WebKit-https.git
8 years ago2011-06-30 Rafael Brandao <rafael.lobo@openbossa.org>
commit-queue@webkit.org [Thu, 30 Jun 2011 22:39:13 +0000 (22:39 +0000)]
2011-06-30  Rafael Brandao  <rafael.lobo@openbossa.org>

        Reviewed by Andreas Kling.

        [Qt] build-webkit warning: cannot find file ElementShadowPseudoId.h
        https://bugs.webkit.org/show_bug.cgi?id=63747

        Removed reference to ElementShadowPseudoId.h that no longer exists.

        * WebCore.pro:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90174 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoFix clang build.
andersca@apple.com [Thu, 30 Jun 2011 22:29:01 +0000 (22:29 +0000)]
Fix clang build.

Remove the complete Widget constructor from the .exp.in file; since Widget is an abstract class,
its complete constructor can never be called and clang has stopped emitting it altogether.

* WebCore.exp.in:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90173 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoMake TestFailures never treat a too-many-failures build as a last-passing build for...
aroben@apple.com [Thu, 30 Jun 2011 22:28:32 +0000 (22:28 +0000)]
Make TestFailures never treat a too-many-failures build as a last-passing build for a test

Fixes <http://webkit.org/b/63758> TestFailures sometimes incorrectly says that a build was
the last passing build for a test, when in fact the test wasn't run due to too many failures
causing old-run-webkit-tests to exit early

Reviewed by Anders Carlsson.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
(LayoutTestHistoryAnalyzer.prototype._incorporateBuildHistory): If there were too many
failures in this build, assume that all the tests failed in the last build would have failed
in this build, too, had they been run. Introduced some new local variables to reduce typing,
and used them throughout.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
(LayoutTestResultsLoader.prototype.start): Fix a regression from r89841 where we wouldn't
correctly remember whether old-run-webkit-tests exited early due to too many failures for
builds older than r89610. Also removed an unused parameter while I was at it.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90172 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Levi Weintraub <leviw@chromium.org>
leviw@chromium.org [Thu, 30 Jun 2011 22:15:25 +0000 (22:15 +0000)]
2011-06-30  Levi Weintraub  <leviw@chromium.org>

        Reviewed by Eric Seidel.

        Switch paintingRootRect and RenderObject repaint methods to new layout types
        https://bugs.webkit.org/show_bug.cgi?id=63746

        Switching paintingRootRect and RenderObject repaint functions over to the
        new layout unit abstraction.

        No new tests, no functionality changes.

        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::repaintDuringLayoutIfMoved):
        * rendering/RenderBox.h:
        * rendering/RenderObject.cpp:
        (WebCore::RenderObject::paintingRootRect):
        (WebCore::RenderObject::paint):
        (WebCore::RenderObject::repaintUsingContainer):
        (WebCore::RenderObject::repaintRectangle):
        (WebCore::RenderObject::repaintAfterLayoutIfNeeded):
        (WebCore::RenderObject::repaintDuringLayoutIfMoved):
        * rendering/RenderObject.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90169 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoFix the Windows build.
mrowe@apple.com [Thu, 30 Jun 2011 22:10:32 +0000 (22:10 +0000)]
Fix the Windows build.

* UIProcess/API/C/win/WKTextChecker.h:
* UIProcess/API/C/win/WKView.h:
* UIProcess/win/WebTextCheckerClient.h:
* UIProcess/win/WebUndoClient.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90168 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Cary Clark <caryclark@google.com>
commit-queue@webkit.org [Thu, 30 Jun 2011 22:00:00 +0000 (22:00 +0000)]
2011-06-30  Cary Clark  <caryclark@google.com>

        Reviewed by James Robinson.

        Use Skia if Skia on Mac Chrome is enabled
        https://bugs.webkit.org/show_bug.cgi?id=62999

        * wtf/Platform.h:
        Add switch to use Skia if, externally,
        Skia has been enabled by a gyp define.
2011-06-30  Cary Clark  <caryclark@google.com>

        Reviewed by James Robinson.

        Use Skia if Skia on Mac Chrome is enabled
        https://bugs.webkit.org/show_bug.cgi?id=62999

        No new tests. This does not affect existing
        functionality.

        * WebCore.gyp/WebCore.gyp:
        Include Skia and related files and exclude CG
        and related files when building Chromium for Skia
        on the Mac.

        * loader/cache/CachedFont.cpp:
        Rename CHROME to CHROMIUM.

        * platform/chromium/DragImageRef.h:
        * platform/graphics/FloatPoint.h:
        * platform/graphics/FloatRect.h:
        * platform/graphics/FloatSize.h:
        * platform/graphics/FontPlatformData.h:
        (WebCore::FontPlatformData::hash):
        * platform/graphics/GlyphBuffer.h:
        (WebCore::GlyphBuffer::advanceAt):
        (WebCore::GlyphBuffer::add):
        (WebCore::GlyphBuffer::expandLastAdvance):
        * platform/graphics/IntPoint.h:
        * platform/graphics/IntRect.h:
        * platform/graphics/IntSize.h:
        * platform/graphics/SimpleFontData.h:
        * platform/graphics/cg/FloatPointCG.cpp:
        * platform/graphics/cg/FloatRectCG.cpp:
        * platform/graphics/cg/FloatSizeCG.cpp:
        * platform/graphics/cg/IntPointCG.cpp:
        * platform/graphics/cg/IntRectCG.cpp:
        * platform/graphics/cg/IntSizeCG.cpp:
        * platform/graphics/mac/FontCustomPlatformData.cpp:
        (WebCore::FontCustomPlatformData::~FontCustomPlatformData):
        (WebCore::createFontCustomPlatformData):
        * platform/graphics/mac/FontCustomPlatformData.h:
        (WebCore::FontCustomPlatformData::FontCustomPlatformData):
        Ditto.
2011-06-30  Cary Clark  <caryclark@google.com>

        Reviewed by James Robinson.

        Use Skia if Skia on Mac Chrome is enabled
        https://bugs.webkit.org/show_bug.cgi?id=62999

        * WebKit.gyp:
        If building for Skia, exclude CG files.
        If not, exclude Skia files.

        * features.gypi:
        Enable Skia on Chromium Mac if so instructed by the gyp define.
        Allow Skia to be hardware accelerated on the Mac.

        * public/WebCommon.h:
        If Skia is defined on the Mac, define it in the WebXXX
        interfaces as well.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90167 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Julien Chaffraix <jchaffraix@webkit.org>
jchaffraix@webkit.org [Thu, 30 Jun 2011 21:51:17 +0000 (21:51 +0000)]
2011-06-30  Julien Chaffraix  <jchaffraix@webkit.org>

        Reviewed by Nikolas Zimmermann.

        Update SVG position values on SVG DOM updates
        https://bugs.webkit.org/show_bug.cgi?id=62439

        * svg/custom/crash-textPath-attributes-expected.txt: Added.
        * svg/custom/crash-textPath-attributes.html: Added.
        * svg/custom/resources/crash-textPath-attributes-iframe.svg: Added.
2011-06-30  Julien Chaffraix  <jchaffraix@webkit.org>

        Reviewed by Nikolas Zimmermann.

        Update SVG position values on SVG DOM updates
        https://bugs.webkit.org/show_bug.cgi?id=62439

        Test: svg/custom/crash-textPath-attributes.html

        * rendering/svg/RenderSVGInline.cpp:
        (WebCore::RenderSVGInline::destroy): Notify our containing RenderSVGText that it needs
        to update its positioning information.

        * rendering/svg/SVGInlineFlowBox.cpp:
        (WebCore::SVGInlineFlowBox::calculateBoundaries): Check the type of the InlineBox
        like the rest of the code (fixes an ASSERT_NOT_REACHED in InlineBox::calculateBoudaries).

        * rendering/svg/SVGTextLayoutAttributesBuilder.cpp:
        (WebCore::SVGTextLayoutAttributesBuilder::buildLayoutAttributesForTextSubtree): Clear
        our cached layout attributes every time we invalidate them. This avoids keeping stale
        attribute that have a backpointer to a RenderObject.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90166 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoMake TestFaiulres only load old-results directories as needed
aroben@apple.com [Thu, 30 Jun 2011 21:49:42 +0000 (21:49 +0000)]
Make TestFaiulres only load old-results directories as needed

Fixes <http://webkit.org/b/63752> Tester pages on TestFailures page load very slowly

Reviewed by Anders Carlsson.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.getBuildNames): Moved up into the API section of the class. Now just
calls through to _getBuildNamesFromResultsDirectory.
(Builder.prototype.getOldBuildNames): Added. Just calls through to
_getBuildNamesFromResultsDirectory.
(Builder.prototype._getBuildNamesFromResultsDirectory): Renamed from getBuildNames. Now
takes the directory URL as an argument and only fetches that single URL.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
(LayoutTestHistoryAnalyzer.prototype.start): Moved most logic to _analyzeBuilds. First
analyzes builds from Builder.getBuildNames, then from Builder.getOldBuildNames if needed.
(LayoutTestHistoryAnalyzer.prototype._analyzeBuilds): Moved logic here from start. (Most
changes are just indentation.) Now takes a callback to call when we've finished analyzing
all builds in buildNames so that we can try to fetch more build names if needed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90165 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Sreeram Ramachandran <sreeram@chromium.org>
commit-queue@webkit.org [Thu, 30 Jun 2011 21:46:50 +0000 (21:46 +0000)]
2011-06-30  Sreeram Ramachandran  <sreeram@chromium.org>

        Reviewed by Adam Barth.

        Suppress modal JavaScript/HTML dialogs during unload events
        https://bugs.webkit.org/show_bug.cgi?id=56397

        Adds Chromium-specific test expectations that show alerts in unload
        handlers being replaced by console messages.

        * fast/loader/page-dismissal-modal-dialogs-expected.txt: Added.
        * fast/loader/page-dismissal-modal-dialogs.html: Added.
        * fast/loader/resources/page-dismissal-modal-dialogs-iframe.html: Added.
        * platform/chromium-mac/fast/history/timed-refresh-in-cached-frame-expected.txt:
        * platform/chromium-win/fast/history/timed-refresh-in-cached-frame-expected.txt:
        * platform/chromium/fast/dom/Geolocation/notimer-after-unload-expected.txt: Added.
        * platform/chromium/fast/events/onbeforeunload-focused-iframe-expected.txt: Added.
        * platform/chromium/fast/events/onunload-clears-onbeforeunload-expected.txt: Added.
        * platform/chromium/fast/events/onunload-expected.txt: Added.
        * platform/chromium/fast/events/onunload-not-on-body-expected.txt: Added.
        * platform/chromium/fast/events/onunload-window-property-expected.txt: Added.
        * platform/chromium/fast/events/pageshow-pagehide-on-back-uncached-expected.txt: Added.
        * platform/chromium/fast/history/timed-refresh-in-cached-frame-expected.txt: Added.
        * platform/chromium/fast/loader/frames-with-unload-handlers-in-page-cache-expected.txt: Added.
        * platform/chromium/fast/loader/page-dismissal-modal-dialogs-expected.txt: Added.
        * platform/chromium/fast/loader/recursive-before-unload-crash-expected.txt: Added.
2011-06-30  Sreeram Ramachandran  <sreeram@chromium.org>

        Reviewed by Adam Barth.

        Suppress modal JavaScript/HTML dialogs during unload events
        https://bugs.webkit.org/show_bug.cgi?id=56397

        Allows clients to specify whether to show alerts during unload handlers.
        When checking for whether a page dismissal event is being dispatched,
        it's important to check all frames on the page (otherwise it becomes a
        loophole easily exploited).

        Test: fast/loader/page-dismissal-modal-dialogs.html

        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::FrameLoader):
        (WebCore::FrameLoader::stopLoading):
        (WebCore::FrameLoader::loadURL):
        (WebCore::FrameLoader::loadWithDocumentLoader):
        (WebCore::FrameLoader::stopAllLoaders):
        (WebCore::FrameLoader::fireBeforeUnloadEvent):
        * loader/FrameLoader.h:
        (WebCore::FrameLoader::pageDismissalEventBeingDispatched):
        * loader/cache/CachedResourceLoader.cpp:
        (WebCore::CachedResourceLoader::requestImage):
        * page/Chrome.cpp:
        (WebCore::canRunModalIfDuringPageDismissal):
        (WebCore::Chrome::canRunModalNow):
        (WebCore::Chrome::runJavaScriptAlert):
        (WebCore::Chrome::runJavaScriptConfirm):
        (WebCore::Chrome::runJavaScriptPrompt):
        * page/Chrome.h:
        * page/ChromeClient.h:
        (WebCore::ChromeClient::shouldRunModalDialogDuringPageDismissal):
        * page/DOMWindow.cpp:
        (WebCore::DOMWindow::showModalDialog):
2011-06-30  Sreeram Ramachandran  <sreeram@chromium.org>

        Reviewed by Adam Barth.

        Suppress modal JavaScript/HTML dialogs during unload events
        https://bugs.webkit.org/show_bug.cgi?id=56397

        Implementation of the new shouldRunModalDialogDuringPageDismissal() API
        to block alerts during unload handlers. Logs such events to the console
        and updates histograms.

        * src/ChromeClientImpl.cpp:
        (WebKit::ChromeClientImpl::shouldRunModalDialogDuringPageDismissal):
        * src/ChromeClientImpl.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90164 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoUpdate MiniBrowser and WebKitTestRunner to use the client version constants.
mrowe@apple.com [Thu, 30 Jun 2011 21:46:39 +0000 (21:46 +0000)]
Update MiniBrowser and WebKitTestRunner to use the client version constants.

Reviewed by Anders Carlsson.

* MiniBrowser/gtk/BrowserWindow.c:
(browserWindowLoaderClientInit):
(browserWindowUIClientInit):
* MiniBrowser/gtk/WebBundle/WebBundleMain.c:
(didCreatePage):
(WKBundleInitialize):
* MiniBrowser/gtk/main.c:
(createWKContextWithInjectedBundle):
* MiniBrowser/mac/AppDelegate.m:
(-[BrowserAppDelegate init]):
* MiniBrowser/mac/BrowserWindowController.m:
(-[BrowserWindowController awakeFromNib]):
* MiniBrowser/mac/WebBundle/WebBundleMain.m:
(WKBundleInitialize):
* MiniBrowser/win/BrowserView.cpp:
(BrowserView::create):
* WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
(WTR::InjectedBundle::initialize):
* WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
(WTR::InjectedBundlePage::InjectedBundlePage):
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::createOtherPage):
(WTR::TestController::initialize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90163 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoTeach APIClient to correctly handle multiple versions of a client interface.
mrowe@apple.com [Thu, 30 Jun 2011 21:46:35 +0000 (21:46 +0000)]
Teach APIClient to correctly handle multiple versions of a client interface.

Reviewed by Anders Carlsson.

If a client struct is not the current version then we can't copy it via assignment
since that will read past the end of the smaller, older struct. To deal with this
APIClient needs to be aware of the sizes of the older client version structs so that
it can copy only the appropriate amount of the struct.

* Shared/APIClient.h:
(WebKit::APIClient::initialize): Assign through to our client if the new in client is the
latest version. If an older client version was passed then we zero out our client and memcpy
the appropriate number of bytes in to it. This ensures that any new members in the client
are initialized to 0.
* Shared/APIClientTraits.cpp:
* Shared/APIClientTraits.h: Client interfaces are the size of their struct unless otherwise stated.
* WebKit2.xcodeproj/project.pbxproj:
* WebProcess/InjectedBundle/API/c/WKBundlePage.h: Bump the version and add comments indicating
which components belong to which versions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90162 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoMake the APIClient template aware of the latest client interface version number.
mrowe@apple.com [Thu, 30 Jun 2011 21:46:31 +0000 (21:46 +0000)]
Make the APIClient template aware of the latest client interface version number.

Reviewed by Anders Carlsson.

This will be used in an upcoming patch.

* Shared/APIClient.h:
* UIProcess/WebContextInjectedBundleClient.h:
* UIProcess/WebCookieManagerProxyClient.h:
* UIProcess/WebDatabaseManagerProxyClient.h:
* UIProcess/WebDownloadClient.h:
* UIProcess/WebFindClient.h:
* UIProcess/WebFormClient.h:
* UIProcess/WebGeolocationProvider.h:
* UIProcess/WebHistoryClient.h:
* UIProcess/WebIconDatabaseClient.h:
* UIProcess/WebLoaderClient.h:
* UIProcess/WebPageContextMenuClient.h:
* UIProcess/WebPolicyClient.h:
* UIProcess/WebResourceLoadClient.h:
* UIProcess/WebUIClient.h:
* WebProcess/InjectedBundle/InjectedBundleClient.h:
* WebProcess/InjectedBundle/InjectedBundlePageContextMenuClient.h:
* WebProcess/InjectedBundle/InjectedBundlePageEditorClient.h:
* WebProcess/InjectedBundle/InjectedBundlePageFormClient.h:
* WebProcess/InjectedBundle/InjectedBundlePageFullScreenClient.h:
* WebProcess/InjectedBundle/InjectedBundlePageLoaderClient.h:
* WebProcess/InjectedBundle/InjectedBundlePagePolicyClient.h:
* WebProcess/InjectedBundle/InjectedBundlePageResourceLoadClient.h:
* WebProcess/InjectedBundle/InjectedBundlePageUIClient.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90161 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoWebKit2 client interfaces should have constant declaring latest version.
mrowe@apple.com [Thu, 30 Jun 2011 21:46:23 +0000 (21:46 +0000)]
WebKit2 client interfaces should have constant declaring latest version.

Reviewed by Anders Carlsson.

* UIProcess/API/C/WKContext.h:
* UIProcess/API/C/WKCookieManager.h:
* UIProcess/API/C/WKDatabaseManager.h:
* UIProcess/API/C/WKGeolocationManager.h:
* UIProcess/API/C/WKIconDatabase.h:
* UIProcess/API/C/WKPage.h:
* WebProcess/InjectedBundle/API/c/WKBundle.h:
* WebProcess/InjectedBundle/API/c/WKBundlePage.h:
* WebProcess/InjectedBundle/API/c/WKBundlePageOverlay.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90160 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Juan C. Montemayor <jmont@apple.com>
commit-queue@webkit.org [Thu, 30 Jun 2011 21:41:58 +0000 (21:41 +0000)]
2011-06-30  Juan C. Montemayor  <jmont@apple.com>

        Reviewed by Geoffrey Garen.

        Web Inspector fails to display source for eval with syntax error
        https://bugs.webkit.org/show_bug.cgi?id=63583

        Added new LayoutTest to check that the correct line number is added to a
        syntax error inside an eval statement.

        * fast/js/eval-contained-syntax-error-expected.txt: Added.
        * fast/js/eval-contained-syntax-error.html: Added.
        * fast/js/script-tests/eval-contained-syntax-error.js: Added.
        * platform/chromium/fast/js/eval-contained-syntax-error-expected.txt: Added.
2011-06-30  Juan C. Montemayor  <jmont@apple.com>

        Reviewed by Geoffrey Garen.

        Web Inspector fails to display source for eval with syntax error
        https://bugs.webkit.org/show_bug.cgi?id=63583

        Web Inspector now displays a link to an eval statement that contains
        a syntax error.

        * parser/Parser.h:
        (JSC::isEvalNode):
        (JSC::EvalNode):
        (JSC::Parser::parse):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90159 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Nate Chapin <japhet@chromium.org>
japhet@chromium.org [Thu, 30 Jun 2011 21:37:56 +0000 (21:37 +0000)]
2011-06-30  Nate Chapin  <japhet@chromium.org>

        Unreviewed, chromium test update after r90148.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90158 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Filip Pizlo <fpizlo@apple.com>
commit-queue@webkit.org [Thu, 30 Jun 2011 21:30:57 +0000 (21:30 +0000)]
2011-06-30  Filip Pizlo  <fpizlo@apple.com>

        Reviewed by Gavin Barraclough.

        X86Assembler does not encode byte registers in 64-bit mode correctly.
        https://bugs.webkit.org/show_bug.cgi?id=63665

        * assembler/X86Assembler.h:
        (JSC::X86Assembler::testb_rr):
        (JSC::X86Assembler::X86InstructionFormatter::oneByteOp8):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90157 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Julien Chaffraix <jchaffraix@webkit.org>
jchaffraix@webkit.org [Thu, 30 Jun 2011 20:50:16 +0000 (20:50 +0000)]
2011-06-30  Julien Chaffraix  <jchaffraix@webkit.org>

        Reviewed by Nikolas Zimmermann.

        Assertion failure in RenderSVGInlineText::characterStartsNewTextChunk
        https://bugs.webkit.org/show_bug.cgi?id=63076

        * svg/custom/crash-text-in-textpath-expected.txt: Added.
        * svg/custom/crash-text-in-textpath.svg: Added.
        Original crashing test case.

        * svg/custom/text-node-in-text-invalidated-expected.txt: Added.
        * svg/custom/text-node-in-text-invalidated.svg: Added.
        This test case was not crashing. However it is good to make sure this change
        did not regress that.
2011-06-30  Julien Chaffraix  <jchaffraix@webkit.org>

        Reviewed by Nikolas Zimmermann.

        Assertion failure in RenderSVGInlineText::characterStartsNewTextChunk
        https://bugs.webkit.org/show_bug.cgi?id=63076

        Tests: svg/custom/crash-text-in-textpath.svg
               svg/custom/text-node-in-text-invalidated.svg

        The problem was that we did not call setNeedsPositionUpdate on RenderSVGText. When
        doing our layout, we would not update the attributes on our SVGRenderInlineText as
        we would not lay it out.

        This was caused by childrenChanged being overridden on SVGTextPositioningElement but
        not on SVGTextPathElement.

        As both classes shared the same mother class, it made sense to move the logic here.
        There should be no other side effects as SVGTextPathElement and SVGTextPositioningElement
        are the only classes deriving from SVGTextContentElement.

        * svg/SVGTextContentElement.cpp:
        (WebCore::SVGTextContentElement::childrenChanged): Moved this method from SVGTextPositioningElement.
        * svg/SVGTextContentElement.h:
        * svg/SVGTextPositioningElement.cpp:
        (WebCore::SVGTextPositioningElement::svgAttributeChanged): Updated after updatePositioningValuesInRenderer
        removal, replaced by RenderSVGText::locateRenderSVGTextAncestor.
        * svg/SVGTextPositioningElement.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90156 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoBuild fix.
andersca@apple.com [Thu, 30 Jun 2011 20:22:22 +0000 (20:22 +0000)]
Build fix.

* UIProcess/win/WebPopupMenuProxyWin.h:
(WebKit::WebPopupMenuProxyWin::enclosingScrollableArea):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90155 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Patrick Gansterer <paroga@webkit.org>
paroga@webkit.org [Thu, 30 Jun 2011 20:17:38 +0000 (20:17 +0000)]
2011-06-30  Patrick Gansterer  <paroga@webkit.org>

        Unreviewed build fix for !ENABLE(DATABASE) after r84789.

        * bindings/js/JSDOMBinding.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90154 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoIdentify revisions that modified failing tests on TestFailures page
aroben@apple.com [Thu, 30 Jun 2011 20:03:33 +0000 (20:03 +0000)]
Identify revisions that modified failing tests on TestFailures page

Fixes <http://webkit.org/b/63716> TestFailures page should try to infer which commit caused
a failure by looking at commit logs

Reviewed by David Kilzer.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
(.existing-bugs-list, .suspect-revisions-list): Made this rule apply to the suspect
revisions list, too.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
(Trac): Initialize our memory cache.
(Trac.prototype.getCommitDataForRevisionRange): Added. Fetches the commit log in RSS form,
then parses out the revision number, Trac's idea of the commit title, and the files modified
by the commit. (To get the modified files, we rely on the commit message including a
prepare-ChangeLog-style file list.)
(Trac.prototype.logURL): Added a new formatAsRSS parameter.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Utilities.js:
(removePathExtension): Added. Returns a new string with the last dot and everything after it
removed.
(sorted): Added sortFunction parameter.
(Node.prototype.removeAllChildren): Added. Does what it says.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._displayBuilder): Updated for change to _domForRegressionRange.
(ViewController.prototype._domForRegressionRange): Changed argument order to match
_domForNewAndExistingBugs. Now takes the list of failing tests and uses it to figure out
which revisions modified the failing tests and displays the suspect revisions in a list.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90153 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Thu, 30 Jun 2011 19:40:43 +0000 (19:40 +0000)]
2011-06-30  Anders Carlsson  <andersca@apple.com>

        Reviewed by Dan Bernstein.

        More ScrollableArea scaffolding
        https://bugs.webkit.org/show_bug.cgi?id=63678

        Add an enclosingScrollableArea stub.

        * src/WebScrollbarImpl.cpp:
        (WebKit::WebScrollbarImpl::enclosingScrollableArea):
        * src/WebScrollbarImpl.h:
2011-06-30  Anders Carlsson  <andersca@apple.com>

        Reviewed by Dan Bernstein.

        More ScrollableArea scaffolding
        https://bugs.webkit.org/show_bug.cgi?id=63678

        * WebScrollBar.h:
        (WebScrollBar::enclosingScrollableArea):
        Always return 0, WebScrollbar has no concept of an enclosing scrollable area.
2011-06-30  Anders Carlsson  <andersca@apple.com>

        Reviewed by Dan Bernstein.

        More ScrollableArea scaffolding
        https://bugs.webkit.org/show_bug.cgi?id=63678

        Add a way to get the enclosing scrollable area for a given scrollable area.
        Also, add ScrollableArea::isPinnedInDirectionOfScrollDelta.

        * page/FrameView.cpp:
        (WebCore::FrameView::enclosingScrollableArea):
        Return 0 for now.

        * page/FrameView.h:
        Add FrameView::enclosingScrollableArea.

        * platform/ScrollableArea.cpp:
        (WebCore::ScrollableArea::isPinnedInDirectionOfScrollDelta):
        Return whether the area is pinned in the direction of the scroll delta.

        * platform/ScrollableArea.h:
        Add new member functions.

        * platform/chromium/FramelessScrollView.cpp:
        (WebCore::FramelessScrollView::enclosingScrollableArea):
        * platform/chromium/FramelessScrollView.h:
        Add stub.

        * platform/win/PopupMenuWin.h:
        (WebCore::PopupMenuWin::enclosingScrollableArea):
        Always return 0.

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::enclosingScrollableLayer):
        Add helper function.

        (WebCore::RenderLayer::scrollByRecursively):
        Use the new enclosingScrollableLayer helper function.

        (WebCore::RenderLayer::enclosingScrollableArea):
        Return the enclosing scrollable area.

        * rendering/RenderLayer.h:
        Add new member function.

        * rendering/RenderListBox.cpp:
        (WebCore::RenderListBox::enclosingScrollableArea):
        Return 0 for now.

        * rendering/RenderListBox.h:
        Add new member function.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90152 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Eric Seidel <eric@webkit.org>
eric@webkit.org [Thu, 30 Jun 2011 19:34:20 +0000 (19:34 +0000)]
2011-06-30  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Rename should_kill_dump_render_tree to driver_needs_restart in preparation for supporting WebKitTestRunner
        https://bugs.webkit.org/show_bug.cgi?id=63743

        * Scripts/webkitpy/layout_tests/layout_package/test_failures.py:
        * Scripts/webkitpy/layout_tests/layout_package/worker.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90151 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Martin Robinson <mrobinson@igalia.com>
mrobinson@webkit.org [Thu, 30 Jun 2011 19:31:01 +0000 (19:31 +0000)]
2011-06-30  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Anders Carlsson.

        [GTK] Crash observed with nspluginwrapper and flash
        https://bugs.webkit.org/show_bug.cgi?id=62249

        Added a test which verifies that WebKit does not crash when InvalidateRect
        is called with a null instance.

        * platform/gtk/plugins/invalidate-rect-with-null-npp-argument-expected.txt: Added.
        * platform/gtk/plugins/invalidate-rect-with-null-npp-argument.html: Added.
2011-06-30  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Anders Carlsson.

        [GTK] Crash observed with nspluginwrapper and flash
        https://bugs.webkit.org/show_bug.cgi?id=62249

        Test: plugins/invalidate-rect-with-null-npp-argument.html

        * plugins/npapi.cpp:
        (NPN_InvalidateRect): Guard against null instances here.
2011-06-30  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Anders Carlsson.

        [GTK] Crash observed with nspluginwrapper and flash
        https://bugs.webkit.org/show_bug.cgi?id=62249

        * WebProcess/Plugins/Netscape/NetscapeBrowserFuncs.cpp:
        (WebKit::NPN_InvalidateRect): Guard against null instances here.
2011-06-30  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Anders Carlsson.

        [GTK] Crash observed with nspluginwrapper and flash
        https://bugs.webkit.org/show_bug.cgi?id=62249

        Added a TestNetscapePlugin test which verifies that WebKit properly
        handles situations where InvalidateRect is called with a null instance.

        * DumpRenderTree/TestNetscapePlugIn/Tests/x11/CallInvalidateRectWithNullNPPArgument.cpp: Added.
        (CallInvalidateRectWithNullNPPArgument::CallInvalidateRectWithNullNPPArgument):
        (CallInvalidateRectWithNullNPPArgument::NPP_New):
        * GNUmakefile.am: Add the new file to sources list.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90148 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Levi Weintraub <leviw@chromium.org>
leviw@chromium.org [Thu, 30 Jun 2011 19:16:53 +0000 (19:16 +0000)]
2011-06-30  Levi Weintraub  <leviw@chromium.org>

        Reviewed by Eric Seidel.

        Switch availableWidth/Height, logicalLeft/RightOffsetForLine to new layout types
        https://bugs.webkit.org/show_bug.cgi?id=63671

        Switch availableWidth/Height, logicalLeft/RightOffsetForLine functions over to the
        new layout unit abstraction.

        No new tests, no functionality changes.

        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::logicalLeftOffsetForLine):
        (WebCore::RenderBlock::logicalRightOffsetForLine):
        (WebCore::RenderBlock::availableLogicalWidthForLine):
        (WebCore::RenderBlock::availableLogicalWidth):
        * rendering/RenderBlock.h:
        (WebCore::RenderBlock::logicalRightOffsetForLine):
        (WebCore::RenderBlock::logicalLeftOffsetForLine):
        (WebCore::RenderBlock::startOffsetForLine):
        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::availableLogicalHeight):
        (WebCore::RenderBox::availableLogicalHeightUsing):
        * rendering/RenderBox.h:
        (WebCore::RenderBox::availableLogicalWidth):
        (WebCore::RenderBox::availableWidth):
        (WebCore::RenderBox::availableHeight):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90144 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Thu, 30 Jun 2011 19:07:23 +0000 (19:07 +0000)]
2011-06-30  Adam Barth  <abarth@webkit.org>

        Remove useless ASSERT that breaks the build.

        * dom/StyledElement.cpp:
        (WebCore::parseColorStringWithCrazyLegacyRules):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90141 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Eric Seidel <eric@webkit.org>
eric@webkit.org [Thu, 30 Jun 2011 18:56:27 +0000 (18:56 +0000)]
2011-06-30  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Remove redundant tests in TestFailuresTest
        https://bugs.webkit.org/show_bug.cgi?id=63742

        These tests are all already covered by TestFailuresTest.test_loads.
        I also added the missing "T" from TestFailuresest. :)

        * Scripts/webkitpy/layout_tests/layout_package/test_failures_unittest.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90140 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Tab Atkins <jackalmage@gmail.com>
commit-queue@webkit.org [Thu, 30 Jun 2011 18:41:30 +0000 (18:41 +0000)]
2011-06-30  Tab Atkins  <jackalmage@gmail.com>

        Reviewed by Adam Barth.

        Fix legacy color attribute parsing to match HTML spec
        https://bugs.webkit.org/show_bug.cgi?id=63029

        Tests <body bgcolor> parsing.

        * fast/dom/attribute-legacy-colors-expected.txt: Added.
        * fast/dom/attribute-legacy-colors.html: Added.
        * fast/dom/script-tests/attribute-legacy-colors.js: Added.
2011-06-30  Tab Atkins  <jackalmage@gmail.com>

        Reviewed by Adam Barth.

        Fix legacy color attribute parsing to match HTML spec
        https://bugs.webkit.org/show_bug.cgi?id=63029

link: http://www.whatwg.org/specs/web-apps/current-work/complete/common-microsyntaxes.html#rules-for-parsing-a-legacy-color-value
        Fix legacy color attribute parsing (<body bgcolor>, <font color>, etc.) to match the HTML spec and more closely match other browsers.

        Test: fast/dom/attribute-legacy-colors.html

        * dom/StyledElement.cpp:
        (WebCore::StyledElement::addCSSColor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90139 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Thu, 30 Jun 2011 18:37:16 +0000 (18:37 +0000)]
2011-06-30  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        REGRESSION(r84294): new-run-webkit-tests results.html generate links to diffs.html or diff.png that don't exist
        https://bugs.webkit.org/show_bug.cgi?id=63494

        This patch does two things:

        1) Refactor diff_image to return the image diff instead of writing the
        diff to a file.  This fixes a bunch of hacks and disentangles a bunch
        of code.

        2) When there's a checksum mismatch but not image diff, we no longer
        report an IMAGE failure to results.html.  That fixes the bug in
        question because results.html won't try to link to a non-existent image
        diff.

        * Scripts/webkitpy/common/system/executive_mock.py:
        * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:
        * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:
        * Scripts/webkitpy/layout_tests/port/base.py:
        * Scripts/webkitpy/layout_tests/port/chromium.py:
        * Scripts/webkitpy/layout_tests/port/chromium_unittest.py:
        * Scripts/webkitpy/layout_tests/port/port_testcase.py:
        * Scripts/webkitpy/layout_tests/port/test.py:
        * Scripts/webkitpy/layout_tests/port/webkit.py:
        * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:
        * Scripts/webkitpy/to_be_moved/rebaseline_chromium_webkit_tests.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90138 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-28 Hans Wennborg <hans@chromium.org>
hans@chromium.org [Thu, 30 Jun 2011 18:32:30 +0000 (18:32 +0000)]
2011-06-28  Hans Wennborg  <hans@chromium.org>

        Reviewed by Tony Chang.

        IndexedDB: Prepare for running layout tests with LevelDB
        https://bugs.webkit.org/show_bug.cgi?id=63593

        Migration of data from SQLite to LevelDB must be done before the
        m_backingStore map is checked for an open LevelDB, because if a
        previous layout test has used LevelDB, it will be in the
        m_backingStore map.

        Covered by existing layout tests.

        * storage/IDBFactoryBackendImpl.cpp:
        (WebCore::IDBFactoryBackendImpl::open):
2011-06-28  Hans Wennborg  <hans@chromium.org>

        Reviewed by Tony Chang.

        IndexedDB: Prepare for running layout tests with LevelDB
        https://bugs.webkit.org/show_bug.cgi?id=63593

        WebIDBFactory should decide what DefaultBackingStore means.
        Also fix handling of layout test and incognito mode properly.

        * src/WebIDBFactoryImpl.cpp:
        (WebKit::WebIDBFactoryImpl::open):
2011-06-28  Hans Wennborg  <hans@chromium.org>

        Reviewed by Tony Chang.

        IndexedDB: Prepare for running layout tests with LevelDB
        https://bugs.webkit.org/show_bug.cgi?id=63593

        The TestShell must always provide a temporary folder for LevelDB,
        not just when the backing store type is overridden.

        * DumpRenderTree/chromium/LayoutTestController.cpp:
        (LayoutTestController::setOverrideIndexedDBBackingStore):
        (LayoutTestController::clearAllDatabases):
        * DumpRenderTree/chromium/LayoutTestController.h:
        * DumpRenderTree/chromium/TestShell.cpp:
        (TestShell::TestShell):
        * DumpRenderTree/chromium/TestShell.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90137 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Thu, 30 Jun 2011 18:22:00 +0000 (18:22 +0000)]
2011-06-30  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Yury Semikhatsky.

        Web Inspector: introduce "filter" icons for timeline panel and elements sidebar.
        https://bugs.webkit.org/show_bug.cgi?id=63731

        * inspector/front-end/EventListenersSidebarPane.js:
        * inspector/front-end/Images/paneFilterButtons.png: Added.
        * inspector/front-end/Images/statusbarButtonGlyphs.png:
        * inspector/front-end/StylesSidebarPane.js:
        * inspector/front-end/inspector.css:
        (.pane > .title > select):
        (.pane > .title > select.select-settings):
        (.pane > .title > select.select-filter):
        (.timeline-filter-status-bar-item .glyph):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90136 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-28 Hans Wennborg <hans@chromium.org>
hans@chromium.org [Thu, 30 Jun 2011 18:16:31 +0000 (18:16 +0000)]
2011-06-28  Hans Wennborg  <hans@chromium.org>

        Reviewed by Tony Chang.

        IndexedDB: Support for deleteIndex with LevelDB backing store
        https://bugs.webkit.org/show_bug.cgi?id=63539

        Covered by storage/indexeddb/mozilla/remove-index.html.

        * storage/IDBLevelDBBackingStore.cpp:
        (WebCore::IDBLevelDBBackingStore::deleteIndex):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90135 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Nate Chapin <japhet@chromium.org>
japhet@chromium.org [Thu, 30 Jun 2011 17:57:47 +0000 (17:57 +0000)]
2011-06-30  Nate Chapin  <japhet@chromium.org>

        Unreviewed, remove passing tests from chromium test expectations.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90134 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoReinstate WKSecurityOriginGetHost and WKSecurityOriginGetProtocol as wrappers around...
mrowe@apple.com [Thu, 30 Jun 2011 17:38:49 +0000 (17:38 +0000)]
Reinstate WKSecurityOriginGetHost and WKSecurityOriginGetProtocol as wrappers around the correctly-named functions.

Reviewed by Anders Carlsson.

They're needed for ABI compatibility.

* Shared/API/c/WKSecurityOrigin.cpp:
(WKSecurityOriginGetHost):
(WKSecurityOriginGetProtocol):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90133 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Thu, 30 Jun 2011 17:24:26 +0000 (17:24 +0000)]
2011-06-30  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        [reviewtool] Expanding lines "below" does not remove purple context line
        https://bugs.webkit.org/show_bug.cgi?id=63695

        Honestly, this is the last bug in the review tool that bothers me.

        * PrettyPatch/PrettyPatch.rb:
            - Jump straight to version 42, because, well, you know.
        * code-review.js:
            - In addition to removing the context bar when we expand all, we
              now remove it when we expand below (but not when we expand
              above).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90132 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoANGLE shouldn't try to use internal SDKs. It doesn't need them!
mrowe@apple.com [Thu, 30 Jun 2011 17:04:46 +0000 (17:04 +0000)]
ANGLE shouldn't try to use internal SDKs. It doesn't need them!

Rubber-stamped by Dan Bernstein.

* Configurations/Base.xcconfig:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90131 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Abhishek Arya <inferno@chromium.org>
inferno@chromium.org [Thu, 30 Jun 2011 17:00:48 +0000 (17:00 +0000)]
2011-06-30  Abhishek Arya  <inferno@chromium.org>

        Reviewed by Ryosuke Niwa.

        Crash when calling DOMSubtreeModified event when extracting range
        contents.
        https://bugs.webkit.org/show_bug.cgi?id=63650

        Convert a few nodes to RefPtrs and add commonRoot verification checks
        for Range::processContents.

        Tests: fast/dom/Range/range-extract-contents-event-fire-crash.html
               fast/dom/Range/range-extract-contents-event-fire-crash2.html

        * dom/Range.cpp:
        (WebCore::childOfCommonRootBeforeOffset):
        (WebCore::Range::processContents):
        (WebCore::Range::processContentsBetweenOffsets):
        (WebCore::Range::processAncestorsAndTheirSiblings):
2011-06-29  Abhishek Arya  <inferno@chromium.org>

        Reviewed by Ryosuke Niwa.

        Crash when calling DOMSubtreeModified event when extracting range
        contents.
        https://bugs.webkit.org/show_bug.cgi?id=63650

        * fast/dom/Range/range-extract-contents-event-fire-crash-expected.txt: Added.
        * fast/dom/Range/range-extract-contents-event-fire-crash.html: Added.
        * fast/dom/Range/range-extract-contents-event-fire-crash2-expected.txt: Added.
        * fast/dom/Range/range-extract-contents-event-fire-crash2.html: Added.
        * fast/dom/Range/range-extractContents.html: remove the appending of fragment
        in this crasher test since we now refptr the nodes and leftContents will be visible.
        This crasher test does not need to show the extractContents fragment.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90130 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoAdd +[WebView _allowsRoundingHacks]
mitz@apple.com [Thu, 30 Jun 2011 16:53:52 +0000 (16:53 +0000)]
Add +[WebView _allowsRoundingHacks]
https://bugs.webkit.org/show_bug.cgi?id=63734

Reviewed by Adele Peterson.

Source/WebCore:

* WebCore.exp.in: Export TextRun::allowsRoundingHacks().
* platform/graphics/TextRun.cpp:
(WebCore::TextRun::allowsRoundingHacks): Added this accessor.
* platform/graphics/TextRun.h:

Source/WebKit/mac:

* WebView/WebView.mm:
(+[WebView _allowsRoundingHacks]): Added this accessor.
* WebView/WebViewPrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90129 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Nate Chapin <japhet@chromium.org>
japhet@chromium.org [Thu, 30 Jun 2011 16:52:27 +0000 (16:52 +0000)]
2011-06-30  Nate Chapin  <japhet@chromium.org>

        Unreviewed, chromium test expectations update after r90101.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90128 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Patrick Gansterer <paroga@webkit.org>
paroga@webkit.org [Thu, 30 Jun 2011 16:48:48 +0000 (16:48 +0000)]
2011-06-30  Patrick Gansterer  <paroga@webkit.org>

        Reviewed by Adam Roben.

        [WIN] Don't redefine the keyboard codes
        https://bugs.webkit.org/show_bug.cgi?id=63726

        Only define the codes when undefined.
        r64957 fixed this problem for COMPILER(MINGW) but not for OS(WINCE).

        * platform/WindowsKeyboardCodes.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90127 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Yury Semikhatsky <yurys@chromium.org>
yurys@chromium.org [Thu, 30 Jun 2011 16:39:21 +0000 (16:39 +0000)]
2011-06-30  Yury Semikhatsky  <yurys@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: hide "Enabled background events collection" option in context menu until the setting is persisted
        https://bugs.webkit.org/show_bug.cgi?id=63708

        The option has no much sense if it is not preserved between browser restarts.

        * inspector/front-end/NetworkPanel.js:
        (WebInspector.NetworkPanel.prototype._contextMenu):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90126 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoTools: [WK2] Create a general wk2 specific LayoutTests/platform/wk2 platform
ossy@webkit.org [Thu, 30 Jun 2011 16:28:10 +0000 (16:28 +0000)]
Tools: [WK2] Create a general wk2 specific LayoutTests/platform/wk2 platform
qt-wk2, win-wk2, mac-wk2 use this common Skipped list too
https://bugs.webkit.org/show_bug.cgi?id=61785

Patch by Zsolt Fehér <h490267@stud.u-szeged.hu> on 2011-06-30
Reviewed by Csaba Osztrogonác.

* Scripts/old-run-webkit-tests:

LayoutTests: [WK2] Create a general wk2 specific LayoutTests/platform/wk2 platform
Common Skipped list copied from LayoutTests/platform/mac-wk2/Skipped.
https://bugs.webkit.org/show_bug.cgi?id=61785

Patch by Zsolt Fehér <h490267@stud.u-szeged.hu> on 2011-06-30
Reviewed by Csaba Osztrogonác.

* platform/mac-wk2/Skipped:
* platform/wk2/Skipped: Copied from LayoutTests/platform/mac-wk2/Skipped.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90125 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Patrick Gansterer <paroga@webkit.org>
paroga@webkit.org [Thu, 30 Jun 2011 16:05:11 +0000 (16:05 +0000)]
2011-06-30  Patrick Gansterer  <paroga@webkit.org>

        Reviewed by Darin Adler.

        Remove dependnecy on IDL generation for disabled features in JSC bindings
        https://bugs.webkit.org/show_bug.cgi?id=63718

        Move the #if ENABLE() line directly after #include "config.h".
        This enables us to compile all files even when we don't generate the IDL output.

        * bindings/js/IDBBindingUtilities.cpp:
        * bindings/js/JSConsoleCustom.cpp:
        * bindings/js/JSCustomSQLStatementErrorCallback.cpp:
        * bindings/js/JSCustomXPathNSResolver.cpp:
        * bindings/js/JSDOMApplicationCacheCustom.cpp:
        * bindings/js/JSDeviceMotionEventCustom.cpp:
        * bindings/js/JSDeviceOrientationEventCustom.cpp:
        * bindings/js/JSInjectedScriptHostCustom.cpp:
        * bindings/js/JSInjectedScriptManager.cpp:
        * bindings/js/JSInspectorFrontendHostCustom.cpp:
        * bindings/js/JSJavaScriptCallFrameCustom.cpp:
        * bindings/js/JSSQLResultSetRowListCustom.cpp:
        * bindings/js/JSSQLTransactionCustom.cpp:
        * bindings/js/JSSQLTransactionSyncCustom.cpp:
        * bindings/js/JSSVGElementInstanceCustom.cpp:
        * bindings/js/JSStorageCustom.cpp:
        * bindings/js/JSTouchCustom.cpp:
        * bindings/js/JSTouchListCustom.cpp:
        * bindings/js/JSXPathResultCustom.cpp:
        * bindings/js/JavaScriptCallFrame.cpp:
        * bindings/js/PageScriptDebugServer.cpp:
        * bindings/js/ScriptDebugServer.cpp:
        * bindings/js/ScriptGCEvent.cpp:
        * bindings/js/ScriptProfile.cpp:
        * bindings/js/WorkerScriptDebugServer.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90124 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoUse a colon to separate test names from failure info on TestFailures
aroben@apple.com [Thu, 30 Jun 2011 15:30:52 +0000 (15:30 +0000)]
Use a colon to separate test names from failure info on TestFailures

This matches how the "Possibly Flaky Tests" list works, and is much prettier when the
failure info itself includes parentheses.

Fixes <http://webkit.org/b/63722> Crashing symbols on TestFailures page look ugly due to
double parentheses

Reviewed by Darin Adler.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForFailedTest): Use a colon instead of surrounding the failure
info in parentheses.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90123 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoRobustify PersistentCache against cached data that contains the string ': '
aroben@apple.com [Thu, 30 Jun 2011 15:30:43 +0000 (15:30 +0000)]
Robustify PersistentCache against cached data that contains the string ': '

Fixes <http://webkit.org/b/63721> Exceptions thrown while loading TestFailures page due to
passing a bad string to JSON.parse

Reviewed by Darin Adler.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/PersistentCache.js:
(PersistentCache._parseDateAndJSONFromString): Manually split the string on the separator
using indexOf and substring. (String.prototype.split always splits on all instances of the
separator.)

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90122 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoInclude the crashing symbol in crash logs links on TestFailures
aroben@apple.com [Thu, 30 Jun 2011 15:30:32 +0000 (15:30 +0000)]
Include the crashing symbol in crash logs links on TestFailures

Fixes <http://webkit.org/b/63465> Links to crash logs on TestFailures page should include
the crashing symbol (like links in results.html do)

Reviewed by David Kilzer.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
(LayoutTestResultsLoader.prototype.start): Bumped the cache version because we now store
crashing symbols for crashing tests. Renamed testsForResultTable to parseResultTable because
it now returns more than just the test names. Specifically, it now looks for crash log links
and extracts the crashing symbol name from them. Updated callers of parseResultTable to
match its new behavior. Changed to store the crashing symbol along with the failure type in
the data we pass to the callback for tests which crashed.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
(code): Make <code> elements a little smaller because their contents can be quite long.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForFailureDiagnosis): Include the crashing symbol inside a
<code> element in the link, if there is a crashing symbol.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90121 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoUse objects instead of strings to represent a test result in TestFailures code
aroben@apple.com [Thu, 30 Jun 2011 15:30:18 +0000 (15:30 +0000)]
Use objects instead of strings to represent a test result in TestFailures code

This will eventually allow us to store more than just the type of failure for each test.
(E.g., we can store the name of the crashing symbol for tests which crashed.)

Prep work for <http://webkit.org/b/63465> Links to crash logs on TestFailures page should
include the crashing symbol (like links in results.html do)

Reviewed by David Kilzer.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Builder.js:
(Builder.prototype.failureDiagnosisTextAndURL): Changed to expect a testResult object
instead of just a failureType string.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/FlakyLayoutTestDetector.js:
(FlakyLayoutTestDetector.prototype.incorporateTestResults): Changed to store a
testResult-like object for passing tests.
(FlakyLayoutTestDetector.prototype.flakinessExamples): Changed to expect testResult-like
objects.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestResultsLoader.js:
(LayoutTestResultsLoader.prototype.start): Store a version number along with the cached data
so we can throw away cached data that's in an old format. Store a testResult object for each
test instead of just its failure type.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForFailedTest):
(ViewController.prototype._domForFailureDiagnosis):
Changed to expect testResult objects instead of failureType strings.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90120 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Alexander Pavlov <apavlov@chromium.org>
apavlov@chromium.org [Thu, 30 Jun 2011 14:45:21 +0000 (14:45 +0000)]
2011-06-30  Alexander Pavlov  <apavlov@chromium.org>

        Reviewed by Kent Tamura.

        [Chromium] Autofill suggestions appear in upper left corner after input change
        https://bugs.webkit.org/show_bug.cgi?id=63702

        This change fixes a few popup layout issues, some of which have existed for quite a while:
        - the absence of the ChromeClientImpl::screenToWindow() implementation (the method used to always return (0, 0));
        - the confusion of window and screen coordinates passed into the autofill popup client's setWindowRect() method;
        - the use of the current frameRect() width as the target element's width (which was wrong when refreshing
          a popup resized during the initial layout (e.g. made wider than the target element)).

        No new tests, as this is a change to platform-specific widget code. The test case is provided in the bug description.

        * platform/chromium/PopupMenuChromium.cpp:
        (WebCore::PopupContainer::layoutAndGetRTLOffset): Use original target element width, not frameRect().width(),
        since the latter gets updated if the popup is not the same width as the target element.
        (WebCore::PopupContainer::refresh): Restore only the popup bounds rather than its original rectangle,
        since it sometimes breaks the invalidation region and layout.
2011-06-30  Alexander Pavlov  <apavlov@chromium.org>

        Reviewed by Kent Tamura.

        [Chromium] Autofill suggestions appear in upper left corner after input change
        https://bugs.webkit.org/show_bug.cgi?id=63702

        * src/ChromeClientImpl.cpp:
        (WebKit::ChromeClientImpl::screenToWindow): Implemented.
        * src/WebViewImpl.cpp:
        (WebKit::WebViewImpl::refreshAutoFillPopup): Pass screen (not window) coordinates into setWindowRect().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90119 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoShow full commit logs when visiting Trac from TestFailures page
aroben@apple.com [Thu, 30 Jun 2011 14:30:19 +0000 (14:30 +0000)]
Show full commit logs when visiting Trac from TestFailures page

Fixes <http://webkit.org/b/63715> Links to Trac on TestFailures page
should show full commit logs

Reviewed by David Kilzer.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/Trac.js:
(Trac.prototype.logURL): Added new 'showFullCommitLogs' parameter. When specified, we add
verbose=on to the log URL to tell Trac to show full commit logs.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype._domForRegressionRange): Specify true for showFullCommitLogs.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90118 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Sheriff Bot <webkit.review.bot@gmail.com>
commit-queue@webkit.org [Thu, 30 Jun 2011 14:11:21 +0000 (14:11 +0000)]
2011-06-30  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r90102.
        http://trac.webkit.org/changeset/90102
        https://bugs.webkit.org/show_bug.cgi?id=63714

        Lots of tests asserting beneath
        SVGSMILElement::findInstanceTime (Requested by aroben on
        #webkit).

        * wtf/StdLibExtras.h:
        (WTF::binarySearch):
2011-06-30  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r90102.
        http://trac.webkit.org/changeset/90102
        https://bugs.webkit.org/show_bug.cgi?id=63714

        Lots of tests asserting beneath
        SVGSMILElement::findInstanceTime (Requested by aroben on
        #webkit).

        * svg/animation/SVGSMILElement.cpp:
        (WebCore::SVGSMILElement::findInstanceTime):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90117 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Andrey Kosyakov <caseq@chromium.org>
caseq@chromium.org [Thu, 30 Jun 2011 14:03:29 +0000 (14:03 +0000)]
2011-06-30  Andrey Kosyakov  <caseq@chromium.org>

        Reviewed by Yury Semikhatsky.

        Web Inspector: remove ExtensionAPISchema.json
        https://bugs.webkit.org/show_bug.cgi?id=63687

        * inspector/front-end/ExtensionAPISchema.json: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90116 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoAdd a link to file a bug about TestFailures to the TestFailures page
aroben@apple.com [Thu, 30 Jun 2011 14:00:30 +0000 (14:00 +0000)]
Add a link to file a bug about TestFailures to the TestFailures page

Fixes <http://webkit.org/b/63466> TestFailures page should include a link to file a bug
about TestFailures itself

Reviewed by Anders Carlsson.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/TestFailures.css:
(aside): Added some styles for the new link. It is positioned in the upper-right and
hopefully not to glaringly obvious.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController.prototype.loaded): Set up our main content element and auxiliary UI.

(ViewController.prototype._displayBuilder):
(ViewController.prototype._displayTesters):
Modify the main content element instead of the entire body, as there are other things in the
body that we want to preserve.

(ViewController.prototype._domForAuxiliaryUIElements): Create and return an <aside> element
that contains a link to file a bug about the TestFailures page.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90115 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoIgnore callbacks that correspond to old navigations of the TestFailures page
aroben@apple.com [Thu, 30 Jun 2011 14:00:16 +0000 (14:00 +0000)]
Ignore callbacks that correspond to old navigations of the TestFailures page

Fixes <http://webkit.org/b/63711> TestFailures page sometimes ignores back/forward
navigations

Reviewed by Anders Carlsson.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/LayoutTestHistoryAnalyzer.js:
(LayoutTestHistoryAnalyzer.prototype.start): Check the return value of the callback. If the
callback returns false, stop fetching more data.

* BuildSlaveSupport/build.webkit.org-config/public_html/TestFailures/ViewController.js:
(ViewController): Initialize new _navigationID member. This member is used to assign each
navigation by the user a unique ID.
(ViewController.prototype.parseHash): Increment the navigation ID.

(ViewController.prototype._displayBuilder):
(ViewController.prototype._displayTesters):
At the start of these functions, record the current navigation ID. Bail out from callbacks
if the navigation ID has changed, indicating that the user has navigated elsewhere.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90114 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Fabrizio Machado <fabrizio.machado@nokia.com>
commit-queue@webkit.org [Thu, 30 Jun 2011 13:59:41 +0000 (13:59 +0000)]
2011-06-30  Fabrizio Machado  <fabrizio.machado@nokia.com>

        Reviewed by Benjamin Poulain.

        [Qt] QML Webview causes performance drops
        https://bugs.webkit.org/show_bug.cgi?id=57554

        Remove tiledBackingStoreEnabled from QML WebView.

        * declarative/qdeclarativewebview.cpp:
        (QDeclarativeWebView::setPage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90113 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Ilya Tikhonovsky <loislo@chromium.org>
loislo@chromium.org [Thu, 30 Jun 2011 13:43:59 +0000 (13:43 +0000)]
2011-06-30  Ilya Tikhonovsky  <loislo@chromium.org>

        Unreviewed build fix for chromium.

        * inspector/CodeGeneratorInspector.pm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90112 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Ilya Tikhonovsky <loislo@chromium.org>
loislo@chromium.org [Thu, 30 Jun 2011 13:38:00 +0000 (13:38 +0000)]
2011-06-30  Ilya Tikhonovsky  <loislo@chromium.org>

        Unreviewed build fix for mac.

        * inspector/CodeGeneratorInspector.pm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90111 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Ilya Tikhonovsky <loislo@chromium.org>
loislo@chromium.org [Thu, 30 Jun 2011 13:22:29 +0000 (13:22 +0000)]
2011-06-30  Ilya Tikhonovsky  <loislo@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: Protocol: error object needs some improvement.
        https://bugs.webkit.org/show_bug.cgi?id=63707

        Current implementation is not useful in the code.
        We need to have toString implementation with nice formatting
        I'd like to use error.message property as the main place for the error's text.

        * inspector/elements/set-attribute.html:
        * inspector/report-protocol-errors-expected.txt:
        * inspector/report-protocol-errors.html:
2011-06-30  Ilya Tikhonovsky  <loislo@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: Protocol: error object needs some improvement.
        https://bugs.webkit.org/show_bug.cgi?id=63707

        Current implementation is not useful in the code.
        We need to have toString implementation with nice formatting
        I'd like to use error.message property as the main place for the error's text.

        * inspector/CodeGeneratorInspector.pm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90110 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Carlos Garcia Campos <cgarcia@igalia.com>
carlosgc@webkit.org [Thu, 30 Jun 2011 13:02:47 +0000 (13:02 +0000)]
2011-06-30  Carlos Garcia Campos  <cgarcia@igalia.com>

        Reviewed by Martin Robinson.

        [GTK] Add back/forward menu to MiniBrowser toolbar
        https://bugs.webkit.org/show_bug.cgi?id=63445

        * MiniBrowser/gtk/BrowserWindow.c:
        (browser_window_init): Use a GtkMenuToolButton for navigation
        widgets.
        (browserWindowConstructed): Get the page BackForwardList.
        (browserWindowHistoryItemActivated): Go to activated history item.
        (browserWindowHistoryItemSelected): Show the url of currently
        selected history item in the statusbar.
        (createGtkActionFromBackForwardItem): Create a GtkAction from a
        WKBackForwardListItemRef.
        (browserWindowCreateMenuItemFromBackForwardItem): Create a
        GtkMenuItem from a WKBackForwardListItemRef.
        (browserWindowCreateBackForwardMenu): Create a GtkMenu with the
        given history items.
        (browserWindowUpdateNavigationActions): Update navigation widgets
        sensitivity and history menus.
        (didChangeBackForwardList): Call
        browserWindowUpdateNavigationActions().
        (browserWindowLoaderClientInit): Add didChangeBackForwardList
        callback.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90109 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Dmitry Lomov <dslomov@google.com>
commit-queue@webkit.org [Thu, 30 Jun 2011 12:55:39 +0000 (12:55 +0000)]
2011-06-30  Dmitry Lomov  <dslomov@google.com>

        Reviewed by Adam Roben.

        https://bugs.webkit.org/show_bug.cgi?id=63660
        This enables unit-tests step on Windows and MacOS SnowLeopard.

        * BuildSlaveSupport/build.webkit.org-config/master.cfg:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90108 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Mikhail Naganov <mnaganov@chromium.org>
mnaganov@chromium.org [Thu, 30 Jun 2011 12:10:13 +0000 (12:10 +0000)]
2011-06-30  Mikhail Naganov  <mnaganov@chromium.org>

        Reviewed by Yury Semikhatsky.

        Web Inspector: [Chromium] Support annotations for arrays in heap profiler.
        https://bugs.webkit.org/show_bug.cgi?id=63690

        * inspector/front-end/DetailedHeapshotGridNodes.js:
        (WebInspector.HeapSnapshotGenericObjectNode.prototype.get data):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90107 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Patrick Gansterer <paroga@webkit.org>
paroga@webkit.org [Thu, 30 Jun 2011 11:41:14 +0000 (11:41 +0000)]
2011-06-30  Patrick Gansterer  <paroga@webkit.org>

        Unreviewed build fix for !ENABLE(SVG_ANIMATION) after r88663, r89187, r89220, r89431 and r89661.

        * svg/SVGAnimatedAngle.h:
        * svg/SVGAnimatedLength.h:
        * svg/SVGAnimatedNumber.h:
        * svg/SVGAnimatedRect.h:
        * svg/SVGAnimatedString.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90106 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Gabor Loki <loki@webkit.org>
loki@webkit.org [Thu, 30 Jun 2011 11:19:40 +0000 (11:19 +0000)]
2011-06-30 Gabor Loki <loki@webkit.org>

        Reviewed by Csaba Osztrogonác.

        Fixed several style and typo issues after r90091.
        https://bugs.webkit.org/show_bug.cgi?id=63692

        Rename yStart to extremaStartY during computing extremas

        * platform/graphics/filters/FEMorphology.cpp:
        (WebCore::FEMorphology::platformApplyGeneric):
        * platform/graphics/filters/FEMorphology.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90105 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago Unreviewed, skip a failing media test on GTK because of
philn@webkit.org [Thu, 30 Jun 2011 11:08:57 +0000 (11:08 +0000)]
    Unreviewed, skip a failing media test on GTK because of
        https://bugs.webkit.org/show_bug.cgi?id=63699

        * platform/gtk/Skipped: skip media/media-blocked-by-willsendrequest.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90104 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Kent Tamura <tkent@chromium.org>
tkent@chromium.org [Thu, 30 Jun 2011 11:00:40 +0000 (11:00 +0000)]
2011-06-30  Kent Tamura  <tkent@chromium.org>

        [Chromium] Resolve a conflict.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90103 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Oliver Varga <Varga.Oliver@stud.u-szeged.hu>
commit-queue@webkit.org [Thu, 30 Jun 2011 10:56:49 +0000 (10:56 +0000)]
2011-06-30  Oliver Varga  <Varga.Oliver@stud.u-szeged.hu>

        Reviewed by Nikolas Zimmermann.

        Speed up SVGSMILElement::findInstanceTime.
        https://bugs.webkit.org/show_bug.cgi?id=61025

        Add a new parameter to StdlibExtras.h::binarySerarch function
        to also handle cases when the array does not contain the key value.
        This is needed for an svg function.

        * wtf/StdLibExtras.h:
        (WTF::binarySearch):
2011-06-30  Oliver Varga  <Varga.Oliver@stud.u-szeged.hu>

        Reviewed by Nikolas Zimmermann.

        Speed up SVGSMILElement::findInstanceTime.
        https://bugs.webkit.org/show_bug.cgi?id=61025

        Replace the linear search to binary search on ordered list because
        the previous searches from the beginning was not efficient.

        No new tests this is only a performance tweak.

        * svg/animation/SVGSMILElement.cpp:
        (WebCore::extractTimeFromVector):
        (WebCore::SVGSMILElement::findInstanceTime):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90102 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Kentaro Hara <haraken@google.com>
commit-queue@webkit.org [Thu, 30 Jun 2011 10:40:32 +0000 (10:40 +0000)]
2011-06-30  Kentaro Hara  <haraken@google.com>

        Reviewed by Kent Tamura.

        Change the label of an HTML5 file chooser button to "Choose Files"
        https://bugs.webkit.org/show_bug.cgi?id=49245

        We should notify capability of multiple files to users.
        input-file-label.html is a test that checks a label of the file chooser
        with/without 'multiple' attribute. input-file-label.html will fail in Chromium
        until a planned Chromium patch is landed. input-file-re-render.html will
        also fail in Chromium, gtk and MacLeopard until we update the expected images
        for the test.

        * fast/forms/input-file-label-expected.txt: Added.
        * fast/forms/input-file-label.html: Added.
        * platform/chromium/test_expectations.txt:
        * platform/mac/fast/forms/input-file-re-render-expected.png:
        * platform/mac/fast/forms/input-file-re-render-expected.txt:
2011-06-30  Kentaro Hara  <haraken@google.com>

        Reviewed by Kent Tamura.

        Change the label of an HTML5 file chooser button to "Choose Files"
        https://bugs.webkit.org/show_bug.cgi?id=49245

        We should notify capability of multiple files to users.

        Test: fast/forms/input-file-label.html

        * English.lproj/Localizable.strings:
        * html/FileInputType.cpp:
        (WebCore::UploadButtonElement::createForMultiple): Constructs a multiple file chooser button.
        (WebCore::FileInputType::createShadowSubtree): Constructs a single/multiple file chooser button depending on 'multiple' attribute.
        (WebCore::FileInputType::multipleAttributeChanged): Updates the label of the file chooser button.
        * html/FileInputType.h:
        * html/HTMLInputElement.cpp:
        (WebCore::HTMLInputElement::parseMappedAttribute): Updates the shadow tree of the file chooser button when 'multiple' attribute is set or unset.
        * html/InputType.cpp:
        (WebCore::InputType::multipleAttributeChanged): Virtual method to be overridden by FileInputType.
        * html/InputType.h:
        * platform/DefaultLocalizationStrategy.cpp:
        (WebCore::DefaultLocalizationStrategy::fileButtonChooseFileLabel): Updated the description for "Choose File".
        (WebCore::DefaultLocalizationStrategy::fileButtonChooseMultipleFilesLabel): Returns a string "Choose Files".
        * platform/DefaultLocalizationStrategy.h:
        * platform/LocalizationStrategy.h:
        * platform/LocalizedStrings.cpp:
        (WebCore::fileButtonChooseMultipleFilesLabel): Wrapper method for getting the "Choose Files" label.
        * platform/LocalizedStrings.h:
        * platform/brew/LocalizedStringsBrew.cpp:
        (WebCore::fileButtonChooseMultipleFilesLabel): Returns a "Choose Files" label.
        * platform/efl/LocalizedStringsEfl.cpp:
        (WebCore::fileButtonChooseMultipleFilesLabel): Returns a "Choose Files" label.
        * platform/gtk/LocalizedStringsGtk.cpp:
        (WebCore::fileButtonChooseMultipleFilesLabel): Returns a "Choose Files" label.
        * platform/haiku/LocalizedStringsHaiku.cpp:
        (WebCore::fileButtonChooseMultipleFilesLabel): Returns a "Choose Files" label.
        * platform/wx/LocalizedStringsWx.cpp:
        (WebCore::fileButtonChooseMultipleFilesLabel): Returns a "Choose Files" label.
2011-06-30  Kentaro Hara  <haraken@google.com>

        Reviewed by Kent Tamura.

        Change the label of an HTML5 file chooser button to "Choose Files"
        https://bugs.webkit.org/show_bug.cgi?id=49245

        We should notify capability of multiple files to users.

        Test: fast/forms/input-file-label.html

        * DEPS:
        * public/WebLocalizedString.h: Depends on chromium_rev>=91051.
        * src/LocalizedStrings.cpp:
        (WebCore::fileButtonChooseMultipleFilesLabel): Returns a "Choose Files" label.
2011-06-30  Kentaro Hara  <haraken@google.com>

        Reviewed by Kent Tamura.

        Change the label of an HTML5 file chooser button to "Choose Files"
        https://bugs.webkit.org/show_bug.cgi?id=49245

        We should notify capability of multiple files to users.

        Test: fast/forms/input-file-label.html

        * WebCoreSupport/WebPlatformStrategies.cpp:
        (WebPlatformStrategies::fileButtonChooseFileLabel): Updated the description for "Choose File".
        (WebPlatformStrategies::fileButtonChooseMultipleFilesLabel):  Returns a "Choose Files" label.
        * WebCoreSupport/WebPlatformStrategies.h:
2011-06-30  Kentaro Hara  <haraken@google.com>

        Reviewed by Kent Tamura.

        Change the label of an HTML5 file chooser button to "Choose Files"
        https://bugs.webkit.org/show_bug.cgi?id=49245

        We should notify capability of multiple files to users.

        Test: fast/forms/input-file-label.html

        * WebCoreSupport/PlatformStrategiesWinCE.cpp:
        (PlatformStrategiesWinCE::fileButtonChooseFileLabel): Updated the description for "Choose File".
        (PlatformStrategiesWinCE::fileButtonChooseMultipleFilesLabel): Returns a "Choose Files" label.
        * WebCoreSupport/PlatformStrategiesWinCE.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90101 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Yuta Kitamura <yutak@chromium.org>
yutak@chromium.org [Thu, 30 Jun 2011 09:31:01 +0000 (09:31 +0000)]
2011-06-30  Yuta Kitamura  <yutak@chromium.org>

        Reviewed by Kent Tamura.

        WebSocket: Don't use script-tests for WebSocket tests
        https://bugs.webkit.org/show_bug.cgi?id=63691

        Move content of script-tests/XXX.js into XXX.html.

        No changes were made to script code. Hence, these tests are not using
        window.jsTestAsync and finishJSTest() for now. I will fix this later.

        * http/tests/websocket/tests/alert-in-event-handler.html:
        * http/tests/websocket/tests/bad-handshake-crash.html:
        * http/tests/websocket/tests/bad-sub-protocol-control-chars.html:
        * http/tests/websocket/tests/bad-sub-protocol-empty.html:
        * http/tests/websocket/tests/bad-sub-protocol-non-ascii.html:
        * http/tests/websocket/tests/bufferedAmount-after-close.html:
        * http/tests/websocket/tests/cross-origin.html:
        * http/tests/websocket/tests/error-detect.html:
        * http/tests/websocket/tests/frame-length-longer-than-buffer.html:
        * http/tests/websocket/tests/frame-length-skip.html:
        * http/tests/websocket/tests/handshake-challenge-randomness.html:
        * http/tests/websocket/tests/handshake-error.html:
        * http/tests/websocket/tests/handshake-fail-by-cross-origin.html:
        * http/tests/websocket/tests/handshake-fail-by-no-cr.html:
        * http/tests/websocket/tests/handshake-fail-by-sub-protocol-mismatch.html:
        * http/tests/websocket/tests/long-invalid-header.html:
        * http/tests/websocket/tests/null-character.html:
        * http/tests/websocket/tests/script-tests/TEMPLATE.html: Removed.
        * http/tests/websocket/tests/script-tests/alert-in-event-handler.js: Removed.
        * http/tests/websocket/tests/script-tests/bad-handshake-crash.js: Removed.
        * http/tests/websocket/tests/script-tests/bad-sub-protocol-control-chars.js: Removed.
        * http/tests/websocket/tests/script-tests/bad-sub-protocol-empty.js: Removed.
        * http/tests/websocket/tests/script-tests/bad-sub-protocol-non-ascii.js: Removed.
        * http/tests/websocket/tests/script-tests/bufferedAmount-after-close.js: Removed.
        * http/tests/websocket/tests/script-tests/cross-origin.js: Removed.
        * http/tests/websocket/tests/script-tests/error-detect.js: Removed.
        * http/tests/websocket/tests/script-tests/frame-length-longer-than-buffer.js: Removed.
        * http/tests/websocket/tests/script-tests/frame-length-skip.js: Removed.
        * http/tests/websocket/tests/script-tests/handshake-challenge-randomness.js: Removed.
        * http/tests/websocket/tests/script-tests/handshake-error.js: Removed.
        * http/tests/websocket/tests/script-tests/handshake-fail-by-cross-origin.js: Removed.
        * http/tests/websocket/tests/script-tests/handshake-fail-by-no-cr.js: Removed.
        * http/tests/websocket/tests/script-tests/handshake-fail-by-sub-protocol-mismatch.js: Removed.
        * http/tests/websocket/tests/script-tests/long-invalid-header.js: Removed.
        * http/tests/websocket/tests/script-tests/null-character.js: Removed.
        * http/tests/websocket/tests/script-tests/send-throw.js: Removed.
        * http/tests/websocket/tests/script-tests/send.js: Removed.
        * http/tests/websocket/tests/script-tests/simple.js: Removed.
        * http/tests/websocket/tests/script-tests/sub-protocol-with-space.js: Removed.
        * http/tests/websocket/tests/script-tests/sub-protocol.js: Removed.
        * http/tests/websocket/tests/script-tests/unicode.js: Removed.
        * http/tests/websocket/tests/script-tests/url-no-trailing-slash.js: Removed.
        * http/tests/websocket/tests/script-tests/url-parsing.js: Removed.
        * http/tests/websocket/tests/script-tests/url-with-credential.js: Removed.
        * http/tests/websocket/tests/script-tests/url-with-empty-query.js: Removed.
        * http/tests/websocket/tests/script-tests/url-with-query-for-no-query.js: Removed.
        * http/tests/websocket/tests/script-tests/url-with-query.js: Removed.
        * http/tests/websocket/tests/script-tests/websocket-event-target.js: Removed.
        * http/tests/websocket/tests/script-tests/websocket-protocol-ignored.js: Removed.
        * http/tests/websocket/tests/send-throw.html:
        * http/tests/websocket/tests/send.html:
        * http/tests/websocket/tests/simple.html:
        * http/tests/websocket/tests/sub-protocol-with-space.html:
        * http/tests/websocket/tests/sub-protocol.html:
        * http/tests/websocket/tests/unicode.html:
        * http/tests/websocket/tests/url-no-trailing-slash.html:
        * http/tests/websocket/tests/url-parsing.html:
        * http/tests/websocket/tests/url-with-credential.html:
        * http/tests/websocket/tests/url-with-empty-query.html:
        * http/tests/websocket/tests/url-with-query-for-no-query.html:
        * http/tests/websocket/tests/url-with-query.html:
        * http/tests/websocket/tests/websocket-event-target.html:
        * http/tests/websocket/tests/websocket-protocol-ignored.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90100 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Yuta Kitamura <yutak@chromium.org>
yutak@chromium.org [Thu, 30 Jun 2011 09:22:49 +0000 (09:22 +0000)]
2011-06-30  Yuta Kitamura  <yutak@chromium.org>

        Reviewed by Kent Tamura.

        WebSocket: Fix style issues of pywebsocket handlers
        https://bugs.webkit.org/show_bug.cgi?id=63689

        Fix minor style errors such as wrong indentation or wrong number of blank lines.

        * http/tests/websocket/tests/bad-handshake-crash_wsh.py:
        * http/tests/websocket/tests/close-on-unload_wsh.py:
        * http/tests/websocket/tests/echo-challenge_wsh.py:
        * http/tests/websocket/tests/echo-cookie_wsh.py:
        * http/tests/websocket/tests/frame-length-longer-than-buffer_wsh.py:
        * http/tests/websocket/tests/frame-length-overflow_wsh.py:
        * http/tests/websocket/tests/frame-length-skip_wsh.py:
        * http/tests/websocket/tests/handshake-fail-by-maxlength_wsh.py:
        Use sequence repetition pattern instead of a loop.
        * http/tests/websocket/tests/handshake-fail-by-no-connection-header_wsh.py:
        * http/tests/websocket/tests/handshake-fail-by-no-cr_wsh.py:
        * http/tests/websocket/tests/handshake-fail-by-no-upgrade-header_wsh.py:
        * http/tests/websocket/tests/handshake-fail-by-prepended-null_wsh.py:
        * http/tests/websocket/tests/hanging-handshake_wsh.py:
        * http/tests/websocket/tests/long-invalid-header_wsh.py:
        * http/tests/websocket/tests/send2_wsh.py:
        * http/tests/websocket/tests/unknown-frame-type_wsh.py:
        * http/tests/websocket/tests/workers/resources/echo-challenge_wsh.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90099 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Ryosuke Niwa <rniwa@webkit.org>
rniwa@webkit.org [Thu, 30 Jun 2011 09:14:17 +0000 (09:14 +0000)]
2011-06-30  Ryosuke Niwa  <rniwa@webkit.org>

        Reviewed by Kent Tamura.

        Replace isNodeInTextFormControl(Node*) by enclosingTextFormControl(const Position&)
        https://bugs.webkit.org/show_bug.cgi?id=63672

        Replaced isNodeInTextFormControl(Node*) by enclosingTextFormControl(const Position&).
        Also added a safe toTextFormControl to HTMLFormControlElement.h.

        * editing/DeleteSelectionCommand.cpp:
        (WebCore::DeleteSelectionCommand::doApply): Calls enclosingTextFormControl instead of manually
        checking the tag name of shadowAncestorNode.
        * editing/Editor.cpp:
        (WebCore::Editor::selectionForCommand): Rewritten using enclosingTextFormControl and toTextFormControl.
        (WebCore::Editor::cut): Calls enclosingTextFormControl instead of isNodeInTextFormControl.
        (WebCore::Editor::copy): Ditto.
        (WebCore::Editor::setBaseWritingDirection): Calls toTextFormControl instead of manually checking tag name.
        * editing/FrameSelection.cpp:
        (WebCore::FrameSelection::isInPasswordField): Calls enclosingTextFormControl.
        * editing/ReplaceSelectionCommand.cpp:
        (WebCore::ReplaceSelectionCommand::doApply): Ditto.
        * editing/htmlediting.cpp:
        (WebCore::enclosingTextFormControl): Added.
        * editing/htmlediting.h:
        * html/HTMLFormControlElement.h:
        (WebCore::toTextFormControl): Added.
        * page/DragController.cpp:
        (WebCore::DragController::startDrag): Calls enclosingTextFormControl instead of isNodeInTextFormControl.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90098 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Kent Tamura <tkent@chromium.org>
tkent@chromium.org [Thu, 30 Jun 2011 09:08:07 +0000 (09:08 +0000)]
2011-06-30  Kent Tamura  <tkent@chromium.org>

        [Mac-wk2] Update expectation files for r90089.
        https://bugs.webkit.org/show_bug.cgi?id=61415

        * platform/mac-wk2/fast/forms/textfield-overflow-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90097 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-27 Pavel Podivilov <podivilov@chromium.org>
podivilov@chromium.org [Thu, 30 Jun 2011 08:18:28 +0000 (08:18 +0000)]
2011-06-27  Pavel Podivilov  <podivilov@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: source code is not highlighted after drawer height animations.
        https://bugs.webkit.org/show_bug.cgi?id=62860

        * inspector/front-end/Drawer.js:
        (WebInspector.Drawer.prototype._animateDrawerHeight.animationFinished):
        (WebInspector.Drawer.prototype._animateDrawerHeight):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90096 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Balazs Kelemen <kbalazs@webkit.org>
kbalazs@webkit.org [Thu, 30 Jun 2011 08:00:31 +0000 (08:00 +0000)]
2011-06-30  Balazs Kelemen  <kbalazs@webkit.org>

        Reviewed by Kenneth Rohde Christiansen.

        [Qt][WK2] Implement the rest of WTR::PlatformWebView
        https://bugs.webkit.org/show_bug.cgi?id=63630

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90095 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 MORITA Hajime <morrita@google.com>
morrita@google.com [Thu, 30 Jun 2011 07:43:39 +0000 (07:43 +0000)]
2011-06-30  MORITA Hajime  <morrita@google.com>

        Reviewed by Dimitri Glazkov.

        Crash if ShadowRoot has a text node.
        https://bugs.webkit.org/show_bug.cgi?id=63607

        * fast/dom/shadow/shadow-root-text-child-expected.txt: Added.
        * fast/dom/shadow/shadow-root-text-child.html: Added.
2011-06-30  MORITA Hajime  <morrita@google.com>

        Reviewed by Dimitri Glazkov.

        Crash if ShadowRoot has a text node.
        https://bugs.webkit.org/show_bug.cgi?id=63607

        Node::styleForRenderer() for Text node wasn't shadow-aware.
        This change allow styleForRenderer() to know its visual parent by
        passing NodeRenderingContext.

        Changes other than NodeRenderingContext and Element is just a
        follow up for change above.

        * dom/Element.cpp: use NodeRenderingContext to resolve parent node.
        (WebCore::Element::recalcStyle):
        * dom/Node.cpp:
        (WebCore::Node::styleForRenderer): Added NodeRenderingContext as a parameter.
        * dom/Node.h:
        * dom/NodeRenderingContext.cpp:
        (WebCore::NodeRendererFactory::createRendererAndStyle):
        * html/HTMLOptGroupElement.cpp:
        (WebCore::HTMLOptGroupElement::attach):
        * html/HTMLOptionElement.cpp:
        (WebCore::HTMLOptionElement::attach):
        * html/HTMLTitleElement.cpp:
        (WebCore::HTMLTitleElement::textWithDirection):
        * html/shadow/TextControlInnerElements.cpp:
        (WebCore::TextControlInnerElement::styleForRenderer):
        (WebCore::TextControlInnerTextElement::styleForRenderer):
        * html/shadow/TextControlInnerElements.h:

        Test: fast/dom/shadow/shadow-root-text-child.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90094 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Leandro Pereira <leandro@profusion.mobi>
commit-queue@webkit.org [Thu, 30 Jun 2011 07:42:09 +0000 (07:42 +0000)]
2011-06-30  Leandro Pereira  <leandro@profusion.mobi>

        Reviewed by Simon Hausmann.

        generate-webkitversion.pl should not "use Switch"
        https://bugs.webkit.org/show_bug.cgi?id=63628

        Switch got deprecated on recent Perl versions and the script wasn't
        using the switch statement.

        * scripts/generate-webkitversion.pl: Remove "use Switch;" line.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90093 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Kent Tamura <tkent@chromium.org>
tkent@chromium.org [Thu, 30 Jun 2011 07:42:07 +0000 (07:42 +0000)]
2011-06-30  Kent Tamura  <tkent@chromium.org>

        [GTK][Mac][Win] Update expectation files for r90089.

        * platform/gtk/fast/block/float/032-expected.txt:
        * platform/gtk/fast/css/input-search-padding-expected.txt:
        * platform/gtk/fast/css/pseudo-cache-stale-expected.txt:
        * platform/gtk/fast/css/text-input-with-webkit-border-radius-expected.txt:
        * platform/gtk/fast/forms/box-shadow-override-expected.txt:
        * platform/gtk/fast/forms/control-restrict-line-height-expected.txt:
        * platform/gtk/fast/forms/input-appearance-height-expected.txt:
        * platform/gtk/fast/forms/input-appearance-spinbutton-disabled-readonly-expected.txt:
        * platform/gtk/fast/forms/input-appearance-spinbutton-layer-expected.txt:
        * platform/gtk/fast/forms/input-appearance-spinbutton-visibility-expected.txt:
        * platform/gtk/fast/forms/input-text-word-wrap-expected.txt:
        * platform/gtk/fast/forms/placeholder-position-expected.txt:
        * platform/gtk/fast/forms/placeholder-pseudo-style-expected.txt:
        * platform/gtk/fast/forms/placeholder-set-value-expected.txt:
        * platform/gtk/fast/forms/search-cancel-button-style-sharing-expected.txt:
        * platform/gtk/fast/forms/search-display-none-cancel-button-expected.txt:
        * platform/gtk/fast/forms/search-placeholder-value-changed-expected.txt:
        * platform/gtk/fast/forms/search-rtl-expected.txt:
        * platform/gtk/fast/forms/search-styled-expected.txt:
        * platform/gtk/fast/forms/search-vertical-alignment-expected.txt:
        * platform/gtk/fast/forms/searchfield-heights-expected.txt:
        * platform/gtk/fast/forms/textfield-overflow-expected.txt:
        * platform/gtk/fast/repaint/search-field-cancel-expected.txt:
        * platform/gtk/fast/replaced/width100percent-searchfield-expected.txt:
        * platform/gtk/fast/table/colspanMinWidth-vertical-expected.txt:
        * platform/mac/fast/block/float/032-expected.txt:
        * platform/win/fast/forms/search-cancel-button-style-sharing-expected.txt:
        * platform/win/fast/forms/search-display-none-cancel-button-expected.txt:
        * platform/win/fast/forms/search-placeholder-value-changed-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90092 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-30 Piroska András <Piroska.Andras@stud.u-szeged.hu>
commit-queue@webkit.org [Thu, 30 Jun 2011 07:20:50 +0000 (07:20 +0000)]
2011-06-30  Piroska András  <Piroska.Andras@stud.u-szeged.hu>

        Reviewed by Dirk Schulze.

        Apply the ParallelJobs support to FEMorphology
        https://bugs.webkit.org/show_bug.cgi?id=63064

        The FEMorphology filter of SVG can consume lots of resources if it is
        applied to a large area. The computation can be distributed to multiple
        cores if the architecture supports.
        The average performance progression is 20-25% on dual-core machines.

        * platform/graphics/filters/FEMorphology.cpp:
        (WebCore::FEMorphology::platformApplyGeneric): Apply the filter between yStart and yEnd interval
        (WebCore::FEMorphology::platformApplyWorker): Call the generic apply from a parallel worker
        (WebCore::FEMorphology::platformApply):
        (WebCore::FEMorphology::apply): Move the functionality into platformApply
        (WebCore::FEMorphology::externalRepresentation): Remove whitespaces
        * platform/graphics/filters/FEMorphology.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90091 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[Qt] Unreviewed. Update platform specific expected files after r90089.
ossy@webkit.org [Thu, 30 Jun 2011 07:14:45 +0000 (07:14 +0000)]
[Qt] Unreviewed. Update platform specific expected files after r90089.

* platform/qt/fast/css/pseudo-cache-stale-expected.png: Added.
* platform/qt/fast/css/pseudo-cache-stale-expected.txt:
* platform/qt/fast/css/text-input-with-webkit-border-radius-expected.png: Added.
* platform/qt/fast/css/text-input-with-webkit-border-radius-expected.txt:
* platform/qt/fast/forms/control-restrict-line-height-expected.png:
* platform/qt/fast/forms/control-restrict-line-height-expected.txt:
* platform/qt/fast/forms/placeholder-set-value-expected.png: Added.
* platform/qt/fast/forms/placeholder-set-value-expected.txt:
* platform/qt/fast/forms/search-cancel-button-style-sharing-expected.png: Added.
* platform/qt/fast/forms/search-cancel-button-style-sharing-expected.txt:
* platform/qt/fast/forms/search-display-none-cancel-button-expected.png: Added.
* platform/qt/fast/forms/search-display-none-cancel-button-expected.txt:
* platform/qt/fast/forms/search-placeholder-value-changed-expected.png: Added.
* platform/qt/fast/forms/search-placeholder-value-changed-expected.txt:
* platform/qt/fast/forms/search-rtl-expected.png: Added.
* platform/qt/fast/forms/search-rtl-expected.txt:
* platform/qt/fast/forms/search-styled-expected.png: Added.
* platform/qt/fast/forms/search-styled-expected.txt:
* platform/qt/fast/forms/search-vertical-alignment-expected.png: Added.
* platform/qt/fast/forms/search-vertical-alignment-expected.txt:
* platform/qt/fast/replaced/width100percent-searchfield-expected.png:
* platform/qt/fast/replaced/width100percent-searchfield-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90090 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Kent Tamura <tkent@chromium.org>
tkent@chromium.org [Thu, 30 Jun 2011 06:30:23 +0000 (06:30 +0000)]
2011-06-29  Kent Tamura  <tkent@chromium.org>

        Reviewed by Dimitri Glazkov.

        Avoid custom layout code of RenderTextControlSingleLine
        https://bugs.webkit.org/show_bug.cgi?id=61415

        Update tests including <input type=search> or <input type=number>. Their
        internal structures are changed.
        Slight position changes for search result buttons and search cancel
        buttons are expected.

        * fast/css/text-input-with-webkit-border-radius-expected.txt:
        * platform/chromium/test_expectations.txt:
        * platform/mac/fast/css/input-search-padding-expected.png:
        * platform/mac/fast/css/input-search-padding-expected.txt:
        * platform/mac/fast/css/pseudo-cache-stale-expected.txt:
        * platform/mac/fast/forms/box-shadow-override-expected.txt:
        * platform/mac/fast/forms/control-restrict-line-height-expected.txt:
        * platform/mac/fast/forms/input-appearance-height-expected.txt:
        * platform/mac/fast/forms/input-appearance-spinbutton-disabled-readonly-expected.txt:
        * platform/mac/fast/forms/input-appearance-spinbutton-expected.txt:
        * platform/mac/fast/forms/input-appearance-spinbutton-layer-expected.txt:
        * platform/mac/fast/forms/input-appearance-spinbutton-up-expected.txt:
        * platform/mac/fast/forms/input-appearance-spinbutton-visibility-expected.txt:
        * platform/mac/fast/forms/placeholder-position-expected.txt:
        * platform/mac/fast/forms/placeholder-pseudo-style-expected.txt:
        * platform/mac/fast/forms/placeholder-set-value-expected.txt:
        * platform/mac/fast/forms/search-cancel-button-style-sharing-expected.txt:
        * platform/mac/fast/forms/search-display-none-cancel-button-expected.txt:
        * platform/mac/fast/forms/search-placeholder-value-changed-expected.txt:
        * platform/mac/fast/forms/search-rtl-expected.txt:
        * platform/mac/fast/forms/search-styled-expected.txt:
        * platform/mac/fast/forms/search-vertical-alignment-expected.png:
        * platform/mac/fast/forms/search-vertical-alignment-expected.txt:
        * platform/mac/fast/forms/searchfield-heights-expected.txt:
        * platform/mac/fast/repaint/search-field-cancel-expected.png:
        * platform/mac/fast/repaint/search-field-cancel-expected.txt:
        * platform/mac/fast/replaced/width100percent-searchfield-expected.txt:
        * platform/mac/fast/table/colspanMinWidth-vertical-expected.png:
        * platform/mac/fast/table/colspanMinWidth-vertical-expected.txt:
2011-06-29  Kent Tamura  <tkent@chromium.org>

        Reviewed by Dimitri Glazkov.

        Avoid custom layout code of RenderTextControlSingleLine
        https://bugs.webkit.org/show_bug.cgi?id=61415

        If a text field has a search results button, a search cancel
        button, a speech button, or a spin button, the editable inner text
        element and the buttons are wrapped by a flexible box container,
        and the editable inner text element has flexibility. As the result
        of applying the flexible box, we don't need any horizontal layout
        code in RenderTextControlSingleLine except the inner spin button.

        Also, we remove custom hit-testing code in nodeAtPoint(). If a
        user clicks on a position left of the inner editable text but not
        on the search results button, the search results button doesn't
        work though it worked.

        * css/html.css:
        (input::-webkit-textfield-decoration-container):
          New style definition for the "container" element.
        (input[type="search"]::-webkit-search-cancel-button):
          It should be a block element because it's a child of the flexible box.
          Also, this doesn't have flexibility.
        (input[type="search"]::-webkit-search-decoration): ditto.
        (input[type="search"]::-webkit-search-results-decoration): ditto.
        (input[type="search"]::-webkit-search-results-button): ditto.
        (input::-webkit-inner-spin-button): ditto.
        (input::-webkit-input-speech-button): ditto.
        * html/HTMLInputElement.cpp:
        (WebCore::HTMLInputElement::containerElement):
          Add an accessor of the flexible box container.
          It just calls InputType::containerElement().
        * html/HTMLInputElement.h: Declare it.
        * html/InputType.h:
        (WebCore::InputType::containerElement): New accessor for a shadow element.
        * html/SearchInputType.cpp:
        (WebCore::SearchInputType::SearchInputType):
          m_innerBlock is moved to TextFieldInputType.
        (WebCore::SearchInputType::needsContainer):
          A search field always needs the flexible box container because
          of the results button and the cancel button.
        (WebCore::SearchInputType::createShadowSubtree):
          Use TextFieldInputType::createShadowSubtree(), then modify the tree.
        (WebCore::SearchInputType::destroyShadowSubtree):
          m_innerBlock is moved to TextFieldInputType.
        * html/SearchInputType.h:
        * html/TextFieldInputType.cpp:
        (WebCore::TextFieldInputType::TextFieldInputType):
          Initialize the new members.
        (WebCore::TextFieldInputType::needsContainer):
          Returns true if x-webkit-speech. We check a spin button separately.
        (WebCore::TextFieldInputType::createShadowSubtree):
          Wrap everything with the flexible box container if needsContainer() or we need a spin button.
        (WebCore::TextFieldInputType::containerElement): Added an accessor.
        (WebCore::TextFieldInputType::innerBlockElement): ditto.
        (WebCore::TextFieldInputType::innerTextElement): ditto.
        (WebCore::TextFieldInputType::innerSpinButtonElement): ditto.
        (WebCore::TextFieldInputType::speechButtonElement): ditto.
        (WebCore::TextFieldInputType::destroyShadowSubtree):
          Clear references to new shadow nodes.
        * html/TextFieldInputType.h:
        * rendering/RenderTextControlSingleLine.cpp:
        (WebCore::RenderTextControlSingleLine::containerElement): Added an accessor.
        (WebCore::RenderTextControlSingleLine::innerTextElement):
        (WebCore::RenderTextControlSingleLine::layout):
         - Correct heights check
          Comparing with height(), not contentHeigth(), was incorrect.
         - Adjust the container height like the editable inner text.
         - Remove width adjustment
         - Remove the speech button layout code
          The flexible box layout is enough.
        (WebCore::RenderTextControlSingleLine::nodeAtPoint):
          Remove custom code for results/cancel/speech buttons, and
          Simplify comments.
        (WebCore::RenderTextControlSingleLine::forwardEvent):
          Remove custom forwarding code for results/cancel buttons.
        (WebCore::RenderTextControlSingleLine::styleDidChange):
          Clear the style of the container because we update it in layout().
        (WebCore::RenderTextControlSingleLine::hasControlClip):
          Decoration buttons should be clipped by the container.
        (WebCore::RenderTextControlSingleLine::controlClipRect): ditto.
        (WebCore::RenderTextControlSingleLine::createInnerTextStyle):
          We don't need to use display:inline-block any more.
        (WebCore::RenderTextControlSingleLine::createInnerBlockStyle):
          The text wrapper element should have flexibility.
        (WebCore::RenderTextControlSingleLine::textBlockInsetTop):
          Updated for the tree structure change.
        * rendering/RenderTextControlSingleLine.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90089 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Kent Tamura <tkent@chromium.org>
tkent@chromium.org [Thu, 30 Jun 2011 05:39:40 +0000 (05:39 +0000)]
2011-06-29  Kent Tamura  <tkent@chromium.org>

        [Mac] Attempt to fix Leopard build.

        * platform/mac/CursorMac.mm:
        (WebCore::Cursor::ensurePlatformCursor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90088 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Darin Adler <darin@apple.com>
darin@apple.com [Thu, 30 Jun 2011 05:12:28 +0000 (05:12 +0000)]
2011-06-29  Darin Adler  <darin@apple.com>

        Reviewed by Dan Bernstein.

        [Mac] Use system cursors instead of custom bitmaps where possible
        https://bugs.webkit.org/show_bug.cgi?id=63679

        * WebCoreSupport/WebSystemInterface.mm:
        (InitWebCoreSystemInterface): Added WKCursor.
2011-06-29  Darin Adler  <darin@apple.com>

        Reviewed by Dan Bernstein.

        [Mac] Use system cursors instead of custom bitmaps where possible
        https://bugs.webkit.org/show_bug.cgi?id=63679

        * Resources/crossHairCursor.png: Removed.
        * Resources/notAllowedCursor.png: Removed.
        * WebCore.exp.in: Added wkCursor.
        * WebCore/WebCore.gyp/WebCore.gyp: Removed the two now-unused PNG files.
        * WebCore.xcodeproj/project.pbxproj: Ditto.

        * platform/mac/CursorMac.mm:
        (WebCore::Cursor::ensurePlatformCursor): Use NSCursor API for Cross, Hand,
        VerticalText, ContextMenu, Copy, and NotAllowed. Use wkCursor for Move,
        Alias, Progress, and the resize and panning cursors.

        * platform/mac/WebCoreSystemInterface.h: Added wkCursor.
        * platform/mac/WebCoreSystemInterface.mm: Ditto.
2011-06-29  Darin Adler  <darin@apple.com>

        Reviewed by Dan Bernstein.

        [Mac] Use system cursors instead of custom bitmaps where possible
        https://bugs.webkit.org/show_bug.cgi?id=63679

        * WebProcess/WebCoreSupport/mac/WebSystemInterface.mm:
        (InitWebCoreSystemInterface): Added WKCursor.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90087 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Dirk Pranke <dpranke@chromium.org>
dpranke@chromium.org [Thu, 30 Jun 2011 04:16:18 +0000 (04:16 +0000)]
2011-06-29  Dirk Pranke  <dpranke@chromium.org>

        Reviewed by Adam Barth.

        test-webkitpy is spamming lots of messages about updating test expectations
        https://bugs.webkit.org/show_bug.cgi?id=63680

        This patch changes the passing_run() function to capture the
        output to stdout and stderr and assert that it is empty, and
        also specifies --no-new-test-results by default to avoid getting
        "generating baseline" messages for the expected-missing results.

        Also check that we are logging messages when we generate new
        baselines.

        Also fix a minor bug in the way test_tolerance() was
        checking for whether the tests were passing or not.

        * Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90086 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Yuta Kitamura <yutak@chromium.org>
yutak@chromium.org [Thu, 30 Jun 2011 03:40:01 +0000 (03:40 +0000)]
2011-06-29  Yuta Kitamura  <yutak@chromium.org>

        Reviewed by Kent Tamura.

        DRT: Make Hixie76WebSocketProtocolEnabled preference flag configurable from LayoutTestController
        https://bugs.webkit.org/show_bug.cgi?id=63532

        LayoutTestController::overridePreference() will be used to change the value of
        the Hixie76WebSocketProtocolEnabled preference flag from layout tests. In this way,
        we can avoid writing a lot of boilerplate code in each port's DumpRenderTree, but
        at least we need to reset preference values in DumpRenderTree every time we start
        a new test because preference values changed in some test should not be leaked
        to the next test.

        * DumpRenderTree/chromium/LayoutTestController.cpp:
        (LayoutTestController::overridePreference):
        The name of preference key is defined in WebKit/{mac/WebView,win}/WebPreferenceKeysPrivate.h.
        * DumpRenderTree/chromium/WebPreferences.cpp:
        (WebPreferences::reset):
        Set the default value. This will be called from TestShell::resetTestController().
        (WebPreferences::applyTo):
        * DumpRenderTree/chromium/WebPreferences.h:
        * DumpRenderTree/mac/DumpRenderTree.mm:
        (resetDefaultsToConsistentValues): Set the default value.
        * DumpRenderTree/win/DumpRenderTree.cpp:
        (resetDefaultsToConsistentValues): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90085 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Thu, 30 Jun 2011 03:34:18 +0000 (03:34 +0000)]
2011-06-29  Simon Fraser  <simon.fraser@apple.com>

        Fix the test description.

        * fast/layers/negative-scroll-positions-expected.txt:
        * fast/layers/script-tests/negative-scroll-positions.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90084 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Kenichi Ishibashi <bashi@chromium.org>
bashi@chromium.org [Thu, 30 Jun 2011 03:07:40 +0000 (03:07 +0000)]
2011-06-29  Kenichi Ishibashi  <bashi@chromium.org>

        Reviewed by Adam Barth.

        check-webkit-style fails to detect https://webkit.org/b/XXXXX as a bug number
        https://webkit.org/b/57360

        Modify the regular expression to match https scheme.

        * Scripts/webkitpy/common/checkout/changelog_unittest.py: Added a test case for https://webkit.org/b/XXXXX.
        * Scripts/webkitpy/common/config/urls.py: Modified bug_url_short to match https scheme.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90083 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Keishi Hattori <keishi@webkit.org>
keishi@webkit.org [Thu, 30 Jun 2011 03:05:03 +0000 (03:05 +0000)]
2011-06-29  Keishi Hattori  <keishi@webkit.org>

        Reviewed by Adam Barth.

        Remove extra space at EOL in view source mode
        https://bugs.webkit.org/show_bug.cgi?id=63620

        * fast/frames/resources/viewsource-frame-5.html: Added.
        * fast/frames/viewsource-empty-attribute-value-expected.txt: Regenerated.
        * fast/frames/viewsource-linebreak-expected.txt: Added.
        * fast/frames/viewsource-linebreak.html: Added. Test that there are no extra spaces at EOL.
        And that the multiple consecutive line breaks are maintained in the frame.
2011-06-29  Keishi Hattori  <keishi@webkit.org>

        Reviewed by Adam Barth.

        Remove extra space at EOL in view source mode
        https://bugs.webkit.org/show_bug.cgi?id=63620

        Test: fast/frames/viewsource-linebreak.html

        * css/view-source.css:
        (tbody:last-child .webkit-line-content:empty:before): If we don't do this
        the last line will be empty and the height will be shorter than the rest.
        * html/HTMLViewSourceDocument.cpp:
        (WebCore::HTMLViewSourceDocument::addSource): If source file is empty, add an empty line.
        (WebCore::HTMLViewSourceDocument::finishLine): Call when you are done with the current line.
        Adds a BR element if the line is empty and sets current to tbody.
        (WebCore::HTMLViewSourceDocument::addText):
        * html/HTMLViewSourceDocument.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90082 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Thu, 30 Jun 2011 02:30:02 +0000 (02:30 +0000)]
2011-06-29  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        new-run-webkit-tests: images page switch actual and expected images before they're loaded
        https://bugs.webkit.org/show_bug.cgi?id=63199

        We now preload the images and keep the DOM nodes in memory instead of
        going back to the MemoryCache all the time.  (Also, remove some nutty
        </img> HTML.)

        * Scripts/webkitpy/layout_tests/layout_package/test_result_writer.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90081 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Kangil Han <kangil.han@samsung.com>
commit-queue@webkit.org [Thu, 30 Jun 2011 02:06:39 +0000 (02:06 +0000)]
2011-06-29  Kangil Han  <kangil.han@samsung.com>

        Reviewed by Kenneth Rohde Christiansen.

        [EFL] Add APIs to get/set view mode
        https://bugs.webkit.org/show_bug.cgi?id=63229

        Implement public API for getting/setting the view-mode media feature as specified in W3C.

        * ewk/ewk_view.cpp:
        (ewk_view_mode_set):
        (ewk_view_mode_get):
        * ewk/ewk_view.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90080 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Thu, 30 Jun 2011 02:02:04 +0000 (02:02 +0000)]
2011-06-29  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        new-run-webkit-tests should tell you when it creates new expectations
        https://bugs.webkit.org/show_bug.cgi?id=63675

        Unforunately, this patch creates a bunch of test-webkitpy spam because
        we don't understand how Python logging works well enough to stop the
        spam.  :(

        * Scripts/webkitpy/layout_tests/layout_package/single_test_runner.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90079 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Eric Seidel <eric@webkit.org>
eric@webkit.org [Thu, 30 Jun 2011 01:48:44 +0000 (01:48 +0000)]
2011-06-29  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Refactor WebKitDriver.run_test into smaller pieces
        https://bugs.webkit.org/show_bug.cgi?id=63673

        There is no functional change here.  Just moving code around.
        This is in preparation for adding support for #CRASHED and #CRASHED - WebProcess

        * Scripts/webkitpy/layout_tests/port/webkit.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90078 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Eric Seidel <eric@webkit.org>
eric@webkit.org [Thu, 30 Jun 2011 01:36:03 +0000 (01:36 +0000)]
2011-06-29  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Simplify skipped file finding in preparation for adding wk2 skipped list fallback
        https://bugs.webkit.org/show_bug.cgi?id=63501

        Re-landing after fixing a bug in WebKitPort.baseline_search_path which broke
        the rebaselineserver unit tests.  Also fixed nits brought up by Dirk after previous landing.

        The goal was to add support for wk2 skipped lists.
        However, I found that our skipped list computation was a
        manual hack (which only worked for the mac port).

        So I fixed a FIXME to move the skipped list finding
        into WebKitPort instead of MacPort.
        Doing so required the concept of a "port_name", since previously
        the only name accessible from a port object was name()
        which includes many things beyond "mac" or "chromium", etc.

        Eventually I believe we'll want to re-think the way that we pass
        in a port_name argument to Port subclasses and expect them to parse
        it.  But for now I just added a cls.port_name variable which contains
        the static information needed to compute wk2 names as well as
        compute Skipped list fallback which works for Mac/Win/Qt and Gtk.

        In order to test my new _skipped_file_search_paths method, I
        fixed another FIXME by making it return relative paths.

        I also fixed the test_expectations_path code in WebKitPort to use port_name.
        It was using name() which would return PORT-VERSION so MacPort was overriding
        it to use just PORT.  After fixing test_expectations_path to use port_name
        (and making it aware of webkit2) I was able to remove the MacPort implementation.

        * Scripts/webkitpy/layout_tests/port/base.py:
         - Add port_name() to access "mac", since name() returns "mac-leopard" etc.
         - Document that real_name() seems to have no purpose.
        * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
         - Add _parse_port_name(), eventually we might call this from WebKitPort directly.
        * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
         - Add _parse_port_name.
        * Scripts/webkitpy/layout_tests/port/chromium_win.py:
         - Add _parse_port_name.
        * Scripts/webkitpy/layout_tests/port/gtk.py:
        * Scripts/webkitpy/layout_tests/port/mac.py:
         - Move Skipped-file finding code down to WebKitPort
        * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
        * Scripts/webkitpy/layout_tests/port/qt.py:
        * Scripts/webkitpy/layout_tests/port/webkit.py:
        * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90077 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Sheriff Bot <webkit.review.bot@gmail.com>
commit-queue@webkit.org [Thu, 30 Jun 2011 01:11:42 +0000 (01:11 +0000)]
2011-06-29  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r90070.
        http://trac.webkit.org/changeset/90070
        https://bugs.webkit.org/show_bug.cgi?id=63676

        Broke test-webkitpy rebaseline tests (Requested by eseidel2 on
        #webkit).

        * Scripts/webkitpy/layout_tests/port/base.py:
        * Scripts/webkitpy/layout_tests/port/chromium.py:
        * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
        * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
        * Scripts/webkitpy/layout_tests/port/chromium_win.py:
        * Scripts/webkitpy/layout_tests/port/gtk.py:
        * Scripts/webkitpy/layout_tests/port/mac.py:
        * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
        * Scripts/webkitpy/layout_tests/port/qt.py:
        * Scripts/webkitpy/layout_tests/port/webkit.py:
        * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:
        * Scripts/webkitpy/layout_tests/port/win.py:
        * Scripts/webkitpy/tool/commands/rebaselineserver_unittest.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90076 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago[WinCE] Correct logical state of WINCE test.
bfulgham@webkit.org [Thu, 30 Jun 2011 00:46:42 +0000 (00:46 +0000)]
[WinCE] Correct logical state of WINCE test.

* platform/win/ScrollbarThemeWin.cpp: The #if OS(WINCE)
  should have been #if !OS(WINCE)
(WebCore::ScrollbarThemeWin::paintTrackPiece):
(WebCore::ScrollbarThemeWin::paintButton):
(WebCore::ScrollbarThemeWin::paintThumb):
* rendering/RenderThemeWin.cpp: ditto.
(WebCore::drawControl):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90075 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Thu, 30 Jun 2011 00:42:39 +0000 (00:42 +0000)]
2011-06-29  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        new-run-webkit-tests doesn't support sample-on-timeout
        https://bugs.webkit.org/show_bug.cgi?id=56731

        * Scripts/webkitpy/layout_tests/port/server_process.py:
        * Scripts/webkitpy/layout_tests/port/server_process_unittest.py:
        * Scripts/webkitpy/layout_tests/run_webkit_tests.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90074 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Justin Garcia <justin.garcia@apple.com>
justin.garcia@apple.com [Thu, 30 Jun 2011 00:30:49 +0000 (00:30 +0000)]
2011-06-29  Justin Garcia  <justin.garcia@apple.com>

        Reviewed by Enrica Casucci.

        I forgot to include these (reviewed) style fix-ups in my previous check-in.

        * editing/htmlediting.cpp:
        (WebCore::indexForVisiblePosition):
        (WebCore::visiblePositionForIndex):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90073 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Justin Garcia <justin.garcia@apple.com>
justin.garcia@apple.com [Thu, 30 Jun 2011 00:10:53 +0000 (00:10 +0000)]
2011-06-29  Justin Garcia  <justin.garcia@apple.com>

        Reviewed by Enrica Casucci.

        https://bugs.webkit.org/show_bug.cgi?id=62922
        indexForVisiblePosition(const VisiblePosition& visiblePosition) does not consider shadow content

        VisiblePositions can be inside web form text regions, which use shadow trees. Made indexForVisiblePosition
        aware of this, and added a new parameter to obtain the scope for a VisiblePosition, in addition to its index.

        Added visiblePositionForIndex to go in the opposite direction, taking into account the scope
        used to compute the index.

        These two functions use TextIterators to convert between VisiblePositions and indices. But
        TextIterator iteration using TextIteratorEmitsCharactersBetweenAllVisiblePositions does not
        exactly match VisiblePosition iteration, so using them to preserve a selection during an
        editing operation is unreliable. This can be seen in the expected results for:

        editing/execCommand/indent-pre-list.html
        editing/execCommand/crash-indenting-list-item.html

        TextIterator's TextIteratorEmitsCharactersBetweenAllVisiblePositions mode needs to be fixed, or
        these functions need to be changed to iterate using actual VisiblePositions. See:

        https://bugs.webkit.org/show_bug.cgi?id=63590
        TextIterators in TextIteratorEmitsCharactersBetweenAllVisiblePositions do not exactly match VisiblePositions

        Also:

        https://bugs.webkit.org/show_bug.cgi?id=63592
        Use visiblePositionForIndex and indexForVisiblePosition everywhere that TextIterators are used to convert between VisiblePositions and indices

        No new tests added because indexForVisiblePosition is currently only used for editing operations
        that cannot be performed inside web form fields.

        * editing/ApplyBlockElementCommand.cpp:
        (WebCore::ApplyBlockElementCommand::doApply):
        * editing/InsertListCommand.cpp:
        (WebCore::InsertListCommand::doApply):
        * editing/htmlediting.cpp:
        (WebCore::indexForVisiblePosition):
        (WebCore::visiblePositionForIndex):
        * editing/htmlediting.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90072 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Dimitri Glazkov <dglazkov@chromium.org>
dglazkov@chromium.org [Thu, 30 Jun 2011 00:03:36 +0000 (00:03 +0000)]
2011-06-29  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Kent Tamura.

        Remove ElementWithPseudoId, since now an Element can store it all by itself.
        https://bugs.webkit.org/show_bug.cgi?id=63657

        Refactoring, covered by existing tests.

        * CMakeLists.txt: Removed ElementWithPseudoId from build.
        * GNUmakefile.list.am: Ditto.
        * WebCore.gypi: Ditto.
        * WebCore.order: Ditto.
        * WebCore.pro: Ditto.
        * WebCore.vcproj/WebCore.vcproj: Ditto.
        * WebCore.xcodeproj/project.pbxproj: Ditto.
        * html/ColorInputType.cpp:
        (WebCore::ColorInputType::createShadowSubtree): Changed to use plain-old HTMLDivElement.
        * html/RangeInputType.cpp:
        (WebCore::RangeInputType::createShadowSubtree): Ditto.
        * html/ValidationMessage.cpp:
        (WebCore::ValidationMessage::buildBubbleTree): Ditto.
        * html/shadow/ElementWithPseudoId.cpp: Removed.
        * html/shadow/ElementWithPseudoId.h: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90071 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Eric Seidel <eric@webkit.org>
eric@webkit.org [Thu, 30 Jun 2011 00:02:16 +0000 (00:02 +0000)]
2011-06-29  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Simplify skipped file finding in preparation for adding wk2 skipped list fallback
        https://bugs.webkit.org/show_bug.cgi?id=63501

        The goal was to add support for wk2 skipped lists.
        However, I found that our skipped list computation was a
        manual hack (which only worked for the mac port).

        So I fixed a FIXME to move the skipped list finding
        into WebKitPort instead of MacPort.
        Doing so required the concept of a "port_name", since previously
        the only name accessible from a port object was name()
        which includes many things beyond "mac" or "chromium", etc.

        Eventually I believe we'll want to re-think the way that we pass
        in a port_name argument to Port subclasses and expect them to parse
        it.  But for now I just added a cls.port_name variable which contains
        the static information needed to compute wk2 names as well as
        compute Skipped list fallback which works for Mac/Win/Qt and Gtk.

        In order to test my new _skipped_file_search_paths method, I
        fixed another FIXME by making it return relative paths.

        I also fixed the test_expectations_path code in WebKitPort to use port_name.
        It was using name() which would return PORT-VERSION so MacPort was overriding
        it to use just PORT.  After fixing test_expectations_path to use port_name
        (and making it aware of webkit2) I was able to remove the MacPort implementation.

        * Scripts/webkitpy/layout_tests/port/base.py:
         - Add port_name() to access "mac", since name() returns "mac-leopard" etc.
         - Document that real_name() seems to have no purpose.
        * Scripts/webkitpy/layout_tests/port/chromium_linux.py:
         - Add _parse_port_name(), eventually we might call this from WebKitPort directly.
        * Scripts/webkitpy/layout_tests/port/chromium_mac.py:
         - Add _parse_port_name.
        * Scripts/webkitpy/layout_tests/port/chromium_win.py:
         - Add _parse_port_name.
        * Scripts/webkitpy/layout_tests/port/gtk.py:
        * Scripts/webkitpy/layout_tests/port/mac.py:
         - Move Skipped-file finding code down to WebKitPort
        * Scripts/webkitpy/layout_tests/port/mac_unittest.py:
        * Scripts/webkitpy/layout_tests/port/qt.py:
        * Scripts/webkitpy/layout_tests/port/webkit.py:
        * Scripts/webkitpy/layout_tests/port/webkit_unittest.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90070 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Levi Weintraub <leviw@chromium.org>
leviw@chromium.org [Wed, 29 Jun 2011 23:59:49 +0000 (23:59 +0000)]
2011-06-29  Levi Weintraub  <leviw@chromium.org>

        Reviewed by Eric Seidel.

        Switch computeWidth and computeHeight functions to new layout types
        https://bugs.webkit.org/show_bug.cgi?id=63655

        Switch computeWidth and computeHeight functions over to the new layout unit abstraction.

        Making a point to mark rounding that should eventually be removed by linking to a new
        tracking bug: 63645.

        No new tests, no functionality changes.

        * rendering/AutoTableLayout.cpp:
        (WebCore::AutoTableLayout::computePreferredLogicalWidths):
        * rendering/AutoTableLayout.h:
        * rendering/FixedTableLayout.cpp:
        (WebCore::FixedTableLayout::computePreferredLogicalWidths):
        * rendering/FixedTableLayout.h:
        * rendering/LayoutTypes.h:
        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::computePreferredLogicalWidths):
        (WebCore::RenderBlock::computeInlinePreferredLogicalWidths):
        (WebCore::RenderBlock::computeBlockPreferredLogicalWidths):
        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::computeLogicalWidthUsing):
        (WebCore::RenderBox::computeLogicalHeight):
        (WebCore::RenderBox::computeLogicalHeightUsing):
        (WebCore::RenderBox::computePercentageLogicalHeight):
        (WebCore::RenderBox::computeReplacedLogicalWidth):
        (WebCore::RenderBox::computeReplacedLogicalWidthRespectingMinMaxWidth):
        (WebCore::RenderBox::computeReplacedLogicalWidthUsing):
        (WebCore::RenderBox::computeReplacedLogicalHeight):
        (WebCore::RenderBox::computeReplacedLogicalHeightRespectingMinMaxHeight):
        (WebCore::RenderBox::computeReplacedLogicalHeightUsing):
        * rendering/RenderBox.h:
        * rendering/RenderIFrame.cpp:
        (WebCore::RenderIFrame::computeLogicalWidth):
        * rendering/RenderImage.cpp:
        (WebCore::RenderImage::computeReplacedLogicalWidth):
        (WebCore::RenderImage::computeReplacedLogicalHeight):
        * rendering/RenderImage.h:
        * rendering/RenderListBox.cpp:
        (WebCore::RenderListBox::computePreferredLogicalWidths):
        * rendering/RenderListMarker.cpp:
        (WebCore::RenderListMarker::computePreferredLogicalWidths):
        * rendering/RenderMenuList.cpp:
        (WebCore::RenderMenuList::computePreferredLogicalWidths):
        * rendering/RenderReplaced.cpp:
        (WebCore::RenderReplaced::computeReplacedLogicalWidth):
        (WebCore::RenderReplaced::computeReplacedLogicalHeight):
        (WebCore::RenderReplaced::computePreferredLogicalWidths):
        * rendering/RenderReplaced.h:
        * rendering/RenderSlider.cpp:
        (WebCore::RenderSlider::computePreferredLogicalWidths):
        * rendering/RenderTable.cpp:
        (WebCore::RenderTable::computeLogicalWidth):
        * rendering/RenderTextControl.cpp:
        (WebCore::RenderTextControl::computePreferredLogicalWidths):
        * rendering/RenderVideo.cpp:
        (WebCore::RenderVideo::computeReplacedLogicalWidth):
        (WebCore::RenderVideo::computeReplacedLogicalHeight):
        * rendering/RenderVideo.h:
        * rendering/svg/RenderSVGForeignObject.cpp:
        (WebCore::RenderSVGForeignObject::computeLogicalWidth):
        (WebCore::RenderSVGForeignObject::computeLogicalHeight):
        * rendering/svg/RenderSVGRoot.cpp:
        (WebCore::RenderSVGRoot::computePreferredLogicalWidths):
        (WebCore::RenderSVGRoot::computeIntrinsicWidth):
        (WebCore::RenderSVGRoot::computeIntrinsicHeight):
        (WebCore::RenderSVGRoot::computeReplacedLogicalWidth):
        (WebCore::RenderSVGRoot::computeReplacedLogicalHeight):
        (WebCore::RenderSVGRoot::layout):
        * rendering/svg/RenderSVGRoot.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90069 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-23 Abhishek Arya <inferno@chromium.org>
inferno@chromium.org [Wed, 29 Jun 2011 23:58:25 +0000 (23:58 +0000)]
2011-06-23  Abhishek Arya  <inferno@chromium.org>

        Reviewed by James Robinson.

        In RenderBlock, RenderWidget and RenderReplaced destroy functions,
        call dirtyLinesFromChangedChild to tell our parent that we are going away.
        https://bugs.webkit.org/show_bug.cgi?id=60307

        Test: fast/block/child-not-removed-from-parent-lineboxes-crash.html
              fast/block/block-not-removed-from-parent-lineboxes-crash.html

        * rendering/RenderBlock.cpp:
        (WebCore::RenderBlock::destroy):
        * rendering/RenderReplaced.cpp:
        (WebCore::RenderReplaced::destroy):
        * rendering/RenderReplaced.h:
        * rendering/RenderWidget.cpp:
        (WebCore::RenderWidget::destroy):
2011-06-23  Abhishek Arya  <inferno@chromium.org>

        Reviewed by James Robinson.

        Tests that we do not crash when failing to remove the linebreak
        object(widget, block, replaced, etc) from parent block lineboxes. The
        testcase tries to fuzz over the list of html and svg tags.
        https://bugs.webkit.org/show_bug.cgi?id=60307

        * fast/block/block-not-removed-from-parent-lineboxes-crash-expected.txt: Added.
        * fast/block/block-not-removed-from-parent-lineboxes-crash.html: Added.
        * fast/block/child-not-removed-from-parent-lineboxes-crash-expected.txt: Added.
        * fast/block/child-not-removed-from-parent-lineboxes-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90068 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years ago2011-06-29 Alan Stearns <stearns@adobe.com>
commit-queue@webkit.org [Wed, 29 Jun 2011 23:54:37 +0000 (23:54 +0000)]
2011-06-29  Alan Stearns  <stearns@adobe.com>

        Reviewed by Tony Chang.

        Modify Skipped files so that future css regions layout tests are
        skipped by default.

        [CSS Regions] Add support for skipped tests
        https://bugs.webkit.org/show_bug.cgi?id=63632

        * fast/regions/flow-content-basic.html: Added.
        * platform/chromium/test_expectations.txt:
        * platform/gtk/Skipped:
        * platform/mac/Skipped:
        * platform/qt/Skipped:
        * platform/win/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90067 268f45cc-cd09-0410-ab3c-d52691b4dbfc

8 years agoSpeculative WinCE build correction.
bfulgham@webkit.org [Wed, 29 Jun 2011 23:52:08 +0000 (23:52 +0000)]
Speculative WinCE build correction.

* rendering/RenderThemeWin.cpp:
(WebCore::drawControl): Don't call setRGBABitmapAlpha
  for WinCE case.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@90066 268f45cc-cd09-0410-ab3c-d52691b4dbfc