WebKit-https.git
11 years ago2009-04-28 Gavin Barraclough <barraclough@apple.com>
barraclough@apple.com [Tue, 28 Apr 2009 08:08:07 +0000 (08:08 +0000)]
2009-04-28  Gavin Barraclough  <barraclough@apple.com>

        Rubber stamped by Maciej "Henry Morgan" Stachowiak.

        Enable YARR.
        (Again.)

        * wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42923 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Gavin Barraclough <barraclough@apple.com>
barraclough@apple.com [Tue, 28 Apr 2009 07:10:34 +0000 (07:10 +0000)]
2009-04-27  Gavin Barraclough  <barraclough@apple.com>

        Reviewed by Maciej Stachowiak.

        Tweak a loop condition to keep GCC happy,
        some GCCs seem to be having issues with this. :-/

        * bytecompiler/BytecodeGenerator.cpp:
        (JSC::BytecodeGenerator::breakTarget):
        * wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42922 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Sam Weinig <sam@webkit.org>
weinig@apple.com [Tue, 28 Apr 2009 06:05:03 +0000 (06:05 +0000)]
2009-04-27  Sam Weinig  <sam@webkit.org>

        Reviewed by Mark Rowe.

        Disable test intermittent failing test.

        * svg/custom/filter-source-alpha.svg: Removed.
        * svg/custom/filter-source-alpha.svg-disabled: Copied from svg/custom/filter-source-alpha.svg.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42921 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Pierre d'Herbemont <pdherbemont@apple.com>
simon.fraser@apple.com [Tue, 28 Apr 2009 05:14:16 +0000 (05:14 +0000)]
2009-04-27  Pierre d'Herbemont  <pdherbemont@apple.com>

        Reviewed by Simon Fraser

        Fix video-played.html layout test.

        https://bugs.webkit.org/show_bug.cgi?id=25424

        We implement playForMillisecs() which makes sure the video
        will have been playing for a minimum time.

        Note, this does not ensure a strict synchronization between the time
        asked and the media playback. But given that media playback is the
        slow part we just need to compensate for the delay involved by playback.
        (And not the other way around.)

        Moreover, we split the test in 3 in order to be more gentle which
        slower machines that may tend to timeout more easily on those tests.

        * media/video-played-collapse-expected.txt: Added.
        * media/video-played-collapse.html: Added.
        * media/video-played-expected.txt:
        * media/video-played-reset-expected.txt: Added.
        * media/video-played-reset.html: Added.
        * media/video-played.html:
        - Split tests in multiple files.
        - Makes sure we always play as much as requested by using
        playForMillisecs();
        * media/video-played.js: Added.
        (testRanges):
        (nextTest):
        (pause):
        (canplay):
        (willCreateNewRange):
        (startPlayingInNewRange):
        (startPlaying):
        (secToMilli):
        (milliToSecs):
        (playForMillisecs.callPauseIfTimeIsReached):
        (playForMillisecs):
        (videoPlayedMain):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42920 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Eric Roman <eroman@chromium.org>
dglazkov@chromium.org [Tue, 28 Apr 2009 03:39:20 +0000 (03:39 +0000)]
2009-04-27  Eric Roman  <eroman@chromium.org>

        Reviewed by Darin Fisher.

        Fix a compile breakage.
        <https://bugs.webkit.org/show_bug.cgi?id=25384>

        * bindings/v8/custom/V8DOMWindowCustom.cpp:
        (WebCore::CALLBACK_FUNC_DECL):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42919 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Eric Carlson <eric.carlson@apple.com>
eric.carlson@apple.com [Tue, 28 Apr 2009 01:01:00 +0000 (01:01 +0000)]
2009-04-27  Eric Carlson  <eric.carlson@apple.com>

        Reviewed by Darin Adler.

        <rdar://problem/6822344> Use of QTMovieCurrentSizeAttribute generates exception and will crash

        QTMovieCurrentSizeAttribute generates an exception with some versions of QTKit, so calculate a
        multiplier to scale from natural size to current size when a movie is opened and use that to
        return the correct value from the naturalSize() method.

        * platform/graphics/mac/MediaPlayerPrivateQTKit.h:
        * platform/graphics/mac/MediaPlayerPrivateQTKit.mm:
        (WebCore::MediaPlayerPrivate::MediaPlayerPrivate): Initialize m_scaleFactor.
        (WebCore::MediaPlayerPrivate::naturalSize): Return naturalSize transformed by initial scale.
        (WebCore::MediaPlayerPrivate::cacheMovieScale): New, calculate difference between initial
        size and natural size so naturalSize() accounts for non-identity movie matrix.
        (WebCore::MediaPlayerPrivate::updateStates): Call cacheMovieScale when load state reaches
        QTMovieLoadStateLoaded for the first time.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42918 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
bdakin@apple.com [Tue, 28 Apr 2009 00:54:57 +0000 (00:54 +0000)]
WebCore:

2009-04-27  Beth Dakin  <bdakin@apple.com>

        Reviewed by Dave Hyatt.

        Fix for <rdar://problem/6709057> REGRESSION (4PB-TOT): http://www.winway.com/main3/support/faq.aspx selector doesn't look right or work

        The problem here is that we had a self-painting layer nested inside
        a non-self-painting layer. We ended up ignoring the self-painting
        layer both during painting and hit-testing. This patch corrects
        that.

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::paintLayer):
        (WebCore::RenderLayer::hitTestLayer):

LayoutTests:

2009-04-27  Beth Dakin  <bdakin@apple.com>

        Reviewed by Dave Hyatt.

        Test for <rdar://problem/6709057> REGRESSION (4PB-TOT): http://www.winway.com/main3/support/faq.aspx selector doesn't look right or work

        * fast/css/nested-layers-with-hover.html: Added.
        * platform/mac/fast/css/nested-layers-with-hover-expected.checksum: Added.
        * platform/mac/fast/css/nested-layers-with-hover-expected.png: Added.
        * platform/mac/fast/css/nested-layers-with-hover-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42917 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago If it has been set to allow any HTTPS certificates for this host, set
adachan@apple.com [Tue, 28 Apr 2009 00:23:03 +0000 (00:23 +0000)]
    If it has been set to allow any HTTPS certificates for this host, set
        kCFStreamSSLAllowsExpiredCertificates to true to allow expired
        certificate, and set kCFStreamSSLValidatesCertificateChain false to skip
        certificate chain validation. (<rdar://problem/6382059>)

        Reviewed by John Sullivan.

        * platform/network/cf/ResourceHandleCFNet.cpp:
        (WebCore::makeFinalRequest):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42916 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-23 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Tue, 28 Apr 2009 00:12:31 +0000 (00:12 +0000)]
2009-04-23  Simon Fraser  <simon.fraser@apple.com>

        https://bugs.webkit.org/show_bug.cgi?id=22242

        Add pixel test for zoomed video with controls, with and without a transform.

        * media/video-zoom-controls.html: Added.
        * platform/mac/media/video-zoom-controls-expected.checksum: Added.
        * platform/mac/media/video-zoom-controls-expected.png: Added.
        * platform/mac/media/video-zoom-controls-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42915 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Tue, 28 Apr 2009 00:08:33 +0000 (00:08 +0000)]
2009-04-27  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Adele Peterson

        <rdar://problem/6822225> Full page zooming of <video> element in the browser window doesn't work well

        Fix a 'size changed' notification thrash that caused standalone video elements
        to continually resize after full page zooming, by using the movie's natual size (which is independent
        of zooming), rather than its current size (which is not).

        Note that this regresses <https://bugs.webkit.org/show_bug.cgi?id=25029>, so we have to
        disable the media/video-size-intrinsic-scale.html test. However, we have to stop using
        QTMovieCurrentSizeAttribute anyway; this will be addressed, and the test re-enabled via
        <rdar://problem/6822344>.

        * platform/graphics/mac/MediaPlayerPrivateQTKit.mm:
        (WebCore::MediaPlayerPrivate::naturalSize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42914 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Mon, 27 Apr 2009 23:58:18 +0000 (23:58 +0000)]
2009-04-27  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Darin Adler

        <rdar://problem/6642221> REGRESSION: With the Movie widget, movie trailer fails to
        load if movie had been previously viewed

        Fix the Dashboard quirk for display:none plug-ins by moving the code from HTMLObjectElement
        to HTMLEmebedElement. It has to be in HTMLEmbedElement because the content we care about uses <embed>.

        * html/HTMLEmbedElement.cpp:
        (WebCore::HTMLEmbedElement::rendererIsNeeded):
        * html/HTMLObjectElement.cpp:
        (WebCore::HTMLObjectElement::rendererIsNeeded):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42913 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Sam Weinig <sam@webkit.org>
weinig@apple.com [Mon, 27 Apr 2009 23:33:57 +0000 (23:33 +0000)]
2009-04-27  Sam Weinig  <sam@webkit.org>

        Reviewed by Geoffrey Garen.

        Move more window event related code into DOMWindow.

        * dom/Document.cpp:
        (WebCore::Document::implicitClose):
        (WebCore::Document::setWindowAttributeEventListener):
        (WebCore::Document::dispatchWindowEvent):
        (WebCore::Document::dispatchLoadEvent):
        * dom/Document.h:
        * dom/Element.cpp:
        * dom/Element.h:
        * dom/Node.cpp:
        * dom/Node.h:
        * html/HTMLBodyElement.cpp:
        (WebCore::HTMLBodyElement::parseMappedAttribute):
        * html/HTMLFrameSetElement.cpp:
        (WebCore::HTMLFrameSetElement::parseMappedAttribute):
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::stopLoading):
        * page/DOMWindow.cpp:
        (WebCore::DOMWindow::dispatchAllPendingUnloadEvents):
        (WebCore::DOMWindow::postMessageTimerFired):
        (WebCore::DOMWindow::dispatchEvent):
        (WebCore::DOMWindow::dispatchEventWithDocumentAsTarget):
        (WebCore::DOMWindow::dispatchLoadEvent):
        (WebCore::DOMWindow::dispatchUnloadEvent):
        (WebCore::DOMWindow::dispatchBeforeUnloadEvent):
        * page/DOMWindow.h:
        * page/Frame.cpp:
        (WebCore::Frame::shouldClose):
        * svg/SVGSVGElement.cpp:
        (WebCore::SVGSVGElement::parseMappedAttribute):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42912 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
adele@apple.com [Mon, 27 Apr 2009 23:14:33 +0000 (23:14 +0000)]
WebCore:

2009-04-27  Douglas R. Davidson  <ddavidso@apple.com>

        Implement the editing mechanisms and add context menu items for
        <rdar://problem/6724106> WebViews need to implement text checking

        Reviewed by Justin Garcia.

        * editing/Editor.cpp:
        * editing/Editor.h:
        * editing/TypingCommand.cpp:
        * loader/EmptyClients.h:
        * page/ContextMenuController.cpp:
        * page/EditorClient.h:
        * page/mac/WebCoreViewFactory.h:
        * platform/ContextMenu.cpp:
        * platform/ContextMenuItem.h:
        * platform/LocalizedStrings.h:
        * platform/mac/LocalizedStringsMac.mm:

WebKit:

2009-04-27  Douglas R. Davidson  <ddavidso@apple.com>

        Add localizable strings for
        <rdar://problem/6724106> WebViews need to implement text checking

        Reviewed by Justin Garcia.

        * English.lproj/Localizable.strings:

WebKit/mac:

2009-04-27  Douglas R. Davidson  <ddavidso@apple.com>

        Add the various switches and context menu items needed for
        <rdar://problem/6724106> WebViews need to implement text checking
        and adopt updatePanels in place of old SPI _updateGrammar.

        Reviewed by Justin Garcia.

        * WebCoreSupport/WebContextMenuClient.mm:
        (fixMenusReceivedFromOldClients):
        * WebCoreSupport/WebEditorClient.h:
        * WebCoreSupport/WebEditorClient.mm:
        (WebEditorClient::uppercaseWord):
        (WebEditorClient::lowercaseWord):
        (WebEditorClient::capitalizeWord):
        (WebEditorClient::showSubstitutionsPanel):
        (WebEditorClient::substitutionsPanelIsShowing):
        (WebEditorClient::toggleSmartInsertDelete):
        (WebEditorClient::isAutomaticQuoteSubstitutionEnabled):
        (WebEditorClient::toggleAutomaticQuoteSubstitution):
        (WebEditorClient::isAutomaticLinkDetectionEnabled):
        (WebEditorClient::toggleAutomaticLinkDetection):
        (WebEditorClient::isAutomaticDashSubstitutionEnabled):
        (WebEditorClient::toggleAutomaticDashSubstitution):
        (WebEditorClient::isAutomaticTextReplacementEnabled):
        (WebEditorClient::toggleAutomaticTextReplacement):
        (WebEditorClient::isAutomaticSpellingCorrectionEnabled):
        (WebEditorClient::toggleAutomaticSpellingCorrection):
        (WebEditorClient::checkTextOfParagraph):
        * WebCoreSupport/WebViewFactory.mm:
        (-[WebViewFactory contextMenuItemTagShowColors]):
        (-[WebViewFactory contextMenuItemTagCorrectSpellingAutomatically]):
        (-[WebViewFactory contextMenuItemTagSubstitutionsMenu]):
        (-[WebViewFactory contextMenuItemTagShowSubstitutions:]):
        (-[WebViewFactory contextMenuItemTagSmartCopyPaste]):
        (-[WebViewFactory contextMenuItemTagSmartQuotes]):
        (-[WebViewFactory contextMenuItemTagSmartDashes]):
        (-[WebViewFactory contextMenuItemTagSmartLinks]):
        (-[WebViewFactory contextMenuItemTagTextReplacement]):
        (-[WebViewFactory contextMenuItemTagTransformationsMenu]):
        (-[WebViewFactory contextMenuItemTagMakeUpperCase]):
        (-[WebViewFactory contextMenuItemTagMakeLowerCase]):
        (-[WebViewFactory contextMenuItemTagCapitalize]):
        * WebView/WebHTMLView.mm:
        (-[WebHTMLView validateUserInterfaceItemWithoutDelegate:]):
        (-[WebHTMLView isAutomaticQuoteSubstitutionEnabled]):
        (-[WebHTMLView setAutomaticQuoteSubstitutionEnabled:]):
        (-[WebHTMLView toggleAutomaticQuoteSubstitution:]):
        (-[WebHTMLView isAutomaticLinkDetectionEnabled]):
        (-[WebHTMLView setAutomaticLinkDetectionEnabled:]):
        (-[WebHTMLView toggleAutomaticLinkDetection:]):
        (-[WebHTMLView isAutomaticDashSubstitutionEnabled]):
        (-[WebHTMLView setAutomaticDashSubstitutionEnabled:]):
        (-[WebHTMLView toggleAutomaticDashSubstitution:]):
        (-[WebHTMLView isAutomaticTextReplacementEnabled]):
        (-[WebHTMLView setAutomaticTextReplacementEnabled:]):
        (-[WebHTMLView toggleAutomaticTextReplacement:]):
        (-[WebHTMLView isAutomaticSpellingCorrectionEnabled]):
        (-[WebHTMLView setAutomaticSpellingCorrectionEnabled:]):
        (-[WebHTMLView toggleAutomaticSpellingCorrection:]):
        * WebView/WebHTMLViewInternal.h:
        * WebView/WebPreferenceKeysPrivate.h:
        * WebView/WebUIDelegatePrivate.h:
        * WebView/WebView.mm:
        (-[WebViewPrivate init]):
        (-[WebView validateUserInterfaceItemWithoutDelegate:]):
        (-[WebView setGrammarCheckingEnabled:]):
        (-[WebView isAutomaticQuoteSubstitutionEnabled]):
        (-[WebView isAutomaticLinkDetectionEnabled]):
        (-[WebView isAutomaticDashSubstitutionEnabled]):
        (-[WebView isAutomaticTextReplacementEnabled]):
        (-[WebView isAutomaticSpellingCorrectionEnabled]):
        (-[WebView setAutomaticQuoteSubstitutionEnabled:]):
        (-[WebView toggleAutomaticQuoteSubstitution:]):
        (-[WebView setAutomaticLinkDetectionEnabled:]):
        (-[WebView toggleAutomaticLinkDetection:]):
        (-[WebView setAutomaticDashSubstitutionEnabled:]):
        (-[WebView toggleAutomaticDashSubstitution:]):
        (-[WebView setAutomaticTextReplacementEnabled:]):
        (-[WebView toggleAutomaticTextReplacement:]):
        (-[WebView setAutomaticSpellingCorrectionEnabled:]):
        (-[WebView toggleAutomaticSpellingCorrection:]):
        * WebView/WebViewPrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42911 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Sam Weinig <sam@webkit.org>
weinig@apple.com [Mon, 27 Apr 2009 23:12:07 +0000 (23:12 +0000)]
2009-04-27  Sam Weinig  <sam@webkit.org>

        Reviewed by Geoffrey Garen.

        Fix for <rdar://problem/6772944> REGRESSION (r42223): PLT 2% slower, i-Bench 3% slower

        Be slightly less eager in collecting after page close/navigation by increasing
        the time before collection from 0 seconds to .5 seconds.

        3% speedup on the PLT.

        * bindings/js/GCController.cpp:
        (WebCore::GCController::garbageCollectSoon):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42910 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Consolidate runtime application checks for Apple Mail and Safari
ddkilzer@apple.com [Mon, 27 Apr 2009 23:06:46 +0000 (23:06 +0000)]
    Consolidate runtime application checks for Apple Mail and Safari

        Reviewed by Mark Rowe and Darin Adler.

WebCore:

        * WebCore.base.exp: Added exports.
        * WebCore.xcodeproj/project.pbxproj: Added
        RuntimeApplicationChecks.{h|mm} source files to the project.
        * page/mac/EventHandlerMac.mm:
        (WebCore::EventHandler::needsKeyboardEventDisambiguationQuirks):
        Removed local checkedSafari and isSafari variables and switched
        to use applicationIsSafari().
        * platform/mac/RuntimeApplicationChecks.h: Added.
        * platform/mac/RuntimeApplicationChecks.mm: Added.
        (WebCore::applicationIsAppleMail): Runtime application check for
        Apple Mail.
        (WebCore::applicationIsSafari): Runtime application check for
        Safari.

WebKit:

        * WebCoreSupport/WebContextMenuClient.mm:
        (isAppleMail): Removed.
        (fixMenusToSendToOldClients): Switched to use
        applicationIsAppleMail().
        * WebView/WebFrame.mm:
        (-[WebFrame reload]): Switched to use applicationIsSafari().
        * WebView/WebPDFView.mm:
        (-[WebPDFView menuForEvent:]): Ditto.
        * WebView/WebResource.mm:
        (+[WebResource _needMailThreadWorkaroundIfCalledOffMainThread]):
        Switched to use applicationIsAppleMail().
        * WebView/WebView.mm:
        (runningLeopardMail): Ditto.
        (runningTigerMail): Ditto.
        (-[WebView _needsKeyboardEventDisambiguationQuirks]): Switched
        to use applicationIsSafari().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42909 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoBug 25374: AX: No AXValue change sent when text is auto-inserted
cfleizach@apple.com [Mon, 27 Apr 2009 22:34:41 +0000 (22:34 +0000)]
Bug 25374: AX: No AXValue change sent when text is auto-inserted
https://bugs.webkit.org/show_bug.cgi?id=25374

Updated to fix regression test failures

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42908 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-16 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Mon, 27 Apr 2009 21:56:05 +0000 (21:56 +0000)]
2009-04-16  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        https://bugs.webkit.org/show_bug.cgi?id=21546
        [GTK] ATK accessibility enhancements

        Based on a patch by Alp Toker.

        Cover more WebCore role -> ATK role conversions.

        * page/gtk/AccessibilityObjectWrapperAtk.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42907 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-18 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Mon, 27 Apr 2009 21:40:31 +0000 (21:40 +0000)]
2009-04-18  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        https://bugs.webkit.org/show_bug.cgi?id=21546
        [GTK] ATK accessibility enhancements

        Do not call ATK setters in the getters, just return the
        value. Among other things calling the setter will emit the notify
        signal, which is wrong.

        * page/gtk/AccessibilityObjectWrapperAtk.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42906 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-16 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Mon, 27 Apr 2009 21:33:55 +0000 (21:33 +0000)]
2009-04-16  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        https://bugs.webkit.org/show_bug.cgi?id=21546
        [GTK] ATK accessibility enhancements

        Based on a patch by Alp Toker.

        Implement AtkImage interface.

        * page/gtk/AccessibilityObjectWrapperAtk.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42905 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Mon, 27 Apr 2009 21:25:02 +0000 (21:25 +0000)]
2009-04-27  Gustavo Noronha Silva  <gns@gnome.org>

        Unreviewed debug build fix.

        * platform/gtk/gtk2drawing.c:
        (moz_gtk_toggle_paint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42904 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Dimitri Glazkov <dglazkov@chromium.org>
dglazkov@chromium.org [Mon, 27 Apr 2009 20:12:04 +0000 (20:12 +0000)]
2009-04-27  Dimitri Glazkov  <dglazkov@chromium.org>

        Build fix.

        https://bugs.webkit.org/show_bug.cgi?id=25382
        Remove direct reference to CallFrame.h include to pacify Win build.

        * bindings/js/JSDOMBinding.h: Removed ScriptState.h include.
        * bindings/js/ScriptState.h: Replaced CallFrame.h with JSDOMBinding.h

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42903 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoBug 25374: AX: No AXValue change sent when text is auto-inserted
cfleizach@apple.com [Mon, 27 Apr 2009 19:49:23 +0000 (19:49 +0000)]
Bug 25374: AX: No AXValue change sent when text is auto-inserted
https://bugs.webkit.org/show_bug.cgi?id=25374

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42902 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Dimitri Glazkov <dglazkov@chromium.org>
dglazkov@chromium.org [Mon, 27 Apr 2009 19:46:00 +0000 (19:46 +0000)]
2009-04-27  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=25382
        Move scriptStateFromNode to ScriptState.

        * bindings/js/JSDOMBinding.cpp: Moved scriptStateFromNode to ScriptState.
        * bindings/js/JSDOMBinding.h: Ditto.
        * bindings/js/ScriptState.cpp:
        (WebCore::scriptStateFromNode): Added.
        * bindings/js/ScriptState.h: Ditto.
        * bindings/v8/ScriptState.cpp:
        (WebCore::scriptStateFromNode): Added.
        * bindings/v8/ScriptState.h: Ditto.
        * dom/NodeFilter.h: Removed JSDOMBinding header include.
        * dom/NodeIterator.h: Ditto.
        * dom/TreeWalker.h: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42901 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoBug 25428: 5 crashes in DumpRenderTree at com.apple.WebCore • -[AccessibilityObjectWr...
cfleizach@apple.com [Mon, 27 Apr 2009 19:44:36 +0000 (19:44 +0000)]
Bug 25428: 5 crashes in DumpRenderTree at com.apple.WebCore • -[AccessibilityObjectWrapper accessibilityAttributeValue:] + 830
https://bugs.webkit.org/show_bug.cgi?id=25428

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42900 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Mon, 27 Apr 2009 19:33:51 +0000 (19:33 +0000)]
2009-04-27  Anders Carlsson  <andersca@apple.com>

        Reviewed by Adam Roben.

        Update the example plug-in for new changes to the Cocoa event model.

        * NetscapeCocoaPlugin/main.m:
        (handleDraw):
        (handleMouseEvent):
        (NPP_HandleEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42899 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Fix the Tiger build.
kdecker@apple.com [Mon, 27 Apr 2009 18:54:33 +0000 (18:54 +0000)]
    Fix the Tiger build.

        * Plugins/WebNetscapePluginView.mm: The base class of WebNetscapePluginView.mm, WebBaseNetscapePluginView, already implemented
        the two methods below. But the Tiger compiler didn't know that.
        (-[WebNetscapePluginView webView]):
        (-[WebNetscapePluginView webFrame]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42898 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Anders Carlsson.
kdecker@apple.com [Mon, 27 Apr 2009 17:50:14 +0000 (17:50 +0000)]
    Reviewed by Anders Carlsson.

        <rdar://problem/6352982>

        * Plugins/WebBaseNetscapePluginView.mm: Removed checkIfAllowedToLoadURL:frame:callbackFunc, cancelCheckIfAllowedToLoadURL, and
        _webPluginContainerCancelCheckIfAllowedToLoadRequest from the base class. These methods now exist in the subclass WebNetscapePluginView.

        Added WebNetscapeContainerCheckContextInfo, which is used as a "contextInfo" object in
        -[WebNetscapePluginView checkIfAllowedToLoadURL:frame:callbackFunc:]
        * Plugins/WebNetscapeContainerCheckContextInfo.h: Added.
        * Plugins/WebNetscapeContainerCheckContextInfo.mm: Added.
        (-[WebNetscapeContainerCheckContextInfo initWithCheckRequestID:callbackFunc:]): Added desiginated initializer.
        (-[WebNetscapeContainerCheckContextInfo checkRequestID]): Added. Returns the checkRequestID.
        (-[WebNetscapeContainerCheckContextInfo callback]): Added. Returns the callback.

        * Plugins/WebNetscapePluginView.h: Added two new ivars: _containerChecksInProgress and _currentContainerCheckRequestID.
        * Plugins/WebNetscapePluginView.mm: #import both WebPluginContainerCheck.h and WebNetscapeContainerCheckContextInfo.h

        (-[WebNetscapePluginView checkIfAllowedToLoadURL:frame:callbackFunc:]): Added. This is the implementation of WKN_CheckIfAllowedToLoadURL.
        Here, we increment the request ID and start the container check.

        (-[WebNetscapePluginView _containerCheckResult:contextInfo:]): Added. This is a callback method for WebPluginContainerCheck.
        It's where we actually call into the plug-in and provide the allow-or-deny result.

        (-[WebNetscapePluginView cancelCheckIfAllowedToLoadURL:]): Added. This is the implementation of WKN_CancelCheckIfAllowedToLoadURL.
        Here we lookup the check, cancel it, and remove it from _containerChecksInProgress.

        (-[WebNetscapePluginView _webPluginContainerCancelCheckIfAllowedToLoadRequest:]): Added. WebPluginContainerCheck automatically calls
        this method after invoking our _containerCheckResult: selector. It works this way because calling -[WebPluginContainerCheck cancel] allows
        it to do it's teardown process.

        (-[WebNetscapePluginView fini]): Release _containerChecksInProgress ivar.

        * Plugins/WebPluginContainerCheck.h: Removed initWithRequest: method from header; no client was using this method directly.
        * Plugins/WebPluginContainerCheck.mm:
        (+[WebPluginContainerCheck checkWithRequest:target:resultObject:selector:controller:contextInfo:]): Added optional contextInfo parameter.
        (-[WebPluginContainerCheck _continueWithPolicy:]): If there's a contextInfo object, pass it as a parameter to resultSelector.
        (-[WebPluginContainerCheck cancel]): Release _contextInfo ivar.
        (-[WebPluginContainerCheck contextInfo]): Added new method. Returns the contextInfo object, if one so exists.
        * Plugins/WebPluginController.mm:
        (-[WebPluginController _webPluginContainerCheckIfAllowedToLoadRequest:inFrame:resultObject:selector:]): Pass nil for contextInfo. WebKit-style
        plug-ins do not need additional context information.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42897 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Darin Adler.
ap@webkit.org [Mon, 27 Apr 2009 16:57:13 +0000 (16:57 +0000)]
    Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=25399
        <rdar://problem/6633943> REGRESSION: Many crashes reported accessing Lexis/Nexis database,
        beneath WebCore::Cache::evict

        The crash happened because a cached resource handle was removed from a document's cached
        resources map twice recursively, so a destructor was called for a value in a deleted bucket.
        The first call was from Cache::evict, and when destroying CachedResourceHandle destroyed
        CachedResource, DocLoader::removeCachedResource() was called again, with HashMap being in
        an inconsistent state.

        I couldn't fully reconstruct the loading sequence to make a test.

        * loader/Cache.cpp:
        (WebCore::Cache::revalidateResource): Assert that the resource being revalidated is in cache
        (it makes no sense to revalidate one that isn't).
        (WebCore::Cache::evict): Don't remove the resource from document's map. Removing a resource
        from the cache in no way implies that documents no longer use the old version. This fixes the
        crash, and also fixes many cases of resource content being unavailable in Web Inspector.

        * loader/CachedResource.h:
        (WebCore::CachedResource::setInCache): When bringing a revalidated resource back to cache,
        reset m_isBeingRevalidated to maintain the invariant of resources being revalidated never
        being in cache. This fixes another assertion I saw on LexisNexis search: in rare cases,
        switchClientsToRevalidatedResource() results in the same resource being requested again,
        but we were only enforcing CachedResource invariants after calling this function.
        (WebCore::CachedResource::unregisterHandle): Assert that the counter doesn't underflow.

        * loader/DocLoader.cpp: (WebCore::DocLoader::removeCachedResource): Assert that the passed
        resource is removed, not some other resource that happens to have the same URL (this used to
        fail on LexisNexis search before this patch).

        * loader/ImageDocument.cpp: (WebCore::ImageTokenizer::write): Replaced ASSERT_NOT_REACHED
        with notImplemented(). This method can be legally called via document.write(), and should
        work. LexisNexis takes this code path, but apparently has a fallback for Safari, so it
        doesn't affect site functionality.

        * loader/CachedResource.cpp:
        (WebCore::CachedResource::clearResourceToRevalidate): Don't assert that m_resourceToRevalidate
        is being revalidated - this may no longer be true, because we now reset this member in
        CachedResource::setInCache().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42896 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Sam Weinig.
mitz@apple.com [Mon, 27 Apr 2009 16:24:55 +0000 (16:24 +0000)]
    Reviewed by Sam Weinig.

        - make BidiContext a RefCounted class, starting with a refcount of 1,
          and share the root BidiContexts.

        * platform/graphics/GraphicsContext.cpp:
        (WebCore::GraphicsContext::drawBidiText): Use BidiContext::create().
        * platform/text/BidiContext.cpp:
        (WebCore::BidiContext::create): Added. For the four "root" contexts,
        returns a reference to a shared static BidiContext.
        * platform/text/BidiContext.h:
        (WebCore::BidiContext::BidiContext):
        * platform/text/BidiResolver.h:
        (WebCore::::commitExplicitEmbedding): Use BidiContext::create().
        * rendering/bidi.cpp:
        (WebCore::RenderBlock::determineStartPosition): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42895 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Mon, 27 Apr 2009 15:23:54 +0000 (15:23 +0000)]
2009-04-27  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Rubber-stamped by Gustavo Noronha.

        Build fix for r42893: change GDK_exclaim to GDK_exclam

        * platform/chromium/KeyCodeConversionGtk.cpp:
        (WebCore::windowsKeyCodeForKeyEvent):
        * platform/gtk/KeyEventGtk.cpp:
        (WebCore::windowsKeyCodeForKeyEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42894 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Fumitoshi Ukai <ukai@google.com>
jmalonzo@webkit.org [Mon, 27 Apr 2009 15:08:48 +0000 (15:08 +0000)]
2009-04-27  Fumitoshi Ukai  <ukai@google.com>

        Reviewed by Xan Lopez.

        Fix wrong key code conversion.

        Missing conversion for GDK_exlaim ('!').
        Parens are opposite: GDK_parenleft is open paren '('
        and GDK_parenright is close paren ')'.
        https://bugs.webkit.org/show_bug.cgi?id=25367

        * platform/chromium/KeyCodeConversionGtk.cpp:
        (WebCore::windowsKeyCodeForKeyEvent):
        * platform/gtk/KeyEventGtk.cpp:
        (WebCore::windowsKeyCodeForKeyEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42893 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Mon, 27 Apr 2009 15:08:33 +0000 (15:08 +0000)]
2009-04-27  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Reviewed by Gustavo Noronha.

        [Gtk] gtkdrawing update
        https://bugs.webkit.org/show_bug.cgi?id=25408

        Update to the 25 Apr 2009 version of gtk2drawing.c and
        gtkdrawing.h from Mozilla (http://hg.mozilla.org/mozilla-central)

        * platform/gtk/gtk2drawing.c:
        (setup_widget_prototype):
        (moz_gtk_get_combo_box_entry_arrow):
        (ensure_toolbar_widget):
        (ensure_tree_header_cell_widget):
        (moz_gtk_init):
        (moz_gtk_button_get_inner_border):
        (moz_gtk_toggle_paint):
        (moz_gtk_caret_paint):
        (moz_gtk_entry_paint):
        (moz_gtk_tree_header_cell_paint):
        (moz_gtk_combo_box_paint):
        (moz_gtk_toolbar_separator_paint):
        (moz_gtk_menu_separator_paint):
        (moz_gtk_get_widget_border):
        (moz_gtk_get_tab_scroll_arrow_size):
        (moz_gtk_get_toolbar_separator_width):
        (moz_gtk_get_menu_separator_height):
        (moz_gtk_widget_paint):
        (moz_gtk_shutdown):
        * platform/gtk/gtkdrawing.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42892 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Mon, 27 Apr 2009 15:08:18 +0000 (15:08 +0000)]
2009-04-27  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Reviewed by Gustavo Noronha.

        [GTK] Make layout tests / jsc tests usable without build-webkit
        https://bugs.webkit.org/show_bug.cgi?id=21784

        This requires WEBKITOUTPUTDIR to be set to the build directory.
        Set the location of the product dir to the location pointed to by
        WEBKITOUTPUTDIR.

        * Scripts/webkitdirs.pm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42891 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Mon, 27 Apr 2009 14:54:03 +0000 (14:54 +0000)]
2009-04-27  Gustavo Noronha Silva  <gns@gnome.org>

        Reviewed by Adam Roben.

        Use libsoup's facilities to resolve relative URLs, instead, since
        it is a more robust method.

        * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
        (LayoutTestController::queueLoad):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42890 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Mon, 27 Apr 2009 14:53:41 +0000 (14:53 +0000)]
2009-04-27  Gustavo Noronha Silva  <gns@gnome.org>

        Reviewed by Xan Lopez.

        https://bugs.webkit.org/show_bug.cgi?id=25409
        [GTK] incorrectly fallsback on cancellation

        Check whether we should fallback also on didFailLoad, so that we
        don't get error pages showing up in parts of the content of some
        sites because of, for instance, location changes.

        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        (WebKit::FrameLoaderClient::dispatchDidFailLoad):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42889 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Mon, 27 Apr 2009 14:53:17 +0000 (14:53 +0000)]
2009-04-27  Gustavo Noronha Silva  <gns@gnome.org>

        Reviewed by Xan Lopez.

        https://bugs.webkit.org/show_bug.cgi?id=25409
        [GTK] incorrectly fallsback on cancellation

        Correctly handle cancellation and frame loading interruption
        caused by policy change.  This also fixes
        fast/frames/location-change.html.

        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        (WebKit::FrameLoaderClient::shouldFallBack):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42888 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-27 Ariya Hidayat <ariya.hidayat@nokia.com>
ariya@webkit.org [Mon, 27 Apr 2009 13:55:35 +0000 (13:55 +0000)]
2009-04-27  Ariya Hidayat  <ariya.hidayat@nokia.com>

        Reviewed by Tor Arne Vestbø.

        https://bugs.webkit.org/show_bug.cgi?id=18475

        [Qt] Widget painting should follow the layout direction (LTR, RTL)
        of the element style, not the application layout direction.

        * platform/qt/RenderThemeQt.cpp:
        (WebCore::RenderThemeQt::applyTheme):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42887 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Windows Debug build fix
aroben@apple.com [Mon, 27 Apr 2009 13:27:56 +0000 (13:27 +0000)]
    Windows Debug build fix

        Not sure why the buildbots weren't affected by this problem.

        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj: Let VS
        re-order the file list, and added JavaScriptCore[_debug].def to the
        project. This was not necessary for the fix, but made making the fix
        easier.

        * JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore_debug.def:
        Removed a function that no longer exists.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42886 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-26 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Mon, 27 Apr 2009 05:58:33 +0000 (05:58 +0000)]
2009-04-26  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Mark Rowe.

        Plug memory leak.

        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        (WebKit::FrameLoaderClient::dispatchDidFailLoad):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42885 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-26 Gavin Barraclough <barraclough@apple.com>
barraclough@apple.com [Mon, 27 Apr 2009 05:17:10 +0000 (05:17 +0000)]
2009-04-26  Gavin Barraclough  <barraclough@apple.com>

        Reviewed by Weinig Sam.

        Fix for https://bugs.webkit.org/show_bug.cgi?id=25416
        "Cached prototype accesses unsafely hoist property storage load above structure checks."

        Do not hoist the load of the pointer to the property storage array.

        No performance impact.

        * jit/JITPropertyAccess.cpp:
        (JSC::JIT::privateCompileGetByIdProto):
        (JSC::JIT::privateCompileGetByIdProtoList):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42884 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-26 Gavin Barraclough <barraclough@apple.com>
barraclough@apple.com [Mon, 27 Apr 2009 04:32:07 +0000 (04:32 +0000)]
2009-04-26  Gavin Barraclough  <barraclough@apple.com>

        Reviewed by Geoffrey "Gaffe or energy?" Garen.

        Randomize address requested by ExecutableAllocatorFixedVMPool.

        * jit/ExecutableAllocatorFixedVMPool.cpp:
        (JSC::FixedVMPoolAllocator::FixedVMPoolAllocator):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42883 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
weinig@apple.com [Mon, 27 Apr 2009 00:03:06 +0000 (00:03 +0000)]
WebCore:

2009-04-26  Sam Weinig  <sam@webkit.org>

        Reviewed by Dan Bernstein.

        Fix for https://bugs.webkit.org/show_bug.cgi?id=25412
        The Window object should be an EventTarget

        Test: fast/dom/Window/dispatchEvent.html

        * bindings/js/JSEventTarget.cpp:
        (WebCore::toJS): Add DOMWindow case.
        (WebCore::toEventTarget): Ditto.

        * dom/EventTarget.cpp:
        (WebCore::EventTarget::toDOMWindow): Added.
        * dom/EventTarget.h:

        * page/DOMWindow.cpp:
        (WebCore::DOMWindow::scriptExecutionContext): Added. Returns the document.
        (WebCore::DOMWindow::dispatchEvent): Added.
        * page/DOMWindow.h:
        (WebCore::DOMWindow::toDOMWindow):
        (WebCore::DOMWindow::refEventTarget):
        (WebCore::DOMWindow::derefEventTarget):
        * page/DOMWindow.idl:

LayoutTests:

2009-04-26  Sam Weinig  <sam@webkit.org>

        Reviewed by Dan Bernstein.

        Tests for https://bugs.webkit.org/show_bug.cgi?id=25412
        The Window object should be an EventTarget

        * fast/dom/Window/dispatchEvent-expected.txt: Added.
        * fast/dom/Window/dispatchEvent.html: Added.
        * fast/dom/Window/resources/dispatchEvent.js: Added.
        * fast/dom/Window/window-properties-expected.txt
        * platform/mac-leopard/fast/dom/Window/window-properties-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42882 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoJavaScriptCore:
weinig@apple.com [Sun, 26 Apr 2009 21:09:29 +0000 (21:09 +0000)]
JavaScriptCore:

2009-04-26  Sam Weinig  <sam@webkit.org>

        Reviewed by Eric Seidel.

        Remove scons-based build system.

        * JavaScriptCore.scons: Removed.

WebCore:

2009-04-26  Sam Weinig  <sam@webkit.org>

        Reviewed by Eric Seidel.

        Remove scons-based build system.

        * WebCore.scons: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42881 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-26 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Sun, 26 Apr 2009 18:11:06 +0000 (18:11 +0000)]
2009-04-26  Gustavo Noronha Silva  <gns@gnome.org>

        Reviewed by Xan Lopez.

        Resolve relative URLs when queueing loads. This fixes
        post-goback-same-url.html timing out.

        * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
        (LayoutTestController::queueLoad):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42880 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-26 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Sun, 26 Apr 2009 17:23:19 +0000 (17:23 +0000)]
2009-04-26  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        https://bugs.webkit.org/show_bug.cgi?id=25391
        G_SIGNAL_ACTION flag in WebKitWebView signals

        Most of the signals in WebKitGTK+ have the G_SIGNAL_ACTION flag,
        but I don't think this makes sense. The flag is meant for those
        signals that can be emitted by the user to trigger some action in
        the view, as if it were a normal API, which is not the case for
        most of them.

        * webkit/webkitdownload.cpp:
        (_WebKitDownloadPrivate::webkit_download_class_init):
        * webkit/webkitwebframe.cpp:
        * webkit/webkitwebinspector.cpp:
        (_WebKitWebInspectorPrivate::webkit_web_inspector_class_init):
        * webkit/webkitwebview.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42879 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-26 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Sun, 26 Apr 2009 17:11:24 +0000 (17:11 +0000)]
2009-04-26  Xan Lopez  <xlopez@igalia.com>

        Unreviewed, actually apply the right patch.

        * webkit/webkitwebview.cpp:
        (webkit_web_view_move_cursor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42878 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-26 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Sun, 26 Apr 2009 17:05:07 +0000 (17:05 +0000)]
2009-04-26  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        Pass a variable to hold the return value of the 'move-cursor'
        signal, otherwise a crash can happen.

        * webkit/webkitwebview.cpp:
        (webkit_web_view_move_cursor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42877 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-26 Darin Adler <darin@apple.com>
darin@apple.com [Sun, 26 Apr 2009 14:25:16 +0000 (14:25 +0000)]
2009-04-26  Darin Adler  <darin@apple.com>

        Suggested by Darin Fisher.

        Improve the fix for bug 25355: Crash when Creating New Tab or New Window when set to open Same Page
        https://bugs.webkit.org/show_bug.cgi?id=25355
        rdar://problem/6823543

        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::loadItem): Allow shouldScroll to be true even if m_currentHistoryItem is 0.
        Not sure if when this case can really arise in practice, but it's good to match the original
        logic more closely.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42876 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Dirk Schulze <krit@webkit.org>
krit@webkit.org [Sun, 26 Apr 2009 07:01:59 +0000 (07:01 +0000)]
2009-04-25  Dirk Schulze  <krit@webkit.org>

        Reviewed by Oliver Hunt.

        A width or height of zero for the destination or source rect of an
        image causes a not invertible pattern matrix.
        This crahes newer versions of Cairo and give some graphic gliches in
        Canvas.
        With this patch we check if there is something to draw and return if not.

        * platform/graphics/cairo/ImageCairo.cpp:
        (WebCore::BitmapImage::draw):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42875 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Sun, 26 Apr 2009 04:35:43 +0000 (04:35 +0000)]
2009-04-25  Gustavo Noronha Silva  <gns@gnome.org>

        Go back to disabling accessibility tests as a whole, since we have
        no real way of checking them, and no accessibilityController
        implementation.

        * platform/gtk/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42874 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Sun, 26 Apr 2009 04:25:54 +0000 (04:25 +0000)]
2009-04-25  Gustavo Noronha Silva  <gns@gnome.org>

        Disable tests that need eventSender, not yet implemented in our
        DumpRenderTree.

        * platform/gtk/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42873 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Sun, 26 Apr 2009 04:11:49 +0000 (04:11 +0000)]
2009-04-24  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Darin Adler

        https://bugs.webkit.org/show_bug.cgi?id=25390

        Fix point mapping and hit testing through transforms and perspective
        with ENABLE_3D_RENDERING. Previously the code did the more efficient
        move(), rather than getTransformFromContainer(), when the object had
        no transform. However, this skipped the application of perspective
        when the immediate child of the perspective element had no transform
        itself.

        Test: transforms/3d/point-mapping/3d-point-mapping-3.html

        * rendering/RenderBox.cpp:
        (WebCore::RenderBox::mapLocalToContainer):
        (WebCore::RenderBox::mapAbsoluteToLocalPoint):
        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::createLocalTransformState):
        Call shouldUseTransformFromContainer() to determine if we need to
        go through the more expensive getTransformFromContainer() code
        path.

        * rendering/RenderObject.h:
        * rendering/RenderObject.cpp:
        (WebCore::RenderObject::shouldUseTransformFromContainer):
        New method that indicates whether we need use getTransformFromContainer()
        when mapping points through renderers, and hit testing.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42872 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Brady Eidson <beidson@apple.com>
beidson@apple.com [Sat, 25 Apr 2009 23:30:55 +0000 (23:30 +0000)]
2009-04-25  Brady Eidson  <beidson@apple.com>

        Reviewed by Sam Weinig

        Some *obvious* style cleanup in my last patch.

        * History/WebBackForwardList.mm:
        (bumperCarBackForwardHackNeeded):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42871 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Brady Eidson <beidson@apple.com>
beidson@apple.com [Sat, 25 Apr 2009 23:24:46 +0000 (23:24 +0000)]
2009-04-25  Brady Eidson  <beidson@apple.com>

        Reviewed by Oliver Hunt

        <rdar://problem/6817607> BumperCar 2.2 crashes going back (invalid WebHistoryItem)

        BumperCar was holding a pointer to a WebHistoryItem they never retain, then later
        tried to go to it.  In some cases it would be dealloc'ed first.
        When WebHistoryItems were pure Objective-C they probably got away with this more often.
        With the WebCore/Obj-C mixed WebHistoryItems it's more likely to crash.

        * History/WebBackForwardList.mm:
        (bumperCarBackForwardHackNeeded):
        (-[WebBackForwardList backListWithLimit:]):  If this is BumperCar, hang on to the
          NSArray of WebHistoryItems until the next time this method is called.
        (-[WebBackForwardList forwardListWithLimit:]):  Ditto.

        * Misc/WebKitVersionChecks.h: Added WEBKIT_FIRST_VERSION_WITHOUT_BUMPERCAR_BACK_FORWARD_QUIRK.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42870 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
adele@apple.com [Sat, 25 Apr 2009 19:00:43 +0000 (19:00 +0000)]
WebCore:

2009-04-25  Adele Peterson  <adele@apple.com>

        Reviewed by Oliver Hunt.

        Fix for <rdar://problem/6712771> REGRESSION(34681): Text is no longer underlined after delete
        https://bugs.webkit.org/show_bug.cgi?id=25396

        Test: editing/deleting/delete-br-013.html

        * editing/CompositeEditCommand.cpp: (WebCore::CompositeEditCommand::moveParagraphs):
         Only preserve an empty paragraph's style when moving paragraphs around if the selection is still
         in an empty paragraph after the move occurs.  This was causing the empty paragraph's style to overwrite
         the previous paragraph's style during a delete of an empty paragraph.

LayoutTests:

2009-04-25  Adele Peterson  <adele@apple.com>

        Reviewed by Oliver Hunt.

        Test for <rdar://problem/6712771> REGRESSION(34681): Text is no longer underlined after delete
        https://bugs.webkit.org/show_bug.cgi?id=25396

        * editing/deleting/delete-br-013.html: Added.
        * platform/mac/editing/deleting/delete-br-013-expected.checksum: Added.
        * platform/mac/editing/deleting/delete-br-013-expected.png: Added.
        * platform/mac/editing/deleting/delete-br-013-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42869 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Eric Carlson <eric.carlson@apple.com>
eric.carlson@apple.com [Sat, 25 Apr 2009 17:48:56 +0000 (17:48 +0000)]
2009-04-25  Eric Carlson  <eric.carlson@apple.com>

        Reviewed by Darin Adler.

        <rdar://problem/6808171> REGRESSION (3-4): Standalone media documents don't
        properly display non-linear media

        If a media engine claims it can support the MIME type, WebCore now creates a <video>
        element for document mode media files instead of an <embed> element. Because WebCore's
        QuickTime backed media players do not support every kind of media the QuickTime plug-in
        supports, and because it is not always possible to tell what type of media is in a file
        without opening and parsing it, some types of media that used to be displayed properly
        by a plug-in are no longer supported properly. To fix this, if the media engine
        sees that it can not completely support a media file it now informs the media
        document, which replaces the <video> element with an <embed>.

        r42301 landed support for OSX. This patch modifies those changes slighly and adds support
        for Windows.

        * loader/MediaDocument.cpp:
        (WebCore::MediaDocument::MediaDocument): Initialize m_replaceMediaElementTimer.
        (WebCore::MediaDocument::mediaElementSawUnsupportedTracks): Don't replace the <video>
        element immediately.
        (WebCore::MediaDocument::replaceMediaElementTimerFired): Renamed from replaceVideoWithEmbed.
        Set body margin-width and margin-height to 0 as that is what a PluginDocument uses.
        * loader/MediaDocument.h:

        * platform/graphics/mac/MediaPlayerPrivateQTKit.mm:
        (WebCore::MediaPlayerPrivate::loadStateChanged): Don't do anything if m_hasUnsupportedTracks.
        (WebCore::MediaPlayerPrivate::rateChanged): Ditto.
        (WebCore::MediaPlayerPrivate::timeChanged): Ditto.
        (WebCore::MediaPlayerPrivate::didEnd): Ditto.
        (WebCore::MediaPlayerPrivate::repaint): Ditto.
        (WebCore::MediaPlayerPrivate::paint): Ditto.
        (WebCore::MediaPlayerPrivate::sawUnsupportedTracks): Set m_hasUnsupportedTracks before callback.

        * platform/graphics/win/MediaPlayerPrivateQuickTimeWin.cpp:
        (WebCore::MediaPlayerPrivate::MediaPlayerPrivate): Initialize m_totalTrackCount and m_hasUnsupportedTracks.
        (WebCore::MediaPlayerPrivate::updateStates): Call sawUnsupportedTracks if the movie has unsupported
        media type(s) or if it fails completely.
        (WebCore::MediaPlayerPrivate::sawUnsupportedTracks): New, disable the movie object and tell the
        media player client we won't play this movie correctly.
        (WebCore::MediaPlayerPrivate::didEnd): Don't do anything if m_hasUnsupportedTracks.
        (WebCore::MediaPlayerPrivate::setSize): Ditto.
        (WebCore::MediaPlayerPrivate::setVisible): Ditto.
        (WebCore::MediaPlayerPrivate::paint): Ditto.
        (WebCore::MediaPlayerPrivate::movieEnded): Ditto.
        (WebCore::MediaPlayerPrivate::movieLoadStateChanged): Ditto.
        (WebCore::MediaPlayerPrivate::movieTimeChanged): Ditto.
        (WebCore::MediaPlayerPrivate::movieNewImageAvailable): Ditto.
        * platform/graphics/win/MediaPlayerPrivateQuickTimeWin.h:

        * platform/graphics/win/QTMovieWin.cpp:
        (QTMovieWinPrivate::task): Stop the task timer if we were disabled during the load state
        change callback.
        (QTMovieWinPrivate::drawingComplete): Don't do anything if disabled.
        (QTMovieWin::QTMovieWin): Initialize m_disabled.
        (QTMovieWin::disableUnsupportedTracks): Return total number of tracks.
        (QTMovieWin::setDisabled): New, set m_disabled flag.
        * platform/graphics/win/QTMovieWin.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42868 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>
kov@webkit.org [Sat, 25 Apr 2009 14:02:26 +0000 (14:02 +0000)]
2009-04-25  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Reviewed by Xan Lopez.

        https://bugs.webkit.org/show_bug.cgi?id=24786
        WebKitDownload sometimes suggests peculiar filenames

        Refactor the emission of the download-requested signal so that we
        have less code duplication.

        * WebCoreSupport/ContextMenuClientGtk.cpp:
        (WebKit::ContextMenuClient::downloadURL):
        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        (WebKit::FrameLoaderClient::startDownload):
        * webkit/webkitprivate.h:
        * webkit/webkitwebview.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42867 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Sat, 25 Apr 2009 09:19:42 +0000 (09:19 +0000)]
2009-04-25  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Reviewed by Xan Lopez.

        [GTK] Error reporting
        https://bugs.webkit.org/show_bug.cgi?id=18344

        Fix the SOUP resource handle to report SOUP_HTTP_ERROR for Soup
        errors and G_IO_ERROR for gio errors.

        * platform/network/soup/ResourceHandleSoup.cpp:
        (WebCore::finishedCallback):
        (WebCore::ResourceHandle::startHttp):
        (WebCore::ResourceHandle::start):
        (WebCore::readCallback):
        (WebCore::openCallback):
        (WebCore::queryInfoCallback):
        (WebCore::ResourceHandle::startGio):

2009-04-25  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Reviewed by Xan Lopez.

        [GTK] Error reporting
        https://bugs.webkit.org/show_bug.cgi?id=18344

        Update FrameLoad errors to use WebKitErrors.

        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        (WebKit::FrameLoaderClient::cancelledError):
        (WebKit::FrameLoaderClient::blockedError):
        (WebKit::FrameLoaderClient::cannotShowURLError):
        (WebKit::FrameLoaderClient::interruptForPolicyChangeError):
        (WebKit::FrameLoaderClient::cannotShowMIMETypeError):
        (WebKit::FrameLoaderClient::fileDoesNotExistError):
        (WebKit::FrameLoaderClient::pluginWillHandleLoadError):
        (WebKit::FrameLoaderClient::shouldFallBack):
        * webkit/webkiterror.h: Added.
        * webkit/webkiterror.cpp: Added.

2009-04-25  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Reviewed by Xan Lopez.

        [GTK] Error reporting
        https://bugs.webkit.org/show_bug.cgi?id=18344

        Add webkiterror to the build.

        * GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42866 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Sat, 25 Apr 2009 09:19:20 +0000 (09:19 +0000)]
2009-04-25  Jan Michael Alonzo  <jmalonzo@webkit.org>

         Reviewed by Xan Lopez.

         [GTK] Error reporting
         https://bugs.webkit.org/show_bug.cgi?id=18344

         Add the default error page for installation.

         * GNUmakefile.am:

2009-04-25  Jan Michael Alonzo  <jmalonzo@webkit.org>

         Reviewed by Xan Lopez.

         Display a default error page for load errors.

         * WebCoreSupport/FrameLoaderClientGtk.cpp:
         (WebKit::FrameLoaderClient::dispatchDidFailLoad):
         * resources/error.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42865 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Sat, 25 Apr 2009 09:19:03 +0000 (09:19 +0000)]
2009-04-25  Jan Michael Alonzo  <jmalonzo@webkit.org>

         Reviewed by Xan Lopez.

         [GTK] Error reporting
         https://bugs.webkit.org/show_bug.cgi?id=18344

         Add a WebKitWebFrame API to load alternate content for unreachable URLs.
         Also add a new signal "load-error" for handling load errors.

         * WebCoreSupport/FrameLoaderClientGtk.cpp:
         (WebKit::FrameLoaderClient::postProgressFinishedNotification):
         (WebKit::FrameLoaderClient::dispatchDidFailLoading):
         (WebKit::FrameLoaderClient::dispatchDidFailProvisionalLoad):
         (WebKit::FrameLoaderClient::dispatchDidFailLoad):
         * webkit/webkitwebframe.cpp:
         * webkit/webkitwebframe.h:
         * webkit/webkitwebview.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42864 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoBuild fix.
oliver@apple.com [Sat, 25 Apr 2009 09:04:23 +0000 (09:04 +0000)]
Build fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42863 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Sat, 25 Apr 2009 08:44:47 +0000 (08:44 +0000)]
2009-04-25  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Reviewed by Oliver Hunt.

        Gtk build fix - check if we have MADV_FREE before using it.

        * interpreter/RegisterFile.cpp:
        (JSC::RegisterFile::releaseExcessCapacity):
        * wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42862 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-25 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>
kov@webkit.org [Sat, 25 Apr 2009 04:03:06 +0000 (04:03 +0000)]
2009-04-25  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Reviewed by Mark Rowe.

        Only set the new URL once for the request. Doing it a second time
        after the call to willSendRequest was causing crashes when
        redirected requests got cancelled.

        * platform/network/soup/ResourceHandleSoup.cpp:
        (WebCore::restartedCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42861 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agowx build fix. Switching JSCore from a static lib to a dynamic lib to match the Apple...
kevino@webkit.org [Sat, 25 Apr 2009 03:44:38 +0000 (03:44 +0000)]
wx build fix. Switching JSCore from a static lib to a dynamic lib to match the Apple build and fix symbol exports.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42860 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agowx build fix. Adding ScriptEventListener.cpp.
kevino@webkit.org [Sat, 25 Apr 2009 03:19:15 +0000 (03:19 +0000)]
wx build fix. Adding ScriptEventListener.cpp.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42859 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Laszlo Gombos <laszlo.1.gombos@nokia.com>
jmalonzo@webkit.org [Sat, 25 Apr 2009 00:27:26 +0000 (00:27 +0000)]
2009-04-24  Laszlo Gombos  <laszlo.1.gombos@nokia.com>

        Rubber-stamped by Mark Rowe.

        https://bugs.webkit.org/show_bug.cgi?id=25337
        Move ThreadingQt.cpp under the qt directory.

        * JavaScriptCore.pri:
        * wtf/ThreadingQt.cpp: Removed.
        * wtf/qt/ThreadingQt.cpp: Copied from JavaScriptCore/wtf/ThreadingQt.cpp.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42858 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Laszlo Gombos <laszlo.1.gombos@nokia.com>
jmalonzo@webkit.org [Sat, 25 Apr 2009 00:18:37 +0000 (00:18 +0000)]
2009-04-24  Laszlo Gombos  <laszlo.1.gombos@nokia.com>

        Rubber-stamped by Mark Rowe.

        https://bugs.webkit.org/show_bug.cgi?id=25338
        Move ThreadingGtk.cpp under the gtk directory.

        * GNUmakefile.am:
        * wtf/ThreadingGtk.cpp: Removed.
        * wtf/gtk/ThreadingGtk.cpp: Copied from JavaScriptCore/wtf/ThreadingGtk.cpp.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42857 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Sat, 25 Apr 2009 00:05:38 +0000 (00:05 +0000)]
2009-04-24  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Rubber-stamped by Gustavo Noronha.

        Fix braces to comply with style guidelines.

        * WebCoreSupport/PasteboardHelperGtk.cpp:
        (WebKit::PasteboardHelperGtk::getClipboard):
        (WebKit::PasteboardHelperGtk::getCopyTargetList):
        (WebKit::PasteboardHelperGtk::getPasteTargetList):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42856 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Sergio García-Cuevas <sergio_gcg@telefonica.net>
jmalonzo@webkit.org [Sat, 25 Apr 2009 00:05:21 +0000 (00:05 +0000)]
2009-04-24  Sergio García-Cuevas  <sergio_gcg@telefonica.net>

        Reviewed by Gustavo Noronha.

        https://bugs.webkit.org/show_bug.cgi?id=17267
        [GTK] Primary selection/clipboard support

        Copy the link location to the primary selection as well as the
        clipboard selection when using the "copy link selection" context
        menu entry.

        * platform/gtk/PasteboardGtk.cpp:
        (WebCore::Pasteboard::writeURL):
        * platform/gtk/PasteboardHelper.h:

2009-04-24  Sergio García-Cuevas  <sergio_gcg@telefonica.net>

        Reviewed by Gustavo Noronha.

        https://bugs.webkit.org/show_bug.cgi?id=17267
        [GTK] Primary selection/clipboard support

        Add a method for getting the primary selection.

        * WebCoreSupport/PasteboardHelperGtk.cpp:
        (WebKit::PasteboardHelperGtk::getPrimary):
        * WebCoreSupport/PasteboardHelperGtk.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42855 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Simon Fraser.
eric@webkit.org [Fri, 24 Apr 2009 23:54:33 +0000 (23:54 +0000)]
    Reviewed by Simon Fraser.

        https://bugs.webkit.org/show_bug.cgi?id=25383
        Changing SVG size via JS does not update CSS width/height

        Add an ugly hack to fix CSS width/height updates from JS.

        We can't easily call setAttribute when width/height changes
        as that would cause infinite recursion (as we'd try to set
        the animation storage from the attributes)

        For now, when we get notification that JS changed width/height
        we copy what parseMappedAttribute does and update the CSS values as well.

        A better fix for this would be:
        https://bugs.webkit.org/show_bug.cgi?id=25383

        * svg/SVGSVGElement.cpp:
        (WebCore::updateCSSForAttribute):
        (WebCore::SVGSVGElement::svgAttributeChanged):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42854 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Gavin Barraclough <barraclough@apple.com>
barraclough@apple.com [Fri, 24 Apr 2009 23:51:47 +0000 (23:51 +0000)]
2009-04-24  Gavin Barraclough  <barraclough@apple.com>

        Reviewed by Sam "Wesley" Weinig.

        Improve performance to YARR interpreter.
        (From about 3x slower than PCRE on regex-dna to about 30% slower).

        * yarr/RegexCompiler.cpp:
        (JSC::Yarr::RegexPatternConstructor::setupAlternativeOffsets):
        * yarr/RegexInterpreter.cpp:
        (JSC::Yarr::Interpreter::checkCharacter):
        (JSC::Yarr::Interpreter::checkCasedCharacter):
        (JSC::Yarr::Interpreter::backtrackPatternCharacter):
        (JSC::Yarr::Interpreter::backtrackPatternCasedCharacter):
        (JSC::Yarr::Interpreter::matchParentheticalAssertionBegin):
        (JSC::Yarr::Interpreter::matchParentheticalAssertionEnd):
        (JSC::Yarr::Interpreter::backtrackParentheticalAssertionBegin):
        (JSC::Yarr::Interpreter::backtrackParentheticalAssertionEnd):
        (JSC::Yarr::Interpreter::matchDisjunction):
        (JSC::Yarr::Interpreter::interpret):
        (JSC::Yarr::ByteCompiler::atomPatternCharacter):
        (JSC::Yarr::ByteCompiler::atomParenthesesSubpatternBegin):
        (JSC::Yarr::ByteCompiler::atomParentheticalAssertionBegin):
        (JSC::Yarr::ByteCompiler::closeAlternative):
        (JSC::Yarr::ByteCompiler::closeBodyAlternative):
        (JSC::Yarr::ByteCompiler::atomParenthesesEnd):
        (JSC::Yarr::ByteCompiler::regexBegin):
        (JSC::Yarr::ByteCompiler::regexEnd):
        (JSC::Yarr::ByteCompiler::alterantiveBodyDisjunction):
        (JSC::Yarr::ByteCompiler::alterantiveDisjunction):
        (JSC::Yarr::ByteCompiler::emitDisjunction):
        * yarr/RegexInterpreter.h:
        (JSC::Yarr::ByteTerm::):
        (JSC::Yarr::ByteTerm::ByteTerm):
        (JSC::Yarr::ByteTerm::BodyAlternativeBegin):
        (JSC::Yarr::ByteTerm::BodyAlternativeDisjunction):
        (JSC::Yarr::ByteTerm::BodyAlternativeEnd):
        (JSC::Yarr::ByteTerm::AlternativeBegin):
        (JSC::Yarr::ByteTerm::AlternativeDisjunction):
        (JSC::Yarr::ByteTerm::AlternativeEnd):
        (JSC::Yarr::ByteTerm::SubpatternBegin):
        (JSC::Yarr::ByteTerm::SubpatternEnd):
        * yarr/RegexJIT.cpp:
        (JSC::Yarr::RegexGenerator::generateParentheticalAssertion):
        * yarr/RegexPattern.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42853 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Kevin McCullough <kmccullough@apple.com>
kmccullough@apple.com [Fri, 24 Apr 2009 23:41:05 +0000 (23:41 +0000)]
2009-04-24  Kevin McCullough  <kmccullough@apple.com>

        Reviewed by Dan Bernstein.

        We should not show the deletion UI for blockquotes in mail but they do
        show now that we only check for one visible border.

        * editing/DeleteButtonController.cpp:
        (WebCore::isDeletableElement):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42852 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Rob Raguet-Schofield <ragfield@gmail.com>
mrowe@apple.com [Fri, 24 Apr 2009 23:28:45 +0000 (23:28 +0000)]
2009-04-24  Rob Raguet-Schofield  <ragfield@gmail.com>

        Rubber-stamped by Mark Rowe.

        * wtf/CurrentTime.h: Fix a typo in a comment.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42851 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago - address Hyatt's review comments on the last patch
mitz@apple.com [Fri, 24 Apr 2009 23:05:21 +0000 (23:05 +0000)]
    - address Hyatt's review comments on the last patch

        * rendering/RenderLayer.cpp:
        (WebCore::performOverlapTests):
        (WebCore::RenderLayer::paintLayer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42850 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Dave Hyatt.
mitz@apple.com [Fri, 24 Apr 2009 23:00:05 +0000 (23:00 +0000)]
    Reviewed by Dave Hyatt.

        - fix https://bugs.webkit.org/show_bug.cgi?id=5909
          <rdar://problem/5863349> overlapping element leaves trail when
          scrolling iframe

        * WebCore.xcodeproj/project.pbxproj: Added OverlapTestRequestClient.h.

        * page/FrameView.cpp:
        (WebCore::FrameView::reset): Initialize m_isOverlapped.
        (WebCore::FrameView::useSlowRepaints): Also check m_isOverlapped.
        (WebCore::FrameView::removeSlowRepaintObject): Use useSlowRepaints()
        so overlapping is taken into account.
        (WebCore::FrameView::setIsOverlapped): Added. Sets m_isOverlapped and
        enables or disables copy-on-scroll accordingly.

        * page/FrameView.h:
        Added setIsOverlapped() and m_isOverlapped.

        * rendering/OverlapTestRequestClient.h: Added. During foreground painting,
        an OverlapTestRequestClient can make a request to test if any subsequently-
        painted layers overlap a rect. The test result is delivered to the
        requestClient via setOverlapTestResult().
        (WebCore::OverlapTestRequestClient::~OverlapTestRequestClient):

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::paint): Pass an OverlapTestRequestMap to
        paintLayer(). When done painting, inform any remaining requestClients that
        they are not overlapped.

        (WebCore::RenderLayer::paintLayer): Take an OverlapTestRequestMap and
        pass it down to child and reflection layers. Inform requestClients if this
        layer overlaps their requested rects.

        * rendering/RenderLayer.h:

        * rendering/RenderObject.h:
        (WebCore::RenderObject::PaintInfo::PaintInfo): Added an
        OverlapTestRequestMap member.

        * rendering/RenderReplica.cpp:
        (WebCore::RenderReplica::paint): For now, pass a 0 OverlapTestRequestMap
        to paintLayer().

        * rendering/RenderWidget.cpp:
        (WebCore::RenderWidget::paint): For FrameViews, request an overlap test
        with the frame.
        (WebCore::RenderWidget::setOverlapTestResult): Call
        FrameView::setIsOverlapped() with the result of the overlap test.

        * rendering/RenderWidget.h: Made RenderWidget an OverlapTestRequestClient.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42849 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Pierre d'Herbemont <pdherbemont@apple.com>
simon.fraser@apple.com [Fri, 24 Apr 2009 22:52:45 +0000 (22:52 +0000)]
2009-04-24  Pierre d'Herbemont  <pdherbemont@apple.com>

        Reviewed by Simon Fraser

        https://bugs.webkit.org/show_bug.cgi?id=25380

        Fixed a race condition on the first 'played' attribute test start.

        Previously we were starting nearly right away to play, and probably
        before receiving the 'canplay' event. Now we ensure we'll wait for
        that 'canplay' event before attempting to play.

        * media/video-played-expected.txt: Updated
        * media/video-played.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42848 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Geoffrey Garen <ggaren@apple.com>
ggaren@apple.com [Fri, 24 Apr 2009 22:45:04 +0000 (22:45 +0000)]
2009-04-24  Geoffrey Garen  <ggaren@apple.com>

        Rubber stamped by Eric Seidel.

        Updated test to turn the buildbot green after Eric's last check-in.

        * platform/mac/svg/W3C-SVG-1.1/animate-elem-36-t-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42847 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 David Hyatt <hyatt@apple.com>
hyatt@apple.com [Fri, 24 Apr 2009 22:30:31 +0000 (22:30 +0000)]
2009-04-24  David Hyatt  <hyatt@apple.com>

        Reviewed by Simon Fraser, Dan Bernstein.

        Speed up the PLT by devirtualizing the height() function on InlineBox.  virtualizing it in order to make the height() computation dynamic caused
        a ~0.5% slowdown.  This patch does the following to get the speed back:

        (a) Devirtualizes isText and forces inline box creators to set the bit as needed.  This actually resulted in simplified code, since ListMarkerBox could
        then be removed.
        (b) Reduces the height() call count.  In some cases the code was repeatedly calling height(), which used to be fine when the function was just an inline
        member variable access.  The call sites have been patched to cut down on extra height() calls now that it is more expensive.
        (c) Devirtualize height() except on SVG boxes.  For all non-SVG, the height() function on InlineBox handles the computation.  For SVG boxes, a new bit has
        been set on InlineBoxes (isSVG()) that indicates that the virtual svgBoxHeight() function should be called to retrieve the height instead.

        * GNUmakefile.am:
        * WebCore.pro:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * WebCoreSources.bkl:
        * rendering/InlineBox.cpp:
        (WebCore::InlineBox::height):
        * rendering/InlineBox.h:
        (WebCore::InlineBox::InlineBox):
        (WebCore::InlineBox::isSVG):
        (WebCore::InlineBox::setIsSVG):
        (WebCore::InlineBox::isText):
        (WebCore::InlineBox::setIsText):
        (WebCore::InlineBox::svgBoxHeight):
        * rendering/InlineFlowBox.cpp:
        (WebCore::InlineFlowBox::placeBoxesVertically):
        (WebCore::InlineFlowBox::paintBoxDecorations):
        (WebCore::InlineFlowBox::paintMask):
        * rendering/InlineFlowBox.h:
        * rendering/InlineTextBox.cpp:
        * rendering/InlineTextBox.h:
        * rendering/ListMarkerBox.cpp: Removed.
        * rendering/ListMarkerBox.h: Removed.
        * rendering/RenderListMarker.cpp:
        (WebCore::RenderListMarker::createInlineBox):
        * rendering/RenderSVGInline.cpp:
        (WebCore::RenderSVGInline::createFlowBox):
        * rendering/RenderSVGInlineText.cpp:
        (WebCore::RenderSVGInlineText::createTextBox):
        * rendering/RenderSVGText.cpp:
        (WebCore::RenderSVGText::createRootBox):
        * rendering/RenderText.cpp:
        (WebCore::RenderText::createInlineTextBox):
        * rendering/RootInlineBox.cpp:
        * rendering/RootInlineBox.h:
        * rendering/SVGInlineFlowBox.h:
        (WebCore::SVGInlineFlowBox::svgBoxHeight):
        * rendering/SVGInlineTextBox.h:
        (WebCore::SVGInlineTextBox::svgBoxHeight):
        * rendering/SVGRootInlineBox.h:
        (WebCore::SVGRootInlineBox::svgBoxHeight):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42846 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Adam Roben.
eric@webkit.org [Fri, 24 Apr 2009 22:25:20 +0000 (22:25 +0000)]
    Reviewed by Adam Roben.

        Add PrettyPatch support to run-webkit-tests

        * Scripts/run-webkit-tests:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42845 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoBuild fix
oliver@apple.com [Fri, 24 Apr 2009 22:09:42 +0000 (22:09 +0000)]
Build fix

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42844 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Dimitri Glazkov <dglazkov@chromium.org>
dglazkov@chromium.org [Fri, 24 Apr 2009 22:00:44 +0000 (22:00 +0000)]
2009-04-24  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=25370
        Introduce ScriptEventListener to refactor away direct references to
        JSLazyEventListener in WebCore code.

        * GNUmakefile.am: Added ScriptEventListener to project.
        * WebCore.pro: Ditto.
        * WebCore.vcproj/WebCore.vcproj: Ditto.
        * WebCore.xcodeproj/project.pbxproj: Ditto.
        * bindings/js/JSDOMGlobalObject.cpp: Removed unused JSLazyEventListener include.
        * bindings/js/JSLazyEventListener.cpp: Moved createAttributeEventListener helpers
            to ScriptEventListener.
        * bindings/js/JSLazyEventListener.h: Ditto.
        * bindings/js/ScriptController.cpp: Removed unused JSLazyEventListener include.
        * bindings/js/ScriptEventListener.cpp: Added.
        * bindings/js/ScriptEventListener.h: Added.
        * bindings/v8/ScriptEventListener.cpp: Added.
        * bindings/v8/ScriptEventListener.h: Added.
        * dom/Document.cpp: Renamed JSLazyEventListener include to ScriptEventListener.
        * html/HTMLBodyElement.cpp: Ditto.
        * html/HTMLButtonElement.cpp: Ditto.
        * html/HTMLElement.cpp: Ditto.
        * html/HTMLFormElement.cpp: Ditto.
        * html/HTMLFrameElementBase.cpp: Ditto.
        * html/HTMLFrameSetElement.cpp: Ditto.
        * html/HTMLImageElement.cpp: Ditto.
        * html/HTMLInputElement.cpp: Ditto.
        * html/HTMLObjectElement.cpp: Ditto.
        * html/HTMLScriptElement.cpp: Ditto.
        * html/HTMLSelectElement.cpp: Ditto.
        * html/HTMLTextAreaElement.cpp: Ditto.
        * svg/SVGElement.cpp: Ditto.
        * svg/SVGSVGElement.cpp: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42843 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago<rdar://problem/6050421> JavaScript register file should remap to release physical...
oliver@apple.com [Fri, 24 Apr 2009 21:53:46 +0000 (21:53 +0000)]
<rdar://problem/6050421> JavaScript register file should remap to release physical pages accumulated during deep recursion

Reviewed by Geoff Garen

We now track the maximum extent of the RegisterFile, and when we reach the final
return from JS (so the stack portion of the registerfile becomes empty) we see
if that extent is greater than maxExcessCapacity.  If it is we use madvise or
VirtualFree to release the physical pages that were backing the excess.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42842 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Darin Adler <darin@apple.com>
darin@apple.com [Fri, 24 Apr 2009 21:52:16 +0000 (21:52 +0000)]
2009-04-24  Darin Adler  <darin@apple.com>

        Reviewed by Sam Weinig.

        Bug 25355: Crash when Creating New Tab or New Window when set to open Same Page
        https://bugs.webkit.org/show_bug.cgi?id=25355
        rdar://problem/6823543

        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::loadItem): Added a null check.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42841 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Simon Fraser.
eric@webkit.org [Fri, 24 Apr 2009 21:34:52 +0000 (21:34 +0000)]
    Reviewed by Simon Fraser.

        Fix RenderSVGImage to dump with full bounds and style information.
        https://bugs.webkit.org/show_bug.cgi?id=25377

        No functional changes, we're just now including more info in render tree dumps.

        * rendering/RenderObject.h:
        (WebCore::RenderObject::isSVGImage):
        * rendering/RenderSVGImage.h:
        (WebCore::RenderSVGImage::isSVGImage):
        * rendering/RenderTreeAsText.cpp:
        (WebCore::write):
        * rendering/SVGRenderTreeAsText.cpp:
        (WebCore::operator<<):
        (WebCore::write):
        * rendering/SVGRenderTreeAsText.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42840 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Fri, 24 Apr 2009 21:30:48 +0000 (21:30 +0000)]
2009-04-24  Anders Carlsson  <andersca@apple.com>

        Reviewed by Darin Adler.

        Fix <rdar://problem/6761635>.

        Make sure to keep an extra reference to the instance proxy in case the plug-in host crashes while
        we're waiting for a reply.

        * Plugins/Hosted/NetscapePluginInstanceProxy.mm:
        (WebKit::NetscapePluginInstanceProxy::destroy):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42839 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 David Levin <levin@chromium.org>
levin@chromium.org [Fri, 24 Apr 2009 21:18:24 +0000 (21:18 +0000)]
2009-04-24  David Levin  <levin@chromium.org>

        Reviewed by NOBODY (chromium build fix).

        Rename "isInline" => "isAttribute" in v8 bindings (jsc was done in r42699).

        * bindings/v8/V8AbstractEventListener.cpp:
        (WebCore::V8AbstractEventListener::V8AbstractEventListener):
        (WebCore::V8AbstractEventListener::invokeEventHandler):
        * bindings/v8/V8AbstractEventListener.h:
        * bindings/v8/V8EventListenerList.cpp:
        (WebCore::V8EventListenerList::add):
        (WebCore::V8EventListenerList::remove):
        (WebCore::V8EventListenerList::clear):
        * bindings/v8/V8LazyEventListener.h:
        (WebCore::V8LazyEventListener::virtualisAttribute):
        * bindings/v8/custom/V8CustomEventListener.cpp:
        (WebCore::V8EventListener::V8EventListener):
        * bindings/v8/custom/V8CustomEventListener.h:
        (WebCore::V8EventListener::create):
        (WebCore::V8EventListener::virtualisAttribute):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42838 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Johan Bilien <jobi@via.ecp.fr>
xan@webkit.org [Fri, 24 Apr 2009 21:16:45 +0000 (21:16 +0000)]
2009-04-24  Johan Bilien  <jobi@via.ecp.fr>

        Reviewed by Xan Lopez.

        https://bugs.webkit.org/show_bug.cgi?id=25373
        Popup menu has the wrong size when popped up near the edge
        of the screen

        * webkit/webkitwebview.cpp:
        (webkit_web_view_forward_context_menu_event): Let GTK place
        the popup menu instead of trying to do better and failing

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42837 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Pierre d'Herbemont <pdherbemont@apple.com>
simon.fraser@apple.com [Fri, 24 Apr 2009 20:48:08 +0000 (20:48 +0000)]
2009-04-24  Pierre d'Herbemont  <pdherbemont@apple.com>

        Reviewed by Simon Fraser

        <rdar://problem/6805928> Re-export layout test MPEG-4 file to remove unusual edit

        Use better mp4 test file for media tests. The original mp4 file had an unusual
        edit (non-zero start time) which is not supported by all media engines. The new
        movie should also give more useful pixel test results.

        * media/content/test.mp4: Update the movie.
        * media/progress-event-total-expected.txt: Update test results to match movie duration and
        content changes.
        * media/progress-event-total.html: Ditto.
        * media/video-currentTime-set-expected.txt: Ditto.
        * media/video-loop-expected.txt: Ditto.
        * platform/mac/media/video-aspect-ratio-expected.checksum: Ditto.
        * platform/mac/media/video-aspect-ratio-expected.png: Ditto.
        * platform/mac/media/video-controls-rendering-expected.checksum: Ditto.
        * platform/mac/media/video-controls-rendering-expected.png: Ditto.
        * platform/mac/media/video-display-toggle-expected.checksum: Ditto.
        * platform/mac/media/video-display-toggle-expected.png: Ditto.
        * platform/mac/media/video-empty-source-expected.checksum: Ditto.
        * platform/mac/media/video-layer-crash-expected.checksum: Ditto.
        * platform/mac/media/video-layer-crash-expected.png: Ditto.
        * platform/mac/media/video-transformed-expected.checksum: Ditto.
        * platform/mac/media/video-transformed-expected.png: Ditto.
        * platform/mac/media/video-zoom-expected.checksum: Ditto.
        * platform/mac/media/video-zoom-expected.png: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42836 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Fix nightly for Windows x64 users.
sfalken@apple.com [Fri, 24 Apr 2009 20:08:16 +0000 (20:08 +0000)]
    Fix nightly for Windows x64 users.
        Installed browser location is written to a slightly different registry key in that case.

        * FindSafari/FindSafari.cpp:
        (getInstalledWebKitDirectory):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42835 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Kevin McCullough <kmccullough@apple.com>
kmccullough@apple.com [Fri, 24 Apr 2009 19:13:52 +0000 (19:13 +0000)]
2009-04-24  Kevin McCullough  <kmccullough@apple.com>

        When I moved the files I forgot to update the location of the image!

        * platform/mac/editing/deleting/deletionUI-differing-background-expected.txt: Added.
        * platform/mac/editing/deleting/deletionUI-differing-background.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42834 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Nate Chapin <japhet@google.com>
dglazkov@chromium.org [Fri, 24 Apr 2009 18:43:47 +0000 (18:43 +0000)]
2009-04-24  Nate Chapin  <japhet@google.com>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=25343
        Fix Chromium/Skia bug where -webkit-box-shadow with 0,0 offset
        ignores blur.

        * WebCore\platform\graphics\skia\GraphicsContextSkia.cpp
         (WebCore::GraphicsContext::setPlatformShadow): Add check against
         blur before clearing looper.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42833 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Kevin McCullough <kmccullough@apple.com>
kmccullough@apple.com [Fri, 24 Apr 2009 18:41:20 +0000 (18:41 +0000)]
2009-04-24  Kevin McCullough  <kmccullough@apple.com>

        - Build fix.  Removed this test until I can figure out why it's failing.

        * platform/mac-leopard/editing/deleting/deletionUI-differing-background-expected.txt: Removed.
        * platform/mac/editing/deleting/deletionUI-differing-background-expected.txt: Removed.
        * platform/mac/editing/deleting/deletionUI-differing-background.html: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42832 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
dglazkov@chromium.org [Fri, 24 Apr 2009 18:40:28 +0000 (18:40 +0000)]
WebCore:

2009-04-24  Eric Roman  <eroman@chromium.org>

        Reviewed by Darin Adler.

        Initialize TextIterator::m_textLength to 0.

        This assures that TextIterator::length() will return 0 for cases when TextIterator's constructor returns early (because there is nothing to iterate over in the range).

        <https://bugs.webkit.org/show_bug.cgi?id=25335>

        Test: editing/selection/doubleclick-whitespace-img-crash.html

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

LayoutTests:

2009-04-24  Eric Roman  <eroman@chromium.org>

        Reviewed by Darin Adler.

        Add a test for <https://bugs.webkit.org/show_bug.cgi?id=25335>.

        To pass, this test must not access invalid memory when run (won't necessarily manifest as a crash for failures).

        * editing/selection/doubleclick-whitespace-img-crash-expected.txt: Added.
        * editing/selection/doubleclick-whitespace-img-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42831 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Fri, 24 Apr 2009 18:38:41 +0000 (18:38 +0000)]
2009-04-24  Simon Fraser  <simon.fraser@apple.com>

        Source changes reviewed by Darin Adler

        https://bugs.webkit.org/show_bug.cgi?id=22242

        Update WebKitSystemInterface for Mac with fixes for video controller drawing.

        * libWebKitSystemInterfaceLeopard.a:
        * libWebKitSystemInterfaceTiger.a:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42830 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Fumitoshi Ukai <ukai@google.com>
dglazkov@chromium.org [Fri, 24 Apr 2009 18:36:08 +0000 (18:36 +0000)]
2009-04-24  Fumitoshi Ukai  <ukai@google.com>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=25329
        Add indexed access method in v8 binding of ClientRectList.

        * bindings/v8/custom/V8ClientRectListCustom.cpp: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42829 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
dglazkov@chromium.org [Fri, 24 Apr 2009 18:32:48 +0000 (18:32 +0000)]
WebCore:

2009-04-24  Fumitoshi Ukai  <ukai@google.com>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=25257
        Use null for unspecified event listener attributes of XMLHttpRequest
        and XMLHttpRequestUpload in v8 binding.

        cf. http://www.whatwg.org/specs/web-apps/current-work/#event-handler-attributes
        An event handler attribute, unless otherwise specified, can either
        have the value null or be set to a Function object.
        Initially, an event handler attribute must be set to null.

        Test: fast/xmlhttprequest/xmlhttprequest-default-attributes.html

        * bindings/v8/custom/V8XMLHttpRequestCustom.cpp:
        (WebCore::ACCESSOR_GETTER): Changed to use v8::Null().
        * bindings/v8/custom/V8XMLHttpRequestUploadCustom.cpp:
        (WebCore::ACCESSOR_GETTER): Ditto.

LayoutTests:

2009-04-24  Fumitoshi Ukai  <ukai@google.com>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=25257
        Verify XMLHttpRequest has correct attribute value by default.

        * fast/xmlhttprequest/resources/xmlhttprequest-default-attributes.js: Added.
        * fast/xmlhttprequest/xmlhttprequest-default-attributes-expected.txt: Added.
        * fast/xmlhttprequest/xmlhttprequest-default-attributes.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42828 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Christian Dywan <christian@twotoasts.de>
christian@webkit.org [Fri, 24 Apr 2009 18:02:12 +0000 (18:02 +0000)]
2009-04-24  Christian Dywan  <christian@twotoasts.de>

        Rubber-stamped by Gustavo Noronha.

        * webkit/webkitwebsettings.cpp:
        (_WebKitWebSettingsPrivate::webkit_web_settings_class_init): Reword the
        description of the enable-spell-checking property and correct the
        mentioning of the default language used.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42827 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Kevin McCullough <kmccullough@apple.com>
kmccullough@apple.com [Fri, 24 Apr 2009 17:42:15 +0000 (17:42 +0000)]
2009-04-24  Kevin McCullough  <kmccullough@apple.com>

        Speculative fix for leopard build-bot until I have a fix on leopard.

        * platform/mac-leopard/editing: Added.
        * platform/mac-leopard/editing/deleting: Added.
        * platform/mac-leopard/editing/deleting/deletionUI-differing-background-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42826 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Darin Adler <darin@apple.com>
darin@apple.com [Fri, 24 Apr 2009 16:38:36 +0000 (16:38 +0000)]
2009-04-24  Darin Adler  <darin@apple.com>

        Reviewed by Adam Roben.

        <rdar://problem/6663836> CrashTracer: crashes in Safari at com.apple.WebCore • WebCore::reportException + 1513

        * bindings/js/JSDOMBinding.cpp:
        (WebCore::reportException): Added a null check.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42825 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-04-24 Brady Eidson <beidson@apple.com>
beidson@apple.com [Fri, 24 Apr 2009 16:13:15 +0000 (16:13 +0000)]
2009-04-24  Brady Eidson  <beidson@apple.com>

        Reviewed by Dan Bernstein

        Currently working on a bug where a WebHistoryItem was being used after being dealloc'ed.
        I added this assertion to help catch the case as soon as it happens instead of random
        issues downstream.  Figured it's worth checking in by itself.

        * History/WebHistoryItem.mm:
        (core): ASSERT that the WebCore::HistoryItem inside this WebHistoryItem is supposed to have
          this WebHistoryItem as a wrapper.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@42824 268f45cc-cd09-0410-ab3c-d52691b4dbfc