[WebKit-https.git] / Source / WebCore / ChangeLog

2014-02-14  Benjamin Poulain  <benjamin@webkit.org>

        Make code generation of the unoptimized pseudo classes separate
        https://bugs.webkit.org/show_bug.cgi?id=128704

        Reviewed by Andreas Kling.

        Mapping the unoptimized selectors when generating the code was useful for debugging
        but those cases are so common that looping over unoptimizedPseudoCheckers was more
        costly than anticipated.

        This patch moves the unoptimized pseudo selector in their own array to simplify code generation.

        * cssjit/SelectorCompiler.cpp:
        (WebCore::SelectorCompiler::addPseudoType):
        (WebCore::SelectorCompiler::SelectorCodeGenerator::SelectorCodeGenerator):
        (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementMatching):

2014-02-14  Myles C. Maxfield  <mmaxfield@apple.com>

        Implement text-decoration-skip: auto
        https://bugs.webkit.org/show_bug.cgi?id=128786

        Reviewed by Dean Jackson.

        http://lists.w3.org/Archives/Public/www-style/2014Feb/0485.html
        We are updating the initial value of text-decoration-skip to a new value, "auto".
        On Mac + iOS, this will have the same behavior as "skip".

        Test: fast/css3-text/css3-text-decoration/text-decoration-skip/text-decoration-skip-default.html

        * css/CSSComputedStyleDeclaration.cpp:
        (WebCore::renderTextDecorationSkipFlagsToCSSValue):
        * css/CSSParser.cpp:
        (WebCore::CSSParser::parseTextDecorationSkip):
        * css/DeprecatedStyleBuilder.cpp:
        (WebCore::valueToDecorationSkip):
        * rendering/InlineTextBox.cpp:
        (WebCore::InlineTextBox::paintDecoration):
        * rendering/style/RenderStyle.h:
        * rendering/style/RenderStyleConstants.h:

2014-02-13  Simon Fraser  <simon.fraser@apple.com>

        Give ScrollingTree(State)Nodes a reference to another layer, which is used for moving overflow:scroll contents around
        https://bugs.webkit.org/show_bug.cgi?id=128790

        Reviewed by Beth Dakin.

        Have scrolling tree nodes and state nodes track another layer, the
        "scrolled contents layer", for accelerated overflow:scroll.
        
        When making ScrollingTreeScrollingNodes for overflow:scroll, the node's
        layer will point to the composited element's primary layer, and its
        scrolledContentsLayer to the layer that gets moved around by scrolling.
        
        Do some other cleanup on AsyncScrollingCoordinator, removing
        functions that just called through to nodes.

        * WebCore.exp.in:
        * page/scrolling/AsyncScrollingCoordinator.cpp:
        (WebCore::AsyncScrollingCoordinator::frameViewLayoutUpdated):
        (WebCore::AsyncScrollingCoordinator::frameViewNonFastScrollableRegionChanged):
        (WebCore::AsyncScrollingCoordinator::frameViewRootLayerDidChange):
        (WebCore::AsyncScrollingCoordinator::updateScrollingNode):
        (WebCore::AsyncScrollingCoordinator::updateViewportConstrainedNode):
        (WebCore::AsyncScrollingCoordinator::recomputeWheelEventHandlerCountForFrameView):
        * page/scrolling/AsyncScrollingCoordinator.h:
        * page/scrolling/ScrollingCoordinator.h:
        (WebCore::ScrollingCoordinator::updateScrollingNode):
        * page/scrolling/ScrollingStateScrollingNode.cpp:
        (WebCore::ScrollingStateScrollingNode::ScrollingStateScrollingNode):
        (WebCore::ScrollingStateScrollingNode::setScrolledContentsLayer):
        (WebCore::ScrollingStateScrollingNode::setCounterScrollingLayer):
        (WebCore::ScrollingStateScrollingNode::setHeaderLayer):
        (WebCore::ScrollingStateScrollingNode::setFooterLayer):
        * page/scrolling/ScrollingStateScrollingNode.h:
        * page/scrolling/ios/ScrollingTreeScrollingNodeIOS.h:
        * page/scrolling/ios/ScrollingTreeScrollingNodeIOS.mm:
        (WebCore::ScrollingTreeScrollingNodeIOS::updateBeforeChildren):
        * page/scrolling/mac/ScrollingTreeScrollingNodeMac.h:
        * page/scrolling/mac/ScrollingTreeScrollingNodeMac.mm:
        (WebCore::ScrollingTreeScrollingNodeMac::updateBeforeChildren):
        * rendering/RenderLayerCompositor.cpp:
        (WebCore::RenderLayerCompositor::fixedRootBackgroundLayerChanged):

2014-02-14  Chris Fleizach  <cfleizach@apple.com>

        AX: WebKit needs heuristics to differentiate lists used for layout from semantic data lists, similar to the heuristics for layout tables versus data tables.
        https://bugs.webkit.org/show_bug.cgi?id=122320

        Reviewed by Mario Sanchez Prada.

        Many authors use lists for layout, rather than presenting data. Exposing these kinds of lists to accessibility users
        makes the web harder to process and listen, while degrading the importance of real lists.

        This introduces heuristics to filter out layout lists with the following:
           1. If it's a named list, like ol or aria=list, then it's a list.
             1a. Unless the list has no children, then it's not a list.
           2. If it displays visible list markers, it's a list.
           3. If it does not display list markers and has only one child, it's not a list.
           4. If it does not have any listitem children, it's not a list.
           5. Otherwise it's a list (for now).

        Test: accessibility/list-detection.html

        * accessibility/AccessibilityList.cpp:
        (WebCore::AccessibilityList::determineAccessibilityRole):
        (WebCore::AccessibilityList::roleValue):
        * accessibility/AccessibilityList.h:

2014-02-14  Brendan Long  <b.long@cablelabs.com>

        Use AtomicString arguments in TrackPrivateBaseClient callbacks
        https://bugs.webkit.org/show_bug.cgi?id=128765

        Reviewed by Eric Carlson.

        No new tests because this doesn't change functionality.

        * html/track/AudioTrack.cpp:
        (WebCore::AudioTrack::idChanged):
        (WebCore::AudioTrack::labelChanged):
        (WebCore::AudioTrack::languageChanged):
        * html/track/AudioTrack.h:
        * html/track/InbandTextTrack.cpp:
        (WebCore::InbandTextTrack::InbandTextTrack):
        (WebCore::InbandTextTrack::idChanged):
        (WebCore::InbandTextTrack::labelChanged):
        (WebCore::InbandTextTrack::languageChanged):
        * html/track/InbandTextTrack.h:
        * html/track/VideoTrack.cpp:
        (WebCore::VideoTrack::idChanged):
        (WebCore::VideoTrack::labelChanged):
        (WebCore::VideoTrack::languageChanged):
        * html/track/VideoTrack.h:
        * platform/graphics/TrackPrivateBase.h:

2014-02-14  Renata Hodovan  <rhodovan.u-szeged@partner.samsung.com>

        ASSERTION FAILED: comparePositions(newEnd, newStart) >= 0 in WebCore::ApplyStyleCommand::updateStartEnd
        https://bugs.webkit.org/show_bug.cgi?id=121791

        Reviewed by Darin Adler.

        If WebCore::ApplyStyleCommand::applyBlockStyle() creates a TextIterator for a range
        that has an element with ReplacedElement rendering object, then a ',' is emitted in the
        constructor of TextIterator. Due to this comma the end of the run range can be at the
        wrong position, what makes the assertion fire. This situation can be handled the same
        way in TextIterator::rangeFromLocationAndLength() as we do in case of the emitted '\n's.

        Test: editing/execCommand/remove-formatting-from-iframe-in-button.html

        * editing/TextIterator.cpp:
        (WebCore::TextIterator::rangeFromLocationAndLength):

2014-02-14  Andrei Bucur  <abucur@adobe.com>

        [CSS Regions] visibility: hidden on a region should hide its content
        https://bugs.webkit.org/show_bug.cgi?id=128814

        Reviewed by Mihnea Ovidenie.

        The patch properly checks that the content of a region layer should be painted before actually
        calling the paint function (for example, when visibility: hidden is specified on a region, its
        content should be hidden).

        Test: fast/regions/visibility-hidden.html

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::paintLayerContents):

2014-02-14  László Langó  <llango.u-szeged@partner.samsung.com>

        Follow-up fix after r164036.

        Reviewed by Csaba Osztrogonác.

        * dom/Document.cpp:
        (WebCore::Document::create):

2014-02-13  Byungseon Shin  <sun.shin@lge.com>

        [MSE] Move PublicURLManager shutdown logic so ActiveDOMObjects associated with public URLs won't leak.
        https://bugs.webkit.org/show_bug.cgi?id=128532

        Reviewed by Jer Noble.

        This fixes a leak of DOM objects by breaking the circular reference 
        between Document, PublicURLManager, and MediaSource. 
        Instead of clearing PublicURLManager at destruction-time, 
        which is delayed indefinitely because of the circular reference, 
        clear the PublicURLManager during ActiveDOMObject::stop().

        Frome Blink r151890 by <acolwell@chromium.org>
        <https://src.chromium.org/viewvc/blink?view=rev&revision=151890>

        * dom/ScriptExecutionContext.cpp:
        (WebCore::ScriptExecutionContext::~ScriptExecutionContext):
        (WebCore::ScriptExecutionContext::publicURLManager):
        * html/DOMURL.h:
        * html/PublicURLManager.cpp:
        (WebCore::PublicURLManager::create):
        (WebCore::PublicURLManager::PublicURLManager):
        (WebCore::PublicURLManager::registerURL):
        (WebCore::PublicURLManager::stop):
        * html/PublicURLManager.h:

2014-02-13  Myles C. Maxfield  <mmaxfield@apple.com>

        Remove position:sticky runtime flag
        https://bugs.webkit.org/show_bug.cgi?id=128774

        Reviewed by Simon Fraser.

        Rollout of r128663

        No new tests are necessary because there is no behavior change.

        * css/CSSParser.cpp:
        (WebCore::CSSParserContext::CSSParserContext):
        (WebCore::operator==):
        (WebCore::isValidKeywordPropertyAndValue):
        * css/CSSParserMode.h:
        * dom/Document.cpp:
        * dom/Document.h:
        * page/Settings.cpp:
        (WebCore::Settings::Settings):
        * page/Settings.h:

2014-02-04  Gustavo Noronha Silva  <gns@gnome.org>

        [GTK][CMake] Generate GObject DOM bindings .symbols files
        https://bugs.webkit.org/show_bug.cgi?id=126210

        Reviewed by Martin Robinson.

        No new tests. Build change only.

        * CMakeLists.txt: list Quota module files and IDLs even if the feature
        is disabled, for GTK, since we rely on that for our DOM bindings.
        * PlatformGTK.cmake: add a new target to check for DOM symbols API by
        running the new python script.
        * bindings/gobject/GNUmakefile.am: call the new python script instead
        of using a custom rule.

2014-02-13  Myles C. Maxfield  <mmaxfield@apple.com>

        text-decoration-skip: ink skips randomly when using SVG fonts
        https://bugs.webkit.org/show_bug.cgi?id=128709

        Reviewed by Simon Fraser.

        This patch simply disables skipping underlines being used with SVG fonts.
        It's a stopgap until we can support skipping properly with SVG fonts.

        Test: svg/custom/svg-fonts-skip-ink.html

        * platform/graphics/cocoa/FontPlatformDataCocoa.mm:
        (WebCore::FontPlatformData::ctFont):
        * platform/graphics/mac/FontMac.mm:
        (WebCore::Font::dashesForIntersectionsWithRect):

2014-02-13  Daniel Bates  <dabates@apple.com>

        Fix the ARM build when HAVE_ARM_NEON_INTRINSICS is enabled

        Add !HAVE(ARM_NEON_INTRINSICS)-guard around file-level inline function arithmeticSoftware to
        avoid an unused function compiler warning when building with ARM Neon intrinsics.

        * platform/graphics/filters/FEComposite.cpp:

2014-02-13  Brady Eidson  <beidson@apple.com>

        IDB: Three tests crash the DatabaseProcess under ~KeyedDecoder
        https://bugs.webkit.org/show_bug.cgi?id=128763

        Reviewed by Anders Carlsson.

        No new tests (Covered by three existing tests)

        * platform/KeyedCoding.h:
        (WebCore::KeyedEncoder::encodeObjects): If the array is empty we still have to begin() and end() it.

2014-02-13  Yuki Sekiguchi  <yuki.sekiguchi@access-company.com>

        [css3-text] Support -webkit-text-decoration-skip: objects
        https://bugs.webkit.org/show_bug.cgi?id=128723

        Reviewed by Dean Jackson.

        Implemented 'objects' value of text-decoration-skip.
        http://www.w3.org/TR/2013/CR-css-text-decor-3-20130801/#text-decoration-skip

        This is initial value and same behavior without
        ENABLE_CSS3_TEXT_DECORATION_SKIP_INK flag. Threfore, this patch only
        added parser code and constant for 'objects'.

        Changed the initial value of text-decoration-skip.

        The current implementation parses 'none', but the rendering code is
        not match the spec, so this patch added FIXME to the rendering code.

        * css/CSSComputedStyleDeclaration.cpp:
        (WebCore::renderTextDecorationSkipFlagsToCSSValue):
        * css/CSSParser.cpp:
        (WebCore::CSSParser::parseTextDecorationSkip):
        * css/CSSValueKeywords.in:
        * css/DeprecatedStyleBuilder.cpp:
        (WebCore::valueToDecorationSkip):
        * rendering/InlineTextBox.cpp:
        (WebCore::InlineTextBox::paintDecoration):
        * rendering/style/RenderStyle.h:
        * rendering/style/RenderStyleConstants.h:

2014-02-13  Daniel Bates  <dabates@apple.com>

        Fix ARM NEON build following <http://trac.webkit.org/changeset/164058>
        (https://bugs.webkit.org/show_bug.cgi?id=128767)

        I inadvertently didn't include header NEONHelpers.h, which defines WebCore::loadRGBA8AsFloat().

        * platform/graphics/cpu/arm/filters/FECompositeArithmeticNEON.h:

2014-02-13  Daniel Bates  <dabates@apple.com>

        Write FEComposite::computeArithmeticPixelsNeon() in terms of loadRGBA8AsFloat()
        https://bugs.webkit.org/show_bug.cgi?id=128767

        Reviewed by Dean Jackson.

        This patch was inspired by Gabor Rapcsanyi's work in <https://bugs.webkit.org/show_bug.cgi?id=90669>.

        Currently we duplicate code in FEComposite::computeArithmeticPixelsNeon() to read pixel data.
        Instead we should use the convenience function WebCore::loadRGBA8AsFloat. This also resolves
        an uninitialized variable compiler warning in the existing code.

        * platform/graphics/cpu/arm/filters/FECompositeArithmeticNEON.h:
        (WebCore::FEComposite::computeArithmeticPixelsNeon):

2014-02-13  Myles C. Maxfield  <mmaxfield@apple.com>

        During a copy, position:fixed gets converted to position:absolute even if only part of the document is selected
        https://bugs.webkit.org/show_bug.cgi?id=128688

        Reviewed by Ryosuke Niwa.

        Adds a field, m_needsPositionStyleConversion, to StyledMarkupAccumulator.

        Also, renames the ConvertPositionStyleOnCopy setting to ShouldConvertPositionStyleOnCopy

        Test: editing/pasteboard/copy-paste-doesnt-convert-sticky-and-fixed-during-partial-copy.html

        * editing/markup.cpp:
        (WebCore::StyledMarkupAccumulator::StyledMarkupAccumulator):
        (WebCore::StyledMarkupAccumulator::appendElement):
        (WebCore::createMarkupInternal):
        * page/Settings.in:
        * testing/InternalSettings.cpp:
        (WebCore::InternalSettings::Backup::Backup):
        (WebCore::InternalSettings::Backup::restoreTo):
        (WebCore::InternalSettings::setShouldConvertPositionStyleOnCopy):
        * testing/InternalSettings.h:
        * testing/InternalSettings.idl:

2014-02-12  Timothy Hatcher  <timothy@apple.com>

        Enable inspection of recently used IndexedDB databases.

        https://bugs.webkit.org/show_bug.cgi?id=128687

        Reviewed by Brady Eidson & Joseph Pecoraro.

        * Modules/indexeddb/IDBFactory.cpp:
        (WebCore::IDBFactory::getDatabaseNames):
        * Modules/indexeddb/IDBFactoryBackendInterface.h:
        * Modules/indexeddb/leveldb/IDBFactoryBackendLevelDB.cpp:
        (WebCore::IDBFactoryBackendLevelDB::getDatabaseNames):
        * Modules/indexeddb/leveldb/IDBFactoryBackendLevelDB.h:
        Fix up the arguments passed to getDatabaseNames.

        * WebCore.xcodeproj/project.pbxproj:
        Make DOMStringList.h private so WebKit2 can use it.

        * inspector/InspectorIndexedDBAgent.cpp:
        Modernize the loops and remove some PLATFORM(COCOA) ifdefs to
        compile the code. 

        * inspector/protocol/IndexedDB.json: Use number for version to
        allow larger values instead of clamping to int.

2014-02-13  Chang Shu  <cshu@webkit.org>

        Copying (createMarkup) wrapping text results in space between wrapped lines stripped.
        https://bugs.webkit.org/show_bug.cgi?id=63233

        The problem is StyledMarkupAccumulator uses renderedText and the space at
        the end of the text has been stripped when the tag after the text was wrapped.

        Reviewed by Ryosuke Niwa.

        Test: editing/pasteboard/copy-text-with-wrapped-tag.html

        * editing/TextIterator.cpp:
        (WebCore::TextIterator::TextIterator):
        (WebCore::TextIterator::handleTextBox):
        * editing/TextIterator.h:
        Check the special case when the iterator runs over a range that is followed by a non-text box.
        In this case, the possible last space has been collapsed and needs to be restored.
        * editing/markup.cpp:
        (WebCore::StyledMarkupAccumulator::renderedText):
        Check if the range is followed by more nodes and pass this information to the constructed TextIterator.

2014-02-13  Myles C. Maxfield  <mmaxfield@apple.com>

        Gaps for text-decoration-skip: ink are calculated assuming the glyph orientation is the same as the run orientation
        https://bugs.webkit.org/show_bug.cgi?id=128718

        Reviewed by Darin Adler.

        This is a stop-gap patch to make underlines in CJK text not look terrible. A long-term fix is to refactor
        showGlyphsWithAdvances to move rotation logic into a common place that the text-decoration-skip: ink code
        can use.

        Test: fast/css3-text/css3-text-decoration/text-decoration-skip/text-decoration-skip-orientation-upright.html

        * rendering/InlineTextBox.cpp:
        (WebCore::InlineTextBox::paintDecoration):

2014-02-13  Andreas Kling  <akling@apple.com>

        InsertIntoTextNodeCommand should get Settings from the Frame.
        <https://webkit.org/b/128678>

        EditCommands have a frame() accessor that returns a reference,
        so there's no need to go through Document.

        Reviewed by Anders Carlsson.

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

2014-02-13  László Langó  <llango.u-szeged@partner.samsung.com>

        Document should be constructable
        https://bugs.webkit.org/show_bug.cgi?id=115643

        Reviewed by Darin Adler.

        http://www.w3.org/TR/2014/WD-dom-20140204/#interface-document
        Make Document constructable so that one can do "new Document"
        instead of "document.implementation.createHTMLDocument('')".

        Test: fast/dom/Document/document-constructor.html

        * dom/Document.cpp:
        (WebCore::Document::create):
        (WebCore::Document::origin):
        * dom/Document.h:
        * dom/Document.idl:

2014-02-13  Javier Fernandez  <jfernandez@igalia.com>

        [CSS Grid Layout] Rename named areas property
        https://bugs.webkit.org/show_bug.cgi?id=127990

        Reviewed by Sergio Villar Senin.

        From Blink r165891 by <rego@igalia.com>

        The property 'grid-template' has been renamed to 'grid-template-areas'
        in the last two versions of the spec.

        * CMakeLists.txt:
        * GNUmakefile.list.am:
        * WebCore.vcxproj/WebCore.vcxproj:
        * WebCore.vcxproj/WebCore.vcxproj.filters:
        * WebCore.xcodeproj/project.pbxproj:
        * css/CSSComputedStyleDeclaration.cpp:
        (WebCore::ComputedStyleExtractor::propertyValue):
        * css/CSSGridTemplateAreasValue.cpp: Renamed from Source/WebCore/css/CSSGridTemplateValue.cpp.
        (WebCore::CSSGridTemplateAreasValue::CSSGridTemplateAreasValue):
        (WebCore::stringForPosition):
        (WebCore::CSSGridTemplateAreasValue::customCSSText):
        * css/CSSGridTemplateAreasValue.h: Renamed from Source/WebCore/css/CSSGridTemplateValue.h.
        (WebCore::CSSGridTemplateAreasValue::create):
        (WebCore::CSSGridTemplateAreasValue::~CSSGridTemplateAreasValue):
        (WebCore::CSSGridTemplateAreasValue::gridAreaMap):
        (WebCore::CSSGridTemplateAreasValue::rowCount):
        (WebCore::CSSGridTemplateAreasValue::columnCount):
        * css/CSSParser.cpp:
        (WebCore::CSSParser::parseValue):
        (WebCore::CSSParser::parseGridTemplateAreas):
        * css/CSSParser.h:
        * css/CSSPropertyNames.in:
        * css/CSSValue.cpp:
        (WebCore::CSSValue::equals):
        (WebCore::CSSValue::cssText):
        (WebCore::CSSValue::destroy):
        * css/CSSValue.h:
        (WebCore::CSSValue::isGridTemplateAreasValue):
        * css/StyleResolver.cpp:
        (WebCore::StyleResolver::applyProperty):

2014-02-13  Adrian Bunk  <bunk@stusta.de>

        Remove the last remnants of Maemo support
        https://bugs.webkit.org/show_bug.cgi?id=85238

        Reviewed by Ryosuke Niwa.

        * plugins/npapi.h:

2014-02-13  ChangSeok Oh  <changseok.oh@collabora.com>

        Support ANGLE_instanced_arrays for GLES backend.
        https://bugs.webkit.org/show_bug.cgi?id=128579

        Reviewed by Dean Jackson.

        Add some APIs to Extensions3D to support ANGLE_instanced_arrays for GLES backend.
        At the same time, drawArraysInstanced, drawElementsInstanced and vertexAttribDivisor
        are moved from GC3DOpenGL.cpp to GC3DOpenGLCommon.cpp since they could be shareable
        through Extensions3D.

        Covered by fast/canvas/webgl/angle-instanced-arrays.html

        * platform/graphics/Extensions3D.h:
        * platform/graphics/opengl/Extensions3DOpenGL.cpp:
        (WebCore::Extensions3DOpenGL::drawArraysInstanced):
        (WebCore::Extensions3DOpenGL::drawElementsInstanced):
        (WebCore::Extensions3DOpenGL::vertexAttribDivisor):
        * platform/graphics/opengl/Extensions3DOpenGL.h:
        * platform/graphics/opengl/Extensions3DOpenGLES.cpp:
        (WebCore::Extensions3DOpenGLES::Extensions3DOpenGLES):
        (WebCore::Extensions3DOpenGLES::drawArraysInstanced):
        (WebCore::Extensions3DOpenGLES::drawElementsInstanced):
        (WebCore::Extensions3DOpenGLES::vertexAttribDivisor):
        (WebCore::Extensions3DOpenGLES::supportsExtension):
        * platform/graphics/opengl/Extensions3DOpenGLES.h:
        * platform/graphics/opengl/GraphicsContext3DOpenGL.cpp:
        * platform/graphics/opengl/GraphicsContext3DOpenGLCommon.cpp:
        (WebCore::GraphicsContext3D::drawArraysInstanced):
        (WebCore::GraphicsContext3D::drawElementsInstanced):
        (WebCore::GraphicsContext3D::vertexAttribDivisor):
        * platform/graphics/opengl/GraphicsContext3DOpenGLES.cpp:

2014-02-13  Xabier Rodriguez Calvar  <calvaris@igalia.com>

        [GTK] MEDIA_CONTROLS_SCRIPT support
        https://bugs.webkit.org/show_bug.cgi?id=123097

        Reviewed by Jer Noble.

        WebKitGTK+ multimedia controls are now managed from Javascript
        code. Apple controls are kept as common code and GTK+ ones are
        subclassed for the specific behavior.

        Both CMake and Autotools build support is provided.

        * CMakeLists.txt: Added support to build the media controls
        script and their associated files.
        * GNUmakefile.am: Added support to generate the C++ code from the
        Javascript.
        * GNUmakefile.list.am: Added the media controls script associated
        files.
        * Modules/mediacontrols/MediaControlsHost.cpp:
        * Modules/mediacontrols/MediaControlsHost.h:
        * Modules/mediacontrols/MediaControlsHost.idl:
        (WebCore::MediaControlsHost::supportsFullscreen): Added attribute
        to know if the element supports fullscreen.
        * Modules/mediacontrols/mediaControlsApple.js:
        (Controller.prototype.handleWrapperMouseMove):
        (Controller.prototype.handleWrapperMouseOut):
        (Controller.prototype.updatePlaying): Use clear and
        resetHideControlsTimer.
        (Controller.prototype.clearHideControlsTimer): Added.
        (Controller.prototype.resetHideControlsTimer): Added.
        * Modules/mediacontrols/mediaControlsGtk.js: Added.
        (createControls): Calls ControllerGtk.
        (ControllerGtk): Calls the superclass.
        (contains): Defines a function to know if an object is contained
        in an array.
        (ControllerGtk.prototype.inheritFrom): Copies the method of the
        superclass that are not reimplemented in the subclass.
        (ControllerGtk.prototype.createControls): Calls the superclass and
        create the remaining needed elements.
        (ControllerGtk.prototype.configureInlineControls): Configures the
        controls.
        (ControllerGtk.prototype.setStatusHidden): Redefined empty.
        (ControllerGtk.prototype.updateTime): Writes the duration and
        current position. As it was so far, current time also includes
        duration. If current time is bigger than 0 we show that.
        (ControllerGtk.prototype.showCurrentTime): Shows current time and
        hides duration label (that is included in current time).
        (ControllerGtk.prototype.handlePlay): Calls the superclass and
        force showing the current time.
        (ControllerGtk.prototype.handleTimeUpdate): Always update the
        time.
        (ControllerGtk.prototype.handleMuteButtonMouseOver): Shows the
        volume slider.
        (ControllerGtk.prototype.handleVolumeBoxMouseOut): Hides the
        volume slider.
        (ControllerGtk.prototype.addControls): Adds the enclosure instead
        of the panel directly. Panel is, of course, part of the enclosure.
        (ControllerGtk.prototype.updateReadyState): Shows the fullscreen
        button only if fullscreen is supported. Sets the volume in up or
        down mode depending on its position in the document. Updates the
        volume.
        (ControllerGtk.prototype.setControlsType): Creates the controls it
        they were not created before. Unlike Apple ones, WebKitGTK+
        fullscreen and inline controls are the same.
        (ControllerGtk.prototype.updatePlaying): Calls the superclass and
        shows the controls if not playing.
        (ControllerGtk.prototype.handleCaptionButtonClicked): Redefined
        empty. To be coherent with volume, that also shows a popup, this
        is handled with mouseover instead of click.
        (ControllerGtk.prototype.buildCaptionMenu): Calls the superclass
        to build the menu, sets some listeners, centers the popup with the
        captions button, keeps the current height and sets the style to 0,
        which is needed to animate it.
        (ControllerGtk.prototype.destroyCaptionMenu): Hides the caption menu.
        (ControllerGtk.prototype.showCaptionMenu): Resets the height to
        its original. We don't animate it with CSS because we would need
        to specify a height in the style and we don't know it in advance.
        (ControllerGtk.prototype.hideCaptionMenu): Sets height to 0.
        (ControllerGtk.prototype.captionMenuTransitionEnd): When the
        captions menu transtition ends, it is destroyed.
        (ControllerGtk.prototype.handleCaptionButtonMouseOver): Creates
        the caption menu and shows it.
        (ControllerGtk.prototype.handleCaptionButtonMouseOut): Hides the
        captions menu.
        (ControllerGtk.prototype.handleCaptionMouseOut): Hides the
        captions menu.
        * PlatformGTK.cmake: Added WebCore to the target link libraries
        and initialized the variables needed at CMakelists.txt
        * css/mediaControlsGtk.css:
        (audio::-webkit-media-controls-panel)
        (video::-webkit-media-controls-panel): Added transtion based on opacity.
        (video::-webkit-media-controls-panel): Set video opacity to 0.
        (video::-webkit-media-controls-panel.paused): Set video opacity to 1.
        (audio::-webkit-media-controls-panel div.mute-box): Set the same
        style as the mute button to preserve the layout.
        (audio::-webkit-media-controls-panel div.mute-box.hidden): Sets
        the display to none.
        (audio::-webkit-media-controls-mute-button)
        (video::-webkit-media-controls-mute-button): Removed the outline
        and the margin that is managed by the mute box now.
        (audio::-webkit-media-controls-play-button)
        (video::-webkit-media-controls-play-button): Removed the outline.
        (audio::-webkit-media-controls-time-remaining-display)
        (video::-webkit-media-controls-time-remaining-display): Removed
        the display.
        (audio::-webkit-media-controls-current-time-display)
        (video::-webkit-media-controls-current-time-display): Added the
        display block.
        (video::-webkit-media-controls-time-remaining-display): Set
        display none.
        (video::-webkit-media-controls-time-remaining-display.show): Set
        display block.
        (video::-webkit-media-controls-time-remaining-display.hidden): Set
        display none.
        (audio::-webkit-media-controls-timeline)
        (video::-webkit-media-controls-timeline): Removed outline.
        (audio::-webkit-media-controls-volume-slider-container)
        (video::-webkit-media-controls-volume-slider-container): Set
        overflow hidden and set a transition by height.
        (video::-webkit-media-controls-volume-slider-container.hidden):
        Set height 0.
        (video::-webkit-media-controls-volume-slider-container.down): Set
        bottom to be below the panel. Changed the border radius and
        transition accordingly.
        (video::-webkit-media-controls-panel .hidden.down): Sets default
        bottom as 0.
        (audio::-webkit-media-controls-volume-slider)
        (video::-webkit-media-controls-volume-slider): Removed the outline.
        (audio::-webkit-media-controls-toggle-closed-captions-button)
        (video::-webkit-media-controls-toggle-closed-captions-button):
        Removed the background that is painted from C++ and removed the
        outline.
        (video::-webkit-media-controls-closed-captions-container):
        (video::-webkit-media-controls-closed-captions-container h3):
        (video::-webkit-media-controls-closed-captions-container ul):
        (video::-webkit-media-controls-closed-captions-container li):
        (video::-webkit-media-controls-closed-captions-container li.selected): Changed
        the style to make it more coherent with the rest of the controls.
        (audio::-webkit-media-controls-fullscreen-button)
        (video::-webkit-media-controls-fullscreen-button): Removed the outline.
        (audio::-webkit-media-controls-panel button.hidden): Sets the
        display to none.
        * html/HTMLMediaElement.cpp: UserAgentScripts.h is not needed here
        and build cmake build would need for changes to get this compiled.
        * platform/gtk/RenderThemeGtk.cpp:
        (WebCore::nodeHasPseudo): Added. Checks if a node has a certain
        pseudo.
        (WebCore::nodeHasClass): Added. Checks if a node has a certain
        class.
        (WebCore::supportsFocus): Removed some element types as outline is
        now handled in CSS.
        (WebCore::RenderThemeGtk::paintMediaPlayButton): Checks if the
        play button has the class paused to show the play icon instead of
        the pause one.
        (WebCore::RenderThemeGtk::paintMediaToggleClosedCaptionsButton):
        Added. Paints the captions icon.
        (WebCore::RenderThemeGtk::mediaControlsScript): Added. Loads the
        bundled scripts.
        * platform/gtk/RenderThemeGtk.h: Added the mediaControlsScript
        method and declared the redefinition of the method to play the
        captions icon.

2014-02-11  Alexey Proskuryakov  <ap@apple.com>

        Don't crash when SerializedScriptValue deserialization fails
        https://bugs.webkit.org/show_bug.cgi?id=128657

        Reviewed by Oliver Hunt.

        Test: crypto/subtle/postMessage-worker.html

        * bindings/js/JSMessageEventCustom.cpp: (WebCore::JSMessageEvent::data): Added a FIXME.

        * bindings/js/SerializedScriptValue.cpp:
        (WebCore::CloneBase::fail): Don't assert on failure.
        (WebCore::SerializedScriptValue::deserialize): Never return a null JSValue, these
        are not allowed.

2014-02-12  Bem Jones-Bey  <bjonesbe@adobe.com>

        [CSS Shapes] Rename shapeSize and others to make ShapeInfo and friends easier to understand
        https://bugs.webkit.org/show_bug.cgi?id=128685

        Reviewed by Alexandru Chiculita.

        The ShapeInfo hierarchy has grown organically as the spec has changed,
        and the naming has become very stale and confusing. The spec is now in
        Last Call, so it seems like a good time to take a stab towards better
        naming that matches the spec terminology and is also more consistent
        in general.

        No new tests, no behavior change.

        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::imageChanged): Use new names.
        (WebCore::RenderBlock::updateShapeInsideInfoAfterStyleChange): Ditto.
        (WebCore::shapeInfoRequiresRelayout): Ditto.
        (WebCore::RenderBlock::computeShapeSize): Ditto.
        (WebCore::RenderBlock::updateShapesAfterBlockLayout): Ditto.
        * rendering/RenderBlock.h:
        (WebCore::RenderBlock::logicalSizeForChild): This method makes it much
            cleaner to set the reference box size.
        * rendering/RenderBlockFlow.cpp:
        (WebCore::RenderBlockFlow::positionNewFloats): Use new names.
        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::updateShapeOutsideInfoAfterStyleChange): Ditto.
        (WebCore::RenderBox::imageChanged): Ditto.
        * rendering/shapes/ShapeInfo.cpp:
        (WebCore::getShapeImageMarginRect): Rename variables.
        (WebCore::ShapeInfo<RenderType>::computedShape): More renames.
        * rendering/shapes/ShapeInfo.h:
        (WebCore::ShapeInfo::setReferenceBoxLogicalSize): Renamed from
            setShapeSize.
        (WebCore::ShapeInfo::logicalLineTop): Use new names.
        (WebCore::ShapeInfo::logicalLineBottom): Ditto.
        (WebCore::ShapeInfo::shapeContainingBlockLogicalHeight): Ditto.
        (WebCore::ShapeInfo::markShapeAsDirty): Renamed from dirtyShapeSize.
        (WebCore::ShapeInfo::isShapeDirty): Renamed from shapeSizeDirty.
        (WebCore::ShapeInfo::referenceBoxLogicalSize): Renamed from shapeSize.
        (WebCore::ShapeInfo::logicalTopOffset): Use new names.
        (WebCore::ShapeInfo::logicalLeftOffset): Ditto.
        * rendering/shapes/ShapeInsideInfo.cpp:
        (WebCore::ShapeInsideInfo::updateSegmentsForLine): Ditto.
        (WebCore::ShapeInsideInfo::adjustLogicalLineTop): Ditto.
        (WebCore::ShapeInsideInfo::computeFirstFitPositionForFloat): Ditto.
        * rendering/shapes/ShapeInsideInfo.h:
        * rendering/shapes/ShapeOutsideInfo.cpp:
        (WebCore::ShapeOutsideInfo::updateDeltasForContainingBlockLine): Ditto.
        * rendering/shapes/ShapeOutsideInfo.h:

2014-02-12  Brent Fulgham  <bfulgham@apple.com>

        REGRESSION: Crashing/Broken Tests Due To Unexpected 8-bit Character Data
        https://bugs.webkit.org/show_bug.cgi?id=128698

        Reviewed by Tim Horton.

        * platform/graphics/win/UniscribeController.cpp:
        (WebCore::UniscribeController::advance): Make 16-bit copy when needed.
        (WebCore::UniscribeController::shapeAndPlaceItem): Handle 8-bit case when checking
        for word boundaries.

2014-02-12  Benjamin Poulain  <bpoulain@apple.com>

        Document::childrenChanged does not necessarily have a page

        Rubber stamped by Enrica Casucci.

        * dom/Document.cpp:
        (WebCore::Document::childrenChanged):

2014-02-12  Timothy Hatcher  <timothy@apple.com>

        Add a missing ": " between the URL and exception in STDOUT logs.

        https://bugs.webkit.org/show_bug.cgi?id=128689

        Reviewed by Joseph Pecoraro.

        * page/PageConsole.cpp:
        (WebCore::PageConsole::addMessage): Print ": " after calling
        printSourceURLAndPosition now that it does not print it for us.

2014-02-12  Enrica Casucci  <enrica@apple.com>

        WK2: coordinate mapping for frames does not work when the page is scrolled.
        https://bugs.webkit.org/show_bug.cgi?id=128690
        <rdar://problem/16042925>

        Reviewed by Simon Fraser.

        We should not apply the scroll offset when using delegate scrolling.

        * platform/ScrollView.cpp:
        (WebCore::ScrollView::rootViewToContents):
        (WebCore::ScrollView::contentsToRootView):
        (WebCore::ScrollView::rootViewToTotalContents):

2014-02-12  Joseph Pecoraro  <pecoraro@apple.com>

        Web Inspector: Rename PageInjectedScript* to WebInjectedScript*
        https://bugs.webkit.org/show_bug.cgi?id=128660

        Reviewed by Timothy Hatcher.

        Rename the files update build systems and users.

        * CMakeLists.txt:
        * GNUmakefile.list.am:
        * WebCore.vcxproj/WebCore.vcxproj:
        * WebCore.vcxproj/WebCore.vcxproj.filters:
        * WebCore.xcodeproj/project.pbxproj:
        * inspector/CommandLineAPIModule.cpp:
        (WebCore::CommandLineAPIModule::host):
        * inspector/InspectorAllInOne.cpp:
        * inspector/InspectorController.cpp:
        (WebCore::InspectorController::InspectorController):
        * inspector/InspectorController.h:
        * inspector/InspectorHeapProfilerAgent.cpp:
        (WebCore::InspectorHeapProfilerAgent::InspectorHeapProfilerAgent):
        * inspector/InspectorHeapProfilerAgent.h:
        * inspector/InspectorProfilerAgent.cpp:
        (WebCore::PageProfilerAgent::PageProfilerAgent):
        (WebCore::InspectorProfilerAgent::create):
        (WebCore::WorkerProfilerAgent::WorkerProfilerAgent):
        (WebCore::InspectorProfilerAgent::InspectorProfilerAgent):
        * inspector/InspectorProfilerAgent.h:
        * inspector/PageConsoleAgent.cpp:
        (WebCore::PageConsoleAgent::PageConsoleAgent):
        (WebCore::PageConsoleAgent::addInspectedNode):
        * inspector/PageConsoleAgent.h:
        * inspector/WebConsoleAgent.cpp:
        (WebCore::WebConsoleAgent::WebConsoleAgent):
        (WebCore::WebConsoleAgent::frameWindowDiscarded):
        (WebCore::WebConsoleAgent::addInspectedHeapObject):
        * inspector/WebConsoleAgent.h:
        * inspector/WebInjectedScriptHost.cpp: Renamed from Source/WebCore/inspector/PageInjectedScriptHost.cpp.
        (WebCore::WebInjectedScriptHost::type):
        (WebCore::WebInjectedScriptHost::isHTMLAllCollection):
        * inspector/WebInjectedScriptHost.h: Renamed from Source/WebCore/inspector/PageInjectedScriptHost.h.
        * inspector/WebInjectedScriptManager.cpp: Renamed from Source/WebCore/inspector/PageInjectedScriptManager.cpp.
        (WebCore::WebInjectedScriptManager::WebInjectedScriptManager):
        (WebCore::WebInjectedScriptManager::disconnect):
        (WebCore::WebInjectedScriptManager::didCreateInjectedScript):
        (WebCore::WebInjectedScriptManager::discardInjectedScriptsFor):
        * inspector/WebInjectedScriptManager.h: Renamed from Source/WebCore/inspector/PageInjectedScriptManager.h.
        * inspector/WorkerConsoleAgent.cpp:
        (WebCore::WorkerConsoleAgent::WorkerConsoleAgent):
        * inspector/WorkerConsoleAgent.h:
        * inspector/WorkerInspectorController.cpp:
        (WebCore::WorkerInspectorController::WorkerInspectorController):
        * inspector/WorkerInspectorController.h:

2014-02-12  Brent Fulgham  <bfulgham@apple.com>

        Unreviewed iOS Build fix after r163975.

        * page/EventHandler.cpp: Use wheel event stubs if !PLATFORM(MAC) || PLATFORM(IOS).

2014-02-12  Gavin Barraclough  <barraclough@apple.com>

        Clean up PageThrottler interface
        https://bugs.webkit.org/show_bug.cgi?id=128677

        Reviewed by Benjamin Poulain.

        Currently, responsibility for throttling DOM timers & suspending animations is split
        between the Page & the PageThrottler. Clarify by making Page responsible for suspending
        animations (PageThrottler is now purely related to aspects of timer throttling), and
        move all timer throttling policy to the PageThrottler, with a single function on Page to
        enable (Page::setTimerThrottlingEnabled).

        Also, transmit the full ViewState to the PageThrottler (not just the IsVisuallyIdle flag),
        and distinguish between media & page-load activity.

        * WebCore.exp.in:
            - removed setDOMTimerAlignmentInterval.
        * html/HTMLMediaElement.cpp:
        (WebCore::HTMLMediaElement::parseAttribute):
            - createActivityToken -> mediaActivityToken
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::started):
            - createActivityToken -> pageLoadActivityToken
        * page/Page.cpp:
        (WebCore::Page::Page):
            - added m_timerThrottlingEnabled, made PageThrottler a member
        (WebCore::Page::setIsVisuallyIdleInternal):
            - update animiation suspension.
        (WebCore::Page::setTimerThrottlingEnabled):
            - setTimerAlignmentInterval -> setTimerThrottlingEnabled
        (WebCore::Page::setViewState):
            - pass viewState to PageThrottler.
        (WebCore::Page::setIsVisibleInternal):
            - don't update timer throttling.
        * page/Page.h:
        (WebCore::Page::pageThrottler):
            - made PageThrottler a member.
        (WebCore::Page::timerAlignmentInterval):
            - inlined.
        * page/PageThrottler.cpp:
        (WebCore::PageThrottler::PageThrottler):
            - initialize ViewState.
        (WebCore::PageThrottler::~PageThrottler):
            - clean up by calling setTimerThrottlingEnabled directly.
        (WebCore::PageThrottler::hiddenPageDOMTimerThrottlingStateChanged):
            - moved from Page.
        (WebCore::PageThrottler::mediaActivityToken):
        (WebCore::PageThrottler::pageLoadActivityToken):
            - from Page::createActivityToken
        (WebCore::PageThrottler::throttlePage):
        (WebCore::PageThrottler::unthrottlePage):
            - don't throttle animations here.
        (WebCore::PageThrottler::setViewState):
        (WebCore::PageThrottler::setIsVisible):
            - added, throttle Dom timers.
        * page/PageThrottler.h:
        * page/Settings.cpp:
        (WebCore::Settings::setHiddenPageDOMTimerThrottlingEnabled):
            - removed setDOMTimerAlignmentInterval, hiddenPageDOMTimerThrottlingStateChanged moved to PageThrottler.
        * page/Settings.h:
            - removed setDOMTimerAlignmentInterval.

2014-02-12  Benjamin Poulain  <bpoulain@apple.com>

        [WK2][iOS] Add back the special viewport for the old xhtml mobile doctype
        https://bugs.webkit.org/show_bug.cgi?id=128639

        Reviewed by Andreas Kling.

        * WebCore.exp.in:
        * dom/Document.cpp:
        (WebCore::Document::childrenChanged):
        Document::setDocType() has been removed and the doctype update code with it.
        Add a call to didReceiveDocType() to ensure the viewport is updated when the doctype is parsed.

        * loader/EmptyClients.h:
        * page/Chrome.cpp:
        (WebCore::Chrome::didReceiveDocType):
        * page/ChromeClient.h:
        * page/ViewportConfiguration.cpp:
        (WebCore::ViewportConfiguration::xhtmlMobileParameters):
        * page/ViewportConfiguration.h:

2014-02-12  Alexey Proskuryakov  <ap@apple.com>

        Mountain Lion build fix.

        * WebCore.exp.in: Only export WebCrypto symbols on 10.9+, because SUBTLE_CRYPTO
        is not enabled on 10.8.

2014-02-12  Alexey Proskuryakov  <ap@apple.com>

        Wrap WebCrypto keys in SerializedScriptValue
        https://bugs.webkit.org/show_bug.cgi?id=128680

        Reviewed by Anders Carlsson.

        Test: crypto/subtle/rsa-indexeddb.html

        Added Mac code to wrap a key with AES-GCM. We then serialize it into a plist,
        because more custom formats would be crazy (even the custom format in SerializedScriptValue
        makes me nervous, we'll certainly need to change CryptoKey in the future).

        * WebCore.exp.in:
        * WebCore.xcodeproj/project.pbxproj:
        * crypto/CommonCryptoUtilities.h:
        * crypto/SerializedCryptoKeyWrap.h: Added.
        * crypto/mac/SerializedCryptoKeyWrapMac.mm: Added.
        (WebCore::vectorFromNSData):
        (WebCore::wrapSerializedCryptoKey):
        (WebCore::unwrapSerializedCryptoKey):

2014-02-12  Brent Fulgham  <bfulgham@apple.com>

        Wheel events don't latch to inner scrollable elements 
        https://bugs.webkit.org/show_bug.cgi?id=128225
        <rdar://problem/12183688>

        Reviewed by Simon Fraser

        * WebCore.exp.in: Add declarations for new scrolledToTop, scrolledToBottom, scrolledToLeft, and scrolledToRight.
        * page/EventHandler.cpp:
        (WebCore::EventHandler::EventHandler):
        (WebCore::EventHandler::clear):
        (WebCore::findScrollableContainer): New helper function to locate first node
        in enclosing region of document that is capable of handling mouse wheel events.
        (WebCore::isAtMaxDominantScrollPosition): Predicate to check if the scrollable
        area is at the limit we will hit based on scroll direction.
        (WebCore::EventHandler::handleWheelEvent): Identify the case where we have hit
        the end of a scroll, and treat that as a valid 'handled' case. If the scroll event
        is just starting, treat end-of-scroll as unhandled so the parent element can
        handle things.
        * page/EventHandler.h:
        * page/scrolling/ScrollingTree.cpp:
        (WebCore::ScrollingTree::shouldHandleWheelEventSynchronously): Use new methods
        on the PlatformWheelEvent class.
        (WebCore::ScrollingTree::setOrClearLatchedNode): Ditto
        * platform/PlatformWheelEvent.h:
        (WebCore::PlatformWheelEvent::shouldConsiderLatching): Moved implementation from ScrollingTree.
        (WebCore::PlatformWheelEvent::shouldClearLatchedNode): Ditto
        * platform/ScrollableArea.cpp:
        (WebCore::ScrollableArea::scrolledToTop): Added
        (WebCore::ScrollableArea::scrolledToBottom):Added
        (WebCore::ScrollableArea::scrolledToLeft): Added
        (WebCore::ScrollableArea::scrolledToRight): Added
        * platform/ScrollableArea.h:
        * rendering/RenderListBox.cpp:
        (WebCore::RenderListBox::scrolledToTop): Added
        (WebCore::RenderListBox::scrolledToBottom): Added
        (WebCore::RenderListBox::scrolledToLeft): Added
        (WebCore::RenderListBox::scrolledToRight): Added
        * rendering/RenderListBox.h: Changed to public inheritance of ScrollableArea to
        allow generic use of this type in scroll wheel logic.

2014-02-12  Brendan Long  <b.long@cablelabs.com>

        Implement DataCue for metadata cues
        https://bugs.webkit.org/show_bug.cgi?id=128402

        Reviewed by Eric Carlson.

        Test: media/track/track-datacue.html

        * CMakeLists.txt: Add DataCue.
        * DerivedSources.cpp: Add JSDataCue.
        * DerivedSources.make: Same.
        * GNUmakefile.list.am: Add DataCue and JSDataCue.
        * WebCore.vcxproj/WebCore.vcxproj: Add DataCue.
        * WebCore.vcxproj/WebCore.vcxproj.filters: Same.
        * WebCore.xcodeproj/project.pbxproj: Add DataCue and JSDataCue.
        * bindings/js/JSTextTrackCueCustom.cpp:
        (WebCore::toJS): Add DataCue wrapper.
        * html/track/DataCue.cpp: Copied from Source/WebCore/html/track/TextTrack.idl.
        * html/track/DataCue.h: Copied from Source/WebCore/html/track/TextTrack.idl.
        * html/track/DataCue.idl: Copied from Source/WebCore/html/track/TextTrack.idl.
        * html/track/InbandGenericTextTrack.cpp:
        (WebCore::InbandGenericTextTrack::addGenericCue): Pass ASSERT_NO_EXCEPTION to addCue().
        * html/track/InbandWebVTTTextTrack.cpp:
        (WebCore::InbandWebVTTTextTrack::newCuesParsed): Same.
        * html/track/TextTrack.cpp:
        (WebCore::TextTrack::addCue): Throw an exception if DataCue is added to a track that isn't kind="metadata".
        * html/track/TextTrack.h: Add exception code parameter to addCue().
        * html/track/TextTrack.idl: Add [RaisesException] to addCue().
        * html/track/TextTrackCue.h: Add Data CueType.

2014-02-12  Zalan Bujtas  <zalan@apple.com>

        Subpixel layout: Clean up LayoutPoint class.
        https://bugs.webkit.org/show_bug.cgi?id=128515

        Reviewed by Simon Fraser.

        Remove redundant functions (keep the more explicit ones).

        No change in functionality.

        * page/FrameView.cpp:
        (WebCore::FrameView::viewportConstrainedVisibleContentRect):
        * platform/graphics/LayoutPoint.h:
        (WebCore::toLayoutPoint):
        (WebCore::toLayoutSize):
        (WebCore::roundedIntPoint):
        * platform/graphics/ca/GraphicsLayerCA.cpp:
        (WebCore::GraphicsLayerCA::updateContentsRects):
        * rendering/LayoutState.cpp:
        (WebCore::LayoutState::LayoutState):
        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::addFocusRingRects):
        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::offsetFromContainer):
        * rendering/RenderInline.cpp:
        (WebCore::RenderInline::addFocusRingRects):
        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::updateLayerPosition):
        (WebCore::accumulateOffsetTowardsAncestor):
        (WebCore::RenderLayer::paintBackgroundForFragments):
        (WebCore::RenderLayer::paintForegroundForFragmentsWithPhase):
        (WebCore::RenderLayer::paintOutlineForFragments):
        (WebCore::RenderLayer::paintMaskForFragments):
        (WebCore::RenderLayer::paintOverflowControlsForFragments):
        * rendering/RenderScrollbarPart.cpp:
        (WebCore::RenderScrollbarPart::paintIntoRect):

2014-02-12  Zan Dobersek  <zdobersek@igalia.com>

        [CoordinatedGraphics] Move CoordinatedGraphicsScene, CoordinatedLayerTreeHostProxy to std::function
        https://bugs.webkit.org/show_bug.cgi?id=128473

        Reviewed by Anders Carlsson.


        Move the CoordinatedGraphicsScene class to using std::function instead of WTF::Functional and std::bind
        instead of WTF::bind. The function wrapper is now moved through function calls and not passed by reference,
        and lambda functions are inlined into the dispatchOnMainThread() calls, with the CoordinatedGraphicsScene
        refcount-protected.

        * platform/graphics/texmap/coordinated/CoordinatedGraphicsScene.cpp:
        (WebCore::CoordinatedGraphicsScene::dispatchOnMainThread):
        (WebCore::CoordinatedGraphicsScene::paintToCurrentGLContext):
        (WebCore::CoordinatedGraphicsScene::commitSceneState):
        (WebCore::CoordinatedGraphicsScene::syncRemoteContent):
        (WebCore::CoordinatedGraphicsScene::purgeGLResources):
        (WebCore::CoordinatedGraphicsScene::commitScrollOffset):
        (WebCore::CoordinatedGraphicsScene::appendUpdate):
        (WebCore::CoordinatedGraphicsScene::setActive):
        * platform/graphics/texmap/coordinated/CoordinatedGraphicsScene.h:

2014-02-12  Eric Carlson  <eric.carlson@apple.com>

        Cleanup the code added for https://bugs.webkit.org/show_bug.cgi?id=128125.

        Not reviewed.

        * html/HTMLMediaSession.cpp:
        (WebCore::restrictionName): Don't need a "break" after a "return" in a case statement.
        * platform/audio/MediaSession.cpp:
        (WebCore::stateName): Ditto.
        * platform/audio/ios/MediaSessionManagerIOS.mm:

2014-02-12  Andreas Kling  <akling@apple.com>

        RenderNamedFlowThread should only support RenderElement children.
        <https://webkit.org/b/128675>

        Tighten up flow-thread rendering so that it only supports element
        children directly. This means we don't have to worry about text
        renderers on this code path.

        Reviewed by Antti Koivisto.

        * rendering/RenderElement.cpp:
        (WebCore::RenderElement::insertedIntoTree):
        (WebCore::RenderElement::willBeRemovedFromTree):
        (WebCore::RenderElement::willBeDestroyed):
        * rendering/RenderNamedFlowThread.cpp:
        (WebCore::RenderNamedFlowThread::nextRendererForElement):
        (WebCore::RenderNamedFlowThread::addFlowChild):
        (WebCore::RenderNamedFlowThread::removeFlowChild):
        * rendering/RenderNamedFlowThread.h:
        * rendering/RenderObject.cpp:
        (WebCore::RenderObject::willBeDestroyed):
        (WebCore::RenderObject::insertedIntoTree):
        (WebCore::RenderObject::willBeRemovedFromTree):
        * style/StyleResolveTree.cpp:
        (WebCore::Style::createRendererIfNeeded):

2014-02-12  Joseph Pecoraro  <pecoraro@apple.com>

        Web Inspector: Modernize missed inspector files
        https://bugs.webkit.org/show_bug.cgi?id=128661

        Reviewed by Anders Carlsson.

        Add final, override, and use nullptr where appropriate.

        * inspector/InspectorCSSAgent.cpp:
        (WebCore::InspectorCSSAgent::asCSSStyleRule):
        (WebCore::InspectorCSSAgent::discardAgent):
        (WebCore::InspectorCSSAgent::disable):
        (WebCore::InspectorCSSAgent::getInlineStylesForNode):
        (WebCore::InspectorCSSAgent::getComputedStyleForNode):
        (WebCore::InspectorCSSAgent::asInspectorStyleSheet):
        (WebCore::InspectorCSSAgent::elementForId):
        (WebCore::InspectorCSSAgent::viaInspectorStyleSheet):
        (WebCore::InspectorCSSAgent::assertStyleSheetForId):
        (WebCore::InspectorCSSAgent::buildObjectForRule):
        (WebCore::InspectorCSSAgent::buildObjectForAttributesStyle):
        * inspector/InspectorFrontendClientLocal.h:

2014-02-11  Brent Fulgham  <bfulgham@apple.com>

        Remove some unintended copies in ranged for loops
        https://bugs.webkit.org/show_bug.cgi?id=128644

        Reviewed by Anders Carlsson.

        * css/StyleResolver.cpp:
        (WebCore::StyleResolver::loadPendingSVGDocuments): Avoid creating/destroying
        RefPtrs in loop.

2014-02-12  Raphael Kubo da Costa  <raphael.kubo.da.costa@intel.com>

        Update the HTML Media Capture implementation.
        https://bugs.webkit.org/show_bug.cgi?id=118465

        Reviewed by Darin Adler.

        Make the implementation in WebKit compatible with the 2013-05-09
        version of the spec, which makes the "capture" attribute a boolean
        instead of an enum.

        Covered by fast/forms/file/file-input-capture.html.

        * html/FileInputType.cpp:
        (WebCore::FileInputType::handleDOMActivateEvent):
        * html/HTMLInputElement.cpp:
        (WebCore::HTMLInputElement::capture): Renamed to shouldUseMediaCapture().
        (WebCore::HTMLInputElement::shouldUseMediaCapture): Return a bool.
        * html/HTMLInputElement.h:
        * html/HTMLInputElement.idl: Turn the `capture' attribute into a
        reflective boolean instead of a DOMString.
        * platform/FileChooser.h:

2014-02-12  Radu Stavila  <stavila@adobe.com>

        [CSS Regions] Remove unused method in RenderFlowThread
        https://bugs.webkit.org/show_bug.cgi?id=128373

        Reviewed by Mihnea Ovidenie.

        After the landing of https://bugs.webkit.org/show_bug.cgi?id=118665, the 
        RenderFlowThread::computeRegionClippingRect method is obsolete.

        No new tests needed, this patch only removes an unused method.

        * rendering/RenderFlowThread.cpp:
        * rendering/RenderFlowThread.h:

2014-02-12  Mihai Tica  <mitica@adobe.com>

        [CSS Element Blending] Implement the software path of -webkit-blend-mode with Core Graphics.
        https://bugs.webkit.org/show_bug.cgi?id=99119

        Reviewed by Simon Fraser.

        This patch adds support for -webkit-blend-mode with Core Graphics.
        The layer promotion code that forced compositing when blending was detected has been removed.
        Remaining work for the software path is to detect and implement isolation of the blending operation:
        as stated in the spec, blending should be limited to the parent stacking context.

        Tests: css3/compositing/blend-mode-blended-element-overlapping-composited-sibling-should-have-compositing-layer.html
               css3/compositing/blend-mode-parent-of-composited-blended-has-layer.html
               css3/compositing/blend-mode-simple-composited.html
               css3/compositing/blend-mode-with-composited-descendant-should-have-layer.html

        * inspector/InspectorLayerTreeAgent.cpp:
        (WebCore::InspectorLayerTreeAgent::reasonsForCompositingLayer): Remove CompositingReasonBlending.
        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::beginTransparencyLayers): Set the blendMode on the GraphicsContext.
        * rendering/RenderLayer.h: paintsWithTransparency should return true when a blendMode is set.
        * rendering/RenderLayerCompositor.cpp: Remove promotion code when a blendMode is detected.
        (WebCore::RenderLayerCompositor::requiresCompositingLayer): Remove CompositingReasonBlending reason.
        (WebCore::RenderLayerCompositor::requiresOwnBackingStore): Remove CompositingReasonBlending reason.
        (WebCore::RenderLayerCompositor::reasonsForCompositing): Remove CompositingReasonBlending reason.
        (WebCore::RenderLayerCompositor::logReasonsForCompositing): Remove CompositingReasonBlending reason.
        * rendering/RenderLayerCompositor.h:
        - Remove the requiresCompositingForBlending method
        - Remove the CompositingReasonBlending from the CompositingReasons enum.

2014-02-11  Zalan Bujtas  <zalan@apple.com>

        Subpixel rendering: Switch repaint rect from IntRect to LayoutRect to be able to
        repaint on device pixel boundaries.
        https://bugs.webkit.org/show_bug.cgi?id=128477

        Reviewed by Simon Fraser.

        RenderLayer needs to be able to repaint on device pixel boundaries. RenderView still
        repaints on integral position.

        No change in functionality.

        * rendering/RenderBlockLineLayout.cpp:
        (WebCore::RenderBlockFlow::layoutRunsAndFloats):
        * rendering/RenderElement.cpp:
        (WebCore::RenderElement::repaintAfterLayoutIfNeeded):
        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::updateLayerPositions):
        (WebCore::RenderLayer::clearRepaintRects):
        (WebCore::RenderLayer::scrollTo):
        (WebCore::RenderLayer::calculateClipRects):
        * rendering/RenderLayerBacking.cpp:
        (WebCore::RenderLayerBacking::setContentsNeedDisplayInRect):
        * rendering/RenderLayerBacking.h:
        * rendering/RenderObject.cpp:
        (WebCore::RenderObject::repaintUsingContainer):
        (WebCore::RenderObject::repaint):
        (WebCore::RenderObject::repaintRectangle):
        * rendering/RenderObject.h:

2014-02-11  Dan Bernstein  <mitz@apple.com>

        iOS build fix fix.

        * rendering/RenderElement.cpp:
        (WebCore::shouldRepaintForImageAnimation):

2014-02-11  Dan Bernstein  <mitz@apple.com>

        iOS build fix.

        * rendering/RenderElement.cpp:
        (WebCore::shouldRepaintForImageAnimation):

2014-02-11  David Kilzer  <ddkilzer@apple.com>

        ContentData equals() methods are not inline-able
        <http://webkit.org/b/128538>

        Reviewed by Darin Adler.

        Get rid of pure virtual equals() method in favor of
        ContentData::Type enum for runtime type information.
        This also lets us devirtualize the isFoo() methods.

        * rendering/style/ContentData.h:
        (WebCore::ContentData::type): Add.
        (WebCore::ContentData::isCounter): Devirtualize.
        (WebCore::ContentData::isImage): Devirtualize.
        (WebCore::ContentData::isQuote): Devirtualize.
        (WebCore::ContentData::isText): Devirtualize.
        (WebCore::ContentData::ContentData): Add.  Include
        ContentData::Type parameter.
        (WebCore::operator==): Add overloaded methods for each subclass.
        Stop using pure virtual equals() method, check type(), and use
        overloaded subclass operator==() methods.
        (WebCore::operator!=): Add overloaded methods for each subclass.

2014-02-11  Andreas Kling  <akling@apple.com>

        Protect some RenderFlowThread functions.
        <https://webkit.org/b/128642>

        Make the RenderFlowThread constructor protected and a handful of
        member functions private.

        Reviewed by Anders Carlsson.

        * rendering/RenderFlowThread.h:

2014-02-11  Andreas Kling  <akling@apple.com>

        Don't allocate RenderNamedFlowThread's child list separately.
        <https://webkit.org/b/128640>

        Since we always create the flow-thread child list, there's no reason
        to put it in a separate heap allocation. Also remove the typedef and
        use auto instead.

        Reviewed by Anders Carlsson.

        * rendering/RenderNamedFlowThread.cpp:
        (WebCore::RenderNamedFlowThread::RenderNamedFlowThread):
        (WebCore::RenderNamedFlowThread::nextRendererForNode):
        (WebCore::RenderNamedFlowThread::addFlowChild):
        (WebCore::RenderNamedFlowThread::removeFlowChild):
        * rendering/RenderNamedFlowThread.h:

2014-02-11  Zalan Bujtas  <zalan@apple.com>

        Subpixel rendering: Make GraphicsLayerClient::paintContents's clip rect subpixel based.
        https://bugs.webkit.org/show_bug.cgi?id=128460

        Reviewed by Simon Fraser.

        GraphicsClient::paintContents takes clipRect as FloatRect now so that we can paint on
        subpixel position.

        No change in functionality.

        * platform/graphics/GraphicsLayerClient.h:
        * platform/graphics/texmap/coordinated/CompositingCoordinator.cpp:
        (WebCore::CompositingCoordinator::paintContents):
        * platform/graphics/texmap/coordinated/CompositingCoordinator.h:
        * rendering/RenderLayer.cpp:
        (WebCore::cornerRect):
        (WebCore::RenderLayer::scrollCornerRect):
        (WebCore::resizerCornerRect):
        (WebCore::RenderLayer::scrollCornerAndResizerRect):
        (WebCore::RenderLayer::verticalScrollbarStart):
        (WebCore::RenderLayer::horizontalScrollbarStart):
        (WebCore::RenderLayer::paintResizer):
        (WebCore::RenderLayer::hitTestOverflowControls):
        * rendering/RenderLayer.h:
        * rendering/RenderLayerBacking.cpp:
        (WebCore::RenderLayerBacking::paintIntoLayer):
        (WebCore::RenderLayerBacking::paintContents):
        * rendering/RenderLayerBacking.h:
        * rendering/RenderLayerCompositor.cpp:
        (WebCore::RenderLayerCompositor::paintContents):
        * rendering/RenderLayerCompositor.h:
        * rendering/RenderWidget.cpp:
        (WebCore::RenderWidget::paint):

2014-02-11  Antti Koivisto  <antti@apple.com>

        GIF animations should be suspended when outside of viewport
        https://bugs.webkit.org/show_bug.cgi?id=128632

        Reviewed by Andreas Kling.
        
        Animations are driven by the paint cycle. Speculative tiles keep animations outside the actual viewport going.
        
        Pause animations when they are outside the viewport by not painting them.

        Test: fast/repaint/no-animation-outside-viewport.html

        * loader/cache/CachedImage.cpp:
        (WebCore::CachedImage::animationAdvanced):
        
            Call animation specific newImageAnimationFrameAvailable instead of the generic notifyObservers.

        * loader/cache/CachedImage.h:
        
            Removed now unnecessary resumeAnimatingImagesForLoader mechanism.
            Remove unnecessary shouldPauseAnimation. Pausing is now always done when by avoiding repaint.

        * loader/cache/CachedImageClient.h:
        (WebCore::CachedImageClient::newImageAnimationFrameAvailable):
        * page/FrameView.cpp:
        (WebCore::FrameView::scrollPositionChanged):
        
            Check if we have image animations to resume when scroll position changes.

        * page/Page.cpp:
        (WebCore::Page::resumeAnimatingImages):
        
            Use the same mechanism when resuming background tabs etc.

        * platform/graphics/BitmapImage.cpp:
        (WebCore::BitmapImage::internalAdvanceAnimation):
        
            Remove the shouldPauseAnimation test, always rely on pausing on invalidation.

        * platform/graphics/ImageObserver.h:
        * rendering/RenderBoxModelObject.cpp:
        * rendering/RenderElement.cpp:
        (WebCore::RenderElement::RenderElement):
        (WebCore::RenderElement::~RenderElement):
        (WebCore::shouldRepaintForImageAnimation):
        
            Factor the pausing conditions to a function. Test that the animation is withing the
            visible rect.

        (WebCore::RenderElement::newImageAnimationFrameAvailable):
        
            Add renderer to the paused animation set if we don't continue the animation.

        (WebCore::RenderElement::repaintForPausedImageAnimationsIfNeeded):
        
            Resume the paused animations by triggering repaint.

        * rendering/RenderElement.h:
        (WebCore::RenderElement::setHasPausedImageAnimations):
        (WebCore::RenderElement::hasPausedImageAnimations):
        * rendering/RenderObject.cpp:
        * rendering/RenderObject.h:
        * rendering/RenderView.cpp:
        (WebCore::RenderView::addRendererWithPausedImageAnimations):
        (WebCore::RenderView::removeRendererWithPausedImageAnimations):
        (WebCore::RenderView::resumePausedImageAnimationsIfNeeded):
        * rendering/RenderView.h:

2014-02-11  Andreas Kling  <akling@apple.com>

        Remove unused RenderNamedFlowThread::previousRendererForNode().
        <https://webkit.org/b/128637>

        Reviewed by Antti Koivisto.

        * rendering/RenderNamedFlowThread.cpp:
        * rendering/RenderNamedFlowThread.h:

2014-02-11  Andreas Kling  <akling@apple.com>

        Move renderNamedFlowThreadWrapper() to RenderElement.
        <https://webkit.org/b/128634>

        This function is only ever called on RenderElements so move it there
        from RenderObject.

        Reviewed by Antti Koivisto.

        * rendering/RenderElement.cpp:
        (WebCore::RenderElement::renderNamedFlowThreadWrapper):
        * rendering/RenderElement.h:
        * rendering/RenderObject.cpp:
        * rendering/RenderObject.h:

2014-02-11  Myles C. Maxfield  <mmaxfield@apple.com>

        Position and thickness of underline as text size changes
        https://bugs.webkit.org/show_bug.cgi?id=16768

        Reviewed by Dean Jackson.

        This patch adopts the iOS codepath for underlines. It also reorganizes
        drawLineForText to avoid a costly global state save & restore.

        Test: fast/css3-text/css3-text-decoration/text-decoration-thickness.html

        * platform/graphics/cg/GraphicsContextCG.cpp:
        (WebCore::computeLineBoundsAndAntialiasingModeForText):
        (WebCore::GraphicsContext::computeLineBoundsForText):
        (WebCore::GraphicsContext::drawLineForText):
        (WebCore::GraphicsContext::drawLinesForText):
        * rendering/InlineTextBox.cpp:
        (WebCore::InlineTextBox::paintDecoration):

2014-02-11  Ryosuke Niwa  <rniwa@webkit.org>

        Frame::rectForSelection shouldn't instantiate FrameSelection
        https://bugs.webkit.org/show_bug.cgi?id=128587

        Reviewed by Enrica Casucci.

        Made VisiblePosition::absoluteCaretBounds more interoperable with the one in FrameSelection and made
        iOS's Frame::rectForScrollToVisible use that function instead.

        The above change allows us to remove:
        - suppressCloseTyping(), restoreCloseTyping(), and m_closeTypingSuppressions in FrameSelection
        - suppressSelectionNotifications() and restoreSelectionNotifications() in EditorClient

        See inline comments below for more details.

        * Source/WebCore/WebCore.exp.in:

        * editing/FrameSelection.cpp:
        (WebCore::FrameSelection::FrameSelection):
        (WebCore::FrameSelection::setSelectionWithoutUpdatingAppearance):
        (WebCore::CaretBase::updateCaretRect):
        (WebCore::FrameSelection::caretRendererWithoutUpdatingLayout):
        (WebCore::DragCaretController::caretRenderer):
        (WebCore::repaintCaretForLocalRect):
        (WebCore::FrameSelection::recomputeCaretRect): Merged FrameSelection::localCaretRect(). Modified
        the code to update caretNode when and only when caret rect is updated. Also added an assertion to
        ensure absoluteCaretBounds() on FrameSelection and VisiblePosition yield the same result.

        (WebCore::CaretBase::paintCaret):
        * editing/FrameSelection.h:

        * editing/VisiblePosition.cpp:
        (WebCore::VisiblePosition::absoluteCaretBounds): Fixed the bug where the old code wasn't respecting
        the convention to use containing block as the renderer to paint caret.

        * editing/htmlediting.cpp:
        (WebCore::caretRendersInsideNode): Moved from FrameSelection.cpp.
        (WebCore::rendererForCaretPainting): Ditto and renamed from caretRenderer.
        (WebCore::localCaretRectInRendererForCaretPainting): Extracted from FrameSelection::updateCaretRect.
        (WebCore::absoluteBoundsForLocalCaretRect): Ditto from CaretBase::absoluteBoundsForLocalRect.
        * editing/htmlediting.h:

        * loader/EmptyClients.h:
        * page/EditorClient.h:
        * page/Frame.h:

        * page/ios/FrameIOS.mm:
        (WebCore::Frame::rectForScrollToVisible): Reimplemented in its simplest form using VisiblePosition's
        absoluteCaretBounds().

2014-02-11  Enrica Casucci  <enrica@apple.com>

        Support WebSelections in WK2 on iOS.
        https://bugs.webkit.org/show_bug.cgi?id=127015
        <rdar://problem/15211964>

        Reviewed by Benjamin Poulain.

        Adding few exports.

        * WebCore.exp.in:

2014-02-11  Andreas Kling  <akling@apple.com>

        CTTE: RenderNamedFlowThread always has a WebKitNamedFlow.
        <https://webkit.org/b/128623>

        Codify the fact that RenderNamedFlowThread always has a corresponding
        WebKitNamedFlow by storing it in a Ref, and adding an accessor that
        returns a reference to get rid of all the ->'s.

        Also removed some unnecessary assertions exposed by this.

        Reviewed by Antti Koivisto.

        * dom/NamedFlowCollection.cpp:
        (WebCore::NamedFlowCollection::ensureFlowWithName):
        * dom/NamedFlowCollection.h:
        * rendering/RenderNamedFlowThread.cpp:
        (WebCore::RenderNamedFlowThread::RenderNamedFlowThread):
        (WebCore::RenderNamedFlowThread::registerNamedFlowContentElement):
        (WebCore::RenderNamedFlowThread::unregisterNamedFlowContentElement):
        (WebCore::RenderNamedFlowThread::flowThreadName):
        (WebCore::RenderNamedFlowThread::dispatchRegionLayoutUpdateEvent):
        (WebCore::RenderNamedFlowThread::dispatchRegionOversetChangeEvent):
        (WebCore::RenderNamedFlowThread::regionLayoutUpdateEventTimerFired):
        (WebCore::RenderNamedFlowThread::regionOversetChangeEventTimerFired):
        (WebCore::RenderNamedFlowThread::setMarkForDestruction):
        (WebCore::RenderNamedFlowThread::resetMarkForDestruction):
        (WebCore::RenderNamedFlowThread::isMarkedForDestruction):
        * rendering/RenderNamedFlowThread.h:

2014-02-10  Myles C. Maxfield  <mmaxfield@apple.com>

        Convert position:fixed property to position:absolute upon copy
        https://bugs.webkit.org/show_bug.cgi?id=128194

        Reviewed by Simon Fraser.

        This adds a Setting and Preference that allows clients to opt-in to this behavior.

        This new behavior is only activated if the entire body is copied. If there is a position:fixed
        element in the copied selection, it is replaced with position:absolute, and a containing
        block (position:relative) is wrapped around the copied text.

        This patch originally converted position:-webkit-sticky to position:relative. However, we
        currently don't support copying and pasting of position:-webkit-sticky content (See below).
        Therefore, this patch only deals with position:fixed.

        Right now we don't copy position:-webkit-sticky. This is because:
        1. When copying styled elements, we parse the style properties again
        2. CSSParserContext has a flag which can disable parsing -webkit-sticky
        3. There are two constructors to CSSParserContext: one that takes a document and sets up the
        aforementioned flag, and a simple one that doesn't take a document and sets all the enableFoo
        flags to false
        4. At the relevant place within copy code, we are far removed from the Document object, so we
        instead call the second constructor, thereby disabling parsing of -webkit-sticky

        Test: editing/pasteboard/copy-paste-converts-sticky-and-fixed.html

        * WebCore.exp.in: Export the Setting setter
        * editing/EditingStyle.cpp:
        (WebCore::EditingStyle::convertFixedAndStickyPosition): Converts a single style
        * editing/EditingStyle.h:
        * editing/markup.cpp:
        (WebCore::StyledMarkupAccumulator::StyledMarkupAccumulator): Remember if we found
        an element which needs the position:relative containing block
        (WebCore::StyledMarkupAccumulator::appendElement): Surround with the position:relative
        containing block if necessary
        (WebCore::createMarkupInternal):
        * page/Settings.cpp:
        (WebCore::Settings::Settings): New setting to opt-in to this new behavior
        (WebCore::Settings::setConvertPositionStyleOnCopy):
        * page/Settings.h:
        (WebCore::Settings::convertPositionStyleOnCopy):
        * testing/InternalSettings.cpp: Allow setting the setting from a Layout Test
        (WebCore::InternalSettings::Backup::Backup):
        (WebCore::InternalSettings::Backup::restoreTo):
        (WebCore::InternalSettings::setConvertPositionStyleOnCopy):
        * testing/InternalSettings.h:
        * testing/InternalSettings.idl:

2014-02-11  Youenn Fablet  <youennf@gmail.com>

        XMLHttpRequest should not send DNT header
        https://bugs.webkit.org/show_bug.cgi?id=128533

        Reviewed by Alexey Proskuryakov.

        Added DNT (Do Not Track) header to the list of forbidden headers.
        Updated http/tests/xmlhttprequest/set-dangerous-headers.html to test that header.

        * xml/XMLHttpRequest.cpp:
        (WebCore::XMLHttpRequestStaticData::XMLHttpRequestStaticData):

2014-02-10  Jeffrey Pfau  <jpfau@apple.com>

        Bring third-party app cache blocking behavior in line with private browsing app cache blocking behavior
        https://bugs.webkit.org/show_bug.cgi?id=128557

        Reviewed by Alexey Proskuryakov.

        * loader/appcache/ApplicationCacheGroup.cpp:
        (WebCore::ApplicationCacheGroup::cacheForMainRequest):
        (WebCore::ApplicationCacheGroup::selectCache):
        (WebCore::ApplicationCacheGroup::selectCacheWithoutManifestURL):
        (WebCore::ApplicationCacheGroup::update):
        * loader/appcache/ApplicationCacheHost.cpp:
        (WebCore::ApplicationCacheHost::maybeLoadMainResource):
        (WebCore::ApplicationCacheHost::maybeLoadFallbackForMainResponse):
        (WebCore::ApplicationCacheHost::maybeLoadFallbackForMainError):
        (WebCore::ApplicationCacheHost::maybeLoadResource):
        (WebCore::ApplicationCacheHost::scheduleLoadFallbackResourceFromApplicationCache):
        (WebCore::ApplicationCacheHost::isApplicationCacheBlockedForRequest):
        * loader/appcache/ApplicationCacheHost.h:

2014-02-11  Brady Eidson  <beidson@apple.com>

        IDB: The test after storage/indexeddb/mozilla/object-identity.html fails in cleanup code
        <rdar://problem/16040663> and https://bugs.webkit.org/show_bug.cgi?id=128621

        Reviewed by Alexey Proskuryakov.

        Covered by storage/indexeddb/mozilla/object-identity.html.

        * Modules/indexeddb/IDBTransactionBackend.cpp:
        (WebCore::IDBTransactionBackend::abort): Clear the m_database pointer before calling the onAbort callback.
        (WebCore::IDBTransactionBackend::commit): Don't run the abort code if there's no m_database pointer.

2014-02-10  Jer Noble  <jer.noble@apple.com>

        [EME][Mac] Move the implementation of CDMPrivateAVFoundation back into MediaPlayerPrivateAVFoundationObjC.
        https://bugs.webkit.org/show_bug.cgi?id=128559

        Reviewed by Dean Jackson.

        To prepare for multiple simultaneous CDMs with muliple MediaPlayer types, move the implementation for
        CDMPrivateAVFoundation back into its media engine.

        * Modules/encryptedmedia/CDMPrivateAVFoundation.mm:
        (WebCore::MediaKeyExceptionToErrorCode): Added.
        (WebCore::CDMSessionAVFoundation::generateKeyRequest): Moved to MediaPlayerPrivateAVFoundationObjC.
        (WebCore::CDMSessionAVFoundation::releaseKeys): Ditto.
        (WebCore::CDMSessionAVFoundation::update): Ditto.
        * platform/graphics/MediaPlayer.cpp:
        (WebCore::MediaPlayer::generateKeyRequest): Added, pass through to MediaPlayerPrivate.
        (WebCore::MediaPlayer::releaseKeys): Ditto.
        (WebCore::MediaPlayer::update): Ditto.
        * platform/graphics/MediaPlayer.h:
        * platform/graphics/MediaPlayerPrivate.h:
        (WebCore::MediaPlayerPrivateInterface::generateKeyRequest): Added.
        (WebCore::MediaPlayerPrivateInterface::releaseKeys): Ditto.
        (WebCore::MediaPlayerPrivateInterface::update): Ditto.
        * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
        * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
        (WebCore::MediaPlayerPrivateAVFoundationObjC::generateKeyRequest): Moved from CDMSessionAVFoundation.
        (WebCore::MediaPlayerPrivateAVFoundationObjC::releaseKeys): Ditto.
        (WebCore::MediaPlayerPrivateAVFoundationObjC::update): Ditto.

2014-01-24  Jer Noble  <jer.noble@apple.com>

        Run UserAgentScripts through jsmin rather than the css preprocessor
        https://bugs.webkit.org/show_bug.cgi?id=127559

        Reviewed by Tim Horton.

        User Agent JavaScript files were being run through the c++ preprocessor to strip out
        comments (and presumably to allow #if ENABLE macros, though that feature is entirely
        unused). This had the side effect of removing important whitespace, namely newlines where
        there would normally be an implicit semicolon.

        Instead, .js files will now be run through the jsmin minifier, used by the inspector.
        Jsmin will also strip comments and whitespace, but in a syntactically aware way which will
        keep newlines when their presence adds an implied semicolon.

        * DerivedSources.make:
        * Scripts/make-js-file-arrays.py: Added.
        (stringifyCodepoint):
        (chunk):
        (main):

2014-02-11  Andy Estes  <aestes@apple.com>

        [Mac] connection:willStopBufferingData: no longer exists in NSURLConnectionDelegate
        https://bugs.webkit.org/show_bug.cgi?id=128583

        Reviewed by Anders Carlsson.

        The delegate method was removed in Snow Leopard.

        * loader/ResourceLoader.cpp:
        * loader/ResourceLoader.h:
        * platform/network/ResourceHandleClient.h:
        * platform/network/mac/WebCoreResourceHandleAsDelegate.mm:
        * platform/network/mac/WebCoreResourceHandleAsOperationQueueDelegate.mm:

2014-02-10  Andy Estes  <aestes@apple.com>

        [Content Filter] Check for NULL before calling dispatch_release()
        https://bugs.webkit.org/show_bug.cgi?id=128576

        Reviewed by Darin Adler.

        m_neFilterSourceQueue will be NULL if NEFilterSource isn't enabled, and
        passing NULL to dispatch_release() is undefined.

        * platform/mac/ContentFilterMac.mm:
        (WebCore::ContentFilter::~ContentFilter):

2014-01-24  Jer Noble  <jer.noble@apple.com>

        [MediaControls] Allow the media controls script to be debuggable by giving it a generated sourceURL
        https://bugs.webkit.org/show_bug.cgi?id=127560

        Reviewed by Eric Carlson.

        When evaluating a script through ScriptController, if that script does not have an
        explicit sourceURL, it will not appear in the resources section of the Web Inspector.
        For debug builds only, give the media controls script a generated (i.e. fake) sourceURL.

        * html/HTMLMediaElement.cpp:
        (WebCore::HTMLMediaElement::parseAttribute):

2014-02-11  Samuel White  <samuel_white@apple.com>

        AX: Add text replacement activity support to NSAccessibilitySelectTextWithCriteriaParameterizedAttribute.
        https://bugs.webkit.org/show_bug.cgi?id=128397

        Reviewed by Chris Fleizach.

        Added text replacement support for the AXSelectTextWithCriteria parameterized attribute.

        No new test. Updated platform/mac/accessibility/select-text.html to test the added functionality.

        * accessibility/AccessibilityObject.cpp:
        (WebCore::AccessibilityObject::selectText):
        * accessibility/AccessibilityObject.h:
        (WebCore::AccessibilitySelectTextCriteria::AccessibilitySelectTextCriteria):
        * accessibility/mac/WebAccessibilityObjectWrapperMac.mm:
        (accessibilitySelectTextCriteriaForCriteriaParameterizedAttribute):

2014-02-11  Roger Fong  <roger_fong@apple.com>

        [Windows] Unreviewed. Speculative test fix.

        * platform/graphics/GraphicsContext.h:
        (WebCore::GraphicsContext::pixelSnappingFactor):
        * platform/graphics/cg/GraphicsContextCG.cpp:
        (WebCore::GraphicsContext::platformInit):

2014-02-11  Dan Bernstein  <mitz@apple.com>

        iOS build fix.

        * platform/ios/TileGrid.mm:
        (WebCore::TileGrid::dropTilesBetweenRects):
        (WebCore::TileGrid::dropDistantTiles):
        (WebCore::TileGrid::dropInvalidTiles):

2014-02-11  James Craig  <jcraig@apple.com>

        Web Inspector: AX: Accessibility Node Inspection
        https://bugs.webkit.org/show_bug.cgi?id=127447

        Reviewed by Timothy Hatcher.

        New methods supporting WebCore::AccessibilityObject::computedRoleString()
        used for Accessibility section in WebInspector Node Inspector. Other updates 
        support the JSON interface for the WebInspectorUI feature.

        Test: accessibility/roles-computedRoleString.html

        * accessibility/AccessibilityObject.cpp:
        (WebCore::initializeRoleMap):
        (WebCore::ariaRoleMap):
        (WebCore::reverseAriaRoleMap):
        (WebCore::AccessibilityObject::ariaRoleToWebCoreRole):
        (WebCore::AccessibilityObject::computedRoleString):
        * accessibility/AccessibilityObject.h:
        * accessibility/mac/WebAccessibilityObjectWrapperMac.mm:
        (-[WebAccessibilityObjectWrapper computedRoleString]):
        (-[WebAccessibilityObjectWrapper accessibilityAttributeValue:]):
        * inspector/InspectorDOMAgent.cpp:
        (WebCore::InspectorDOMAgent::getAccessibilityPropertiesForNode):
        (WebCore::InspectorDOMAgent::buildObjectForAccessibilityProperties):
        * inspector/InspectorDOMAgent.h:
        * inspector/protocol/DOM.json:

2014-02-10  Oliver Hunt  <oliver@apple.com>

        Stop throwing when attempting to read instance properties directly from the prototype
        https://bugs.webkit.org/show_bug.cgi?id=128568

        Reviewed by Mark Lam.

        A number of websites expect to be able to access instance properties
        directly through the prototype.  This is broken behavior but if we
        throw an exception the entire site breaks.  This patch simply makes us
        return undefined when reading, and logs the error to the console.

        * bindings/scripts/CodeGeneratorJS.pm:
        (GenerateImplementation):
        * bindings/scripts/test/JS/JSTestActiveDOMObject.cpp:
        (WebCore::jsTestActiveDOMObjectExcitingAttr):
        * bindings/scripts/test/JS/JSTestEventConstructor.cpp:
        (WebCore::jsTestEventConstructorAttr1):
        (WebCore::jsTestEventConstructorAttr2):
        * bindings/scripts/test/JS/JSTestException.cpp:
        (WebCore::jsTestExceptionName):
        * bindings/scripts/test/JS/JSTestInterface.cpp:
        (WebCore::jsTestInterfaceImplementsStr1):
        (WebCore::jsTestInterfaceImplementsStr2):
        (WebCore::jsTestInterfaceImplementsStr3):
        (WebCore::jsTestInterfaceImplementsNode):
        (WebCore::jsTestInterfaceSupplementalStr1):
        (WebCore::jsTestInterfaceSupplementalStr2):
        (WebCore::jsTestInterfaceSupplementalStr3):
        (WebCore::jsTestInterfaceSupplementalNode):
        * bindings/scripts/test/JS/JSTestObj.cpp:
        (WebCore::jsTestObjReadOnlyLongAttr):
        (WebCore::jsTestObjReadOnlyStringAttr):
        (WebCore::jsTestObjReadOnlyTestObjAttr):
        (WebCore::jsTestObjConstructorTestSubObj):
        (WebCore::jsTestObjTestSubObjEnabledBySettingConstructor):
        (WebCore::jsTestObjEnumAttr):
        (WebCore::jsTestObjByteAttr):
        (WebCore::jsTestObjOctetAttr):
        (WebCore::jsTestObjShortAttr):
        (WebCore::jsTestObjUnsignedShortAttr):
        (WebCore::jsTestObjLongAttr):
        (WebCore::jsTestObjLongLongAttr):
        (WebCore::jsTestObjUnsignedLongLongAttr):
        (WebCore::jsTestObjStringAttr):
        (WebCore::jsTestObjTestObjAttr):
        (WebCore::jsTestObjXMLObjAttr):
        (WebCore::jsTestObjCreate):
        (WebCore::jsTestObjReflectedStringAttr):
        (WebCore::jsTestObjReflectedIntegralAttr):
        (WebCore::jsTestObjReflectedUnsignedIntegralAttr):
        (WebCore::jsTestObjReflectedBooleanAttr):
        (WebCore::jsTestObjReflectedURLAttr):
        (WebCore::jsTestObjReflectedCustomIntegralAttr):
        (WebCore::jsTestObjReflectedCustomBooleanAttr):
        (WebCore::jsTestObjReflectedCustomURLAttr):
        (WebCore::jsTestObjTypedArrayAttr):
        (WebCore::jsTestObjAttrWithGetterException):
        (WebCore::jsTestObjAttrWithSetterException):
        (WebCore::jsTestObjStringAttrWithGetterException):
        (WebCore::jsTestObjStringAttrWithSetterException):
        (WebCore::jsTestObjCustomAttr):
        (WebCore::jsTestObjWithScriptStateAttribute):
        (WebCore::jsTestObjWithScriptExecutionContextAttribute):
        (WebCore::jsTestObjWithScriptStateAttributeRaises):
        (WebCore::jsTestObjWithScriptExecutionContextAttributeRaises):
        (WebCore::jsTestObjWithScriptExecutionContextAndScriptStateAttribute):
        (WebCore::jsTestObjWithScriptExecutionContextAndScriptStateAttributeRaises):
        (WebCore::jsTestObjWithScriptExecutionContextAndScriptStateWithSpacesAttribute):
        (WebCore::jsTestObjWithScriptArgumentsAndCallStackAttribute):
        (WebCore::jsTestObjConditionalAttr1):
        (WebCore::jsTestObjConditionalAttr2):
        (WebCore::jsTestObjConditionalAttr3):
        (WebCore::jsTestObjConditionalAttr4Constructor):
        (WebCore::jsTestObjConditionalAttr5Constructor):
        (WebCore::jsTestObjConditionalAttr6Constructor):
        (WebCore::jsTestObjCachedAttribute1):
        (WebCore::jsTestObjCachedAttribute2):
        (WebCore::jsTestObjAnyAttribute):
        (WebCore::jsTestObjContentDocument):
        (WebCore::jsTestObjMutablePoint):
        (WebCore::jsTestObjImmutablePoint):
        (WebCore::jsTestObjStrawberry):
        (WebCore::jsTestObjStrictFloat):
        (WebCore::jsTestObjDescription):
        (WebCore::jsTestObjId):
        (WebCore::jsTestObjHash):
        (WebCore::jsTestObjReplaceableAttribute):
        (WebCore::jsTestObjNullableDoubleAttribute):
        (WebCore::jsTestObjNullableLongAttribute):
        (WebCore::jsTestObjNullableBooleanAttribute):
        (WebCore::jsTestObjNullableStringAttribute):
        (WebCore::jsTestObjNullableLongSettableAttribute):
        (WebCore::jsTestObjNullableStringValue):
        (WebCore::jsTestObjAttribute):
        (WebCore::jsTestObjAttributeWithReservedEnumType):
        * bindings/scripts/test/JS/JSTestSerializedScriptValueInterface.cpp:
        (WebCore::jsTestSerializedScriptValueInterfaceValue):
        (WebCore::jsTestSerializedScriptValueInterfaceReadonlyValue):
        (WebCore::jsTestSerializedScriptValueInterfaceCachedValue):
        (WebCore::jsTestSerializedScriptValueInterfacePorts):
        (WebCore::jsTestSerializedScriptValueInterfaceCachedReadonlyValue):
        * bindings/scripts/test/JS/JSTestTypedefs.cpp:
        (WebCore::jsTestTypedefsUnsignedLongLongAttr):
        (WebCore::jsTestTypedefsImmutableSerializedScriptValue):
        (WebCore::jsTestTypedefsConstructorTestSubObj):
        (WebCore::jsTestTypedefsAttrWithGetterException):
        (WebCore::jsTestTypedefsAttrWithSetterException):
        (WebCore::jsTestTypedefsStringAttrWithGetterException):
        (WebCore::jsTestTypedefsStringAttrWithSetterException):
        * bindings/scripts/test/JS/JSattribute.cpp:
        (WebCore::jsattributeReadonly):

2014-02-11  Sergio Villar Senin  <svillar@igalia.com>

        [CSS Grid Layout] Support calc() breadth size type
        https://bugs.webkit.org/show_bug.cgi?id=103761

        Reviewed by Dean Jackson.

        We now support using calc() in
        -webkit-grid-template-{columns|rows}. This means that we now
        match the specification and support all the possible ways to
        specify the track breadths.

        This includes a fix to CSSCalculationValue that was converting
        Length values to CalcExpressionNumber and thus removing all the
        info regarding length units (px, em, pt...). That's why things
        like calc(10px + 20%) were shown as calc(10 + 20%) in
        getComputedStyle() output.

        * css/CSSCalculationValue.cpp:
        (WebCore::CSSCalcPrimitiveValue::toCalcValue):
        * css/CSSComputedStyleDeclaration.cpp:
        (WebCore::zoomAdjustedPixelValueForLength):
        (WebCore::specifiedValueForGridTrackBreadth):
        (WebCore::specifiedValueForGridTrackSize):
        * css/StyleResolver.cpp:
        (WebCore::createGridTrackBreadth):
        * rendering/RenderGrid.cpp:
        (WebCore::RenderGrid::computeUsedBreadthOfMinLength):
        (WebCore::RenderGrid::computeUsedBreadthOfMaxLength):
        (WebCore::RenderGrid::computeUsedBreadthOfSpecifiedLength):

2014-02-11  Brent Fulgham  <bfulgham@apple.com>

        Correct some uses of 'auto'
        https://bugs.webkit.org/show_bug.cgi?id=128578

        Reviewed by Anders Carlsson.

        Correct the following instances of "for (auto ...)" to use reference or
        const references to avoid copies.

        * accessibility/AccessibilityObject.cpp:
        (WebCore::AccessibilityObject::elementsFromAttribute):
        * css/CSSGridTemplateValue.cpp:
        (WebCore::stringForPosition):
        * dom/Node.cpp:
        (WebCore::NodeListsNodeData::invalidateCaches):
        * inspector/PageInjectedScriptManager.cpp:
        (WebCore::PageInjectedScriptManager::discardInjectedScriptsFor):
        * page/WheelEventDeltaTracker.cpp:
        (WebCore::WheelEventDeltaTracker::dominantScrollGestureDirection):
        * page/scrolling/ScrollingCoordinator.cpp:
        (WebCore::ScrollingCoordinator::computeNonFastScrollableRegion):
        * platform/graphics/mac/WebLayer.mm:
        (WebCore::drawLayerContents):
        * platform/ios/TileGrid.mm:
        (WebCore::TileGrid::dropTilesBetweenRects):
        (WebCore::TileGrid::dropDistantTiles):
        (WebCore::TileGrid::dropInvalidTiles):
        * rendering/InlineTextBox.cpp:
        (WebCore::translateIntersectionPointsToSkipInkBoundaries):
        * testing/InternalSettings.cpp:
        (WebCore::InternalSettings::Backup::restoreTo):

2014-02-10  David Hyatt  <hyatt@apple.com>

        [New Multicolumn] Make columns work with line grids
        https://bugs.webkit.org/show_bug.cgi?id=128555

        Reviewed by Antti Koivisto.

        Added line-snap-into-columns.html and line-snap-inside-columns.html

        * rendering/LayoutState.cpp:
        (WebCore::LayoutState::LayoutState):
        * rendering/LayoutState.h:
        (WebCore::LayoutState::pageOffset):
        (WebCore::LayoutState::setLineGridPaginationOrigin):
        Change the LayoutState to call into the renderer to compute the line
        grid pagination origin instead of doing it directly. Added the
        appropriate getters and setters to enable the renderer to do this.

        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::computeLineGridPaginationOrigin):
         * rendering/RenderBlock.h:
        An implementation for the old multi-column code. The logic is the
        same with tweaks made now that the method is on the renderer instead.

        * rendering/RenderBlockFlow.cpp:
        (WebCore::RenderBlockFlow::pageLogicalTopForOffset):
        Fix a bug with flow threads and pageLogicalTopForOffset. Normal CSS region-based
        flow threads are never embedded in an enclosing pagination context, so they
        didn't add in the firstPageLogicalTop (or subtract it when computing the region
        hit). Multi-column flow threads do need to subtract out the firstPageLogicalTop,
        since it can occur somewhere on the page.
        
        Ultimately regions will need to get smarter here too if they want to work with
        line grids, but for now I just fixed multi-column.

        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::isUnsplittableForPagination):
        Undo this change, since it prevents the inheritance of line grids into the
        multi-column layout.

        * rendering/RenderMultiColumnFlowThread.cpp:
        (WebCore::RenderMultiColumnFlowThread::computeLineGridPaginationOrigin):
        * rendering/RenderMultiColumnFlowThread.h:
        An implementation for the new multi-column code. The logic is the
        same as the old code. Code duplication is ok, since the old multi-column method
        in RenderBlock will just be deleted once the new code is turned on, and it's easier
        not to intertwine them.

2014-02-11  Radu Stavila  <stavila@adobe.com>

        [CSS Regions] Overflow above the first region is not properly painted for regions with padding
        https://bugs.webkit.org/show_bug.cgi?id=128590

        Reviewed by Andrei Bucur.

        Painting is done using the layer of the region's container, so offsetting using the content box
        of the region itself is incorrect. 

        Test: fast/regions/region-padding-overflow-hidden.html

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::mapLayerClipRectsToFragmentationLayer):
        * rendering/RenderRegion.cpp:
        (WebCore::RenderRegion::regionContainer):
        (WebCore::RenderRegion::regionContainerLayer):
        * rendering/RenderRegion.h:

2014-02-11  Benjamin Poulain  <benjamin@webkit.org>

        querySelector() does not use the compiler correctly
        https://bugs.webkit.org/show_bug.cgi?id=128588

        Reviewed by Andreas Kling.

        * dom/SelectorQuery.cpp:
        (WebCore::SelectorDataList::execute):
        I messed up the refactoring when I landed SelectorQuery. The compiled
        code was not used the first time through SelectorDataList::execute.

2014-02-11  Piotr Grad  <p.grad@samsung.com>

        [GStreamer] High playback rate causes crash
        https://bugs.webkit.org/show_bug.cgi?id=128453

        Reviewed by Philippe Normand.

        To high playback rate passed to GStreamer was causing crash.
        Added guard in setRate method.

        Test: media/video-extreme-playbackrate-crash.html

        * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
        (WebCore::MediaPlayerPrivateGStreamer::setRate):

2014-02-10  Brady Eidson  <beidson@apple.com>

        IndexedDB assertion at IDBTransactionBackend::~IDBTransactionBackend()
        https://bugs.webkit.org/show_bug.cgi?id=128341

        Reviewed by Maciej Stachowiak.

        * Modules/indexeddb/IDBTransactionBackend.cpp:
        (WebCore::IDBTransactionBackend::~IDBTransactionBackend): "Finished" is not the only
          valid value here - In WK2, "Unused" is also valid. Update the ASSERT.

2014-02-10  Alexey Proskuryakov  <ap@apple.com>

        Add hooks for wrapping CryptoKeys in SerializedScriptValue
        https://bugs.webkit.org/show_bug.cgi?id=128567

        Reviewed by Anders Carlsson.

        * bindings/js/SerializedScriptValue.cpp: Changed SerializedScriptValue to serialize
        wrapped keys. Added a version number to crypto key serialization.

        * dom/Document.cpp:
        (WebCore::Document::wrapCryptoKey):
        (WebCore::Document::unwrapCryptoKey):
        * dom/Document.h:
        * dom/ScriptExecutionContext.h:
        * page/ChromeClient.h:
        (WebCore::ChromeClient::wrapCryptoKey):
        (WebCore::ChromeClient::unwrapCryptoKey):
        Hand wrapping/unwrapping over to client code.

        * workers/WorkerGlobalScope.cpp:
        (WebCore::WorkerGlobalScope::wrapCryptoKey):
        (WebCore::WorkerGlobalScope::unwrapCryptoKey):
        * workers/WorkerGlobalScope.h:
        Not implemented in workers. SubtleCrypto is currently not exposed in workers. It used
        to be possible in WebKit implementation to post a CryptoKey to a worker anyway,
        but this doesn't work any more.

2014-02-10  ChangSeok Oh  <changseok.oh@collabora.com>

        Support ANGLE_instanced_arrays for linux
        https://bugs.webkit.org/show_bug.cgi?id=127465

        Reviewed by Martin Robinson.

        Support the instanced drawing WebGL extension, ANGLE_instanced_arrays for linux platform.
        This is a following patch for r162565. Relevant opengl APIs are exposed
        for WebGLRenderingContext to access them.

        Covered by fast/canvas/webgl/angle-instanced-arrays.html

        * html/canvas/ANGLEInstancedArrays.cpp:
        (WebCore::ANGLEInstancedArrays::supported):
        * platform/graphics/OpenGLShims.cpp:
        (WebCore::initializeOpenGLShims):
        * platform/graphics/OpenGLShims.h:
        * platform/graphics/opengl/Extensions3DOpenGL.cpp:
        (WebCore::Extensions3DOpenGL::supportsExtension):
        * platform/graphics/opengl/GraphicsContext3DOpenGL.cpp:
        (WebCore::GraphicsContext3D::drawArraysInstanced):
        (WebCore::GraphicsContext3D::drawElementsInstanced):
        (WebCore::GraphicsContext3D::vertexAttribDivisor):

2014-02-10  Mark Lam  <mark.lam@apple.com>

        Removing limitation on JSLock’s lockDropDepth.
        <https://webkit.org/b/128570>

        Reviewed by Geoffrey Garen.

        No new tests.

        * bindings/js/PageScriptDebugServer.cpp:
        (WebCore::PageScriptDebugServer::runEventLoopWhilePaused):
        * platform/ios/wak/WebCoreThread.mm:
        (SendDelegateMessage):
        (WebThreadRunOnMainThread):
        - No longer need to specify AlwaysDropLocks, because DropAllLocks is
          now always unconditional.

2014-02-10  Benjamin Poulain  <benjamin@webkit.org>

        Clean up MarkupAccumulator::appendCharactersReplacingEntities
        https://bugs.webkit.org/show_bug.cgi?id=128440

        Reviewed by Ryosuke Niwa.

        Some cleanup:
        -Remove the copied code for the loops. Instead, use a template with the loop
         parametrized by the characters type.
        -Move EntityDescription from the header to the implementation.
        -Make the 5 substitution strings compile-time literals. Replacement is not hot enough
         to justify the static here.

        * editing/MarkupAccumulator.cpp:
        (WebCore::appendCharactersReplacingEntitiesInternal):
        (WebCore::MarkupAccumulator::appendCharactersReplacingEntities):
        * editing/MarkupAccumulator.h:

2014-02-10  Benjamin Poulain  <benjamin@webkit.org>

        Add a few pseudo type to the selector compiler through function calls
        https://bugs.webkit.org/show_bug.cgi?id=128514

        Reviewed by Dean Jackson.

        Certain pseudo type checkers rely on virtual function calls. Since those types
        are uncommon, and this code generator cannot currently make virtual function calls,
        the checker are added by generating a function call to a wrapper function.

        To avoid code duplication between SelectorChecker and the compiler, all the relevant checking
        code has been moved to a common header, SelectorCheckerTestFunctions.h.
        SelectorChecker still inline the functions, while the SelectorCompiler generate local static functions
        and generate calls to them as needed.

        * GNUmakefile.list.am:
        * WebCore.vcxproj/WebCore.vcxproj:
        * WebCore.vcxproj/WebCore.vcxproj.filters:
        * WebCore.xcodeproj/project.pbxproj:
        * css/SelectorChecker.cpp:
        (WebCore::SelectorChecker::checkOne):
        * css/SelectorCheckerTestFunctions.h: Added.
        (WebCore::isAutofilled):
        (WebCore::isDefaultButtonForForm):
        (WebCore::isDisabled):
        (WebCore::isEnabled):
        (WebCore::isChecked):
        (WebCore::isInvalid):
        (WebCore::isOptionalFormControl):
        (WebCore::isRequiredFormControl):
        (WebCore::isValid):
        (WebCore::matchesReadOnlyPseudoClass):
        (WebCore::matchesReadWritePseudoClass):
        (WebCore::shouldAppearIndeterminate):
        (WebCore::matchesFullScreenPseudoClass):
        (WebCore::matchesFutureCuePseudoClass):
        (WebCore::matchesPastCuePseudoClass):
        * cssjit/SelectorCompiler.cpp:
        (WebCore::SelectorCompiler::addPseudoType):
        (WebCore::SelectorCompiler::nonConstTestFunctionWrapper):
        (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementMatching):
        (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementFunctionCallTest):
        * dom/Element.h:
        (WebCore::Element::isValidFormControlElement):
        * dom/Node.h:
        (WebCore::Node::toInputElement):
        * html/HTMLFormControlElement.cpp:
        (WebCore::HTMLFormControlElement::isValidFormControlElement):
        * html/HTMLFormControlElement.h:

2014-02-10  Benjamin Poulain  <benjamin@webkit.org>

        Add the basic infrastructure to compile attributes matching in selectors
        https://bugs.webkit.org/show_bug.cgi?id=128484

        Reviewed by Gavin Barraclough.

        Tests: fast/selectors/querySelector-attribute-match-with-child-backtracking.html
               fast/selectors/querySelector-long-attribute-match-with-child-backtracking.html
               fast/selectors/querySelector-long-multiple-attribute-match-with-child-backtracking.html
               fast/selectors/querySelector-multiple-attribute-match-with-child-backtracking.html

        Add the infrastructure to match attributes. This only add basic support for the 'Set' match
        type, the other match types will have to be built on top.

        * cssjit/RegisterAllocator.h:
        (WebCore::RegisterAllocator::availableRegisterCount):
        (WebCore::RegisterAllocator::allocateRegister):
        (WebCore::RegisterAllocator::deallocateRegister):
        (WebCore::RegisterAllocator::reserveCalleeSavedRegisters):
        (WebCore::RegisterAllocator::restoreCalleeSavedRegisters):
        (WebCore::RegisterAllocator::allocatedRegisters):
        (WebCore::RegisterAllocator::RegisterAllocator):
        (WebCore::RegisterAllocator::~RegisterAllocator):
        In the worst case, matching attributes can take up to 10 registers. On x86_64, we have
        8 caller-saved registers. The extra 2 registers are simply added by taking callee-saved registers.

        RegisterAllocator is modified to know support saving and restoring callee saved registers.
        The list of available registers is changed from a vector to a HashSet because the registers
        are removed from arbitrary locations in restoreCalleeSavedRegisters(). The m_allocatedRegisters
        remain a vector since the allocation/release pattern remain ordered.

        * cssjit/SelectorCompiler.cpp:
        (WebCore::SelectorCompiler::SelectorCodeGenerator::SelectorCodeGenerator):

        (WebCore::SelectorCompiler::minimumRegisterRequirements):
        This new utility finds the minimum number of registers needed to compile the input. In most
        cases we have plenty enough. In rare cases we need to saved a few registers to the stack.

        (WebCore::SelectorCompiler::SelectorCodeGenerator::compile):
        (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementDataMatching):
        (WebCore::SelectorCompiler::testIsHTMLFlagOnNode):
        (WebCore::SelectorCompiler::canMatchStyleAttribute):

        (WebCore::SelectorCompiler::SelectorCodeGenerator::generateSynchronizeStyleAttribute):
        (WebCore::SelectorCompiler::SelectorCodeGenerator::generateSynchronizeAllAnimatedSVGAttribute):
        The style attribute and certain SVG attributes can be modified lazily. In those cases,
        the element needs to be updated before querying the attributes.

        (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementAttributesMatching):
        (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementAttributeMatching):
        Generate the attribute matching. For each CSSSelector matching an attribute, we generate
        a loop over all the attributes of the element, maching the particular CSSSelector.
        This makes no attempt at grouping related queries since those do not seem to happen a lot
        in practice.

        * dom/Attribute.h:
        (WebCore::Attribute::nameMemoryOffset):
        (WebCore::Attribute::nameMatchesFilter):
        (WebCore::Attribute::matches):
        * dom/ElementData.h:
        (WebCore::ElementData::isUniqueFlag):
        (WebCore::ElementData::arraySizeAndFlagsMemoryOffset):
        (WebCore::ElementData::styleAttributeIsDirtyFlag):
        (WebCore::ElementData::animatedSVGAttributesAreDirtyFlag):
        (WebCore::ElementData::arraySizeOffset):
        (WebCore::ShareableElementData::attributeArrayMemoryOffset):
        (WebCore::UniqueElementData::attributeVectorMemoryOffset):
        * dom/Node.h:
        (WebCore::Node::flagIsHTML):
        * dom/StyledElement.cpp:
        (WebCore::StyledElement::synchronizeStyleAttributeInternal):
        * dom/StyledElement.h:
        (WebCore::StyledElement::synchronizeStyleAttributeInternal):
        * svg/SVGElement.cpp:
        (WebCore::SVGElement::synchronizeAllAnimatedSVGAttribute):
        (WebCore::SVGElement::synchronizeAnimatedSVGAttribute):
        * svg/SVGElement.h:

2014-02-10  Brady Eidson  <beidson@apple.com>

        IDB: storage/indexeddb/mozilla/indexes.html fails
        <rdar://problem/16031590> and https://bugs.webkit.org/show_bug.cgi?id=128563

        Reviewed by Tim Horton.

        Tested by storage/indexeddb/mozilla/indexes.html (and probably others)

        * Modules/indexeddb/IDBGetResult.h:
        (WebCore::IDBGetResult::IDBGetResult): Add a constructor that takes an IDBKeyData argument.

        * Modules/indexeddb/IDBRequest.cpp:
        (WebCore::IDBRequest::onSuccess): If there’s no keyPath then skip the injection step.

2014-02-10  Benjamin Poulain  <benjamin@webkit.org>

        Speed up DatasetDOMStringMap::item() when the element has multiple attributes
        https://bugs.webkit.org/show_bug.cgi?id=128058

        Reviewed by Darin Adler.

        Accessing data attributes by name through DatasetDOMStringMap involes the conversion
        from JavaScript property name to attribute name (done with propertyNameMatchesAttributeName()).

        When there is a single data attribute, that method is efficient. When there are several attributes,
        comparing names character by character becomes a bottleneck.

        This patch add an efficent path for this case: instead of converting the attribute name on the fly,
        the JavaScript property name is converted to an attribute name so that it can compared by its
        AtomicStringImpl pointer.

        This method puts a lot more pressure on convertPropertyNameToAttributeName()'s speed. The method was
        improved accordingly to compensate for its new caller.

        When enumerating multiple attributes by name, this patch provides about 80% speedup.
        I could not measure any slow down on the simple cases.

        * dom/DatasetDOMStringMap.cpp:
        (WebCore::convertPropertyNameToAttributeName):
        (WebCore::DatasetDOMStringMap::item):
        * dom/ElementData.h:
        (WebCore::AttributeIteratorAccessor::attributeCount):

2014-02-10  Ryosuke Niwa  <rniwa@webkit.org>

        Address the review comments for r163825.

        * html/HTMLTextFormControlElement.cpp:
        (WebCore::HTMLTextFormControlElement::indexForVisiblePosition):
        (WebCore::positionForIndex):

2014-02-10  Simon Fraser  <simon.fraser@apple.com>

        Try to fix the 32-bit build.

        * WebCore.exp.in:

2014-02-10  Filip Pizlo  <fpizlo@apple.com>

        Rename Operations.h to JSCInlines.h
        https://bugs.webkit.org/show_bug.cgi?id=128543

        Rubber stamped by Geoffrey Garen.

        No new tests because no change in behavior.

        * ForwardingHeaders/runtime/JSCInlines.h: Added.
        * bindings/js/JSCryptoAlgorithmBuilder.cpp:
        * bindings/js/JSCryptoKeySerializationJWK.cpp:
        * bindings/js/JSCustomXPathNSResolver.h:
        * bindings/js/JSDOMBinding.h:
        * bindings/js/JSDOMGlobalObject.h:
        * bindings/js/JSDictionary.h:
        * bindings/js/JSMessagePortCustom.cpp:
        * bindings/js/JSMessagePortCustom.h:
        * bindings/js/JSNodeFilterCondition.h:
        * bindings/js/SerializedScriptValue.cpp:
        * bindings/js/WebCoreTypedArrayController.cpp:
        * bridge/c/c_utility.h:
        * bridge/jsc/BridgeJSC.h:
        * dom/CustomEvent.cpp:
        * dom/Node.cpp:
        * html/HTMLCanvasElement.cpp:
        * html/HTMLImageLoader.cpp:
        * html/canvas/WebGLRenderingContext.cpp:
        * inspector/InspectorDOMAgent.cpp:
        * inspector/WebConsoleAgent.cpp:
        * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
        * platform/graphics/filters/FEGaussianBlur.cpp:
        * platform/graphics/filters/FilterEffect.cpp:
        * testing/MockCDM.cpp:
        * xml/XMLHttpRequest.cpp:

2014-02-09  Dean Jackson  <dino@apple.com>

        Update aspect-ratio property to have constraining keywords
        https://bugs.webkit.org/show_bug.cgi?id=128262

        Reviewed by Simon Fraser.

        Add support for "from-dimensions" and "from-intrinsic"
        property values to "-webkit-aspect-ratio". I also changed
        the default value from "none" to "auto", because "none" doesn't
        make sense any more.

        Covered by enhancing existing tests.

        * css/CSSComputedStyleDeclaration.cpp:
        (WebCore::ComputedStyleExtractor::propertyValue): New keywords.
        * css/CSSParser.cpp:
        (WebCore::CSSParser::parseAspectRatio): Support new keywords.
        * css/CSSValueKeywords.in: Add from-dimensions and from-intrinsic.
        * css/DeprecatedStyleBuilder.cpp: This now has to handle the new
        values. I also changed "none" to "auto".
        (WebCore::ApplyPropertyAspectRatio::applyInheritValue):
        (WebCore::ApplyPropertyAspectRatio::applyInitialValue):
        (WebCore::ApplyPropertyAspectRatio::applyValue):
        * rendering/style/RenderStyle.h: hasAspectRatio is now aspectRatioType
        and indicates one of the three keywords, or a specified number.
        * rendering/style/RenderStyleConstants.h: New enum.
        * rendering/style/StyleRareNonInheritedData.cpp:
        (WebCore::StyleRareNonInheritedData::StyleRareNonInheritedData):
        (WebCore::StyleRareNonInheritedData::operator==):
        * rendering/style/StyleRareNonInheritedData.h:

2014-02-10  Benjamin Poulain  <bpoulain@apple.com>

        [WK2] Add support for image document viewport configuration
        https://bugs.webkit.org/show_bug.cgi?id=128565

        Reviewed by Simon Fraser.

        Add proper default settings for displaying images.

        * WebCore.exp.in:
        * page/ViewportConfiguration.cpp:
        (WebCore::ViewportConfiguration::imageDocumentParameters):
        (WebCore::ViewportConfiguration::textDocumentParameters):
        * page/ViewportConfiguration.h:

2014-02-10  Brendan Long  <b.long@cablelabs.com>

        Unreviewed GTK build fix after r163816.

        * GNUmakefile.list.am: Remove HTMLMediaSource.h and HTMLMediaSource.cpp

2014-02-10  Jinwoo Song  <jinwoo7.song@samsung.com>

        Unreviewed EFL WebKit2 build fix after r163816.

        * CMakeLists.txt: Remove HTMLMediaSource.cpp

2014-02-10  Ryosuke Niwa  <rniwa@webkit.org>

        HTMLTextFormControlElement::setSelectionRange shouldn't use VisiblePosition
        https://bugs.webkit.org/show_bug.cgi?id=128478

        Reviewed by Darin Adler.

        Added positionForIndex to compute Position given a selection index. This function doesn't
        synchronously trigger like visiblePositionForIndex.

        Also added assertions in visiblePositionForIndex and visiblePositionForIndex to make sure
        they are inverse of one another.

        * html/HTMLTextFormControlElement.cpp:
        (WebCore::HTMLTextFormControlElement::setSelectionRange): Use positionForIndex. Also removed
        the now tautological assertions since we would never create a position outside the inner text
        element.

        (WebCore::HTMLTextFormControlElement::indexForVisiblePosition): Fixed the bug where this
        function could return a selection index beyond innerTextElement in some types of input
        elements such as search fields. fast/forms/search-disabled-readonly.html hits the newly
        added assertion without this change. Note we can't use visiblePositionForIndex here as that
        would result in a mutual recursion with the assertion in visiblePositionForIndex.

        (WebCore::HTMLTextFormControlElement::visiblePositionForIndex): Added an assertion.

        (WebCore::positionForIndex): Added. It's here with prototype at the beginning of the file
        so that it's right next to HTMLTextFormControlElement::innerText() where we do a similar
        DOM traversal to obtain the inner text value.

2014-02-07  Jeffrey Pfau  <jpfau@apple.com>

        Disable access to application cache when in private browsing
        https://bugs.webkit.org/show_bug.cgi?id=128426

        Reviewed by Alexey Proskuryakov.

        Tests: http/tests/security/appcache-in-private-browsing.html
               http/tests/security/appcache-switching-private-browsing.html

        * loader/appcache/ApplicationCacheGroup.cpp:
        (WebCore::ApplicationCacheGroup::cacheForMainRequest):
        (WebCore::ApplicationCacheGroup::selectCache):
        (WebCore::ApplicationCacheGroup::selectCacheWithoutManifestURL):
        (WebCore::ApplicationCacheGroup::update):
        * loader/appcache/ApplicationCacheHost.cpp:
        (WebCore::ApplicationCacheHost::isApplicationCacheEnabled):

2014-02-10  Jer Noble  <jer.noble@apple.com>

        [MSE] Fix layering violations in MediaSource
        https://bugs.webkit.org/show_bug.cgi?id=128546

        Reviewed by Eric Carlson.

        Code in Modules should be considered part of html/ and should have the
        same layering properties. Get rid of HTMLMediaSource (which was intended
        to allow Modules/mediasource to be considered part of platform/) and add
        a new client interface allowing communication from platform/ -> 
        Modules/mediasource.

        Replace HTMLMediaSource -> MediaSourcePrivateClient:
        * html/HTMLMediaSource.cpp: Removed.
        * html/HTMLMediaSource.h: Removed.
        * platform/graphics/MediaSourcePrivateClient.h: Added.
        (WebCore::MediaSourcePrivateClient::~MediaSourcePrivateClient):

        Move registry support from HTMLMediaSource -> MediaSource.
        * Modules/mediasource/MediaSource.cpp:
        (WebCore::MediaSource::setRegistry): Moved from HTMLMediaSource.cpp.
        * Modules/mediasource/MediaSource.h:
        (WebCore::MediaSource::lookup): Ditto.

        Update references to HTMLMediaSource -> MediaSource:
        * Modules/mediasource/MediaSourceRegistry.cpp:
        (WebCore::MediaSourceRegistry::MediaSourceRegistry): 
        * WebCore.xcodeproj/project.pbxproj:
        * html/HTMLMediaElement.cpp:
        (WebCore::HTMLMediaElement::loadResource):
        * html/HTMLMediaElement.h:

        Update references to HTMLMediaSource -> MediaSourcePrivateClient:
        * platform/graphics/MediaPlayer.cpp:
        (WebCore::NullMediaPlayerPrivate::load):
        (WebCore::MediaPlayer::load): Ditto.
        (WebCore::MediaPlayer::loadWithNextMediaEngine):
        * platform/graphics/MediaPlayer.h:
        * platform/graphics/MediaPlayerPrivate.h:
        * platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.cpp:
        (WebCore::MediaPlayerPrivateAVFoundation::load):
        * platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.h:
        * platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaSourceAVFObjC.h:
        * platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaSourceAVFObjC.mm:
        (WebCore::MediaPlayerPrivateMediaSourceAVFObjC::load):
        * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
        (WebCore::MediaPlayerPrivateGStreamer::load):
        * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:
        * platform/graphics/gstreamer/MediaSourceGStreamer.cpp:
        (WebCore::MediaSourceGStreamer::open):
        * platform/graphics/gstreamer/MediaSourceGStreamer.h:
        * platform/graphics/ios/MediaPlayerPrivateIOS.h:
        * platform/graphics/mac/MediaPlayerPrivateQTKit.h:
        * platform/graphics/mac/MediaPlayerPrivateQTKit.mm:
        (WebCore::MediaPlayerPrivateQTKit::load):
        * platform/mock/mediasource/MockMediaPlayerMediaSource.cpp:
        (WebCore::MockMediaPlayerMediaSource::load):
        * platform/mock/mediasource/MockMediaPlayerMediaSource.h:

2014-02-10  Alexey Proskuryakov  <ap@apple.com>

        Remove some unused functions from SerializedScriptValue
        https://bugs.webkit.org/show_bug.cgi?id=128407

        Reviewed by Oliver Hunt.

        Removed functions that used Deprecated::ScriptValue

        * Modules/indexeddb/IDBObjectStore.cpp:
        (WebCore::IDBObjectStore::put):
        * bindings/js/IDBBindingUtilities.cpp:
        (WebCore::deserializeIDBValue):
        (WebCore::deserializeIDBValueBuffer):
        * bindings/js/SerializedScriptValue.cpp:
        * bindings/js/SerializedScriptValue.h:

2014-02-10  Roger Fong  <roger_fong@apple.com>

        [Windows] Unreviewed test fix.

        * platform/graphics/cg/GraphicsContextCG.cpp: m_pixelSnappingFactor was not set to 1.
        (WebCore::GraphicsContext::platformInit):

2014-02-10  Carlos Garcia Campos  <cgarcia@igalia.com>

        [GLIB] Add GUniqueOutPtr and use it instead of GOwnPtr
        https://bugs.webkit.org/show_bug.cgi?id=127554

        Reviewed by Gustavo Noronha Silva.

        Use GUniqueOutPtr instead of GOwnPtr.

        * GNUmakefile.list.am:
        * PlatformEfl.cmake:
        * PlatformGTK.cmake:
        * platform/audio/gstreamer/AudioDestinationGStreamer.cpp:
        (WebCore::AudioDestinationGStreamer::handleMessage):
        * platform/audio/gstreamer/AudioFileReaderGStreamer.cpp:
        (WebCore::AudioFileReader::handleMessage):
        * platform/glib/BatteryProviderUPower.cpp:
        (BatteryProviderUPower::startUpdating):
        * platform/graphics/gstreamer/GStreamerUtilities.cpp:
        (WebCore::initializeGStreamer):
        * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
        (WebCore::MediaPlayerPrivateGStreamer::handleMessage):
        * platform/graphics/gstreamer/TrackPrivateBaseGStreamer.cpp:
        (WebCore::TrackPrivateBaseGStreamer::getTag):
        * platform/graphics/gtk/ImageBufferGtk.cpp:
        (WebCore::encodeImage):
        (WebCore::ImageBuffer::toDataURL):
        * platform/graphics/gtk/ImageGtk.cpp:
        (WebCore::loadResourceSharedBuffer):
        * platform/gtk/DataObjectGtk.cpp:
        (WebCore::DataObjectGtk::setURIList):
        * platform/gtk/FileSystemGtk.cpp:
        (WebCore::readFromFile):
        * platform/gtk/GamepadsGtk.cpp:
        (WebCore::GamepadDeviceGtk::readCallback):
        * platform/gtk/GtkInputMethodFilter.cpp:
        (WebCore::GtkInputMethodFilter::handlePreeditChanged):
        * platform/gtk/PasteboardHelper.cpp:
        (WebCore::PasteboardHelper::fillDataObjectFromDropData):
        * platform/gtk/RenderThemeGtk.cpp:
        (WebCore::RenderThemeGtk::systemFont):
        * platform/gtk/SharedBufferGtk.cpp:
        (WebCore::SharedBuffer::createWithContentsOfFile):
        * platform/network/gtk/CredentialBackingStore.cpp:
        (WebCore::credentialForChallengeAsyncReadyCallback):
        * platform/network/soup/GOwnPtrSoup.cpp: Removed.
        * platform/network/soup/GOwnPtrSoup.h: Removed.
        * platform/network/soup/ResourceHandleSoup.cpp:
        (WebCore::redirectSkipCallback):
        (WebCore::nextMultipartResponsePartCallback):
        (WebCore::sendRequestCallback):
        (WebCore::addFileToSoupMessageBody):
        (WebCore::createSoupRequestAndMessageForHandle):
        (WebCore::readCallback):
        * platform/network/soup/SocketStreamHandleSoup.cpp:
        (WebCore::SocketStreamHandle::platformSend):
        (WebCore::SocketStreamHandle::platformClose):
        (WebCore::connectedCallback):
        (WebCore::readReadyCallback):
        * platform/network/soup/SoupNetworkSession.cpp:
        (WebCore::SoupNetworkSession::httpProxy):

2014-02-10  Darin Adler  <darin@apple.com>

        Automatically generate isRendererOfType in RENDER_OBJECT_TYPE_CASTS
        https://bugs.webkit.org/show_bug.cgi?id=128520

        Reviewed by Andreas Kling.

        * rendering/RenderObject.h: Updated the RENDER_OBJECT_TYPE_CASTS macro so
        that it generates isRendererOfType specializations as well as type casts
        and also have the type casts use the isRendererOfType function. The macro
        also now uses an argument name of "renderer" and allows for a predicate
        that is more than just a single function call (see RenderTextFragment.h).

        * rendering/RenderBlock.h:
        * rendering/RenderBlockFlow.h:
        * rendering/RenderBox.h:
        * rendering/RenderBoxModelObject.h:
        * rendering/RenderCounter.h:
        * rendering/RenderElement.h:
        * rendering/RenderFrameSet.h:
        * rendering/RenderLayerModelObject.h:
        * rendering/RenderMenuList.h:
        * rendering/RenderNamedFlowThread.h:
        * rendering/RenderRubyRun.h:
        * rendering/RenderTableSection.h:
        * rendering/RenderText.h:
        * rendering/mathml/RenderMathMLBlock.h:
        * rendering/mathml/RenderMathMLOperator.h:
        * rendering/mathml/RenderMathMLToken.h:
        * rendering/svg/RenderSVGRoot.h:
        * rendering/svg/RenderSVGText.h:
        Removed specialization of isRendererOfType, since the RENDER_OBJECT_TYPE_CASTS
        macro handles that now. Also removed some unneeded semicolons.

        * rendering/RenderTextFragment.h: Use RENDER_OBJECT_TYPE_CASTS instead of
        hand written casts, now that it can handle a non-trivial predicate.

        * rendering/RenderCombineText.h:
        * rendering/RenderDetailsMarker.h:
        * rendering/RenderListMarker.h:
        * rendering/RenderVideo.h:
        * rendering/RenderView.h:
        * rendering/mathml/RenderMathMLFraction.h:
        * rendering/mathml/RenderMathMLScripts.h:
        * rendering/mathml/RenderMathMLSpace.h:
        Removed unneeded semicolons.

2014-02-10  Darin Adler  <darin@apple.com>

        Stop using String::deprecatedCharacters to call WTF::Collator
        https://bugs.webkit.org/show_bug.cgi?id=128517

        Reviewed by Alexey Proskuryakov.

        * xml/XSLTUnicodeSort.cpp:
        (WebCore::xsltUnicodeSortFunction): Create the collator in a single line using the
        new constructor that takes a shouldSortLowercaseFirst boolean. Use the new
        collateUTF8 function instead of upconverting UTF-8 strings to UTF-16 as the old code did.

2014-02-10  Changhun Kang  <temoochin@company100.net>

        Remove unnecessary comment lines.
        https://bugs.webkit.org/show_bug.cgi?id=127894

        Reviewed by Darin Adler.

        No new tests. No change in behavior.

        * Modules/websockets/WebSocketHandshake.cpp:
        (WebCore::WebSocketHandshake::clientHandshakeRequest):

2014-02-10  Peter Molnar  <pmolnar.u-szeged@partner.samsung.com>

        Fix EFL build with INSPECTOR disabled
        https://bugs.webkit.org/show_bug.cgi?id=125064

        Reviewed by Csaba Osztrogonác.

        * bindings/js/PageScriptDebugServer.cpp:
        * bindings/js/PageScriptDebugServer.h:
        * bindings/js/WorkerScriptDebugServer.cpp:
        * bindings/js/WorkerScriptDebugServer.h:
        * inspector/PageInjectedScriptManager.h:

2014-02-10  Zan Dobersek  <zdobersek@igalia.com>

        Fix a few mistakes that landed with r163749.

        Rubber-stamped by Carlos Garcia Campos.

        * platform/gtk/GtkTouchContextHelper.cpp:
        (WebCore::GtkTouchContextHelper::handleEvent): When handling GDK_TOUCH_BEGIN, the assertion
        should ensure that the sequence-to-event map doesn't yet contain pairs with the specific
        sequence as the key.

2014-02-10  David Kilzer  <ddkilzer@apple.com>

        Add type-safe casts for ContentData subclasses
        <http://webkit.org/b/128510>

        Reviewed by Darin Adler.

        * css/CSSComputedStyleDeclaration.cpp:
        (WebCore::contentToCSSValue):
        * css/StyleResolver.cpp:
        (WebCore::StyleResolver::loadPendingImages):
        * rendering/RenderElement.cpp:
        (WebCore::RenderElement::createFor):
        * rendering/style/RenderStyle.cpp:
        (WebCore::RenderStyle::setContent):
        - Switch to toFooContentData() methods.

        * rendering/style/ContentData.h:
        - Define toFooContentData() methods.
        - Extract FooContentData::equals() methods so that the
          toFooContentData() methods can be used.

2014-02-09  Dirk Schulze  <dschulze@chromium.org>

        -webkit-clip-path should support fill, stroke, view-box keywords
        https://bugs.webkit.org/show_bug.cgi?id=128393

        Reviewed by Dean Jackson.

        CSS Masking uses the keywords fill, stroke and view-box for SVG shapes.
        This patch updates the parser, style resolver and SVG implementation of
        the prefixed clip-path property.

        Remove the unexposed and obsolete keyword bounding-box.

        Tests: svg/clip-path/clip-path-shape-fill-expected.svg
               svg/clip-path/clip-path-shape-fill.svg
               svg/clip-path/clip-path-shape-stroke-expected.svg
               svg/clip-path/clip-path-shape-stroke.svg
               svg/clip-path/clip-path-shape-view-box-expected.svg
               svg/clip-path/clip-path-shape-view-box.svg

        * css/CSSParser.cpp: Parse fill, stroke, view-box. Remove bounding-box.
        (WebCore::isBoxValue):
        * css/CSSPrimitiveValueMappings.h:
        (WebCore::CSSPrimitiveValue::CSSPrimitiveValue):
        (WebCore::CSSPrimitiveValue::operator LayoutBox):
        * css/CSSValueKeywords.in:
        * css/DeprecatedStyleBuilder.cpp:
        (WebCore::ApplyPropertyClipPath::applyValue):
        * rendering/RenderLayer.cpp:
        (WebCore::computeReferenceBox):
        * rendering/shapes/ShapeInfo.h:
        (WebCore::ShapeInfo::setShapeSize):
        (WebCore::ShapeInfo::logicalTopOffset):
        (WebCore::ShapeInfo::logicalLeftOffset):
        * rendering/style/RenderStyleConstants.h:
        * rendering/svg/SVGRenderingContext.cpp: Use different reference boxes per keyword.
        (WebCore::SVGRenderingContext::prepareToRenderSVGContent):

2014-02-09  Dean Jackson  <dino@apple.com>

        [WebGL] Remove state restorer code
        https://bugs.webkit.org/show_bug.cgi?id=128516

        Reviewed by Dirk Schulze.

        I'm not sure what the point of the WebGLStateRestorer class was. It was
        always constructed with a false parameter, and then wasn't even retained
        in local scope (so wouldn't have worked anyway). Meanwhile there were
        also LOTS of calls to cleanupAfterGraphicsCall, the majority of which
        were a no-op. I replaced the handful of cleanupAfterGraphicsCall(true)
        with the call to markContextChanged().

        * html/canvas/OESVertexArrayObject.cpp:
        (WebCore::OESVertexArrayObject::bindVertexArrayOES):
        * html/canvas/WebGLRenderingContext.cpp:
        (WebCore::WebGLRenderingContext::activeTexture):
        (WebCore::WebGLRenderingContext::attachShader):
        (WebCore::WebGLRenderingContext::bindAttribLocation):
        (WebCore::WebGLRenderingContext::bindBuffer):
        (WebCore::WebGLRenderingContext::bindFramebuffer):
        (WebCore::WebGLRenderingContext::bindRenderbuffer):
        (WebCore::WebGLRenderingContext::bindTexture):
        (WebCore::WebGLRenderingContext::blendColor):
        (WebCore::WebGLRenderingContext::blendEquation):
        (WebCore::WebGLRenderingContext::blendEquationSeparate):
        (WebCore::WebGLRenderingContext::blendFunc):
        (WebCore::WebGLRenderingContext::blendFuncSeparate):
        (WebCore::WebGLRenderingContext::bufferData):
        (WebCore::WebGLRenderingContext::bufferSubData):
        (WebCore::WebGLRenderingContext::checkFramebufferStatus):
        (WebCore::WebGLRenderingContext::clear):
        (WebCore::WebGLRenderingContext::clearColor):
        (WebCore::WebGLRenderingContext::clearDepth):
        (WebCore::WebGLRenderingContext::clearStencil):
        (WebCore::WebGLRenderingContext::colorMask):
        (WebCore::WebGLRenderingContext::compileShader):
        (WebCore::WebGLRenderingContext::compressedTexImage2D):
        (WebCore::WebGLRenderingContext::compressedTexSubImage2D):
        (WebCore::WebGLRenderingContext::copyTexImage2D):
        (WebCore::WebGLRenderingContext::copyTexSubImage2D):
        (WebCore::WebGLRenderingContext::cullFace):
        (WebCore::WebGLRenderingContext::depthFunc):
        (WebCore::WebGLRenderingContext::depthMask):
        (WebCore::WebGLRenderingContext::depthRange):
        (WebCore::WebGLRenderingContext::detachShader):
        (WebCore::WebGLRenderingContext::disable):
        (WebCore::WebGLRenderingContext::disableVertexAttribArray):
        (WebCore::WebGLRenderingContext::validateDrawArrays):
        (WebCore::WebGLRenderingContext::drawArrays):
        (WebCore::WebGLRenderingContext::validateDrawElements):
        (WebCore::WebGLRenderingContext::drawElements):
        (WebCore::WebGLRenderingContext::enable):
        (WebCore::WebGLRenderingContext::enableVertexAttribArray):
        (WebCore::WebGLRenderingContext::finish):
        (WebCore::WebGLRenderingContext::flush):
        (WebCore::WebGLRenderingContext::framebufferRenderbuffer):
        (WebCore::WebGLRenderingContext::framebufferTexture2D):
        (WebCore::WebGLRenderingContext::frontFace):
        (WebCore::WebGLRenderingContext::generateMipmap):
        (WebCore::WebGLRenderingContext::getBufferParameter):
        (WebCore::WebGLRenderingContext::getFramebufferAttachmentParameter):
        (WebCore::WebGLRenderingContext::getParameter):
        (WebCore::WebGLRenderingContext::getProgramParameter):
        (WebCore::WebGLRenderingContext::getProgramInfoLog):
        (WebCore::WebGLRenderingContext::getRenderbufferParameter):
        (WebCore::WebGLRenderingContext::getShaderParameter):
        (WebCore::WebGLRenderingContext::getShaderInfoLog):
        (WebCore::WebGLRenderingContext::getTexParameter):
        (WebCore::WebGLRenderingContext::getUniform):
        (WebCore::WebGLRenderingContext::getUniformLocation):
        (WebCore::WebGLRenderingContext::getVertexAttrib):
        (WebCore::WebGLRenderingContext::getVertexAttribOffset):
        (WebCore::WebGLRenderingContext::hint):
        (WebCore::WebGLRenderingContext::lineWidth):
        (WebCore::WebGLRenderingContext::linkProgram):
        (WebCore::WebGLRenderingContext::pixelStorei):
        (WebCore::WebGLRenderingContext::polygonOffset):
        (WebCore::WebGLRenderingContext::readPixels):
        (WebCore::WebGLRenderingContext::releaseShaderCompiler):
        (WebCore::WebGLRenderingContext::renderbufferStorage):
        (WebCore::WebGLRenderingContext::sampleCoverage):
        (WebCore::WebGLRenderingContext::scissor):
        (WebCore::WebGLRenderingContext::shaderSource):
        (WebCore::WebGLRenderingContext::stencilFunc):
        (WebCore::WebGLRenderingContext::stencilFuncSeparate):
        (WebCore::WebGLRenderingContext::stencilMask):
        (WebCore::WebGLRenderingContext::stencilMaskSeparate):
        (WebCore::WebGLRenderingContext::stencilOp):
        (WebCore::WebGLRenderingContext::stencilOpSeparate):
        (WebCore::WebGLRenderingContext::texImage2DBase):
        (WebCore::WebGLRenderingContext::texImage2D):
        (WebCore::WebGLRenderingContext::texParameter):
        (WebCore::WebGLRenderingContext::texSubImage2DBase):
        (WebCore::WebGLRenderingContext::uniform1f):
        (WebCore::WebGLRenderingContext::uniform1fv):
        (WebCore::WebGLRenderingContext::uniform1i):
        (WebCore::WebGLRenderingContext::uniform1iv):
        (WebCore::WebGLRenderingContext::uniform2f):
        (WebCore::WebGLRenderingContext::uniform2fv):
        (WebCore::WebGLRenderingContext::uniform2i):
        (WebCore::WebGLRenderingContext::uniform2iv):
        (WebCore::WebGLRenderingContext::uniform3f):
        (WebCore::WebGLRenderingContext::uniform3fv):
        (WebCore::WebGLRenderingContext::uniform3i):
        (WebCore::WebGLRenderingContext::uniform3iv):
        (WebCore::WebGLRenderingContext::uniform4f):
        (WebCore::WebGLRenderingContext::uniform4fv):
        (WebCore::WebGLRenderingContext::uniform4i):
        (WebCore::WebGLRenderingContext::uniform4iv):
        (WebCore::WebGLRenderingContext::uniformMatrix2fv):
        (WebCore::WebGLRenderingContext::uniformMatrix3fv):
        (WebCore::WebGLRenderingContext::uniformMatrix4fv):
        (WebCore::WebGLRenderingContext::useProgram):
        (WebCore::WebGLRenderingContext::validateProgram):
        (WebCore::WebGLRenderingContext::vertexAttribPointer):
        (WebCore::WebGLRenderingContext::viewport):
        (WebCore::WebGLRenderingContext::vertexAttribfImpl):
        (WebCore::WebGLRenderingContext::vertexAttribfvImpl):
        (WebCore::WebGLRenderingContext::drawArraysInstanced):
        (WebCore::WebGLRenderingContext::drawElementsInstanced):
        * html/canvas/WebGLRenderingContext.h:

2014-02-09  Dean Jackson  <dino@apple.com>

        Remove unused layer owner from WebGLLayer
        https://bugs.webkit.org/show_bug.cgi?id=128512

        Reviewed by Dirk Schulze.

        The m_layerOwner and ObjC category were not
        used anywhere.

        * platform/graphics/mac/WebGLLayer.h:
        * platform/graphics/mac/WebGLLayer.mm:
        (-[WebGLLayer display]):

2014-02-09  Ryuan Choi  <ryuan.choi@samsung.com>

        [EFL] Remove PageClientEfl
        https://bugs.webkit.org/show_bug.cgi?id=128508

        Reviewed by Andreas Kling.

        PageClientEfl was introduced for accelerated compositing of WebKit1, but
        it's bad idea because it's layer violation.
        Indeed, it has never been used since introduced.

        * platform/Widget.h: Back to PlatformPageClient to Evas_Object.
        * platform/graphics/efl/GraphicsContext3DPrivate.cpp:
        Removed empty method which use PageClientEfl.
        * platform/graphics/efl/GraphicsContext3DPrivate.h: Ditto.

2014-02-09  Tim Horton  <timothy_horton@apple.com>

        [iOS][wk2] ASSERT(WebThreadIsCurrent()) in [WebDisplayLinkHandler handleDisplayLink:]
        https://bugs.webkit.org/show_bug.cgi?id=128490

        Reviewed by Simon Fraser.

        * platform/graphics/ios/DisplayRefreshMonitorIOS.mm:
        (-[WebDisplayLinkHandler handleDisplayLink:]):
        [WebDisplayLinkHandler handleDisplayLink:] has an ASSERT(WebThreadIsCurrent()).

        We don't use the Web Thread in WebKit2, but it looks like isMainThread()
        does the right thing (main thread if web thread is disabled, web thread
        otherwise), so use that instead.

2014-02-09  Anders Carlsson  <andersca@apple.com>

        Add WTF_MAKE_FAST_ALLOCATED to more classes
        https://bugs.webkit.org/show_bug.cgi?id=128506

        Reviewed by Andreas Kling.

        * dom/Node.h:
        * dom/ScriptElement.h:
        * loader/ImageLoader.h:
        * loader/cache/CachedResourceClient.h:
        * platform/TreeShared.h:
        * platform/graphics/GlyphMetricsMap.h:
        * rendering/InlineBox.h:
        * rendering/RenderLayer.h:
        * rendering/RenderObject.h:

2014-02-09  Andreas Kling  <akling@apple.com>

        Reduce ref churn in ChildNodesLazySnapshot iteration.
        <https://webkit.org/b/128492>

        Add a missing release() in ChildNodesLazySnapshot::nextNode()
        so we don't have to churn the ref count.

        Reviewed by Sam Weinig.

        * dom/ContainerNode.h:
        (WebCore::ChildNodesLazySnapshot::nextNode):

2014-02-09  Andreas Kling  <akling@apple.com>

        Reduce ref churn in ChildInsertionNotifier.
        <https://webkit.org/b/128494>

        All callers of notifyNodeInsertedIntoDocument() already hold a strong
        reference on the Node, so there's no need to ref it again inside.

        Reviewed by Anders Carlsson.

        * dom/ContainerNodeAlgorithms.h:
        (WebCore::ChildNodeInsertionNotifier::notifyNodeInsertedIntoDocument):

2014-02-09  Zan Dobersek  <zdobersek@igalia.com>

        Manage ShadowData through std::unique_ptr
        https://bugs.webkit.org/show_bug.cgi?id=128466

        Reviewed by Andreas Kling.

        Use std::unique_ptr instead of OwnPtr to manage ShadowData objects.

        * css/SVGCSSStyleSelector.cpp:
        (WebCore::StyleResolver::applySVGProperty):
        * css/StyleResolver.cpp:
        (WebCore::StyleResolver::applyProperty):
        * page/animation/CSSPropertyAnimation.cpp:
        (WebCore::blendFunc):
        (WebCore::PropertyWrapperShadow::PropertyWrapperShadow):
        (WebCore::PropertyWrapperShadow::blendSimpleOrMatchedShadowLists):
        (WebCore::PropertyWrapperShadow::blendMismatchedShadowLists):
        * rendering/style/RenderStyle.cpp:
        (WebCore::RenderStyle::setTextShadow):
        (WebCore::RenderStyle::setBoxShadow):
        * rendering/style/RenderStyle.h:
        * rendering/style/SVGRenderStyle.h:
        (WebCore::SVGRenderStyle::setShadow):
        * rendering/style/SVGRenderStyleDefs.cpp:
        (WebCore::StyleShadowSVGData::StyleShadowSVGData):
        * rendering/style/SVGRenderStyleDefs.h:
        * rendering/style/ShadowData.cpp:
        (WebCore::ShadowData::ShadowData):
        * rendering/style/ShadowData.h:
        (WebCore::ShadowData::setNext):
        * rendering/style/StyleRareInheritedData.cpp:
        (WebCore::StyleRareInheritedData::StyleRareInheritedData):
        * rendering/style/StyleRareInheritedData.h:
        * rendering/style/StyleRareNonInheritedData.cpp:
        (WebCore::StyleRareNonInheritedData::StyleRareNonInheritedData):
        * rendering/style/StyleRareNonInheritedData.h:

2014-02-09  Carlos Garnacho  <carlosg@gnome.org>

        [GTK] Add touch-events related code to build
        https://bugs.webkit.org/show_bug.cgi?id=98931

        Reviewed by Carlos Garcia Campos.

        Tests in fast/events/touch have been enabled on GTK+ now that touch
        support is implemented, and thus all expected bits are there.

        * platform/gtk/GtkTouchContextHelper.cpp:
        * platform/gtk/GtkTouchContextHelper.h: Add helper object to track all touchpoints, handy
        when creating WebTouchEvents.
        * GNUmakefile.list.am:
        * PlatformGTK.cmake:
        * bindings/gobject/GNUmakefile.am: Add touch related code and idls to build, those are
        necessary now that GTK+ implements touch events.

2014-02-09  Zan Dobersek  <zdobersek@igalia.com>

        Switch the rest of CSS code from OwnPtr over to std::unique_ptr
        https://bugs.webkit.org/show_bug.cgi?id=128471

        Reviewed by Andreas Kling.

        Replace the remaining uses of OwnPtr in the CSS code with std::unique_ptr, covering
        the CSSParser, CSSParserSelector and CSSSelector classes and a few additional, closely related classes.

        * css/CSSGrammar.y.in:
        * css/CSSParser.cpp:
        (WebCore::CSSParser::parseSheet):
        (WebCore::CSSParser::parseDeclaration):
        (WebCore::CSSParser::parseValue):
        (WebCore::CSSParser::markSupportsRuleHeaderStart):
        (WebCore::CSSParser::createKeyframesRule):
        (WebCore::CSSParser::createStyleRule):
        (WebCore::CSSParser::rewriteSpecifiersWithElementName):
        (WebCore::CSSParser::rewriteSpecifiers):
        * css/CSSParser.h:
        * css/CSSParserValues.cpp:
        (WebCore::CSSParserSelector::CSSParserSelector):
        (WebCore::CSSParserSelector::~CSSParserSelector):
        (WebCore::CSSParserSelector::adoptSelectorVector):
        (WebCore::CSSParserSelector::insertTagHistory):
        (WebCore::CSSParserSelector::appendTagHistory):
        (WebCore::CSSParserSelector::prependTagSelector):
        * css/CSSParserValues.h:
        (WebCore::CSSParserSelector::releaseSelector):
        (WebCore::CSSParserSelector::setTagHistory):
        (WebCore::CSSParserSelector::clearTagHistory):
        * css/CSSSelector.cpp:
        (WebCore::CSSSelector::setSelectorList):
        * css/CSSSelector.h:
        * css/CSSSelectorList.cpp:
        (WebCore::CSSSelectorList::adoptSelectorVector):
        * css/CSSSelectorList.h:
        * css/CSSValueList.cpp:
        * css/StyleRule.cpp:
        (WebCore::StyleRuleRegion::StyleRuleRegion):
        * css/StyleRule.h:
        (WebCore::StyleRule::parserAdoptSelectorVector):
        (WebCore::StyleRulePage::parserAdoptSelectorVector):
        (WebCore::StyleRuleRegion::create):

2014-02-09  Zan Dobersek  <zdobersek@igalia.com>

        Manage MutationObserverInterestGroup through std::unique_ptr
        https://bugs.webkit.org/show_bug.cgi?id=128468

        Reviewed by Andreas Kling.

        Use std::unique_ptr instead of OwnPtr for managing MutationObserverInterestGroup objects.

        * css/PropertySetCSSStyleDeclaration.cpp:
        * dom/CharacterData.cpp:
        (WebCore::CharacterData::dispatchModifiedEvent):
        * dom/ChildListMutationScope.cpp:
        (WebCore::ChildListMutationAccumulator::ChildListMutationAccumulator):
        * dom/ChildListMutationScope.h:
        (WebCore::ChildListMutationAccumulator::hasObservers):
        * dom/Element.cpp:
        (WebCore::Element::willModifyAttribute):
        * dom/MutationObserverInterestGroup.cpp:
        (WebCore::MutationObserverInterestGroup::createIfNeeded):
        * dom/MutationObserverInterestGroup.h:
        (WebCore::MutationObserverInterestGroup::createForChildListMutation):
        (WebCore::MutationObserverInterestGroup::createForCharacterDataMutation):
        (WebCore::MutationObserverInterestGroup::createForAttributesMutation):

2014-02-09  Zan Dobersek  <zdobersek@igalia.com>

        Manage MessagePort, MessagePortChannel and friends through std::unique_ptr
        https://bugs.webkit.org/show_bug.cgi?id=128467

        Reviewed by Andreas Kling.

        Use std::unique_ptr instead of OwnPtr to manage MessagePort, MessagePortArray, MessagePortChannel
        and MessagePortChannelArray objects.

        * bindings/js/JSMessageEventCustom.cpp:
        (WebCore::handleInitMessageEvent):
        * dom/MessageEvent.cpp:
        (WebCore::MessageEvent::MessageEvent):
        (WebCore::MessageEvent::initMessageEvent):
        * dom/MessageEvent.h:
        (WebCore::MessageEvent::create):
        * dom/MessagePort.cpp:
        (WebCore::MessagePort::postMessage):
        (WebCore::MessagePort::disentangle):
        (WebCore::MessagePort::entangle):
        (WebCore::MessagePort::dispatchMessages):
        (WebCore::MessagePort::disentanglePorts):
        (WebCore::MessagePort::entanglePorts):
        * dom/MessagePort.h:
        * dom/MessagePortChannel.h:
        * dom/default/PlatformMessagePortChannel.cpp:
        (WebCore::PlatformMessagePortChannel::EventData::EventData):
        (WebCore::MessagePortChannel::createChannel):
        (WebCore::MessagePortChannel::postMessageToRemote):
        (WebCore::MessagePortChannel::tryGetMessageFromRemote):
        * dom/default/PlatformMessagePortChannel.h:
        (WebCore::PlatformMessagePortChannel::EventData::channels):
        * page/DOMWindow.cpp:
        (WebCore::PostMessageTimer::PostMessageTimer):
        (WebCore::PostMessageTimer::event):
        (WebCore::DOMWindow::postMessage):
        * workers/DedicatedWorkerGlobalScope.cpp:
        (WebCore::DedicatedWorkerGlobalScope::postMessage):
        * workers/DefaultSharedWorkerRepository.cpp:
        (WebCore::SharedWorkerConnectTask::create):
        (WebCore::SharedWorkerConnectTask::SharedWorkerConnectTask):
        (WebCore::SharedWorkerConnectTask::performTask):
        (WebCore::SharedWorkerScriptLoader::SharedWorkerScriptLoader):
        (WebCore::SharedWorkerScriptLoader::notifyFinished):
        (WebCore::DefaultSharedWorkerRepository::workerScriptLoaded):
        (WebCore::DefaultSharedWorkerRepository::connectToWorker):
        * workers/DefaultSharedWorkerRepository.h:
        * workers/SharedWorker.cpp:
        (WebCore::SharedWorker::create):
        * workers/SharedWorkerGlobalScope.cpp:
        (WebCore::createConnectEvent):
        * workers/SharedWorkerRepository.cpp:
        (WebCore::SharedWorkerRepository::connect):
        * workers/SharedWorkerRepository.h:
        * workers/Worker.cpp:
        (WebCore::Worker::postMessage):
        * workers/WorkerGlobalScopeProxy.h:
        * workers/WorkerMessagingProxy.cpp:
        (WebCore::MessageWorkerGlobalScopeTask::create):
        (WebCore::MessageWorkerGlobalScopeTask::MessageWorkerGlobalScopeTask):
        (WebCore::MessageWorkerGlobalScopeTask::performTask):
        (WebCore::MessageWorkerTask::create):
        (WebCore::MessageWorkerTask::MessageWorkerTask):
        (WebCore::MessageWorkerTask::performTask):
        (WebCore::WorkerMessagingProxy::postMessageToWorkerObject):
        (WebCore::WorkerMessagingProxy::postMessageToWorkerGlobalScope):
        * workers/WorkerMessagingProxy.h:
        * workers/WorkerObjectProxy.h:

2014-02-08  Ryosuke Niwa  <rniwa@webkit.org>

        Cleanup the interface of FrameSelection
        https://bugs.webkit.org/show_bug.cgi?id=128481

        Reviewed by Andreas Kling.

        Removed FrameSelection::end() as intended in r163232, and the stale declaration of
        paintDragCaret() which was supposed to be removed when we extracted DragCaretController.

        Renamed caretRenderer() to caretRendererWithoutUpdatingLayout() to clarify the contract
        as the only caller of this function is RenderBlock::paintCaret. Also renamed bounds()
        to selectionBounds() to make it more easily identifiable / grep'able.

        In addition, made recomputeCaretRect() and invalidateCaretRect() private.

        * WebCore.exp.in:
        * editing/FrameSelection.cpp:
        (WebCore::FrameSelection::caretRendererWithoutUpdatingLayout):
        (WebCore::FrameSelection::selectionBounds):
        (WebCore::FrameSelection::revealSelection):
        * editing/FrameSelection.h:
        (WebCore::FrameSelection::setCaretVisible):
        * page/DragController.cpp:
        (WebCore::dragLocForSelectionDrag):
        * page/FrameSnapshotting.cpp:
        (WebCore::snapshotSelection):
        * page/win/FrameWin.cpp:
        (WebCore::imageFromSelection):
        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::paintCaret):
        * testing/Internals.cpp:
        (WebCore::Internals::selectionBounds):

2014-02-08  Dan Bernstein  <mitz@apple.com>

        Reverted part of r163734, because the assertion it enabled still fails when running
        loader/load-defer-resume-crash.html

        * loader/DocumentLoader.cpp:
        (WebCore::DocumentLoader::dataReceived):

2014-02-08  Ryosuke Niwa  <rniwa@webkit.org>

        EFL build fix after r163729.

        * Modules/indexeddb/IDBDatabase.cpp:
        (WebCore::IDBDatabase::version):

2014-02-08  Dan Bernstein  <mitz@apple.com>

        Remove outdated workarounds in DocumentLoader::dataReceived
        https://bugs.webkit.org/show_bug.cgi?id=128465

        Reviewed by Andreas Kling.

        * loader/DocumentLoader.cpp:
        (WebCore::DocumentLoader::dataReceived):

2014-02-08  Alexey Proskuryakov  <ap@apple.com>

        Remove some unused functions from SerializedScriptValue
        https://bugs.webkit.org/show_bug.cgi?id=128407

        Reviewed by Oliver Hunt.

        Removed more unused code, particularly in API helpers. Renamed one serialize()
        function to create(), because it does the same thing as other create() functions.

        * Modules/indexeddb/IDBObjectStore.cpp:
        (WebCore::IDBObjectStore::put):
        * bindings/js/SerializedScriptValue.cpp:
        (WebCore::SerializedScriptValue::create):
        (WebCore::SerializedScriptValue::deserialize):
        * bindings/js/SerializedScriptValue.h:

2014-02-08  Brady Eidson  <beidson@apple.com>

        IDB: storage/indexeddb/mozilla/versionchange-abort.html fails
        <rdar://problem/16018887> and https://bugs.webkit.org/show_bug.cgi?id=128442

        Reviewed by Dan Bernstein.

        Tested by storage/indexeddb/mozilla/versionchange-abort.html (and probably others)

        * Modules/indexeddb/IDBDatabase.cpp:
        (WebCore::IDBDatabase::version): If the version is NoIntVersion, return DefaultIntVersion to script.

2014-02-08  Brady Eidson  <beidson@apple.com>

        IDB: storage/indexeddb/mozilla/cursors.html fails
        <rdar://problem/16017998> and https://bugs.webkit.org/show_bug.cgi?id=128423

        Reviewed by Dan Bernstein.

        Tested by storage/indexeddb/mozilla/cursors.html (And probably others)

        * Modules/indexeddb/IDBTransactionBackendOperations.cpp:
        (WebCore::OpenCursorOperation::perform): Distinguish between an error while opening the cursor
          and opening a cursor that points to no records.

2014-02-08  Andreas Kling  <akling@apple.com>

        Remove unused ChromeClient::layoutUpdated().
        <https://webkit.org/b/128470>

        Nobody listens for this callback anymore so remove it.

        Reviewed by Anders Carlsson.

        * page/Chrome.cpp:
        * page/Chrome.h:
        * page/ChromeClient.h:
        * page/FrameView.cpp:
        (WebCore::FrameView::layout):

2014-02-08  Dan Bernstein  <mitz@apple.com>

        Stop using PLATFORM(MAC) in WebCore except where it means “OS X but not iOS”
        https://bugs.webkit.org/show_bug.cgi?id=128464

        Reviewed by Anders Carlsson.

        * Modules/webaudio/AudioContext.cpp: Changed PLATFORM(MAC) to PLATFORM(COCOA).
        (WebCore::AudioContext::constructCommon):
        * accessibility/AccessibilityNodeObject.cpp: Ditto.
        (WebCore::AccessibilityNodeObject::visibleText):
        * accessibility/AccessibilityObject.h: Ditto.
        * accessibility/AccessibilityRenderObject.cpp: Ditto.
        (WebCore::AccessibilityRenderObject::clickPoint):
        (WebCore::AccessibilityRenderObject::visiblePositionForPoint):
        (WebCore::AccessibilityRenderObject::addChildren):
        * accessibility/AccessibilityRenderObject.h: Ditto.
        * bindings/js/JSInspectorFrontendHostCustom.cpp:
        (WebCore::JSInspectorFrontendHost::platform): Made it explicit that this function returns
        "mac" when built for iOS.
        * bindings/js/ScriptController.cpp: Changed PLATFORM(MAC) to PLATFORM(COCOA).
        (WebCore::ScriptController::ScriptController):
        * bindings/js/ScriptController.h: Ditto.
        * crypto/CryptoKey.cpp: Changed PLATFORM(MAC) to OS(DARWIN) && !PLATFORM(EFL) &&
        !PLATFORM(GTK) when checking for whether to use CoreCrypto.
        * crypto/keys/CryptoKeyRSA.h: Ditto.
        * dom/Clipboard.cpp: Changed PLATFORM(MAC) to PLATFORM(COCOA).
        * dom/Document.cpp: Ditto.
        (WebCore::Document::implicitClose):
        * dom/KeyboardEvent.h: Ditto.
        * editing/Editor.cpp: Ditto.
        (WebCore::Editor::cut):
        (WebCore::Editor::copy):
        (WebCore::Editor::copyImage):
        * editing/Editor.h: Ditto.
        * editing/FrameSelection.h: Ditto.
        * editing/TypingCommand.cpp: Ditto.
        (WebCore::TypingCommand::typingAddedToOpenCommand):
        * history/HistoryItem.h: Ditto.
        * html/HTMLMediaElement.cpp: Ditto.
        (WebCore::HTMLMediaElement::parseAttribute):
        * html/HTMLMediaElement.h: Ditto.
        * html/HTMLPlugInElement.cpp: Ditto.
        (WebCore::registeredPluginReplacements):
        * html/HTMLPlugInImageElement.cpp: Ditto.
        (WebCore::HTMLPlugInImageElement::setDisplayState):
        * html/HTMLSelectElement.cpp: Ditto.
        (WebCore::HTMLSelectElement::listBoxDefaultEventHandler):
        * html/shadow/MediaControlElements.cpp: Ditto.
        (WebCore::MediaControlToggleClosedCaptionsButtonElement::MediaControlToggleClosedCaptionsButtonElement):
        (WebCore::MediaControlToggleClosedCaptionsButtonElement::defaultEventHandler):
        * html/shadow/MediaControlElements.h: Ditto.
        * inspector/InspectorIndexedDBAgent.cpp: Ditto.
        * loader/CookieJar.cpp: Ditto.
        * loader/DocumentLoader.cpp: Ditto.
        (WebCore::DocumentLoader::dataReceived):
        * loader/DocumentLoader.h: Ditto.
        (WebCore::DocumentLoader::didTellClientAboutLoad):
        * loader/EmptyClients.h: Ditto.
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::loadArchive): Added !PLATFORM(IOS) around Mac-only workaround.
        (WebCore::FrameLoader::defaultObjectContentType): Changed PLATFORM(MAC) to PLATFORM(COCOA).
        (WebCore::FrameLoader::subresourceCachePolicy): Ditto.
        * loader/ResourceBuffer.h: Changed PLATFORM(MAC) to USE(FOUNDATION) around functions that
        are implemented using Foundation.
        * loader/ResourceLoadScheduler.cpp: Changed PLATFORM(MAC) to PLATFORM(COCOA).
        (WebCore::ResourceLoadScheduler::scheduleLoad):
        * loader/ResourceLoader.cpp: Ditto.
        (WebCore::ResourceLoader::didReceiveAuthenticationChallenge):
        * loader/ResourceLoader.h: Ditto.
        * loader/archive/cf/LegacyWebArchive.cpp: Ditto.
        * loader/cache/CachedResource.cpp: Changed PLATFORM(MAC) to USE(FOUNDATION) around function
        that is implemented using other USE(FOUNDATION)-guarded code.
        * loader/cache/CachedResource.h: Ditto.
        * page/Chrome.h: Changed PLATFORM(MAC) to PLATFORM(COCOA).
        * page/ChromeClient.h: Ditto.
        * page/ContextMenuClient.h: Ditto.
        * page/ContextMenuController.cpp: Ditto.
        (WebCore::ContextMenuController::contextMenuItemSelected):
        (WebCore::ContextMenuController::createAndAppendFontSubMenu):
        (WebCore::ContextMenuController::createAndAppendSpellingAndGrammarSubMenu):
        (WebCore::ContextMenuController::populate):
        (WebCore::ContextMenuController::checkOrEnableIfNeeded):
        * page/DragClient.h: Ditto.
        * page/DragController.cpp: Ditto.
        (WebCore::dragLocForDHTMLDrag):
        (WebCore::dragLocForSelectionDrag):
        (WebCore::DragController::startDrag):
        (WebCore::DragController::doImageDrag):
        * page/EditorClient.h: Ditto.
        * page/EventHandler.cpp: Ditto.
        (WebCore::EventHandler::EventHandler):
        (WebCore::EventHandler::handleMouseDraggedEvent):
        (WebCore::EventHandler::logicalScrollRecursively):
        (WebCore::EventHandler::clearDragState):
        (WebCore::EventHandler::handleWheelEvent):
        (WebCore::EventHandler::defaultWheelEventHandler):
        (WebCore::EventHandler::keyEvent):
        (WebCore::EventHandler::eventInvertsTabsToLinksClientCallResult):
        * page/EventHandler.h: Ditto.
        * page/FrameView.cpp: Ditto.
        (WebCore::FrameView::layout):
        * page/Page.h: Ditto.
        * page/Settings.cpp: Ditto.
        * page/Settings.h: Ditto.
        * page/scrolling/ScrollingCoordinator.cpp: Ditto.
        * page/scrolling/ScrollingCoordinator.h: Ditto.
        * page/scrolling/ScrollingStateNode.h: Removed unused #include.
        * page/scrolling/ScrollingThread.cpp: Changed PLATFORM(MAC) to PLATFORM(COCOA).
        (WebCore::ScrollingThread::createThreadIfNeeded):
        * page/scrolling/ScrollingThread.h: Ditto.
        * platform/network/cf/ResourceRequest.h: Added !PLATFORM(IOS) around
        applyWebArchiveHackForMail.
        * platform/network/cf/ResourceRequestCFNet.cpp:
        (WebCore::ResourceRequest::applyWebArchiveHackForMail): Ditto.
        * platform/network/mac/ResourceRequestMac.mm:
        (WebCore::ResourceRequest::applyWebArchiveHackForMail): Ditto.
        * plugins/PluginViewNone.cpp: Changed PLATFORM(MAC) to PLATFORM(COCOA)
        * rendering/RenderBox.cpp: Ditto.
        (WebCore::RenderBox::logicalScroll):
        * rendering/RenderLayerBacking.cpp: Ditto.
        (WebCore::RenderLayerBacking::createGraphicsLayer):
        (WebCore::RenderLayerBacking::createPrimaryGraphicsLayer):
        * rendering/RenderMenuList.h: Ditto.
        * rendering/RenderText.cpp: Ditto.
        (WebCore::RenderText::previousOffsetForBackwardDeletion):
        * rendering/break_lines.cpp: Removed unused #include.
        * testing/js/WebCoreTestSupport.h: Changed PLATFORM(MAC) to PLATFORM(COCOA).
        * xml/XSLStyleSheetLibxslt.cpp: Changed PLATFORM(MAC) to OS(DARWIN) && !PLATFORM(EFL) &&
        !PLATFORM(GTK) around soft-linking libxslt.
        * xml/XSLTExtensions.cpp: Ditto.
        * xml/XSLTProcessorLibxslt.cpp: Ditto.
        * xml/XSLTUnicodeSort.cpp: Ditto.

2014-02-08  Andreas Kling  <akling@apple.com>

        Remove unused ChromeClient::formStateDidChange().
        <https://webkit.org/b/128469>

        Nobody listens for this callback anymore so remove it and stop
        spamming no-op virtual dispatches in forms code.

        Reviewed by Anders Carlsson.

        * html/FileInputType.cpp:
        (WebCore::FileInputType::setFiles):
        * html/HTMLFormControlElementWithState.cpp:
        * html/HTMLFormControlElementWithState.h:
        * html/HTMLInputElement.cpp:
        (WebCore::HTMLInputElement::updateType):
        (WebCore::HTMLInputElement::setValue):
        (WebCore::HTMLInputElement::setValueFromRenderer):
        * html/HTMLSelectElement.cpp:
        (WebCore::HTMLSelectElement::updateListBoxSelection):
        (WebCore::HTMLSelectElement::selectOption):
        * html/HTMLTextAreaElement.cpp:
        (WebCore::HTMLTextAreaElement::updateValue):
        (WebCore::HTMLTextAreaElement::setValueCommon):
        * loader/EmptyClients.h:
        * page/ChromeClient.h:

2014-02-08  Chris J. Shull  <chrisjshull@gmail.com>

        Web Inspector: Find evaluates attributes in a case sensitive manner
        https://bugs.webkit.org/show_bug.cgi?id=128405

        Reviewed by Timothy Hatcher.

        Changed matchesAttribute to ignore case.

        Updated existing test with additional cases: 
        inspector-protocol/dom/dom-search.html

        * inspector/InspectorNodeFinder.cpp:
        (WebCore::InspectorNodeFinder::matchesAttribute):

2014-02-08  Andreas Kling  <akling@apple.com>

        Remove unused FrameLoaderClient::dispatchWillOpenSocketStream().
        <https://webkit.org/b/128472>

        Nobody listens for this callback anymore so remove it.

        Reviewed by Anders Carlsson.

        * Modules/websockets/WebSocketChannel.cpp:
        (WebCore::WebSocketChannel::willOpenSocketStream):
        * loader/FrameLoaderClient.h:

2014-02-08  Ryosuke Niwa  <rniwa@webkit.org>

        Split UserTriggered into FireSelectEvent and RevealSelection for selection options
        https://bugs.webkit.org/show_bug.cgi?id=128441

        Reviewed by Darin Adler.

        Split UserTriggered by FireSelectEvent and RevealSelection for selection options.

        Also added defaultSetSelectionOptions() to abstract away the default options.

        * editing/AlternativeTextController.cpp:
        (WebCore::AlternativeTextController::respondToUnappliedSpellCorrection):
        * editing/Editor.cpp:
        (WebCore::Editor::unappliedEditing):
        (WebCore::Editor::reappliedEditing):
        * editing/FrameSelection.cpp:
        (WebCore::FrameSelection::moveTo):
        (WebCore::FrameSelection::setSelectionByMouseIfDifferent): UserTriggered | DoNotRevealSelection
        is replaced by FireSelectEvent.
        (WebCore::FrameSelection::setSelectionWithoutUpdatingAppearance): Check options & FireSelectEvent
        instead of options & UserTriggered.
        (WebCore::FrameSelection::setSelection): Check options & RevealSelection instead of
        options & UserTriggered && !(options & DoNotRevealSelection).
        (WebCore::FrameSelection::prepareForDestruction):
        (WebCore::FrameSelection::setBase):
        (WebCore::FrameSelection::setExtent):
        (WebCore::FrameSelection::selectAll):
        (WebCore::FrameSelection::wordSelectionContainingCaretSelection):
        * editing/FrameSelection.h:
        (WebCore::FrameSelection::defaultSetSelectionOptions): Added.
        * editing/SpellingCorrectionCommand.cpp:
        (WebCore::SpellingCorrectionCommand::doApply):
        * html/HTMLTextFormControlElement.cpp:
        (WebCore::HTMLTextFormControlElement::selectionChanged): Renamed the argument.
        * html/HTMLTextFormControlElement.h:

2014-02-08  Zan Dobersek  <zdobersek@igalia.com>

        Move TreeScope, IdTargetObserverRegistry to std::unique_ptr
        https://bugs.webkit.org/show_bug.cgi?id=127276

        Reviewed by Andreas Kling.

        Replace uses of OwnPtr in the TreeScope and IdTargetObserverRegistry classes with std::unique_ptr.

        * dom/IdTargetObserverRegistry.cpp:
        (WebCore::IdTargetObserverRegistry::addObserver):
        * dom/IdTargetObserverRegistry.h:
        (WebCore::IdTargetObserverRegistry::IdTargetObserverRegistry):
        * dom/TreeScope.cpp:
        (WebCore::TreeScope::TreeScope):
        (WebCore::TreeScope::destroyTreeScopeData):
        (WebCore::TreeScope::addElementById):
        (WebCore::TreeScope::addElementByName):
        (WebCore::TreeScope::addImageMap):
        (WebCore::TreeScope::labelElementForId):
        * dom/TreeScope.h:
        (WebCore::TreeScope::shouldCacheLabelsByForAttribute):

2014-02-08  Anders Carlsson  <andersca@apple.com>

        Slight CTTE in PingLoader
        https://bugs.webkit.org/show_bug.cgi?id=128462

        Reviewed by Dan Bernstein.

        PingLoader always wants a non-null frame.

        * html/HTMLAnchorElement.cpp:
        (WebCore::HTMLAnchorElement::sendPings):
        * html/parser/XSSAuditorDelegate.cpp:
        (WebCore::XSSAuditorDelegate::didBlockScript):
        * inspector/InspectorInstrumentation.h:
        (WebCore::InspectorInstrumentation::continueAfterPingLoader):
        * loader/PingLoader.cpp:
        (WebCore::PingLoader::loadImage):
        (WebCore::PingLoader::sendPing):
        (WebCore::PingLoader::sendViolationReport):
        (WebCore::PingLoader::createPingLoader):
        (WebCore::PingLoader::PingLoader):
        * loader/PingLoader.h:
        * loader/cache/CachedResourceLoader.cpp:
        (WebCore::CachedResourceLoader::requestImage):
        * page/ContentSecurityPolicy.cpp:
        (WebCore::ContentSecurityPolicy::reportViolation):

2014-02-08  Dan Bernstein  <mitz@apple.com>

        Remove client-drawn highlights (-webkit-highlight, WebHTMLHighlighter)
        https://bugs.webkit.org/show_bug.cgi?id=128456

        Reviewed by Anders Carlsson.

        Updated fast/css/getComputedStyle and svg/css results.

        * css/CSSComputedStyleDeclaration.cpp:
        (WebCore::ComputedStyleExtractor::propertyValue): Removed CSSPropertyWebKitHighlight case.
        * css/CSSParser.cpp:
        (WebCore::CSSParser::parseValue): Ditto.
        * css/CSSPropertyNames.in: Removed -webkit-highlight.
        * css/DeprecatedStyleBuilder.cpp:
        (WebCore::DeprecatedStyleBuilder::DeprecatedStyleBuilder): Removed
        CSSPropertyWebKitHighlight handler.
        * css/StyleResolver.cpp:
        (WebCore::StyleResolver::applyProperty): Removed CSSPropertyWebKitHighlight case.
        * page/Chrome.cpp: Removed customHighlightRect and paintCustomHighlight.
        * page/ChromeClient.h: Ditto.
        * rendering/InlineTextBox.cpp:
        (WebCore::InlineTextBox::paint): Removed painting custom highlight.
        * rendering/InlineTextBox.h:
        * rendering/RenderBlockLineLayout.cpp:
        (WebCore::RenderBlockFlow::createLineBoxesFromBidiRuns): Removed adding overflow for custom
        highlights.
        * rendering/RenderBox.cpp: Removed paintCustomHighlight.
        * rendering/RenderBox.h:
        * rendering/RenderImage.cpp:
        (WebCore::RenderImage::paintReplaced): Removed painting custom highlight.
        * rendering/RenderListMarker.cpp:
        (WebCore::RenderListMarker::paint): Ditto.
        * rendering/RenderSnapshottedPlugIn.cpp:
        (WebCore::RenderSnapshottedPlugIn::paintSnapshot): Ditto.
        * rendering/RenderWidget.cpp:
        (WebCore::RenderWidget::paint): Ditto.
        * rendering/RootInlineBox.cpp:
        (WebCore::RootInlineBox::paint): Ditto.
        * rendering/RootInlineBox.h:
        * rendering/style/RenderStyle.cpp:
        (WebCore::RenderStyle::changeRequiresLayout): Removed highlight comparison.
        * rendering/style/RenderStyle.h:
        * rendering/style/StyleRareInheritedData.cpp:
        (WebCore::StyleRareInheritedData::StyleRareInheritedData): Removed initializer.
        (WebCore::StyleRareInheritedData::operator==): Removed highlight comparison.
        * rendering/style/StyleRareInheritedData.h: Removed highlight member variable.

2014-02-08  Dan Bernstein  <mitz@apple.com>

        One more build fix fix.

        * accessibility/ios/WebAccessibilityObjectWrapperIOS.mm:
        (-[WebAccessibilityObjectWrapper arrayOfTextForTextMarkers:attributed:]):

2014-02-08  Dan Bernstein  <mitz@apple.com>

        More (and more correct) iOS build fix after r163712.

        * accessibility/ios/WebAccessibilityObjectWrapperIOS.mm:
        (AXAttributedStringAppendText):
        (-[WebAccessibilityObjectWrapper arrayOfTextForTextMarkers:attributed:]):
        * page/ios/FrameIOS.mm:
        (WebCore::Frame::wordsInCurrentParagraph):

2014-02-08  Dan Bernstein  <mitz@apple.com>

        iOS build fix after r163712.

        * page/ios/FrameIOS.mm:
        (WebCore::Frame::indexCountOfWordPrecedingSelection):
        (WebCore::Frame::wordsInCurrentParagraph):

2014-02-08  Darin Adler  <darin@apple.com>

        Change TextIterator to use StringView, preparing to wean it from deprecatedCharacters
        https://bugs.webkit.org/show_bug.cgi?id=128233

        Reviewed by Anders Carlsson.

        * accessibility/AccessibilityNodeObject.cpp: Removed unneeded TextIterator.h include.

        * accessibility/AccessibilityObject.cpp:
        (WebCore::AccessibilityObject::hasMisspelling): Updated to use StringView for checkSpelling.

        * accessibility/mac/WebAccessibilityObjectWrapperMac.mm:
        (AXAttributeStringSetSpelling): Changed to take a StringView.
        (AXAttributedStringAppendText): Ditto.
        (-[WebAccessibilityObjectWrapper doAXAttributedStringForTextMarkerRange:]): Pass StringView.

        * editing/Editor.cpp:
        (WebCore::Editor::misspelledWordAtCaretOrRange): Pass StringView.
        (WebCore::Editor::misspelledSelectionString): Ditto.
        (WebCore::Editor::markAllMisspellingsAndBadGrammarInRanges): Ditto.

        * editing/TextCheckingHelper.cpp:
        (WebCore::findGrammaticalErrors): Renamed this function. Changed to use StringView.
        (WebCore::findMisspellings): Use StringView.
        (WebCore::TextCheckingHelper::findFirstMisspelling): Ditto. Also separated out assertions
        that were asserting multiple things with &&.
        (WebCore::TextCheckingHelper::findFirstMisspellingOrBadGrammar): Ditto.
        (WebCore::TextCheckingHelper::findFirstGrammarDetail): Ditto.
        (WebCore::TextCheckingHelper::findFirstBadGrammar): Ditto.
        (WebCore::TextCheckingHelper::isUngrammatical): Ditto.
        (WebCore::TextCheckingHelper::guessesForMisspelledOrUngrammaticalRange): Ditto.
        (WebCore::checkTextOfParagraph): Ditto.

        * editing/TextCheckingHelper.h: Made TextCheckingParagraph::text public. Deleted
        TextCheckingParagraph::textDeprecatedCharacters. Added comments about additional
        TextCheckingParagraph refinements. Changed checkTextOfParagraph to take a client
        reference instead of pointer and StringView instead of characters pointer with length.

        * editing/TextIterator.cpp:
        (WebCore::TextIterator::appendTextToStringBuilder): Use data members directly,
        rather than using functions, since we already checked m_textCharacters for null.
        (WebCore::CharacterIterator::string): Use text() instead of characters().
        (WebCore::WordAwareIterator::WordAwareIterator): Removed initialization of
        m_previousText now that it's a StringView rather than a pointer.
        (WebCore::WordAwareIterator::advance): Use TextIterator::text instead of
        TextIterator::characters. Also added a FIXME about a fundamental problem
        with the implementation of this class!
        (WebCore::WordAwareIterator::length): Updated for m_previousText change.
        (WebCore::WordAwareIterator::text): Replaced WordAwareIterator::characters with this.
        (WebCore::SearchBuffer::append): Changed to take a StringView.
        (WebCore::SearchBuffer::prependContext): Ditto.
        (WebCore::SearchBuffer::isWordStartMatch): Use StringView.
        (WebCore::SearchBuffer::search): Ditto.
        (WebCore::findPlainText): Ditto.

        * editing/TextIterator.h: Added TextIterator::text that returns a StringView, and
        renamed TextIterator::characters to TextIterator::deprecatedTextIteratorCharacters
        (easy to search for in source code). Added SimplifiedBackwardsTextIterator::text
        and removed SimplifiedBackwardsTextIterator::characters. Added CharacterIterator::text,
        and removed CharacterIterator::characters. Added WordAwareIterator::text and removed
        WorkdAwareIterator::characters. Changed WordAwareIterator data members to use StringView.

        * editing/VisibleSelection.cpp:
        (WebCore::VisibleSelection::appendTrailingWhitespace): Use TextIterator::text instead
        of TextIterator::characters.

        * editing/VisibleUnits.cpp:
        (WebCore::previousBoundary): Updated to use StringView.
        (WebCore::nextBoundary): Ditto.
        (WebCore::startWordBoundary): Ditto.
        (WebCore::startOfWord): Ditto.
        (WebCore::endWordBoundary): Ditto.
        (WebCore::endOfWord): Fixed formatting.
        (WebCore::previousWordPositionBoundary): Updated to use StringView.
        (WebCore::previousWordPosition): Fixed formatting and got rid of local variable.
        (WebCore::nextWordPositionBoundary): Updated to use StringView.
        (WebCore::nextWordPosition): Fixed formatting and got rid of local variable.
        (WebCore::inSameLine): Fixed formatting.
        (WebCore::isStartOfLine): Ditto.
        (WebCore::isEndOfLine): Ditto.
        (WebCore::absoluteLineDirectionPointToLocalPointInBlock): Changed to take a reference.
        (WebCore::previousLinePosition): Fixed formatting chand changed to pass a reference.
        (WebCore::nextLinePosition): Ditto.
        (WebCore::startSentenceBoundary): Updated to use StringView and got rid of a local.
        (WebCore::startOfSentence): Fixed formatting.
        (WebCore::endSentenceBoundary): Updated to use StringView and got rid of a local.
        (WebCore::endOfSentence): Fixed formatting.
        (WebCore::previousSentencePositionBoundary): Updated to use StringView and got rid of
        a local.
        (WebCore::previousSentencePosition): Ditto.
        (WebCore::nextSentencePositionBoundary): Ditto.
        (WebCore::nextSentencePosition): Fixed formatting.
        (WebCore::endOfParagraph): Ditto.
        (WebCore::inSameParagraph): Ditto.
        (WebCore::isStartOfParagraph): Ditto.
        (WebCore::isEndOfParagraph): Ditto.
        (WebCore::inSameBlock): Ditto.
        (WebCore::isStartOfBlock): Ditto.
        (WebCore::isEndOfBlock): Ditto.
        (WebCore::startOfDocument): Ditto.
        (WebCore::endOfDocument): Ditto.
        (WebCore::inSameDocument): Ditto.
        (WebCore::isStartOfDocument): Ditto.
        (WebCore::isEndOfDocument): Ditto.
        (WebCore::isEndOfEditableOrNonEditableContent): Ditto.

        * loader/EmptyClients.h: Use StringView.

        * platform/mac/HTMLConverter.mm:
        (+[WebHTMLConverter editingAttributedStringFromRange:]): Use StringView.

        * platform/text/TextBoundaries.cpp:
        (WebCore::endOfFirstWordBoundaryContext): Use StringView and unsigned.
        (WebCore::startOfLastWordBoundaryContext): Ditto.

        * platform/text/TextBoundaries.h: Change interfaces to use StringView,
        and in some cases, unsigned instead of int. All call sites were better off
        with unsigned.

        * platform/text/TextCheckerClient.h: Use StringView.

        * platform/text/mac/TextBoundaries.mm: Changed conditionals to say
        USE(APPKIT) instead of PLATFORM(IOS), since that's the real issue.
        (WebCore::isSkipCharacter): Tweaked formatting.
        (WebCore::isWhitespaceCharacter): Ditto.
        (WebCore::isWordDelimitingCharacter): Ditto, also removed local variable.
        (WebCore::isSymbolCharacter): Ditto.
        (WebCore::tokenizerForString): Ditto.
        (WebCore::findSimpleWordBoundary): Use StringView. Also changed to mostly
        use unsigned instead of int.
        (WebCore::findComplexWordBoundary): Use StringView. Also restructured to
        be much more readable, with early returns and such.
        (WebCore::findWordBoundary): Use StringView and unsigned.
        (WebCore::findEndWordBoundary): Removed redudant copy of the findWordBoundary
        function and changed this to just call findWordBoundary. The reason this
        function exists is to optimize this case for some non-Mac, non-iOS platforms.
        We can always do that for Mac and/or iOS later if we like.
        (WebCore::findNextWordFromIndex): Use StringView. Also use wordBreakIterator
        instead of using UBreakIterator directly so we get a cached iterator instead
        of creating and destroying a new one each time this function is called.

        * bindings/objc/DOMUIKitExtensions.mm: Removed unneeded includes.
        * dom/Element.cpp: Ditto.
        * dom/PositionIterator.cpp: Ditto.
        * editing/ApplyBlockElementCommand.cpp: Ditto.
        * editing/IndentOutdentCommand.cpp: Ditto.
        * editing/InsertListCommand.cpp: Ditto.
        * editing/markup.cpp: Ditto.
        * html/HTMLElement.cpp: Ditto.
        * html/HTMLTextAreaElement.cpp: Ditto.
        * page/Frame.cpp: Ditto.
        * rendering/RenderTextControl.cpp: Ditto.

2014-02-08  Andreas Kling  <akling@apple.com>

        Remove two unused function declarations.

        Scrub "highQualityRepaintTimerFired" declarations from RenderImage
        and RenderBoxModelObject since they're not actually defined.

        * rendering/RenderBoxModelObject.h:
        * rendering/RenderImage.h:

2014-02-08  Andreas Kling  <akling@apple.com>

        CTTE: SVGTRefTargetEventListener is always owned by SVGTRefElement.
        <https://webkit.org/b/128432>

        Tighten up the relationship between SVGTRefElement and its internal
        event listener helper by storing the listener in a Ref, and making
        the listeners backpointer to the element be a reference.

        Reviewed by Anders Carlsson.

        * svg/SVGTRefElement.cpp:
        (WebCore::SVGTRefTargetEventListener::create):
        (WebCore::SVGTRefTargetEventListener::SVGTRefTargetEventListener):
        (WebCore::SVGTRefTargetEventListener::operator==):
        (WebCore::SVGTRefTargetEventListener::handleEvent):
        (WebCore::SVGTRefElement::SVGTRefElement):
        * svg/SVGTRefElement.h:

2014-02-07  Andreas Kling  <akling@apple.com>

        Devirtualize RenderBlockFlowRareData.
        <https://webkit.org/b/128427>

        This class had a virtual destructor for no reason. Removing it
        shrinks RenderBlockFlowRareData by 8 bytes.

        Reviewed by Anders Carlsson.

        * rendering/RenderBlockFlow.h:
        (WebCore::RenderBlockFlow::RenderBlockFlowRareData::~RenderBlockFlowRareData):

2014-02-08  Andreas Kling  <akling@apple.com>

        Use renderer iterators in two more places.
        <https://webkit.org/b/128371>

        Reviewed by Antti Koivisto.

        * dom/Position.cpp:
        (WebCore::Position::primaryDirection):

            Use lineageOfType instead of walking the parent chain.

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::insertOnlyThisLayer):

            Use childrenOfType instead of walking the children.

2014-02-07  David Kilzer  <ddkilzer@apple.com>

        [ASan] Use new/delete in PODFreeListArena
        <http://webkit.org/b/128437>

        Reviewed by Oliver Hunt.

        * platform/PODFreeListArena.h:
        (WebCore::PODFreeListArena::allocateObject): Use new when
        ADDRESS_SANITIZER is defined.
        (WebCore::PODFreeListArena::freeObject): Use delete when
        ADDRESS_SANITIZER is defined.

2014-02-07  Ryosuke Niwa  <rniwa@webkit.org>

        Merge updateSelectionCachesIfSelectionIsInsideTextFormControl into setSelectionWithoutUpdatingAppearance
        https://bugs.webkit.org/show_bug.cgi?id=128439

        Reviewed by Anders Carlsson.

        FrameSelection::selectAll had a superfluous call to updateSelectionCachesIfSelectionIsInsideTextFormControl
        because it wasn't setting UserTriggered option to avoid revealing selection.

        Call setSelection with UserTriggered and recently added DoNotRevealSelection option and merge
        updateSelectionCachesIfSelectionIsInsideTextFormControl into setSelectionWithoutUpdatingAppearance.

        Also rename local variables in setSelectionWithoutUpdatingAppearance, newSelection to
        newSelectionPossiblyWithoutDirection and s to newSelection so that they're self explanatory.

        In addition, we now update the input element's selection caches before calling
        selectFrameElementInParentIfFullySelected but this should be fine because selection cannot simultaneously
        select the entire document and be inside a text form control.

        * editing/FrameSelection.cpp:
        (WebCore::FrameSelection::setSelectionWithoutUpdatingAppearance):
        (WebCore::FrameSelection::selectAll):

2014-02-07  Ryosuke Niwa  <rniwa@webkit.org>

        EFL build fix attempt after r163662.

        * Modules/indexeddb/leveldb/IDBLevelDBCoding.cpp:
        (WebCore::IDBLevelDBCoding::encodeIDBKey):

2014-02-07  Ryosuke Niwa  <rniwa@webkit.org>

        FrameSelection's destructor shouldn't notify accessibility
        https://bugs.webkit.org/show_bug.cgi?id=128421

        Reviewed by Benjamin Poulain.

        Extracted setSelectionWithoutUpdatingAppearance out of setSelection and called it in prepareForDestruction
        instead of setting DoNotUpdateAppearance option. This new function doesn't reveal selection or notify
        accessibility code in addition to not updating appearance.

        Note that all implementations of notifyAccessibilityForSelectionChange in FrameSelectionAtk.cpp and
        FrameSelectionMac.mm exit early when the selection type is not caret or either start or end is null,
        which is already the case inside FrameSelection's destructor. In addition, revealSelection is never called
        when selection change was not triggered by user so there should be no behavioral change from this patch.

        * editing/FrameSelection.cpp:
        (WebCore::FrameSelection::setSelectionWithoutUpdatingAppearance):
        (WebCore::FrameSelection::setSelection):
        (WebCore::FrameSelection::prepareForDestruction):
        * editing/FrameSelection.h:
        (WebCore::FrameSelection::notifyAccessibilityForSelectionChange): Added the trivial implementation in the case
        accessibility is disabled to tidy up call sites.

2014-02-07  Martin Robinson  <mrobinson@igalia.com>

        Build fix for the GTK+ CMake build

        * PlatformGTK.cmake: VTTCue.idl was unintentionally added to the list of GObject DOM bindings
        output files. Remove it.

2014-02-07  Dirk Schulze  <krit@webkit.org>

        Per CSSOM, computed rect() function values must be comma separated
        https://bugs.webkit.org/show_bug.cgi?id=128401

        Reviewed by Simon Fraser.