CSS box-shadow default color should be something other than transparent
https://bugs.webkit.org/show_bug.cgi?id=58511

Patch by Dave Tharp <dtharp@codeaurora.org> on 2012-04-10
Reviewed by Simon Fraser.

Source/WebCore:

If box-shadow or text-shadow color is not specified, color defaults to
the value specified in the 'color' property of the element (as specified in W3C
spec).

Test: fast/text/text-shadow-no-default-color.html. Also existing
tests ietestcenter/css3/box-shadow-002 and ietestcenter/css3/box-shadow-003

* css/CSSStyleSelector.cpp:
(WebCore::CSSStyleSelector::collectMatchingRulesForList):

LayoutTests:

Added new test for the case where no color is specified for text
shadow. This test and two others are marked as expected fail pending
rebaseline.

* fast/text/text-shadow-no-default-color.html: Added.
* platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113770 268f45cc-cd09-0410-ab3c-d52691b4dbfc

AudioParam must support connections from audio-rate signals
https://bugs.webkit.org/show_bug.cgi?id=83524

Source/WebCore:

Reviewed by Eric Carlson.

In the Web Audio API, it's possible to connect one AudioNode to another AudioNode.
Similary we should allow an AudioNode to connect to an AudioParam, thus controlling
a parameter with an audio-rate signal.  This is important in many audio processing
applications.

Test: webaudio/audioparam-connect-audioratesignal.html

Simple method name change of AudioParam::hasTimelineValues() to AudioParam::hasSampleAccurateValues().
* Modules/webaudio/AudioGainNode.cpp:
(WebCore::AudioGainNode::process):

* Modules/webaudio/AudioNode.cpp:
(WebCore::AudioNode::connect): Add connect() method from AudioNode -> AudioParam.
(WebCore):
(WebCore::AudioNode::disconnect):
(WebCore::AudioNode::finishDeref):
Use AudioNodeOutput::disconnectAll() instead of AudioNodeOutput::disconnectAllInputs().
* Modules/webaudio/AudioNode.h: Add connect() method from AudioNode -> AudioParam.
(WebCore):
(AudioNode):
* Modules/webaudio/AudioNode.idl: Add connect() method from AudioNode -> AudioParam.

Implement support for an AudioNodeOutput to fanout to multiple AudioParams.
* Modules/webaudio/AudioNodeOutput.cpp:
(WebCore::AudioNodeOutput::AudioNodeOutput):
(WebCore::AudioNodeOutput::updateRenderingState): Update rendering state related to AudioParams.
(WebCore::AudioNodeOutput::pull): pull() must now take into account fanout to AudioParams for in-place processing.
(WebCore::AudioNodeOutput::fanOutCount):
(WebCore):
(WebCore::AudioNodeOutput::paramFanOutCount): New method keeping track of number of connections to AudioParams.
(WebCore::AudioNodeOutput::renderingParamFanOutCount): New method keeping track of number of connections to AudioParams for rendering.
(WebCore::AudioNodeOutput::addParam): Add a connection to an AudioParam.
(WebCore::AudioNodeOutput::removeParam): Remove a connection to an AudioParam.
(WebCore::AudioNodeOutput::disconnectAllParams): Remove all connections to AudioParams.
(WebCore::AudioNodeOutput::disconnectAll): New method to disconnect all AudioNodeInputs and AudioParams.
* Modules/webaudio/AudioNodeOutput.h:
(AudioNodeOutput):

Allow an AudioParam to accept a connection from an AudioNodeOutput, thus being controlled
by an audio-rate signal.
* Modules/webaudio/AudioParam.cpp:
(WebCore::AudioParam::calculateSampleAccurateValues): Calculates sample-accurate values from timeline or an AudioNode.
(WebCore):
(WebCore::AudioParam::calculateAudioRateSignalValues): Calculates sample-accurate values from an AudioNode.
(WebCore::AudioParam::calculateTimelineValues): Calculates sample-accurate values scheduled on the timeline.
(WebCore::AudioParam::connect): Connect from an AudioNodeOutput for control from an audio-rate signal.
(WebCore::AudioParam::disconnect): Disconnect from an AudioNodeOutput.
* Modules/webaudio/AudioParam.h:
(WebCore):
(WebCore::AudioParam::AudioParam):
(WebCore::AudioParam::hasSampleAccurateValues): Change name from hasTimelineValues() and return true
either if we have timeline values or if we've been connected from an AudioNode.
(AudioParam):

Simple method name change of AudioParam::hasTimelineValues() to AudioParam::hasSampleAccurateValues().
* Modules/webaudio/Oscillator.cpp:
(WebCore::Oscillator::calculateSampleAccuratePhaseIncrements):
(WebCore::Oscillator::process):

LayoutTests:

Reviewed by Eric Carlson.

* webaudio/audioparam-connect-audioratesignal-expected.txt: Added.
* webaudio/audioparam-connect-audioratesignal.html: Added.
* webaudio/resources/audio-testing.js:
(createLinearRampBuffer):
(createConstantBuffer):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113769 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[CMake][WIN] Add /MP flag when using a compiler with support for it
https://bugs.webkit.org/show_bug.cgi?id=83577

Reviewed by Daniel Bates.

* Source/cmake/OptionsWindows.cmake:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113767 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[CMake] Add missing DERIVED_SOURCES_WEBKIT_DIR variable
https://bugs.webkit.org/show_bug.cgi?id=83573

Reviewed by Daniel Bates.

We need this variable when ports want to generate files in the WebKit target.

* CMakeLists.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113766 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove unused data member from Lexer class
https://bugs.webkit.org/show_bug.cgi?id=83429

Reviewed by Kentaro Hara.

I noticed that m_delimited was "write-only", so I deleted it.

* parser/Lexer.cpp:
(JSC::Lexer::setCode): Removed code to set m_delimited.
(JSC::Lexer::parseIdentifier): Ditto.
(JSC::Lexer::parseIdentifierSlowCase): Ditto.
(JSC::Lexer::lex): Ditto.
* parser/Lexer.h: Deleted m_delimited.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113765 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[CMake] Enable USE_FOLDERS property
https://bugs.webkit.org/show_bug.cgi?id=83571

Reviewed by Daniel Bates.

.:

Setting the FOLDER property on targets gives more structure
to the generated Visual Studio solutions.
This does not affect other CMake generators.

* Source/cmake/OptionsCommon.cmake:

Source/JavaScriptCore:

Setting the FOLDER property on targets gives more structure
to the generated Visual Studio solutions.
This does not affect other CMake generators.

* CMakeLists.txt:
* shell/CMakeLists.txt:

Source/WebCore:

Setting the FOLDER property on targets gives more structure
to the generated Visual Studio solutions.
This does not affect other CMake generators.

* CMakeLists.txt:

Source/WebKit:

Setting the FOLDER property on targets gives more structure
to the generated Visual Studio solutions.
This does not affect other CMake generators.

* CMakeLists.txt:

Source/WebKit2:

Setting the FOLDER property on targets gives more structure
to the generated Visual Studio solutions.
This does not affect other CMake generators.

* CMakeLists.txt:

Source/WTF:

Setting the FOLDER property on targets gives more structure
to the generated Visual Studio solutions.
This does not affect other CMake generators.

* wtf/CMakeLists.txt:

Tools:

Setting the FOLDER property on targets gives more structure
to the generated Visual Studio solutions.
This does not affect other CMake generators.

* DumpRenderTree/efl/CMakeLists.txt:
* EWebLauncher/CMakeLists.txt:
* WinCELauncher/CMakeLists.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113764 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[CMake] CMake SVG Code Generation fails to generate code for extra defines
https://bugs.webkit.org/show_bug.cgi?id=83562

Patch by Dominik Röttsches <dominik.rottsches@linux.intel.com> on 2012-04-10
Reviewed by Rob Buis.

Fixing CMakeLists.txt to generate correct extra defines for SVG code generator.

* CMakeLists.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113763 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Crash due to captions list not updated after section recalc.
https://bugs.webkit.org/show_bug.cgi?id=83552

Reviewed by Julien Chaffraix.

Source/WebCore:

Test: fast/table/table-caption-not-removed-crash.html

* rendering/RenderTable.cpp:
(WebCore::RenderTable::addChild): no need to set the need for
section recalc. It was needed in old code when we had more than
one caption and we need to call section recalc to destroy the other
captions.
(WebCore::RenderTable::recalcSections): need to rebuild captions list.
This is how the old code worked before r100177. Basically, children can
moved without calling RenderTable::removeChild, so we should depend on
recalcSections to update our captions list. Also, fix a style nit of aligning
case labels with the switch statement.

LayoutTests:

* fast/table/table-caption-not-removed-crash-expected.txt: Added.
* fast/table/table-caption-not-removed-crash.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113759 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[BlackBerry] Crash when tearing down web page rendered using WebPageCompositor
https://bugs.webkit.org/show_bug.cgi?id=83567

Patch by Arvid Nilsson <anilsson@rim.com> on 2012-04-10
Reviewed by Antonio Gomes.

Fixed by checking if there's a layer renderer before calling into it.

* Api/WebPageCompositor.cpp:
(BlackBerry::WebKit::WebPageCompositorPrivate::releaseLayerResources):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113757 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, rolling out r113611.
http://trac.webkit.org/changeset/113611
https://bugs.webkit.org/show_bug.cgi?id=71541

Broke fast/forms/basic-textareas.html

Source/WebCore:

* rendering/RenderBox.h:
(RenderBox):
* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::hasHorizontalOverflow):
(WebCore::RenderLayer::hasVerticalOverflow):

LayoutTests:

* fast/overflow/overflow-auto-destroy-scroll-after-resizing-expected.html: Removed.
* fast/overflow/overflow-auto-destroy-scroll-after-resizing.html: Removed.
* platform/efl/Skipped:
* platform/mac/Skipped:
* platform/qt/Skipped:
* platform/win/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113756 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, rolling out r113734.
http://trac.webkit.org/changeset/113734
https://bugs.webkit.org/show_bug.cgi?id=83606

causing a bunch of unrelated test failures (Requested by
simonjam on #webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-04-10

Source/WebCore:

* Modules/battery/BatteryController.cpp:
* Modules/battery/BatteryController.h:
(BatteryController):
* Modules/battery/BatteryManager.cpp:
(WebCore::BatteryManager::charging):
(WebCore::BatteryManager::chargingTime):
(WebCore::BatteryManager::dischargingTime):
(WebCore::BatteryManager::level):
* WebCore.gyp/WebCore.gyp:
* WebCore.gypi:

Source/WebKit/chromium:

* WebKit.gyp:
* features.gypi:
* public/WebBatteryStatus.h: Removed.
* public/WebBatteryStatusClient.h: Removed.
* public/WebView.h:
(WebKit):
(WebView):
* public/WebViewClient.h:
(WebKit):
* src/BatteryClientImpl.cpp: Removed.
* src/BatteryClientImpl.h: Removed.
* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::WebViewImpl):
* src/WebViewImpl.h:
(WebKit):
(WebViewImpl):

LayoutTests:

* platform/chromium/fast/dom/navigator-detached-no-crash-expected.txt:
* platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113755 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add setJSWrapperForActiveDOMNode and use it for Nodes that are also ActiveDOMObjects
https://bugs.webkit.org/show_bug.cgi?id=83528

Reviewed by Kentaro Hara.

Instead of using a run-time call to isActiveNode to determine which
map to put a Node wrapper in, generate the proper call in the CodeGenerator.

This was originally part of r112318, which got rolled out due to OOM concerns.
I'm splitting it into smaller pieces so that each can be landed and
watched for issues seperately.

No new tests, no change in behavior.

* bindings/scripts/CodeGeneratorV8.pm:
(GenerateConstructorCallback): Use new GetDomMapName function to
figure out which setJSWrapper to call.
(GenerateNamedConstructorCallback): ditto.
(GetDomMapFunction): Delegate to GetDomMapName for logic.
(GetDomMapName): New helper factored out of GetDomMapFunction.
* bindings/v8/V8DOMWrapper.cpp:
(WebCore::V8DOMWrapper::setJSWrapperForDOMNode): Assert !isActiveNode instead of branching on it.
(WebCore::V8DOMWrapper::setJSWrapperForActiveDOMNode): New method split ouf of the above.
Assert isActiveNode instead of branching on it.
* bindings/v8/V8DOMWrapper.h:
(V8DOMWrapper):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113754 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Pass PropertyHandler by reference in CSSStyleApplyProperty.h.
https://bugs.webkit.org/show_bug.cgi?id=83551

Reviewed by Kentaro Hara.

No new tests / cleanup only.

* css/CSSStyleApplyProperty.h:
(WebCore::CSSStyleApplyProperty::setPropertyHandler):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113753 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Pass FontDescription as const reference instead of by value in CSSStyleSelector.h.
https://bugs.webkit.org/show_bug.cgi?id=83548

Reviewed by Kentaro Hara.

FontDescription is relatively large, so pass-by-reference is preferred.

No new tests / code cleanup only.

* css/CSSStyleSelector.h:
(WebCore::CSSStyleSelector::fontDescription):
(WebCore::CSSStyleSelector::parentFontDescription):
(WebCore::CSSStyleSelector::setFontDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113752 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add TestNode.idl to run-binding-tests
https://bugs.webkit.org/show_bug.cgi?id=83599

Reviewed by Adam Barth.

TestNode.idl is a minimized version of Node.idl with enough attributes
to exercise CodeGenerator.pm's handling of Node subtypes.

* bindings/scripts/test/CPP/WebDOMTestNode.cpp: Added.
* bindings/scripts/test/CPP/WebDOMTestNode.h: Added.
* bindings/scripts/test/GObject/WebKitDOMTestNode.cpp: Added.
* bindings/scripts/test/GObject/WebKitDOMTestNode.h: Added.
* bindings/scripts/test/GObject/WebKitDOMTestNodePrivate.h: Added.
* bindings/scripts/test/JS/JSTestNode.cpp: Added.
* bindings/scripts/test/JS/JSTestNode.h: Added.
* bindings/scripts/test/ObjC/DOMTestNode.h: Added.
* bindings/scripts/test/ObjC/DOMTestNode.mm: Added.
* bindings/scripts/test/ObjC/DOMTestNodeInternal.h: Added.
* bindings/scripts/test/TestNode.idl: Added.
* bindings/scripts/test/V8/V8TestNode.cpp: Added.
* bindings/scripts/test/V8/V8TestNode.h: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113751 268f45cc-cd09-0410-ab3c-d52691b4dbfc

msqrt's implied mrow should do operator stretching
https://bugs.webkit.org/show_bug.cgi?id=82353

Patch by David Barton <dbarton@mathscribe.com> on 2012-04-10
Reviewed by Julien Chaffraix.

Source/WebCore:

An <msqrt> with more than one child is supposed to surround them with an implied <mrow>.
This patch does this, and thus includes vertical operator stretching. We also fix some
problems with the old RenderMathMLSquareRoot.cpp: possibly shared RenderStyles were
being modified, and layout() was using heights of children without checking whether they
needed layout themselves first.

To leave room for the radical sign, we use intrinsic padding. This is simpler and more
efficient than using anonymous RenderStyles.

Added tests to mo-stretch.html and style.xhtml. Regression tests are in roots.xhtml.

* css/mathml.css:
(msqrt):
* rendering/RenderBoxModelObject.h:
(RenderBoxModelObject):
* rendering/mathml/RenderMathMLBlock.cpp:
(WebCore::RenderMathMLBlock::RenderMathMLBlock):
(WebCore::RenderMathMLBlock::paddingTop):
(WebCore::RenderMathMLBlock::paddingBottom):
(WebCore::RenderMathMLBlock::paddingLeft):
(WebCore::RenderMathMLBlock::paddingRight):
(WebCore::RenderMathMLBlock::paddingBefore):
(WebCore::RenderMathMLBlock::paddingAfter):
(WebCore::RenderMathMLBlock::paddingStart):
(WebCore::RenderMathMLBlock::paddingEnd):
* rendering/mathml/RenderMathMLBlock.h:
(RenderMathMLBlock):
* rendering/mathml/RenderMathMLRow.cpp:
(WebCore::RenderMathMLRow::RenderMathMLRow):
(WebCore::RenderMathMLRow::createAnonymousWithParentRenderer):
* rendering/mathml/RenderMathMLRow.h:
(RenderMathMLRow):
* rendering/mathml/RenderMathMLSquareRoot.cpp:
(WebCore):
(WebCore::RenderMathMLSquareRoot::addChild):
(WebCore::RenderMathMLSquareRoot::computePreferredLogicalWidths):
(WebCore::RenderMathMLSquareRoot::computeLogicalHeight):
(WebCore::RenderMathMLSquareRoot::layout):
(WebCore::RenderMathMLSquareRoot::paint):
* rendering/mathml/RenderMathMLSquareRoot.h:
(RenderMathMLSquareRoot):

LayoutTests:

* mathml/presentation/mo-stretch.html:
* mathml/presentation/style.xhtml:
* platform/mac/mathml/presentation/attributes-expected.txt:
* platform/mac/mathml/presentation/mo-stretch-expected.png:
* platform/mac/mathml/presentation/mo-stretch-expected.txt:
* platform/mac/mathml/presentation/roots-expected.txt:
* platform/mac/mathml/presentation/style-expected.png:
* platform/mac/mathml/presentation/style-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113749 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Don't allow fallthrough for CSSPropertyBaselineShift in CSSComputedStyleDeclaration::getSVGPropertyCSSValue().
https://bugs.webkit.org/show_bug.cgi?id=83536

Reviewed by Daniel Bates.

Add a return statement to the case CSSPropertyBaselineShift so that we don't fall through to the next case statement.

No new tests / code cleanup from coverity static analysis.

* css/SVGCSSComputedStyleDeclaration.cpp:
(WebCore::CSSComputedStyleDeclaration::getSVGPropertyCSSValue):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113748 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2012-04-10  Alejandro G. Castro  <alex@igalia.com>

        [GTK] Build fix TextureMapper compilation and solve warning.

        Reviewed by Martin Robinson.

        * platform/graphics/texmap/TextureMapperGL.cpp:
        (WebCore::BitmapTextureGL::updateContents):
        * platform/graphics/texmap/TextureMapperShaderManager.h:
        (WebCore::TextureMapperShaderManager::getShaderProgram):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113746 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add fast/events/drop-handler-should-not-stop-navigate.html.

Like fast/events/drop-with-file-paths.html, this expects eventSender.beginDragWithFiles to be implemented.

* platform/wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113744 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add fast/events/drop-with-file-paths.html to the WebKit2 Skipped list.

* platform/wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113743 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[GTK] Separate image encoding from dataURL construction
https://bugs.webkit.org/show_bug.cgi?id=83152

Reviewed by Martin Robinson.

Remove the implicit assumption that a dataURL is the only desired output format
of the image encoding phase.

No new tests, refactoring only, covered by existing canvas tests.

* platform/graphics/gtk/ImageBufferGtk.cpp:
(WebCore::encodeImage): Output the encoded image to the provided GOwnPtr<gchar>&
buffer. Update GTK document reference to a valid URL. Round the quality argument
to an int like the toDataURL() implementations of other ports. Ditch the success
variable; instead test the GError* error return to indicate success.
(WebCore):
(WebCore::ImageBuffer::toDataURL): Format the dataURL encoding of the mimeType
encoded image data buffer herein.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113742 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add test expectations for Mac debug builds.

Unreviewed.

* platform/chromium-mac-snowleopard/fast/dom/shadow: Added.
* platform/chromium-mac-snowleopard/fast/dom/shadow/selection-in-shadow-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113741 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Defer texture id allocation for copies until the actual copy executes
https://bugs.webkit.org/show_bug.cgi?id=83514

Reviewed by Adrienne Walker.

Source/WebCore:

This allows for queuing a texture copy operation into a ManagedTexture whose texture ID has not yet been
allocated. The ManagedTexture* serves as a promise that an ID will be available when the copy is executed.

* platform/graphics/chromium/Canvas2DLayerChromium.cpp:
(WebCore::Canvas2DLayerChromium::updateCompositorResources):
* platform/graphics/chromium/cc/CCSingleThreadProxy.cpp:
(WebCore::CCSingleThreadProxy::doCommit):
* platform/graphics/chromium/cc/CCTextureUpdater.cpp:
(WebCore::CCTextureUpdater::CCTextureUpdater):
(WebCore::CCTextureUpdater::appendManagedCopy):
(WebCore):
(WebCore::CCTextureUpdater::hasMoreUpdates):
(WebCore::CCTextureUpdater::update):
(WebCore::CCTextureUpdater::clear):
* platform/graphics/chromium/cc/CCTextureUpdater.h:
(CCTextureUpdater):
(ManagedCopyEntry):
* platform/graphics/chromium/cc/CCThreadProxy.cpp:
(WebCore::CCThreadProxy::beginFrameCompleteOnImplThread):
(WebCore::CCThreadProxy::scheduledActionUpdateMoreResources):

Source/WebKit/chromium:

* tests/CCTiledLayerTestCommon.cpp:
* tests/CCTiledLayerTestCommon.h:
* tests/Canvas2DLayerChromiumTest.cpp:
* tests/TiledLayerChromiumTest.cpp:
(WTF::TEST):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113740 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add ref test expected result.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113739 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Source/WebCore: https://bugs.webkit.org/show_bug.cgi?id=83595
<rdar://problem/10443278> Overlapping text in table cell across column break

If a cell's height changes such that it is taller than the overall row height because
of pagination, then grow the entire row height to enclose the cell instead of shrinking
the cell.

Reviewed by Dan Bernstein.

Added fast/multicol/cell-shrinkback.html

* rendering/RenderTableSection.cpp:
(WebCore::RenderTableSection::layoutRows):

LayoutTests: https://bugs.webkit.org/show_bug.cgi?id=83595

Add new layout test for cells paginating across columns.

Reviewed by Dan Bernstein.

* fast/multicol/cell-shrinkback.html: Added.
* fast/multicol/cell-shrinkback-expected.html: Added.
* platform/mac/fast/multicol/table-vertical-align-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113738 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove unnecessary null check in void SimplifyMarkupCommand::doApply().
https://bugs.webkit.org/show_bug.cgi?id=83535

Reviewed by Kentaro Hara.

No new tests / code cleanup only.

currentNode cannot be null within the loop body.
It is dereferenced before and after the removed line without checking.
Additionally I have added an assertion to express this loop invariant.

* editing/SimplifyMarkupCommand.cpp:
(WebCore::SimplifyMarkupCommand::doApply):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113737 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add Encrypted Media Extensions methods to HTMLMediaElement
https://bugs.webkit.org/show_bug.cgi?id=82971

Patch by David Dorwin <ddorwin@chromium.org> on 2012-04-10
Reviewed by Adam Barth.

The extensions are behind the ENABLE(ENCRYPTED_MEDIA) feature define.
Implementation is based on v0.1 of the draft proposal at
http://dvcs.w3.org/hg/html-media/raw-file/tip/encrypted-media/encrypted-media.html.

Source/WebCore:

Tests: media/encrypted-media/encrypted-media-not-loaded.html
       media/encrypted-media/encrypted-media-syntax.html

* bindings/generic/RuntimeEnabledFeatures.cpp:
(WebCore):
* bindings/generic/RuntimeEnabledFeatures.h:
(RuntimeEnabledFeatures):
(WebCore::RuntimeEnabledFeatures::webkitEncryptedMediaEnabled):
(WebCore::RuntimeEnabledFeatures::setWebkitEncryptedMediaEnabled):
* html/HTMLMediaElement.cpp:
(WebCore):
(WebCore::exceptionCodeForMediaKeyException):
(WebCore::HTMLMediaElement::webkitGenerateKeyRequest):
(WebCore::HTMLMediaElement::webkitAddKey):
(WebCore::HTMLMediaElement::webkitCancelKeyRequest):
(WebCore::HTMLMediaElement::setVolume):
(WebCore::HTMLMediaElement::togglePlayState):
(WebCore::HTMLMediaElement::beginScrubbing):
(WebCore::HTMLMediaElement::addTextTrack):
(WebCore::HTMLMediaElement::textTracks):
(WebCore::HTMLMediaElement::showingTrackWithSameKind):
(WebCore::HTMLMediaElement::didAddTrack):
* html/HTMLMediaElement.h:
* html/HTMLMediaElement.idl:
* platform/graphics/MediaPlayer.cpp:
(NullMediaPlayerPrivate):
(WebCore):
(WebCore::MediaPlayer::generateKeyRequest):
(WebCore::MediaPlayer::addKey):
(WebCore::MediaPlayer::cancelKeyRequest):
* platform/graphics/MediaPlayer.h:
* platform/graphics/MediaPlayerPrivate.h:
(MediaPlayerPrivateInterface):
(WebCore::MediaPlayerPrivateInterface::addKey):
(WebCore::MediaPlayerPrivateInterface::generateKeyRequest):
(WebCore::MediaPlayerPrivateInterface::cancelKeyRequest):

Source/WebKit/chromium:

* features.gypi:
* public/WebMediaPlayer.h:
(WebKit):
(WebMediaPlayer):
(WebKit::WebMediaPlayer::generateKeyRequest):
(WebKit::WebMediaPlayer::addKey):
(WebKit::WebMediaPlayer::cancelKeyRequest):
* public/WebRuntimeFeatures.h:
(WebRuntimeFeatures):
* src/AssertMatchingEnums.cpp:
* src/WebMediaPlayerClientImpl.cpp:
(WebKit):
(WebKit::WebMediaPlayerClientImpl::generateKeyRequest):
(WebKit::WebMediaPlayerClientImpl::addKey):
(WebKit::WebMediaPlayerClientImpl::cancelKeyRequest):
* src/WebMediaPlayerClientImpl.h:
(WebMediaPlayerClientImpl):
* src/WebRuntimeFeatures.cpp:
(WebKit::WebRuntimeFeatures::isMediaStreamEnabled):
(WebKit::WebRuntimeFeatures::enableEncryptedMedia):
(WebKit):
(WebKit::WebRuntimeFeatures::isEncryptedMediaEnabled):

Tools:

* DumpRenderTree/chromium/TestShell.cpp:
(TestShell::TestShell):

LayoutTests:

* media/encrypted-media/encrypted-media-not-loaded-expected.txt: Added.
* media/encrypted-media/encrypted-media-not-loaded.html: Added.
* media/encrypted-media/encrypted-media-syntax-expected.txt: Added.
* media/encrypted-media/encrypted-media-syntax.html: Added.
* platform/chromium/test_expectations.txt:
* platform/efl/Skipped:
* platform/gtk/Skipped:
* platform/mac/Skipped:
* platform/qt/Skipped:
* platform/win/Skipped:
* platform/wincairo/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113736 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Add Battery Status API support.

Source/WebCore:

Add Battery Status API support to chromium.
https://bugs.webkit.org/show_bug.cgi?id=83284

Patch by Sadrul Habib Chowdhury <sadrul@chromium.org> on 2012-04-10
Reviewed by Adam Barth.

This change is covered by tests in batterystatus/.

* Modules/battery/BatteryController.cpp:
(WebCore::BatteryController::updateBatteryStatus):
(WebCore):
* Modules/battery/BatteryController.h:
(BatteryController):
* Modules/battery/BatteryManager.cpp:
(WebCore::BatteryManager::charging):
(WebCore::BatteryManager::chargingTime):
(WebCore::BatteryManager::dischargingTime):
(WebCore::BatteryManager::level):
* WebCore.gyp/WebCore.gyp:
* WebCore.gypi:

Source/WebKit/chromium:

The battery-client (BatteryClientChromium) will notify the client (WebBatteryStatusClient) when to start or stop
sending battery status updates. The client, in response will send the notifications to WebViewImpl, which sends
this update information to the WebCore::BatteryClient, and it triggers the appropriate javascript-callbacks.
The spec is at http://www.w3.org/TR/2011/WD-battery-status-20111129/.
https://bugs.webkit.org/show_bug.cgi?id=83284

Patch by Sadrul Habib Chowdhury <sadrul@chromium.org> on 2012-04-10
Reviewed by Adam Barth.

* WebKit.gyp:
* features.gypi:
* public/WebBatteryStatus.h: Added.
(WebKit):
(WebBatteryStatus):
(WebKit::WebBatteryStatus::WebBatteryStatus):
* public/WebBatteryStatusClient.h: Added.
(WebKit):
(WebBatteryStatusClient):
(WebKit::WebBatteryStatusClient::~WebBatteryStatusClient):
* public/WebView.h:
(WebKit):
(WebView):
(WebKit::WebView::updateBatteryStatus):
* public/WebViewClient.h:
(WebKit):
(WebViewClient):
(WebKit::WebViewClient::batteryStatusClient):
* src/BatteryClientImpl.cpp: Added.
(WebKit):
(WebKit::BatteryClientImpl::BatteryClientImpl):
(WebKit::BatteryClientImpl::updateBatteryStatus):
(WebKit::BatteryClientImpl::setController):
(WebKit::BatteryClientImpl::startUpdating):
(WebKit::BatteryClientImpl::stopUpdating):
(WebKit::BatteryClientImpl::batteryControllerDestroyed):
* src/BatteryClientImpl.h: Added.
(WebKit):
(BatteryClientImpl):
(WebKit::BatteryClientImpl::~BatteryClientImpl):
* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::WebViewImpl):
(WebKit):
(WebKit::WebViewImpl::updateBatteryStatus):
* src/WebViewImpl.h:
(WebKit):
(WebViewImpl):

LayoutTests:

Add Battery Status API support to chromium.
https://bugs.webkit.org/show_bug.cgi?id=83284

Patch by Sadrul Habib Chowdhury <sadrul@chromium.org> on 2012-04-10
Reviewed by Adam Barth.

* platform/chromium/fast/dom/navigator-detached-no-crash-expected.txt:
* platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113734 268f45cc-cd09-0410-ab3c-d52691b4dbfc

getShaderFormatPrecision should return sensible values
https://bugs.webkit.org/show_bug.cgi?id=83520

Reviewed by Kenneth Russell.

Source/WebCore:

* platform/graphics/opengl/GraphicsContext3DOpenGL.cpp: Fix getShaderPrecisionFormat values.
(WebCore::GraphicsContext3D::getShaderPrecisionFormat):
* platform/graphics/qt/GraphicsContext3DQt.cpp: Ditto.
(WebCore::GraphicsContext3D::getShaderPrecisionFormat):

LayoutTests:

* fast/canvas/webgl/shader-precision-format-expected.txt:
* fast/canvas/webgl/shader-precision-format.html: Add tests to make sure getShaderPrecisionFormat returns sensible values.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113733 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Add Galaxy Nexus to the list of emulated user agents
https://bugs.webkit.org/show_bug.cgi?id=83589

The new record contains the device display metrics and the User Agent string.

Reviewed by Pavel Feldman.

* inspector/front-end/SettingsScreen.js:
(WebInspector.SettingsScreen.prototype._createUserAgentSelectRowElement.get const):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113731 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add missing methods and operators to FractionalLayoutUnit
https://bugs.webkit.org/show_bug.cgi?id=82403

Reviewed by Eric Seidel.

Add missing methods and operators and from branch version of
FractionalLayoutUnit. Also fix overflow assertion and add const keyword
to a couple of methods.

No new tests.

* platform/FractionalLayoutUnit.h:
(WebCore::FractionalLayoutUnit::abs):
(FractionalLayoutUnit):
(WebCore::FractionalLayoutUnit::ceil):
(WebCore::FractionalLayoutUnit::round):
(WebCore::FractionalLayoutUnit::floor):
(WebCore::FractionalLayoutUnit::isInBounds):
(WebCore::operator<):
(WebCore):
(WebCore::operator-):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113730 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2012-04-10  Zhenyao Mo  <zmo@google.com>

Unreviewed, roll chromium to 131469

* DEPS:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113729 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebAudio: propagate a silence hint through the AudioNode graph.
https://bugs.webkit.org/show_bug.cgi?id=74553

Reviewed by Chris Rogers.

No new tests; optimization of existing code path, so covered by existing tests.

Introduce the concept of a "silent" channel:
* platform/audio/AudioChannel.h:
(WebCore::AudioChannel::AudioChannel):
(WebCore::AudioChannel::set): Clear silent bit.
(WebCore::AudioChannel::zero): Set silent bit.
(WebCore::AudioChannel::clearSilentFlag): Clear silent bit.
(WebCore::AudioChannel::isSilent): Accessor.

Optimize a few channel operations when the source or destination channels are silent.
* platform/audio/AudioChannel.cpp:
(WebCore::AudioChannel::scale): No-op on a silent channel.
(WebCore::AudioChannel::copyFrom): zero() when source is silent.
(WebCore::AudioChannel::copyFromRange): possibly zero() when source is silent.
(WebCore::AudioChannel::sumFrom): No-op when source is silent.
(WebCore::AudioChannel::maxAbsValue): 0 on a silent channel.

Optimize a few bus operations when the source or destination channels are silent.
* platform/audio/AudioBus.cpp:
(WebCore::AudioBus::processWithGainFromMonoStereo): No-op if source is silent and either
    the destination bus is silent, or the output is not summed to the destination.
(WebCore::AudioBus::copyWithSampleAccurateGainValuesFrom): zero() if the source is silent
    and the lengths of the gain values, source, and destination match.
(WebCore::AudioBus::createBySampleRateConverting): Return an empty bus if the source is silent.
(WebCore::AudioBus::createByMixingToMono): Ditto; clear the destination's silent bit otherwise.
(WebCore::AudioBus::isSilent): Return whether all channels are silent.
(WebCore::AudioBus::clearSilentFlag): Clear silent bit of constituent channels.
* platform/audio/AudioBus.h:

Make sure that classes which generate audio clear their busses' silent bit.
* webaudio/AudioBufferSourceNode.cpp:
(WebCore::AudioBufferSourceNode::process): Ditto.
(WebCore::AudioBufferSourceNode::renderFromBuffer): Ditto.
(WebCore::AudioBufferSourceNode::propagatesSilence): Propagate silence only when the source node
    has nothing left to render.
* webaudio/AudioBufferSourceNode.h:
(WebCore::AudioBufferSourceNode::playbackState): Made const correct.
(WebCore::AudioBufferSourceNode::isPlaying): Added simple accessor.
(WebCore::AudioBufferSourceNode::hasFinished): Ditto.
* webaudio/AudioDestinationNode.h:
(WebCore::AudioDestinationNode::currentSampleFrame): Made const correct.
(WebCore::AudioDestinationNode::currentTime): Ditto.

Audio nodes should not process audio data when the input is silent and the nodes will propagate silences.
* webaudio/AudioNode.cpp:
(WebCore::AudioNode::processIfNecessary):
(WebCore::AudioNode::inputsAreSilent): Convenience function which walk over the node's inputs.
(WebCore::AudioNode::silenceOutputs): Ditto.
(WebCore::AudioNode::unsilenceOutputs): Ditto.
* webaudio/AudioNode.h:
(WebCore::AudioNode::propagatesSilence):

These Nodes can generate audio when given silent input, so return false from propagatesSilence.
* Modules/webaudio/Oscillator.h:
(WebCore::Oscillator::propagatesSilence): Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113728 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Chromium] Unreviewed, add baselines for fast/dom/shadow/selection-in-shadow.html

* platform/chromium/fast/dom/shadow/selection-in-shadow-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113727 268f45cc-cd09-0410-ab3c-d52691b4dbfc

It should be possible to see why a code block was not compiled by the DFG
https://bugs.webkit.org/show_bug.cgi?id=83553

Reviewed by Geoff Garen.

If DFG_ENABLE(DEBUG_VERBOSE) and a code block is rejected, then print the
opcode that caused the rejection.

* dfg/DFGCapabilities.cpp:
(JSC::DFG::debugFail):
(DFG):
(JSC::DFG::canHandleOpcodes):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113726 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed - adding myself as a committer.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113725 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL][DRT] Catch the "title,changed" signal
https://bugs.webkit.org/show_bug.cgi?id=82174

Patch by Christophe Dumez <christophe.dumez@intel.com> on 2012-04-10
Reviewed by Gustavo Noronha Silva.

Tools:

EFL's DumpRenderTree now catches the "title,changed" signal, and
prints out the needed information.

* DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
(DumpRenderTreeChrome::onTitleChanged):

LayoutTests:

Unskip tests which requires EFL's DumpRenderTree to catch the
"title,changed" signal and print out information, now that it
is implemented.

* platform/efl/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113724 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Pressing enter on blank line after bullet deletes entire bulleted line.
https://bugs.webkit.org/show_bug.cgi?id=82690

Reviewed by Enrica Casucci.

In CompositeEditCommand::breakOutOfEmptyListItem(), it checks the empty list item's renderer's siblings
to decide which empty part of the list should be removed. However, if the empty list item's renderer is
wrapped by an anonymous block, e.g. <ul><li>hello</li><br>^</ul>, the check may fail. So, we use isListItem
on the empty list node's siblings instead.

Source/WebCore:

No new tests : added new test case in the existing test (break-out-of-empty-list-item.html)

* editing/CompositeEditCommand.cpp:
(WebCore::CompositeEditCommand::breakOutOfEmptyListItem):

LayoutTests:

* editing/execCommand/break-out-of-empty-list-item-expected.txt:
* editing/execCommand/script-tests/break-out-of-empty-list-item.js:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113723 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL] DRT should support LayoutTestController's willSendRequestReturnsNull()
https://bugs.webkit.org/show_bug.cgi?id=82443

Patch by Christophe Dumez <christophe.dumez@intel.com> on 2012-04-10
Reviewed by Philippe Normand.

Tools:

EFL's DRT needs to support LayoutTestController's
willSendRequestReturnsNull() and return NULL when expected. This
allows for several tests to be removed from the skip list.

* DumpRenderTree/efl/DumpRenderTreeChrome.cpp:
(DumpRenderTreeChrome::onWillSendRequest):

LayoutTests:

Unskip tests which depend on EFL's DumpRenderTree support for
LayoutTestController's willSendRequestReturnsNull(), now that it is
implemented.

* platform/efl/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113722 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: Inspect Element - Edit as HTML - scrolling bug
https://bugs.webkit.org/show_bug.cgi?id=83563

mousedown in the scrollbar focuses the Elements tree outline container DIV element,
so the corresponding mousedown event should be consumed in the AT_TARGET phase.

Reviewed by Yury Semikhatsky.

* inspector/front-end/ElementsTreeOutline.js:
(WebInspector.ElementsTreeElement.prototype._startEditingAsHTML.consume):
(WebInspector.ElementsTreeElement.prototype._startEditingAsHTML.dispose):
(WebInspector.ElementsTreeElement.prototype._startEditingAsHTML):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113721 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][WK2] Assertion failure on loading new page after panning/zooming
https://bugs.webkit.org/show_bug.cgi?id=83049

Patch by Dinu Jacob <dinu.jacob@nokia.com> on 2012-04-10
Reviewed by Kenneth Rohde Christiansen.

When handling contents size change, set the visible content rectangle immediately,
instead of animating to it.

* UIProcess/qt/QtViewportInteractionEngine.cpp:
(WebKit::QtViewportInteractionEngine::itemSizeChanged):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113720 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL] LayoutTestController needs implementation of addUserStyleSheet
https://bugs.webkit.org/show_bug.cgi?id=82446

Source/WebKit/efl:

Add missing implementation addUserStyleSheet to EFL's
DumpRenderTreeSupport.

Patch by Sudarsana Nagineni <sudarsana.nagineni@linux.intel.com> on 2012-04-10
Reviewed by Antonio Gomes.

* WebCoreSupport/DumpRenderTreeSupportEfl.cpp:
(DumpRenderTreeSupportEfl::addUserStyleSheet):
* WebCoreSupport/DumpRenderTreeSupportEfl.h:

Tools:

Adding missing implementation addUserStyleSheet to EFL's LayoutTestController
so that we can unskip related tests from the skip list.

Patch by Sudarsana Nagineni <sudarsana.nagineni@linux.intel.com> on 2012-04-10
Reviewed by Antonio Gomes.

* DumpRenderTree/efl/LayoutTestControllerEfl.cpp:
(LayoutTestController::addUserStyleSheet): Implemented.

LayoutTests:

Unskipping the following test cases:
userscripts/mixed-case-stylesheet.html
userscripts/user-style-all-frames.html
userscripts/simple-stylesheet.html
userscripts/script-run-at-end.html

Patch by Sudarsana Nagineni <sudarsana.nagineni@linux.intel.com> on 2012-04-10
Reviewed by Antonio Gomes.

* platform/efl/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113719 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Adding myself to committers.py!

Unreviewed update to committers.py.

* Scripts/webkitpy/common/config/committers.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113718 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][WK2] Unreviewed gardening, skip a test to try to paint the bot greener.

* platform/qt-5.0-wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113717 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove the unused function WebCore::toIconIndex
https://bugs.webkit.org/show_bug.cgi?id=83037

Patch by Dinu Jacob <dinu.jacob@nokia.com> on 2012-04-10
Reviewed by Kentaro Hara.

* WebCore.order:
* dom/IconURL.cpp:
* dom/IconURL.h:
(WebCore):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113716 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][WK2] Title in MiniBrowser is not updated for a page with no title
https://bugs.webkit.org/show_bug.cgi?id=82483

Patch by Dinu Jacob <dinu.jacob@nokia.com> on 2012-04-10
Reviewed by Kenneth Rohde Christiansen.

Source/WebKit2:

titleChanged signal is not emitted on comitting a new load.

* UIProcess/API/qt/tests/html/basic_page.html:
* UIProcess/API/qt/tests/qquickwebview/tst_qquickwebview.cpp:
(tst_QQuickWebView):
(tst_QQuickWebView::titleUpdate):
* UIProcess/qt/QtWebPageLoadClient.cpp:
(QtWebPageLoadClient::didCommitLoadForFrame):

Tools:

Set window title to default if there is no page title.

* MiniBrowser/qt/BrowserWindow.cpp:
(BrowserWindow::BrowserWindow):
(BrowserWindow::onTitleChanged):
* MiniBrowser/qt/BrowserWindow.h:
(BrowserWindow):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113715 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][WK2] Implement PageClient::isViewWindowActive()
https://bugs.webkit.org/show_bug.cgi?id=81143

Reviewed by Kenneth Rohde Christiansen.

Source/WebKit2:

This patch implements PageClient::isViewWindowActive()
now that QQuickCanvas::isActive() is available (from QWindow).

* UIProcess/qt/QtPageClient.cpp:
(QtPageClient::isViewWindowActive):

Tools:

Fix the WrapperWindow from our PlatformWebView
so it correctly creates the platform related
stuff for QWindow, which is not created unless
QWindow::setVisible() or QWindow::show() are called.

* WebKitTestRunner/qt/PlatformWebViewQt.cpp:
(WTR::PlatformWebView::PlatformWebView):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113714 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Wk2] Gradening. Skip failing test.
https://bugs.webkit.org/show_bug.cgi?id=83581

Patch by Ádám Kallai <kadam@inf.u-szeged.hu> on 2012-04-10
Reviewed by Csaba Osztrogonác.

* platform/wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113713 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Unreviewed gardening. Skip a failing, a timeouting and a crashing test to paint the bots green.

* platform/qt-5.0-wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113712 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: make error a string on the front-end side, not an object.
https://bugs.webkit.org/show_bug.cgi?id=83570

Reviewed by Yury Semikhatsky.

Source/WebCore:

We already expect it to be a string in all the code, we never use the error code in it.

* inspector/front-end/InspectorBackend.js:
(InspectorBackendClass.prototype.dispatch):

LayoutTests:

* inspector/debugger/debugger-set-breakpoint-regex-expected.txt:
* inspector/debugger/debugger-set-breakpoint-regex.html:
* inspector/elements/resolve-node-blocked-expected.txt:
* inspector/elements/set-attribute.html:
* inspector/profiler/heap-snapshot-inspect-dom-wrapper-expected.txt:
* inspector/report-protocol-errors-expected.txt:
* inspector/report-protocol-errors.html:
* inspector/styles/set-property-boundaries-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113711 268f45cc-cd09-0410-ab3c-d52691b4dbfc

getElementsByTagName unable to find SVG camelCase elements imported into HTML
https://bugs.webkit.org/show_bug.cgi?id=46800

Reviewed by Nikolas Zimmermann.

Source/WebCore:

Implement getElementsByTagName according to the algorithm in
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-getelementsbytagname.
If the owner document of the element is html, the new class HTMLTagNodeList deals with
finding the match(es) based on this algorithm. If the owner document is non HTML, the existing
TagNodeList is used.

Implement getElementsByTagNameNS according to the algorithm in
http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#concept-getelementsbytagnamens.
The existing TagNodeList implements this algorithm.

Cached results may have less hits after this change. Tag names that are equal when case
insensitive but different when case sensitive can't share the result due to above algorithms, i.e.
they would match differently in the non html part of the document.

Tests: fast/dom/getElementsByTagName-localName-matching.html
       fast/dom/getElementsByTagName-localName-matching2.xhtml
       svg/dom/getElementsByTagName-localName-matching.html

* dom/Node.cpp:
(WebCore::Node::getElementsByTagName): create TagNodeList depending on owner document type.
(WebCore::Node::getElementsByTagNameNS): do not lowercase localName for html.
* dom/TagNodeList.cpp:
(WebCore::HTMLTagNodeList::HTMLTagNodeList): add specialized TagNodeList for html matching rules.
(WebCore):
(WebCore::HTMLTagNodeList::nodeMatches):
* dom/TagNodeList.h:
(TagNodeList):
(HTMLTagNodeList):
(WebCore::HTMLTagNodeList::create):

LayoutTests:

Add tests for getElementsByTagName(NS) in html, xhtml, and svg documents.

* fast/dom/getElementsByTagName-localName-matching-expected.txt: Added.
* fast/dom/getElementsByTagName-localName-matching.html: Added.
* fast/dom/getElementsByTagName-localName-matching2-expected.txt: Added.
* fast/dom/getElementsByTagName-localName-matching2.xhtml: Added.
* svg/dom/getElementsByTagName-localName-matching-expected.txt: Added.
* svg/dom/getElementsByTagName-localName-matching.html: Added.
* svg/dom/script-tests/getElementsByTagName-localName-matching.js: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113710 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: [Device Metrics] Introduce the "Fit window" option
https://bugs.webkit.org/show_bug.cgi?id=83280

Source/WebCore:

This change adds the "Fit window" option to the Settings pane of Web Inspector, which lets users
resize the FrameView to fit the WebViewImpl size. When the user resizes the browser window
in the device metrics override mode with the option enabled, the FrameView follows
to match WebViewImpl by at least one dimension.

Reviewed by Pavel Feldman.

* English.lproj/localizedStrings.js:
* inspector/Inspector.json:
* inspector/InspectorClient.h:
(WebCore::InspectorClient::overrideDeviceMetrics):
(WebCore::InspectorClient::autoZoomPageToFitWidth):
* inspector/InspectorInstrumentation.cpp:
(WebCore::InspectorInstrumentation::shouldApplyScreenWidthOverrideImpl):
(WebCore):
(WebCore::InspectorInstrumentation::shouldApplyScreenHeightOverrideImpl):
* inspector/InspectorInstrumentation.h:
(InspectorInstrumentation):
(WebCore::InspectorInstrumentation::shouldApplyScreenWidthOverride):
(WebCore):
(WebCore::InspectorInstrumentation::shouldApplyScreenHeightOverride):
* inspector/InspectorPageAgent.cpp:
(PageAgentState):
(WebCore::InspectorPageAgent::restore):
(WebCore::InspectorPageAgent::disable):
(WebCore::InspectorPageAgent::setDeviceMetricsOverride):
(WebCore::InspectorPageAgent::updateViewMetrics):
* inspector/InspectorPageAgent.h:
* inspector/front-end/Settings.js:
* inspector/front-end/SettingsScreen.js:
(WebInspector.SettingsScreen.prototype._createDeviceMetricsElement):
* inspector/front-end/UserAgentSupport.js:
* inspector/front-end/helpScreen.css:
(.help-content input[type=checkbox]):
* page/DOMWindow.cpp:
(WebCore::DOMWindow::innerHeight):
(WebCore::DOMWindow::innerWidth):

Source/WebKit/chromium:

Implement the scaling of the FrameView to the WebViewImpl size, handle WebViewImpl resizes.

Reviewed by Pavel Feldman.

* src/InspectorClientImpl.cpp:
(WebKit::InspectorClientImpl::overrideDeviceMetrics):
* src/InspectorClientImpl.h:
(InspectorClientImpl):
* src/WebDevToolsAgentImpl.cpp:
(WebKit::DeviceMetricsSupport::DeviceMetricsSupport):
(WebKit::DeviceMetricsSupport::setDeviceMetrics):
(WebKit::DeviceMetricsSupport::autoZoomPageToFitWidthOnNavigation):
(DeviceMetricsSupport):
(WebKit::DeviceMetricsSupport::autoZoomPageToFitWidth):
(WebKit::DeviceMetricsSupport::webViewResized):
(WebKit::DeviceMetricsSupport::applySizeOverrideIfNecessary):
(WebKit::DeviceMetricsSupport::ensureOriginalZoomFactor):
(WebKit::DeviceMetricsSupport::restore):
(WebKit::DeviceMetricsSupport::scaledEmulatedFrameSize):
(WebKit::DeviceMetricsSupport::forcedScrollbarDimensions):
(WebKit::DeviceMetricsSupport::applySizeOverrideInternal):
(WebKit::WebDevToolsAgentImpl::webViewResized):
(WebKit):
(WebKit::WebDevToolsAgentImpl::overrideDeviceMetrics):
(WebKit::WebDevToolsAgentImpl::autoZoomPageToFitWidth):
* src/WebDevToolsAgentImpl.h:
(WebDevToolsAgentImpl):
* src/WebDevToolsAgentPrivate.h:
(WebDevToolsAgentPrivate):
* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::resize):

LayoutTests:

Update the test to follow the protocol change and test the "Fit window" parameter.

Reviewed by Pavel Feldman.

* inspector/styles/override-screen-size-expected.txt:
* inspector/styles/override-screen-size.html:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113709 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: use maxJSObjectId that is provided by back-end.
https://bugs.webkit.org/show_bug.cgi?id=82451

Summary view can filter objects in snapshot. It uses maxJSObjectId for this.
There was no such field in the profile header at the
moment but I've landed a patch in downstream.

Reviewed by Yury Semikhatsky.

* bindings/js/ScriptHeapSnapshot.h:
(WebCore):
(WebCore::ScriptHeapSnapshot::maxSnapshotJSObjectId):
* bindings/v8/ScriptHeapSnapshot.cpp:
(WebCore::ScriptHeapSnapshot::maxSnapshotJSObjectId):
(WebCore):
* bindings/v8/ScriptHeapSnapshot.h:
(WebCore):
(ScriptHeapSnapshot):
* inspector/InspectorProfilerAgent.cpp:
(WebCore::InspectorProfilerAgent::createSnapshotHeader):
* inspector/front-end/DetailedHeapshotView.js:
(WebInspector.HeapSnapshotConstructorsDataGrid):
(WebInspector.HeapSnapshotConstructorsDataGrid.prototype.setDataSource):
(WebInspector.HeapSnapshotConstructorsDataGrid.prototype.populateChildren):
(WebInspector.HeapSnapshotConstructorsDataGrid.prototype._filterSelectIndexChanged):
(WebInspector.DetailedHeapshotView.prototype._changeFilter):
* inspector/front-end/HeapSnapshot.js:
(WebInspector.HeapSnapshot.prototype.updateStaticData):
* inspector/front-end/HeapSnapshotProxy.js:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113708 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: annotate front-end for newer closure compiler.
https://bugs.webkit.org/show_bug.cgi?id=83478

Reviewed by Yury Semikhatsky.

This change mostly adds annotations to bound functions and !!bool expressions.
As a result, it makes closure compiler v20120305 happy.

* inspector/Inspector.json:
* inspector/front-end/BreakpointManager.js:
* inspector/front-end/CSSStyleModel.js:
* inspector/front-end/ConsoleView.js:
(WebInspector.ConsoleView.prototype.evalInInspectedWindow.evalCallback):
(WebInspector.ConsoleView.prototype.evalInInspectedWindow):
* inspector/front-end/DOMAgent.js:
(WebInspector.DOMNode.prototype.removeAttribute):
(WebInspector.DOMAgent.prototype.pushNodeToFrontend):
* inspector/front-end/DebuggerModel.js:
* inspector/front-end/DebuggerPresentationModel.js:
(WebInspector.DebuggerPresentationModelResourceBinding.prototype.canSetContent):
* inspector/front-end/ElementsPanel.js:
* inspector/front-end/ExtensionServer.js:
(WebInspector.ExtensionServer.prototype._onEvaluateOnInspectedPage):
* inspector/front-end/ObjectPropertiesSection.js:
* inspector/front-end/RemoteObject.js:
* inspector/front-end/Resource.js:
(WebInspector.ResourceDomainModelBinding.prototype.canSetContent):
* inspector/front-end/ResourceUtils.js:
* inspector/front-end/ScopeChainSidebarPane.js:
* inspector/front-end/Script.js:
(WebInspector.Script.prototype.editSource):
* inspector/front-end/ScriptFormatter.js:
(WebInspector.ScriptFormatter.prototype.get _worker):
* inspector/front-end/WorkerManager.js:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113707 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Gardening after r113701.

Patch by Ádám Kallai <kadam@inf.u-szeged.hu> on 2012-04-10
Reviewed by Csaba Osztrogonác.

* platform/wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113706 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, skip a failing test and rebaseline another one after r113611.

* platform/gtk/Skipped:
* platform/gtk/scrollbars/scrollbars-on-positioned-content-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113705 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL] Unreviewed. Gardening Add 19 missing expectations.

* platform/efl/Skipped:
* platform/efl/http/tests/loading/simple-subframe-expected.txt: Added.
* platform/efl/http/tests/local/file-url-sent-as-referer-expected.txt: Added.
* platform/efl/http/tests/misc/acid2-expected.txt: Added.
* platform/efl/http/tests/misc/error404-expected.txt: Added.
* platform/efl/http/tests/misc/favicon-as-image-expected.txt: Added.
* platform/efl/http/tests/misc/frame-access-during-load-expected.txt: Added.
* platform/efl/http/tests/misc/generated-content-inside-table-expected.txt: Added.
* platform/efl/http/tests/misc/iframe404-expected.txt: Added.
* platform/efl/http/tests/misc/location-replace-crossdomain-expected.txt: Added.
* platform/efl/http/tests/misc/object-embedding-svg-delayed-size-negotiation-2-expected.txt: Added.
* platform/efl/http/tests/misc/object-embedding-svg-delayed-size-negotiation-expected.txt: Added.
* platform/efl/http/tests/misc/slow-loading-image-in-pattern-expected.txt: Added.
* platform/efl/http/tests/navigation/error404-basic-expected.txt: Added.
* platform/efl/http/tests/navigation/error404-goback-expected.txt: Added.
* platform/efl/http/tests/navigation/error404-subframeload-expected.txt: Added.
* platform/efl/http/tests/navigation/javascriptlink-frames-expected.txt: Added.
* platform/efl/http/tests/navigation/postredirect-basic-expected.txt: Added.
* platform/efl/http/tests/navigation/postredirect-goback1-expected.txt: Added.
* platform/efl/http/tests/uri/css-href-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113704 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt][WK2] TestNetscapePlugin is broken
https://bugs.webkit.org/show_bug.cgi?id=83024

Reviewed by Csaba Osztrogonác.

Tools:

* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
Fix X11 availability check. embedded is always true with QPA.
* Tools.pro:
Build the test plugin if plugins are enabled for WebKit2.
* qmake/mkspecs/features/features.prf:
* qmake/mkspecs/features/functions.prf:
Added a convenience function to determine availability
of X11 libraries.

LayoutTests:

Remove plugins from skiplist and put the rest
of failing tests to their group.

* platform/qt-5.0-wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113703 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[WIN] Fix build without precompiled header.

* DumpRenderTree/cg/ImageDiffCG.cpp: Define max as max as we do for min.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113702 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Gardening. Skip failing tests because of missing DnD support and ENABLE(SHADOW_DOM) is disabled.
[WK2] Skip tests because of missing layoutTestController.setAlwaysAcceptCookies().

Patch by Ádám Kallai <kadam@inf.u-szeged.hu> on 2012-04-10
Reviewed by Csaba Osztrogonác.

* platform/qt/Skipped:
* platform/wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113701 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Web Inspector: searching document with no documentElement results in 0x0 access.
https://bugs.webkit.org/show_bug.cgi?id=83483

Reviewed by Yury Semikhatsky.

Source/WebCore:

Added the null check.

Test: inspector/elements/dom-search-crash.html

* inspector/InspectorDOMAgent.cpp:
(WebCore::InspectorDOMAgent::performSearch):

LayoutTests:

* inspector/elements/dom-search-crash-expected.txt: Added.
* inspector/elements/dom-search-crash.html: Added.
* inspector/elements/resources/dom-search-crash-iframe.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113700 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, GTK baselines for two new tests.

* platform/gtk/fast/events/drop-with-file-paths-expected.txt: Added.
* platform/gtk/media/nodesFromRect-shadowContent-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113698 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[WK2][GTK] FullScreen signals
https://bugs.webkit.org/show_bug.cgi?id=76166

Reviewed by Gustavo Noronha Silva.

* GNUmakefile.am: Add new files to compilation.
* UIProcess/API/C/gtk/WKFullScreenClientGtk.cpp:
(WKViewSetFullScreenClientGtk): Initialize WebKitWebViewBase
fullscreen client.
* UIProcess/API/C/gtk/WKFullScreenClientGtk.h:
* UIProcess/API/gtk/WebKitFullscreenClient.cpp: Added.
(willEnterFullScreen): Call webkitWebViewEnterFullScreen().
(willExitFullScreen): Call webkitWebViewLeaveFullScreen().
(attachFullScreenClientToView): Initialize FullScreenClient adding
implementations for willEnterFullScreen and willExitFullScreen callbacks.
* UIProcess/API/gtk/WebKitFullscreenClient.h: Added.
* UIProcess/API/gtk/WebKitPrivate.h:
* UIProcess/API/gtk/WebKitWebView.cpp:
(webkitWebViewConstructed): Attach fullscreen client to view.
(webkit_web_view_class_init): Add WebKitWebView::enter-fullscreen
and WebKitWebView::leave-fullscreen signals.
(webkitWebViewEnterFullScreen): Emit
WebKitWebView::enter-fullscreen signal.
(webkitWebViewLeaveFullScreen): Emit
WebKitWebView::leave-fullscreen signal.
* UIProcess/API/gtk/WebKitWebView.h:
* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseKeyPressEvent): Only return early when leaving
fullscreen, otherwise let the view process the key pressed.
(webkitWebViewBaseEnterFullScreen): Return early if
willEnterFullScreen callback is handled and returns false.
(webkitWebViewBaseExitFullScreen): Return early if
willExitFullScreen callback is handled and returns false.
(webkitWebViewBaseInitializeFullScreenClient): Initialize the
fullscreen client.
* UIProcess/API/gtk/WebKitWebViewBasePrivate.h:
* UIProcess/API/gtk/WebKitWebViewPrivate.h:
* UIProcess/API/gtk/tests/TestWebKitWebView.cpp:
(testWebViewFullScreen):
(beforeAll):
* UIProcess/API/gtk/tests/WebViewTest.cpp:
(WebViewTest::keyStroke): Helper function to synthesize key
press/release events.
* UIProcess/API/gtk/tests/WebViewTest.h:
* UIProcess/API/gtk/webkit2marshal.list:
* UIProcess/gtk/WebFullScreenClientGtk.cpp:
(WebKit::WebFullScreenClientGtk::willEnterFullScreen): Call
willEnterFullScreen callback if defined.
(WebKit::WebFullScreenClientGtk::willExitFullScreen): Call
willExitFullScreen callback if defined.
* UIProcess/gtk/WebFullScreenClientGtk.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113697 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[CMake] Add missing source files to build system

Source/WebCore:

* CMakeLists.txt:

Source/WebKit2:

* CMakeLists.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113696 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed. Fixed AppleWin and Chromium-win build failure.

Renamed 'exceptionNameDescription' to per-file variable name to avoid
name conflict.

* Modules/indexeddb/IDBDatabaseException.cpp:
(WebCore::IDBDatabaseException::initializeDescription):
* Modules/webdatabase/SQLException.cpp:
(WebCore::SQLException::initializeDescription):
* dom/EventException.cpp:
(WebCore::EventException::initializeDescription):
* dom/RangeException.cpp:
(WebCore::RangeException::initializeDescription):
* fileapi/FileException.cpp:
(WebCore::FileException::initializeDescription):
* fileapi/OperationNotAllowedException.cpp:
(WebCore::OperationNotAllowedException::initializeDescription):
* svg/SVGException.cpp:
(WebCore::SVGException::initializeDescription):
* xml/XPathException.cpp:
(WebCore::XPathException::initializeDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113695 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed: adding myself to committers.py.

* Scripts/webkitpy/common/config/committers.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113694 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed. Fixed AppleWin and Chromium-win build failure.

Renamed 'exceptions' to 'exceptionNameDescription' to avoid
variable name conflict.

* Modules/indexeddb/IDBDatabaseException.cpp:
(WebCore::IDBDatabaseException::initializeDescription):
* Modules/webdatabase/SQLException.cpp:
(WebCore::SQLException::initializeDescription):
* dom/EventException.cpp:
(WebCore::EventException::initializeDescription):
* dom/RangeException.cpp:
(WebCore::RangeException::initializeDescription):
* fileapi/FileException.cpp:
(WebCore::FileException::initializeDescription):
* fileapi/OperationNotAllowedException.cpp:
(WebCore::OperationNotAllowedException::initializeDescription):
* svg/SVGException.cpp:
(WebCore::SVGException::initializeDescription):
* xml/XPathException.cpp:
(WebCore::XPathException::initializeDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113693 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, GTK rebaseline.

* platform/gtk/editing/pasteboard/interchange-newline-2-expected.txt:
* platform/gtk/editing/selection/editable-html-element-expected.txt:
* platform/gtk/fast/overflow/hidden-scrollbar-resize-expected.txt:
* platform/gtk/fast/repaint/overflow-outline-repaint-expected.txt:
* platform/gtk/http/tests/xmlhttprequest/send-array-buffer-expected.txt:
* platform/gtk/http/tests/xmlhttprequest/send-undefined-and-null-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113692 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed: adding myself to committers.py.

* Scripts/webkitpy/common/config/committers.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113691 268f45cc-cd09-0410-ab3c-d52691b4dbfc

combine two arrays in XPathException into one
https://bugs.webkit.org/show_bug.cgi?id=83442

Patch by Lu Guanqun <guanqun.lu@intel.com> on 2012-04-09
Reviewed by Kentaro Hara.

No new tests required.

* xml/XPathException.cpp:
(XPathExceptionNameDescription):
(WebCore):
(WebCore::XPathException::initializeDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113690 268f45cc-cd09-0410-ab3c-d52691b4dbfc

combine exceptionNames and exceptionDescriptions in IDBDatabaseException.cpp into one array
https://bugs.webkit.org/show_bug.cgi?id=83433

Patch by Lu Guanqun <guanqun.lu@intel.com> on 2012-04-09
Reviewed by Kentaro Hara.

No new tests required.

* Modules/indexeddb/IDBDatabaseException.cpp:
(IDBDatabaseExceptionNameDescription):
(WebCore):
(WebCore::IDBDatabaseException::initializeDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113689 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Chromium] Remove unused variable, WebViewImpl::m_lastMousePosition.
https://bugs.webkit.org/show_bug.cgi?id=83550

Reviewed by Kentaro Hara.

* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::WebViewImpl):
(WebKit::WebViewImpl::mouseMove):
* src/WebViewImpl.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113688 268f45cc-cd09-0410-ab3c-d52691b4dbfc

combine two arrays in SQLException.cpp into one
https://bugs.webkit.org/show_bug.cgi?id=83452

Patch by Lu Guanqun <guanqun.lu@intel.com> on 2012-04-09
Reviewed by Kentaro Hara.

No new tests required.

* Modules/webdatabase/SQLException.cpp:
(SQLExceptionNameDescription):
(WebCore):
(WebCore::SQLException::initializeDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113687 268f45cc-cd09-0410-ab3c-d52691b4dbfc

combine two arrays in EventException.cpp into one
https://bugs.webkit.org/show_bug.cgi?id=83451

Patch by Lu Guanqun <guanqun.lu@intel.com> on 2012-04-09
Reviewed by Kentaro Hara.

No new tests required.

* dom/EventException.cpp:
(EventExceptionNameDescription):
(WebCore):
(WebCore::EventException::initializeDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113686 268f45cc-cd09-0410-ab3c-d52691b4dbfc

combine two arrays in RangeException.cpp into one
https://bugs.webkit.org/show_bug.cgi?id=83450

Patch by Lu Guanqun <guanqun.lu@intel.com> on 2012-04-09
Reviewed by Kentaro Hara.

No new tests required.

* dom/RangeException.cpp:
(RangeExceptionNameDescription):
(WebCore):
(WebCore::RangeException::initializeDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113685 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Texmap] Improve TextureMapperGL readability
https://bugs.webkit.org/show_bug.cgi?id=83477

Change the debugging macros for TextureMapperGL, so that the GL_CMD call can include a semicolon.
Modify some if blocks to have an early return.

Reviewed by Martin Robinson.

No new functionality.

* platform/graphics/texmap/TextureMapperGL.cpp:
(WebCore::TextureMapperGLData::SharedGLData::scissorClip):
(WebCore::TextureMapperGLData::SharedGLData::applyCurrentClip):
(WebCore::TextureMapperGLData::SharedGLData::~SharedGLData):
(WebCore::TextureMapperGLData::initializeStencil):
(WebCore::TextureMapperGL::beginPainting):
(WebCore::TextureMapperGL::endPainting):
(WebCore::TextureMapperGL::drawTexture):
(WebCore::BitmapTextureGL::didReset):
(WebCore::BitmapTextureGL::updateContents):
(WebCore::BitmapTextureGL::initializeStencil):
(WebCore::BitmapTextureGL::bind):
(WebCore::BitmapTextureGL::~BitmapTextureGL):
(WebCore::TextureMapperGL::bindSurface):
(WebCore::TextureMapperGL::beginClip):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113684 268f45cc-cd09-0410-ab3c-d52691b4dbfc

combine two arrays in FileException.cpp into one
https://bugs.webkit.org/show_bug.cgi?id=83449

Patch by Lu Guanqun <guanqun.lu@intel.com> on 2012-04-09
Reviewed by Kentaro Hara.

No new tests required.

* fileapi/FileException.cpp:
(FileExceptionNameDescription):
(WebCore):
(WebCore::FileException::initializeDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113683 268f45cc-cd09-0410-ab3c-d52691b4dbfc

combine two arrays in OperationNotAllowedException.cpp into one
https://bugs.webkit.org/show_bug.cgi?id=83445

Patch by Lu Guanqun <guanqun.lu@intel.com> on 2012-04-09
Reviewed by Kentaro Hara.

No new tests required.

* fileapi/OperationNotAllowedException.cpp:
(OperationNotAllowedExceptionNameDescription):
(WebCore):
(WebCore::OperationNotAllowedException::initializeDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113682 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Assert triggers VisibleSelection::adjustSelectionToAvoidCrossingShadowBoundaries.
https://bugs.webkit.org/show_bug.cgi?id=83118

Reviewed by Hajime Morita.

Source/WebCore:

VisibleSelection::adjustSelectionToAvoidCrossingShadowBoundaries did not work correctly.
There was a case that selection crosses shadow boundaries.

This patch introduces a class TreeScopeAdjuster, which enables us to adjust Node or Position
into a specified tree scope.

Test: fast/dom/shadow/selections-in-shadow.html

* CMakeLists.txt:
* GNUmakefile.list.am:
* Target.pri:
* WebCore.ext.in:
* WebCore.gypi:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* dom/DOMAllInOne.cpp:
* dom/TreeScope.h:
(TreeScope):
* dom/TreeScopeAdjuster.cpp: Added.
(WebCore):
(WebCore::TreeScopeAdjuster::TreeScopeAdjuster):
(WebCore::TreeScopeAdjuster::ancestorInThisScope):
  Returns the node itself if it's in the same tree scope.
  Otherwise, this method checks the shadow ancestor of the node recursively.
  If no corresponding node is found, 0 will be returned.
(WebCore::TreeScopeAdjuster::adjustPositionBefore):
(WebCore::TreeScopeAdjuster::adjustPositionAfter):
* dom/TreeScopeAdjuster.h: Added.
(WebCore):
(TreeScopeAdjuster):
(WebCore::TreeScopeAdjuster::treeScope):
* editing/VisibleSelection.cpp:
(WebCore::VisibleSelection::adjustSelectionToAvoidCrossingShadowBoundaries):

LayoutTests:

This tests checks all the combination of element not in a shadow tree, element in a shadow tree,
element in a nested shadow tree, element in a non-youngest shadow tree, element in a nested shadow tree
in non-youngest shadow treee.

* fast/dom/resources/event-sender-util.js: Added.
(mouseMoveToElem):
* fast/dom/shadow/selections-in-shadow-expected.txt: Added.
* fast/dom/shadow/selections-in-shadow.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113681 268f45cc-cd09-0410-ab3c-d52691b4dbfc

combine two arrays in SVGException.cpp into one
https://bugs.webkit.org/show_bug.cgi?id=83444

Patch by Lu Guanqun <guanqun.lu@intel.com> on 2012-04-09
Reviewed by Kentaro Hara.

No new tests required.

* svg/SVGException.cpp:
(SVGExceptionNameDescription):
(WebCore):
(WebCore::SVGException::initializeDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113680 268f45cc-cd09-0410-ab3c-d52691b4dbfc

combine two arrays in XMLHttpRequestException.cpp into one
https://bugs.webkit.org/show_bug.cgi?id=83443

Patch by Lu Guanqun <guanqun.lu@intel.com> on 2012-04-09
Reviewed by Kentaro Hara.

No new tests requied.

* xml/XMLHttpRequestException.cpp:
(XMLHttpRequestExceptionNameDescription):
(WebCore):
(WebCore::XMLHttpRequestException::initializeDescription):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113679 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[WK2] Enable using a single ShareableBitmap for multiple updates
https://bugs.webkit.org/show_bug.cgi?id=83424

Source/WebCore:

Modify BitmapTexture::updateContents to include a source offset.
This allows us to update a texture from a sub-image.

Reviewed by Kenneth Rohde Christiansen.

Tested by existing API tests.

* platform/graphics/texmap/TextureMapper.h:
(BitmapTexture):
* platform/graphics/texmap/TextureMapperGL.cpp:
(WebCore::BitmapTextureGL::didReset):
(WebCore::driverSupportsBGRASwizzling):
(WebCore::BitmapTextureGL::updateContents):
* platform/graphics/texmap/TextureMapperGL.h:
(WebCore::BitmapTextureGL::textureTarget):
(BitmapTextureGL):
* platform/graphics/texmap/TextureMapperImageBuffer.cpp:
(WebCore::BitmapTextureImageBuffer::updateContents):
* platform/graphics/texmap/TextureMapperImageBuffer.h:
(BitmapTextureImageBuffer):

Source/WebKit2:

Reviewed by Kenneth Rohde Christiansen.

Enabled creating a GraphicsContext that references a rect inside ShareableBitmap.
Added bitmapOffset to UpdateInfo, to allow updates to reference an offset inside the bitmap.
Added UpdateAtlas, a class that manages available rects in a larger ShareableBitmap.

In this iteration, UpdateAtlas has a simple behavior where a rect inside the bitmap is
either available or used. When the buffers are swapped, all used buffers become available
again. A future enhancement might allow triple-buffering, where available rects can be
updated while other rects from the same atlas are in use.

Added the necessary plumbing to allow Qt's UI-side compositing to take advantage of
UpdateAtlas. LayerTreeHostQt creates an atlas per type of bitmap (i.e. one opaque and one
alpha-enabled atlas). When a tile wants to update and there's no available scratch-buffer,
the update would be postponed to the next frame, creating a tiling-artifact in low memory
situations.

* Shared/UpdateInfo.cpp:
(WebKit::UpdateInfo::encode):
(WebKit::UpdateInfo::decode):
* Shared/UpdateInfo.h:
(UpdateInfo):
* Target.pri:
* UIProcess/LayerTreeHostProxy.cpp:
(WebKit::LayerTreeHostProxy::updateTileForLayer):
* UIProcess/WebLayerTreeRenderer.cpp:
(WebKit::WebLayerTreeRenderer::syncLayerParameters):
(WebKit::WebLayerTreeRenderer::updateTile):
(WebKit::WebLayerTreeRenderer::createImage):
* UIProcess/WebLayerTreeRenderer.h:
(TileUpdate):
(WebKit::WebLayerTreeRenderer::TileUpdate::TileUpdate):
(WebLayerTreeRenderer):
* UIProcess/qt/LayerBackingStore.cpp:
(WebKit::LayerBackingStoreTile::swapBuffers):
(WebKit::LayerBackingStoreTile::setBackBuffer):
(WebKit::LayerBackingStore::createTile):
* UIProcess/qt/LayerBackingStore.h:
(LayerBackingStoreTile):
(LayerBackingStore):
* WebProcess/WebCoreSupport/WebGraphicsLayer.cpp:
(WebCore::WebGraphicsLayer::tiledBackingStoreBackgroundColor):
(WebCore::WebGraphicsLayer::beginContentUpdate):
(WebCore):
* WebProcess/WebCoreSupport/WebGraphicsLayer.h:
(WebGraphicsLayerClient):
(WebGraphicsLayer):
* WebProcess/WebPage/TiledBackingStoreRemoteTile.cpp:
(WebKit::TiledBackingStoreRemoteTile::updateBackBuffer):
* WebProcess/WebPage/TiledBackingStoreRemoteTile.h:
(TiledBackingStoreRemoteTileClient):
* WebProcess/WebPage/UpdateAtlas.cpp: Added.
(WebKit):
(WebKit::UpdateAtlas::UpdateAtlas):
(WebKit::nextPowerOfTwo):
(WebKit::UpdateAtlas::buildLayoutIfNeeded):
(WebKit::UpdateAtlas::findAvailableIndex):
(WebKit::UpdateAtlas::didSwapBuffers):
(WebKit::UpdateAtlas::acquireScratchBuffer):
(WebKit::UpdateAtlas::offsetForIndex):
* WebProcess/WebPage/UpdateAtlas.h: Added.
(WebCore):
(WebKit):
(UpdateAtlas):
(WebKit::UpdateAtlas::bitmap):
(WebKit::UpdateAtlas::size):
(WebKit::UpdateAtlas::flags):
* WebProcess/WebPage/qt/LayerTreeHostQt.cpp:
(WebKit::LayerTreeHostQt::adoptImageBackingStore):
(WebKit::LayerTreeHostQt::renderNextFrame):
(WebKit::LayerTreeHostQt::purgeBackingStores):
(WebKit):
(WebKit::LayerTreeHostQt::getAtlas):
(WebKit::LayerTreeHostQt::beginContentUpdate):
* WebProcess/WebPage/qt/LayerTreeHostQt.h:
(LayerTreeHostQt):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113678 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Viewport is not filled when out of texture memory on mac
https://bugs.webkit.org/show_bug.cgi?id=83351

Reviewed by Adrienne Walker.

Source/Platform:

* chromium/public/WebLayerTreeView.h:
(WebLayerTreeView):

Source/WebCore:

Currently we add gutter quads on the NonCompositedContentHost layer,
which lies above another visible layer - the rubberband layer on mac.
For this reason, on mac, gutter quads were disabled, as well as forcing
the NCCH layer to draw, in order to make the rubberband layer appear.

We move the logic for adding gutter quads into CCRenderPass, and add
gutter quads for all pixels that are visible through the viewport.
This allows us to stop special-casing the NCCH layer, and allows us
to skip drawing the layer equally with other layers. We remove the
backgroundCoversViewport() flag entirely.

In order to do this, we fix a bug in the occlusion tracker, that
allowed opaque() layers with skipsDraw to occlude, by making
the visibleContentOpaqueRegion() method on the layer classes
also return the opaque region for non-tiled layers, always use
its value in the occlusion tracker.

Unit test: CCLayerTreeHostImplTest.viewportCovered

* platform/graphics/chromium/LayerChromium.cpp:
(WebCore::LayerChromium::LayerChromium):
(WebCore::LayerChromium::pushPropertiesTo):
(WebCore::LayerChromium::visibleContentOpaqueRegion):
(WebCore):
* platform/graphics/chromium/LayerChromium.h:
(LayerChromium):
* platform/graphics/chromium/TiledLayerChromium.cpp:
(WebCore::TiledLayerChromium::prepareToUpdateTiles):
(WebCore::TiledLayerChromium::visibleContentOpaqueRegion):
* platform/graphics/chromium/cc/CCLayerImpl.cpp:
(WebCore::CCLayerImpl::CCLayerImpl):
(WebCore::CCLayerImpl::visibleContentOpaqueRegion):
(WebCore):
* platform/graphics/chromium/cc/CCLayerImpl.h:
(WebCore::CCLayerImpl::appendQuads):
(CCLayerImpl):
* platform/graphics/chromium/cc/CCLayerTreeHost.cpp:
(WebCore::CCLayerTreeHost::CCLayerTreeHost):
(WebCore::CCLayerTreeHost::finishCommitOnImplThread):
* platform/graphics/chromium/cc/CCLayerTreeHost.h:
(WebCore::CCLayerTreeHost::setBackgroundColor):
(CCLayerTreeHost):
* platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp:
(WebCore::CCLayerTreeHostImpl::calculateRenderPasses):
* platform/graphics/chromium/cc/CCLayerTreeHostImpl.h:
(WebCore::CCLayerTreeHostImpl::backgroundColor):
(WebCore::CCLayerTreeHostImpl::setBackgroundColor):
(CCLayerTreeHostImpl):
* platform/graphics/chromium/cc/CCOcclusionTracker.cpp:
(WebCore::computeOcclusionBehindLayer):
(WebCore::::markOccludedBehindLayer):
* platform/graphics/chromium/cc/CCOcclusionTracker.h:
(CCOcclusionTrackerBase):
(WebCore::CCOcclusionTrackerBase::computeVisibleRegionInScreen):
* platform/graphics/chromium/cc/CCRenderPass.cpp:
(WebCore::CCRenderPass::appendQuadsToFillScreen):
(WebCore):
* platform/graphics/chromium/cc/CCRenderPass.h:
(WebCore):
(CCRenderPass):
* platform/graphics/chromium/cc/CCTiledLayerImpl.cpp:
(WebCore::CCTiledLayerImpl::appendQuads):
(WebCore::CCTiledLayerImpl::visibleContentOpaqueRegion):
* platform/graphics/chromium/cc/CCTiledLayerImpl.h:
(WebCore::CCTiledLayerImpl::skipsDraw):
(CCTiledLayerImpl):

Source/WebKit/chromium:

* src/NonCompositedContentHost.cpp:
(WebKit::NonCompositedContentHost::NonCompositedContentHost):
(WebKit::NonCompositedContentHost::setBackgroundColor):
* src/WebLayerTreeView.cpp:
(WebKit::WebLayerTreeView::setBackgroundColor):
(WebKit):
* src/WebViewImpl.cpp:
(WebKit::WebViewImpl::setBackgroundColor):
(WebKit):
(WebKit::WebViewImplContentPainter::paint):
(WebKit::WebViewImpl::setIsAcceleratedCompositingActive):
* src/WebViewImpl.h:
(WebCore):
(WebViewImpl):
* tests/CCLayerImplTest.cpp:
(WebCore::TEST):
* tests/CCLayerTreeHostImplTest.cpp:
(WebKitTests::BlendStateCheckLayer::appendQuads):
(WebKitTests::TEST_F):
(WebKitTests):
* tests/CCLayerTreeHostTest.cpp:
(WTF):
(CCLayerTreeHostTestCommit):
(WTF::CCLayerTreeHostTestCommit::CCLayerTreeHostTestCommit):
(WTF::CCLayerTreeHostTestCommit::beginTest):
(WTF::CCLayerTreeHostTestCommit::commitCompleteOnCCThread):
(WTF::CCLayerTreeHostTestCommit::afterTest):
(WTF::TEST_F):
(TestLayerChromium):
* tests/CCOcclusionTrackerTest.cpp:
(WebKitTests::TestContentLayerChromium::TestContentLayerChromium):
(WebKitTests::TestContentLayerChromium::visibleContentOpaqueRegion):
(WebKitTests::TestContentLayerChromium::setOpaqueContentsRect):
(TestContentLayerChromium):
(WebKitTests::TestContentLayerImpl::TestContentLayerImpl):
(TestContentLayerImpl):
(WebKitTests::TestContentLayerImpl::visibleContentOpaqueRegion):
(WebKitTests::TestContentLayerImpl::setOpaqueContentsRect):
* tests/CCTiledLayerImplTest.cpp:
* tests/LayerChromiumTest.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113677 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Don't let CSSPropertyWebkitPerspective dereference primitiveValue without null check.
https://bugs.webkit.org/show_bug.cgi?id=83538

Reviewed by Daniel Bates.

No new tests / code cleanup only.

Coverity pointed out that we potentially dereference primitiveValue here without checking for null.
I've added an early out for that case to make sure it can't ever happen. I don't know if it's actually
possible to exercise that code path or not - probably the parser prevents it from being hit in practice.

* css/CSSStyleSelector.cpp:
(WebCore::CSSStyleSelector::collectMatchingRulesForList):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113676 268f45cc-cd09-0410-ab3c-d52691b4dbfc

<webkit.org/b/83539> Web Inspector: ASSERT attempting to unbind null contentDocument

Avoid calling unbind with a null object. The HTMLFrameOwnerElement's
contentDocument can be null.

Patch by Joseph Pecoraro <pecoraro@apple.com> on 2012-04-09
Reviewed by Timothy Hatcher.

* inspector/InspectorDOMAgent.cpp:
(WebCore::InspectorDOMAgent::unbind):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113675 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[BlackBerry] Parsed Cookie's m_isMaxAgeSet is not needed.
https://bugs.webkit.org/show_bug.cgi?id=83457

Patch by Jason Liu <jason.liu@torchmobile.com.cn> on 2012-04-09
Reviewed by Rob Buis.

When m_expiry is not -1, it must be set by setExpiry or setMaxAge. setExpiry will return when m_expiry
has been set. This ensures Max-Age's precedence.
And m_isMaxAgeSet is always false when the cookie is from database. This is not right.

So we use m_expiry instead of m_isMaxAgeSet.

No new tests. Refactor.

* platform/blackberry/ParsedCookie.cpp:
(WebCore::ParsedCookie::ParsedCookie):
(WebCore::ParsedCookie::setExpiry):
(WebCore::ParsedCookie::setMaxAge):
* platform/blackberry/ParsedCookie.h:
(ParsedCookie):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113674 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fixed ordering of deleting CCInputHandler and CCLayerTreeHostImpl in layerTreeHostClosedOnImplThread
https://bugs.webkit.org/show_bug.cgi?id=83488

Patch by Min Qin <qinmin@google.com> on 2012-04-09
Reviewed by James Robinson.

CCInputHandler(WebCompositorInputHandlerImpl) has a pointer to CCLayerTreeHostImpl.
So we should delete CCInputHandler earlier as otherwise that pointer could get misused after LayerTreeHostImpl is deleted.
Just fixing a potential NPE error, no behavior change.

* platform/graphics/chromium/cc/CCThreadProxy.cpp:
(WebCore::CCThreadProxy::layerTreeHostClosedOnImplThread):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113673 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[BlackBerry] Update about:config page
https://bugs.webkit.org/show_bug.cgi?id=83515

Reviewed by Rob Buis.

* WebCoreSupport/AboutData.cpp:
(WebCore::configPage):
* WebCoreSupport/AboutDataEnableFeatures.in:
* WebCoreSupport/AboutDataHaveFeatures.in:
* WebCoreSupport/AboutDataUseFeatures.in:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113672 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, rolling out r113656.
http://trac.webkit.org/changeset/113656
https://bugs.webkit.org/show_bug.cgi?id=83542

test is broken on chromium-win (Requested by simonjam on
#webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-04-09

Source/WebCore:

* Modules/intents/Intent.cpp:
(WebCore::Intent::Intent):
* Modules/intents/Intent.h:
(Intent):
* Modules/intents/Intent.idl:
* bindings/scripts/CodeGeneratorV8.pm:
(GenerateParametersCheck):
* bindings/scripts/test/V8/V8TestSerializedScriptValueInterface.cpp:

Source/WebKit/chromium:

* public/WebIntent.h:
(WebIntent):
* src/WebIntent.cpp:

Tools:

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

LayoutTests:

* webintents/web-intents-invoke-port-expected.txt: Removed.
* webintents/web-intents-invoke-port.html: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113671 268f45cc-cd09-0410-ab3c-d52691b4dbfc

ASSERTION FAILED: !attached() in Node::attach.
https://bugs.webkit.org/show_bug.cgi?id=80726

Reviewed by Adam Barth.

Source/WebCore:

While parsing XML document, prevent attaching the leaf text node
back to document, if its parent is not attached.

Test: fast/dom/text-node-attach-crash.xhtml

* xml/parser/XMLDocumentParser.cpp:
(WebCore::XMLDocumentParser::exitText):

LayoutTests:

* fast/dom/text-node-attach-crash-expected.txt: Added.
* fast/dom/text-node-attach-crash.xhtml: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113670 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Chromium] Unreviewed gardening.

* platform/chromium-mac-snowleopard/fast/forms/basic-textareas-expected.txt: Added.
* platform/chromium-mac-snowleopard/media/audio-repaint-expected.png:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113669 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[chromium] Replica layers are not drawn, so should not be painted
https://bugs.webkit.org/show_bug.cgi?id=83504

Reviewed by James Robinson.

We attempt to paint replica layers when we are painting the mask layers
in CCLayerTreeHost.cpp.

This is useless code, replica layers have drawsContent() == false, and
have an empty visibleLayerRect, so the paint functions end up early-
outting. But it is a waste of time to try, and confusing to have
this attempted in the code at all, since the replica layers are never
drawn. They exist only for their transforms.

Also the dimensions used for the replica mask are different from the
surface mask, which is confusing and unnecessary as well. The mask is
applied to the same surface contents as the surface mask would be, so
only the surface's contentBounds need to be painted in the replica
mask.

Covered by existing tests.

* platform/graphics/chromium/cc/CCLayerTreeHost.cpp:
(WebCore::CCLayerTreeHost::paintMasksForRenderSurface):
(WebCore::CCLayerTreeHost::paintLayerContents):
* platform/graphics/chromium/cc/CCLayerTreeHost.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113668 268f45cc-cd09-0410-ab3c-d52691b4dbfc

ShadowRoot should have selection attribute.
https://bugs.webkit.org/show_bug.cgi?id=82429

Reviewed by Hajime Morita.

Source/WebCore:

This patch makes ShadowRoot have selection attribute. Currently the selection returns
the same object as what window.getSelection() returns.

We will extend DOMSelection to have TreeScope. It will adjust Node into TreeScope.
This will be tracked in Bug 82698.

Test: fast/dom/shadow/selection-in-shadow.html

* dom/ShadowRoot.cpp:
(WebCore::ShadowRoot::selection):
(WebCore):
* dom/ShadowRoot.h:
(WebCore):
(ShadowRoot):
* dom/ShadowRoot.idl:

LayoutTests:

Checks ShadowRoot.selection returns the same selection as what window.getSelection() returns.

* fast/dom/shadow/selection-in-shadow.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113667 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL] Gardening in fast/text/whitespace.

The last commit to these baselines had some weird fonts for some
reason, fix the expected sizes.

* platform/efl/fast/text/whitespace/002-expected.png:
* platform/efl/fast/text/whitespace/002-expected.txt:
* platform/efl/fast/text/whitespace/003-expected.png:
* platform/efl/fast/text/whitespace/003-expected.txt:
* platform/efl/fast/text/whitespace/004-expected.png:
* platform/efl/fast/text/whitespace/004-expected.txt:
* platform/efl/fast/text/whitespace/005-expected.png:
* platform/efl/fast/text/whitespace/005-expected.txt:
* platform/efl/fast/text/whitespace/006-expected.png:
* platform/efl/fast/text/whitespace/006-expected.txt:
* platform/efl/fast/text/whitespace/007-expected.png:
* platform/efl/fast/text/whitespace/007-expected.txt:
* platform/efl/fast/text/whitespace/008-expected.png:
* platform/efl/fast/text/whitespace/008-expected.txt:
* platform/efl/fast/text/whitespace/009-expected.png:
* platform/efl/fast/text/whitespace/009-expected.txt:
* platform/efl/fast/text/whitespace/010-expected.png:
* platform/efl/fast/text/whitespace/010-expected.txt:
* platform/efl/fast/text/whitespace/011-expected.png:
* platform/efl/fast/text/whitespace/011-expected.txt:
* platform/efl/fast/text/whitespace/012-expected.png:
* platform/efl/fast/text/whitespace/012-expected.txt:
* platform/efl/fast/text/whitespace/013-expected.png:
* platform/efl/fast/text/whitespace/013-expected.txt:
* platform/efl/fast/text/whitespace/014-expected.png:
* platform/efl/fast/text/whitespace/014-expected.txt:
* platform/efl/fast/text/whitespace/015-expected.png:
* platform/efl/fast/text/whitespace/015-expected.txt:
* platform/efl/fast/text/whitespace/016-expected.png:
* platform/efl/fast/text/whitespace/016-expected.txt:
* platform/efl/fast/text/whitespace/018-expected.png:
* platform/efl/fast/text/whitespace/018-expected.txt:
* platform/efl/fast/text/whitespace/020-expected.png:
* platform/efl/fast/text/whitespace/020-expected.txt:
* platform/efl/fast/text/whitespace/021-expected.png:
* platform/efl/fast/text/whitespace/021-expected.txt:
* platform/efl/fast/text/whitespace/025-expected.png:
* platform/efl/fast/text/whitespace/025-expected.txt:
* platform/efl/fast/text/whitespace/027-expected.png:
* platform/efl/fast/text/whitespace/027-expected.txt:
* platform/efl/fast/text/whitespace/030-expected.png:
* platform/efl/fast/text/whitespace/030-expected.txt:
* platform/efl/fast/text/whitespace/normal-after-nowrap-breaking-expected.png:
* platform/efl/fast/text/whitespace/normal-after-nowrap-breaking-expected.txt:
* platform/efl/fast/text/whitespace/pre-break-word-expected.png:
* platform/efl/fast/text/whitespace/pre-break-word-expected.txt:
* platform/efl/fast/text/whitespace/pre-wrap-last-char-expected.png:
* platform/efl/fast/text/whitespace/pre-wrap-last-char-expected.txt:
* platform/efl/fast/text/whitespace/pre-wrap-line-test-expected.png:
* platform/efl/fast/text/whitespace/pre-wrap-line-test-expected.txt:
* platform/efl/fast/text/whitespace/pre-wrap-overflow-selection-expected.png:
* platform/efl/fast/text/whitespace/pre-wrap-overflow-selection-expected.txt:
* platform/efl/fast/text/whitespace/pre-wrap-spaces-after-newline-expected.png:
* platform/efl/fast/text/whitespace/pre-wrap-spaces-after-newline-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113666 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Replace numeric_limits<LayoutUnit>::min/max with constants
https://bugs.webkit.org/show_bug.cgi?id=83497

Reviewed by Eric Seidel.

Replace all uses of numeric_limits<LayoutUnit>::min/max with
MIN_LAYOUT_UNIT and MAX_LAYOUT_UNIT respectively in preparation for the
switch to subpixel layout.

Also rename zeroLayoutUnit to ZERO_LAYOUT_UNIT to be consistent and
change it to be a define as to avoid global initializer warnings once we
change the definition of LayoutUnit from int to FractionalLayoutUnit.

No new tests, no change in functionality.

* dom/ElementRareData.h:
(WebCore::defaultMinimumSizeForResizing):
* editing/FrameSelection.cpp:
(WebCore::NoXPosForVerticalArrowNavigation):
* platform/graphics/transforms/TransformationMatrix.cpp:
(WebCore::clampEdgeValue):
(WebCore::TransformationMatrix::clampedBoundsOfProjectedQuad):
* rendering/AutoTableLayout.cpp:
(WebCore::AutoTableLayout::computePreferredLogicalWidths):
* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::placeBoxesInBlockDirection):
(WebCore::InlineFlowBox::addBorderOutsetVisualOverflow):
(WebCore::InlineFlowBox::paintFillLayer):
(WebCore::InlineFlowBox::paintBoxDecorations):
(WebCore::InlineFlowBox::paintMask):
(WebCore::InlineFlowBox::computeOverAnnotationAdjustment):
* rendering/LayoutTypes.h:
(WebCore):
* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::MarginInfo::MarginInfo):
(WebCore::RenderBlock::computeInitialRegionRangeForBlock):
(WebCore::RenderBlock::adjustFloatingBlock):
(WebCore::RenderBlock::paintColumnRules):
(WebCore::RenderBlock::removeFloatingObject):
(WebCore::RenderBlock::nextFloatLogicalBottomBelow):
(WebCore::RenderBlock::markLinesDirtyInBlockRange):
(WebCore::RenderBlock::clearFloats):
(WebCore::RenderBlock::addOverhangingFloats):
(WebCore::RenderBlock::getClearDelta):
(WebCore::RenderBlock::computeInlinePreferredLogicalWidths):
(WebCore::getHeightForLineCount):
(WebCore::RenderBlock::borderFitAdjust):
(WebCore::RenderBlock::adjustForUnsplittableChild):
* rendering/RenderBlock.h:
(WebCore::RenderBlock::paginationStrut):
(WebCore::RenderBlock::pageLogicalOffset):
* rendering/RenderBlockLineLayout.cpp:
(WebCore::LineLayoutState::updateRepaintRangeFromBox):
(WebCore::RenderBlock::checkFloatsInCleanLine):
(WebCore::RenderBlock::addOverflowFromInlineChildren):
* rendering/RenderBox.cpp:
(WebCore::RenderBox::shrinkLogicalWidthToAvoidFloats):
(WebCore::RenderBox::positionForPoint):
* rendering/RenderBox.h:
(WebCore::RenderBox::minYVisualOverflow):
(WebCore::RenderBox::minXVisualOverflow):
(RenderBox):
* rendering/RenderBoxModelObject.cpp:
(WebCore::RenderBoxModelObject::offsetLeft):
(WebCore::RenderBoxModelObject::offsetTop):
(WebCore::RenderBoxModelObject::paintFillLayerExtended):
* rendering/RenderDeprecatedFlexibleBox.cpp:
(WebCore::RenderDeprecatedFlexibleBox::layoutHorizontalBox):
(WebCore::RenderDeprecatedFlexibleBox::layoutVerticalBox):
(WebCore::RenderDeprecatedFlexibleBox::allowedChildFlex):
* rendering/RenderFieldset.cpp:
(WebCore::RenderFieldset::paintBoxDecorations):
(WebCore::RenderFieldset::paintMask):
* rendering/RenderFlexibleBox.cpp:
(WebCore::RenderFlexibleBox::repositionLogicalHeightDependentFlexItems):
(WebCore::RenderFlexibleBox::lineBreakLength):
(WebCore::RenderFlexibleBox::alignChildren):
(WebCore::RenderFlexibleBox::applyStretchAlignmentToChild):
* rendering/RenderFlowThread.cpp:
(WebCore::RenderFlowThread::layout):
(WebCore::RenderFlowThread::computeLogicalWidth):
* rendering/RenderInline.cpp:
(WebCore::RenderInline::linesVisualOverflowBoundingBox):
* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::resize):
* rendering/RenderListMarker.cpp:
(WebCore::RenderListMarker::updateMargins):
* rendering/RenderMarquee.cpp:
(WebCore::RenderMarquee::computePosition):
* rendering/RenderObject.cpp:
(WebCore::RenderObject::repaintAfterLayoutIfNeeded):
* rendering/RenderReplaced.cpp:
(WebCore::RenderReplaced::computeReplacedLogicalWidth):
(WebCore::RenderReplaced::computePreferredLogicalWidths):
* rendering/RenderTable.cpp:
(WebCore::RenderTable::convertStyleLogicalWidthToComputedWidth):
(WebCore::RenderTable::layout):
* rendering/RenderTable.h:
(WebCore::RenderTable::bordersPaddingAndSpacingInRowDirection):
* rendering/RenderText.cpp:
(WebCore::RenderText::linesVisualOverflowBoundingBox):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113665 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[EFL] Gardening in transforms/2d.

Mostly account for the changes caused by the jhbuild and font
commits.

* platform/efl/transforms/2d/compound-transforms-vs-containers-expected.png:
* platform/efl/transforms/2d/hindi-rotated-expected.png:
* platform/efl/transforms/2d/transform-borderbox-expected.png:
* platform/efl/transforms/2d/transform-fixed-container-expected.png:
* platform/efl/transforms/2d/transform-origin-borderbox-expected.png:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113664 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Clean up LayoutUnit usage in DOMNodeHighlighter
https://bugs.webkit.org/show_bug.cgi?id=83507

Reviewed by Eric Seidel.

Ensuring DOMNodeHighlighter uses pixel snapping that matches that done in the RenderTree. Changes include:
- Reverting drawSubstring to use an IntPoint with pixel snapped values instead of a LayoutPoint.
- Reverting borderBox to an IntRect since it's in absolute coordinates. See https://trac.webkit.org/wiki/LayoutUnit.
- Frames are still laid out with integers. Cleaning up mainFrameOffset to reflect that.

No new tests. No change in behavior.

* inspector/DOMNodeHighlighter.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113663 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add missing IndexedDB files to CMake build
https://bugs.webkit.org/show_bug.cgi?id=83495

Patch by Jeff Rogers <jrogers@rim.com> on 2012-04-09
Reviewed by Rob Buis.

* CMakeLists.txt:
* UseJSC.cmake:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113662 268f45cc-cd09-0410-ab3c-d52691b4dbfc

https://bugs.webkit.org/show_bug.cgi?id=83411
 All Mac bots occasionally fail tons of media/ sputnik/ and svg/ tests

This is a speculative fix.
* platform/mac/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113661 268f45cc-cd09-0410-ab3c-d52691b4dbfc

If a callback constructor returns a C++ null, throw a type error.
https://bugs.webkit.org/show_bug.cgi?id=83537

Rubber Stamped by Geoff Garen.

* API/JSCallbackConstructor.cpp:
(JSC::constructJSCallback):
    - If a callback constructor returns a C++ null, throw a type error.
* API/tests/testapi.c:
(Base_returnHardNull):
* API/tests/testapi.js:
    - Add a test case for callback constructors that return a C++ null.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@113660 268f45cc-cd09-0410-ab3c-d52691b4dbfc