Unreviewed build fix for EFL port.
[WebKit-https.git] / Source / WebCore / ChangeLog
index 2dedf4a..57db63f 100644 (file)
+2013-09-13  Ryuan Choi  <ryuan.choi@samsung.com>
+
+        Unreviewed build fix for EFL port.
+
+        * editing/efl/EditorEfl.cpp:
+        * platform/efl/PasteboardEfl.cpp:
+        (WebCore::Pasteboard::read):
+        (WebCore::Pasteboard::write):
+
+2013-09-13  Sam Weinig  <sam@webkit.org>
+
+        Fix a few lingering ScriptStates.
+
+        Reviewed by Andy Estes.
+
+        * Modules/indexeddb/IDBCursor.cpp:
+        (WebCore::IDBCursor::update):
+        * Modules/indexeddb/IDBObjectStore.cpp:
+        (WebCore::IDBObjectStore::add):
+        (WebCore::IDBObjectStore::put):
+        * bindings/cpp/WebNativeNodeFilterCondition.cpp:
+        (WebNativeNodeFilterCondition::acceptNode):
+        * bindings/cpp/WebNativeNodeFilterCondition.h:
+        * dom/CustomElementConstructor.cpp:
+        (WebCore::CustomElementConstructor::create):
+        * dom/CustomElementConstructor.h:
+        * dom/CustomElementRegistry.cpp:
+        (WebCore::CustomElementRegistry::registerElement):
+        * dom/CustomElementRegistry.h:
+        * plugins/PluginView.cpp:
+        (WebCore::PluginView::performRequest):
+
+2013-09-13  Darin Adler  <darin@apple.com>
+
+        Rename confusingly named "runtimeObject" functions
+        https://bugs.webkit.org/show_bug.cgi?id=121339
+
+        Also CTTE: HTMLElement.
+
+        Reviewed by Sam Weinig.
+
+        * bindings/js/JSHTMLAppletElementCustom.cpp:
+        (WebCore::JSHTMLAppletElement::putDelegate): Renamed.
+        (WebCore::JSHTMLAppletElement::getCallData): Renamed.
+        * bindings/js/JSHTMLEmbedElementCustom.cpp:
+        (WebCore::JSHTMLEmbedElement::putDelegate): Renamed.
+        (WebCore::JSHTMLEmbedElement::getCallData): Renamed.
+        * bindings/js/JSHTMLObjectElementCustom.cpp:
+        (WebCore::JSHTMLObjectElement::putDelegate): Renamed.
+        (WebCore::JSHTMLObjectElement::getCallData): Renamed.
+
+        * bindings/js/JSPluginElementFunctions.cpp:
+        (WebCore::isPluginElement): Changed to take HTMLElement&.
+        (WebCore::pluginInstance): Ditto.
+        (WebCore::pluginScriptObjectFromPluginViewBase): Updated for name and type change.
+        (WebCore::pluginScriptObject): Ditto.
+        (WebCore::pluginElementPropertyGetter): Ditto.
+        (WebCore::pluginElementCustomGetOwnPropertySlot): Ditto.
+        (WebCore::pluginElementCustomPut): Ditto.
+        (WebCore::pluginElementGetCallData): Ditto.
+
+        * bindings/js/JSPluginElementFunctions.h: Updated names and types.
+
+        * bindings/objc/WebScriptObject.mm:
+        (+[WebScriptObject _convertValueToObjcValue:JSC::originRootObject:rootObject:]):
+        Updated for type change.
+        * bridge/runtime_method.cpp:
+        (JSC::callRuntimeMethod): Ditto.
+
+2013-09-13  Sam Weinig  <sam@webkit.org>
+
+        There is no need for any custom bindings for Console
+        https://bugs.webkit.org/show_bug.cgi?id=121338
+
+        Reviewed by Joseph Pecoraro.
+
+        * GNUmakefile.list.am:
+        * Target.pri:
+        * UseJSC.cmake:
+        * WebCore.vcxproj/WebCore.vcxproj:
+        * WebCore.vcxproj/WebCore.vcxproj.filters:
+        * WebCore.xcodeproj/project.pbxproj:
+        * bindings/js/JSBindingsAllInOne.cpp:
+        Remove JSConsoleCustom.cpp from project files.
+
+        * bindings/js/JSConsoleCustom.cpp:
+        Removed.
+
+        * page/Console.cpp:
+        (WebCore::Console::profile):
+        (WebCore::Console::profileEnd):
+        * page/Console.h:
+        Swap the order of the arguments to appease the bindings generator.
+
+        * page/Console.idl:
+        Add the [CallWith=ScriptState] attribute to pass the ExecState
+        and [TreatNullAs=NullString, TreatUndefinedAs=NullString] to get
+        the right stringification.
+
+2013-09-13  Sam Weinig  <sam@webkit.org>
+
+        Part 2 of Replace ScriptState with JSC::ExecState
+        https://bugs.webkit.org/show_bug.cgi?id=121336
+
+        * dom/Document.h:
+        There is no WebCore::JSC namespace!
+
+2013-09-13  Sam Weinig  <sam@webkit.org>
+
+        Replace ScriptState with JSC::ExecState
+        https://bugs.webkit.org/show_bug.cgi?id=121336
+
+        Reviewed by Darin Adler.
+
+        * Modules/indexeddb/IDBCursor.h:
+        * Modules/indexeddb/IDBObjectStore.h:
+        * bindings/js/JSInjectedScriptHostCustom.cpp:
+        * bindings/js/JSInjectedScriptManager.cpp:
+        * bindings/js/JSNodeCustom.cpp:
+        * bindings/js/JSNodeFilterCondition.h:
+        * bindings/js/ScriptController.cpp:
+        * bindings/js/ScriptDebugServer.cpp:
+        * bindings/js/ScriptDebugServer.h:
+        * bindings/js/ScriptEventListener.cpp:
+        * bindings/js/ScriptEventListener.h:
+        * bindings/js/ScriptFunctionCall.cpp:
+        * bindings/js/ScriptFunctionCall.h:
+        * bindings/js/ScriptObject.cpp:
+        * bindings/js/ScriptObject.h:
+        * bindings/js/ScriptProfiler.cpp:
+        * bindings/js/ScriptProfiler.h:
+        * bindings/js/ScriptState.cpp:
+        * bindings/js/ScriptState.h:
+        * bindings/js/ScriptValue.cpp:
+        * bindings/js/ScriptValue.h:
+        * bindings/js/SerializedScriptValue.cpp:
+        * bindings/js/SerializedScriptValue.h:
+        * dom/Document.cpp:
+        * dom/Document.h:
+        * dom/NodeFilter.cpp:
+        * dom/NodeFilter.h:
+        * dom/NodeFilterCondition.cpp:
+        * dom/NodeFilterCondition.h:
+        * dom/NodeIterator.cpp:
+        * dom/NodeIterator.h:
+        * dom/ScriptExecutionContext.cpp:
+        * dom/ScriptExecutionContext.h:
+        * dom/Traversal.cpp:
+        * dom/Traversal.h:
+        * dom/TreeWalker.cpp:
+        * dom/TreeWalker.h:
+        * inspector/ConsoleMessage.cpp:
+        * inspector/ConsoleMessage.h:
+        * inspector/InjectedScript.h:
+        * inspector/InjectedScriptBase.cpp:
+        * inspector/InjectedScriptBase.h:
+        * inspector/InjectedScriptCanvasModule.cpp:
+        * inspector/InjectedScriptCanvasModule.h:
+        * inspector/InjectedScriptHost.cpp:
+        * inspector/InjectedScriptHost.h:
+        * inspector/InjectedScriptManager.cpp:
+        * inspector/InjectedScriptManager.h:
+        * inspector/InjectedScriptModule.cpp:
+        * inspector/InjectedScriptModule.h:
+        * inspector/InspectorAgent.cpp:
+        * inspector/InspectorCanvasAgent.cpp:
+        * inspector/InspectorCanvasAgent.h:
+        * inspector/InspectorConsoleAgent.cpp:
+        * inspector/InspectorConsoleAgent.h:
+        * inspector/InspectorConsoleInstrumentation.h:
+        * inspector/InspectorDOMAgent.cpp:
+        * inspector/InspectorDebuggerAgent.cpp:
+        * inspector/InspectorDebuggerAgent.h:
+        * inspector/InspectorFrontendClientLocal.cpp:
+        * inspector/InspectorFrontendHost.cpp:
+        * inspector/InspectorIndexedDBAgent.cpp:
+        * inspector/InspectorInstrumentation.cpp:
+        * inspector/InspectorInstrumentation.h:
+        * inspector/PageConsoleAgent.cpp:
+        * inspector/PageDebuggerAgent.cpp:
+        * inspector/PageRuntimeAgent.cpp:
+        * inspector/PageRuntimeAgent.h:
+        * inspector/ScriptArguments.cpp:
+        * inspector/ScriptArguments.h:
+        * inspector/ScriptDebugListener.h:
+        * inspector/WorkerDebuggerAgent.cpp:
+        * inspector/WorkerRuntimeAgent.cpp:
+        * page/Console.cpp:
+        * page/Console.h:
+        * page/ContentSecurityPolicy.cpp:
+        * page/ContentSecurityPolicy.h:
+        * page/PageConsole.cpp:
+        * page/PageConsole.h:
+        * workers/WorkerGlobalScope.cpp:
+        * workers/WorkerGlobalScope.h:
+
+2013-09-13  Sam Weinig  <sam@webkit.org>
+
+        Chrome, DragController and FocusController should hold onto a Page&
+        https://bugs.webkit.org/show_bug.cgi?id=121325
+
+        Reviewed by Anders Carlsson.
+
+        * editing/FrameSelection.cpp:
+        * editing/FrameSelection.h:
+        * page/Chrome.cpp:
+        (WebCore::Chrome::Chrome):
+        (WebCore::Chrome::scroll):
+        (WebCore::Chrome::createWindow):
+        (WebCore::canRunModalIfDuringPageDismissal):
+        (WebCore::Chrome::runBeforeUnloadConfirmPanel):
+        (WebCore::Chrome::runJavaScriptAlert):
+        (WebCore::Chrome::runJavaScriptConfirm):
+        (WebCore::Chrome::runJavaScriptPrompt):
+        (WebCore::Chrome::mouseDidMoveOverElement):
+        (WebCore::Chrome::setToolTip):
+        (WebCore::Chrome::windowScreenDidChange):
+        * page/Chrome.h:
+        * page/DragController.cpp:
+        * page/DragController.h:
+        * page/mac/DragControllerMac.mm:
+        * page/FocusController.cpp:
+        * page/FocusController.h:
+        * page/Page.cpp:
+        (WebCore::Page::Page):
+        * page/PageGroupLoadDeferrer.cpp:
+        * page/PageGroupLoadDeferrer.h:
+        Take references where possible.
+
+2013-09-13  Anders Carlsson  <andersca@apple.com>
+
+        Use nullptr instead of 0 in calls to HashMap::add
+        https://bugs.webkit.org/show_bug.cgi?id=121322
+
+        Reviewed by Sam Weinig.
+
+        * bridge/IdentifierRep.cpp:
+        (WebCore::IdentifierRep::get):
+        * css/CSSFontFaceSource.cpp:
+        (WebCore::CSSFontFaceSource::getFontData):
+        * css/CSSFontSelector.cpp:
+        (WebCore::CSSFontSelector::getFontFace):
+        * css/CSSSegmentedFontFace.cpp:
+        (WebCore::CSSSegmentedFontFace::getFontData):
+        * css/CSSValuePool.cpp:
+        (WebCore::CSSValuePool::createFontFamilyValue):
+        (WebCore::CSSValuePool::createFontFaceValue):
+        * dom/ChildListMutationScope.cpp:
+        (WebCore::ChildListMutationAccumulator::getOrCreate):
+        * dom/Document.cpp:
+        (WebCore::Document::getCSSCanvasElement):
+        * dom/NodeRareData.h:
+        (WebCore::NodeListsNodeData::addCacheWithAtomicName):
+        (WebCore::NodeListsNodeData::addCacheWithName):
+        (WebCore::NodeListsNodeData::addCacheWithQualifiedName):
+        * loader/appcache/ApplicationCacheStorage.cpp:
+        (WebCore::ApplicationCacheStorage::findOrCreateCacheGroup):
+        * page/PageGroup.cpp:
+        (WebCore::PageGroup::pageGroup):
+        * page/animation/AnimationController.cpp:
+        (WebCore::AnimationControllerPrivate::ensureCompositeAnimation):
+        * platform/mac/ScrollbarThemeMac.mm:
+        (WebCore::ScrollbarThemeMac::registerScrollbar):
+        * storage/StorageNamespaceImpl.cpp:
+        (WebCore::StorageNamespaceImpl::localStorageNamespace):
+
+2013-09-13  Sam Weinig  <sam@webkit.org>
+
+        ContextMenuController should hold onto a Page&
+        https://bugs.webkit.org/show_bug.cgi?id=121320
+
+        Reviewed by Anders Carlsson.
+
+        * page/ContextMenuController.cpp:
+        (WebCore::ContextMenuController::ContextMenuController):
+        (WebCore::ContextMenuController::~ContextMenuController):
+        (WebCore::ContextMenuController::showContextMenu):
+        (WebCore::ContextMenuController::contextMenuItemSelected):
+        (WebCore::ContextMenuController::checkOrEnableIfNeeded):
+        (WebCore::ContextMenuController::showContextMenuAt):
+        * page/ContextMenuController.h:
+        Store the Page and ContextMenuClient as references.
+
+        * page/Page.cpp:
+        (WebCore::Page::Page):
+        Pass the Page and ContextMenuClient as references and switch to using createOwned<>().
+
+2013-09-13  Tim Horton  <timothy_horton@apple.com>
+
+        Apply OVERRIDE and FINAL to Image subclasses and clean them up
+        https://bugs.webkit.org/show_bug.cgi?id=121086
+
+        Reviewed by Alexey Proskuryakov.
+
+        Everything other than the copious application of OVERRIDE and FINAL
+        is explicitly called out below.
+
+        * WebCore.exp.in:
+        Export the correct version of getCGImageRef for use by WebKit2.
+
+        * platform/graphics/BitmapImage.cpp:
+        * platform/graphics/BitmapImage.h:
+        Drive-by inline isBitmapImage.
+        Move create(HBITMAP) to be near the other create() functions.
+        Move all of the members private.
+        Devirtualize frameCount().
+        Move notSolidColor() near the other solid color functions.
+
+        * platform/graphics/CrossfadeGeneratedImage.cpp:
+        (WebCore::CrossfadeGeneratedImage::CrossfadeGeneratedImage):
+        (WebCore::CrossfadeGeneratedImage::drawPattern):
+        * platform/graphics/GeneratorGeneratedImage.cpp:
+        (WebCore::GeneratorGeneratedImage::draw):
+        (WebCore::GeneratorGeneratedImage::drawPattern):
+        Use size() and setContainerSize() instead of directly accessing a
+        superclass' (now private, previously protected) members.
+
+        * platform/graphics/CrossfadeGeneratedImage.h:
+        * platform/graphics/GeneratedImage.h:
+        Make m_size private, I don't think we do protected members.
+        
+        * platform/graphics/GeneratorGeneratedImage.h:
+        Make all the members private.
+
+        * svg/graphics/SVGImage.h:
+        * svg/graphics/SVGImageForContainer.h:
+
+2013-09-13  Alexey Proskuryakov  <ap@apple.com>
+
+        [Mac] WebCore Web Inspector files are only copied as part of clean build
+        https://bugs.webkit.org/show_bug.cgi?id=121316
+
+        Reviewed by Timothy Hatcher.
+
+        Removed insufficient dependencies to make the phase run unconditionally.
+
+        * WebCore.xcodeproj/project.pbxproj:
+
+2013-09-13  Eric Carlson  <eric.carlson@apple.com>
+
+        CodeGeneratorJS doesn't generate header includes for sequence<type>
+        https://bugs.webkit.org/show_bug.cgi?id=121314
+
+        Reviewed by Dean Jackson.
+
+        * bindings/scripts/CodeGeneratorJS.pm:
+        (SkipIncludeHeader): Don't return false for sequence and array types.
+        (AddIncludesForType): Do add includes for sequence and array types.
+        (GenerateCallbackImplementation):
+        * bindings/scripts/test/JS/JSTestTypedefs.cpp: Regenerated
+        * bindings/scripts/test/JS/JSTestTypedefs.h: Ditto.
+        * bindings/scripts/test/TestTypedefs.idl: Add a method that takes a sequence<TestEventTarget>.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        CTTE: RenderSVGResourcePattern always has an SVGPatternElement.
+        <https://webkit.org/b/121313>
+
+        Reviewed by Anders Carlsson.
+
+        This renderer is never anonymous. Tighten things up with a patternElement()
+        reference getter. Removed some unnecessary null checks and assertions.
+
+2013-09-13  Darin Adler  <darin@apple.com>
+
+        Remove layering-violating Pasteboard::plainText function (for all platforms now)
+        https://bugs.webkit.org/show_bug.cgi?id=121312
+
+        Reviewed by Antti Koivisto.
+
+        * editing/Editor.cpp:
+        (WebCore::Editor::pasteAsPlainTextWithPasteboard): Call readPlainTextFromPasteboard
+        on all platforms, not just Mac and EFL. Also call canSmartReplaceWithPasteboard with
+        a reference instead of a pointer.
+        (WebCore::Editor::readPlainTextFromPasteboard): Added. This is a cross-platform
+        version of what was previously a Mac-only function.
+        (WebCore::Editor::plainTextFromPasteboard): Added. This is the simple non-Mac version
+        of this function.
+        (WebCore::Editor::pasteWithPasteboard): Updated to call canSmartReplaceWithPasteboard
+        with a reference.
+        (WebCore::Editor::canSmartReplaceWithPasteboard): Updated to take a reference.
+
+        * editing/Editor.h: Make readPlainTextFromPasteboard cross-platform instead of
+        Mac-only. Also changed argument to canSmartReplaceWithPasteboard to be a reference.
+        Also added private plainTextFromPasteboard function.
+
+        * editing/mac/EditorMac.mm:
+        (WebCore::Editor::pasteWithPasteboard): Updated to call canSmartReplaceWithPasteboard
+        with a reference.
+        (WebCore::Editor::plainTextFromPasteboard): Added. Contains the Mac-specific parts
+        of the readPlainTextFromPasteboard function.
+
+        * platform/Pasteboard.h: Fixed a conditional in the PasteboardPlainText struct to
+        omit iOS. Removed the layering-voiolating plainText function.
+
+        * platform/blackberry/PasteboardBlackBerry.cpp:
+        (WebCore::Pasteboard::read): Renamed from plainText and updated to new calling
+        convention.
+        * platform/gtk/PasteboardGtk.cpp:
+        (WebCore::Pasteboard::read): Ditto.
+        * platform/ios/PasteboardIOS.mm:
+        (WebCore::Pasteboard::read): Ditto.
+        * platform/nix/PasteboardNix.cpp:
+        (WebCore::Pasteboard::read): Ditto.
+        * platform/qt/PasteboardQt.cpp:
+        (WebCore::Pasteboard::read): Ditto.
+        * platform/win/PasteboardWin.cpp:
+        (WebCore::Pasteboard::read): Ditto.
+
+2013-09-13  Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
+
+        [GTK] Move to the new web inspector
+        https://bugs.webkit.org/show_bug.cgi?id=120647
+
+        Reviewed by Carlos Garcia Campos.
+
+        No tests for the new inspector UI as of yet.
+
+        * GNUmakefile.am: no longer install the deprecated inspector UI files.
+        * WebCore.exp.in: add SchemeRegistry symbol that is now used.
+
+2013-09-13  Darin Adler  <darin@apple.com>
+
+        Try to fix Qt build.
+
+        * platform/qt/PasteboardQt.cpp:
+        (WebCore::Pasteboard::write): Update variable name at one place I missed.
+
+2013-09-13  Arvid Nilsson  <anilsson@rim.com>
+
+        [BlackBerry] Scrolling and zooming is slow when Flash plugin visible
+        https://bugs.webkit.org/show_bug.cgi?id=120618
+
+        Reviewed by Anders Carlsson.
+        Internally reviewed by Jacky Jiang.
+
+        JIRA 490823
+        Compositing Flash plugin content triggered a workaround for a graphics
+        driver quirk. The workaround was actually related to WebGL, not Flash,
+        back when we used glCopyTexImage to implement WebGL double buffering
+        (see PR 148933 and 150228 for more details of the old implementation).
+
+        Originally, both WebGL and plugin layers added themselves to the
+        LayerRenderer::m_layersLockingTextureResources list. Now only plugin
+        layers use this list.
+        The glCopyTexImage code was removed long ago, and WebGL now uses the
+        EGLImage-based double buffering approach implemented in
+        EGLImageLayerWebKitThread and EGLImageLayerCompositingThreadClient.
+
+        Fixed by removing the now obsolete workaround, a call to "glFinish".
+        This considerably improves the frame rate of zooming and scrolling when
+        Flash content is visible.
+
+        Improvement in frame rate not appropriate to test using layout tests,
+        could easily result in flaky test.
+
+        * platform/graphics/blackberry/LayerRenderer.cpp:
+        (WebCore::LayerRenderer::compositeLayers):
+
+2013-09-13  Zoltan Horvath  <zoltan@webkit.org>
+
+        Move LineLayoutState class to LineLayoutState.h
+        https://bugs.webkit.org/show_bug.cgi?id=121263
+
+        Reviewed by Ryosuke Niwa.
+
+        I added bitfields, reordered the members and a functions too.
+
+        No new tests, no behavior change.
+
+        * rendering/LineLayoutState.h: Added.
+        (WebCore::LineLayoutState::LineLayoutState):
+        (WebCore::LineLayoutState::markForFullLayout):
+        (WebCore::LineLayoutState::isFullLayout):
+        (WebCore::LineLayoutState::usesRepaintBounds):
+        (WebCore::LineLayoutState::setRepaintRange):
+        (WebCore::LineLayoutState::updateRepaintRangeFromBox):
+        (WebCore::LineLayoutState::endLineMatched):
+        (WebCore::LineLayoutState::setEndLineMatched):
+        (WebCore::LineLayoutState::checkForFloatsFromLastLine):
+        (WebCore::LineLayoutState::setCheckForFloatsFromLastLine):
+        (WebCore::LineLayoutState::lineInfo):
+        (WebCore::LineLayoutState::endLineLogicalTop):
+        (WebCore::LineLayoutState::setEndLineLogicalTop):
+        (WebCore::LineLayoutState::endLine):
+        (WebCore::LineLayoutState::setEndLine):
+        (WebCore::LineLayoutState::lastFloat):
+        (WebCore::LineLayoutState::setLastFloat):
+        (WebCore::LineLayoutState::floats):
+        (WebCore::LineLayoutState::floatIndex):
+        (WebCore::LineLayoutState::setFloatIndex):
+        (WebCore::LineLayoutState::adjustedLogicalLineTop):
+        (WebCore::LineLayoutState::setAdjustedLogicalLineTop):
+        (WebCore::LineLayoutState::flowThread):
+        (WebCore::LineLayoutState::setFlowThread):
+        * rendering/RenderBlockLineLayout.cpp: Remove LineLayoutState.
+
+2013-09-13  Darin Adler  <darin@apple.com>
+
+        Remove layering-violating Pasteboard::writeURL (for all platforms now)
+        https://bugs.webkit.org/show_bug.cgi?id=121306
+
+        Reviewed by Antti Koivisto.
+
+        * editing/Editor.cpp:
+        (WebCore::Editor::copyURL): Move the contents of the writeURLToPasteboard function
+        in here.
+
+        * editing/Editor.h: Removed Mac-specific writeURLToPasteboard, added private
+        Mac-specific fillInUserVisibleForm function for the one Mac-specific part of writing URLs.
+
+        * editing/efl/EditorEfl.cpp: Deleted Editor::writeURLToPasteboard.
+
+        * editing/mac/EditorMac.mm: Deleted Editor::writeURLToPasteboard, moving the code
+        into copyURL.
+        (WebCore::Editor::fillInUserVisibleForm): Added.
+
+        * page/DragController.cpp:
+        (WebCore::DragController::startDrag): Call copyURL instead of writeURL or
+        writeURLToPasteboard.
+
+        * platform/Pasteboard.h: Fixed up the PasteboardURL struct so it's suitable for
+        cross-platform use. Removed the writeURL function. Removed the Frame* argument from
+        the Windows-specific writeURLToDataObject function.
+
+        * platform/blackberry/PasteboardBlackBerry.cpp:
+        (WebCore::Pasteboard::write): Renamed/refactored from writeURL.
+        * platform/gtk/PasteboardGtk.cpp:
+        (WebCore::Pasteboard::write): Ditto.
+        * platform/nix/PasteboardNix.cpp:
+        (WebCore::Pasteboard::write): Ditto.
+        * platform/qt/PasteboardQt.cpp:
+        (WebCore::Pasteboard::write): Ditto.
+        * platform/win/PasteboardWin.cpp:
+        (WebCore::Pasteboard::writeURLToDataObject): Removed unused Frame* argument.
+        (WebCore::Pasteboard::write): Renamed/refactored from writeURL.
+
+2013-09-13  Chris Fleizach  <cfleizach@apple.com>
+
+        AX: AXPress does not work in SVG
+        https://bugs.webkit.org/show_bug.cgi?id=121248
+
+        Reviewed by Mario Sanchez Prada.
+
+        Implement accessKeyAction() on SVGElements, which is what accessibility code
+        uses in order to activate Elements.
+
+        Test: accessibility/svg-element-press.html
+
+        * svg/SVGElement.cpp:
+        (WebCore::SVGElement::accessKeyAction):
+        * svg/SVGElement.h:
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderSVGTransformableContainer is never anonymous.
+        <https://webkit.org/b/121304>
+
+        Reviewed by Anders Carlsson.
+
+        This renderer is never anonymous. Tighten things up with a graphicsElement()
+        reference getter.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderSVGImage always has a SVGImageElement.
+        <https://webkit.org/b/121301>
+
+        Reviewed by Anders Carlsson.
+
+        This renderer is never anonymous, and always has a corresponding SVGImageElement.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] Tighter element types for RenderSVGShape and subclasses.
+        <https://webkit.org/b/121302>
+
+        Reviewed by Antti Koivisto.
+
+        Codify the following:
+
+        - RenderSVGPath always has an SVGGraphicsElement.
+        - RenderSVGEllipse always has an SVGGraphicsElement.
+        - RenderSVGRect always has an SVGRectElement.
+        - RenderSVGShape always has an SVGGraphicsElement.
+
+        None of these renderers are ever anonymous, so delete element() and provide
+        strongly typed reference getters instead.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderSVGResourceFilterPrimitive always has an SVGFilterPrimitiveStandardAttributes.
+        <https://webkit.org/b/121300>
+
+        Reviewed by Anders Carlsson.
+
+        This renderer is never anonymous, and always has a corresponding SVGFilterPrimitiveStandardAttributes element. 
+
+2013-09-12  Darin Adler  <darin@apple.com>
+
+        Finish removing layer-violating code from Pasteboard (just for Mac during this first step)
+        https://bugs.webkit.org/show_bug.cgi?id=121279
+
+        Reviewed by Anders Carlsson.
+
+        * editing/Editor.h: Added webContentFromPasteboard member function, private nested class
+        WebContentReader, private createFragmentForImageResourceAndAddResource and
+        createFragmentAndAddResources member functions.
+
+        * editing/mac/EditorMac.mm: Re-sorted includes and added some.
+        (WebCore::Editor::pasteWithPasteboard): Call the new webContentFromPasteboard function.
+        (WebCore::Editor::insertParagraphSeparatorInQuotedContent): Comment tweak.
+        (WebCore::styleForSelectionStart): Removed an unneeded call to String::impl.
+        (WebCore::Editor::readPlainTextFromPasteboard): Updated to use an isURL boolean instead
+        of a separate string named "url".
+        (WebCore::Editor::WebContentReader::readWebArchive): Added. Contains reworked code that
+        was originally in Pasteboard::documentFragment.
+        (WebCore::Editor::WebContentReader::readFilenames): Ditto.
+        (WebCore::Editor::WebContentReader::readHTML): Ditto.
+        (WebCore::Editor::WebContentReader::readRTFD): Ditto.
+        (WebCore::Editor::WebContentReader::readRTF): Ditto.
+        (WebCore::Editor::WebContentReader::readImage): Ditto.
+        (WebCore::Editor::WebContentReader::readURL): Ditto.
+        (WebCore::Editor::WebContentReader::readPlainText): Ditto.
+        (WebCore::Editor::webContentFromPasteboard): Added. Uses WebContentReader to read the
+        content from the pasteboard.
+        (WebCore::Editor::createFragmentForImageResourceAndAddResource): Added. Helper used
+        by code above.
+        (WebCore::Editor::createFragmentAndAddResources): Ditto.
+
+        * platform/Pasteboard.h: Added new PasteboardWebContentReader abstract class so
+        client can get the various types of pasteboard content and stop on the kind it
+        can handle. Improved comments and formatting a bit. Removed documentFragment
+        when compiling for PLATFORM(MAC).
+
+        * platform/mac/DragDataMac.mm:
+        (WebCore::DragData::asFragment): Changed to use Editor::webContentFromPasteboard.
+
+        * platform/mac/PasteboardMac.mm:
+        (WebCore::Pasteboard::read): Updated the plain text overload of this for changes
+        to the struct that holds the result. Added a new web content overload of this,
+        which retains the pasteboard-reading part of Pasteboard::documentFragment, but
+        omits the document fragment construction part.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderSVGText always has a SVGTextElement.
+        <https://webkit.org/b/121299>
+
+        Reviewed by Antti Koivisto.
+
+        This renderer is never anonymous, and always has a corresponding <svg> element.
+        Tighten this up and remove some pointless assertions.
+
+2013-09-13  Allan Sandfeld Jensen  <allan.jensen@digia.com>
+
+        Remove support for QXmlStream as the XML parser.
+        https://bugs.webkit.org/show_bug.cgi?id=121229
+
+        Reviewed by Andreas Kling.
+
+        Remove the implementation and all the #ifdef related to it.
+
+        * Target.pri:
+        * WebCore.pri:
+        * dom/TransformSource.h:
+        * xml/XMLErrors.h:
+        * xml/XSLStyleSheet.h:
+        * xml/XSLTProcessor.h:
+        * xml/parser/XMLDocumentParser.cpp:
+        (WebCore::XMLDocumentParser::enterText):
+        (WebCore::toString):
+        (WebCore::XMLDocumentParser::exitText):
+        (WebCore::XMLDocumentParser::insertErrorMessageBlock):
+        * xml/parser/XMLDocumentParser.h:
+        * xml/parser/XMLDocumentParserQt.cpp: Removed.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderSVGViewportContainer always has a SVGSVGElement.
+        <https://webkit.org/b/121297>
+
+        Reviewed by Antti Koivisto.
+
+        This renderer is never anonymous, and always has a corresponding <svg> element.
+        Tighten this up and remove some pointless null checks and type checks.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderSVGRoot always has a SVGSVGElement.
+        <https://webkit.org/b/121293>
+
+        Reviewed by Antti Koivisto.
+
+        This renderer is never anonymous, and always has a corresponding <svg> element.
+        Tighten this up and remove some pointless null checks and assertions.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        Remove support for anonymous deprecated flexboxes.
+        <https://webkit.org/b/121288>
+
+        Reviewed by Antti Koivisto.
+
+        There are no remaining internal clients of RenderDeprecatedFlexibleBox,
+        Nuke a FIXME from r143643 where the move to basing renderers on new flexbox started.
+
+        This was the only user of anonymous deprecated flexbox, so removed the ability to
+        construct those.
+
+        Also removed some Chrome-related feature observation gunk from constructor.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderMultiColumnBlock is never anonymous.
+        <https://webkit.org/b/121292>
+
+        Reviewed by Antti Koivisto.
+
+        Hide element() and provide existingElement() instead, returning an Element&.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderGrid is never anonymous.
+        <https://webkit.org/b/121290>
+
+        Reviewed by Antti Koivisto.
+
+        Hide element() and provide existingElement() instead, returning an Element&.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderTableCaption is never anonymous.
+        <https://webkit.org/b/121289>
+
+        Reviewed by Antti Koivisto.
+
+        Hide element() and provide existingElement() instead, returning an Element&.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderTableCol is never anonymous.
+        <https://webkit.org/b/121286>
+
+        Reviewed by Antti Koivisto.
+
+        Hide element() and provide existingElement() instead, returning an Element&.
+        Removed one bogus null check that was exposed.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderListItem is never anonymous.
+        <https://webkit.org/b/121285>
+
+        Reviewed by Antti Koivisto.
+
+        Hide element() and provide existingElement() instead, returning an Element&.
+        Removed some bogus assertions that were exposed by this.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] Tighter element types for RenderSVGInline and subclasses.
+        <https://webkit.org/b/121283>
+
+        Reviewed by Antti Koivisto.
+
+        Codify the following:
+
+        - RenderSVGInline always has an SVGGraphicsElement.
+        - RenderSVGTextPath always has an SVGTextPathElement.
+        - RenderSVGTSpan always has an SVGTextPositioningElement.
+
+        None of these renderers are ever anonymous, so delete element() and provide
+        strongly typed reference getters instead.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderFileUploadControl always has a HTMLInputElement.
+        <https://webkit.org/b/121278>
+
+        Reviewed by Darin Adler.
+
+        RenderFileUploadControl is never anonymous and always has a corresponding HTMLInputElement.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderFieldset always has a HTMLFieldSetElement.
+        <https://webkit.org/b/121277>
+
+        Reviewed by Darin Adler.
+
+        RenderFieldset is never anonymous and always has a corresponding HTMLFieldSetElement.
+
+2013-09-13  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderButton always has a HTMLFormControlElement.
+        <https://webkit.org/b/121281>
+
+        Reviewed by Antti Koivisto.
+
+        RenderButton is never anonymous and has either a <button> or an <input> element,
+        and their nearest shared ancestor is HTMLFormControlElement.
+
+2013-09-13  Darin Adler  <darin@apple.com>
+
+        Give actions in Range class NON_SHOUTING_NAMES
+        https://bugs.webkit.org/show_bug.cgi?id=121280
+
+        Reviewed by Andreas Kling.
+
+        * dom/Range.cpp:
+        (WebCore::Range::deleteContents):
+        (WebCore::Range::processContents):
+        (WebCore::Range::processContentsBetweenOffsets):
+        (WebCore::Range::processNodes):
+        (WebCore::Range::processAncestorsAndTheirSiblings):
+        (WebCore::Range::extractContents):
+        (WebCore::Range::cloneContents):
+        * dom/Range.h:
+        Just let do-webcore-rename do its thing.
+        - DELETE_CONTENTS -> Delete
+        - EXTRACT_CONTENTS -> Extract
+        - CLONE_CONTENTS -> Clone
+
+2013-09-12  Santosh Mahto  <santosh.ma@samsung.com>
+
+        [EFL]Background Style of element not visible because of  default theme style
+        https://bugs.webkit.org/show_bug.cgi?id=117405
+
+        Reviewed by Gyuyoung Kim.
+
+        When select element is styled with css background property(color/image) the styling
+        gets hidden by Combo box non transparent image. This patch handle the
+        case and transparent images are used when select image is styled. So
+        now background-color and background-image are visible across the Combo box
+        
+        The code is added to support css styling of other theme control,
+        Basically an extra signal is emitted to edje which change the state 
+        of edje object whenever the control is styled.
+
+        Covered by existing test case, updated the test expected image.
+
+        * platform/efl/DefaultTheme/CMakeLists.txt:
+        * platform/efl/DefaultTheme/widget/combo/combo.edc:
+        * platform/efl/DefaultTheme/widget/combo/combo_focus_button_transparent.png: Added.
+        * platform/efl/DefaultTheme/widget/combo/combo_focus_transparent.png: Added.
+        * platform/efl/DefaultTheme/widget/combo/combo_hover_button_transparent.png: Added.
+        * platform/efl/DefaultTheme/widget/combo/combo_hover_transparent.png: Added.
+        * platform/efl/DefaultTheme/widget/combo/combo_normal_button_transparent.png: Added.
+        * platform/efl/DefaultTheme/widget/combo/combo_normal_transparent.png: Added.
+        * platform/efl/DefaultTheme/widget/combo/combo_press_button_transparent.png: Added.
+        * platform/efl/DefaultTheme/widget/combo/combo_press_transparent.png: Added.
+        * platform/efl/RenderThemeEfl.cpp:
+        (WebCore::RenderThemeEfl::applyEdjeStateFromForm):
+        (WebCore::RenderThemeEfl::isControlStyled):
+        (WebCore::RenderThemeEfl::paintThemePart):
+        * platform/efl/RenderThemeEfl.h:
+
+2013-09-12  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderMenuList always has a HTMLSelectElement.
+        <https://webkit.org/b/121276>
+
+        Reviewed by Anders Carlsson.
+
+        This renderer is never anonymous and always has a corresponding HTMLSelectElement.
+        Deleted element() in favor of HTMLSelectElement& selectElement().
+
+2013-09-12  Andreas Kling  <akling@apple.com>
+
+        [CTTE] Tighten RenderTextControl element typing.
+        <https://webkit.org/b/121275>
+
+        Reviewed by Anders Carlsson.
+
+        Codify these invariants:
+
+        - RenderTextControl always has a HTMLTextFormControl.
+        - RenderSearchField always has a HTMLInputElement.
+        - RenderTextControlSingleLine always has a HTMLInputElement.
+
+        None of these renderers are ever anonymous. Deleted element() and added
+        strongly typed reference getters instead.
+
+2013-09-12  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderListMarker is always anonymous and owned by RenderListItem.
+        <https://webkit.org/b/121274>
+
+        Reviewed by Anders Carlsson.
+
+        Store a RenderListItem& instead of a pointer in RenderListMarker.
+        Deleted the element() function since list markers are always anonymous.
+
+2013-09-12  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderListBox's element is always a HTMLSelectElement.
+        <https://webkit.org/b/121273>
+
+        Reviewed by Anders Carlsson.
+
+        Hide element() on RenderListBox and make selectElement() return a reference
+        since this renderer cannot be anonymous.
+
+2013-09-12  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderHTMLCanvas's element is always a HTMLCanvasElement.
+        <https://webkit.org/b/121272>
+
+        Reviewed by Anders Carlsson.
+
+        Add RenderHTMLCanvas::canvasElement(), hiding element().
+        This function also returns a reference since this renderer cannot be anonymous.
+
+2013-09-12  Andreas Kling  <akling@apple.com>
+
+        [CTTE] RenderTextControlMultiLine's element is always a HTMLTextAreaElement.
+        <https://webkit.org/b/121271>
+
+        Reviewed by Anders Carlsson.
+
+        Add RenderTextControlMultiLine::textAreaElement(), hiding element().
+        This function also returns a reference since this renderer cannot be anonymous.
+
+2013-09-12  Tim Horton  <timothy_horton@apple.com>
+
+        [mac] Cache rendered image in PDFDocumentImage
+        https://bugs.webkit.org/show_bug.cgi?id=121207
+
+        Reviewed by Simon Fraser.
+
+        Tests: fast/images/pdf-as-image-too-big.html
+
+        * loader/cache/CachedImage.cpp:
+        (WebCore::CachedImage::createImage):
+        PDFDocumentImage takes a ImageObserver now so that it can report
+        decoded data size changes to the memory cache.
+
+        * platform/graphics/Image.h:
+        (WebCore::Image::isPDFDocumentImage): Added.
+
+        * platform/graphics/cg/PDFDocumentImage.cpp:
+        (WebCore::PDFDocumentImage::PDFDocumentImage):
+        PDFDocumentImage takes a ImageObserver now so that it can report
+        decoded data size changes to the memory cache.
+
+        (WebCore::PDFDocumentImage::applyRotationForPainting):
+        Fix up some comments, and use GraphicsContext instead of CG API.
+
+        (WebCore::PDFDocumentImage::cacheParametersMatch):
+        Determine whether our cached image is still valid, given the new
+        destination's size, CTM scale, and source rect.
+
+        (WebCore::transformContextForPainting): Added.
+
+        (WebCore::PDFDocumentImage::updateCachedImageIfNeeded):
+        Cache a rendered bitmap of the PDF. Invalidate the cache if cacheParametersMatch
+        decides that the parameters don't match, unless we're painting in low quality mode,
+        in which case we'll scale the existing image (and then fully repaint when the
+        high-quality repaint timer fires).
+        Inform the memory cache of our new size.
+
+        (WebCore::PDFDocumentImage::draw):
+        Update the cached image if needed.
+        Paint the cached image into the context if it's available (which it might not be,
+        if the image is way too big and the allocation fails). Otherwise, paint straight
+        into the context as we previously did.
+
+        (WebCore::PDFDocumentImage::destroyDecodedData):
+        Throw away the cached image if requested.
+
+        (WebCore::PDFDocumentImage::decodedSize):
+        (WebCore::PDFDocumentImage::drawPDFPage):
+        Drive-by use GraphicsContext instead of CG directly.
+
+        * platform/graphics/cg/PDFDocumentImage.h:
+        (WebCore::PDFDocumentImage::create):
+        Override isPDFDocumentImage().
+        Add storage for the cached image buffer and various cache parameters.
+
+        * rendering/ImageQualityController.cpp:
+        (WebCore::ImageQualityController::shouldPaintAtLowQuality):
+        PDFDocumentImage is also interested in/capable of low-quality painting now.
+
+2013-09-12  Tim Horton  <timothy_horton@apple.com>
+
+        Image doesn't always repaint at high quality in all tiles after a live resize
+        https://bugs.webkit.org/show_bug.cgi?id=121244
+
+        Reviewed by Darin Adler.
+
+        Currently, ImageQualityController removes an image from its low-quality-images
+        list from inside shouldPaintAtLowQuality, if this is the first paint outside
+        of a live resize, but does not force the renderer to repaint in its entirety.
+
+        However, there's no guarantee we've invalidated the whole renderer, so this can
+        leave some parts of the image painted in low-quality.
+
+        This just removes a short-circuit, instead using the ordinary high-quality-repaint
+        timer to ensure that the entire renderer is repainted.
+
+        No new test; all attempts have failed, as this depends on tiled drawing
+        and a live resize occurring.
+
+        * rendering/ImageQualityController.cpp:
+        (WebCore::ImageQualityController::shouldPaintAtLowQuality):
+
+2013-09-12  Brent Fulgham  <bfulgham@apple.com>
+
+        [EFL] Speculative build fix after r155638.
+
+        * platform/efl/PasteboardEfl.cpp: Add include for PassRefPtr, which
+        is no longer implicitly included.
+
+2013-09-12  Simon Fraser  <simon.fraser@apple.com>
+
+        Avoid extra scrollbar-related layouts for overlay scrollbars
+        https://bugs.webkit.org/show_bug.cgi?id=121267
+
+        Reviewed by Beth Dakin.
+
+        If ScrollView::updateScrollbars() detected that scrollbars were added
+        and removed, it would call contentsResized(), which calls setNeedsLayout(),
+        followed by visibleContentsResized() which would trigger layout. There is no
+        point doing this with overlay scrollbars, so avoid it by having 
+        setHas*Scrollbar() return true if the addition/removal of a scrollbar changed
+        the available width.
+        
+        No tests: we can't test overlay scrollbars in tests.
+
+        * page/FrameView.cpp:
+        (WebCore::FrameView::setContentsSize): Drive-by assertion that
+        checks that the unsigned m_deferSetNeedsLayouts doesn't wrap when
+        decremented.
+        * platform/ScrollView.cpp:
+        (WebCore::ScrollView::setHasHorizontalScrollbar): Return true if the addition/removal
+        changed available space.
+        (WebCore::ScrollView::setHasVerticalScrollbar): Ditto.
+        (WebCore::ScrollView::updateScrollbars): Only set sendContentResizedNotification
+        if available space was changed by addition/removal of scrollbars.
+        * platform/ScrollView.h:
+
+2013-09-12  Zoltan Horvath  <zoltan@webkit.org>
+
+        [CSS Shapes] Rename shapeContainingBlockHeight to shapeContainingBlockLogicalHeight
+        https://bugs.webkit.org/show_bug.cgi?id=121252
+
+        Reviewed by Oliver Hunt.
+
+        No new tests, no behavior change.
+
+        * rendering/RenderBlockLineLayout.cpp:
+        (WebCore::pushShapeContentOverflowBelowTheContentBox):
+        * rendering/shapes/ShapeInfo.h:
+        (WebCore::ShapeInfo::shapeContainingBlockLogicalHeight):
+
+2013-09-12  Samuel White  <samuel_white@apple.com>
+
+        <meter> element AXValue is listed as a writable value
+        https://bugs.webkit.org/show_bug.cgi?id=117650
+
+        Reviewed by Chris Fleizach.
+
+        No new tests, updated existing accessibility/meter-element.html test
+        to check writability. Changed meter element value to be unwritable.
+
+        * accessibility/AccessibilityRenderObject.cpp:
+        (WebCore::AccessibilityRenderObject::canSetValueAttribute):
+
+2013-09-12  Renata Hodovan  <reni@webkit.org>
+
+        [Qt] SHOULD NEVER BE REACHED is touched WebCore::InputType::createStepRange
+        https://bugs.webkit.org/show_bug.cgi?id=121148
+
+        Reviewed by Kent Tamura.
+
+        Non-steppable input types must not rendered as slider even if its webkit-apperance style
+        property is set to slider-vertical/horizontal.
+
+        Test: platform/qt/fast/forms/range/slider-crash-on-input.html
+
+        * platform/qt/RenderThemeQStyle.cpp:
+        (WebCore::RenderThemeQStyle::paintSliderTrack):
+
+2013-09-12  Ryosuke Niwa  <rniwa@webkit.org>
+
+        Qt and Windows build fix.
+
+        * page/animation/CSSPropertyAnimation.cpp:
+
 2013-09-11  Ryosuke Niwa  <rniwa@webkit.org>
 
         Encapsulate globals in CSSPropertyAnimation.cpp