Use bloom filter for descendant selector filtering
https://bugs.webkit.org/show_bug.cgi?id=53880

Reviewed by Maciej Stachowiak.

Source/JavaScriptCore:

Implement a bloom filter with k=2 and 8 bit counting.

* GNUmakefile.am:
* JavaScriptCore.gypi:
* JavaScriptCore.vcproj/WTF/WTF.vcproj:
* JavaScriptCore.xcodeproj/project.pbxproj:
* wtf/BloomFilter.h: Added.
(WTF::BloomFilter::maximumCount):
(WTF::BloomFilter::BloomFilter):
(WTF::BloomFilter::mayContain):
(WTF::BloomFilter::add):
(WTF::BloomFilter::remove):
(WTF::BloomFilter::firstSlot):
(WTF::BloomFilter::secondSlot):
(WTF::::add):
(WTF::::remove):
(WTF::::clear):
(WTF::::likelyEmpty):
(WTF::::isClear):

Source/WebCore:

Bloom filter is faster than a hash set in this kind of use.

Shark thinks this speeds up style matching by ~30% on sites
with lots of descendant selectors.

* ForwardingHeaders/wtf/BloomFilter.h: Added.
* css/CSSStyleSelector.cpp:
(WebCore::collectElementIdentifierHashes):
(WebCore::CSSStyleSelector::pushParent):
(WebCore::CSSStyleSelector::popParent):
(WebCore::CSSStyleSelector::fastRejectSelector):
(WebCore::RuleData::collectDescendantSelectorIdentifierHashes):
* css/CSSStyleSelector.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77777 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Antti Koivisto.

        media/video-document-types.html sometimes crashes the WebProcess
        https://bugs.webkit.org/show_bug.cgi?id=53884

        No new tests. Already covered by media/video-document-types.html

        * html/HTMLMediaElement.cpp:
        (WebCore::HTMLMediaElement::userCancelledLoad): Stop the load timer
        when cancelling the load.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77776 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Patrick Gansterer  <paroga@webkit.org>

        Reviewed by Adam Roben.

        [WIN] Use WCHAR instead of TCHAR
        https://bugs.webkit.org/show_bug.cgi?id=53863

        We always use the UNICODE versions of windows functions, so
        the usage of TCHAR makes no sense and mixing them is bad style.

        * platform/graphics/win/IconWin.cpp:
        (WebCore::Icon::createIconForFiles):
        * platform/graphics/win/QTMovie.cpp:
        * platform/graphics/win/QTMovieGWorld.cpp:
        (QTMovieGWorld::fullscreenWndProc):
        (QTMovieGWorld::enterFullscreen):
        * platform/graphics/win/SimpleFontDataCGWin.cpp:
        (WebCore::SimpleFontData::platformInit):
        * platform/graphics/win/SimpleFontDataCairoWin.cpp:
        * platform/graphics/win/SimpleFontDataWin.cpp:
        * platform/graphics/wince/SimpleFontDataWinCE.cpp:
        * platform/win/BString.cpp:
        * platform/win/ClipboardWin.cpp:
        (WebCore::filesystemPathFromUrlOrTitle):
        (WebCore::createGlobalHDropContent):
        (WebCore::createGlobalImageFileDescriptor):
        * platform/win/ContextMenuWin.cpp:
        * platform/win/CursorWin.cpp:
        (WebCore::loadSharedCursor):
        * platform/win/FileChooserWin.cpp:
        (WebCore::FileChooser::basenameForWidth):
        * platform/win/PopupMenuWin.cpp:
        (WebCore::PopupMenuWin::popupClassName):
        (WebCore::PopupMenuWin::show):
        * platform/win/PopupMenuWin.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77775 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Patrick Gansterer  <paroga@webkit.org>

        Reviewed by Andreas Kling.

        Replace FLT_MAX with std::numeric_limits
        https://bugs.webkit.org/show_bug.cgi?id=53861

        Also move EFL's userIdleTime from TemporaryLinkStubs into SystemTimeEfl.

        * html/NumberInputType.cpp:
        * html/parser/HTMLParserIdioms.cpp:
        (WebCore::parseToDoubleForNumberType):
        * platform/brew/SystemTimeBrew.cpp:
        (WebCore::userIdleTime):
        * platform/efl/SystemTimeEfl.cpp:
        (WebCore::userIdleTime):
        * platform/efl/TemporaryLinkStubs.cpp:
        * platform/graphics/ca/GraphicsLayerCA.cpp:
        (WebCore::GraphicsLayerCA::setupAnimation):
        * platform/win/SystemTimeWin.cpp:
        (WebCore::userIdleTime):
        * platform/wx/SystemTimeWx.cpp:
        (WebCore::userIdleTime):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77774 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Robert Hogan  <robert@webkit.org>

        Reviewed by Andreas Kling.

        [Qt] Remove unnecessary QRect assignment
        https://bugs.webkit.org/show_bug.cgi?id=53873

        * Api/qwebframe.cpp:
        (QWebFramePrivate::renderRelativeCoords):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77773 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Robert Hogan  <robert@webkit.org>

        Reviewed by Andreas Kling.

        [Qt] Clear page's groupName even when not in DRT
        https://bugs.webkit.org/show_bug.cgi?id=53874

        Matches behaviour of other ports.

        * WebCoreSupport/ChromeClientQt.cpp:
        (WebCore::ChromeClientQt::closeWindowSoon):
2011-02-06  Robert Hogan  <robert@webkit.org>

        Reviewed by Andreas Kling.

        [Qt] Clear page's groupName even when not in DRT
        https://bugs.webkit.org/show_bug.cgi?id=53874

        * DumpRenderTree/qt/DumpRenderTreeQt.cpp:
        (WebCore::DumpRenderTree::windowCloseRequested):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77772 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Patrick Gansterer  <paroga@webkit.org>

        Reviewed by Andreas Kling.

        [CMake] Add dependencies for Visual Studio projects
        https://bugs.webkit.org/show_bug.cgi?id=53773

        Add a JavaScriptCore dependecy to WebCore, so CMake can
        generate the correct build order for the solution.

        * CMakeLists.txt:
2011-02-06  Patrick Gansterer  <paroga@webkit.org>

        Reviewed by Andreas Kling.

        [CMake] Add dependencies for Visual Studio projects
        https://bugs.webkit.org/show_bug.cgi?id=53773

        Add a WebCore dependecy to WebKit, so CMake can
        generate the correct build order for the solution.

        Remove JavaScriptCore dependecy, since WebCore already depends on it.

        * CMakeLists.txt:
2011-02-06  Patrick Gansterer  <paroga@webkit.org>

        Reviewed by Andreas Kling.

        [CMake] Add dependencies for Visual Studio projects
        https://bugs.webkit.org/show_bug.cgi?id=53773

        Add a WebKit dependecy to WinCELauncher, so CMake can
        generate the correct build order for the solution.

        * CMakeListsWinCE.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77771 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Patrick Gansterer  <paroga@webkit.org>

        Reviewed by Andreas Kling.

        [CMake] Unify ENABLE_SVG sections
        https://bugs.webkit.org/show_bug.cgi?id=53778

        * CMakeLists.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77770 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Martin Robinson  <mrobinson@igalia.com>

        Reclassify a failing test and unskip another which should be passing now.

        * platform/gtk/Skipped: Update the skipped list.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77769 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Andreas Kling  <kling@webkit.org>

        Reviewed by Dirk Schulze.

        Refactor Font::normalizeSpaces() to avoid an extra String copy.
        https://bugs.webkit.org/show_bug.cgi?id=53871

        Have normalizeSpaces() take UChar*,length arguments instead of a
        String, since all call sites were creating temporaries from these
        components anyway.

        Font::normalizeSpaces() now always makes 1 String copy,
        instead of 1 in the best case and 2 in the worst case.

        * platform/graphics/Font.cpp:
        (WebCore::Font::normalizeSpaces):
        * platform/graphics/Font.h:
        * platform/graphics/qt/FontQt.cpp:
        (WebCore::drawTextCommon):
        (WebCore::Font::floatWidthForSimpleText):
        (WebCore::Font::floatWidthForComplexText):
        (WebCore::Font::offsetForPositionForSimpleText):
        (WebCore::Font::offsetForPositionForComplexText):
        (WebCore::Font::selectionRectForSimpleText):
        (WebCore::Font::selectionRectForComplexText):
        * svg/SVGFont.cpp:
        (WebCore::SVGTextRunWalker::walk):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77768 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Andreas Kling  <kling@webkit.org>

        Reviewed by Antonio Gomes.

        [Qt] MiniBrowser: Clean up handling of titleChanged() signal
        https://bugs.webkit.org/show_bug.cgi?id=53869

        * MiniBrowser/qt/BrowserView.cpp:
        (BrowserView::BrowserView): Remove unnecessary titleChanged() handling.
        * MiniBrowser/qt/BrowserWindow.cpp:
        (BrowserWindow::BrowserWindow): Connect titleChanged() directly to setWindowTitle().
        * MiniBrowser/qt/BrowserWindow.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77767 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Andreas Kling  <kling@webkit.org>

        Reviewed by Dirk Schulze.

        Fix potential buffer overrun in SVGTextRunWalker::walk()
        https://bugs.webkit.org/show_bug.cgi?id=53870

        A new String was created from a UChar* with a 'length' argument
        that could be greater than the number of UChars available.

        * svg/SVGFont.cpp:
        (WebCore::SVGTextRunWalker::walk):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77766 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Benjamin Poulain  <ikipou@gmail.com>

        Reviewed by Andreas Kling.

        [Qt] WebKit does not build on Mac with WebKit 2
        https://bugs.webkit.org/show_bug.cgi?id=53380

        * Platform/CoreIPC/qt/ConnectionQt.cpp: Use fcntl() to set close on exec so
        we do not rely on recent kernel.
        (CoreIPC::Connection::readyReadHandler):
        * Platform/qt/SharedMemoryQt.cpp: ditto
        (WebKit::SharedMemory::create):
        (WebKit::SharedMemory::createHandle):
        * UIProcess/Launcher/qt/ProcessLauncherQt.cpp: Guard prctl() since it is
        a Linux specific function.
        (WebKit::QtWebProcess::setupChildProcess):
        * UIProcess/DrawingAreaProxy.h: Differentiate between building on Mac
        and the Mac port.
        * UIProcess/DrawingAreaProxyImpl.cpp: ditto
        * UIProcess/WebPageProxy.cpp: ditto
        (WebKit::WebPageProxy::didReceiveMessage):
        * WebProcess/WebPage/DrawingArea.cpp: ditto
        (WebKit::DrawingArea::create):
        * WebProcess/WebPage/DrawingArea.h:
        * WebProcess/WebPage/DrawingAreaImpl.cpp: ditto
        * WebProcess/WebPage/WebPage.cpp: ditto
        (WebKit::WebPage::didReceiveMessage):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77765 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Robert Hogan  <robert@webkit.org>

        Unreviewed, skip failing test on Qt bot.

        [Qt]  fast/notifications/notifications-document-close-crash.html fails after r77738
        https://bugs.webkit.org/show_bug.cgi?id=53868

        * platform/qt/Skipped: Skip fast/notifications/notifications-document-close-crash.html

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77764 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-06  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Dan Bernstein.

        WebKitTestRunner needs layoutTestController.setWillSendRequestReturnsNull
        https://bugs.webkit.org/show_bug.cgi?id=42690
        <rdar://problem/8213851>

        * platform/mac-wk2/Skipped: Unskip tests that now pass, except for the ones that
        mysteriously fail.
2011-02-06  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Dan Bernstein.

        WebKitTestRunner needs layoutTestController.setWillSendRequestReturnsNull
        https://bugs.webkit.org/show_bug.cgi?id=42690
        <rdar://problem/8213851>

        * WebKitTestRunner/InjectedBundle/Bindings/LayoutTestController.idl:
        * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
        (WTR::InjectedBundlePage::willSendRequestForFrame):
        * WebKitTestRunner/InjectedBundle/LayoutTestController.cpp:
        (WTR::LayoutTestController::LayoutTestController):
        * WebKitTestRunner/InjectedBundle/LayoutTestController.h:
        (WTR::LayoutTestController::willSendRequestReturnsNull):
        (WTR::LayoutTestController::setWillSendRequestReturnsNull):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77763 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Dan Bernstein.

        REGRESSION: New drawing area is causing some WebProcess crashes in layout tests
        https://bugs.webkit.org/show_bug.cgi?id=53865
        <rdar://problem/8963148>

        * WebProcess/WebPage/DrawingAreaImpl.cpp:
        (WebKit::DrawingAreaImpl::enterAcceleratedCompositingMode): Handle the case
        where this is called repeatedly without exiting accelerated composting mode
        first.
        (WebKit::DrawingAreaImpl::exitAcceleratedCompositingMode): Handle the case
        of being called more than once.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77759 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Dimitri Glazkov  <dglazkov@chromium.org>

        [Chromium] Add baselines for the test added in r77690.

        * platform/chromium-mac/media/video-frame-accurate-seek-expected.checksum: Added.
        * platform/chromium-mac/media/video-frame-accurate-seek-expected.png: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77758 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Dimitri Glazkov  <dglazkov@chromium.org>

        Remove passing canvas tests from expectations.

        * platform/chromium/test_expectations.txt: Removed passing tests.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77757 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Dimitri Glazkov  <dglazkov@chromium.org>

        Remove passing tests whose results match mac/ and mac-leopard.

        * platform/chromium/test_expectations.txt: Removed passing tests.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77756 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        Remove unused parameter related to XSSAuditor
        https://bugs.webkit.org/show_bug.cgi?id=53862

        * WebCore.exp.in:
        * bindings/ScriptControllerBase.cpp:
        (WebCore::ScriptController::executeScript):
        (WebCore::ScriptController::executeIfJavaScriptURL):
        * bindings/ScriptControllerBase.h:
        * bindings/js/ScriptController.cpp:
        (WebCore::ScriptController::evaluateInWorld):
        (WebCore::ScriptController::evaluate):
        (WebCore::ScriptController::executeScriptInWorld):
        * bindings/js/ScriptController.h:
        * bindings/v8/ScriptController.cpp:
        (WebCore::ScriptController::evaluate):
        * bindings/v8/ScriptController.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77755 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Dimitri Glazkov  <dglazkov@chromium.org>

        Remove results for failing SVG tests. I will add the new ones after the
        bots tell me which ones aren't matching mac/ or mac-leopard/.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77754 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Adam Barth  <abarth@webkit.org>

        Fix Qt build.

        * WebCore.pro:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77753 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        Delete XSSAuditor
        https://bugs.webkit.org/show_bug.cgi?id=53859

        The job of the XSSAuditor is now done by the XSSFilter.  In the future,
        we might rename XSSFilter to XSSAuditor.

        * Android.mk:
        * CMakeLists.txt:
        * GNUmakefile.am:
        * WebCore.gypi:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * bindings/ScriptControllerBase.cpp:
        (WebCore::ScriptController::executeIfJavaScriptURL):
        * bindings/ScriptControllerBase.h:
        * bindings/js/ScriptController.cpp:
        (WebCore::ScriptController::ScriptController):
        (WebCore::ScriptController::evaluateInWorld):
        * bindings/js/ScriptController.h:
        * bindings/js/ScriptEventListener.cpp:
        (WebCore::createAttributeEventListener):
        * bindings/v8/ScriptController.cpp:
        (WebCore::ScriptController::ScriptController):
        (WebCore::ScriptController::evaluate):
        * bindings/v8/ScriptController.h:
        * bindings/v8/ScriptEventListener.cpp:
        (WebCore::createAttributeEventListener):
        * dom/Document.cpp:
        (WebCore::Document::implicitOpen):
        (WebCore::Document::processBaseElement):
        * dom/ScriptableDocumentParser.cpp:
        (WebCore::ScriptableDocumentParser::ScriptableDocumentParser):
        * dom/ScriptableDocumentParser.h:
        * html/parser/HTMLDocumentParser.cpp:
        (WebCore::HTMLDocumentParser::shouldLoadExternalScriptFromSrc):
        * loader/SubframeLoader.cpp:
        (WebCore::SubframeLoader::requestObject):
        (WebCore::SubframeLoader::loadMediaPlayerProxyPlugin):
        * page/XSSAuditor.cpp: Removed.
        * page/XSSAuditor.h: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77752 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Anders Carlsson  <andersca@apple.com>

        Reviewed by Darin Adler.

        Need a way to resize and scroll the page at the same time
        https://bugs.webkit.org/show_bug.cgi?id=53814
        <rdar://problem/8498008>

        Export a function needed by WebKit2.

        * WebCore.exp.in:
2011-02-05  Anders Carlsson  <andersca@apple.com>

        Reviewed by Darin Adler.

        Need a way to resize and scroll the page at the same time
        https://bugs.webkit.org/show_bug.cgi?id=53814
        <rdar://problem/8498008>

        Add -[WKView setFrame:andScrollBy:].

        * UIProcess/API/mac/WKView.h:
        * UIProcess/API/mac/WKView.mm:
        (-[WKView setFrame:andScrollBy:]):
        Stash away the scroll offset and call setFrame:.

        (-[WKView _switchToDrawingAreaTypeIfNecessary:DrawingAreaInfo::]):
        (-[WKView _setDrawingAreaSize:]):
        * UIProcess/API/qt/qwkpage.cpp:
        (QWKPage::setViewportSize):
        DrawingAreaProxy::setSize now takes a scroll offset.

        * UIProcess/DrawingAreaProxy.cpp:
        (WebKit::DrawingAreaProxy::setSize):
        This now takes a scroll offset and stores it in m_scrollOffset.

        * UIProcess/DrawingAreaProxy.h:
        * UIProcess/DrawingAreaProxyImpl.cpp:
        (WebKit::DrawingAreaProxyImpl::sendSetSize):
        Send along the scroll offset and set it back to zero.

        * UIProcess/LayerBackedDrawingAreaProxy.cpp:
        (WebKit::LayerBackedDrawingAreaProxy::didSetSize):
        * UIProcess/TiledDrawingAreaProxy.cpp:
        (WebKit::TiledDrawingAreaProxy::didSetSize):
        * UIProcess/win/WebView.cpp:
        (WebKit::WebView::onSizeEvent):
        (WebKit::WebView::switchToDrawingAreaTypeIfNecessary):
        DrawingAreaProxy::setSize now takes a scroll offset.

        * WebProcess/WebPage/DrawingArea.h:
        (WebKit::DrawingArea::setSize):
        Add a scrollOffset parameter.

        * WebProcess/WebPage/DrawingArea.messages.in:
        Add a scroll offset parameter to the SetSize message.

        * WebProcess/WebPage/DrawingAreaImpl.cpp:
        (WebKit::DrawingAreaImpl::setSize):
        Call WebPage::scrollMainFrameIfNotAtMaxScrollPosition.

        * WebProcess/WebPage/DrawingAreaImpl.h:
        * WebProcess/WebPage/WebPage.cpp:
        (WebKit::WebPage::scrollMainFrameIfNotAtMaxScrollPosition):
        Compute a scroll offset and call setScrollPosition on the main frame view.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77751 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Patrick Gansterer  <paroga@webkit.org>

        Reviewed by Alexey Proskuryakov.

        Use Document::encoding() instead of DocumentWriter::encoding()
        https://bugs.webkit.org/show_bug.cgi?id=53398

        This change makes the code more readable, when a subresource
        inherits the character encoding of the document.

        * dom/ProcessingInstruction.cpp:
        (WebCore::ProcessingInstruction::checkStyleSheet):
        * dom/ScriptElement.cpp:
        (WebCore::ScriptElement::scriptCharset):
        * html/HTMLLinkElement.cpp:
        (WebCore::HTMLLinkElement::process):
        * loader/cache/CachedResourceLoader.cpp:
        (WebCore::CachedResourceLoader::requestPreload):
        * platform/network/FormDataBuilder.cpp:
        (WebCore::FormDataBuilder::encodingFromAcceptCharset):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77750 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Patrick Gansterer  <paroga@webkit.org>

        Reviewed by Alexey Proskuryakov.

        Add addition test for inherited encoding
        https://bugs.webkit.org/show_bug.cgi?id=53398

        Check if the right encoding is used, after document.charset was set.

        * fast/dom/Document/document-charset.html: Added doctype to make it pass in IE9 out of the box.
        * fast/encoding/css-charset-default-expected.txt:
        * fast/encoding/css-charset-default.xhtml:
        * fast/encoding/external-script-charset-expected.txt:
        * fast/encoding/external-script-charset-koi-8.js: Added.
        * fast/encoding/external-script-charset.js:
        * fast/encoding/external-script-charset.xhtml:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77749 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        Introduce URLString to represent a canonicalized URL represented by a string
        https://bugs.webkit.org/show_bug.cgi?id=53847

        Instead of using String all over the codebase to represent a URL,
        instead we should have a type to represent a canonicalized URL in the
        form of a string.  This patch introduces that type.

        * GNUmakefile.am:
        * WebCore.gypi:
        * WebCore.xcodeproj/project.pbxproj:
        * platform/KURL.cpp:
        (WebCore::KURL::KURL):
        * platform/KURL.h:
        (WebCore::KURL::urlString):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77748 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Martin Robinson  <mrobinson@igalia.com>

        JSC::Bindings m_rootObject->isValid() assert fails when running layout tests
        https://bugs.webkit.org/show_bug.cgi?id=53716

        * platform/gtk/Skipped: Skip a test hitting this assertion.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77747 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Robert Hogan  <robert@webkit.org>

        Reviewed by Antonio Gomes.

        [Qt] Style widgets not rendering
        https://bugs.webkit.org/show_bug.cgi?id=53849

        * platform/qt/Skipped: Unskip plugins/application-plugins-plugins-disabled.html
2011-02-05  Robert Hogan  <robert@webkit.org>

        Reviewed by Antonio Gomes.

        [Qt] Style widgets not rendering
        https://bugs.webkit.org/show_bug.cgi?id=53849

        * html/HTMLObjectElement.cpp:
        (WebCore::HTMLObjectElement::hasValidClassId): x-qt-styled-widget is a valid class id too

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77746 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r77720.
        http://trac.webkit.org/changeset/77720
        https://bugs.webkit.org/show_bug.cgi?id=53854

        "Broke nrwt on Chromium win." (Requested by dglazkov|away on
        #webkit).

        * Scripts/webkitpy/common/system/executive.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77745 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Nikolas Zimmermann  <nzimmermann@rim.com>

        Not reviewed.

        small text which is scaled to be large renders pixelated
        https://bugs.webkit.org/show_bug.cgi?id=12448

        Update Snow Leopard SVG pixel test baseline after r77485.

        * platform/mac/svg/W3C-SVG-1.1-SE/coords-dom-01-f-expected.png:
        * platform/mac/svg/W3C-SVG-1.1-SE/coords-dom-02-f-expected.png:
...

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77744 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Eric Seidel  <eric@webkit.org>

        Unreviewed.

        Safari should lowercase schemes in cannonicalized urls to match every other browser
        https://bugs.webkit.org/show_bug.cgi?id=53848

        Checkin results I forgot to update in my previous commit. All progressions.

        * fast/dom/anchor-origin-expected.txt:
        * http/tests/security/dataURL/xss-DENIED-to-data-url-sub-frame-uppercase-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77743 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Jochen Eisinger  <jochen@chromium.org>

        Reviewed by Adam Barth.

        Add ContentSecurityPolicy object to Document and pass the X-WebKit-CSP header from the MainResourceLoader.
        https://bugs.webkit.org/show_bug.cgi?id=53685

        * WebCore.xcodeproj/project.pbxproj:
        * dom/Document.h:
        (WebCore::Document::contentSecurityPolicy):
        * loader/MainResourceLoader.cpp:
        (WebCore::MainResourceLoader::didReceiveResponse):
        * page/ContentSecurityPolicy.cpp:
        (WebCore::ContentSecurityPolicy::didReceiveHeader):
        * page/ContentSecurityPolicy.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77742 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Safari should lowercase schemes in cannonicalized urls to match every other browser
        https://bugs.webkit.org/show_bug.cgi?id=53848

        Update our results now that we pass more tests.

        * fast/url/file-expected.txt:
        * fast/url/file-http-base-expected.txt:
        * fast/url/path-url-expected.txt:
        * fast/url/relative-expected.txt:
        * fast/url/scheme-expected.txt:
2011-02-05  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Safari should lowercase schemes in cannonicalized urls to match every other browser
        https://bugs.webkit.org/show_bug.cgi?id=53848

        We're clearly the odd man out here.  See results from every browser at:
        https://github.com/abarth/url-spec/blob/master/tests/gurl-results/by-browser.txt

        * platform/KURL.cpp:
        (WebCore::KURL::parse):
         - Just lowercase the scheme when copying.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77741 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Optimize matching of descendant selectors
https://bugs.webkit.org/show_bug.cgi?id=49876
<rdar://problem/8772822>

Reviewed by Dave Hyatt.

During style recalculation, maintain a filter of tags, ids and classes seen in ancestor elements.
Use the filter to quickly reject descendant and child selectors when doing style matching.

This speeds up style recalculations 3-6x on many major web sites.

* css/CSSStyleSelector.cpp:
(WebCore::RuleData::RuleData):
(WebCore::RuleData::descendantSelectorIdentifierHashes):
(WebCore::collectElementIdentifiers):
(WebCore::CSSStyleSelector::pushParent):
(WebCore::CSSStyleSelector::popParent):
(WebCore::CSSStyleSelector::fastRejectSelector):
(WebCore::CSSStyleSelector::matchRulesForList):
(WebCore::RuleData::collectDescendantSelectorIdentifierHashes):
* css/CSSStyleSelector.h:
(WebCore::CSSStyleSelector::ParentStackFrame::ParentStackFrame):
* dom/Element.cpp:
(WebCore::StyleSelectorParentPusher::StyleSelectorParentPusher):
(WebCore::StyleSelectorParentPusher::push):
(WebCore::StyleSelectorParentPusher::~StyleSelectorParentPusher):
(WebCore::Element::attach):
(WebCore::Element::recalcStyle):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77740 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Add performance tests for URL parsing
        https://bugs.webkit.org/show_bug.cgi?id=53845

        The final-url-en URL corpus is from:
        http://corpus.leeds.ac.uk/internet.html

        There are also other language corpuses which we may want to use in the future.

        The usage instructions were as follows:
        "URL lists and other resources: you can freely use them in your research
        provided that you supply a link to this website: http://corpus.leeds.ac.uk/."
        Done.

        * Parser/simple-url.html: Added.
         On my machine:
         Firefox: 255ms
         TOT WebKit: 378ms
         Chrome: 286ms
         Opera: 2830ms (no, that is not a typo)
        * Parser/url-parser.html: Added.
         Firefox: 381ms
         TOT WebKit: 216ms
         Chrome: 131ms
         Opera: 1383ms (again, not a typo)
        * Parser/resources/final-url-en: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77739 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Nate Chapin  <japhet@chromium.org>

        Reviewed by Adam Barth.

        Refactor: NotificationCenter shouldn't hold its own copy of the ScriptExecutionContext*
        when it inherits one from ActiveDOMObject.
        https://bugs.webkit.org/show_bug.cgi?id=53815

        * bindings/js/JSDesktopNotificationsCustom.cpp:
        (WebCore::JSNotificationCenter::requestPermission):
        * bindings/v8/custom/V8NotificationCenterCustom.cpp:
        (WebCore::V8NotificationCenter::requestPermissionCallback):
        * notifications/NotificationCenter.cpp:
        (WebCore::NotificationCenter::NotificationCenter):
        (WebCore::NotificationCenter::checkPermission):
        (WebCore::NotificationCenter::requestPermission):
        (WebCore::NotificationCenter::disconnectFrame):
        * notifications/NotificationCenter.h:
        (WebCore::NotificationCenter::createHTMLNotification):
        (WebCore::NotificationCenter::createNotification):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77738 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-05  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r77725.
        http://trac.webkit.org/changeset/77725
        https://bugs.webkit.org/show_bug.cgi?id=53844

        It broke Qt minimal build (Requested by Ossy_ on #webkit).

        * QtTestBrowser/launcherwindow.cpp:
        (LauncherWindow::createChrome):
        * QtTestBrowser/launcherwindow.h:
        (WindowOptions::WindowOptions):
        * QtTestBrowser/main.cpp:
        (LauncherApplication::handleUserOptions):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77737 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Jer Noble  <jer.noble@apple.com>

        Adding test to Qt Skipped list.

        [Qt] media/video-frame-accurate-seek.html fails
        https://bugs.webkit.org/show_bug.cgi?id=53843

        * platform/qt/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77735 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Joone Hur  <joone.hur@collabora.co.uk>

        Unreviewed, Fix compilation warning.

        Constructor initialization list should follow the declaration order.

        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        (WebKit::FrameLoaderClient::FrameLoaderClient): m_hasRepresentation is moved to
        the last position in the initialization list.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77734 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Versioning.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77732 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Adam Barth  <abarth@webkit.org>

        Reviewed by Maciej Stachowiak.

        Crash in WebCore::TextEncoding::decode below XSSFilter::init
        https://bugs.webkit.org/show_bug.cgi?id=53837

        Test that we can successfully execute a JavaScript URL when it isn't
        blocked by the filter.

        * http/tests/security/xssAuditor/non-block-javascript-url-frame-expected.txt: Added.
        * http/tests/security/xssAuditor/non-block-javascript-url-frame.html: Added.
2011-02-04  Adam Barth  <abarth@webkit.org>

        Reviewed by Maciej Stachowiak.

        Crash in WebCore::TextEncoding::decode below XSSFilter::init
        https://bugs.webkit.org/show_bug.cgi?id=53837

        Add missing null check.

        Test: http/tests/security/xssAuditor/non-block-javascript-url-frame.html

        * html/parser/XSSFilter.cpp:
        (WebCore::XSSFilter::init):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77730 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dan Bernstein.

        Crashes in ShadowBlur via WebKit2 FindController
        https://bugs.webkit.org/show_bug.cgi?id=53830

        Fix a crash cause by re-entering ShadowBlur, and add assertions to
        detect when it happens.

        The re-entrancy occurred when drawRectShadowWithTiling() filled
        the interior of the shadow with fillRect() on the context
        which still had the shadow state set. This would make another ShadowBlur
        on the stack and call into the code again, potentially blowing away
        the image buffer.

        Fix by turning off shadows in the destination context while we're
        drawing the tiled shadow. The non-tiled code path already did this.

        Not testable because CSS shadows clip out the inside of the rect
        being shadowed, and SVG uses fillPath, even for rects.

        * platform/graphics/ShadowBlur.cpp:
        (WebCore::ScratchBuffer::ScratchBuffer):
        (WebCore::ScratchBuffer::getScratchBuffer):
        (WebCore::ScratchBuffer::scheduleScratchBufferPurge):
        (WebCore::ShadowBlur::ShadowBlur):
        (WebCore::ShadowBlur::drawRectShadowWithTiling):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77729 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Adam Barth.

        Skip a new test that uses EventSender
        https://bugs.webkit.org/show_bug.cgi?id=53841

        * platform/mac-wk2/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77728 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Adam Barth.

        [GTK] plugins/plugin-document-back-forward.html fails
        https://bugs.webkit.org/show_bug.cgi?id=53833

        * platform/gtk/Skipped: Unskip test which is now passing.
2011-02-04  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Adam Barth.

        [GTK] plugins/plugin-document-back-forward.html fails
        https://bugs.webkit.org/show_bug.cgi?id=53833

        Add the new "alert on load" functionality for the Unix version of the
        test plugin. This functionality was originally introduced in r77706.

        * DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp:
        (webkit_test_plugin_new_instance): Add new plugin fuctionality.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77727 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Carlos Garcia Campos  <cgarcia@igalia.com>

        Reviewed by Martin Robinson.

        Update expected results reflecting the change to search icon rendering.

        * platform/gtk/fast/css/input-search-padding-expected.checksum:
        * platform/gtk/fast/css/input-search-padding-expected.png:
        * platform/gtk/fast/css/input-search-padding-expected.txt:
        * platform/gtk/fast/forms/search-styled-expected.checksum:
        * platform/gtk/fast/forms/search-styled-expected.png:
        * platform/gtk/fast/forms/search-styled-expected.txt:
        * platform/gtk/fast/forms/search-vertical-alignment-expected.checksum:
        * platform/gtk/fast/forms/search-vertical-alignment-expected.png:
        * platform/gtk/fast/forms/search-zoomed-expected.checksum:
        * platform/gtk/fast/forms/search-zoomed-expected.png:
        * platform/gtk/fast/forms/search-zoomed-expected.txt:
        * platform/gtk/fast/forms/searchfield-heights-expected.checksum:
        * platform/gtk/fast/forms/searchfield-heights-expected.png:
        * platform/gtk/fast/forms/searchfield-heights-expected.txt:
2011-02-04  Carlos Garcia Campos  <cgarcia@igalia.com>

        Reviewed by Martin Robinson.

        [GTK] Don't use a fixed size for search field icons
        https://bugs.webkit.org/show_bug.cgi?id=50624

        Use the parent input content box to make sure the icon fits in the
        search field, scaling it down when needed.

        * platform/gtk/RenderThemeGtk.cpp:
        (WebCore::paintGdkPixbuf):
        (WebCore::getIconSizeForPixelSize):
        (WebCore::adjustSearchFieldIconStyle):
        (WebCore::RenderThemeGtk::adjustSearchFieldResultsDecorationStyle):
        (WebCore::centerRectVerticallyInParentInputElement):
        (WebCore::RenderThemeGtk::paintSearchFieldResultsDecoration):
        (WebCore::RenderThemeGtk::adjustSearchFieldCancelButtonStyle):
        (WebCore::RenderThemeGtk::paintSearchFieldCancelButton):
        (WebCore::RenderThemeGtk::paintCapsLockIndicator):
        (WebCore::RenderThemeGtk::paintMediaButton):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77726 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Kundu Suchismita  <suchi.kundu@nokia.com>

        Reviewed by Andreas Kling.

        Local Storage settings can be enable/disable from "Develop" menu
        https://bugs.webkit.org/show_bug.cgi?id=52296

        * QtTestBrowser/launcherwindow.cpp:
        (LauncherWindow::createChrome):
        (LauncherWindow::toggleLocalStorage):
        (LauncherWindow::toggleOfflineStorageDatabase):
        (LauncherWindow::toggleOfflineWebApplicationCache):
        (LauncherWindow::setOfflineStorageDefaultQuota):
        * QtTestBrowser/launcherwindow.h:
        (WindowOptions::WindowOptions):
        * QtTestBrowser/main.cpp:
        (LauncherApplication::handleUserOptions):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77725 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Hironori Bono  <hbono@chromium.org>

        Reviewed by Adam Barth.

        [chromium] JPEG corruption
        https://bugs.webkit.org/show_bug.cgi?id=53250

        This change adds a new script test that we can render a CMYK JPEG with a
        color profile correctly, i.e. we can render it without color corruption.

        * fast/images/cmyk-jpeg-with-color-profile-expected.txt: Added.
        * fast/images/cmyk-jpeg-with-color-profile.html: Added.
        * fast/images/resources/cmyk-jpeg.jpg: Added.
        * fast/images/script-tests/cmyk-jpeg-with-color-profile.js: Added.
        (image.onload):
2011-02-04  Hironori Bono  <hbono@chromium.org>

        Reviewed by Adam Barth.

        [chromium] JPEG corruption
        https://bugs.webkit.org/show_bug.cgi?id=53250

        Same as gray-scale JPEGs, we convert the colors of CMYK JPEGs with color
        profiles from CMYK to RGB twice and it causes color corruption. This
        change suppresses the color profiles for CMYK JPEGs same as gray-scale
        ones.

        Test: fast/images/cmyk-jpeg-with-color-profile.html

        * platform/image-decoders/jpeg/JPEGImageDecoder.cpp:
        (WebCore::JPEGImageReader::decode):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77724 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Martin Robinson  <mrobinson@igalia.com>

        Update pixel results for the fast/forms directory in preparation for landing
        fixes to input type=search icon sizing. Also add some missing results. These
        results are out of date because of changes to font hinting some time ago.

        * platform/gtk/fast/forms/
        * platform/gtk/fast/forms/input-placeholder-visibility-1-expected.checksum: Added.
        * platform/gtk/fast/forms/input-placeholder-visibility-1-expected.png: Added.
        * platform/gtk/fast/forms/input-placeholder-visibility-1-expected.txt: Added.
        * platform/gtk/fast/forms/input-placeholder-visibility-2-expected.checksum: Added.
        * platform/gtk/fast/forms/input-placeholder-visibility-2-expected.png: Added.
        * platform/gtk/fast/forms/input-placeholder-visibility-2-expected.txt: Added.
        * platform/gtk/fast/forms/input-placeholder-visibility-3-expected.checksum: Added.
        * platform/gtk/fast/forms/input-placeholder-visibility-3-expected.png: Added.
        * platform/gtk/fast/forms/input-placeholder-visibility-3-expected.txt: Added.
        * platform/gtk/fast/forms/textarea-placeholder-visibility-1-expected.checksum: Added.
        * platform/gtk/fast/forms/textarea-placeholder-visibility-1-expected.png: Added.
        * platform/gtk/fast/forms/textarea-placeholder-visibility-1-expected.txt: Added.
        * platform/gtk/fast/forms/textarea-placeholder-visibility-2-expected.checksum: Added.
        * platform/gtk/fast/forms/textarea-placeholder-visibility-2-expected.png: Added.
        * platform/gtk/fast/forms/textarea-placeholder-visibility-2-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77723 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Adam Barth  <abarth@webkit.org>

        [Windows 7 Release Tests] plugins/plugin-document-back-forward.html failing since introduction in r77706
        https://bugs.webkit.org/show_bug.cgi?id=53831

        Stop dumping the child frames because that's not needed for this test
        and cause extra output on Windows.

        * plugins/plugin-document-back-forward.html:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77722 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Adam Barth  <abarth@webkit.org>

        [GTK] plugins/plugin-document-back-forward.html fails
        https://bugs.webkit.org/show_bug.cgi?id=53833

        Add plugins/plugin-document-back-forward.html to skipped list while we
        investigate the issue.

        * platform/gtk/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77721 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Tony Chang  <tony@chromium.org>

        Reviewed by Eric Seidel.

        get test-webkitpy running on win32 python
        https://bugs.webkit.org/show_bug.cgi?id=53822

        The test harness crashes with a WindowsError because it can't find
        'svn' when using subprocess.Popen.  This gets us past the error
        so we can see the failing tests on the Chromium Win Release Tests
        bot.

        * Scripts/webkitpy/common/system/executive.py:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77720 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Matthew Delaney  <mdelaney@apple.com>

        Reviewed by Dan Bernstein.

        Move canvas layout test expectations into more appropriate locations
        https://bugs.webkit.org/show_bug.cgi?id=53145

        * platform/mac-snowleopard/canvas/philip/tests/2d.drawImage.broken-expected.txt: Removed.
        * ...
        * platform/mac-snowleopard/canvas/philip/tests/type.prototype-expected.txt: Removed.
        * platform/mac/canvas: Added.
        * platform/mac/canvas/philip: Added.
        * platform/mac/canvas/philip/tests: Added.
        * platform/mac/canvas/philip/tests/2d.drawImage.broken-expected.txt: Copied from platform/mac-snowleopard/canvas/philip/tests/2d.drawImage.broken-expected.txt.
        * ...
        * platform/mac/canvas/philip/tests/type.prototype-expected.txt: Copied from platform/mac-snowleopard/canvas/philip/tests/type.prototype-expected.txt.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77719 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Windows 7 Release Tests] plugins/plugin-document-back-forward.html failing since
introduction in r77706.
https://bugs.webkit.org/show_bug.cgi?id=53831

Unreviewed. Add expected (failing?) results to get the Windows 7 Release Test bots green.

* platform/win/plugins/plugin-document-back-forward-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77718 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Martin Robinson  <mrobinson@igalia.com>

        Unskip some tests which should now be passing.

        * platform/gtk/Skipped: Unskip tests.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77717 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Xiyuan Xia  <xiyuan@chromium.org>

        Reviewed by Tony Chang.

        [Chromium] Option text in select popup does not align with menulist button text
        https://bugs.webkit.org/show_bug.cgi?id=53632

        This makes clientPaddingLeft and  clientPaddingRight return
        the additional m_innerBlock's padding so that the popup item text
        aligns with the menulist button text.

        * rendering/RenderMenuList.cpp:
        (WebCore::RenderMenuList::clientPaddingLeft):
        (WebCore::RenderMenuList::clientPaddingRight):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77716 268f45cc-cd09-0410-ab3c-d52691b4dbfc

REGRESSION: Horizontal scrollbar thumbs leave artifacts over page content when scrolling vertically
<rdar://problem/8962457>

Reviewed by Sam Weinig and Beth Dakin.

* platform/ScrollView.cpp:
(WebCore::ScrollView::scrollContents):
Subtract scrollbars from the scroll view rect if overlay scrollers are enabled.

* platform/ScrollableArea.cpp:
(WebCore::ScrollableArea::setScrollOffsetFromAnimation):
Make sure to invalidate both scrollbars if overlay scrollers are enabled.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77715 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Xan Lopez.

        [GTK] WebKitWebFrame can return a stale frame name when calling webkit_web_frame_get_name
        https://bugs.webkit.org/show_bug.cgi?id=53797

        * platform/gtk/Skipped: Unskip a test which is now passing.
2011-02-04  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Xan Lopez.

        [GTK] WebKitWebFrame can return a stale frame name when calling webkit_web_frame_get_name
        https://bugs.webkit.org/show_bug.cgi?id=53797

        When the frame name changes between different calls to webkit_web_frame_get_name
        on the same frame, return the new frame name instead of the one valid during the
        previous call.

        * webkit/webkitwebframe.cpp:
        (webkit_web_frame_init): Initialize the frame name to 0.
        (webkit_web_frame_get_name): Check the current frame name first before
        returning the cached value.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77714 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebKit2: Crash when InjectedBundlePageLoaderClient::willSendRequestForFrame returns a null WKURLRequestRef
https://bugs.webkit.org/show_bug.cgi?id=53829

Reviewed by Maciej Stachowiak.

Store the result of m_client.willSendRequestForFrame in a RefPtr, and null check it before
calling resourceRequest() on it.

* WebProcess/InjectedBundle/InjectedBundlePageLoaderClient.cpp:
(WebKit::InjectedBundlePageLoaderClient::willSendRequestForFrame):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77713 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Martin Robinson  <mrobinson@igalia.com>

        Rebaseline a few tests after r77665 and add a GTK-specific baseline
        after r77690.

        * platform/gtk/fast/blockflow/Kusa-Makura-background-canvas-expected.txt:
        * platform/gtk/fast/blockflow/japanese-ruby-horizontal-bt-expected.txt:
        * platform/gtk/fast/blockflow/japanese-ruby-vertical-lr-expected.txt:
        * platform/gtk/fast/blockflow/japanese-ruby-vertical-rl-expected.txt:
        * platform/gtk/fast/ruby/nested-ruby-expected.txt:
        * platform/gtk/fast/text/emphasis-avoid-ruby-expected.txt:
        * platform/gtk/media/video-frame-accurate-seek-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77712 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Windows 7 Release Tests] 5 ruby tests failing on Windows after r77665
https://bugs.webkit.org/show_bug.cgi?id=53820

Unreviewed. Landing expected results that are probably correct (given that the change in
77665 was to make the default ruby text size smaller by default).

* platform/win/fast/blockflow/Kusa-Makura-background-canvas-expected.checksum:
* platform/win/fast/blockflow/Kusa-Makura-background-canvas-expected.png:
* platform/win/fast/blockflow/Kusa-Makura-background-canvas-expected.txt:
* platform/win/fast/blockflow/japanese-ruby-horizontal-bt-expected.checksum:
* platform/win/fast/blockflow/japanese-ruby-horizontal-bt-expected.png:
* platform/win/fast/blockflow/japanese-ruby-horizontal-bt-expected.txt:
* platform/win/fast/blockflow/japanese-ruby-vertical-lr-expected.checksum:
* platform/win/fast/blockflow/japanese-ruby-vertical-lr-expected.png:
* platform/win/fast/blockflow/japanese-ruby-vertical-lr-expected.txt:
* platform/win/fast/blockflow/japanese-ruby-vertical-rl-expected.checksum:
* platform/win/fast/blockflow/japanese-ruby-vertical-rl-expected.png:
* platform/win/fast/blockflow/japanese-ruby-vertical-rl-expected.txt:
* platform/win/fast/ruby/nested-ruby-expected.checksum:
* platform/win/fast/ruby/nested-ruby-expected.png:
* platform/win/fast/ruby/nested-ruby-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77711 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Gustavo Noronha Silva.

        [GTK] fast/events/pagehide-timeout.html fails
        https://bugs.webkit.org/show_bug.cgi?id=53771

        * platform/gtk/Skipped: Unskip a test which is now passing.
2011-02-04  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Gustavo Noronha Silva.

        [GTK] fast/events/pagehide-timeout.html fails
        https://bugs.webkit.org/show_bug.cgi?id=53771

        Add the document browser cache model. This model is optimized for viewing
        a series of local files, while the document viewer is optimized for only
        viewing one local file. Improve the documentation on the mapping of these
        enum values to actual cache settings.

        * webkit/webkitglobals.cpp:
        (webkit_set_cache_model): Add support for the document browser cache model and
        improve comments in this section better explaining where the magic numbers come from.
        * webkit/webkitglobals.h: Added an enum value for the document browser cache model.
2011-02-04  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Gustavo Noronha Silva.

        [GTK] fast/events/pagehide-timeout.html fails
        https://bugs.webkit.org/show_bug.cgi?id=53771

        Add knowledge of the document browser cache model the DRT. When resetting
        WebKit settings to consistent values between tests, default to the document
        browser cache model.

        * DumpRenderTree/gtk/DumpRenderTree.cpp:
        (resetDefaultsToConsistentValues): Reset the cache model to the document browser
        cache model between tests.
        * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
        (LayoutTestController::setCacheModel): An int value of 2 here corresponds to the
        document browser cache model.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77710 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix build.

* WebProcess/mac/WebProcessMac.mm:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77709 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Anders Carlsson  <andersca@apple.com>

        Reviewed by Mark Rowe.

        Link opened in background tab fails to paint when activated if it completes loading before tab is selected
        https://bugs.webkit.org/show_bug.cgi?id=53827
        <rdar://problem/8962355>

        * WebProcess/WebPage/DrawingAreaImpl.cpp:
        (WebKit::DrawingAreaImpl::resumePainting):
        Invalidate the entire page.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77708 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix <rdar://problem/8961227> WebKit2 does not work with symlinked home directories.

Reviewed by Sam Weinig.

* WebProcess/mac/WebProcessMac.mm:
(WebKit::WebProcess::platformSetCacheModel):
Don't adopt an autoreleased pointer, this would lead to a crash.

(WebKit::appendSandboxParameterPath):
(WebKit::appendSandboxParameterConfPath):
(WebKit::initializeSandbox):
Clean up the sandbox parameter code and make sure that realpath is called for all parameters.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77707 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        PluginDocuments don't create widgets for plugins on back/forward
        https://bugs.webkit.org/show_bug.cgi?id=53474

        Test that plugin loads when in a plugin document on back-forward.

        * plugins/plugin-document-back-forward-expected.txt: Added.
        * plugins/plugin-document-back-forward.html: Added.
2011-02-04  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        PluginDocuments don't create widgets for plugins on back/forward
        https://bugs.webkit.org/show_bug.cgi?id=53474

        Long ago, PluginDocument always caused the HTMLEmbedElement to create
        its widget synchronously during a post-layout task.  Recently, however,
        some changes to the HistroyController caused layout on back/forward to
        become slightly more complicated (and added an extra level of recursion
        to layout).  This extra level of recursion triggered the "I've recursed
        too many times" condition in the post-layout task queue, causing the
        FrameView to run the remainder of the tasks asynchronously.
        Unfortunately, that broke PluginDocument because it needs its the
        HTMLEmbedElement's updateWidget task to run synchronously.

        This patch adds a mechanism for "kicking off" the pending post-layout
        tasks synchronously (instead of waiting for the timer to fire).
        PluginDocument then uses that facility to ensure that the
        HTMLEmbedElement's updateWidget task happens.

        Test: plugins/plugin-document-back-forward.html

        * html/PluginDocument.cpp:
        (WebCore::PluginDocumentParser::appendBytes):
        * page/FrameView.cpp:
        (WebCore::FrameView::flushAnyPendingPostLayoutTasks):
        * page/FrameView.h:
2011-02-04  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        PluginDocuments don't create widgets for plugins on back/forward
        https://bugs.webkit.org/show_bug.cgi?id=53474

        Teach the test plugin how to call alert on load.

        * DumpRenderTree/TestNetscapePlugIn/main.cpp:
        (NPP_New):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77706 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Charlie Reis  <creis@chromium.org>

        Reviewed by Mihai Parparita.

        Crash in WebCore::HistoryController::itemsAreClones
        https://bugs.webkit.org/show_bug.cgi?id=52819

        Tests that navigating back and forward between hash items works.

        * fast/history/history-back-forward-within-subframe-hash.html: Added.
        * fast/history/history-back-forward-within-subframe-hash-expected.txt: Added.
2011-02-04  Charlie Reis  <creis@chromium.org>

        Reviewed by Mihai Parparita.

        Crash in WebCore::HistoryController::itemsAreClones
        https://bugs.webkit.org/show_bug.cgi?id=52819

        Avoids deleting the current HistoryItem while it is still in use.
        Ensures that provisional items are committed for same document navigations.
        Ensures that error pages are committed on back/forward navigations.
        Also removes unneeded sanity checks used for diagnosing the problem.

        * loader/HistoryController.cpp:
        * loader/HistoryController.h:
2011-02-04  Charlie Reis  <creis@chromium.org>

        Reviewed by Mihai Parparita.

        Crash in WebCore::HistoryController::itemsAreClones
        https://bugs.webkit.org/show_bug.cgi?id=52819

        Removes unneeded sanity checks used for diagnosing a memory error.

        * src/WebFrameImpl.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77705 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        [Chromium] Add expectations for the test, added in r77690.

        * platform/chromium-win/media/video-frame-accurate-seek-expected.checksum: Added.
        * platform/chromium-win/media/video-frame-accurate-seek-expected.png: Added.
        * platform/chromium-win/media/video-frame-accurate-seek-expected.txt: Added.
        * platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77704 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Sandbox violations when visiting a site using HTTP authentication
<rdar://problem/8951079>

Reviewed by Mark Rowe.

* WebProcess/com.apple.WebProcess.sb:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77703 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Carol Szabo  <carol.szabo@nokia.com>

        Reviewed by David Hyatt.

        Code Changes.

        CSS 2.1 failure: content-*
        https://bugs.webkit.org/show_bug.cgi?id=52126

        * fast/css/counters/content-021-expected.txt: Added.
        * fast/css/counters/content-021.html: Added.
        This is a copy of the test with the same name from the
        official css test suite, adapted for DumpRenderTree.
2011-02-04  Carol Szabo  <carol.szabo@nokia.com>

        Reviewed by David Hyatt.

        Code Changes.

        CSS 2.1 failure: content-*
        https://bugs.webkit.org/show_bug.cgi?id=52126

        Test: fast/css/counters/content-021.html

        * rendering/CounterNode.cpp:
        (showCounterTree):
        Made parameter const because it is supposed to be so.
        * rendering/RenderCounter.cpp:
        (WebCore::previousInPreOrder):
        (WebCore::previousSiblingOrParent):
        (WebCore::parentElement):
        (WebCore::areRenderersElementsSiblings):
        (WebCore::nextInPreOrder):
        Added these local helper functions to help navigate the DOM tree
        enriched with :before and :after pseudo elements.
        (WebCore::planCounter):
        Fixed bug that would create a repeat counter for second and
        subsequent renderers associated with the same DOM element.
        (WebCore::findPlaceForCounter):
        (WebCore::makeCounterNode):
        Changed to use the new tree navigation functions described above
        instead of the Renderer Tree navigation functions.
        (WebCore::RenderCounter::rendererSubtreeAttached):
        (WebCore::RenderCounter::rendererStyleChanged):
        Optimized to not bother about counters until the renderers are
        finally attached.
        (showRendererTree):
        (showNodeTree):
        Debug helper functions used to debug Counter bugs.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77702 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=53823
        Print layout sometimes seen on screen

        <rdar://problem/8948742> REGRESSION (WebKit2) - When I print a page, I visually see it re-layout
        <rdar://problem/8899389> REGRESSION(WebKit2): Wrong content can appear when print dialog is up

        * UIProcess/API/mac/PageClientImpl.mm: (WebKit::PageClientImpl::setAutodisplay): Removed a
        branch that didn't do anything useful.

        * UIProcess/API/mac/WKPrintingView.h:
        * UIProcess/API/mac/WKPrintingView.mm:
        (-[WKPrintingView _suspendAutodisplay]):
        (-[WKPrintingView _delayedResumeAutodisplayTimerFired]):
        (-[WKPrintingView _delayedResumeAutodisplay]):
        (-[WKPrintingView knowsPageRange:]):
        (-[WKPrintingView beginDocument]):
        (-[WKPrintingView endDocument]):
        When page setup changes, or when going from print preview to actual printing, AppKit calls
        -endDocument just to begin it again. Delay resuming autodisplay until the next run loop
        iteration.

        * WebProcess/WebCoreSupport/WebChromeClient.cpp:
        (WebKit::WebChromeClient::invalidateContentsAndWindow):
        (WebKit::WebChromeClient::invalidateContentsForSlowScroll):
        Tightened it up from other side - we should never send pixels to UI process while in
        printing mode. This is not 100% reliable, because relayout doesn't happen immediately,
        and we can still have print layout in screen mode - but it helps.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77701 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        [Chromium] Removed Chromium-specific baselines that are no longer necessary.

        * platform/chromium-mac-leopard/fast/blockflow/japanese-ruby-vertical-lr-expected.checksum: Removed.
        * platform/chromium-mac-leopard/fast/blockflow/japanese-ruby-vertical-lr-expected.png: Removed.
        * platform/chromium-mac-leopard/fast/blockflow/japanese-ruby-vertical-rl-expected.checksum: Removed.
        * platform/chromium-mac-leopard/fast/blockflow/japanese-ruby-vertical-rl-expected.png: Removed.
        * platform/chromium-mac/fast/blockflow/Kusa-Makura-background-canvas-expected.checksum: Removed.
        * platform/chromium-mac/fast/blockflow/Kusa-Makura-background-canvas-expected.png: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77700 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Oliver Hunt.

        Rolled back in r77612 with ASSERT/crash fixed.
        https://bugs.webkit.org/show_bug.cgi?id=53759

        Don't shrink the heap to 0 unconditionally. Instead, shrink to 1 if
        necessary. For now, the heap assumes that it always has at least one
        block live.

        * runtime/Heap.cpp:
        (JSC::Heap::Heap):
        (JSC::Heap::reset):
        * runtime/Heap.h:
        * runtime/MarkedSpace.cpp:
        (JSC::MarkedSpace::allocate):
        (JSC::MarkedSpace::shrinkBlocks):
        (JSC::MarkedSpace::sweep):
        (JSC::MarkedSpace::reset):
        * runtime/MarkedSpace.h:
        (JSC::MarkedSpace::highWaterMark):
        (JSC::MarkedSpace::setHighWaterMark):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77699 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Typo fix.

* html/parser/XSSFilter.cpp:
(WebCore::HTMLNames::containsJavaScriptURL):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77698 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Make an infinite loop introduced in r77454 finite.

Reviewed by Anders Carlsson.

* html/parser/XSSFilter.cpp:
(WebCore::HTMLNames::containsJavaScriptURL):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77697 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        Update Leopard pixel results for fast/box-shadow.

        * platform/mac-leopard/fast/box-shadow/basic-shadows-expected.checksum:
        * platform/mac-leopard/fast/box-shadow/basic-shadows-expected.png:
        * platform/mac-leopard/fast/box-shadow/inset-expected.checksum:
        * platform/mac-leopard/fast/box-shadow/inset-expected.png:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77696 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix for <rdar://problem/8961580>.
Work around for <rdar://problem/8957845>.

Reviewed by Mark Rowe.

* WebProcess/com.apple.WebProcess.sb:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77695 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Darin Adler  <darin@apple.com>

        Reviewed by Dan Bernstein.

        WebKit2 doesn't hide mouse pointer on keypress
        https://bugs.webkit.org/show_bug.cgi?id=53821
        rdar://problem/8630895

        * UIProcess/PageClient.h: Renamed didNotHandleKeyEvent to
        doneWithKeyEvent and added a bool wasEventHandled argument.
        * UIProcess/WebPageProxy.cpp:
        (WebKit::WebPageProxy::didReceiveEvent): Changed code to call
        doneWithKeyEvent and call even if event was handled.

        * UIProcess/API/mac/PageClientImpl.h: Updated for above change.
        * UIProcess/API/mac/PageClientImpl.mm:
        (WebKit::PageClientImpl::doneWithKeyEvent): Ditto. But also
        call [NSCursor setHiddenUntilMouseMoves:YES] when a key down
        event was handled. This matches what -[WebHTMLView keyDown:] does.
        * UIProcess/API/qt/qwkpage.cpp:
        (QWKPagePrivate::doneWithKeyEvent): Ditto.
        * UIProcess/API/qt/qwkpage_p.h: Ditto.
        * UIProcess/win/WebView.cpp:
        (WebKit::WebView::doneWithKeyEvent): Ditto.
        * UIProcess/win/WebView.h: Ditto.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77694 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Jer Noble  <jer.noble@apple.com>

        Reviewed by Eric Carlson.

        Frame accurate seeking isn't always accurate
        https://bugs.webkit.org/show_bug.cgi?id=52697

        New pixel test that seeks to specific currentTimes and records the rendered frame.

        * media/content/test-25fps.mp4: Added.
        * media/content/test-25fps.ogv: Added.
        * media/video-frame-accurate-seek-expected.txt: Added.
        * media/video-frame-accurate-seek.html: Added.
        * platform/mac/media/video-frame-accurate-seek-expected.checksum: Added.
        * platform/mac/media/video-frame-accurate-seek-expected.png: Added.
2011-02-04  Jer Noble  <jer.noble@apple.com>

        Reviewed by Eric Carlson.

        Frame accurate seeking isn't always accurate
        https://bugs.webkit.org/show_bug.cgi?id=52697

        Test: media/video-frame-accurate-seek.html

        Make seeking slightly more accurate by rounding instead of truncating
        when converting from seconds-in-float to time/timeScale.

        * platform/graphics/mac/MediaPlayerPrivateQTKit.mm:
        (WebCore::MediaPlayerPrivateQTKit::createQTTime):
        * platform/graphics/win/MediaPlayerPrivateQuickTimeVisualContext.cpp:
        (WebCore::MediaPlayerPrivateQuickTimeVisualContext::mediaTimeForTimeValue):
        * platform/graphics/win/QTMovie.cpp:
        (QTMovie::setCurrentTime):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77690 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Jeremy Orlow  <jorlow@chromium.org>

        Reviewed by Nate Chapin.

        Second IndexedDB events overhaul patch
        https://bugs.webkit.org/show_bug.cgi?id=53813

        * storage/indexeddb/request-event-propagation-expected.txt: Added.
        * storage/indexeddb/request-event-propagation.html: Added.
2011-02-04  Jeremy Orlow  <jorlow@chromium.org>

        Reviewed by Nate Chapin.

        Second IndexedDB events overhaul patch
        https://bugs.webkit.org/show_bug.cgi?id=53813

        If an IDBRequest has a IDBTransaction, it should propogate
        events through it. In order to do this, IDBRequest needs to
        hold a transaction frontend object rather than a backend one.

        Test: storage/indexeddb/request-event-propagation.html

        * storage/IDBCursor.cpp:
        (WebCore::IDBCursor::IDBCursor):
        * storage/IDBCursor.h:
        (WebCore::IDBCursor::create):
        * storage/IDBDatabase.cpp:
        (WebCore::IDBDatabase::setSetVersionTransaction):
        (WebCore::IDBDatabase::createObjectStore):
        (WebCore::IDBDatabase::deleteObjectStore):
        * storage/IDBDatabase.h:
        * storage/IDBIndex.cpp:
        (WebCore::IDBIndex::IDBIndex):
        (WebCore::IDBIndex::openCursor):
        (WebCore::IDBIndex::openKeyCursor):
        (WebCore::IDBIndex::get):
        (WebCore::IDBIndex::getKey):
        * storage/IDBIndex.h:
        (WebCore::IDBIndex::create):
        * storage/IDBObjectStore.cpp:
        (WebCore::IDBObjectStore::IDBObjectStore):
        (WebCore::IDBObjectStore::get):
        (WebCore::IDBObjectStore::add):
        (WebCore::IDBObjectStore::put):
        (WebCore::IDBObjectStore::deleteFunction):
        (WebCore::IDBObjectStore::createIndex):
        (WebCore::IDBObjectStore::deleteIndex):
        (WebCore::IDBObjectStore::openCursor):
        * storage/IDBObjectStore.h:
        (WebCore::IDBObjectStore::create):
        * storage/IDBRequest.cpp:
        (WebCore::IDBRequest::create):
        (WebCore::IDBRequest::IDBRequest):
        (WebCore::IDBRequest::resetReadyState):
        (WebCore::IDBRequest::onSuccess):
        (WebCore::IDBRequest::dispatchEvent):
        * storage/IDBRequest.h:
        * storage/IDBTransaction.cpp:
        (WebCore::IDBTransaction::IDBTransaction):
        (WebCore::IDBTransaction::objectStore):
        (WebCore::IDBTransaction::contextDestroyed):
        (WebCore::IDBTransaction::enqueueEvent):
        * storage/IDBTransaction.h:
        * storage/IDBTransaction.idl:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77689 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Jeremy Orlow  <jorlow@chromium.org>

        Reviewed by Nate Chapin.

        First step towards event propogation within IndexedDB
        https://bugs.webkit.org/show_bug.cgi?id=53795

        This is the first step towards implementing
        http://www.w3.org/Bugs/Public/show_bug.cgi?id=11348
        within IndexedDB. I've created a method that knows how
        to capture and bubble (based on Node's dispatchGenericEvent).
        I've then changed IDBRequest to use it.

        The only functional change is that preventDefault now must
        be called in error events to prevent the transaction from
        being aborted. The tests reflect this change and there's one
        specific test to look at this behavior.

        * storage/indexeddb/cursor-index-delete-expected.txt:
        * storage/indexeddb/cursor-index-delete.html:
        * storage/indexeddb/database-quota-expected.txt:
        * storage/indexeddb/database-quota.html:
        * storage/indexeddb/duplicates-expected.txt:
        * storage/indexeddb/duplicates.html:
        * storage/indexeddb/error-causes-abort-by-default-expected.txt: Copied from LayoutTests/storage/indexeddb/cursor-index-delete-expected.txt.
        * storage/indexeddb/error-causes-abort-by-default.html: Added.
        * storage/indexeddb/index-basics-expected.txt:
        * storage/indexeddb/index-basics.html:
        * storage/indexeddb/objectstore-autoincrement-expected.txt:
        * storage/indexeddb/objectstore-autoincrement.html:
        * storage/indexeddb/objectstore-basics-expected.txt:
        * storage/indexeddb/objectstore-basics.html:
2011-02-04  Jeremy Orlow  <jorlow@chromium.org>

        Reviewed by Nate Chapin.

        First step towards event propogation within IndexedDB
        https://bugs.webkit.org/show_bug.cgi?id=53795

        This is the first step towards implementing
        http://www.w3.org/Bugs/Public/show_bug.cgi?id=11348
        within IndexedDB. I've created a method that knows how
        to capture and bubble (based on Node's dispatchGenericEvent).
        I've then changed IDBRequest to use it.

        The only functional change is that preventDefault now must
        be called in error events to prevent the transaction from
        being aborted. The tests reflect this change and there's one
        specific test to look at this behavior.

        Test: storage/indexeddb/error-causes-abort-by-default.html

        * storage/IDBAbortEvent.cpp:
        (WebCore::IDBAbortEvent::create):
        (WebCore::IDBAbortEvent::IDBAbortEvent):
        * storage/IDBAbortEvent.h:
        * storage/IDBCompleteEvent.cpp:
        (WebCore::IDBCompleteEvent::create):
        (WebCore::IDBCompleteEvent::IDBCompleteEvent):
        * storage/IDBCompleteEvent.h:
        * storage/IDBErrorEvent.cpp:
        (WebCore::IDBErrorEvent::IDBErrorEvent):
        * storage/IDBEvent.cpp:
        (WebCore::IDBEvent::IDBEvent):
        (WebCore::IDBEvent::dispatch):
        * storage/IDBEvent.h:
        * storage/IDBRequest.cpp:
        (WebCore::IDBRequest::dispatchEvent):
        * storage/IDBRequest.h:
        * storage/IDBSuccessEvent.cpp:
        (WebCore::IDBSuccessEvent::IDBSuccessEvent):
        * storage/IDBTransaction.cpp:
        (WebCore::IDBTransaction::onAbort):
        (WebCore::IDBTransaction::onComplete):
        * storage/IDBTransaction.h:
        (WebCore::IDBTransaction::backend):
        * storage/IDBTransactionBackendImpl.cpp:
        (WebCore::IDBTransactionBackendImpl::taskTimerFired):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77688 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Daniel Cheng  <dcheng@chromium.org>

        Reviewed by Dmitry Titov.

        Clone WebClipboard to be frame-specific.
        https://bugs.webkit.org/show_bug.cgi?id=53727

        For drop operations, Chrome currently snapshots the data and copies it
        into the renderer process. As we add more supported drag data types, the
        copy will become increasingly expensive. Instead, we'd like to snapshot
        data in the browser to reduce the amount of data copied and to support
        Blob in DataTransferItem. In order to allow this, we associated
        WebClipboard with a frame so it can correctly route its IPCs to the
        corresponding Chromium host.

        No new tests because no new functionality.

        * platform/chromium/ChromiumDataObject.cpp:
        (WebCore::ChromiumDataObject::createReadable):
        * platform/chromium/ChromiumDataObject.h:
        * platform/chromium/ClipboardChromium.cpp:
        (WebCore::ClipboardChromium::create):
        * platform/chromium/PlatformBridge.h:
        * platform/chromium/ReadableDataObject.cpp:
        (WebCore::ReadableDataObject::create):
        (WebCore::ReadableDataObject::ReadableDataObject):
        (WebCore::ReadableDataObject::getData):
        (WebCore::ReadableDataObject::urlTitle):
        (WebCore::ReadableDataObject::htmlBaseUrl):
        (WebCore::ReadableDataObject::filenames):
        (WebCore::ReadableDataObject::ensureTypeCacheInitialized):
        * platform/chromium/ReadableDataObject.h:
2011-02-04  Daniel Cheng  <dcheng@chromium.org>

        Reviewed by Dmitry Titov.

        Clone WebClipboard to be frame-specific.
        https://bugs.webkit.org/show_bug.cgi?id=53727

        For drop operations, Chrome currently snapshots the data and copies it
        into the renderer process. As we add more supported drag data types, the
        copy will become increasingly expensive. Instead, we'd like to snapshot
        data in the browser to reduce the amount of data copied and to support
        Blob in DataTransferItem. In order to allow this, we associated
        WebClipboard with a frame so it can correctly route its IPCs to the
        corresponding Chromium host.

        * public/WebFrameClient.h:
        (WebKit::WebFrameClient::clipboard):
        * src/PlatformBridge.cpp:
        (WebCore::getClipboard):
        (WebCore::PlatformBridge::clipboardReadAvailableTypes):
        (WebCore::PlatformBridge::clipboardReadData):
        (WebCore::PlatformBridge::clipboardReadFilenames):
        * src/WebFrameImpl.cpp:
        (WebKit::WebFrameImpl::fromFrame):
        * src/WebFrameImpl.h:
        * src/WebViewImpl.cpp:
        (WebKit::WebViewImpl::dragTargetDragEnterNew):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77687 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Chromium] Remove Chromium-specific baselines that we no longer need.

* platform/chromium-mac-leopard/fast/ruby/nested-ruby-expected.checksum: Removed.
* platform/chromium-mac-leopard/fast/ruby/nested-ruby-expected.png: Removed.
* platform/chromium-mac-leopard/fast/ruby/ruby-beforeafter-expected.checksum: Removed.
* platform/chromium-mac-leopard/fast/ruby/ruby-beforeafter-expected.png: Removed.
* platform/chromium-mac-leopard/fast/ruby/ruby-run-break-expected.checksum: Removed.
* platform/chromium-mac-leopard/fast/ruby/ruby-run-break-expected.png: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77686 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Jeremy Orlow  <jorlow@chromium.org>

        Revert https://bugs.webkit.org/show_bug.cgi?id=53795

        * storage/indexeddb/cursor-index-delete-expected.txt:
        * storage/indexeddb/cursor-index-delete.html:
        * storage/indexeddb/cursor-update-expected.txt:
        * storage/indexeddb/cursor-update.html:
        * storage/indexeddb/database-quota-expected.txt:
        * storage/indexeddb/database-quota.html:
        * storage/indexeddb/duplicates-expected.txt:
        * storage/indexeddb/duplicates.html:
        * storage/indexeddb/error-causes-abort-by-default-expected.txt: Removed.
        * storage/indexeddb/error-causes-abort-by-default.html: Removed.
        * storage/indexeddb/index-basics-expected.txt:
        * storage/indexeddb/index-basics.html:
        * storage/indexeddb/objectstore-autoincrement-expected.txt:
        * storage/indexeddb/objectstore-autoincrement.html:
        * storage/indexeddb/objectstore-basics-expected.txt:
        * storage/indexeddb/objectstore-basics.html:
2011-02-04  Jeremy Orlow  <jorlow@chromium.org>

        Revert https://bugs.webkit.org/show_bug.cgi?id=53795

        * storage/IDBAbortEvent.cpp:
        (WebCore::IDBAbortEvent::create):
        (WebCore::IDBAbortEvent::IDBAbortEvent):
        * storage/IDBAbortEvent.h:
        * storage/IDBCompleteEvent.cpp:
        (WebCore::IDBCompleteEvent::create):
        (WebCore::IDBCompleteEvent::IDBCompleteEvent):
        * storage/IDBCompleteEvent.h:
        * storage/IDBErrorEvent.cpp:
        (WebCore::IDBErrorEvent::IDBErrorEvent):
        * storage/IDBEvent.cpp:
        (WebCore::IDBEvent::IDBEvent):
        * storage/IDBEvent.h:
        * storage/IDBRequest.cpp:
        (WebCore::IDBRequest::dispatchEvent):
        * storage/IDBRequest.h:
        * storage/IDBSuccessEvent.cpp:
        (WebCore::IDBSuccessEvent::IDBSuccessEvent):
        * storage/IDBTransaction.cpp:
        (WebCore::IDBTransaction::onAbort):
        (WebCore::IDBTransaction::onComplete):
        * storage/IDBTransaction.h:
        * storage/IDBTransactionBackendImpl.cpp:
        (WebCore::IDBTransactionBackendImpl::taskTimerFired):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77685 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        [Chromium] Marked test, added in r77680 as failing due to lighttpd.

        * platform/chromium/test_expectations.txt: Marked.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77684 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        Update Leopard baselines for fast/blockflow, this time correctly.

        * platform/mac-leopard/fast/blockflow/Kusa-Makura-background-canvas-expected.checksum:
        * platform/mac-leopard/fast/blockflow/Kusa-Makura-background-canvas-expected.png:
        * platform/mac-leopard/fast/blockflow/Kusa-Makura-background-canvas-expected.txt: Removed.
        * platform/mac-leopard/fast/blockflow/fieldsets-expected.png:
        * platform/mac-leopard/fast/blockflow/japanese-ruby-horizontal-bt-expected.checksum: Removed.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-horizontal-bt-expected.png: Removed.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-horizontal-bt-expected.txt: Removed.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-lr-expected.checksum:
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-lr-expected.png:
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-lr-expected.txt: Removed.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-rl-expected.checksum:
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-rl-expected.png:
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-rl-expected.txt: Removed.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77683 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dirk Pranke  <dpranke@chromium.org>

        Unreviewed, expectations change.

        Merge in some GPU expectations that got dropped in r77666, get lint to pass.

        * platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77682 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        Add Leopard-specific pixel results for fast/ruby.

        * platform/mac-leopard/fast/ruby/nested-ruby-expected.checksum: Added.
        * platform/mac-leopard/fast/ruby/nested-ruby-expected.png: Added.
        * platform/mac-leopard/fast/ruby/ruby-beforeafter-expected.checksum: Added.
        * platform/mac-leopard/fast/ruby/ruby-beforeafter-expected.png: Added.
        * platform/mac-leopard/fast/ruby/ruby-empty-rt-expected.checksum:
        * platform/mac-leopard/fast/ruby/ruby-empty-rt-expected.png:
        * platform/mac-leopard/fast/ruby/ruby-length-expected.checksum:
        * platform/mac-leopard/fast/ruby/ruby-length-expected.png:
        * platform/mac-leopard/fast/ruby/ruby-run-break-expected.checksum:
        * platform/mac-leopard/fast/ruby/ruby-run-break-expected.png:
        * platform/mac-leopard/fast/ruby/ruby-runs-expected.checksum:
        * platform/mac-leopard/fast/ruby/ruby-runs-expected.png:
        * platform/mac-leopard/fast/ruby/ruby-runs-spans-expected.checksum:
        * platform/mac-leopard/fast/ruby/ruby-runs-spans-expected.png:
        * platform/mac-leopard/fast/ruby/ruby-simple-expected.checksum:
        * platform/mac-leopard/fast/ruby/ruby-simple-expected.png:
        * platform/mac-leopard/fast/ruby/ruby-simple-rp-expected.checksum:
        * platform/mac-leopard/fast/ruby/ruby-simple-rp-expected.png:
        * platform/mac-leopard/fast/ruby/ruby-trailing-expected.checksum:
        * platform/mac-leopard/fast/ruby/ruby-trailing-expected.png:
        * platform/mac-leopard/fast/ruby/rubyDOM-insert-rt-expected.checksum:
        * platform/mac-leopard/fast/ruby/rubyDOM-insert-rt-expected.png:
        * platform/mac-leopard/fast/ruby/rubyDOM-insert-text1-expected.checksum:
        * platform/mac-leopard/fast/ruby/rubyDOM-insert-text1-expected.png:
        * platform/mac-leopard/fast/ruby/rubyDOM-insert-text2-expected.checksum:
        * platform/mac-leopard/fast/ruby/rubyDOM-insert-text2-expected.png:
        * platform/mac-leopard/fast/ruby/rubyDOM-insert-text3-expected.checksum:
        * platform/mac-leopard/fast/ruby/rubyDOM-insert-text3-expected.png:
        * platform/mac-leopard/fast/ruby/rubyDOM-remove-rt1-expected.checksum:
        * platform/mac-leopard/fast/ruby/rubyDOM-remove-rt1-expected.png:
        * platform/mac-leopard/fast/ruby/rubyDOM-remove-rt2-expected.checksum:
        * platform/mac-leopard/fast/ruby/rubyDOM-remove-rt2-expected.png:
        * platform/mac-leopard/fast/ruby/rubyDOM-remove-text1-expected.checksum:
        * platform/mac-leopard/fast/ruby/rubyDOM-remove-text1-expected.png:
        * platform/mac-leopard/fast/ruby/rubyDOM-remove-text2-expected.checksum:
        * platform/mac-leopard/fast/ruby/rubyDOM-remove-text2-expected.png:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77681 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Martin Galpin  <martin@66laps.com>

        Reviewed by David Levin.

        CORS origin header not set on GET when a preflight request is required.
        https://bugs.webkit.org/show_bug.cgi?id=50773

        * http/tests/xmlhttprequest/cross-origin-preflight-get-expected.txt: Added.
        * http/tests/xmlhttprequest/cross-origin-preflight-get.html: Added.
        * http/tests/xmlhttprequest/resources/cross-origin-preflight-get.php: Added.
2011-02-04  Martin Galpin  <martin@66laps.com>

        Reviewed by David Levin.

        CORS origin header not set on GET when a preflight request is required.
        https://bugs.webkit.org/show_bug.cgi?id=50773

        Test: http/tests/xmlhttprequest/cross-origin-preflight-get.html

        * loader/DocumentThreadableLoader.cpp:
        (WebCore::DocumentThreadableLoader::preflightSuccess):
        Explicitly set the request origin after a preflight request succeeds.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77680 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Jeremy Orlow  <jorlow@chromium.org>

        Reviewed by Nate Chapin.

        First step towards event propogation within IndexedDB
        https://bugs.webkit.org/show_bug.cgi?id=53795

        This is the first step towards implementing
        http://www.w3.org/Bugs/Public/show_bug.cgi?id=11348
        within IndexedDB. I've created a method that knows how
        to capture and bubble (based on Node's dispatchGenericEvent).
        I've then changed IDBRequest to use it.

        The only functional change is that preventDefault now must
        be called in error events to prevent the transaction from
        being aborted. The tests reflect this change and there's one
        specific test to look at this behavior.

        * storage/indexeddb/cursor-index-delete-expected.txt:
        * storage/indexeddb/cursor-index-delete.html:
        * storage/indexeddb/database-quota-expected.txt:
        * storage/indexeddb/database-quota.html:
        * storage/indexeddb/duplicates-expected.txt:
        * storage/indexeddb/duplicates.html:
        * storage/indexeddb/error-causes-abort-by-default-expected.txt: Copied from LayoutTests/storage/indexeddb/cursor-index-delete-expected.txt.
        * storage/indexeddb/error-causes-abort-by-default.html: Added.
        * storage/indexeddb/index-basics-expected.txt:
        * storage/indexeddb/index-basics.html:
        * storage/indexeddb/objectstore-autoincrement-expected.txt:
        * storage/indexeddb/objectstore-autoincrement.html:
        * storage/indexeddb/objectstore-basics-expected.txt:
        * storage/indexeddb/objectstore-basics.html:
2011-02-04  Jeremy Orlow  <jorlow@chromium.org>

        Reviewed by Nate Chapin.

        First step towards event propogation within IndexedDB
        https://bugs.webkit.org/show_bug.cgi?id=53795

        This is the first step towards implementing
        http://www.w3.org/Bugs/Public/show_bug.cgi?id=11348
        within IndexedDB. I've created a method that knows how
        to capture and bubble (based on Node's dispatchGenericEvent).
        I've then changed IDBRequest to use it.

        The only functional change is that preventDefault now must
        be called in error events to prevent the transaction from
        being aborted. The tests reflect this change and there's one
        specific test to look at this behavior.

        Test: storage/indexeddb/error-causes-abort-by-default.html

        * storage/IDBAbortEvent.cpp:
        (WebCore::IDBAbortEvent::create):
        (WebCore::IDBAbortEvent::IDBAbortEvent):
        * storage/IDBAbortEvent.h:
        * storage/IDBCompleteEvent.cpp:
        (WebCore::IDBCompleteEvent::create):
        (WebCore::IDBCompleteEvent::IDBCompleteEvent):
        * storage/IDBCompleteEvent.h:
        * storage/IDBErrorEvent.cpp:
        (WebCore::IDBErrorEvent::IDBErrorEvent):
        * storage/IDBEvent.cpp:
        (WebCore::IDBEvent::IDBEvent):
        (WebCore::IDBEvent::dispatch):
        * storage/IDBEvent.h:
        * storage/IDBRequest.cpp:
        (WebCore::IDBRequest::dispatchEvent):
        * storage/IDBRequest.h:
        * storage/IDBSuccessEvent.cpp:
        (WebCore::IDBSuccessEvent::IDBSuccessEvent):
        * storage/IDBTransaction.cpp:
        (WebCore::IDBTransaction::onAbort):
        (WebCore::IDBTransaction::onComplete):
        * storage/IDBTransaction.h:
        (WebCore::IDBTransaction::backend):
        * storage/IDBTransactionBackendImpl.cpp:
        (WebCore::IDBTransactionBackendImpl::taskTimerFired):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77679 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Steve Falkenburg  <sfalken@apple.com>

        Reviewed by Adam Roben.

        WebKit2 clients fail to launch on Windows if client exe is in a different directory from WebKit
        https://bugs.webkit.org/show_bug.cgi?id=53810
        <rdar://problem/8946085>

        If WebKit2WebProcess.exe is in a different directory from the client process, and that
        client process uses CreateProcess to launch WebKit2WebProcess, the DLL search path may
        end up excluding the directory containing WebKit. To address this, we use SetDllDirectory
        to explicitly include the WebKit directory in the DLL search path.

        * win/MainWin.cpp:
        (wWinMain): Use SetDllDirectory and LoadLibrary to load WebKit.dll from WebKit2WebProcess.exe
        * win/WebKit2WebProcessCommon.vsprops: Don't link directly to WebKit. Link against shlwapi.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77678 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Steve Falkenburg  <sfalken@apple.com>

        Reviewed by Adam Roben.

        Bump Windows minimum system requirement to XPSP2
        https://bugs.webkit.org/show_bug.cgi?id=53807

        * win/tools/vsprops/common.vsprops:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77677 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        [Chromium] Updated Win and Linux baselines after r77665.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77676 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Xiaomei Ji  <xji@chromium.org>

        Unreviewed.

        Rebaseline chromium after r77654.
        text-align attribute in <select> is ignored both in button text and drop-down
        items.

        * platform/chromium-linux/fast/text/international/bidi-menulist-expected.checksum:
        * platform/chromium-linux/fast/text/international/bidi-menulist-expected.png:
        * platform/chromium-linux/fast/text/international/bidi-menulist-expected.txt:
        * platform/chromium-win/fast/text/international/bidi-menulist-expected.checksum:
        * platform/chromium-win/fast/text/international/bidi-menulist-expected.png:
        * platform/chromium-win/fast/text/international/bidi-menulist-expected.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77675 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        Add Leopard-specific pixel results for fast/blockflow.

        * platform/mac-leopard/fast/blockflow/Kusa-Makura-background-canvas-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/Kusa-Makura-background-canvas-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/Kusa-Makura-background-canvas-expected.txt: Added.
        * platform/mac-leopard/fast/blockflow/block-level-images-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/block-level-images-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/border-vertical-lr-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/border-vertical-lr-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/english-bt-text-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/english-bt-text-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/english-lr-text-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/english-lr-text-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/japanese-lr-selection-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/japanese-lr-selection-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/japanese-lr-text-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/japanese-lr-text-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/japanese-rl-selection-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/japanese-rl-selection-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/japanese-rl-text-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/japanese-rl-text-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-horizontal-bt-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-horizontal-bt-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-horizontal-bt-expected.txt: Added.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-lr-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-lr-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-lr-expected.txt: Added.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-rl-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-rl-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/japanese-ruby-vertical-rl-expected.txt: Added.
        * platform/mac-leopard/fast/blockflow/vertical-align-table-baseline-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/vertical-align-table-baseline-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/vertical-baseline-alignment-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/vertical-baseline-alignment-expected.png: Added.
        * platform/mac-leopard/fast/blockflow/vertical-font-fallback-expected.checksum: Added.
        * platform/mac-leopard/fast/blockflow/vertical-font-fallback-expected.png: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77674 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        Remove platform-specific baselines, since we match platform/mac here.

        * platform/chromium-mac/fast/blockflow/box-shadow-horizontal-bt-expected.checksum: Removed.
        * platform/chromium-mac/fast/blockflow/box-shadow-horizontal-bt-expected.png: Removed.
        * platform/chromium-mac/fast/blockflow/box-shadow-vertical-lr-expected.checksum: Removed.
        * platform/chromium-mac/fast/blockflow/box-shadow-vertical-lr-expected.png: Removed.
        * platform/chromium-mac/fast/blockflow/box-shadow-vertical-rl-expected.checksum: Removed.
        * platform/chromium-mac/fast/blockflow/box-shadow-vertical-rl-expected.png: Removed.
        * platform/chromium/test_expectations.txt:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77673 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add a test case for encode/decode of FormData/FormDataElement

If the fix made in r77401 is broken, this new test will crash.

Fixes <http://webkit.org/b/53616> <rdar://problem/8949883>.

Reviewed by Sam Weinig.

Tools:

* TestWebKitAPI/Test.h: Moved the bulk of TEST_ASSERT into a new _TEST_ASSERT_HELPER macro.
Added a new TEST_ASSERT_RETURN macro that can be used in functions with a return value.

* TestWebKitAPI/Tests/WebKit2/RestoreSessionStateContainingFormData.cpp: Added.
(TestWebKitAPI::didFinishLoadForFrame): Record that the load is finished.
(TestWebKitAPI::setPageLoaderClient): Hook up the loader client.
(TestWebKitAPI::createSessionStateContainingFormData): Load simple-form.html, submit the
form, an return the session state data.
(TestWebKitAPI::TEST): Create some session state that contains form data from one WKPage,
and restore it into another WKPage.

* TestWebKitAPI/Tests/WebKit2/simple-form.html: Added.

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
* TestWebKitAPI/win/copy-resources.cmd:
Added new files.

Source/WebKit2:

Allow passing a null filter callback to WKPageCopySessionState

* UIProcess/cf/WebBackForwardListCF.cpp:
(WebKit::WebBackForwardList::createCFDictionaryRepresentation): Check if the filter is null
before calling it.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77672 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Move code to run JavaScript tests into its own files

This will allow other tests to use this mechanism.

Fixes <http://webkit.org/b/53806> SpacebarScrolling should share its JavaScript-fu with the
world

Reviewed by Sam Weinig.

* TestWebKitAPI/JavaScriptTest.cpp: Added.
* TestWebKitAPI/JavaScriptTest.h: Added.
Moved code here...

* TestWebKitAPI/Tests/WebKit2/SpacebarScrolling.cpp: ...from here.

* TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
* TestWebKitAPI/win/TestWebKitAPI.vcproj:
Added the new files.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77671 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        [Chromium] Update mac expectations.

        * platform/chromium/test_expectations.txt: Updated.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77670 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Csaba Osztrogonác.

        [Qt]REGRESSION(r76951): media/controls-without-preload.html fails
        https://bugs.webkit.org/show_bug.cgi?id=53674

        * platform/qt/Skipped:
2011-02-04  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Csaba Osztrogonác.

        [Qt]REGRESSION(r76951): media/controls-without-preload.html fails
        https://bugs.webkit.org/show_bug.cgi?id=53674

        * css/mediaControlsQt.css:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@77669 268f45cc-cd09-0410-ab3c-d52691b4dbfc