[WK2][EFL] creating dummy class derived from ResourceHandleClient for Efl download module.
https://bugs.webkit.org/show_bug.cgi?id=76171

Patch by Keunsoon Lee <keunsoon.lee@samsung.com> on 2012-02-08
Reviewed by Andreas Kling.

FileDownloaderEfl is a practical class to operate file download.
It communicates with Download class to start download
and to notify downloading states to user(e.g. client application).
Besides FileDownloaderEfl derives from ResourceHandleClient
and receives chunked data from ResourceHandle directly
or redirects data flow which is started from MainResourceLoader
by mean of exchanging ResourceHandle's existing client for FileDownloaderEfl.

This patch has only dummy functions to make easy to review.
Working patch will be uploaded on another bug thread.

* PlatformEfl.cmake: added new file to compile.
* WebProcess/Downloads/Download.h: added FileDownloaderEfl's instance as a form of OwnPtr<FileDownloaderEfl>.
(WebKit):
(Download):
* WebProcess/Downloads/efl/DownloadEfl.cpp:
(WebKit::Download::start): calling to FileDownloaderEfl::start() to start download for passing ResourceRequest.
* WebProcess/Downloads/efl/FileDownloaderEfl.cpp: Added. a practical class to operate file download.
(WebKit):
(WebKit::FileDownloaderEfl::create): create function, which returns PassOwnPtr<FileDownloaderEfl>.
(WebKit::FileDownloaderEfl::FileDownloaderEfl): constructor, which receives Download* as a parameter.
(WebKit::FileDownloaderEfl::~FileDownloaderEfl): deconstructor.
(WebKit::FileDownloaderEfl::start): a function to start downloading for passed ResourceRequest.
It does nothing for now, but shows how Download class can call FileDownloaderEfl's function.
(WebKit::FileDownloaderEfl::didReceiveResponse): virtual function for ResourceHandleClient.
It will receive response header information from ResourceHandle.
(WebKit::FileDownloaderEfl::didReceiveData): virtual function for ResourceHandleClient.
It will receive chunk data from ResourceHandle.
(WebKit::FileDownloaderEfl::didFinishLoading): virtual function for ResourceHandleClient.
It will be notified loading is finished from ResourceHandle.
(WebKit::FileDownloaderEfl::didFail): virtual function for ResourceHandleClient.
It will be notified loading is fail with ResourceError from ResourceHandle.
(WebKit::FileDownloaderEfl::shouldUseCredentialStorage): virtual function for ResourceHandleClient.
It returns whether to use credential storage or not.
(WebKit::FileDownloaderEfl::didReceiveAuthenticationChallenge): virtual function for ResourceHandleClient.
It will receive AuthenticationChallenge.
(WebKit::FileDownloaderEfl::didCancelAuthenticationChallenge): virtual function for ResourceHandleClient.
It will be notified AuthenticationChallenge is canceled.
(WebKit::FileDownloaderEfl::receivedCancellation): virtual function for ResourceHandleClient.
* WebProcess/Downloads/efl/FileDownloaderEfl.h: Added.
(WebCore):
(WebKit):
(FileDownloaderEfl):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107180 268f45cc-cd09-0410-ab3c-d52691b4dbfc

refactor RenderFlexibleBox to return preferred sizes of all children
https://bugs.webkit.org/show_bug.cgi?id=78169

Reviewed by Ojan Vafai.

This is in preparation for multi-line flexbox. We need the preferred
size of each child so we can compute where the multi-line breaks happen.

No new tests, just refactoring.

* rendering/RenderFlexibleBox.cpp:
(WebCore::RenderFlexibleBox::preferredMainAxisContentExtentForChild): Renamed from preferredMainAxisContentExtentForFlexItem because
other methods are ForChild rather than ForFlexItem.
(WebCore::RenderFlexibleBox::layoutFlexItems):
(WebCore::RenderFlexibleBox::computeMainAxisPreferredSizes):
(WebCore::RenderFlexibleBox::runFreeSpaceAllocationAlgorithm):
* rendering/RenderFlexibleBox.h:
(RenderFlexibleBox):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107179 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Stop calling Element::ensureShadowRoot() if it is used in construction phase.
https://bugs.webkit.org/show_bug.cgi?id=77929

Patch by Shinya Kawanaka <shinyak@google.com> on 2012-02-08
Reviewed by Hajime Morita.

ShadowRoot's life cycle can be consufing If Element::ensureShadowRoot() is used.
So we want to remove Element::ensureShadowRoot().
This patch replaces Element::ensureShadowRoot() if it is used in object construction phase.

No new tests, no change in behavior.

* dom/ShadowRoot.cpp:
(WebCore::ShadowRoot::create):
  Initialize exception code before calling appendChild.
* html/HTMLDetailsElement.cpp:
(WebCore::HTMLDetailsElement::createShadowSubtree):
* html/HTMLKeygenElement.cpp:
(WebCore::HTMLKeygenElement::HTMLKeygenElement):
* html/HTMLMeterElement.cpp:
(WebCore::HTMLMeterElement::createShadowSubtree):
* html/HTMLProgressElement.cpp:
(WebCore::HTMLProgressElement::createShadowSubtree):
* html/HTMLSummaryElement.cpp:
(WebCore::HTMLSummaryElement::createShadowSubtree):
* html/HTMLTextAreaElement.cpp:
(WebCore::HTMLTextAreaElement::createShadowSubtree):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107178 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Chromium] Avoid unnecessary full tile updates without breaking atomicity of commits.
https://bugs.webkit.org/show_bug.cgi?id=76740

Patch by David Reveman <reveman@chromium.org> on 2012-02-08
Reviewed by James Robinson.

Source/WebCore:

Allow the final batch of texture uploads to be performed without
allocating new textures and re-painting complete tiles.

This patch is tested by the following unit test:
- CCLayerTreeHostTestAtomicCommitWithPartialUpdate.runMultiThread
- TiledLayerChromiumTest.partialUpdates

* platform/graphics/chromium/TiledLayerChromium.cpp:
(WebCore::UpdatableTile::UpdatableTile):
(UpdatableTile):
(WebCore::TiledLayerChromium::updateCompositorResources):
(WebCore):
(WebCore::TiledLayerChromium::tileOnlyNeedsPartialUpdate):
(WebCore::TiledLayerChromium::tileNeedsBufferedUpdate):
(WebCore::TiledLayerChromium::prepareToUpdateTiles):
* platform/graphics/chromium/TiledLayerChromium.h:
(TiledLayerChromium):
* platform/graphics/chromium/cc/CCLayerTreeHost.cpp:
(WebCore::CCLayerTreeHost::CCLayerTreeHost):
(WebCore::CCLayerTreeHost::initializeLayerRenderer):
(WebCore::CCLayerTreeHost::updateLayers):
(WebCore::CCLayerTreeHost::requestPartialTextureUpdate):
(WebCore):
* platform/graphics/chromium/cc/CCLayerTreeHost.h:
(WebCore::CCSettings::CCSettings):
(CCSettings):
(CCLayerTreeHost):
* platform/graphics/chromium/cc/CCProxy.h:
(CCProxy):
* platform/graphics/chromium/cc/CCSingleThreadProxy.cpp:
(WebCore::CCSingleThreadProxy::doCommit):
* platform/graphics/chromium/cc/CCSingleThreadProxy.h:
(WebCore::CCSingleThreadProxy::maxPartialTextureUpdates):
* platform/graphics/chromium/cc/CCTextureUpdater.cpp:
(WebCore::CCTextureUpdater::append):
(WebCore):
(WebCore::CCTextureUpdater::appendPartial):
(WebCore::CCTextureUpdater::hasMoreUpdates):
(WebCore::CCTextureUpdater::update):
(WebCore::CCTextureUpdater::clear):
* platform/graphics/chromium/cc/CCTextureUpdater.h:
(CCTextureUpdater):
* platform/graphics/chromium/cc/CCThreadProxy.cpp:
(WTF):
(WebCore::CCThreadProxy::scheduledActionUpdateMoreResources):
(WebCore::CCThreadProxy::maxPartialTextureUpdates):
* platform/graphics/chromium/cc/CCThreadProxy.h:
(CCThreadProxy):

Source/WebKit/chromium:

Add CCLayerTreeHostTestAtomicCommitWithPartialUpdate and
TiledLayerChromiumTest.partialUpdates tests that verifies
atomicity of commits with partial updates.

* tests/CCLayerTreeHostTest.cpp:
(WTF::CCLayerTreeHostTestAtomicCommit::CCLayerTreeHostTestAtomicCommit):
(WTF::CCLayerTreeHostTestAtomicCommit::beginTest):
(WTF::CCLayerTreeHostTestAtomicCommit::commitCompleteOnCCThread):
(WTF::CCLayerTreeHostTestAtomicCommit::layout):
(CCLayerTreeHostTestAtomicCommit):
(CCLayerTreeHostTestAtomicCommitWithPartialUpdate):
(WTF::CCLayerTreeHostTestAtomicCommitWithPartialUpdate::CCLayerTreeHostTestAtomicCommitWithPartialUpdate):
(WTF::CCLayerTreeHostTestAtomicCommitWithPartialUpdate::beginTest):
(WTF::CCLayerTreeHostTestAtomicCommitWithPartialUpdate::commitCompleteOnCCThread):
(WTF::CCLayerTreeHostTestAtomicCommitWithPartialUpdate::drawLayersOnCCThread):
(WTF::CCLayerTreeHostTestAtomicCommitWithPartialUpdate::layout):
(WTF::CCLayerTreeHostTestAtomicCommitWithPartialUpdate::afterTest):
(WTF):
(WTF::TEST_F):
* tests/TiledLayerChromiumTest.cpp:
(WTF::FakeLayerTextureUpdater::Texture::Texture):
(WTF::FakeLayerTextureUpdater::Texture::updateRect):
(Texture):
(WTF::FakeLayerTextureUpdater::FakeLayerTextureUpdater):
(FakeLayerTextureUpdater):
(WTF::FakeLayerTextureUpdater::updateCount):
(WTF::FakeLayerTextureUpdater::clearUpdateCount):
(WTF::FakeLayerTextureUpdater::updateRect):
(WTF::FakeLayerTextureUpdater::createTexture):
(FakeTiledLayerChromium):
(WTF::TEST):
(WTF):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107177 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Replacement text should be available from the marker.
https://bugs.webkit.org/show_bug.cgi?id=77934

.:

Reviewed by Kent Tamura.

* Source/autotools/symbols.filter: Added an exporting symbol.

Source/WebCore:

Reviewed by Kent Tamura.

On spellchecking, TextCheckingResult can contain a replacement text
which is usable both for an automatic replacement and for showing a suggestion.

But when marking a misspelled word ragarding to returned
TextCheckingResult, Editor uses only the misspelled range data and
discards the replacement value. Then it asks the same value again
when showing suggestion/autocorrection.

It would be great if the marker holds the replacement text
and Editor can use it on suggesting a correction, without any re-request.
This is especially true in the case when it needs IPC messaging for spellchecking:
We can save one round-trip by this technique.

Here is actual change:

- Passed the replacement text to addMarker() for for misspelling markers.
  Note that this is done only for the unified checker path because legacy
  TextCheckerClient API doesn't provide such a replacement.
- Added an Internals API to retrieve a description text on a marker.

Test: editing/spelling/spelling-marker-description.html

* WebCore.exp.in:
* editing/Editor.cpp:
(WebCore::Editor::markAndReplaceFor):
* testing/Internals.cpp:
(WebCore::Internals::markerAt):
(WebCore):
(WebCore::Internals::markerRangeForNode):
(WebCore::Internals::markerDescriptionForNode):
* testing/Internals.h:
(WebCore):
(Internals):
* testing/Internals.idl:

Source/WebKit/chromium:

didFinishCheckingText() accidentally dropped the description text
given from the backend. Fixed to copy it on the API-to-Core
convertion. The test won't work without this fix.

Reviewed by Kent Tamura.

* src/WebTextCheckingCompletionImpl.cpp:
(WebKit::toCoreResults):

Source/WebKit/win:

Reviewed by Kent Tamura.

* WebKit.vcproj/WebKit_Cairo.def:
* WebKit.vcproj/WebKit_Cairo_debug.def:

Source/WebKit2:

Reviewed by Kent Tamura.

* win/WebKit2.def:
* win/WebKit2CFLite.def:

Tools:

Made the mock requestCheckingOfText() implementation to return
a replacement text for each misspelled word if available.

Reviewed by Kent Tamura.

* DumpRenderTree/chromium/WebViewHost.cpp:
(WebViewHost::finishLastTextCheck):

LayoutTests:

Reviewed by Kent Tamura.

* editing/spelling/spelling-marker-description-expected.txt: Added.
* editing/spelling/spelling-marker-description.html: Added.
* platform/gtk/Skipped:
* platform/mac/Skipped:
* platform/qt/Skipped:
* platform/win/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107176 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove more cruft now that HTMLIsIndexElement is gone
https://bugs.webkit.org/show_bug.cgi?id=77887

Reviewed by Darin Adler.

Just removing dead code, thus no tests.
There is likely more to remove after this.

* CMakeLists.txt:
* GNUmakefile.list.am:
* Target.pri:
* WebCore.gypi:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* html/HTMLInputElement.cpp:
* html/HTMLInputElement.h:
(HTMLInputElement):
* html/InputType.cpp:
(WebCore::createInputTypeFactoryMap):
* html/InputType.h:
(InputTypeNames):
* html/IsIndexInputType.cpp: Removed.
* html/IsIndexInputType.h: Removed.
* rendering/HitTestResult.cpp:
(WebCore::HitTestResult::isContentEditable):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107175 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[V8][Chromium] Race between worker accessing WebDatabase and frame closing down.
https://bugs.webkit.org/show_bug.cgi?id=78150
Handle case when the frame is shut down under us gracefully.

Patch by Dmitry Lomov <dslomov@chromium.org> on 2012-02-08
Reviewed by David Levin.

* src/DatabaseObserver.cpp:
(WebCore::DatabaseObserver::canEstablishDatabase): Handle WebWorkerBase::view returning 0.
* src/IDBFactoryBackendProxy.cpp:
(WebKit::IDBFactoryBackendProxy::openFromWorker): Handle WebWorkerBase::view returning 0.
* src/WebWorkerClientImpl.cpp:
(WebKit::WebWorkerClientImpl::allowFileSystem):
(WebKit::WebWorkerClientImpl::allowDatabase):
(WebKit::WebWorkerClientImpl::view):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107174 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Increased style sharing for elements with presentation attributes.
<http://webkit.org/b/78199>

Reviewed by Antti Koivisto.

When determining whether two elements can share style, we can do a lot better.
Instead of comparing the attribute maps for exact equality, do a property-by-property
comparison of the attributeStyle() and the additionalAttributeStyle() (if any.)

This increases our style sharing hit rate and shaves 100ms off of each cycle on
Chromium's "Moz" page cycler test on my machine.

The function that compares attribute styles has O(n^2) runtime in the worst case,
where n is the number of properties in the styles. However, given the low number of
properties found in attribute styles, this should be fine, and it doesn't seem to
heat up in profiles.

* css/CSSStyleSelector.cpp:
(WebCore::attributeStylesEqual):
(WebCore::CSSStyleSelector::canShareStyleWithElement):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107173 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix the caculation of preDelayFrames in DynamicsCompressorKernel
https://bugs.webkit.org/show_bug.cgi?id=78057

Patch by Raymond Liu <raymond.liu@intel.com> on 2012-02-08
Reviewed by Chris Rogers.

No new tests required.

* platform/audio/DynamicsCompressorKernel.cpp:
(WebCore::DynamicsCompressorKernel::setPreDelayTime):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107172 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][DRT] DumpRenderTreeQt should support --no-timeout and --timeout options
https://bugs.webkit.org/show_bug.cgi?id=78146

Patch by Jesus Sanchez-Palencia <jesus.palencia@openbossa.org> on 2012-02-08
Reviewed by Ryosuke Niwa.

Our DumpRenderTree should support --no-timeout and --timeout options in order
to be able to use run-perf-tests and have a Performance Bot.
This patch adds setTimeout and setShouldTimeout functions to our LayoutTestController
and the necessary code to handle such command line arguments to our DumpRenderTree.

* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::setTimeout):
(WebCore):
(WebCore::DumpRenderTree::setShouldTimeout):
* DumpRenderTree/qt/DumpRenderTreeQt.h:
(DumpRenderTree):
* DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::LayoutTestController):
(LayoutTestController::waitUntilDone):
(LayoutTestController::notifyDone):
* DumpRenderTree/qt/LayoutTestControllerQt.h:
(LayoutTestController::setTimeout):
(LayoutTestController::setShouldTimeout):
(LayoutTestController):
* DumpRenderTree/qt/main.cpp:
(isOption):
(printUsage):
(main):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107171 268f45cc-cd09-0410-ab3c-d52691b4dbfc

DOM mutations should not be delivered on worker threads
https://bugs.webkit.org/show_bug.cgi?id=77898

Reviewed by Dmitry Titov.

Source/WebCore:

In V8RecursionScope, only call WebKitMutationObserver::deliverAllMutations
if in a Document context.

This is accomplished through a change to V8Proxy::instrumentedCallFunction
(which now takes a Frame* instead of a Page*), requiring an update to all
callers of that function (accounting for the majority of files changed
in this patch).

Added ASSERT(isMainThread()) in a deliverAllMutations to confirm that
it's no longer called on worker threads, and in enqueueMutationRecord,
where the same global store of active observers is accessed.

See also http://crbug.com/112586, where the problem was initially
reported.

* bindings/v8/ScriptFunctionCall.cpp:
(WebCore::ScriptCallback::call):
* bindings/v8/V8NodeFilterCondition.cpp:
(WebCore::V8NodeFilterCondition::acceptNode):
* bindings/v8/V8Proxy.cpp:
(WebCore::V8Proxy::runScript):
(WebCore::V8Proxy::callFunction):
(WebCore::V8Proxy::instrumentedCallFunction):
* bindings/v8/V8Proxy.h:
(WebCore):
(V8Proxy):
* bindings/v8/V8RecursionScope.cpp:
(WebCore::V8RecursionScope::didLeaveScriptContext):
* bindings/v8/V8RecursionScope.h:
(WebCore):
(WebCore::V8RecursionScope::V8RecursionScope):
(V8RecursionScope):
(WebCore::V8RecursionScope::~V8RecursionScope):
* bindings/v8/V8WindowErrorHandler.cpp:
(WebCore::V8WindowErrorHandler::callListenerFunction):
* bindings/v8/custom/V8CustomVoidCallback.cpp:
(WebCore::invokeCallback):
* bindings/v8/custom/V8CustomXPathNSResolver.cpp:
(WebCore::V8CustomXPathNSResolver::lookupNamespaceURI):
* dom/WebKitMutationObserver.cpp:
(WebCore::WebKitMutationObserver::enqueueMutationRecord):
(WebCore::WebKitMutationObserver::deliverAllMutations):

Source/WebKit/chromium:

* src/WebDevToolsFrontendImpl.cpp:
(WebKit::WebDevToolsFrontendImpl::dispatchOnInspectorFrontend):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107170 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[BlackBerry] Upstream InspectorClientBlackBerry.{h, cpp}
https://bugs.webkit.org/show_bug.cgi?id=78082

Reviewed by Rob Buis.

Initial upstream, no new tests.

* blackberry/WebCoreSupport/InspectorClientBlackBerry.cpp: Added.
* blackberry/WebCoreSupport/InspectorClientBlackBerry.h: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107169 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Don't use the wheel event handler count to track if a page has horizontal scrollbars
https://bugs.webkit.org/show_bug.cgi?id=78192

Reviewed by Andreas Kling.

Source/WebCore:

Stop calling Document::didAddWheelEventHandler and Document::didRemoveWheelEventHandler when
adding and removing scrollbars.

* page/FrameView.cpp:
* page/FrameView.h:
(FrameView):
* rendering/RenderLayer.cpp:
* rendering/RenderLayer.h:

Source/WebKit2:

Prior to this change, we were incrementing and decrementing the wheel event handler count
whenever a scrollable area gained or lost a horizontal scrollbar, so we could use the count
to determine if Safari can handle horizontal wheel events directly or whether they have to be sent
to the web process first.

What this meant was that whenever a page had horizontal scrollbars we'd have to send all scroll wheel events
to the main thread instead of the scrolling thread, regardless of whether there were any wheel event handlers.

After this change, we traverse the tree of scrollable areas after every layout and check if any of them
have a horizontal scrollbar. (We still also check if there are wheel event handlers).

If traversing the tree after every layout is deemed to slow we can go back to caching the number of horizontal scrollbars
in a page, but the number of subframes in a page is usually very small and the number of other scrollable areas is even smaller.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::WebPageProxy):
(WebKit::WebPageProxy::willHandleHorizontalScrollEvents):
* UIProcess/WebPageProxy.h:
(WebKit::WebPageProxy::setCanShortCircuitHorizontalWheelEvents):
(WebPageProxy):
* UIProcess/WebPageProxy.messages.in:
* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::numWheelEventHandlersChanged):
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::dispatchDidLayout):
* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::WebPage):
(WebKit::WebPage::numWheelEventHandlersChanged):
(WebKit):
(WebKit::hasEnabledHorizontalScrollbar):
(WebKit::pageContainsAnyHorizontalScrollbars):
(WebKit::WebPage::recomputeShortCircuitHorizontalWheelEventsState):
* WebProcess/WebPage/WebPage.h:
(WebPage):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107168 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Roll Chromium DEPS
https://bugs.webkit.org/show_bug.cgi?id=78152

Patch by Scott Graham <scottmg@chromium.org> on 2012-02-08
Reviewed by Adam Barth.

* DEPS:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107166 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Rubber-stamped by Martin Robinson.

* gtk/common.py:
(get_build_path): Move test for current directory being a valid
directory up so it takes precedence over WebKitBuild/Release and
WebKitBuild/Debug when they exist as well.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107165 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[GTK] doc rebasing does not respect DESTDIR
https://bugs.webkit.org/show_bug.cgi?id=78177

Reviewed by Martin Robinson.

* GNUmakefile.am: Pass DESTDIR on to generate-gtkdoc, when
calling it for rebasing.
* gtk/generate-gtkdoc:
(get_common_options): Handle the new --virtual-root option.
* gtk/gtkdoc.py:
(GTKDoc.rebase_installed_docs): If a virtual-root has been given, pass
it on to gtkdoc-rebase as dest-dir, and prefix the htmldir with it.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107164 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Source/WebCore: Implement reverse animation direction

Implement reverse animation direction
https://bugs.webkit.org/show_bug.cgi?id=60525

Implement reverse and alternate-reverse direction.

Reviewed by Dean Jackson.

Tests: animations/animation-direction-alternate-reverse.html
       animations/animation-direction-reverse.html
       animations/fill-mode-reverse.html

* css/CSSParser.cpp:
(WebCore::CSSParser::parseAnimationDirection):
* css/CSSStyleSelector.cpp:
(WebCore::CSSStyleSelector::mapAnimationDirection):
* css/CSSValueKeywords.in:
* page/WebKitAnimation.cpp:
(WebCore::WebKitAnimation::direction):
* page/WebKitAnimation.h:
* page/animation/AnimationBase.cpp:
(WebCore::AnimationBase::fractionalTime):
* platform/animation/Animation.h:
(Animation):
* platform/graphics/texmap/TextureMapperAnimation.cpp:
(WebCore):
(WebCore::shouldReverseAnimationValue):
(WebCore::normalizedAnimationValue):

LayoutTests: Implement reverse animation direction
https://bugs.webkit.org/show_bug.cgi?id=60525

Add test for the new animation-direction values.

Reviewed by Dean Jackson.

* animations/animation-direction-alternate-reverse-expected.txt: Added.
* animations/animation-direction-alternate-reverse.html: Added.
* animations/animation-direction-reverse-expected.txt: Added.
* animations/animation-direction-reverse.html: Added.
* animations/fill-mode-reverse-expected.txt: Added.
* animations/fill-mode-reverse.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107162 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Avoid creating a temporary GraphicsContext3D if someone requests the WebView's GraphicsContext3D before initialization is complete
https://bugs.webkit.org/show_bug.cgi?id=78154

Patch by James Robinson <jamesr@chromium.org> on 2012-02-08
Reviewed by Kenneth Russell.

* platform/graphics/chromium/cc/CCSingleThreadProxy.cpp:
(WebCore::CCSingleThreadProxy::context):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107161 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Source/WebCore: https://bugs.webkit.org/show_bug.cgi?id=78157

Make multi-column layout work with line grids that are outside of the multi-column
block.

Reviewed by Dan Bernstein.

Added new tests in fast/line-grid.

* rendering/LayoutState.cpp:
(WebCore::LayoutState::LayoutState):
(WebCore::LayoutState::propagateLineGridInfo):
(WebCore::LayoutState::establishLineGrid):
(WebCore):
(WebCore::LayoutState::computeLineGridPaginationOrigin):
* rendering/LayoutState.h:
(WebCore::LayoutState::LayoutState):
(WebCore::LayoutState::lineGrid):
(WebCore::LayoutState::lineGridOffset):
(WebCore::LayoutState::lineGridPaginationOrigin):
(WebCore::LayoutState::needsBlockDirectionLocationSetBeforeLayout):
(LayoutState):
* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::layoutBlockChildren):
* rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlock::layoutInlineChildren):
* rendering/RenderView.h:
(WebCore::RenderView::pushLayoutState):
* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::lineGridSnapAdjustment):

LayoutTests: https://bugs.webkit.org/show_bug.cgi?id=78157

Make multi-column layout work with line grids that are outside of the multi-column
block.

Reviewed by Dan Bernstein.

* fast/line-grid/line-grid-into-columns.html: Added.
* platform/mac/fast/line-grid/line-grid-into-columns-expected.png: Added.
* platform/mac/fast/line-grid/line-grid-into-columns-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107160 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Improve http-equiv content-language parsing
https://bugs.webkit.org/show_bug.cgi?id=77724

Patch by Matt Falkenhagen <falken@chromium.org> on 2012-02-08
Reviewed by Alexey Proskuryakov.

Add more tests for mapping http-equiv content-language to -webkit-locale. The desired behavior of some
of these is not yet decided, so some of the tests currently fail and the expectations may change. See comments in bug and in the tests.

* fast/text/content-language-case-insensitivity-expected.txt:
* fast/text/content-language-case-insensitivity.html: Style fixes.
* fast/text/content-language-empty-expected.txt: Added.
* fast/text/content-language-empty.html: Added.
* fast/text/content-language-late-expected.txt: Added.
* fast/text/content-language-late.html: Added.
* fast/text/content-language-mapped-to-webkit-locale-expected.txt:
* fast/text/content-language-mapped-to-webkit-locale.html: Style fixes.
* fast/text/content-language-multiple-expected.txt: Added.
* fast/text/content-language-multiple.html: Added.
* fast/text/content-language-no-content-expected.txt: Added.
* fast/text/content-language-no-content.html: Added.
* fast/text/content-language-only-whitespace-expected.txt: Added.
* fast/text/content-language-only-whitespace.html: Added.
* fast/text/content-language-with-whitespace-expected.txt: Added.
* fast/text/content-language-with-whitespace.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107159 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[BlackBerry] Remove EditCommandBlackBerry.cpp from build system
https://bugs.webkit.org/show_bug.cgi?id=78078

Reviewed by Rob Buis.

WebKit/blackberry/WebCoreSupport/EditCommandBlackBerry.cpp which hasn't
been upstreamed has been removed from the BlackBerry porting internally.
So we should remove it from the build system.

* blackberry/CMakeListsBlackBerry.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107158 268f45cc-cd09-0410-ab3c-d52691b4dbfc

GeneratorGeneratedImage::drawPattern does not factor in its destination context's scale when generating its image tiles

https://bugs.webkit.org/show_bug.cgi?id=67729
<rdar://problem/10087050>

Reviewed by Beth Dakin.

No new tests, current pixel tests will cover this. Though some pixel results might improve to become less pixel-y.

* platform/graphics/GeneratorGeneratedImage.cpp:
(WebCore::GeneratorGeneratedImage::draw): Updated context to destContext for consistency.
(WebCore::GeneratorGeneratedImage::drawPattern): Taught drawPattern about the destination
scale factor to avoid having low-res generated images such as gradients in certain cases.
* platform/graphics/GraphicsContext.cpp:
(WebCore::GraphicsContext::createCompatibleBuffer): Have the image buffer match the
context acceleration setting as well.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107157 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove Python 2.5 support from WebKit
https://bugs.webkit.org/show_bug.cgi?id=71593

Reviewed by Eric Seidel.

This is the last vestige of our Python 2.5 support.

* Scripts/webkitpy/tool/commands/queues_unittest.py:
(AbstractQueueTest.test_log_from_script_error_for_upload):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107156 268f45cc-cd09-0410-ab3c-d52691b4dbfc

version_check.py should inform users that we don't support Python 2.5 anymore
https://bugs.webkit.org/show_bug.cgi?id=78179

Reviewed by Eric Seidel.

We don't support Python 2.5 anymore.

* Scripts/webkitpy/common/version_check.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107155 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove empty directory.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107154 268f45cc-cd09-0410-ab3c-d52691b4dbfc

test_configuration.py shouldn't re-implement itertools
https://bugs.webkit.org/show_bug.cgi?id=78178

Reviewed by Eric Seidel.

* Scripts/webkitpy/layout_tests/models/test_configuration.py:
(TestConfigurationConverter.to_specifiers_list):
(TestConfigurationConverter.to_specifiers_list.try_collapsing):
(TestConfigurationConverter.to_specifiers_list.try_abbreviating):
(TestConfigurationConverter):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107153 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening.

* platform/mac-future/svg/carto.net: Added.
* platform/mac-future/svg/carto.net/scrollbar-expected.png: Added.
* platform/mac/svg/carto.net/scrollbar-expected.png: Removed.
* platform/chromium-mac-leopard/svg/carto.net/scrollbar-expected.png:
* platform/chromium-mac-leopard/svg/carto.net/scrollbar-expected.txt: Added.
Updated the baseline here, thanks to the text on the button...

* platform/chromium/test_expectations.txt:
Marked one test as failing as some output don't make much sense.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107152 268f45cc-cd09-0410-ab3c-d52691b4dbfc

webkitpy should reply upon the multiprocessing package existing
https://bugs.webkit.org/show_bug.cgi?id=78176

Reviewed by Eric Seidel.

Now that we don't support Python 2.5, this import can't fail.

* Scripts/webkitpy/layout_tests/controllers/manager_worker_broker.py:
(get):
(_Process):
(_Process.__init__):
(_Process.run):
* Scripts/webkitpy/layout_tests/controllers/manager_worker_broker_unittest.py:
(FunctionTests.test_get__processes):
(MultiProcessBrokerTests.setUp):
* Scripts/webkitpy/layout_tests/port/base.py:
(Port.__init__):
(Port.default_worker_model):
* Scripts/webkitpy/layout_tests/port/chromium_mac.py:
(ChromiumMacPort.check_build):
* Scripts/webkitpy/layout_tests/port/port_testcase.py:
(PortTestCase.test_default_worker_model):
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107151 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed typo fix that makes docs build again for the gtk2-based
library.

* gtk/generate-gtkdoc:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107150 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Simplify and correct mutation delivery timing for JSC
https://bugs.webkit.org/show_bug.cgi?id=78172

Reviewed by Adam Barth.

Instead of keeping a static recursion counter in JSMainThreadExecState,
simply wait for a state change from non-null ExecState to null ExecState.
Because s_mainThreadState is initially null, this equivalent to
waiting for s_recursionLevel to rewind to zero.

This also properly handles the usage of JSMainThreadNullState (and
does not do mutation delivery), since that class is only used by
non-JS bindings. Now fast/mutation/end-of-task-delivery.html properly
fails, whereas it was passing before due to usage of the ObjC DOM API
from DumpRenderTree.

* bindings/js/JSMainThreadExecState.cpp:
(WebCore):
* bindings/js/JSMainThreadExecState.h: Added a comment explaining the purpose of JSMainThreadNullState.
(WebCore::JSMainThreadExecState::JSMainThreadExecState):
(WebCore::JSMainThreadExecState::~JSMainThreadExecState):
(JSMainThreadExecState):
(WebCore):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107149 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove [ConvertToString] from CodeGeneratorCPP.pm and rename
it to [ObjCImplementedAsUnsignedLong]
https://bugs.webkit.org/show_bug.cgi?id=78100

Reviewed by Eric Seidel.

Now [ConvertToString] is used by ObjC's HTMLElement.size only.
This patch removes [ConvertToString] code from CodeGeneratorCPP.pm,
and renames [ConvertToString] to [ObjCImplementedAsUnsignedLong].

No tests. No change in behavior.

* bindings/scripts/CodeGeneratorCPP.pm:
(GenerateImplementation):
* bindings/scripts/CodeGeneratorObjC.pm:
(GenerateImplementation):
* html/HTMLInputElement.idl: Removed FIXME comment, because the latest spec
says HTMLInputElement.size should be unsigned long.
(http://www.whatwg.org/specs/web-apps/current-work/multipage/the-input-element.html#the-input-element)

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107148 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Don't re-implement ZipFile.extractall
https://bugs.webkit.org/show_bug.cgi?id=78173

Reviewed by Eric Seidel.

We can use ZipFile.extractall now that we don't support Python 2.5.

* Scripts/webkitpy/common/system/autoinstall.py:
(AutoInstaller._extract_targz):
(AutoInstaller._unzip):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107147 268f45cc-cd09-0410-ab3c-d52691b4dbfc

REGRESSION(r84194): Build fails when turning on ENABLE_FILE_SYSTEM
https://bugs.webkit.org/show_bug.cgi?id=78088

Reviewed by Antonio Gomes.

WebCore::getDOMObjectWrapper was changed to WebCore::wrap in r84194.
Adapt to this change in JSEntryCustom.cpp and JSEntrySyncCustom.cpp.

Build fix, no new tests.

* bindings/js/JSEntryCustom.cpp:
(WebCore::toJS):
* bindings/js/JSEntrySyncCustom.cpp:
(WebCore::toJS):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107146 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove the ospath compat shim from webkitpy
https://bugs.webkit.org/show_bug.cgi?id=78170

Reviewed by Eric Seidel.

We no longer need this compat shim now that we don't support Python 2.5.

* Scripts/webkitpy/common/checkout/scm/git.py:
* Scripts/webkitpy/common/checkout/scm/svn.py:
* Scripts/webkitpy/common/system/filesystem.py:
(FileSystem.relpath):
* Scripts/webkitpy/common/system/filesystem_mock.py:
(MockFileSystem.relpath):
* Scripts/webkitpy/common/system/ospath.py: Removed.
* Scripts/webkitpy/common/system/ospath_unittest.py: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107145 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Build fix attempt after r107053.

* Scripts/webkitpy/performance_tests/perftestsrunner_unittest.py:
(test_collect_tests):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107144 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Chromium] Fix opaque tracking for box shadows and non-composited child elements
https://bugs.webkit.org/show_bug.cgi?id=78073

Patch by Dana Jansens <danakj@chromium.org> on 2012-02-08
Reviewed by Stephen White.

Source/WebCore:

Tests: compositing/culling/scrolled-within-boxshadow.html
       compositing/culling/translated-boxshadow.html

Unit tests: PlatformContextSkiaTest.cpp

When painting a box shadow, a filter is applied to the skia canvas, that can make
pixels painted with an opaque color end up non-opaque. So consider image/mask/color
filters when deciding if a paint is opaque.

Also when painting the background of an element with a box shadow, the background is
painted with a transform on the skia canvas based on the size of the box shadow. This
transform needs to be considered when tracking an opaque paint.

However, when a layer's contentRect position is non-zero, we translate the GraphicsContext
to put the contentRect at 0,0 in the skia canvas. For tracking opaque regions in the resulting
layer, we need to unto this translation. Scaling can also occur which we must undo. So we pass
the transform in to PlatformContextSkia to go from the SkCanvas back to the layer's content
coordinate space. Opaque paints can then be tracked in the layer's content space rather than
in the skia canvas space.

* platform/graphics/chromium/BitmapCanvasLayerTextureUpdater.cpp:
(WebCore::BitmapCanvasLayerTextureUpdater::prepareToUpdate):
* platform/graphics/chromium/CanvasLayerTextureUpdater.cpp:
(WebCore::CanvasLayerTextureUpdater::paintContents):
* platform/graphics/chromium/CanvasLayerTextureUpdater.h:
(WebCore):
(CanvasLayerTextureUpdater):
* platform/graphics/chromium/SkPictureCanvasLayerTextureUpdater.cpp:
(WebCore::SkPictureCanvasLayerTextureUpdater::prepareToUpdate):
* platform/graphics/skia/OpaqueRegionSkia.cpp:
(WebCore::paintIsOpaque):
(WebCore::OpaqueRegionSkia::didDrawRect):
(WebCore::OpaqueRegionSkia::didDrawPath):
(WebCore::OpaqueRegionSkia::didDrawPoints):
(WebCore::OpaqueRegionSkia::didDrawBounded):
(WebCore::OpaqueRegionSkia::didDraw):
(WebCore::OpaqueRegionSkia::markRectAsOpaque):
(WebCore::OpaqueRegionSkia::markRectAsNonOpaque):
* platform/graphics/skia/OpaqueRegionSkia.h:
(WebCore):
(OpaqueRegionSkia):
* platform/graphics/skia/PlatformContextSkia.cpp:
(WebCore::PlatformContextSkia::didDrawRect):
(WebCore::PlatformContextSkia::didDrawPath):
(WebCore::PlatformContextSkia::didDrawPoints):
(WebCore::PlatformContextSkia::didDrawBounded):
* platform/graphics/skia/PlatformContextSkia.h:
(PlatformContextSkia):
(WebCore::PlatformContextSkia::setOpaqueRegionTransform):

Source/WebKit/chromium:

* tests/PlatformContextSkiaTest.cpp:
(WebCore::TEST):
(WebCore):

LayoutTests:

* compositing/culling/scrolled-within-boxshadow-expected.png: Added.
* compositing/culling/scrolled-within-boxshadow-expected.txt: Added.
* compositing/culling/scrolled-within-boxshadow.html: Added.
* compositing/culling/translated-boxshadow-expected.png: Added.
* compositing/culling/translated-boxshadow-expected.txt: Added.
* compositing/culling/translated-boxshadow.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107143 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove simplejson because it's no longer used by webkitpy
https://bugs.webkit.org/show_bug.cgi?id=78164

Reviewed by Eric Seidel.

All the importers of this code were removed in a previous patch.

* Scripts/webkitpy/thirdparty/simplejson: Removed.
* Scripts/webkitpy/thirdparty/simplejson/LICENSE.txt: Removed.
* Scripts/webkitpy/thirdparty/simplejson/README.txt: Removed.
* Scripts/webkitpy/thirdparty/simplejson/__init__.py: Removed.
* Scripts/webkitpy/thirdparty/simplejson/_speedups.c: Removed.
* Scripts/webkitpy/thirdparty/simplejson/decoder.py: Removed.
* Scripts/webkitpy/thirdparty/simplejson/encoder.py: Removed.
* Scripts/webkitpy/thirdparty/simplejson/jsonfilter.py: Removed.
* Scripts/webkitpy/thirdparty/simplejson/ordered_dict.py: Removed.
* Scripts/webkitpy/thirdparty/simplejson/scanner.py: Removed.
* Scripts/webkitpy/thirdparty/simplejson/tool.py: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107141 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening.

* platform/chromium-mac-snowleopard/platform/chromium/compositing/lost-compositor-context-with-video-expected.png: Added.
Updated this baseline. It looks like an opacity difference of 1 that we could't really explain.

* platform/chromium-mac-snowleopard/tables/mozilla/bugs/bug120364-expected.png: Added.
* platform/chromium-mac-snowleopard/tables/mozilla/bugs/bug22019-expected.png: Added.
* platform/chromium-mac-snowleopard/tables/mozilla/bugs/bug23151-expected.png: Added.
* platform/chromium-mac-snowleopard/tables/mozilla/bugs/bug46480-1-expected.png: Added.
* platform/chromium-mac-snowleopard/tables/mozilla/bugs/bug46480-2-expected.png: Added.
* platform/chromium-mac-snowleopard/tables/mozilla/bugs/bug56405-expected.png: Added.
* platform/chromium-mac-snowleopard/tables/mozilla/marvin/x_td_nowrap-expected.png: Added.
* platform/chromium-mac-snowleopard/tables/mozilla_expected_failures/bugs/bug67915-2-expected.png: Added.
* platform/chromium/tables/mozilla/bugs/bug56405-expected.txt: Removed.
Rebaselined those tests on snowleopard as we started taking up the new snow-leopard baselines from r107122.

* platform/chromium/test_expectations.txt:
Mark compositing/iframes/scrolling-iframe.html as flaky as seen on a snow-leopard or windows (debug) bot near you.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107140 268f45cc-cd09-0410-ab3c-d52691b4dbfc

webkitpy still imports simplejson
https://bugs.webkit.org/show_bug.cgi?id=78161

Reviewed by Eric Seidel.

We no longer support Python 2.5, so we can rely upon Python's built-in
JSON package.

* Scripts/webkitpy/common/net/buildbot/buildbot.py:
* Scripts/webkitpy/common/net/resultsjsonparser.py:
* Scripts/webkitpy/layout_tests/controllers/test_expectations_editor.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
* Scripts/webkitpy/layout_tests/layout_package/json_results_generator_unittest.py:
* Scripts/webkitpy/layout_tests/models/test_expectations.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
* Scripts/webkitpy/style/checkers/jsonchecker.py:
* Scripts/webkitpy/tool/servers/gardeningserver_unittest.py:
* Scripts/webkitpy/tool/servers/rebaselineserver_unittest.py:
* Scripts/webkitpy/tool/servers/reflectionhandler.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107139 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[GTK] Fails to build docs with non-standard build directories
https://bugs.webkit.org/show_bug.cgi?id=78118

* gtk/common.py:
(get_build_path): add comment to document changes done in r107098,
as requested by the reviewer and forgotten by me.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107138 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove [CustomHeader] from CanvasPixelArray and rename [CustomHeader] to [JSCustomHeader]
https://bugs.webkit.org/show_bug.cgi?id=78089

Reviewed by Adam Barth.

This patch removes [CustomHeader] from CanvasPixelArray.idl, since CanvasPixelArrayCustom.h
does not exist. (The reason why missing CanvasPixelArrayCustom.h has not caused build failure
is that [CustomHeader] has been JSC-specific and JSC has not enabled CanvasPixelArray.)
Also, this patch renames [CustomHeader] to [JSCustomHeader], since whether a given class
should have custom header or not will depend on JavaScript bindings.

No tests. No change in behavior.

* bindings/scripts/CodeGeneratorJS.pm:
(GenerateHeader):
* dom/Node.idl:
* html/canvas/CanvasPixelArray.idl:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107137 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Stop importing with_statement from the mysterious future
https://bugs.webkit.org/show_bug.cgi?id=78156

Reviewed by Eric Seidel.

The future is now!  (Well, at least 2008, when Python 2.6 was released,
is no longer the __future__.)

* Scripts/webkitpy/common/checkout/changelog_unittest.py:
* Scripts/webkitpy/common/checkout/checkout_unittest.py:
* Scripts/webkitpy/common/checkout/scm/scm_unittest.py:
* Scripts/webkitpy/common/net/credentials_unittest.py:
* Scripts/webkitpy/common/net/file_uploader.py:
* Scripts/webkitpy/common/newstringio_unittest.py:
* Scripts/webkitpy/common/system/autoinstall.py:
* Scripts/webkitpy/common/system/crashlogs.py:
* Scripts/webkitpy/common/system/fileset.py:
* Scripts/webkitpy/common/system/filesystem.py:
* Scripts/webkitpy/common/system/filesystem_unittest.py:
* Scripts/webkitpy/common/system/path.py:
* Scripts/webkitpy/common/thread/threadedmessagequeue.py:
* Scripts/webkitpy/layout_tests/controllers/manager.py:
* Scripts/webkitpy/layout_tests/port/base.py:
(of):
* Scripts/webkitpy/layout_tests/port/gtk.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
* Scripts/webkitpy/style/filereader_unittest.py:
* Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:
* Scripts/webkitpy/to_be_moved/update_webgl_conformance_tests.py:
* Scripts/webkitpy/tool/commands/analyzechangelog.py:
* Scripts/webkitpy/tool/commands/queues.py:
* Scripts/webkitpy/tool/servers/reflectionhandler.py:
* Scripts/webkitpy/tool/steps/preparechangelogforrevert_unittest.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107136 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Dispatch updateViewportArguments(), when Document is finished
restoring from page cache.

https://bugs.webkit.org/show_bug.cgi?id=77943

Patch by Zalan Bujtas <zbujtas@gmail.com> on 2012-02-08
Reviewed by Kenneth Rohde Christiansen.

Move updateViewportArguments() call from setPageInCache() to
documentDidResumeFromPageCache() to ensure, that the Document is
fully resumed from the page cache and attached to the mainframe,
when the viewport arguments are updated.

No tests. No change in behaviour.

* dom/Document.cpp:
(WebCore::Document::setInPageCache):
(WebCore::Document::documentDidResumeFromPageCache):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107135 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening.

* platform/chromium-mac-leopard/svg/repaint/inner-svg-change-viewBox-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/repaint/inner-svg-change-viewBox-expected.png: Added.
Add those 2 new baselines.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107134 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Add a layout test for losing the compositor context with a video playing
https://bugs.webkit.org/show_bug.cgi?id=78060

Reviewed by Eric Carlson.

Adds a test to prevent regressions with video playback when we lose the compositor context. This test loads up a
video, forces a frame to be displayed, simulates a lost context, then finishes the test (which causes another
frame to be displayed).

Prior to r106610 this test times out in debug and crashes in ASAN.

* platform/chromium/compositing/lost-compositor-context-with-video-expected.png: Added.
* platform/chromium/compositing/lost-compositor-context-with-video-expected.txt: Added.
* platform/chromium/compositing/lost-compositor-context-with-video.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107132 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Whoops, fix the build.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107131 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Remove incorrect early exit in CCDamageTracker
https://bugs.webkit.org/show_bug.cgi?id=76924

Reviewed by James Robinson.

Source/WebCore:

New unit test added to CCDamageTrackerTest.cpp

This patch does three things: (1) adds unit test that demonstrates
that early exiting in CCDamageTracker is wrong, (2) removes the
early exit and cleans up the surrounding code, and (3) re-names
several functions in CCDamageTracker so that state updating is
implied by the name, and not just a bad side-effect of the functions.

* platform/graphics/chromium/cc/CCDamageTracker.cpp:
(WebCore::CCDamageTracker::updateDamageTrackingState):
(WebCore::CCDamageTracker::trackDamageFromActiveLayers):
(WebCore::CCDamageTracker::trackDamageFromSurfaceMask):
(WebCore::CCDamageTracker::trackDamageFromLeftoverRects):
* platform/graphics/chromium/cc/CCDamageTracker.h:
(CCDamageTracker):
* platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp:
(WebCore::CCLayerTreeHostImpl::trackDamageForAllSurfaces):

Source/WebKit/chromium:

* tests/CCDamageTrackerTest.cpp:
(WebKitTests::emulateDrawingOneFrame):
(WebKitTests::TEST_F):
(WebKitTests):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107130 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[BlackBerry] Upstream DumpRenderTreeBlackBerry
https://bugs.webkit.org/show_bug.cgi?id=78042

Reviewed by Antonio Gomes.

Upstream more parts of our DRT implementation.

* DumpRenderTree/blackberry/AccessibilityControllerBlackBerry.cpp: Added.
* DumpRenderTree/blackberry/AccessibilityUIElementBlackBerry.cpp: Added.
* DumpRenderTree/blackberry/EventSender.cpp: Added.
* DumpRenderTree/blackberry/EventSender.h: Added.
* DumpRenderTree/blackberry/GCControllerBlackBerry.cpp: Added.
* DumpRenderTree/blackberry/WorkQueueItemBlackBerry.cpp: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107129 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix issue encountered while debugging stacktraces
https://bugs.webkit.org/show_bug.cgi?id=78147

Reviewed by Gavin Barraclough.

Debugging is easier if we always ensure that we have a non-null
inferred name.

* runtime/Executable.cpp:
(JSC::FunctionExecutable::FunctionExecutable):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107128 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Check that we can make the SharedGraphicsContext3D current before returning
https://bugs.webkit.org/show_bug.cgi?id=78142

Reviewed by Stephen White.

If we can't make the context current, we can't use it.

* platform/graphics/gpu/SharedGraphicsContext3D.cpp:
(WebCore::SharedGraphicsContext3D::get):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107127 268f45cc-cd09-0410-ab3c-d52691b4dbfc

updateTopCallframe in the baseline JIT doesn't provide enough information to the stubs
https://bugs.webkit.org/show_bug.cgi?id=78145

Reviewed by Gavin Barraclough.

Fix the updateTopCallFrame helper to store additional information
that becomes necessary when we are trying to provide more stack
frame information.

* interpreter/CallFrame.h:
(JSC::ExecState::bytecodeOffsetForBaselineJIT):
(ExecState):
* jit/JIT.cpp:
(JSC::JIT::privateCompile):
* jit/JIT.h:
(JSC::JIT::compileGetByIdProto):
(JSC::JIT::compileGetByIdSelfList):
(JSC::JIT::compileGetByIdProtoList):
(JSC::JIT::compileGetByIdChainList):
(JSC::JIT::compileGetByIdChain):
(JSC::JIT::compilePutByIdTransition):
(JIT):
* jit/JITInlineMethods.h:
(JSC::JIT::updateTopCallFrame):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107126 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add support for pinch gesture processing in the MT compositor.
https://bugs.webkit.org/show_bug.cgi?id=77804

Patch by Sadrul Habib Chowdhury <sadrul@chromium.org> on 2012-02-08
Reviewed by James Robinson.

* public/WebInputEvent.h:
* src/WebCompositorInputHandlerImpl.cpp:
(WebKit::WebCompositorInputHandlerImpl::WebCompositorInputHandlerImpl):
(WebKit::WebCompositorInputHandlerImpl::handleInputEvent):
* src/WebCompositorInputHandlerImpl.h:
(WebCompositorInputHandlerImpl):
* src/WebInputEventConversion.cpp:
(WebKit::PlatformGestureEventBuilder::PlatformGestureEventBuilder):
* src/WebPopupMenuImpl.cpp:
(WebKit::WebPopupMenuImpl::handleInputEvent):
* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::handleInputEvent):
* tests/WebCompositorInputHandlerImplTest.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107125 268f45cc-cd09-0410-ab3c-d52691b4dbfc

check-webkit-style failing with "Path does not exist."
https://bugs.webkit.org/show_bug.cgi?id=77873

Reviewed by Ojan Vafai.

This change fixes the way the style checker determines which
Port class to use for a given test_expectations.txt path; the
previous version used a heuristic that didn't really work in the
first place.

* Scripts/webkitpy/style/checkers/test_expectations.py:
(TestExpectationsChecker._determine_port_from_expectations_path):
(TestExpectationsChecker.__init__):
* Scripts/webkitpy/style/checkers/test_expectations_unittest.py:
(TestExpectationsTestCase._expect_port_for_expectations_path):
(TestExpectationsTestCase.test_determine_port_from_expectations_path):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107124 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Crash in Node::normalize.
https://bugs.webkit.org/show_bug.cgi?id=78135

Reviewed by Ryosuke Niwa.

No new tests. Original testcase does not reduce to manageable
extent.

* dom/Node.cpp:
(WebCore::Node::normalize):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107123 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Update pixel table test expectations for mac to match new DRT scrollbars
https://bugs.webkit.org/show_bug.cgi?id=76938

Rubber stamped by Eric Seidel.

* platform/mac/tables/mozilla/bugs/bug101674-expected.png:
* platform/mac/tables/mozilla/bugs/bug10296-1-expected.png:
* platform/mac/tables/mozilla/bugs/bug113235-1-expected.png:
* platform/mac/tables/mozilla/bugs/bug113235-3-expected.png:
* platform/mac/tables/mozilla/bugs/bug11944-expected.png:
* platform/mac/tables/mozilla/bugs/bug120364-expected.png:
* platform/mac/tables/mozilla/bugs/bug131020-expected.png:
* platform/mac/tables/mozilla/bugs/bug131020_iframe-expected.png:
* platform/mac/tables/mozilla/bugs/bug137388-2-expected.png:
* platform/mac/tables/mozilla/bugs/bug149275-1-expected.png:
* platform/mac/tables/mozilla/bugs/bug22019-expected.png:
* platform/mac/tables/mozilla/bugs/bug23151-expected.png:
* platform/mac/tables/mozilla/bugs/bug27038-2-expected.png:
* platform/mac/tables/mozilla/bugs/bug29314-expected.png:
* platform/mac/tables/mozilla/bugs/bug32205-2-expected.png:
* platform/mac/tables/mozilla/bugs/bug38916-expected.png:
* platform/mac/tables/mozilla/bugs/bug43039-expected.png:
* platform/mac/tables/mozilla/bugs/bug43854-1-expected.png:
* platform/mac/tables/mozilla/bugs/bug44505-expected.png:
* platform/mac/tables/mozilla/bugs/bug46480-1-expected.png:
* platform/mac/tables/mozilla/bugs/bug46480-2-expected.png:
* platform/mac/tables/mozilla/bugs/bug50695-1-expected.png:
* platform/mac/tables/mozilla/bugs/bug56405-expected.png:
* platform/mac/tables/mozilla/bugs/bug67915-1-expected.png:
* platform/mac/tables/mozilla/bugs/bug73321-expected.png:
* platform/mac/tables/mozilla/bugs/bug92143-expected.png:
* platform/mac/tables/mozilla/bugs/bug96334-expected.png:
* platform/mac/tables/mozilla/bugs/bug96343-expected.png:
* platform/mac/tables/mozilla/marvin/x_table_bgcolor_name-expected.png:
* platform/mac/tables/mozilla/marvin/x_table_bgcolor_rgb-expected.png:
* platform/mac/tables/mozilla/marvin/x_td_bgcolor_name-expected.png:
* platform/mac/tables/mozilla/marvin/x_td_bgcolor_rgb-expected.png:
* platform/mac/tables/mozilla/marvin/x_td_height-expected.png:
* platform/mac/tables/mozilla/marvin/x_td_nowrap-expected.png:
* platform/mac/tables/mozilla/marvin/x_th_bgcolor_name-expected.png:
* platform/mac/tables/mozilla/marvin/x_th_bgcolor_rgb-expected.png:
* platform/mac/tables/mozilla/marvin/x_th_height-expected.png:
* platform/mac/tables/mozilla/marvin/x_th_nowrap-expected.png:
* platform/mac/tables/mozilla/marvin/x_tr_bgcolor_name-expected.png:
* platform/mac/tables/mozilla/marvin/x_tr_bgcolor_rgb-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug101759-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug1055-2-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug131020-3-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug14007-1-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug19526-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug220653-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug22122-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug2479-5-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug32205-1-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug67915-2-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug80762-2-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug85016-expected.png:
* platform/mac/tables/mozilla_expected_failures/bugs/bug89315-expected.png:
* platform/mac/tables/mozilla_expected_failures/core/backgrounds-expected.png:
* platform/mac/tables/mozilla_expected_failures/core/col_span2-expected.png:
* platform/mac/tables/mozilla_expected_failures/core/columns-expected.png:
* platform/mac/tables/mozilla_expected_failures/core/conflicts-expected.png:
* platform/mac/tables/mozilla_expected_failures/core/standards1-expected.png:
* platform/mac/tables/mozilla_expected_failures/other/empty_cells-expected.png:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107122 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Make WebGL context current early to check validity
https://bugs.webkit.org/show_bug.cgi?id=78141

Patch by Antoine Labour <piman@chromium.org> on 2012-02-08
Reviewed by James Robinson.

Covered by existing tests

* html/canvas/WebGLRenderingContext.cpp:
(WebCore):
(WebCore::WebGLRenderingContext::create):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107121 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Disable root layer clears on release builds.
https://bugs.webkit.org/show_bug.cgi?id=77478

Patch by Jonathan Backer <backer@chromium.org> on 2012-02-08
Reviewed by James Robinson.

Source/WebCore:

* platform/graphics/chromium/LayerRendererChromium.cpp:
(WebCore::LayerRendererChromium::clearRenderSurface):
(WebCore::LayerRendererChromium::drawRenderPass):
* platform/graphics/chromium/LayerRendererChromium.h:
(LayerRendererChromium):

LayoutTests:

* platform/chromium-gpu-linux/fast/canvas/canvas-text-alignment-expected.png:
* platform/chromium-gpu-linux/media/video-zoom-expected.png:
* platform/chromium-linux/compositing/direct-image-compositing-expected.png:
* platform/chromium-linux/compositing/geometry/ancestor-overflow-change-expected.png:
* platform/chromium-linux/compositing/geometry/fixed-in-composited-expected.png:
* platform/chromium-linux/compositing/geometry/tall-page-composited-expected.png:
* platform/chromium-linux/compositing/masks/masked-ancestor-expected.png:
* platform/chromium-linux/compositing/masks/multiple-masks-expected.png:
* platform/chromium-linux/compositing/masks/simple-composited-mask-expected.png:
* platform/chromium-linux/compositing/scaling/tiled-layer-recursion-expected.png:
* platform/chromium-linux/platform/chromium/compositing/layout-width-change-expected.png:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107120 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add a content shadow layer to the render layer compositor
https://bugs.webkit.org/show_bug.cgi?id=78133
<rdar://problem/10797742>

Reviewed by Beth Dakin.

Have the render layer compositor optionally create a content shadow layer,
and add a ScrollbarTheme::setUpContentShadowLayer member function that subclasses
can use to set content shadow properties.

* platform/mac/ScrollbarThemeMac.mm:
(WebCore::ScrollbarThemeMac::setUpContentShadowLayer):
Set the layer properties once, and set the shadow path on every call, since we know that this
function will be called every time the size of the content shadow layer changes.

* rendering/RenderLayerCompositor.cpp:
(WebCore::RenderLayerCompositor::updateRootLayerPosition):
Reposition the content shadow layer, and call ScrollbarTheme::setUpContentShadowLayer if the size changes.

(WebCore::RenderLayerCompositor::requiresContentShadowLayer):
Add new helper function.

(WebCore::RenderLayerCompositor::updateOverflowControlsLayers):
Create a content shadow layer if needed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107119 268f45cc-cd09-0410-ab3c-d52691b4dbfc

REGRESSION (r102983): ClicktoFlash drawing of old style youtube embeds missing until resize
https://bugs.webkit.org/show_bug.cgi?id=77167

Reviewed by Eric Seidel.

Source/WebCore:

Test: plugins/layout-in-beforeload-listener-affects-plugin-loading.html

r102983 made FrameView::updateWidgets() check if the DOM node actually
needs a widget update before calling updateWidget(). Due to historical
reasons, however, updateWidget() can be legitimately called twice: once
at attach time for non-Netscape plug-ins and once at layout time for
Netscape plug-ins.

If the widget represents a Netscape plug-in, but updateWidget() is
called for the CreateOnlyNonNetscapePlugins case after the DOM node was
marked as needing an update, updateWidget() will clear the update flag
and prevent a second call to updateWidget() at layout time for the
CreateAnyWidgetType case.

As much as I loathe adding to the code duplication between
HTMLEmbedElement::updateWidget() and HTMLObjectElement::updateWidget(),
the simplest solution seems to be marking the DOM node as needing
update in the case where we are calling updateWidget() for the
CreateOnlyNonNetscapePlugins case and we know we will be loading a
Netscape plug-in.

* html/HTMLEmbedElement.cpp:
(WebCore::HTMLEmbedElement::updateWidget): Call
setNeedsWidgetUpdate(true) if pluginCreationOption is
CreateOnlyNonNetscapePlugins but we will load a Netscape plug-in.
* html/HTMLObjectElement.cpp:
(WebCore::HTMLObjectElement::updateWidget): Ditto.
* html/HTMLPlugInElement.cpp:
(WebCore::HTMLPlugInElement::guardedDispatchBeforeLoadEvent): Remove an
invalid assertion that prevents the layout test from running in a Debug
configuration.

LayoutTests:

* plugins/layout-in-beforeload-listener-affects-plugin-loading-expected.txt: Added.
* plugins/layout-in-beforeload-listener-affects-plugin-loading.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107118 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Make dependency on Accelerate.framework explicit.
https://bugs.webkit.org/show_bug.cgi?id=78131

It's used by platform/audio/mac/FFTFrameMac.cpp. The explicit
dependency is needed to build a libwebkit.dylib.

Patch by Nico Weber <nicolasweber@gmx.de> on 2012-02-08
Reviewed by Tony Chang.

* WebKit.gyp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107117 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Chromium] Don't add selection from password fields to context menu.
https://bugs.webkit.org/show_bug.cgi?id=77733

Patch by Bernhard Bauer <bauerb@google.com> on 2012-02-08
Reviewed by Adam Barth.

* src/ContextMenuClientImpl.cpp:
(WebKit::ContextMenuClientImpl::getCustomMenuFromDefaultItems):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107115 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed rebaselines for vertical text tests.
See https://bugs.webkit.org/show_bug.cgi?id=78120

* platform/chromium-mac-leopard/editing/selection/vertical-lr-ltr-extend-line-backward-br-expected.png:
[...]
* platform/chromium-mac-snowleopard/fast/writing-mode/vertical-font-fallback-expected.png: Added.
* platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107114 268f45cc-cd09-0410-ab3c-d52691b4dbfc

nrwt: make --skip-pixel-test-if-no-baseline option
https://bugs.webkit.org/show_bug.cgi?id=70484

Patch by Fehér Zsolt <feherzs@inf.u-szeged.hu> on 2012-02-08
Reviewed by Dirk Pranke.

* Scripts/webkitpy/layout_tests/controllers/single_test_runner.py:
(SingleTestRunner.__init__):
(SingleTestRunner._should_fetch_expected_checksum):
* Scripts/webkitpy/layout_tests/controllers/worker.py:
(Worker.handle_test_list):
* Scripts/webkitpy/layout_tests/models/test_input.py:
(TestInput.__init__):
* Scripts/webkitpy/layout_tests/port/webkit.py:
(WebKitDriver.cmd_line):
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
(_set_up_derived_options):
(parse_args):
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
(MainTest.test_skip_pixel_test_if_no_baseline_option):
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::TestController):
(WTR::TestController::initialize):
(WTR::TestController::runTest):
* WebKitTestRunner/TestController.h:
(TestController):
* WebKitTestRunner/TestInvocation.cpp:
(WTR::TestInvocation::TestInvocation):
(WTR::TestInvocation::setIsPixelTest):
* WebKitTestRunner/TestInvocation.h:
(WTR::TestInvocation::setSkipPixelTestOption):
(TestInvocation):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107113 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Floated flexboxes render as regular RenderBlocks
https://bugs.webkit.org/show_bug.cgi?id=77909

Reviewed by Eric Seidel.

Source/WebCore:

Add grid/flexbox cases to adjusting the display of floated/positioned
elements. Also, move this logic into a switch statement. This makes
the code more readable and gives compile warnings when new display types
are added that aren't handled here.

Test: css3/flexbox/floated-flexbox.html

* css/CSSStyleSelector.cpp:
(WebCore::adjustDisplay):
(WebCore):
(WebCore::CSSStyleSelector::adjustRenderStyle):

LayoutTests:

* css3/flexbox/floated-flexbox-expected.txt: Added.
* css3/flexbox/floated-flexbox.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107112 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Let -Wglobal-constructors hack in shared build metastasize to mac
https://bugs.webkit.org/show_bug.cgi?id=78130

Patch by Nico Weber <nicolasweber@gmx.de> on 2012-02-08
Reviewed by Tony Chang.

* WebKit.gyp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107111 268f45cc-cd09-0410-ab3c-d52691b4dbfc

r107042 caused compile breakages on chromium try bots
https://bugs.webkit.org/show_bug.cgi?id=78125

Reviewed by Adam Barth.

Replace the copy script with a copies GYP action since the MSVS generator tracks dependencies on a per-file
basis.

* Platform.gyp/Platform.gyp:
* Platform.gyp/copy_webcore_headers.py: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107110 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed gardening.

* platform/chromium-linux/fast/line-grid/line-grid-contains-value-expected.txt: Removed.
* platform/chromium-mac-leopard/fast/line-grid/line-grid-contains-value-expected.png: Added.
* platform/chromium-mac-snowleopard/fast/line-grid/line-grid-contains-value-expected.png: Added.
* platform/chromium-win/fast/line-grid/line-grid-contains-value-expected.png: Added.
* platform/chromium-win/fast/line-grid/line-grid-contains-value-expected.txt: Added.
More rebaseline of those (I pulled the trigger too early).

* platform/chromium/rubberbanding/custom-scrollbars-ne-expected.png:
* platform/chromium/rubberbanding/custom-scrollbars-nw-expected.png:
* platform/chromium/rubberbanding/custom-scrollbars-se-expected.png:
* platform/chromium/rubberbanding/custom-scrollbars-sw-expected.png:
Update the expected files. I don't know how we ended up with a blue color when the html specifies
an orange.

* platform/chromium/test_expectations.txt:
Mark a test as timing out as it has been raising an exception on the bots fairly regularly.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107109 268f45cc-cd09-0410-ab3c-d52691b4dbfc

viewBox on nested SVG causes wrong content size for relative values
https://bugs.webkit.org/show_bug.cgi?id=69459

Reviewed by Nikolas.

Source/WebCore:

In the past we just checked the change of the viewport size of the root SVG element. If the size changed, all childs
with relative length values needed a relayout. We did not consider that we might have other viewports in the document.
Childs with relative lengths had a strange zooming, if just the viewport size of an inner SVG element changed.

With this patch we check if the size of the nearest viewport changes. Is this the case, childs with relative lengths
need a relayout.

Test: inner-svg-change-viewBox.svg

* rendering/svg/RenderSVGContainer.cpp:
(WebCore::RenderSVGContainer::layout):
* rendering/svg/RenderSVGContainer.h:
(RenderSVGContainer):
(WebCore::RenderSVGContainer::determineIfLayoutSizeChanged): Check if we need layout and have relative length values.
* rendering/svg/RenderSVGRoot.cpp:
(WebCore::RenderSVGRoot::layout): Remove resetting 'viewport size changed' flag for code operability. No influence on the layout.
* rendering/svg/RenderSVGViewportContainer.cpp:
(WebCore::RenderSVGViewportContainer::RenderSVGViewportContainer): Add a flag that indicates that the viewport size changes.
(WebCore::RenderSVGViewportContainer::determineIfLayoutSizeChanged): The flag gets set during the layout phase of the SVG element if the size changes.
(WebCore):
* rendering/svg/RenderSVGViewportContainer.h:
(WebCore::RenderSVGViewportContainer::isLayoutSizeChanged): Added getter to get flag status.
(RenderSVGViewportContainer):
(WebCore::toRenderSVGViewportContainer): Added casting function for constant RenderObjects.
(WebCore):
* rendering/svg/SVGRenderSupport.cpp:
(WebCore::layoutSizeOfNearestViewportChanged): Search the nearest viewport and check if the size changed.
(WebCore):
(WebCore::SVGRenderSupport::layoutChildren): Don't check the roots viewport for size changes, but the nearest viewport.
* svg/SVGSVGElement.cpp:
(WebCore::SVGSVGElement::svgAttributeChanged): Added viewBoxAttr to the list of attributes that cause relayout.

LayoutTests:

Test relayout of content of inner SVG on change of relative length values.

* svg/repaint/inner-svg-change-viewBox-expected.png: Added.
* svg/repaint/inner-svg-change-viewBox-expected.txt: Added.
* svg/repaint/inner-svg-change-viewBox.svg: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107108 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Implement new WEBGL compressed texture extensions
https://bugs.webkit.org/show_bug.cgi?id=77066

Source/WebCore:

This removes the old experimental compressed
texture extension and implements the first new
one.

A test is in the WebGL conformance tests in
extensions/webgl-compressed-texture-s3tc.html
and will be copied here in a future patch.

Patch by Gregg Tavares <gman@google.com> on 2012-02-08
Reviewed by Kenneth Russell.

No new tests. Test coming in future patch.

* CMakeLists.txt:
* DerivedSources.make:
* GNUmakefile.list.am:
* Target.pri:
* WebCore.gypi:
* WebCore.xcodeproj/project.pbxproj:
* bindings/js/JSWebGLRenderingContextCustom.cpp:
(WebCore::toJS):
* bindings/v8/custom/V8WebGLRenderingContextCustom.cpp:
(WebCore::toV8Object):
* html/canvas/WebGLCompressedTextureS3TC.cpp: Added.
(WebCore):
(WebCore::WebGLCompressedTextureS3TC::WebGLCompressedTextureS3TC):
(WebCore::WebGLCompressedTextureS3TC::~WebGLCompressedTextureS3TC):
(WebCore::WebGLCompressedTextureS3TC::getName):
(WebCore::WebGLCompressedTextureS3TC::create):
(WebCore::WebGLCompressedTextureS3TC::supported):
* html/canvas/WebGLCompressedTextureS3TC.h: Copied from Source/WebCore/html/canvas/WebGLExtension.h.
(WebCore):
(WebGLCompressedTextureS3TC):
* html/canvas/WebGLCompressedTextureS3TC.idl: Renamed from Source/WebCore/html/canvas/WebGLCompressedTextures.idl.
* html/canvas/WebGLCompressedTextures.cpp: Removed.
* html/canvas/WebGLCompressedTextures.h: Removed.
* html/canvas/WebGLExtension.h:
* html/canvas/WebGLGetInfo.cpp:
(WebCore::WebGLGetInfo::WebGLGetInfo):
(WebCore):
(WebCore::WebGLGetInfo::getWebGLUnsignedIntArray):
* html/canvas/WebGLGetInfo.h:
* html/canvas/WebGLObject.cpp:
* html/canvas/WebGLRenderingContext.cpp:
(WebCore):
(WebCore::WebGLRenderingContext::addCompressedTextureFormat):
(WebCore::WebGLRenderingContext::compressedTexImage2D):
(WebCore::WebGLRenderingContext::compressedTexSubImage2D):
(WebCore::WebGLRenderingContext::getExtension):
(WebCore::WebGLRenderingContext::getParameter):
(WebCore::WebGLRenderingContext::getSupportedExtensions):
(WebCore::WebGLRenderingContext::validateCompressedTexFormat):
(WebCore::WebGLRenderingContext::validateCompressedTexFuncData):
(WebCore::WebGLRenderingContext::validateCompressedTexDimensions):
(WebCore::WebGLRenderingContext::validateCompressedTexSubDimensions):
* html/canvas/WebGLRenderingContext.h:
(WebCore):
(WebGLRenderingContext):
* html/canvas/WebGLRenderingContext.idl:

LayoutTests:

Patch by Gregg Tavares <gman@chromium.org> on 2012-02-08
Reviewed by Kenneth Russell.

* fast/canvas/webgl/constants.html:
* fast/canvas/webgl/gl-get-calls-expected.txt:
* fast/canvas/webgl/gl-get-calls.html:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107107 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: heap snapshot: implement Distance column in Object's retaining tree.
https://bugs.webkit.org/show_bug.cgi?id=78113

Retaining path list was replaced with Retaining tree some time ago.
But it was not so useful when we want to track the retaining path from an object to a DOM Window node.

Drive by fix: sort doesn't work in retaining tree panel.
Drive by fix: save/load child nodes doesn't work for the retaining tree panel.

Reviewed by Yury Semikhatsky.

* inspector/front-end/DetailedHeapshotGridNodes.js:
(WebInspector.HeapSnapshotObjectNode):
(WebInspector.HeapSnapshotObjectNode.prototype._childHashForEntity): save/load children fix
(WebInspector.HeapSnapshotObjectNode.prototype._childHashForNode): save/load children fix
(WebInspector.HeapSnapshotObjectNode.prototype.comparator):
(WebInspector.HeapSnapshotObjectNode.prototype._enhanceData):
* inspector/front-end/DetailedHeapshotView.js:
(WebInspector.HeapSnapshotContainmentDataGrid):
(WebInspector.HeapSnapshotRetainmentDataGrid):
(WebInspector.HeapSnapshotRetainmentDataGrid.prototype._sortFields):
* inspector/front-end/HeapSnapshot.js:
(WebInspector.HeapSnapshotRetainerEdge.prototype.set retainerIndex):
(WebInspector.HeapSnapshotRetainerEdge.prototype.set edgeIndex):
(WebInspector.HeapSnapshotRetainerEdge.prototype.get _node):
(WebInspector.HeapSnapshotRetainerEdge.prototype.get _edge):
(WebInspector.HeapSnapshotNode.prototype.get distanceToWindow):
(WebInspector.HeapSnapshot.prototype._init):
(WebInspector.HeapSnapshot.prototype._buildRetainers):
(WebInspector.HeapSnapshot.prototype._calculateObjectToWindowDistance):
(WebInspector.HeapSnapshot.prototype._bfs):
(WebInspector.HeapSnapshotEdgesProvider.prototype._serialize):
(WebInspector.HeapSnapshotEdgesProvider.prototype.sort.compareEdgeFieldName):
(WebInspector.HeapSnapshotEdgesProvider.prototype.sort.compareNodeField):
(WebInspector.HeapSnapshotEdgesProvider.prototype.sort.compareEdgeAndNode):
(WebInspector.HeapSnapshotEdgesProvider.prototype.sort.compareNodeAndEdge):
(WebInspector.HeapSnapshotEdgesProvider.prototype.sort.compareNodeAndNode):
(WebInspector.HeapSnapshotEdgesProvider.prototype.sort):
(WebInspector.HeapSnapshotNodesProvider.prototype._serialize):
* inspector/front-end/heapProfiler.css:
(.detailed-heapshot-view .data-grid td.distanceToWindow-column):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107106 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[BlackBerry] Upstream DumpRenderTreeBlackBerry
https://bugs.webkit.org/show_bug.cgi?id=78042

Patch by Rob Buis <rbuis@rim.com> on 2012-02-08
Reviewed by Antonio Gomes.

Source/WebKit:

Add abstract interface for our DumpRenderTree solution.

* blackberry/Api/DumpRenderTreeClient.h: Added.

Tools:

Add implementation for our DumpRenderTree solution.

* DumpRenderTree/blackberry/DumpRenderTree.cpp: Added.
* DumpRenderTree/blackberry/DumpRenderTreeBlackBerry.h: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107105 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] svg/zoom/page/zoom-coords-viewattr-01-b.svg fails
https://bugs.webkit.org/show_bug.cgi?id=78128

* platform/qt/Skipped: Skip it until fix.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107104 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed morning gardening.

* platform/chromium-linux/fast/line-grid/line-grid-contains-value-expected.png: Added.
* platform/chromium-linux/fast/line-grid/line-grid-contains-value-expected.txt: Added.
Rebaselined this file (font difference).

* platform/chromium/test_expectations.txt: Marked one test as SLOW on Window.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107103 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix assertion in svg/dom/SVGStyledElement-pendingResource-crash.html
https://bugs.webkit.org/show_bug.cgi?id=78126

Reviewed by Dan Bernstein.

This broke in r106977 when I tried to change an early return into an ASSERT,
so let's bring back the early return.

* page/FrameView.cpp:
(WebCore::FrameView::notifyPageThatContentAreaWillPaint):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107102 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, rolling out r106920, r106924, r106933, r106939,
and r107090.
http://trac.webkit.org/changeset/106920
http://trac.webkit.org/changeset/106924
http://trac.webkit.org/changeset/106933
http://trac.webkit.org/changeset/106939
http://trac.webkit.org/changeset/107090
https://bugs.webkit.org/show_bug.cgi?id=78124

Something is completely wrong this change (Requested by
Ossy_gardener on #webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-02-08

Source/WebCore:

* platform/FileSystem.h:
(WebCore):
* platform/qt/FileSystemQt.cpp:

Source/WebKit2:

* Shared/WebProcessCreationParameters.cpp:
(WebKit::WebProcessCreationParameters::encode):
(WebKit::WebProcessCreationParameters::decode):
* Shared/WebProcessCreationParameters.h:
(WebProcessCreationParameters):
* UIProcess/WebContext.cpp:
(WebKit::WebContext::WebContext):
* UIProcess/qt/WebContextQt.cpp:
(WebKit::WebContext::platformInitializeWebProcess):
* WebProcess/qt/WebProcessQt.cpp:
(WebKit):
(WebKit::WebProcess::platformSetCacheModel):
(WebKit::WebProcess::platformInitializeWebProcess):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107101 268f45cc-cd09-0410-ab3c-d52691b4dbfc

REGRESSION (r106681): Null check missing in [WebFrame(WebInternal) _typingStyle]
https://bugs.webkit.org/show_bug.cgi?id=78080

Source/WebKit/mac:

Reviewed by Reviewed by Andreas Kling, Darin Adler and Dan Bernstein.

* WebView/WebFrame.mm:
(-[WebFrame _typingStyle]):

Tools:

Reviewed by Andreas Kling, Darin Adler and Dan Bernstein.

Add API test.

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/Tests/mac/TypingStyleCrash.mm: Added.
(TestWebKitAPI):
(TestWebKitAPI::TEST):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107100 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Source/WebCore: https://bugs.webkit.org/show_bug.cgi?id=78122

Add support for the "contains" value for line-grid-snap. This value centers the line box in between the
text-top and text-bottom of the minimum number of grid lines that enclose the line box. This is useful for
centering headers in a line grid.

Reviewed by Adam Roben.

Added a new test in fast/line-grid.

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::lineGridSnapAdjustment):

LayoutTests: https://bugs.webkit.org/show_bug.cgi?id=78122

Add a new test that shows the "contains" value in action.

Reviewed by Adam Roben.

* fast/line-grid/line-grid-contains-value.html: Added.
* platform/mac/fast/line-grid/line-grid-contains-value-expected.png: Added.
* platform/mac/fast/line-grid/line-grid-contains-value-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107099 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[GTK] Fails to build docs with non-standard build directories
https://bugs.webkit.org/show_bug.cgi?id=78118

Patch by Gustavo Noronha Silva <gns@gnome.org> on 2012-02-08
Reviewed by Martin Robinson.

* gtk/common.py:
(get_build_path): also try the current directory as a valid build
dir, which makes non-standard build directories such as build-2.0
and build-3.0 work

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107098 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Skia Mac] Make misspelling underline dots unclipped
https://bugs.webkit.org/show_bug.cgi?id=78117
http://code.google.com/p/chromium/issues/detail?id=113154

Reviewed by Stephen White.

No new tests. Existing layout tests cover this.

As is done on the CoreGraphics Mac platform, adjust the
underline width to remove partial dots, not including the
trailing transparent pixel column.

* platform/graphics/skia/GraphicsContextSkia.cpp:
(WebCore::GraphicsContext::drawLineForTextChecking):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107097 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Moved a Mac-specific test to platform/mac.

* fast/text/synthetic-bold-transformed-expected.html: Removed.
* fast/text/synthetic-bold-transformed.html: Removed.
* platform/mac/fast/text/synthetic-bold-transformed-expected.html: Copied from fast/text/synthetic-bold-transformed-expected.html.
* platform/mac/fast/text/synthetic-bold-transformed.html: Copied from fast/text/synthetic-bold-transformed.html.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107096 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Initial upstreaming of input handling for BlackBerry port
https://bugs.webkit.org/show_bug.cgi?id=77992

Patch by Nima Ghanavatian <nghanavatian@rim.com> on 2012-02-08
Reviewed by Rob Buis.

* blackberry/WebKitSupport/InputHandler.cpp: Added.
* blackberry/WebKitSupport/InputHandler.h: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107095 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Chromium: Fix Mac rubber band overhang drawing on composited pages and re-enable tests.
https://bugs.webkit.org/show_bug.cgi?id=78007

This got broken by: http://trac.webkit.org/changeset/105470
And the tests were accidentally(?) disabled in: http://trac.webkit.org/changeset/95191

There is still three tests that fail from the set, but the cause for these is different,
which I'm still investigating.

Patch by Alexei Svitkine <asvitkine@chromium.org> on 2012-02-08
Reviewed by James Robinson.

Source/WebKit/chromium:

* src/NonCompositedContentHost.cpp:
(WebKit::NonCompositedContentHost::NonCompositedContentHost):

LayoutTests:

* platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107094 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: bind entire subtree upon childNodeInserted so that text node were accounted.
https://bugs.webkit.org/show_bug.cgi?id=78116

Reviewed by Yury Semikhatsky.

Source/WebCore:

* inspector/front-end/DOMAgent.js:
(WebInspector.DOMNode):
(WebInspector.DOMDocument):
(WebInspector.DOMAgent.prototype._setDocument):
(WebInspector.DOMAgent.prototype._setDetachedRoot):
(WebInspector.DOMAgent.prototype._setChildNodes):
(WebInspector.DOMAgent.prototype._childNodeRemoved):
(WebInspector.DOMAgent.prototype._unbind):

LayoutTests:

* inspector/elements/insert-node-expected.txt:
* inspector/elements/insert-node.html:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107093 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[GTK] Fix a crash when WebKitWebView is created without a WebContext
https://bugs.webkit.org/show_bug.cgi?id=78104

Reviewed by Philippe Normand.

* UIProcess/API/gtk/WebKitWebView.cpp:
(webkitWebViewSetProperty): Make sure WebKitWebView:web-context
property is initialized to the default web context when a web
context is not passed to g_object_new().
* UIProcess/API/gtk/tests/TestWebKitWebView.cpp:
(testWebViewDefaultContext): Check that a web view created with
g_object_new has the default context.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107092 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Optional out arguments are not supported in the Web Inspector protocol, which breaks the implementation
https://bugs.webkit.org/show_bug.cgi?id=77967

Patch by Peter Rybin <peter.rybin@gmail.com> on 2012-02-08
Reviewed by Yury Semikhatsky.

Condition for RefPtr-based types added. This is more-or-less a hack
and it's should be redone together with the switch to type-safe API.

* inspector/CodeGeneratorInspector.py:
(Generator.process_command):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107091 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][WK2] Compute and set cache capacities using the current CacheModel
https://bugs.webkit.org/show_bug.cgi?id=73918

Patch by Michael Brüning <michael.bruning@nokia.com> on 2012-02-08
Reviewed by Kenneth Rohde Christiansen.

Source/WebKit2:

* UIProcess/WebContext.cpp:
(WebKit::WebContext::WebContext): Set default cacheModel for Qt platform to
CacheModelPrimaryWebBrowser.

LayoutTests:

* platform/qt-wk2/Skipped: Unskip tests skipped for regression from r106920.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107090 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: inspected object wrapper should be released by InjectedScript when popover closes
https://bugs.webkit.org/show_bug.cgi?id=77972

When object popover is shown the object under cursor is resolved and its
wrapper is put into 'popover' object wrapper group. The group is discarded
when the popover closes.

Reviewed by Pavel Feldman.

* bindings/js/ScriptProfiler.cpp:
(WebCore::ScriptProfiler::objectByHeapObjectId):
* bindings/js/ScriptProfiler.h:
(WebCore):
(ScriptProfiler):
* bindings/v8/ScriptProfiler.cpp:
(WebCore::ScriptProfiler::objectByHeapObjectId):
(WebCore):
* bindings/v8/ScriptProfiler.h:
(WebCore):
(ScriptProfiler):
* inspector/Inspector.json:
* inspector/InspectorProfilerAgent.cpp:
(WebCore::InspectorProfilerAgent::getObjectByHeapObjectId):
* inspector/InspectorProfilerAgent.h:
(InspectorProfilerAgent):
* inspector/front-end/DetailedHeapshotGridNodes.js:
(WebInspector.HeapSnapshotGenericObjectNode.prototype.queryObjectContent):
* inspector/front-end/DetailedHeapshotView.js:
(WebInspector.DetailedHeapshotView.prototype._resolveObjectForPopover):
* inspector/front-end/JavaScriptSourceFrame.js:
(WebInspector.JavaScriptSourceFrame):
(WebInspector.JavaScriptSourceFrame.prototype._resolveObjectForPopover):
(WebInspector.JavaScriptSourceFrame.prototype._onHidePopover):
* inspector/front-end/ObjectPopoverHelper.js:
(WebInspector.ObjectPopoverHelper):
(WebInspector.ObjectPopoverHelper.prototype._showObjectPopover):
(WebInspector.ObjectPopoverHelper.prototype._onHideObjectPopover):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107089 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[GTK] Open links in a new window when clicking with the middle button in MiniBrowser
https://bugs.webkit.org/show_bug.cgi?id=78099

Reviewed by Martin Robinson.

* MiniBrowser/gtk/BrowserWindow.c:
(webViewDecidePolicy): Check whether it's a link clicked with the
middle mouse button and load the request in a new window.
(browserWindowConstructed): Connect to
WebKitWebView::decide-policy signal.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107088 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[GTK] Fix WebKitWebView::decide-policy API documentation
https://bugs.webkit.org/show_bug.cgi?id=78101

Reviewed by Martin Robinson.

- It refers to WebKitPolicyClient instead of WebKitWebView
- Trailing ':' is missing
- References @decision as WebKitNavigationPolicyDecision instead
of WebKitPolicyDecision

* UIProcess/API/gtk/WebKitWebView.cpp:
(webkit_web_view_class_init):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107087 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Gtk] atk_text_get_text_at_offset() fails to provide the correct line for list items whose text wraps
https://bugs.webkit.org/show_bug.cgi?id=73431

Reviewed by Chris Fleizach.

Source/WebCore:

Don't replace item's markers with the objectReplacementCharacter
character, as they will be treated in an special way later on.

* accessibility/gtk/WebKitAccessibleInterfaceText.cpp:
(textForRenderer): Don't append the objectReplacementCharacter
character for list item's markers.

Source/WebKit/gtk:

Updated unit test to check text wrapping accross different lines
inside list items with bullet markers.

* tests/testatk.c:
(testWebkitAtkGetTextAtOffsetWithSpecialCharacters): Updated test.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107086 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remote web inspector reentrancy fixes
https://bugs.webkit.org/show_bug.cgi?id=77022

Patch by Leo Franchi <lfranchi@kde.org> on 2012-02-08
Reviewed by Joseph Pecoraro.

* WebCoreSupport/InspectorServerQt.cpp:
(WebCore::InspectorServerRequestHandlerQt::tcpReadyRead):
(WebCore::InspectorServerRequestHandlerQt::webSocketReadyRead):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107085 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, add chromium expectations for fast/forms/select-overflow-scroll* tests.

* platform/chromium-linux/fast/forms/select-overflow-scroll-expected.png: Added.
* platform/chromium-linux/fast/forms/select-overflow-scroll-expected.txt: Added.
* platform/chromium-linux/fast/forms/select-overflow-scroll-inherited-expected.png: Added.
* platform/chromium-linux/fast/forms/select-overflow-scroll-inherited-expected.txt: Added.
* platform/chromium-mac-snowleopard/fast/forms/select-overflow-scroll-expected.png: Added.
* platform/chromium-mac-snowleopard/fast/forms/select-overflow-scroll-expected.txt: Added.
* platform/chromium-mac-snowleopard/fast/forms/select-overflow-scroll-inherited-expected.png: Added.
* platform/chromium-mac-snowleopard/fast/forms/select-overflow-scroll-inherited-expected.txt: Added.
* platform/chromium-win/fast/forms/select-overflow-scroll-expected.png: Added.
* platform/chromium-win/fast/forms/select-overflow-scroll-expected.txt: Added.
* platform/chromium-win/fast/forms/select-overflow-scroll-inherited-expected.png: Added.
* platform/chromium-win/fast/forms/select-overflow-scroll-inherited-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107084 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: do not clear entire tree map upon last element deletion.
https://bugs.webkit.org/show_bug.cgi?id=78112

Reviewed by Yury Semikhatsky.

* inspector/front-end/treeoutline.js:
(TreeOutline.prototype._forgetChildrenRecursive):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107083 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Migrate createObjectURL & revokeObjectURL to static (Class) methods.
https://bugs.webkit.org/show_bug.cgi?id=74386

Patch by Kaustubh Atrawalkar <kaustubh@motorola.com> on 2012-02-08
Reviewed by Kentaro Hara.

Source/WebCore:

Move createObjectURL & revokeObjectURL from DOMURL implementation to
static methods as per specs - http://www.w3.org/TR/FileAPI/#creating-revoking

Test: fast/dom/DOMURL/check-instanceof-domurl-functions.html
Already Existing:
    fast/files/revoke-blob-url.html
    fast/dom/window-domurl-crash.html
    fast/files/apply-blob-url-to-img.html
    fast/files/create-blob-url-crash.html
    fast/files/workers/inline-worker-via-blob-url.html

* GNUmakefile.list.am:
* Target.pri:
* WebCore.gypi:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* dom/ScriptExecutionContext.cpp:
(WebCore::ScriptExecutionContext::~ScriptExecutionContext):
(WebCore::ScriptExecutionContext::fileThread):
(WebCore):
(WebCore::ScriptExecutionContext::publicURLManager):
* dom/ScriptExecutionContext.h:
(WebCore):
(ScriptExecutionContext):
* html/DOMURL.cpp:
(WebCore):
(WebCore::DOMURL::createObjectURL): Changed to static.
(WebCore::DOMURL::revokeObjectURL): ditto.
* html/DOMURL.h:
(DOMURL):
(WebCore::DOMURL::create):
* html/DOMURL.idl:
* html/PublicURLManager.h: Added.
(WebCore):
(PublicURLManager):
(WebCore::PublicURLManager::create):
(WebCore::PublicURLManager::contextDestroyed):
(WebCore::PublicURLManager::blobURLs):
(WebCore::PublicURLManager::streamURLs):
* page/DOMWindow.cpp: Removed object initialization for DOMURL.
(WebCore):
* page/DOMWindow.h: ditto.
(DOMWindow):
* page/DOMWindow.idl:
* workers/WorkerContext.cpp:
(WebCore):
* workers/WorkerContext.h:
(WorkerContext):
* workers/WorkerContext.idl:

LayoutTests:

Added test to check if createObjectURL & revokeObjectURL are static functions.

* fast/dom/DOMURL/check-instanceof-domurl-functions-expected.txt: Added.
* fast/dom/DOMURL/check-instanceof-domurl-functions.html: Added.
* platform/gtk/fast/dom/prototype-inheritance-2-expected.txt: GTK Rebaseline.
* platform/gtk/fast/js/global-constructors-expected.txt: ditto.
* platform/qt/fast/dom/constructed-objects-prototypes-expected.txt: Qt Rebaseline.
* platform/qt/fast/dom/prototype-inheritance-2-expected.txt: ditto.
* platform/qt/fast/js/global-constructors-expected.txt: ditto.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107082 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2012-02-08  Pavel Podivilov  <podivilov@chromium.org>

        Unreviewed, chromium expectations update: mark repaint-on-image-bounds-change.svg as IMAGE+TEXT IMAGE.

        * platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107081 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, update chromium svg expectations after r107067.

* platform/chromium-linux/svg/filters/feImage-late-indirect-update-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/filters/feImage-animated-transform-on-target-rect-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/filters/feImage-change-target-id-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/filters/feImage-late-indirect-update-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/filters/feImage-multiple-targets-id-change-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/filters/feImage-remove-target-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/filters/feImage-target-attribute-change-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/filters/feImage-target-attribute-change-with-use-indirection-2-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/filters/feImage-target-attribute-change-with-use-indirection-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/filters/feImage-target-inline-style-change-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/filters/feImage-target-property-change-expected.png: Added.
* platform/chromium-mac-snowleopard/svg/filters/feImage-target-style-change-expected.png: Added.
* platform/chromium-win/svg/filters/feImage-animated-transform-on-target-rect-expected.png: Added.
* platform/chromium-win/svg/filters/feImage-late-indirect-update-expected.png: Added.
* platform/chromium-win/svg/filters/feImage-multiple-targets-id-change-expected.png: Added.
* platform/chromium-win/svg/filters/feImage-target-attribute-change-expected.png: Added.
* platform/chromium-win/svg/filters/feImage-target-attribute-change-with-use-indirection-2-expected.png: Added.
* platform/chromium-win/svg/filters/feImage-target-attribute-change-with-use-indirection-expected.png: Added.
* platform/chromium-win/svg/filters/feImage-target-inline-style-change-expected.png: Added.
* platform/chromium-win/svg/filters/feImage-target-property-change-expected.png: Added.
* platform/chromium-win/svg/filters/feImage-target-style-change-expected.png: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107080 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Add changes for Spectrum colorpicker
https://bugs.webkit.org/show_bug.cgi?id=75454

Patch by Brian Grinstead <briangrinstead@gmail.com> on 2012-02-01
Reviewed by Pavel Feldman.

* inspector/front-end/Color.js:
(WebInspector.Color.fromRGB):
* inspector/front-end/utilities.js:
(Element.prototype.totalOffsetLeft):
(Element.prototype.totalOffsetTop):
(Element.prototype.totalOffset):
(Element.prototype.scrollOffset):
():

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107079 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2012-02-08  Pavel Podivilov  <podivilov@chromium.org>

        Unreviewed, repaint-on-image-bounds-change.svg still has wrong image in chromium.
        https://bugs.webkit.org/show_bug.cgi?id=78084

        * platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107078 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, GTK rebaseline and skipping 2 new failing tests.

* platform/gtk/Skipped:
* platform/gtk/fast/dom/Window/window-properties-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107077 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Avoid an avalanche of "class" attribute modifications in WatchExpressionsSidebarPane
https://bugs.webkit.org/show_bug.cgi?id=78102

Reviewed by Vsevolod Vlasov.

* inspector/front-end/WatchExpressionsSidebarPane.js:
(WebInspector.WatchExpressionsSection.prototype._updateHoveredElement):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@107076 268f45cc-cd09-0410-ab3c-d52691b4dbfc