdar://problem/9080867> All inspector tests time
38889         out or crash on Windows 7 Release (WebKit2 Tests)
38890
38891         Reviewed by Darin Adler.
38892
38893         * WebKitTestRunner/TestInvocation.cpp:
38894         (WTR::shouldOpenWebInspector): Also look for backslashes, since that's what we'll get on
38895         Windows.
38896
38897 2011-03-03  Adam Roben  <aroben@apple.com>
38898
38899         Make Windows 7 Release (Tests) a core builder
38900
38901         We can probably make Windows XP Debug (Tests) a core builder, too, but we should probably
38902         give it a few more days of greenness first.
38903
38904         Fixes <http://webkit.org/b/55665> Windows 7 Release (Tests) should be a core builder
38905
38906         Reviewed by Andreas Kling.
38907
38908         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
38909         (BuildBot): Added a regexp to match Windows 7 Release (Tests).
38910
38911         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
38912         (BuildBotTest): Updated the list of builders to match the current list on build.webkit.org.
38913         Updated the core regexps to match the buildbot module. Updated the expected results to match
38914         the current set of core builders.
38915
38916 2011-03-03  Benjamin Poulain  <benjamin.poulain@nokia.com>
38917
38918         Reviewed by Andreas Kling.
38919
38920         Adding myself as a reviewer.
38921
38922         * Scripts/webkitpy/common/config/committers.py:
38923
38924 2011-03-03  Lukasz Slachciak  <l.slachciak@samsung.com>
38925
38926         Reviewed by Xan Lopez.
38927
38928         [GTK] Extended application cache database API and added unit tests file.
38929         https://bugs.webkit.org/show_bug.cgi?id=55335
38930
38931         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: removed external declaration
38932         of webkit_application_cache_set_maximum_size due to API exposal.
38933
38934 2011-03-02  Ojan Vafai  <ojan@chromium.org>
38935
38936         Reviewed by Adam Barth.
38937
38938         make webkit-patch upload work for security bugs
38939         https://bugs.webkit.org/show_bug.cgi?id=55655
38940
38941         The problem is that we shell out to prepare-ChangeLog,
38942         which does not authenticate with bugzilla. All it needs from
38943         bugzilla is the bug title though. We just pass it through from
38944         webkit-patch instead.
38945
38946         * Scripts/prepare-ChangeLog:
38947         * Scripts/webkitpy/common/net/bugzilla/bug.py:
38948         * Scripts/webkitpy/tool/steps/preparechangelog.py:
38949
38950 2011-03-01  Ojan Vafai  <ojan@chromium.org>
38951
38952         Reviewed by Tony Chang.
38953
38954         [test-dashboard] make gtest JSON match layout-tests JSON
38955         https://bugs.webkit.org/show_bug.cgi?id=55556
38956
38957         http://trac.webkit.org/changeset/80090 stopped converting time_ms to t.
38958         Make this code match.
38959
38960         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
38961
38962 2011-03-02  Dirk Pranke  <dpranke@chromium.org>
38963
38964         Reviewed by Tony Chang.
38965
38966         NRWT - implement Linux Hardy 64-bit port support. This involves
38967         adding support for non-'x86' architectures and updating the
38968         "deduplicate_tests" script to correctly understand that
38969         port.name() is not always equal to basename(port.baseline_path()).
38970
38971         This change introduces two new port names: 'chromium-linux-x86'
38972         and 'chromium-linux-x86_64'. Using the prior 'chromium-linux'
38973         flag will cause the code to determine which one to use at
38974         runtime, just like we currently do with 'chromium-win'.
38975
38976         https://bugs.webkit.org/show_bug.cgi?id=55535
38977
38978         * Scripts/webkitpy/common/system/executive_mock.py:
38979           Handle 'return_stderr' keyword arg to run_command()
38980         * Scripts/webkitpy/layout_tests/deduplicate_tests.py:
38981         * Scripts/webkitpy/layout_tests/deduplicate_tests_unittest.py:
38982         * Scripts/webkitpy/layout_tests/port/base.py:
38983         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
38984         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py:
38985         * Scripts/webkitpy/layout_tests/port/factory.py:
38986         * Scripts/webkitpy/layout_tests/port/google_chrome.py:
38987
38988 2011-03-02  Kevin Ollivier  <kevino@theolliviers.com>
38989
38990         [wx] Build fix, make sure we put the wxWebKit dylib in the right directory and
38991         update symlinks accordingly.
38992
38993         * wx/packaging/build-mac-installer.py:
38994
38995 2011-03-02  Sheriff Bot  <webkit.review.bot@gmail.com>
38996
38997         Unreviewed, rolling out r80139.
38998         http://trac.webkit.org/changeset/80139
38999         https://bugs.webkit.org/show_bug.cgi?id=55620
39000
39001         failing and crashing tests on Snow Leopard bot (Requested by
39002         mihaip1 on #webkit).
39003
39004         * DumpRenderTree/LayoutTestController.cpp:
39005         (LayoutTestController::staticFunctions):
39006         * DumpRenderTree/LayoutTestController.h:
39007         * DumpRenderTree/chromium/LayoutTestController.cpp:
39008         (LayoutTestController::LayoutTestController):
39009         * DumpRenderTree/chromium/LayoutTestController.h:
39010         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
39011         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
39012         * DumpRenderTree/mac/UIDelegate.h:
39013         * DumpRenderTree/mac/UIDelegate.mm:
39014         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
39015         * DumpRenderTree/qt/LayoutTestControllerQt.h:
39016         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
39017         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
39018
39019 2011-03-02  Adam Roben  <aroben@apple.com>
39020
39021         Wait for force a paint in WebKitTestRunner until we're actually ready to dump the output
39022
39023         This matches DumpRenderTree on Windows.
39024
39025         Fixes <http://webkit.org/b/55469> <rdar://problem/9068539> REGRESSION (r79863): Lots of
39026         dom/html/level2/html/HTMLFrameElement*.html tests crashing in FrameView::paintOverhangAreas
39027         in WebKit2
39028
39029         Reviewed by Beth Dakin.
39030
39031         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
39032         (WTR::InjectedBundlePage::dump): Moved code to force a paint here...
39033         (WTR::InjectedBundlePage::didFinishLoadForFrame): ...from here.
39034
39035 2011-03-02  Daniel Cheng  <dcheng@chromium.org>
39036
39037         Reviewed by David Levin.
39038
39039         Add feature define for data transfer items
39040         https://bugs.webkit.org/show_bug.cgi?id=55510
39041
39042         * Scripts/build-webkit:
39043
39044 2011-03-02  Adam Roben  <aroben@apple.com>
39045
39046         Look in the new-to-SnowLeopard location for crash logs on SnowLeopard and newer
39047
39048         ReportCrash saves logs to ~/Library/Logs/CrashReporter on Leopard, but to
39049         ~/Library/Logs/DiagnosticReports on SnowLeopard. old-run-webkit-tests was only looking in
39050         the former location, but was getting lucky on some SnowLeopard machines because symlinks
39051         were being created there pointing into DiagnosticReports. For machines without these
39052         symlinks, crash logs were not getting captured.
39053
39054         Fixes <http://webkit.org/b/55607> old-run-webkit-tests doesn't capture crash logs on some
39055         SnowLeopard machines
39056
39057         Reviewed by Eric Seidel.
39058
39059         * Scripts/old-run-webkit-tests:
39060         (captureSavedCrashLog): Look in ~/Library/Logs/DiagnosticReports on SnowLeopard and newer.
39061
39062 2011-03-02  Adam Roben  <aroben@apple.com>
39063
39064         Don't try to capture crash logs on platforms that haven't implemented that feature
39065
39066         Fixes <http://webkit.org/b/55504> Use of uninitialized value in numeric lt (<) at
39067         Tools/Scripts/old-run-webkit-tests line 1778 seen on Qt bots after a crash
39068
39069         Reviewed by Joseph Pecoraro.
39070
39071         * Scripts/old-run-webkit-tests:
39072         (captureSavedCrashLog): Bail if we don't have a crash log glob.
39073         (findNewestFileMatchingGlob): Fix a buggy test for when no paths match the glob. This isn't
39074         strictly required to fix bug 55504, but seems worthwhile.
39075
39076 2011-01-26  John Knottenbelt  <jknotten@chromium.org>
39077
39078         Reviewed by Dmitry Titov.
39079
39080         Detach Geolocation from Frame when Page destroyed.
39081         https://bugs.webkit.org/show_bug.cgi?id=52877
39082
39083         Extend the layout test controller to expose the number of pending
39084         geolocation requests, so that we can test that the requests have
39085         been cancelled on page close.
39086
39087         * DumpRenderTree/LayoutTestController.cpp:
39088         (numberOfPendingGeolocationPermissionRequestsCallback):
39089         (LayoutTestController::staticFunctions):
39090         * DumpRenderTree/LayoutTestController.h:
39091         * DumpRenderTree/chromium/LayoutTestController.cpp:
39092         (LayoutTestController::LayoutTestController):
39093         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
39094         * DumpRenderTree/chromium/LayoutTestController.h:
39095         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
39096         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
39097         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
39098         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
39099         * DumpRenderTree/mac/UIDelegate.h:
39100         * DumpRenderTree/mac/UIDelegate.mm:
39101         (-[UIDelegate numberOfPendingGeolocationPermissionRequests]):
39102         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
39103         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
39104         * DumpRenderTree/qt/LayoutTestControllerQt.h:
39105         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
39106         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
39107         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
39108         (LayoutTestController::numberOfPendingGeolocationPermissionRequests):
39109
39110
39111 2011-03-01  Kent Tamura  <tkent@chromium.org>
39112
39113         Reviewed by Dimitri Glazkov.
39114
39115         Assertion fails when a form validation bubble appears
39116         https://bugs.webkit.org/show_bug.cgi?id=55550
39117
39118         Test: fast/forms/interactive-validation-attach-assertion.html
39119
39120         * DumpRenderTree/chromium/WebPreferences.cpp:
39121         (WebPreferences::applyTo): Enable the interactive validation feature by
39122           default. This matches to Chromium browser.
39123
39124 2011-03-01  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
39125
39126         Reviewed by Kent Tamura.
39127
39128         Remove efl port style exception of Webcore/JavaScriptCore
39129         https://bugs.webkit.org/show_bug.cgi?id=55549
39130
39131         WebCore and JavaScriptCore need to adhere WebKit coding style.
39132
39133         * Scripts/webkitpy/style/checker.py:
39134
39135 2011-03-01  Tony Chang  <tony@chromium.org>
39136
39137         Reviewed by Ojan Vafai.
39138
39139         [chromium] clean up temp files on chromium-mac builder too
39140         https://bugs.webkit.org/show_bug.cgi?id=55525
39141
39142         * BuildSlaveSupport/chromium/remove-crash-logs:
39143
39144 2011-02-28  Ojan Vafai  <ojan@chromium.org>
39145
39146         Reviewed by Tony Chang.
39147
39148         wrap json in a function call to afford cross-domain loading
39149         https://bugs.webkit.org/show_bug.cgi?id=55353
39150
39151         Also delete the code that compacts the JSON. It turns out this
39152         breaks the rebaseline tool and is probably a premature optimization anyways.
39153
39154         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
39155         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
39156         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
39157         * Scripts/webkitpy/tool/commands/rebaselineserver.py:
39158
39159 2011-03-01  Sheriff Bot  <webkit.review.bot@gmail.com>
39160
39161         Unreviewed, rolling out r80079.
39162         http://trac.webkit.org/changeset/80079
39163         https://bugs.webkit.org/show_bug.cgi?id=55547
39164
39165         "Broke the Win debug build?" (Requested by dcheng on #webkit).
39166
39167         * Scripts/build-webkit:
39168
39169 2011-03-01  Mark Rowe  <mrowe@apple.com>
39170
39171         Reviewed by Sam Weinig.
39172
39173         Update verification scripts to handle WebKit2.
39174
39175         * Scripts/check-for-global-initializers: Skip object files that use RefCountedLeakChecker.
39176         * Scripts/check-for-inappropriate-files-in-framework: Add some inappropriate file types for
39177         WebKit2.
39178
39179 2011-03-01  Daniel Cheng  <dcheng@chromium.org>
39180
39181         Reviewed by David Levin.
39182
39183         Add feature define for data transfer items
39184         https://bugs.webkit.org/show_bug.cgi?id=55510
39185
39186         * Scripts/build-webkit:
39187
39188 2011-02-23  Joseph Pecoraro  <joepeck@webkit.org>
39189
39190         Reviewed by Timothy Hatcher.
39191
39192         All Console Messages should be passed to ChromeClients.
39193         https://bugs.webkit.org/show_bug.cgi?id=54926
39194
39195         Update DumpRenderTree's UIDelegate to use the new console message delegate.
39196
39197         * DumpRenderTree/mac/UIDelegate.mm:
39198         (-[UIDelegate webView:addMessageToConsole:withSource:]):
39199
39200 2011-03-01  Dirk Pranke  <dpranke@chromium.org>
39201
39202         Reviewed by Adam Barth.
39203
39204         Tragically, we stopped running the python unit tests
39205         on Chromium Win just as I was getting them to be green.
39206         This patch turns them back on.
39207
39208         https://bugs.webkit.org/show_bug.cgi?id=55521
39209
39210         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
39211
39212 2011-03-01  Dirk Pranke  <dpranke@chromium.org>
39213
39214         Reviewed by Adam Roben.
39215
39216         Disable
39217         webkitpy.common.system.executive_unittest.ExecutiveTest.\
39218            test_kill_all fails on win until it can be properly triaged.
39219
39220         https://bugs.webkit.org/show_bug.cgi?id=54790
39221
39222         * Scripts/webkitpy/common/system/executive_unittest.py:
39223
39224     2011-03-01  Helder Correia  <helder@sencha.com>
39225
39226             Unreviewed. Add myself to committers.py.
39227
39228             * Scripts/webkitpy/common/config/committers.py:
39229
39230 2011-03-01  Adam Roben  <aroben@apple.com>
39231
39232         Give up if a crash log for the web process is taking too long to be saved on Windows
39233
39234         This should keep the bots from getting stuck while waiting for a crash log, if writing a
39235         crash log ever hangs or the UI process for some reason doesn't notice it's finished.
39236
39237         Fixes <http://webkit.org/b/55499> WebKitTestRunner can hang forever waiting for a crash log
39238         to be saved for the web process
39239
39240         Reviewed by Steve Falkenburg.
39241
39242         * WebKitTestRunner/win/TestControllerWin.cpp:
39243         (WTR::runRunLoopUntil): Moved code to run the run loop here here from platformRunUntil.
39244         Generalized the code slightly to handle the optional object to wait on.
39245         (WTR::TestController::platformRunUntil): Use the new runRunLoopUntil function, and also use
39246         it (with a timeout) when waiting for a crash log for the web process to be saved. This will
39247         prevent us from waiting forever.
39248
39249 2011-03-01  Adam Roben  <aroben@apple.com>
39250
39251         Give WebKitLauncherWin's PDB file a name different from WebKit's
39252
39253         WebKitLauncherWin (which builds WebKit.exe) and WebKit (which builds WebKit.dll) were
39254         fighting over the same PDB file, meaning that one of them was going to lose and not have
39255         symbols. WebKitLauncherWin's PDB file is now named WebKitLauncherWin.pdb.
39256
39257         Fixes <http://webkit.org/b/55495> Crashes in WebKit.dll have bad backtraces for some test
39258         runs (due to missing symbols)
39259
39260         Reviewed by David Kilzer.
39261
39262         * WebKitLauncherWin/WebKitLauncherWinCommon.vsprops: Use the project name,
39263         "WebKitLauncherWin", instead of the target name, "WebKit", when naming the PDB file.
39264
39265 2011-03-01  Adam Roben  <aroben@apple.com>
39266
39267         Count web process crashes toward the --exit-after-n-crashes-or-timeouts limit
39268
39269         Fixes <http://webkit.org/b/55492> Web process crashes aren't counted toward the
39270         --exit-after-n-crashes-or-timeouts limit
39271
39272         Reviewed by David Kilzer.
39273
39274         * Scripts/old-run-webkit-tests:
39275         (stopRunningTestsEarlyIfNeeded): Include web process crashes in the computation, and print
39276         the number of tests that caused the web process to crash when exiting early.
39277
39278 2011-03-01  Adam Roben  <aroben@apple.com>
39279
39280         Save a crash log when the web process crashes
39281
39282         On Windows, WebKitTestRunner now detects when the web process is crashing and waits to exit
39283         until it has finished crashing, which guarantees that the crash log will have had time to be
39284         saved, too. On Mac, we always wait until ReportCrash has exited before capturing the crash
39285         log, so all we have to do is choose the right crash log out of the CrashReporter directory.
39286
39287         Fixes <http://webkit.org/b/44121> <rdar://problem/8320759> When the web process crashes and
39288         a crash log is being saved, WebKitTestRunner thinks the web process has become unresponsive
39289
39290         Reviewed by Sam Weinig.
39291
39292         * Scripts/old-run-webkit-tests:
39293         (testCrashedOrTimedOut): Don't kill WebKitTestRunner when the web process crashes. It will
39294         kill itself. On Windows, this will cause us to wait until the crash log has been saved. On
39295         Mac, it should have no effect. Capture saved crash logs for web process crashes, too.
39296         (captureSavedCrashLog): Added $webProcessCrashed argument. On Mac, look for
39297         WebProces_*.crash files when the web process crashes.
39298
39299         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
39300         (WTR::InjectedBundle::initialize): Added an initializationUserData argument. Updated for
39301         initializePlatformDefaults -> platformInitialize rename. Pass the initializationUserData
39302         along to platformInitialize.
39303
39304         * WebKitTestRunner/InjectedBundle/InjectedBundle.h: See above.
39305
39306         * WebKitTestRunner/InjectedBundle/InjectedBundleMain.cpp:
39307         (WKBundleInitialize): Pass along the initializationUserData to the InjectedBundle.
39308
39309         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm:
39310         (WTR::InjectedBundle::platformInitialize):
39311         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp:
39312         (WTR::InjectedBundle::platformInitialize):
39313         Updated function signature.
39314
39315         * WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp:
39316         (WTR::exceptionFilter): Added. Tells the UI process we're crashing by signaling the
39317         webProcessCrashingEvent, then lets the crash continue as normal.
39318
39319         (WTR::InjectedBundle::platformInitialize): Hook up exceptionFilter. Retrieve the name of the
39320         event we should use to tell the UI process we're crashing from the initializationUserData,
39321         and get a handle to that event.
39322
39323         * WebKitTestRunner/TestController.cpp:
39324         (WTR::TestController::TestController): Initialize new members.
39325         (WTR::TestController::processDidCrash): Removed unnecessary WKPageRef argument. Changed to
39326         only print the "#CRASHED - WebProcess" message once, since this can be called more than once
39327         when a crash log is being saved on Windows. Exit right away if specified. (This is the
39328         default.)
39329
39330         * WebKitTestRunner/TestController.h: Added new members.
39331
39332         * WebKitTestRunner/win/TestControllerWin.cpp:
39333         (WTR::TestController::platformInitialize): Set up the event the web process will use to tell
39334         us it's crashing.
39335         (WTR::TestController::platformRunUntil): Pass MWMO_INPUTAVAILABLE to
39336         ::MsgWaitForMultipleObjectsEx so we'll process messages that have already been seen by
39337         ::PeekMessage. (This is unrelated to the bug fix.) Notice when the webProcessCrashingEvent
39338         has been signaled. When this happens, print the "#CRASHED - WebProcess" message right away
39339         so the test harness will know the web process has crashed and not try to kill us, then wait
39340         for the web process to finish crashing so a crash log will have time to be saved.
39341         (WTR::toWK): Simple hepler function.
39342         (WTR::TestController::platformInitializeContext): Pass along the name of the event the web
39343         process should use to tell us it is crashing in the context's initialization user data.
39344
39345 2011-03-01  Dimitri Glazkov  <dglazkov@chromium.org>
39346
39347         Reviewed by Tony Gentilcore.
39348
39349         Move the checks to exclude Python/Perl tests for Chromium Win build to the right place.
39350         https://bugs.webkit.org/show_bug.cgi?id=55476
39351
39352         Chromium Win uses BuildAndTestFactory, not TestFactory.
39353
39354         * BuildSlaveSupport/build.webkit.org-config/master.cfg: Moved to the right place.
39355
39356 2011-03-01  Adam Roben  <aroben@apple.com>
39357
39358         Test that the WebKit2 responsiveness timer doesn't fire too early
39359
39360         Test for <http://webkit.org/b/55417> <rdar://problem/9065287> RunLoop::Timer fires 1000x too
39361         early on Windows
39362
39363         Reviewed by Anders Carlsson.
39364
39365         * TestWebKitAPI/PlatformUtilities.h: Added sleep().
39366
39367         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly.cpp: Added.
39368         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
39369         (TestWebKitAPI::didFinishLoadForFrame):
39370         (TestWebKitAPI::processDidBecomeUnresponsive):
39371         (TestWebKitAPI::setInjectedBundleClient):
39372         (TestWebKitAPI::setPageLoaderClient):
39373         Simple helper functions.
39374
39375         (TestWebKitAPI::TEST): Load an HTML file to make sure the web process is initialized. Then
39376         tell the web process to pause and press the spacebar key. The spacebar keypress should cause
39377         the responsiveness timer to start, but the web process should unpause before it has a chance
39378         to fire. Run until the web process has unpaused, and assert that the timer didn't fire.
39379
39380         * TestWebKitAPI/Tests/WebKit2/ResponsivenessTimerDoesntFireEarly_Bundle.cpp: Added.
39381         (TestWebKitAPI::ResponsivenessTimerDoesntFireEarlyTest::ResponsivenessTimerDoesntFireEarlyTest):
39382         Call up to the base class.
39383         (TestWebKitAPI::ResponsivenessTimerDoesntFireEarlyTest::didReceiveMessage): When asked to
39384         pause, sleep for 0.5 seconds, then send back a message saying we paused.
39385
39386         * TestWebKitAPI/mac/PlatformUtilitiesMac.mm:
39387         (TestWebKitAPI::Util::sleep): Added. Calls through to usleep.
39388         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp:
39389         (TestWebKitAPI::Util::sleep): Added. Calls through to ::Sleep.
39390
39391         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
39392         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
39393         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
39394         Added new files.
39395
39396 2011-03-01  Andras Becsi  <abecsi@webkit.org>
39397
39398         Reviewed by Csaba Osztrogonác.
39399
39400         [Qt] Clean up the project files and move common options to WebKit.pri.
39401
39402         * WebKitTestRunner/qt/WebKitTestRunner.pro: Deduplicate.
39403
39404
39405 2011-03-01  Adam Roben  <aroben@apple.com>
39406
39407         Skip another multiprocessing test on Windows
39408
39409         It is sometimes failing (and may be leaving child Python processes around).
39410         <http://webkit.org/b/55087> tracks the failure.
39411
39412         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
39413         (FuncitonTests.test_get__processes): Skip this test on Windows.
39414
39415 2011-03-01  Ilya Tikhonovsky  <loislo@chromium.org>
39416
39417         Reviewed by Pavel Feldman.
39418
39419         Web Inspector: reduce inspector tests flakiness.
39420         https://bugs.webkit.org/show_bug.cgi?id=55401
39421
39422         * DumpRenderTree/chromium/TestShell.cpp:
39423         (TestShell::runFileTest):
39424         * DumpRenderTree/chromium/WebPreferences.cpp:
39425         (WebPreferences::reset):
39426
39427 2011-02-28  Mihai Parparita  <mihaip@chromium.org>
39428
39429         Reviewed by James Robinson.
39430
39431         NRWT should still upload results when it exits early due to too many failures
39432         https://bugs.webkit.org/show_bug.cgi?id=55432
39433
39434         Only keyboard interrupts should prevent results from being uploaded.
39435
39436         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
39437         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
39438
39439 2011-02-28  Alice Liu  <alice.liu@apple.com>
39440
39441         Reviewed by Adam Roben and Gavin Barraclough.
39442
39443         Test for https://bugs.webkit.org/show_bug.cgi?id=54898
39444         InjectedBundleNodeHandle dies too early in WKBundleHitTestResultGetNodeHandle.
39445
39446         * TestWebKitAPI/InjectedBundleController.h:
39447         (TestWebKitAPI::InjectedBundleController::bundle):
39448         * TestWebKitAPI/PlatformWebView.h:
39449         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle.cpp: Added.
39450         (TestWebKitAPI::didFinishLoadForFrame):
39451         (TestWebKitAPI::didReceiveMessageFromInjectedBundle):
39452         (TestWebKitAPI::setPageLoaderClient):
39453         (TestWebKitAPI::TEST):
39454         * TestWebKitAPI/Tests/WebKit2/HitTestResultNodeHandle_Bundle.cpp: Added.
39455         (TestWebKitAPI::HitTestResultNodeHandleTest::HitTestResultNodeHandleTest):
39456         (TestWebKitAPI::HitTestResultNodeHandleTest::getContextMenuFromDefaultMenu):
39457         (TestWebKitAPI::HitTestResultNodeHandleTest::didCreatePage):
39458         * TestWebKitAPI/mac/PlatformWebViewMac.mm:
39459         (TestWebKitAPI::PlatformWebView::simulateRightClick):
39460         * TestWebKitAPI/win/PlatformWebViewWin.cpp:
39461         (TestWebKitAPI::PlatformWebView::simulateRightClick):
39462
39463         Added files to:
39464         * TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
39465         * TestWebKitAPI/win/TestWebKitAPI.vcproj:
39466         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj:
39467
39468 2011-02-28  Adam Roben  <aroben@apple.com>
39469
39470         Start running InjectedBundle API tests on Windows
39471
39472         A new project, TestWebKitAPIInjectedBundle, builds the injected bundle DLL.
39473
39474         Fixes <http://webkit.org/b/55420> run-api-tests doesn't run injected bundle tests on Windows
39475
39476         Reviewed by Darin Adler.
39477
39478         * TestWebKitAPI/Configurations/TestWebKitAPIDebug.vsprops: Added.
39479         * TestWebKitAPI/Configurations/TestWebKitAPIDebugAll.vsprops: Added.
39480         * TestWebKitAPI/Configurations/TestWebKitAPIDebugCairoCFLite.vsprops: Added.
39481         * TestWebKitAPI/Configurations/TestWebKitAPIRelease.vsprops: Added.
39482         * TestWebKitAPI/Configurations/TestWebKitAPIReleaseCairoCFLite.vsprops: Added.
39483         These are just extracted from TestWebKitAPI.vcproj.
39484
39485         * TestWebKitAPI/Configurations/TestWebKitAPIInjectedBundleCommon.vsprops: Added. Just sets
39486         up the DLL's filename.
39487
39488         * TestWebKitAPI/win/PlatformUtilitiesWin.cpp: Fixed DLL name.
39489         (TestWebKitAPI::Util::createInjectedBundlePath): Changed to use kCFAllocatorNull instead of
39490         0 (== kCFAllocatorDefault) so that CF won't try to deallocate our C string.
39491
39492         * TestWebKitAPI/win/TestWebKitAPI.sln: Added TestWebKitAPIInjectedBundle. Removed the bogus
39493         "all" configuration.
39494
39495         * TestWebKitAPI/win/TestWebKitAPI.vcproj: Changed to use the new .vsprops files. Added
39496         injected bundle tests. Let VS resort things.
39497
39498         * TestWebKitAPI/win/TestWebKitAPIInjectedBundle.vcproj: Copied from Tools/TestWebKitAPI/win/TestWebKitAPI.vcproj.
39499
39500 2011-02-28  Adam Roben  <aroben@apple.com>
39501
39502         Add a test for the paint rect passed via WM_PAINT to windowless plugins
39503
39504         Test for <http://webkit.org/b/55365> <rdar://problem/9031089> REGRESSION (r79040): WebKit2:
39505         Large portions of pages with plugins paint black
39506
39507         Reviewed by Anders Carlsson.
39508
39509         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
39510         (PluginTest::NPP_HandleEvent):
39511         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
39512         Added NPP_HandleEvent. Just returns 0 at this level.
39513
39514         * DumpRenderTree/TestNetscapePlugIn/Tests/win/WindowlessPaintRectCoordinates.cpp: Added.
39515         (WindowlessPaintRectCoordinates::WindowlessPaintRectCoordinates): Call up to the base class.
39516         (WindowlessPaintRectCoordinates::NPP_New): Mark ourselves as windowless.
39517         (WindowlessPaintRectCoordinates::NPP_HandleEvent): Check that the paint rect passed via
39518         WM_PAINT has the expected coordinates.
39519
39520         * DumpRenderTree/TestNetscapePlugIn/main.cpp:
39521         (NPP_HandleEvent): Give the PluginTest a chance to handle the event before doing anything
39522         else.
39523
39524         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj: Added new file, let VS
39525         resort things.
39526
39527         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
39528         (WTR::InjectedBundlePage::didFinishLoadForFrame): Added a call to WKBundlePageForceRepaint.
39529         This is roughly equivalent to the call to -[WebView displayIfNeeded] in -[FrameLoadDelegate
39530         webView:didFinishLoadForFrame:] in DumpRenderTree.
39531
39532         * WebKitTestRunner/win/PlatformWebViewWin.cpp:
39533         (WTR::PlatformWebView::PlatformWebView): Tell the WKView it's in a window so that plugins
39534         can start running.
39535
39536 2011-02-28  Adam Roben  <aroben@apple.com>
39537
39538         Tell MiniBrowser's WKViews that they're in a window
39539
39540         Fixes <http://webkit.org/b/55364> Plugins don't work in MiniBrowser on Windows
39541
39542         Reviewed by Anders Carlsson.
39543
39544         * MiniBrowser/win/BrowserView.cpp:
39545         (BrowserView::create): Call WKViewSetIsInWindow after creating the view.
39546
39547 2011-02-28  Alexis Menard  <alexis.menard@openbossa.org>
39548
39549         Reviewed by Csaba Osztrogonác.
39550
39551         The new QML tests require QML_IMPORT_PATH variable to be set in order to
39552         load the proper plugin.
39553
39554         * Scripts/run-qtwebkit-tests:
39555
39556 2011-02-28  Joone Hur  <joone.hur@collabora.co.uk>
39557
39558         Unreviewed, fix spelling mistake.
39559
39560         hardware-accelated => hardware-accelerated.
39561
39562         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
39563
39564 2011-02-28  Kristian Amlie  <kristian.amlie@nokia.com>
39565
39566         Reviewed by Andreas Kling.
39567
39568         Made sure that the build-webkit --qmake argument is always respected.
39569
39570         Previously, it would be respected during the build, but not during
39571         the early feature detection. The build would also fail if qmake was
39572         not in your path, even if you specified --qmake with a valid qmake
39573         binary.
39574
39575         [Qt] WebKit patches required to work with a modularized version of Qt
39576         https://bugs.webkit.org/show_bug.cgi?id=53916
39577
39578         * Scripts/build-webkit:
39579         * Scripts/webkitdirs.pm:
39580
39581 2011-02-27  Ojan Vafai  <ojan@chromium.org>
39582
39583         Fix failing tests from http://trac.webkit.org/changeset/79837.
39584
39585         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
39586         This was just a mistake that would break if someone tried printing with this configuration.
39587         * Scripts/webkitpy/layout_tests/layout_package/printing_unittest.py:
39588         Method signature changed. Didn't notice that this method was being calling in a unittest.
39589
39590 2011-02-24  Hayato Ito  <hayato@chromium.org>
39591
39592         Reviewed by Eric Seidel.
39593
39594         [NRWT] Remove test_types/* classes.
39595
39596         This is the last patch in the series of efforts which tried to remove test_types classes.
39597         A single_test_runner module will take over responsibilities of test_type classes.
39598
39599         From this patch, we won't display statistics per test_type even if a '--verbose' option is given.
39600
39601         https://bugs.webkit.org/show_bug.cgi?id=55123
39602
39603         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
39604         * Scripts/webkitpy/layout_tests/layout_package/test_results.py:
39605         * Scripts/webkitpy/layout_tests/layout_package/test_results_unittest.py:
39606         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
39607         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
39608         * Scripts/webkitpy/layout_tests/test_types/__init__.py: Removed.
39609         * Scripts/webkitpy/layout_tests/test_types/image_diff.py: Removed.
39610         * Scripts/webkitpy/layout_tests/test_types/test_type_base.py: Removed.
39611         * Scripts/webkitpy/layout_tests/test_types/test_type_base_unittest.py: Removed.
39612         * Scripts/webkitpy/layout_tests/test_types/text_diff.py: Removed.
39613
39614 2011-02-25  Ojan Vafai  <ojan@chromium.org>
39615
39616         Reviewed by Tony Chang.
39617
39618         Change results.json format to the one used by unexpected_results.json
39619         https://bugs.webkit.org/show_bug.cgi?id=52267
39620
39621         Also add runtimes in milliseconds to the JSON and make the output format more compact.
39622         Named the file full_results.json to avoid conflicting with the results.json
39623         file the test-results server currently serves up.
39624
39625         * Scripts/webkitpy/layout_tests/layout_package/json_layout_results_generator.py:
39626         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
39627         * Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
39628         * Scripts/webkitpy/layout_tests/layout_package/printing.py:
39629         * Scripts/webkitpy/layout_tests/layout_package/result_summary.py:
39630         * Scripts/webkitpy/layout_tests/layout_package/test_runner.py:
39631
39632 2011-02-27  Adam Roben  <aroben@apple.com>
39633
39634         Followup fix to r79827
39635
39636         * DumpRenderTree/win/UIDelegate.cpp:
39637         (UIDelegate::webViewAddMessageToConsole): Replaced call to urlSuitableForTestResult with
39638         lastPathComponent. This matches what we did here before r79827, and matches Mac.
39639
39640 2011-02-27  Dan Bernstein  <mitz@apple.com>
39641
39642         Reviewed by Adam Roben.
39643
39644         Make bisect-builds print a trac link for the regression range
39645         https://bugs.webkit.org/show_bug.cgi?id=55331
39646
39647         * Scripts/bisect-builds:
39648
39649 2011-02-27  Adam Roben  <aroben@apple.com>
39650
39651         Use iswalpha instead of isalpha when dealing with wchar_ts in EditingDelegate
39652
39653         Fixes <http://webkit.org/b/55062> <rdar://problem/9059907> Crash beneath
39654         EditingDelegate::checkSpellingOfString when running fast/forms/input-text-maxlength.html or
39655         fast/forms/input-text-paste-maxlength.html on Windows with full page heap enabled
39656
39657         Reviewed by Anders Carlsson.
39658
39659         * DumpRenderTree/win/EditingDelegate.cpp:
39660         (indexOfFirstWordCharacter):
39661         (wordLength):
39662         Use iswalpha instead of isalpha, since TCHAR is really just wchar_t.
39663
39664 2011-02-27  Adam Roben  <aroben@apple.com>
39665
39666         Make ResourceLoadDelegate print URLs relative to the main resource on Windows
39667
39668         This matches Mac. (In fact, the code was ported from Mac!)
39669
39670         Fixes <http://webkit.org/b/55328> URLs printed by ResourceLoadDelegate on Windows don't
39671         match Mac
39672
39673         Reviewed by Anders Carlsson.
39674
39675         * DumpRenderTree/win/DumpRenderTree.cpp:
39676         (substringFromIndex): Added. Emulates -[NSString substringFromIndex:].
39677         (urlSuitableForTestResult): Ported code from -[NSURL(DRTExtras)
39678         _drt_descriptionSuitableForTestResult].
39679         (cfStringRefToWString): Moved here from LayoutTestControllerWin.cpp.
39680
39681         * DumpRenderTree/win/DumpRenderTreeWin.h: Added declaration of cfStringRefToWString.
39682
39683         * DumpRenderTree/win/LayoutTestControllerWin.cpp: Moved cfStringRefToWString to
39684         DumpRenderTree.cpp.
39685
39686 2011-02-27  Adam Roben  <aroben@apple.com>
39687
39688         Change the order of Chang Shu's email addresses so the first one is their Bugzilla username
39689
39690         This makes Bugzilla autocompletion work correctly.
39691
39692         * Scripts/webkitpy/common/config/committers.py:
39693
39694 2011-02-27  Patrick Gansterer  <paroga@webkit.org>
39695
39696         Reviewed by Andreas Kling.
39697
39698         [Qt] fast/loader/user-stylesheet-fast-path.html fails
39699         https://bugs.webkit.org/show_bug.cgi?id=50911
39700
39701         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
39702         (LayoutTestController::setUserStyleSheetLocation):
39703         Use QUrl::fromEncoded since the passed string is already encoded.
39704
39705 2011-02-27  Adam Roben  <aroben@apple.com>
39706
39707         Ensure $testResultsDirectory is an absolute path before setting up Windows crash log saving
39708
39709         Windows 7 (and maybe Vista?) launches the post-mortem debugger with a working directory that
39710         is different from old-run-webkit-test's. Because of this, the path to the crash log file
39711         (which is relative to $testResultsDirectory) must be an absolute path. See the bug for more
39712         details.
39713
39714         Fixes <http://webkit.org/b/55318> Crash logs aren't saved on Windows 7 when a relative
39715         --results-directory path is used (like on the test slaves)
39716
39717         Reviewed by Dan Bernstein.
39718
39719         * Scripts/old-run-webkit-tests: Moved the call to setUpWindowsCrashLogSaving after we've
39720         made $testResultsDirectory an absolute path.
39721
39722 2011-02-26  Dominic Mazzoni  <dmazzoni@google.com>
39723
39724         Reviewed by Adam Barth.
39725
39726         m_dumpAccessibilityNotifications should be initialized.
39727         https://bugs.webkit.org/show_bug.cgi?id=54874
39728
39729         * DumpRenderTree/chromium/AccessibilityController.cpp:
39730         (AccessibilityController::AccessibilityController):
39731
39732 2011-02-26  Vsevolod Vlasov  <vsevik@chromium.org>
39733
39734         Reviewed by Pavel Feldman.
39735
39736         DumpRenderTree should reset frame opener between tests.
39737         https://bugs.webkit.org/show_bug.cgi?id=54874
39738
39739         Added clearOpener() calls between tests.
39740         Some tests are using window.open() to change current window location.
39741         This changes window.opener property whose nullity is checked before
39742         showing xml viewer.
39743
39744         * DumpRenderTree/chromium/TestShell.cpp:
39745         (TestShell::resetTestController):
39746         * DumpRenderTree/gtk/DumpRenderTree.cpp:
39747         (resetDefaultsToConsistentValues):
39748         * DumpRenderTree/mac/DumpRenderTree.mm:
39749         (resetWebViewToConsistentStateBeforeTesting):
39750         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
39751         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
39752         * DumpRenderTree/win/DumpRenderTree.cpp:
39753         (resetWebViewToConsistentStateBeforeTesting):
39754         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
39755         (WTR::InjectedBundlePage::reset):
39756
39757 2011-02-26  Yael Aharon  <yael.aharon@nokia.com>
39758
39759         Reviewed by Andreas Kling.
39760
39761         [Qt] Fix the focus in MiniBrowser.
39762         https://bugs.webkit.org/show_bug.cgi?id=55288
39763
39764         Give focus to the web view when a load starts.
39765
39766         * MiniBrowser/qt/BrowserWindow.cpp:
39767         (BrowserWindow::urlChanged):
39768
39769 2011-02-26  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
39770
39771         Reviewed by Adam Barth.
39772
39773         [EFL] Add coding style exceptions for EFL port
39774         https://bugs.webkit.org/show_bug.cgi?id=54733
39775
39776         The EFL APIs use EFL naming style, which includes both lower-cased and camel-cased,
39777         underscore-sparated values. But, style checking rule have found parameter_name errors
39778         in efl directories so far.
39779
39780         * Scripts/webkitpy/style/checker.py: Add prameter_name rule's exception and efl directories.
39781
39782 2011-02-26  David Levin  <levin@chromium.org>
39783
39784         Reviewed by Darin Adler.
39785
39786         check-webkit-style should check for spaces after periods in comments
39787         https://bugs.webkit.org/show_bug.cgi?id=55253
39788
39789         * Scripts/webkitpy/style/checkers/cpp.py: Added the check.
39790         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added tests
39791           and fixed the name of one related test.
39792
39793 2011-02-26  Kenneth Rohde Christiansen  <kenneth@webkit.org>
39794
39795         Reviewed by Andreas Kling.
39796
39797         Make it possible to test the targetdensity-dpi support
39798         https://bugs.webkit.org/show_bug.cgi?id=55142
39799
39800         Test the viewport meta tag feature targetdensity-dpi by
39801         adding extra arguments to dumpConfigurationForViewport
39802
39803         Add the new arguments to the shared header and all
39804         implementation.
39805
39806         * DumpRenderTree/LayoutTestController.cpp:
39807         (dumpConfigurationForViewportCallback):
39808         * DumpRenderTree/LayoutTestController.h:
39809         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
39810         (LayoutTestController::dumpConfigurationForViewport):
39811         * DumpRenderTree/mac/LayoutTestControllerMac.mm:
39812         (LayoutTestController::dumpConfigurationForViewport):
39813         * DumpRenderTree/win/LayoutTestControllerWin.cpp:
39814         (LayoutTestController::dumpConfigurationForViewport):
39815         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
39816         (LayoutTestController::dumpConfigurationForViewport):
39817         * DumpRenderTree/qt/LayoutTestControllerQt.cpp:
39818         (LayoutTestController::dumpConfigurationForViewport):
39819         * DumpRenderTree/qt/LayoutTestControllerQt.h:
39820
39821 2011-02-26  Sheriff Bot  <webkit.review.bot@gmail.com>
39822
39823         Unreviewed, rolling out r79764.
39824         http://trac.webkit.org/changeset/79764
39825         https://bugs.webkit.org/show_bug.cgi?id=55295
39826
39827         "broke Chromium builds" (Requested by rniwa on #webkit).
39828
39829         * DumpRenderTree/chromium/TestShell.cpp:
39830         (TestShell::resetTestController):
39831         * DumpRenderTree/gtk/DumpRenderTree.cpp:
39832         (resetDefaultsToConsistentValues):
39833         * DumpRenderTree/mac/DumpRenderTree.mm:
39834         (resetWebViewToConsistentStateBeforeTesting):
39835         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
39836         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
39837         * DumpRenderTree/win/DumpRenderTree.cpp:
39838         (resetWebViewToConsistentStateBeforeTesting):
39839         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
39840         (WTR::InjectedBundlePage::reset):
39841
39842 2011-02-26  Yongjun Zhang  <yongjun_zhang@apple.com>
39843
39844         Reviewed by David Kilzer.
39845
39846         https://bugs.webkit.org/show_bug.cgi?id=48781
39847
39848         Add a resource load delegate method to query if WebCore should paint the default broken image for failed images.
39849
39850         Add a new resource load client method (shouldPaintBrokenImage).  WebKit client can decide if WebCore
39851         should paint the default broken image when an image fails to load or decode.
39852
39853         * DumpRenderTree/LayoutTestController.cpp:
39854         (LayoutTestController::LayoutTestController):
39855         (setShouldPaintBrokenImageCallback):
39856         (LayoutTestController::staticFunctions):
39857         (LayoutTestController::setShouldPaintBrokenImage):
39858         * DumpRenderTree/LayoutTestController.h:
39859         (LayoutTestController::shouldPaintBrokenImage):
39860         * DumpRenderTree/mac/ResourceLoadDelegate.mm:
39861         (-[ResourceLoadDelegate webView:shouldPaintBrokenImageForURL:]):
39862
39863 2011-02-26  Vsevolod Vlasov  <vsevik@chromium.org>
39864
39865         Reviewed by Pavel Feldman.
39866
39867         DumpRenderTree should reset frame opener between tests.
39868         https://bugs.webkit.org/show_bug.cgi?id=54874
39869
39870         Added clearOpener() calls between tests.
39871         Some tests are using window.open() to change current window location.
39872         This changes window.opener property whose nullity is checked before
39873         showing xml viewer.
39874
39875         * DumpRenderTree/chromium/TestShell.cpp:
39876         (TestShell::resetTestController):
39877         * DumpRenderTree/gtk/DumpRenderTree.cpp:
39878         (resetDefaultsToConsistentValues):
39879         * DumpRenderTree/mac/DumpRenderTree.mm:
39880         (resetWebViewToConsistentStateBeforeTesting):
39881         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
39882         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
39883         * DumpRenderTree/win/DumpRenderTree.cpp:
39884         (resetWebViewToConsistentStateBeforeTesting):
39885         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
39886         (WTR::InjectedBundlePage::reset):
39887
39888 2011-02-25  Lukasz Slachciak  <l.slachciak@samsung.com>
39889
39890         Reviewed by Kenneth Rohde Christiansen.
39891
39892         [EFL] Added EWebLauncher command line option to turn on/off frame flattening
39893         https://bugs.webkit.org/show_bug.cgi?id=54809
39894
39895         * EWebLauncher/main.c: Handling -f option added.
39896         (on_key_down):
39897         (browserCreate):
39898         (main):
39899
39900 2011-02-25  Sheriff Bot  <webkit.review.bot@gmail.com>
39901
39902         Unreviewed, rolling out r79324.
39903         http://trac.webkit.org/changeset/79324
39904         https://bugs.webkit.org/show_bug.cgi?id=55260
39905
39906         Leopard Debug is once again running out of address space while
39907         linking (Requested by aroben on #webkit).
39908
39909         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
39910         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
39911
39912 2011-02-25  Tony Chang  <tony@chromium.org>
39913
39914         Reviewed by James Robinson.
39915
39916         Clean up temp files left by crashing DRTs on chromium-win
39917         https://bugs.webkit.org/show_bug.cgi?id=55182
39918
39919         * BuildSlaveSupport/build.webkit.org-config/master.cfg:
39920         * BuildSlaveSupport/chromium/remove-crash-logs: Added.
39921
39922 2011-02-25  Adam Roben  <aroben@apple.com>
39923
39924         Double-check that modifying the registry worked
39925
39926         On Windows Vista/7 with UAC enabled, regtool will fail to modify the registry, but it will
39927         still return a successful exit code. So we double-check here that the value we tried to
39928         write to the registry was really written.
39929
39930         Fixes <http://webkit.org/b/55227> old-run-webkit-tests falsely claims to be saving crash
39931         logs on Windows Vista/7
39932
39933         Reviewed by Darin Adler.
39934
39935         * Scripts/old-run-webkit-tests:
39936         (writeRegistryString): After trying to write the string value, check that it was really
39937         written.
39938
39939 2011-02-25  Adam Roben  <aroben@apple.com>
39940
39941         Set the 32-bit post-mortem debugger on 64-bit OSes
39942
39943         The post-mortem debugger is controlled by some registry values in
39944         HKLM\SOFTWARE\Microsoft\Windows NT\CurrentVersion\AeDebug. On 64-bit OSes, you can either
39945         set the values in that key, which will be used for 64-bit apps, or in the equivalent key
39946         under Wow6432Node, which will be used for 32-bit apps. Since DumpRenderTree and
39947         WebKitTestRunner are 32-bit, we want to use the latter location.
39948
39949         Fixes <http://webkit.org/b/55225> old-run-webkit-tests sets the post-mortem debugger for
39950         64-bit apps, but should set it for 32-bit apps instead
39951
39952         Reviewed by Darin Adler.
39953
39954         * Scripts/old-run-webkit-tests:
39955         (readRegistryString):
39956         (writeRegistryString):
39957         Pass --wow32 to regtool so it will get/set values beneath Wow6432Node, which is where
39958         Windows looks for the post-mortem debugger values when a 32-bit app crashes on a 64-bit OS.
39959         Passing --wow32 on a 32-bit OS will not have any effect.
39960
39961 2011-02-25  Adam Roben  <aroben@apple.com>
39962
39963         Clean up code to set up crash log saving on Windows
39964
39965         Fixes <http://webkit.org/b/55224> old-run-webkit-test's Windows registry code is ugly
39966
39967         Reviewed by David Kilzer.
39968
39969         * Scripts/old-run-webkit-tests:
39970         (readRegistryString):
39971         (writeRegistryString):
39972         Moved code to read strings from and write strings to the registry here...
39973
39974         (setUpWindowsCrashLogSaving): ...from here.
39975
39976 2011-02-25  Adam Roben  <aroben@apple.com>
39977
39978         Work around Cygwin's crash-suppression behavior
39979
39980         Cygwin calls ::SetErrorMode(SEM_FAILCRITICALERRORS), which any processes it launches will
39981         inherit. This is bad for testing/debugging, as it causes the post-mortem debugger not to be
39982         invoked. (Cygwin does this because it makes crashes more UNIX-y.) We reset the error mode
39983         when our test apps launch to work around Cygwin's behavior.
39984
39985         Fixes <http://webkit.org/b/55222> Test apps crash silently (without invoking post-mortem
39986         debugger) when launched from Cygwin 1.7
39987
39988         Reviewed by Darin Adler.
39989
39990         * DumpRenderTree/win/DumpRenderTree.cpp:
39991         (main):
39992         * TestWebKitAPI/win/main.cpp:
39993         (main):
39994         * WebKitAPITest/main.cpp:
39995         (main):
39996         * WebKitTestRunner/win/TestControllerWin.cpp:
39997         (WTR::TestController::platformInitialize):
39998         Call ::SetErrorMode(0) to undo Cygwin's folly.
39999
40000 2011-02-24  Jocelyn Turcotte  <jocelyn.turcotte@nokia.com>
40001
40002         Reviewed by Andreas Kling.
40003
40004         [Qt] Revert the support for QNAM affined to a different thread.
40005         https://bugs.webkit.org/show_bug.cgi?id=55149
40006
40007         Qt 4.8 will have QNAM use its own thread internally by default,
40008         no need to keep this complexity in WebKit.
40009
40010         This mainly reverts:
40011         http://trac.webkit.org/changeset/73710
40012         http://trac.webkit.org/changeset/73712
40013
40014         * QtTestBrowser/launcherwindow.cpp:
40015         (LauncherWindow::~LauncherWindow):
40016         (LauncherWindow::initializeView):
40017         (LauncherWindow::createChrome):
40018         * QtTestBrowser/launcherwindow.h:
40019         (WindowOptions::WindowOptions):
40020         * QtTestBrowser/webpage.cpp:
40021         (WebPage::WebPage):
40022         * QtTestBrowser/webpage.h:
40023
40024 2011-02-24  Adam Barth  <abarth@webkit.org>
40025
40026         Reviewed by Eric Seidel.
40027
40028         Add a script for creating gypi files
40029         https://bugs.webkit.org/show_bug.cgi?id=55113
40030
40031         This is a trivial script to create gypi files from the source code in
40032         the current directory.
40033
40034         * Scripts/make-gypi: Added.
40035
40036 2011-02-24  Sheriff Bot  <webkit.review.bot@gmail.com>
40037
40038         Unreviewed, rolling out r79628.
40039         http://trac.webkit.org/changeset/79628
40040         https://bugs.webkit.org/show_bug.cgi?id=55195
40041
40042         "broke chromium win tests" (Requested by dpranke on #webkit).
40043
40044         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
40045         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
40046
40047 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
40048
40049         Reviewed by Mihai Parparita.
40050
40051         rebaseline-chromium-webkit-tests can't handle GPU variants
40052         https://bugs.webkit.org/show_bug.cgi?id=55002
40053
40054         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
40055         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
40056
40057 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
40058
40059         Reviewed by Mihai Parparita.
40060
40061         nrwt: log command used to start web server to --verbose
40062
40063         https://bugs.webkit.org/show_bug.cgi?id=55188
40064
40065         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
40066         * Scripts/webkitpy/layout_tests/port/http_server.py:
40067
40068 2011-02-24  Victoria Kirst  <vrk@google.com>
40069
40070         Reviewed by Mihai Parparita.
40071
40072         [chromium] Move media-file.js and video-test.js back to media/ for UILayoutTests
40073         https://bugs.webkit.org/show_bug.cgi?id=55089
40074
40075         This adds an alias into the media/ directory so that http/tests/media
40076         tests can access the media resources when running in an httpd process.
40077
40078         * Scripts/webkitperl/httpd.pm:
40079         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
40080         * Scripts/webkitpy/layout_tests/port/http_server.py:
40081
40082 2011-02-24  Dirk Pranke  <dpranke@chromium.org>
40083
40084         Reviewed by Mihai Parparita.
40085
40086         nrwt: enable multiple processes by default on Chromium Win.
40087
40088         https://bugs.webkit.org/show_bug.cgi?id=55163
40089
40090         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
40091         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
40092
40093 2011-02-24  Sam Weinig  <sam@webkit.org>
40094
40095         Attempt to fix Qt.
40096
40097         * WebKitTestRunner/TestInvocation.cpp:
40098
40099 2011-02-24  James Robinson  <jamesr@chromium.org>
40100
40101         Unreviewed, rolling out r79598.
40102         http://trac.webkit.org/changeset/79598
40103         https://bugs.webkit.org/show_bug.cgi?id=55089
40104
40105         Appears to be causing trouble with httpd on windows
40106
40107         * Scripts/webkitperl/httpd.pm:
40108         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
40109         * Scripts/webkitpy/layout_tests/port/http_server.py:
40110
40111 2011-02-24  Sam Weinig  <sam@webkit.org>
40112
40113         Reviewed by Darin Adler.
40114
40115         Add basic pixel testing support to WebKitTestRunner
40116         https://bugs.webkit.org/show_bug.cgi?id=55161
40117
40118         * Scripts/old-run-webkit-tests:
40119         Don't print color space information when using WebKitTestRunner since
40120         we don't need to change the screen.
40121
40122         * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
40123         (WTR::InjectedBundle::initialize):
40124         Add call to initializePlatformDefaults.
40125
40126         (WTR::InjectedBundle::done):
40127         Return a dictionary on completion with both the text output and pixel dump override.
40128
40129         * WebKitTestRunner/InjectedBundle/InjectedBundle.h:
40130         Add declaration for initializePlatformDefaults.
40131
40132         * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
40133         (WTR::LayoutTestController::LayoutTestController):
40134         (WTR::LayoutTestController::dumpAsText):
40135         * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
40136         (WTR::LayoutTestController::shouldDumpPixels):
40137         Make setting dumpAsText override the dump pixels preference.
40138
40139         * WebKitTestRunner/InjectedBundle/mac/InjectedBundleMac.mm: Added.
40140         (WTR::InjectedBundle::initializePlatformDefaults):
40141         Set platform defaults matching DumpRenderTree.
40142
40143         * WebKitTestRunner/InjectedBundle/qt/InjectedBundleQt.cpp: Added.
40144         (WTR::InjectedBundle::initializePlatformDefaults):
40145         * WebKitTestRunner/InjectedBundle/win/InjectedBundleWin.cpp: Added.
40146         (WTR::InjectedBundle::initializePlatformDefaults):
40147         Add stubs for initializePlatformDefaults.
40148
40149         * WebKitTestRunner/TestController.cpp:
40150         (WTR::TestController::runTest):
40151         (WTR::TestController::runTestingServerLoop):
40152         (WTR::TestController::run):
40153         Parse expected pixel results out of stdin and setup the invocation with it.
40154
40155         * WebKitTestRunner/TestInvocation.cpp:
40156         (WTR::TestInvocation::TestInvocation):
40157         Initialize new members.
40158
40159         (WTR::TestInvocation::~TestInvocation):
40160         (WTR::TestInvocation::setIsPixelTest):
40161         (WTR::sizeWebViewForCurrentTest):
40162         (WTR::TestInvocation::invoke):
40163         (WTR::TestInvocation::dump):
40164         (WTR::TestInvocation::didReceiveMessageFromInjectedBundle):
40165         * WebKitTestRunner/TestInvocation.h:
40166         Dump pixels in addition to text by calling dumpPixelsAndCompareWithExpected.
40167
40168         * WebKitTestRunner/mac/PlatformWebViewMac.mm:
40169         (WTR::PlatformWebView::PlatformWebView):
40170         Use -[NSColorSpace genericRGBColorSpace] instead of the main screen color space.
40171
40172         * WebKitTestRunner/mac/TestInvocationMac.mm: Added.
40173         (WTR::createCGContextFromPlatformView):
40174         (WTR::computeMD5HashStringForContext):
40175         (WTR::dumpBitmap):
40176         (WTR::forceRepaintFunction):
40177         (WTR::TestInvocation::dumpPixelsAndCompareWithExpected):
40178         Add mac specific pixel dumping support. Right now, we always pull pixels
40179         from the window server.
40180
40181         * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj:
40182         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
40183         * WebKitTestRunner/qt/WebKitTestRunner.pro:
40184         * WebKitTestRunner/win/InjectedBundle.vcproj:
40185         * WebKitTestRunner/win/WebKitTestRunner.vcproj:
40186         Add new files.
40187
40188 2011-02-24  Chang Shu  <cshu@webkit.org>
40189
40190         Reviewed by Kenneth Rohde Christiansen.
40191
40192         [Qt] Launch MiniBrowser if the test run is WebKit2.
40193         https://bugs.webkit.org/show_bug.cgi?id=55145
40194
40195         * Scripts/old-run-webkit-tests:
40196         * Scripts/run-launcher:
40197         * Scripts/webkitdirs.pm:
40198
40199 2011-02-24  Victoria Kirst  <vrk@google.com>
40200
40201         Reviewed by Mihai Parparita.
40202
40203         [chromium] Move media-file.js and video-test.js back to media/ for UILayoutTests
40204         https://bugs.webkit.org/show_bug.cgi?id=55089
40205
40206         This adds an alias into the media/ directory so that http/tests/media
40207         tests can access the media resources when running in an httpd process.
40208
40209         * Scripts/webkitperl/httpd.pm:
40210         * Scripts/webkitpy/layout_tests/port/apache_http_server.py:
40211         * Scripts/webkitpy/layout_tests/port/http_server.py:
40212
40213 2011-02-24  Andrew Wilson  <atwilson@chromium.org>
40214
40215         Unreviewed, rolling out r79570.
40216         http://trac.webkit.org/changeset/79570
40217         https://bugs.webkit.org/show_bug.cgi?id=54874
40218
40219         Breaks chromium build because glue/mocks/mock_web_frame.h/cc
40220         was not updated
40221
40222         * DumpRenderTree/chromium/TestShell.cpp:
40223         (TestShell::resetTestController):
40224         * DumpRenderTree/gtk/DumpRenderTree.cpp:
40225         (resetDefaultsToConsistentValues):
40226         * DumpRenderTree/mac/DumpRenderTree.mm:
40227         (resetWebViewToConsistentStateBeforeTesting):
40228         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
40229         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
40230         * DumpRenderTree/win/DumpRenderTree.cpp:
40231         (resetWebViewToConsistentStateBeforeTesting):
40232         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
40233         (WTR::InjectedBundlePage::reset):
40234
40235 2011-02-24  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
40236
40237         Reviewed by Laszlo Gombos.
40238
40239         [Qt] add android 2.2 user agent to QtTestBrowser
40240         https://bugs.webkit.org/show_bug.cgi?id=55085
40241
40242         Credit to Forrest Hodgkins to find right user-agent which works for youtube.
40243         * QtTestBrowser/useragentlist.txt:
40244
40245 2011-02-24  Adam Roben  <aroben@apple.com>
40246
40247         Change FrameLoadDelegate to support any number of delegates with delayed work to process
40248
40249         This makes our behavior match Mac more closely, and allows us to remove an incorrect
40250         assertion that was firing during some tests. (The assertion was claiming that there was
40251         never more than one delegate with delayed work to process, but that was not the case.)
40252
40253         Fixes <http://webkit.org/b/55146> Assertion failure in FrameLoadDelegate::locationChangeDone
40254         when running http/tests/navigation/back-twice-without-commit.html
40255
40256         Reviewed by Eric Carlson.
40257
40258         * DumpRenderTree/win/FrameLoadDelegate.cpp:
40259         (delegatesWithDelayedWork): Added. Returns all FrameLoadDelegates that have delayed work to
40260         process. A single delegate may appear in this Vector more than once (just as, on Mac, a
40261         single delegate may have multiple performSelector requests).
40262         (processWorkTimer): Pass the HWND to ::KillTimer, for pedantic brownie points. Added an
40263         assertion that the timer firing is the shared process work timer. Instead of using the
40264         single, global "delegate waiting for timer" delegate, give all delegates that have delayed
40265         work to process a chance to process their work.
40266         (FrameLoadDelegate::locationChangeDone): If we don't already have an active timer for
40267         processing delayed work, create one. Then add ourselves to the delegatesWithDelayedWork
40268         Vector so our processWork function will be called when the timer fires.
40269
40270 2011-02-24  Vsevolod Vlasov  <vsevik@chromium.org>
40271
40272         Reviewed by Alexey Proskuryakov.
40273
40274         DumpRenderTree should reset frame opener between tests.
40275         https://bugs.webkit.org/show_bug.cgi?id=54874
40276
40277         Added clearOpener() calls between tests.
40278         Some tests are using window.open() to change current window location.
40279         This changes window.opener property whose nullity is checked before
40280         showing xml viewer.
40281
40282         * DumpRenderTree/chromium/TestShell.cpp:
40283         (TestShell::resetTestController):
40284         * DumpRenderTree/gtk/DumpRenderTree.cpp:
40285         (resetDefaultsToConsistentValues):
40286         * DumpRenderTree/mac/DumpRenderTree.mm:
40287         (resetWebViewToConsistentStateBeforeTesting):
40288         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
40289         (WebCore::DumpRenderTree::resetToConsistentStateBeforeTesting):
40290         * DumpRenderTree/win/DumpRenderTree.cpp:
40291         (resetWebViewToConsistentStateBeforeTesting):
40292         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
40293         (WTR::InjectedBundlePage::reset):
40294
40295 2011-02-24  Adam Roben  <aroben@apple.com>
40296
40297         Don't trigger Windows builds for changes to WebCore.exp.in or the top-level GNUmakefile.am
40298
40299         Fixes <http://webkit.org/b/55144> Windows builders should not have built r79343 or r79440,
40300         but did
40301
40302         Reviewed by Anders Carlsson.
40303
40304         * Scripts/webkitpy/common/config/build.py:
40305         (_should_file_trigger_build): Loosened the pattern for GNUmakefile.am to include the
40306         top-level GNUmakefile.am, and loosened the pattern for .exp files to include WebCore.exp.in.
40307
40308         * Scripts/webkitpy/common/config/build_unittest.py:
40309         (ShouldBuildTest): Added some more tests.
40310
40311 2011-02-24  Andras Becsi  <abecsi@webkit.org>
40312
40313         Reviewed by Laszlo Gombos.
40314
40315         [Qt] MinGW build fails to link
40316         https://bugs.webkit.org/show_bug.cgi?id=55050
40317
40318         Prepend the libraries of subcomponents instead of appending them
40319         to fix the library order according to the dependency of the libraries
40320
40321         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
40322         * WebKitTestRunner/qt/WebKitTestRunner.pro:
40323
40324 2011-02-24  Gabor Loki  <loki@webkit.org>
40325
40326         Reviewed by Csaba Osztrogonác.
40327
40328         [Qt] Remove ARMv5 Release buildbot.
40329
40330         Remove ARMv5 Release buildbot, because there is very little difference
40331         between ARMv5 and ARMv7 Release bots, and there is also very little
40332         interest in ARMv5 build slave (compared to ARMv7).
40333
40334         * BuildSlaveSupport/build.webkit.org-config/config.json:
40335         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
40336
40337 2011-02-22  Ojan Vafai  <ojan@chromium.org>
40338
40339         Reviewed by Tony Chang.
40340
40341         fix revision number finding when creating git patches
40342         https://bugs.webkit.org/show_bug.cgi?id=55024
40343
40344         HEAD~n does not walk merge commits correctly. git log however does.
40345
40346         * Scripts/webkitpy/common/checkout/scm.py:
40347         * Scripts/webkitpy/common/checkout/scm_unittest.py:
40348
40349 2011-02-23  Kenneth Russell  <kbr@google.com>
40350
40351         Unreviewed, rolling out r79387.
40352         https://bugs.webkit.org/show_bug.cgi?id=54885
40353
40354         Reapply the code changes from r79038 since they weren't the root
40355         cause of the test timeouts.
40356
40357         * DumpRenderTree/chromium/WebViewHost.cpp:
40358         (WebViewHost::reset):
40359
40360 2011-02-23  Siddharth Mathur  <siddharth.mathur@nokia.com>
40361
40362         Reviewed by Laszlo Gombos.
40363
40364         [Qt] Make sure Symbian binary UIDs are unique
40365
40366         * MiniBrowser/qt/MiniBrowser.pro: Resolve the UID collision with
40367         QtTestBrowser.pro by changing the UID.
40368
40369 2011-02-23  Patrick Gansterer  <paroga@webkit.org>
40370
40371         Reviewed by Darin Adler.
40372
40373         Rename PLATFORM(CF) to USE(CF)
40374         https://bugs.webkit.org/show_bug.cgi?id=53540
40375
40376         * DumpRenderTree/config.h:
40377
40378 2011-02-23  Dirk Pranke  <dpranke@chromium.org>
40379
40380         Reviewed by Ojan Vafai.
40381
40382         nrwt: turn on multiprocessing on chromium_linux, clean up unit tests for worker model defaults
40383         https://bugs.webkit.org/show_bug.cgi?id=55016
40384
40385         * Scripts/webkitpy/layout_tests/port/base.py:
40386         * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
40387         * Scripts/webkitpy/layout_tests/port/chromium_linux_unittest.py: Added.
40388         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
40389         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
40390         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
40391         * Scripts/webkitpy/layout_tests/port/mac.py:
40392         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
40393         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
40394         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
40395
40396 2011-02-23  Tony Chang  <tony@chromium.org>
40397
40398         Reviewed by Kent Tamura.
40399
40400         [chromium] DumpRenderTree --check-layout-test-sys-deps leaks 4 temp dirs
40401         https://bugs.webkit.org/show_bug.cgi?id=55004
40402
40403         * DumpRenderTree/chromium/DumpRenderTree.cpp:
40404         (WebKitSupportTestEnvironment::WebKitSupportTestEnvironment):
40405         (WebKitSupportTestEnvironment::~WebKitSupportTestEnvironment):
40406         (main): Use a C++ object to ensure that TearDownTestEnvironment
40407             gets called when --check-layout-test-sys-deps is called
40408
40409 2011-02-23  Gopal Raghavan  <gopal.1.raghavan@nokia.com>
40410
40411         Reviewed by Anders Carlsson.
40412
40413         [Qt] [Qt] check webkit style fails in qt/tests
40414         https://bugs.webkit.org/show_bug.cgi?id=54998
40415
40416         check-webkit-style was complaining about missing config.h in WebKit/qt/test/qdeclarativewebview. This patch was originally part of 50222. Reviewer suggested to split and create a separate patch for checker.
40417
40418         * Scripts/webkitpy/style/checker.py:
40419
40420 2011-02-22  Vsevolod Vlasov  <vsevik@chromium.org>
40421
40422         Reviewed by Pavel Feldman.
40423
40424         DumpRenderTree should dump as text if test path contains "dumpAsText/"
40425         https://bugs.webkit.org/show_bug.cgi?id=54873
40426
40427         * DumpRenderTree/chromium/LayoutTestController.h:
40428         (LayoutTestController::setShouldDumpAsText):
40429         (LayoutTestController::setShouldGeneratePixelResults):
40430         * DumpRenderTree/chromium/TestShell.cpp:
40431         (TestShell::runFileTest):
40432         * DumpRenderTree/gtk/DumpRenderTree.cpp:
40433         (shouldDumpAsText):
40434         (dump):
40435         (runTest):
40436         * DumpRenderTree/mac/DumpRenderTree.mm:
40437         (shouldDumpAsText):
40438         (runTest):
40439         * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
40440         (WebCore::isDumpAsTextTest):
40441         (WebCore::DumpRenderTree::open):
40442         * DumpRenderTree/win/DumpRenderTree.cpp:
40443         (shouldDumpAsText):
40444         (runTest):
40445         * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
40446         (WTR::InjectedBundlePage::dump):
40447
40448 2011-02-22  Kenneth Russell  <kbr@google.com>
40449
40450         Reviewed by James Robinson.
40451
40452         Many tests started timing out on the chrome mac canaries around r79035
40453         https://bugs.webkit.org/show_bug.cgi?id=54885
40454
40455         Revert the code changes, but not the Chromium DEPS roll, from
40456         r79038 to see whether this clears up the timing-out tests. Was not
40457         able to reproduce the timeouts locally with a Release mode build
40458         of DRT (on Snow Leopard).
40459
40460         * DumpRenderTree/chromium/WebViewHost.cpp:
40461         (WebViewHost::reset):
40462
40463 2011-02-22  Dirk Pranke  <dpranke@chromium.org>
40464
40465         Reviewed by James Robinson.
40466
40467         nrwt: --platform chromium-gpu doesn't work with --worker-model=processes
40468         https://bugs.webkit.org/show_bug.cgi?id=55009
40469
40470         * Scripts/webkitpy/layout_tests/port/chromium_gpu.py:
40471
40472 2011-02-18  Mihai Parparita  <mihaip@chromium.org>
40473
40474         Reviewed by Ojan Vafai.
40475
40476         Default to 500 for --exit-after-n-failures
40477         https://bugs.webkit.org/show_bug.cgi?id=54773
40478
40479         Make 500 the default value for --exit-after-n-failures (ORWT bots
40480         already default to this as of r75726). Similarly, change the default for
40481         --exit-after-n-crashes-or-timeouts to 20.
40482
40483         * Scripts/webkitpy/layout_tests/run_webkit_tests.py:
40484
40485 2011-02-02  Martin Robinson  <mrobinson@igalia.com>
40486
40487         Reviewed by Xan Lopez.
40488
40489         [GTK] DRT needs an implementation of LayoutTestController.setSelectTrailingWhitespace
40490         https://bugs.webkit.org/show_bug.cgi?id=53603
40491
40492         Add support for LayoutTestController.setSelectTrailingWhitespace by calling through
40493         to DumpRenderTreeSupportGtk for this functionality.
40494
40495         * DumpRenderTree/gtk/DumpRenderTree.cpp:
40496         (resetDefaultsToConsistentValues): Always call LayoutTestController.setSelectTrailingWhitespace to
40497         false when reseting values between tests.
40498         * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
40499         (LayoutTestController::setSelectTrailingWhitespaceEnabled): Call LayoutTestController.setSelectTrailingWhitespace
40500         with the appropriate value.
40501
40502 2011-02-22  Adam Roben  <aroben@apple.com>
40503
40504         Don't trigger a Windows build when the top-level ChangeLog changes
40505
40506         r79320 is an example of a change that shouldn't have built on Windows, but did, because of
40507         this bug.
40508
40509         Fixes <http://webkit.org/b/54957> Windows bots build when top-level ChangeLog changes, but
40510         shouldn't
40511
40512         Reviewed by Anders Carlsson.
40513
40514         * Scripts/webkitpy/common/config/build.py:
40515         (_should_file_trigger_build): Modified the ChangeLog pattern to match the top-level
40516         ChangeLog, and re-sorted it in the list.
40517
40518         * Scripts/webkitpy/common/config/build_unittest.py:
40519         (ShouldBuildTest): Added a test that ChangeLogs don't trigger builds.
40520
40521 2011-02-22  Sheriff Bot  <webkit.review.bot@gmail.com>
40522
40523         Unreviewed, rolling out r78691.
40524         http://trac.webkit.org/changeset/78691
40525         https://bugs.webkit.org/show_bug.cgi?id=54956
40526
40527         Leopard Debug is successfully building again (Requested by
40528         aroben on #webkit).
40529
40530         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
40531         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
40532
40533 2011-02-22  Ilya Tikhonovsky  <loislo@chromium.org>
40534
40535         Reviewed by Pavel Feldman.
40536
40537         Web Inspector: flakyness of inspector tests.
40538         https://bugs.webkit.org/show_bug.cgi?id=54729
40539
40540         As far as we have the protocol with sequence numbers we can simplify test harness support
40541         and drop out chomium specific methods.
40542
40543         * DumpRenderTree/DumpRenderTree.gypi:
40544         * DumpRenderTree/chromium/DRTDevToolsAgent.cpp:
40545         (DRTDevToolsAgent::sendMessageToInspectorFrontend):
40546         (DRTDevToolsAgent::runtimePropertyChanged):
40547         (DRTDevToolsAgent::asyncCall):
40548         (DRTDevToolsAgent::call):
40549         * DumpRenderTree/chromium/DRTDevToolsAgent.h:
40550         * DumpRenderTree/chromium/DRTDevToolsCallArgs.cpp: Removed.
40551         * DumpRenderTree/chromium/DRTDevToolsCallArgs.h: Removed.
40552         * DumpRenderTree/chromium/DRTDevToolsClient.cpp:
40553         (DRTDevToolsClient::sendMessageToBackend):
40554         (DRTDevToolsClient::asyncCall):
40555         (DRTDevToolsClient::call):
40556         * DumpRenderTree/chromium/DRTDevToolsClient.h:
40557
40558 2011-02-22  Andras Becsi  <abecsi@webkit.org>
40559
40560         Reviewed by Laszlo Gombos.
40561
40562         [Qt] Redesign the build system
40563         https://bugs.webkit.org/show_bug.cgi?id=51339
40564
40565         Part 2.
40566
40567         Build WebCore as a static library, compile the WebKit API and WebKit2 API
40568         in a final step and link to WebKit2, WebCore and JSC libraries to fix
40569         linking issues resulting from stripped away symbols.
40570
40571         * WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro: Remove superfluous includepaths.
40572
40573 2011-02-21  Lukasz Slachciak  <l.slachciak@samsung.com>
40574
40575         Reviewed by Antonio Gomes.
40576
40577         [EFL] Restored spatial navigation option in EWebLauncher since #18662 bug is closed
40578         https://bugs.webkit.org/show_bug.cgi?id=54806
40579
40580         * EWebLauncher/main.c:
40581         (on_key_down): calling ewk_view_setting_spatial_navigation_set when "F12" pressed added.
40582
40583 2011-02-21  Dirk Pranke  <dpranke@chromium.org>
40584
40585         Reviewed by Kent Tamura.
40586
40587         nrwt: revert behavior on 10.5 to "old-threads". It's possible
40588         that r79062 has introduced a bunch of flakiness, although I
40589         can't explain why right now. This change should restore the
40590         old behavior, and then we can look at the build logs to see
40591         if the tests stop timing out or being flaky.
40592
40593         https://bugs.webkit.org/show_bug.cgi?id=54925
40594
40595         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
40596         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
40597         * Scripts/webkitpy/layout_tests/port/mac.py:
40598         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
40599
40600 2011-02-21  Dirk Pranke  <dpranke@chromium.org>
40601
40602         Reviewed by Dimitri Glazkov.
40603
40604         nrwt: fix typo causing us to try and repeatedly shut down the servers
40605         https://bugs.webkit.org/show_bug.cgi?id=54904
40606
40607         * Scripts/webkitpy/layout_tests/layout_package/worker_mixin.py:
40608
40609 2011-02-21  Alejandro G. Castro  <alex@igalia.com>
40610
40611         Reviewed by Csaba Osztrogonác.
40612
40613         REGRESSION(r79157):
40614         plugins/get-url-with-javascript-destroying-plugin.html make the
40615         next test crash
40616         https://bugs.webkit.org/show_bug.cgi?id=54863
40617
40618         * GNUmakefile.am:
40619
40620 2011-02-20  Dirk Pranke  <dpranke@chromium.org>
40621
40622         Reviewed by Ojan Vafai.
40623
40624         test-webkitpy: stop skipping a bunch of tests on win32
40625         https://bugs.webkit.org/show_bug.cgi?id=54788
40626
40627         * Scripts/webkitpy/common/net/testoutputset_unittest.py:
40628         * Scripts/webkitpy/test/main.py:
40629
40630 2011-02-20  Dirk Pranke  <dpranke@chromium.org>
40631
40632         Reviewed by Ojan Vafai.
40633
40634         Fix various bugs related to handling of dir separators and real
40635         vs. fake filesystems that were causing test-webkitpy to fail
40636         tests on win32. These bugs have a variety of causes but are
40637         lumped together because they should be easy to review.
40638
40639         https://bugs.webkit.org/show_bug.cgi?id=54700
40640
40641         * Scripts/webkitpy/common/net/testoutput.py:
40642           Here we were using os.path.sep but not escaping it prior to
40643           using in a regexp, which  caused bad things on win32.
40644         * Scripts/webkitpy/common/net/testoutput_unittest.py:
40645           Same here.
40646         * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
40647           Here we were using the native filesystem instead of the mock
40648           filesystem, making some tests unnecessarily platform dependent.
40649           Also we weren't setting the default configuration for the
40650           TestChromiumWinPort.
40651         * Scripts/webkitpy/layout_tests/update_webgl_conformance_tests.py:
40652           Here we *shouldn't* use os.path.join(), just so the code becomes
40653           less platform dependent.
40654         * Scripts/webkitpy/style/checker.py:
40655           Here we should be using os.path.sep instead of "/"; the tests
40656           require the native filesystem.
40657         * Scripts/webkitpy/style/checkers/cpp_unittest.py:
40658           Here the underlying code is using os.path.sep, so we probably
40659           should as well.
40660         * Scripts/webkitpy/style/test_expectations_unittest.py:
40661           Here we should be using the filesystem hanging of the port
40662           object (which is the mocked filesystem) rather than the native
40663           one.
40664
40665 2011-02-20  Anders Carlsson  <andersca@apple.com>
40666
40667         Reviewed by Maciej Stachowiak.
40668
40669         Crash when a plug-in requests a javascript: url that destroys the plug-in
40670         https://bugs.webkit.org/show_bug.cgi?id=54837
40671         <rdar://problem/9005475>
40672
40673         Add new plug-in test.
40674
40675         * DumpRenderTree/DumpRenderTree.gypi:
40676         * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
40677         * DumpRenderTree/TestNetscapePlugIn/PluginTest.cpp:
40678         (PluginTest::NPN_GetURL):
40679         * DumpRenderTree/TestNetscapePlugIn/PluginTest.h:
40680         * DumpRenderTree/TestNetscapePlugIn/Tests/GetURLWithJavaScriptURLDestroyingPlugin.cpp: Added.
40681         (GetURLWithJavaScriptURLDestroyingPlugin::GetURLWithJavaScriptURLDestroyingPlugin):
40682         (GetURLWithJavaScriptURLDestroyingPlugin::NPP_New):
40683         * DumpRenderTree/TestNetscapePlugIn/win/TestNetscapePlugin.vcproj:
40684         * DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
40685
40686 2011-02-19  Patrick Gansterer  <paroga@webkit.org>
40687
40688         Reviewed by Andreas Kling.
40689
40690         Make WinCE a core builder
40691         https://bugs.webkit.org/show_bug.cgi?id=54801
40692
40693         * Scripts/webkitpy/common/net/buildbot/buildbot.py:
40694         * Scripts/webkitpy/common/net/buildbot/buildbot_unittest.py:
40695
40696 2011-02-19  Lukasz Slachciak  <l.slachciak@samsung.com>
40697
40698         Reviewed by Andreas Kling.
40699
40700         Fixed problem with launching EWebLauncher with "run-launcher --efl". Incorrect path referring to non-existing .libs directory ws set.
40701         https://webkit.org/b/54778
40702
40703         * Scripts/webkitdirs.pm:
40704
40705 2011-02-18  David Levin  <levin@chromium.org>
40706
40707         Reviewed by Eric Seidel.
40708
40709         check-webkit-style: Misses brace style error when the line with the { has a } in it.
40710         https://bugs.webkit.org/show_bug.cgi?id=54769
40711
40712         * Scripts/webkitpy/style/checkers/cpp.py: Fixed the check for the close brace
40713           to only look after the last open brace, so that the open brace in this line
40714           "} else {" will still be able to trigger the error.
40715         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added related unit tests.
40716
40717 2011-02-18  Zan Dobersek  <zandobersek@gmail.com>
40718
40719         Reviewed by Martin Robinson.
40720
40721         [Gtk] Re-enable meter tag support
40722         https://bugs.webkit.org/show_bug.cgi?id=54762
40723
40724         Meter tag is supported on the Gtk port, enable it by default.
40725
40726         * Scripts/build-webkit:
40727
40728 2011-02-18  David Levin  <levin@chromium.org>
40729
40730         Reviewed by Eric Seidel.
40731
40732         check-webkit-style falsely complains about WebKitGTK+ public headers
40733         https://bugs.webkit.org/show_bug.cgi?id=54650
40734
40735         * Scripts/webkitpy/style/checker.py: Add some exceptions for the gtk files.
40736         * Scripts/webkitpy/style/checkers/cpp.py: Restrict the checks for WEBKIT_API
40737           to the chromium directory (and improved the checks slightly).
40738         * Scripts/webkitpy/style/checkers/cpp_unittest.py: Added corresponding tests.
40739
40740 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
40741
40742         Reviewed by James Robinson.
40743
40744         new-run-webkit-tests: enable multiple processes by default on
40745         mac. This change removes the artificial restrictions we placed
40746         on the # of child processes we used with the old threading
40747         model, and switches to the new message-based model and multiple
40748         processes, where available. If multiple processes are not
40749         available (Leopard / Python 2.5), then we use the 'inline' model
40750         instead of the 'old-threads' model on the 'mac' port or one
40751         process on the 'chromium-mac' port. We need additional testing
40752         to see if the new 'threads' model works reliably and is worth
40753         supporting, or if we should just live with things being slightly
40754         slow.
40755
40756         https://bugs.webkit.org/show_bug.cgi?id=54596
40757
40758         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
40759         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
40760         * Scripts/webkitpy/layout_tests/port/mac.py:
40761         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
40762         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
40763
40764 2011-02-18  Andrew Wilson  <atwilson@chromium.org>
40765
40766         Unreviewed, rolling out r79047.
40767         http://trac.webkit.org/changeset/79047
40768         https://bugs.webkit.org/show_bug.cgi?id=54596
40769
40770         Broke canary bots - please watch the canaries the next time
40771         you land this
40772
40773         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
40774         * Scripts/webkitpy/layout_tests/port/mac.py:
40775
40776 2011-02-18  Emil A Eklund  <eae@chromium.org>
40777
40778         Reviewed by Darin Adler.
40779
40780         Crash in EventHandler::sendContextMenuEventForKey
40781         https://bugs.webkit.org/show_bug.cgi?id=54495
40782
40783         Add support for the context menu key (VK_APPS) to EventSender for
40784         platforms that support the key.
40785
40786         * DumpRenderTree/chromium/EventSender.cpp:
40787         (EventSender::keyDown):
40788         * DumpRenderTree/gtk/EventSender.cpp:
40789         (keyDownCallback):
40790         * DumpRenderTree/qt/EventSenderQt.cpp:
40791         (EventSender::keyDown):
40792         * DumpRenderTree/win/EventSender.cpp:
40793         (keyDownCallback):
40794
40795 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
40796
40797         Reviewed by Ojan Vafai.
40798
40799         new-run-webkit-tests: enable multiple processes by default on mac
40800         https://bugs.webkit.org/show_bug.cgi?id=54596
40801
40802         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
40803         * Scripts/webkitpy/layout_tests/port/mac.py:
40804
40805 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
40806
40807         Reviewed by Adam Roben.
40808
40809         Skip pretty patch unit tests if ruby isn't installed.
40810
40811         https://bugs.webkit.org/show_bug.cgi?id=54699
40812
40813         * Scripts/webkitpy/common/prettypatch_unittest.py:
40814
40815 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
40816
40817         Reviewed by Adam Roben.
40818
40819         Skip mac.test_skipped_file_paths() when running on win32; the
40820         test has hardcoded '/' directory separators and since it's a
40821         port-specific test and we have coverage on other ports it's not
40822         worth it to make the test more generic.
40823
40824         https://bugs.webkit.org/show_bug.cgi?id=54698
40825
40826         * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
40827
40828 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
40829
40830         Reviewed by Adam Roben.
40831
40832         Fix webkitpy.layout_tests.port.config_unittest to work on Win32.
40833
40834         https://bugs.webkit.org/show_bug.cgi?id=54703
40835
40836         * Scripts/webkitpy/layout_tests/port/config_unittest.py:
40837
40838 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
40839
40840         Reviewed by Adam Roben.
40841
40842         Change the default port we're mocking out on win32 when we
40843         run mock_drt_unittest. Normally we'd default to the
40844         port-specific default, but that would be "win", which doesn't
40845         currently work.
40846
40847         https://bugs.webkit.org/show_bug.cgi?id=54708
40848
40849         * Scripts/webkitpy/layout_tests/port/mock_drt_unittest.py:
40850
40851 2011-02-18  Kenneth Russell  <kbr@google.com>
40852
40853         Reviewed by James Robinson.
40854
40855         [chromium] Clean up initialization of minimum timer interval
40856         https://bugs.webkit.org/show_bug.cgi?id=54772
40857
40858         Use new GetForegroundTabTimerInterval function in webkit_support
40859         to reset the minimum timer interval before each test.
40860
40861         * DumpRenderTree/chromium/WebViewHost.cpp:
40862         (WebViewHost::reset):
40863
40864 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
40865
40866         Reviewed by Adam Roben.
40867
40868         r78494 fixed the chromium_win unittest failing under win32,
40869         but not cygwin. This fixes cygwin as well.
40870
40871         https://bugs.webkit.org/show_bug.cgi?id=54652
40872
40873         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
40874
40875 2011-02-18  Dirk Pranke  <dpranke@chromium.org>
40876
40877         Reviewed by Ojan Vafai, Mihai Parparita.
40878
40879         new-run-webkit-tests: r78522 made it impossible to create a
40880         'chromium-win' or 'chromium-mac' port; they would always get the
40881         version tacked on to the end. It turns out that we actually need
40882         to be able to create these ports in order for
40883         rebaseline-chromium-webkit-tests to work correctly given the
40884         broken coverage of our layout bots. As soon as we get SL and
40885         Win 7 bots running reliably on the canaries, we can get rid of
40886         this change. Note that we only need to change the chromium_mac
40887         and win ports because linux doesn't yet use versions and that
40888         script doesn't work with non-chromium ports.
40889
40890         Also, this change fixes a bug where we were using
40891         "platform/chromium-win-win7" and
40892         "platform/chromium-mac-snowleopard" as the baseline dirs for
40893         Win 7 and SL; we should've been using the generic directories
40894         instead since they are the newest versions of the os's.
40895
40896         https://bugs.webkit.org/show_bug.cgi?id=54691
40897
40898         * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
40899         * Scripts/webkitpy/layout_tests/port/chromium_mac_unittest.py:
40900         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
40901         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
40902         * Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
40903
40904 2011-02-16  MORITA Hajime  <morrita@google.com>
40905
40906         Reviewed by Kent Tamura.
40907
40908         [Chromium][DRT] some PostDelayedTask() accesses possibly dangling this pointer.
40909         http://webkit.org/b/53899
40910
40911         Replaced static invocation wrappers with MethodTask objects,
40912         which are removed on WebViewHost destructor.
40913
40914         * DumpRenderTree/chromium/WebViewHost.cpp:
40915         (WebViewHost::requestCheckingOfText):
40916         (WebViewHost::scheduleAnimation):
40917         (WebViewHost::closeWidgetSoon):
40918         * DumpRenderTree/chromium/WebViewHost.h:
40919         (WebViewHost::HostMethodTask::HostMethodTask):
40920         (WebViewHost::HostMethodTask::runIfValid):
40921         (WebViewHost::taskList):
40922
40923 2011-02-07  Hayato Ito  <hayato@chromium.org>
40924
40925         Reviewed by Ojan Vafai.
40926
40927         Get rid of code which writes test results from test_type's
40928         compare_output() method.
40929         Writing test results now happens in single_test_runner, using
40930         newly introduced test_result_writer module.
40931
40932         Subsequent patches will eliminate test_type/* classes.
40933
40934         https://bugs.webkit.org/show_bug.cgi?id=53908
40935
40936         * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
40937         * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py: Added.
40938         * Scripts/webkitpy/layout_tests/test_types/image_diff.py:
40939         * Scripts/webkitpy/layout_tests/test_types/test_type_base.py:
40940         * Scripts/webkitpy/layout_tests/test_types/text_diff.py:
40941
40942 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
40943
40944         Reviewed by Adam Roben.
40945
40946         Add blacklist of remaining webkitpy/* modules that fail on
40947         Win32, so that test-webkitpy will at least run cleanly.
40948
40949         https://bugs.webkit.org/show_bug.cgi?id=54709
40950
40951         * Scripts/webkitpy/layout_tests/test/main.py:
40952
40953 2011-02-17  Simon Fraser  <simon.fraser@apple.com>
40954
40955         Reviewed by Sam Weinig.
40956
40957         Set a preference to ensure that scrollbars are in a predictable
40958         state for DRT.
40959
40960         * DumpRenderTree/mac/DumpRenderTree.mm:
40961         (resetDefaultsToConsistentValues):
40962
40963 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
40964
40965         Reviewed by Adam Roben.
40966
40967         test-webkitpy doesn't work with the multiprocessing module on win32
40968         or cygwin, so we skip it for now.
40969
40970         https://bugs.webkit.org/show_bug.cgi?id=54520
40971
40972         * Scripts/webkitpy/layout_tests/layout_package/manager_worker_broker_unittest.py:
40973         * Scripts/webkitpy/layout_tests/layout_package/run_webkit_tests_unittest.py:
40974
40975 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
40976
40977         Reviewed by Adam Roben.
40978
40979         Fix version, baseline_search_path for chromium-win-win7. Also
40980         fix chromium_win_unittests, and fix port_testcase.make_port()
40981         to correctly use the options passed into it and only set
40982         results_directory() if necessary.
40983
40984         https://bugs.webkit.org/show_bug.cgi?id=54652
40985
40986         * Scripts/webkitpy/layout_tests/port/chromium_win.py:
40987         * Scripts/webkitpy/layout_tests/port/chromium_win_unittest.py:
40988         * Scripts/webkitpy/layout_tests/port/port_testcase.py:
40989
40990 2011-02-17  Dirk Pranke  <dpranke@chromium.org>
40991
40992         Reviewed by Ojan Vafai.
40993
40994         Fix a hang in new-run-webkit-tests when run with
40995         --dry-run --worker-model=processes
40996
40997         https://bugs.webkit.org/show_bug.cgi?id=54595
40998
40999         * Scripts/webkitpy/layout_tests/layout_package/test_runner2.py:
41000         * Scripts/webkitpy/layout_tests/run_webkit_tests_unittest.py:
41001
41002 2011-02-17  Kevin Ollivier  <kevino@theolliviers.com>
41003
41004         [wx] Build fixes after recent changes.
41005
41006         * DumpRenderTree/wx/LayoutTestControllerWx.cpp:
41007         (LayoutTestController::setMinimumTimerInterval):
41008         * wx/build/settings.py:
41009
41010 2011-02-17  Csaba Osztrogonác  <ossy@webkit.org>
41011
41012         Reviewed by Andreas Kling.
41013
41014         [Qt] ImageDiff sometimes hangs
41015         https://bugs.webkit.org/show_bug.cgi?id=54641
41016
41017         * DumpRenderTree/qt/ImageDiff.cpp: Clear actualImage and baselineImage after all comparison.
41018         (main):
41019
41020 2011-02-17  Peter Varga  <pvarga@webkit.org>
41021
41022         Reviewed by Csaba Osztrogonác.
41023
41024         The run-sunspider --v8-suite doesn't work
41025         https://bugs.webkit.org/show_bug.cgi?id=54660
41026
41027         * Scripts/run-sunspider:
41028
41029 2011-02-17  Gabor Rapcsanyi  <rgabor@webkit.org>
41030
41031         Reviewed by Csaba Osztrogonác.
41032
41033         [NRWT] Fix environment setup on Qt port.
41034         https://bugs.webkit.org/show_bug.cgi?id=54549
41035
41036         * Scripts/webkitpy/layout_tests/port/qt.py:
41037
41038 2011-02-16  Yael Aharon  <yael.aharon@nokia.com>
41039
41040         Reviewed by Andreas Kling.
41041
41042         [Qt] Enable tiled backing store by default in MiniBrowser.
41043         https://bugs.webkit.org/show_bug.cgi?id=54581
41044
41045         Instead of a flag to turn it on, a new flag is defined to turn
41046         tiled backing store off.
41047
41048         * MiniBrowser/qt/MiniBrowserApplication.cpp:
41049         (MiniBrowserApplication::handleUserOptions):
41050         * MiniBrowser/qt/MiniBrowserApplication.h:
41051         (WindowOptions::WindowOptions):
41052
41053 2011-02-16  Eric Seidel  <eric@webkit.org>
41054
41055         Reviewed by Mihai Parparita.
41056
41057         flaky test archive uploads are always 22 bytes long
41058         https://bugs.webkit.org/show_bug.cgi?id=54593
41059
41060         Speculative fix for the 22-bytes bug.
41061
41062         * Scripts/webkitpy/tool/bot/flakytestreporter.py:
41063
41064 2011-02-16  Anna Cavender  <annacc@chromium.org>
41065
41066         Reviewed by Mihai Parparita.
41067
41068         http/tests/media should be served over HTTP (not from local file)
41069         https://bugs.webkit.org/show_bug.cgi?id=54028
41070
41071         Removed special file:// treatment of http/tests/media
41072
41073         * Scripts/old-run-webkit-tests:
41074         * Scripts/webkitpy/layout_tests/port/base.py:
41075
41076 2011-02-16  Mihai Parparita  <mihaip@chromium.org>
41077
41078         Reviewed by Ojan Vafai.
41079
41080         ChangeLog.parse_latest_entry_from_file does not handle rolled over ChangeLogs
41081         https://bugs.webkit.org/show_bug.cgi?id=54609
41082
41083         r78737 rolled over ChangeLogs, and webkit-patch complains when landing
41084         patches such as this one that are the first to add an entry to the
41085         ChangeLog (since they don't see a second date line to know where the
41086         entry ends).
41087
41088         * Scripts/webkitpy/common/checkout/changelog.py:
41089         * Scripts/webkitpy/common/checkout/changelog_unittest.py:
41090
41091 == Rolled over to ChangeLog-2011-02-16 ==