GitLab

        Reviewed by Darin Fisher.

        Fix Chromium (V8) build by renaming methods to match changes
        in http://trac.webkit.org/changeset/42377.

        https://bugs.webkit.org/show_bug.cgi?id=25141

        * bindings/v8/ScheduledAction.cpp:
     (WebCore::ScheduledAction::execute): updateRendering() -> updateStyleIfNeeded()
        * bindings/v8/V8AbstractEventListener.cpp:
        (WebCore::V8AbstractEventListener::handleEvent): updateDocumentsRendering() -> updateStyleForAllDocuments()



git-svn-id: svn://svn.chromium.org/blink/trunk@42444 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        * Scripts/make-js-test-wrappers: Added another exception.



git-svn-id: svn://svn.chromium.org/blink/trunk@42443 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=25136
        CRASH in DocumentLoader::removeSubresourceLoader due to null m_frame

        Test: fast/frames/frame-unload-crash.html

        * loader/DocumentLoader.cpp:
        (WebCore::DocumentLoader::updateLoading): Added null check of m_frame.



git-svn-id: svn://svn.chromium.org/blink/trunk@42442 bbb929c8-8fbe-4397-9dbb-9b2b20218538

2009-04-13  Sam Weinig  <sam@webkit.org>

        Reviewed by Tim Hatcher.

        Put the MessagePort constructor back in ENABLE_CHANNEL_MESSAGING.

        * page/DOMWindow.idl:

LayoutTests:

2009-04-13  Sam Weinig  <sam@webkit.org>

        Reviewed by Tim Hatcher.

        Update test results after putting the MessagePort constructor
        back in ENABLE_CHANNEL_MESSAGING.

        * fast/dom/Window/window-properties-expected.txt:
        * fast/js/global-constructors-expected.txt:
        * platform/mac-leopard/fast/dom/Window/window-properties-expected.txt:



git-svn-id: svn://svn.chromium.org/blink/trunk@42441 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Fix GTK build break.

        * platform/gtk/GeolocationServiceGtk.cpp:
        (WebCore::GeolocationServiceGtk::updatePosition):



git-svn-id: svn://svn.chromium.org/blink/trunk@42440 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Darin Adler and Dan Bernstein.

        Assertion in updateStyleForAllDocuments.  Make sure to unscheduleStyleRecalc when an object goes into
        the page cache.  Disallow scheduling of style recalcs on a document that is already in the page cache.
        Schedule a style recalc if needed when an object comes out of the page cache.

        * dom/Document.cpp:
        (WebCore::Document::scheduleStyleRecalc):
        (WebCore::Document::setInPageCache):



git-svn-id: svn://svn.chromium.org/blink/trunk@42439 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Ada Chan.

        * inspector/front-end/ScriptsPanel.js:
        (WebInspector.ScriptsPanel.prototype._addScriptToFilesMenu): Refactor the code
        and sort "(program)" options by the sourceID to maintain some order.
        * inspector/front-end/utilities.js:
        (indexOfObjectInListSortedByFunction): Fix a bug where identical objects would cause
        an incorrect return value that confuses insertionIndexForObjectInListSortedByFunction.
        Just set 'first' and break so the correct result will be returned.

git-svn-id: svn://svn.chromium.org/blink/trunk@42438 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=24770
        
        Add custom code for Coordinates that returns null when required by the
        Geolocation spec.
        
        http://dev.w3.org/geo/api/spec-source.html#coordinates

        * WebCore.xcodeproj/project.pbxproj:
        * bindings/js/JSCoordinatesCustom.cpp: Added.
        (WebCore::JSCoordinates::altitude): returns null if the value cannot be provided.
        (WebCore::JSCoordinates::altitudeAccuracy): Ditto
        (WebCore::JSCoordinates::heading): Ditto
        (WebCore::JSCoordinates::speed): Ditto
        * page/Coordinates.h:
        (WebCore::Coordinates::create): Add 'canProvide' parameters and values
        (WebCore::Coordinates::canProvideAltitude):
        (WebCore::Coordinates::canProvideAltitudeAccuracy):
        (WebCore::Coordinates::canProvideHeading):
        (WebCore::Coordinates::canProvideSpeed):
        (WebCore::Coordinates::Coordinates):
        * page/Coordinates.idl: specify custom getters
        * platform/mac/GeolocationServiceMac.mm: Set the values correctly if they cannot be provided.
        (-[WebCoreCoreLocationObserver locationManager:didUpdateToLocation:fromLocation:]):



git-svn-id: svn://svn.chromium.org/blink/trunk@42437 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Holger Freyther.

        https://bugs.webkit.org/show_bug.cgi?id=22898
        [GTK] need proper API for printing

        Added simple printing unit tests.

        * tests/testwebframe.c:
        (print_requested_cb):
        (print_timeout):
        (test_webkit_web_frame_printing):
        (main):

git-svn-id: svn://svn.chromium.org/blink/trunk@42436 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Holger Freyther.

        https://bugs.webkit.org/show_bug.cgi?id=22898
        [GTK] need proper API for printing

        Implement proper printing API, using the GTK+ printing API.

        * WebCoreSupport/ChromeClientGtk.cpp:
        (WebKit::ChromeClient::print):
        * webkit/webkitprivate.h:
        * webkit/webkitwebframe.cpp:
        * webkit/webkitwebframe.h:
        * webkit/webkitwebview.cpp:

git-svn-id: svn://svn.chromium.org/blink/trunk@42435 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Rubber-stamped by Holger Freyther.

        Reworked the skipped tests listing so that it is easier to find
        tests that are skipped but pass using diff. Enable all tests that
        are passing, and disable failing tests except for a few crashers
        and known regressions. This is an effort to make the build slave
        output more useful when it is back.

        * platform/gtk/Skipped:

git-svn-id: svn://svn.chromium.org/blink/trunk@42434 bbb929c8-8fbe-4397-9dbb-9b2b20218538

2009-04-13  Adele Peterson  <adele@apple.com>

        Reviewed by Darin Adler.

        Fix for <rdar://problem/5510537> Typing return at start of block quote yields empty quoted line

        Test: editing/inserting/5510537.html

        * editing/BreakBlockquoteCommand.cpp: (WebCore::BreakBlockquoteCommand::doApply): If the break insertion position
        is at the beginning of the topmost quote, then just insert the <br> before the blockquote.  Otherwise, if the break
        insertion position is at the beginning of any quote, adjust the position so the break comes before the current quote level
        so we don't end up with an empty line in that quote in addition to the new line we're adding with the <br>.

LayoutTests:

2009-04-13  Adele Peterson  <adele@apple.com>

        Reviewed by Darin Adler.

        Test for <rdar://problem/5510537> Typing return at start of block quote yields empty quoted line

        * editing/inserting/5510537.html: Added.
        * platform/mac/editing/inserting/5510537-expected.checksum: Added.
        * platform/mac/editing/inserting/5510537-expected.png: Added.
        * platform/mac/editing/inserting/5510537-expected.txt: Added.



git-svn-id: svn://svn.chromium.org/blink/trunk@42433 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Dan Bernstein.

        Fix problem when encoding history files with duplicate integer arrays near the end of the file.
        This results in an assertion failure, and if assertions are turned off, corrupted output.

        When writing an integer array it's important not to add an object reference to the
        aggregate buffer. The writing of the array does not depend on the aggregate buffer. But,
        more importantly, it's possible this instance integer array is a duplicate and won't be
        written out. If so, there's no guarantee there's enough space in the aggregate buffer to
        store the object references (the references will be ignored). In some cases the aggregate
        buffer can then overrun the data being written; normally this is prevented by the fact that
        the data being written will include a copy of the aggregate buffer.

        Also removed a bit of unneeded dead code to handle the integer -1.

        * platform/cf/BinaryPropertyList.cpp:
        (WebCore::BinaryPropertyListPlan::BinaryPropertyListPlan): Removed unneeded
        m_integerNegativeOneObjectReference, since property lists support only non-negative integers.
        (WebCore::BinaryPropertyListPlan::writeInteger): Removed support for
        m_integerNegativeOneObjectReference.
        (WebCore::BinaryPropertyListPlan::integerObjectReference): Ditto.
        (WebCore::BinaryPropertyListSerializer::writeIntegerWithoutAddingAggregateObjectReference):
        Added. Factored out most of writeInteger, for use in writeIntegerArray, without calling
        addAggregateObjectReference.
        (WebCore::BinaryPropertyListSerializer::writeInteger): Changed to call the new
        writeIntegerWithoutAddingAggregateObjectReference function.
        (WebCore::BinaryPropertyListSerializer::writeIntegerArray): Call the new
        writeIntegerWithoutAddingAggregateObjectReference function and therefore remove the
        code to save and restore m_currentAggregateBufferByte, which is no longer needed.



git-svn-id: svn://svn.chromium.org/blink/trunk@42432 bbb929c8-8fbe-4397-9dbb-9b2b20218538

git-svn-id: svn://svn.chromium.org/blink/trunk@42431 bbb929c8-8fbe-4397-9dbb-9b2b20218538

2009-04-12  Sam Weinig  <sam@webkit.org>

        Reviewed by Mark Rowe.

        Move the event and crypto attributes and the captureEvents and releaseEvents
        functions out of JSDOMWindowBase by partially generating them.

        * bindings/js/JSDOMGlobalObject.cpp:
        (WebCore::JSDOMGlobalObject::currentEvent): Make const.
        * bindings/js/JSDOMGlobalObject.h:
        * bindings/js/JSDOMWindowBase.cpp:
        * bindings/js/JSDOMWindowCustom.cpp:
        (WebCore::JSDOMWindow::crypto): Not implemented.
        (WebCore::JSDOMWindow::event):
        (WebCore::JSDOMWindow::captureEvents): Not implemented.
        (WebCore::JSDOMWindow::releaseEvents): Ditto.
        * page/DOMWindow.idl:

LayoutTests:

2009-04-12  Sam Weinig  <sam@webkit.org>

        Reviewed by Mark Rowe.

        Update test result for new nop window.crypto behavior.

        * http/tests/security/cross-frame-access-get-expected.txt:



git-svn-id: svn://svn.chromium.org/blink/trunk@42430 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Update results.

        * fast/dom/Window/window-properties-expected.txt:
        * fast/js/global-constructors-expected.txt:
        * platform/mac-leopard/fast/dom/Window/window-properties-expected.txt:



git-svn-id: svn://svn.chromium.org/blink/trunk@42429 bbb929c8-8fbe-4397-9dbb-9b2b20218538

2009-04-12  Sam Weinig  <sam@webkit.org>

        Reviewed by Mark Rowe.

        Move more attributes out of JSDOMWindowBase by
        partially generating them.

        - Expose the MessagePort constructor.

        * bindings/js/JSDOMGlobalObject.h:
        (WebCore::getDOMConstructor):
        * bindings/js/JSDOMWindowBase.cpp:
        (setJSDOMWindowBaseEvent):
        * bindings/js/JSDOMWindowCustom.cpp:
        (WebCore::JSDOMWindow::image):
        (WebCore::JSDOMWindow::option):
        (WebCore::JSDOMWindow::audio):
        (WebCore::JSDOMWindow::webKitPoint):
        (WebCore::JSDOMWindow::webKitCSSMatrix):
        (WebCore::JSDOMWindow::xMLHttpRequest):
        (WebCore::JSDOMWindow::xSLTProcessor):
        (WebCore::JSDOMWindow::messageChannel):
        (WebCore::JSDOMWindow::worker):
        * page/DOMWindow.idl:

LayoutTests:

2009-04-12  Sam Weinig  <sam@webkit.org>

        Reviewed by Mark Rowe.

        Update tests after making MessageChannel completely missing when not
        compiling with support for it (and removing it from the tests because
        we never compile with support for it right now) and exposing the MessagePort
        constructor.

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



git-svn-id: svn://svn.chromium.org/blink/trunk@42428 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Dan Bernstein.

        * inspector/InspectorResource.cpp:
        (WebCore::InspectorResource::updateScriptObject): Fix a typo in the
        lastPathComponent property name.

git-svn-id: svn://svn.chromium.org/blink/trunk@42427 bbb929c8-8fbe-4397-9dbb-9b2b20218538

2009-04-12  Adele Peterson  <adele@apple.com>

        Reviewed by Darin Adler.

        Test: editing/inserting/6609479-1.html

        Follow-up to <rdar://problem/6609479>.

        Before my last change, a Mail blockquote would've been broken up earlier, and you would've been guaranteed that the inserted content
        was at the start of a paragraph.  Now we need to check for that explicitly, and only merge the start of the inserted content if it is 
        at the start of a paragraph.

        * editing/ReplaceSelectionCommand.cpp: (WebCore::ReplaceSelectionCommand::shouldMergeStart):

LayoutTests:

2009-04-12  Adele Peterson  <adele@apple.com>

        Reviewed by Darin Adler.

        Another test for <rdar://problem/6609479>.  This covers inserting HTML in a quoted table.

        * editing/inserting/6609479-1-expected.txt: Added.
        * editing/inserting/6609479-1.html: Added.



git-svn-id: svn://svn.chromium.org/blink/trunk@42426 bbb929c8-8fbe-4397-9dbb-9b2b20218538

2009-04-11  Adele Peterson  <adele@apple.com>

        Reviewed by Dan Bernstein.

        Fix for <rdar://problem/6609479> Pressing return inside a table cell that's inside quoted content will split the table

        Test: editing/inserting/6609479.html

        Don't break a blockquote if pasting or inserting into a table.

        * editing/ReplaceSelectionCommand.cpp: (WebCore::ReplaceSelectionCommand::doApply):
        * editing/TypingCommand.cpp: (WebCore::TypingCommand::insertParagraphSeparatorInQuotedContent):

LayoutTests:

2009-04-11  Adele Peterson  <adele@apple.com>

        Reviewed by Dan Bernstein.

        Tests for <rdar://problem/6609479> Pressing return inside a table cell that's inside quoted content will split the table

        * editing/inserting/5418891.html: Changed test so it doesn't involve tables.
        * editing/inserting/6609479-expected.txt: Added.
        * editing/inserting/6609479.html: Added.
        * platform/mac/editing/inserting/5418891-expected.checksum:
        * platform/mac/editing/inserting/5418891-expected.png:
        * platform/mac/editing/inserting/5418891-expected.txt:



git-svn-id: svn://svn.chromium.org/blink/trunk@42425 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Darin Adler.

        - fix <rdar://problem/6643103> Unscaled values from body.scrollHeight

        Test: fast/css/zoom-body-scroll.html

        Adjust scroll{Left,Top,Width,Height} values for zoom in getters and
        setters.

        * html/HTMLBodyElement.cpp:
        (WebCore::adjustForZoom):
        (WebCore::HTMLBodyElement::scrollLeft):
        (WebCore::HTMLBodyElement::setScrollLeft):
        (WebCore::HTMLBodyElement::scrollTop):
        (WebCore::HTMLBodyElement::setScrollTop):
        (WebCore::HTMLBodyElement::scrollHeight):
        (WebCore::HTMLBodyElement::scrollWidth):

LayoutTests:

        Reviewed by Darin Adler.

        - test for <rdar://problem/6643103> Unscaled values from body.scrollHeight

        * fast/css/zoom-body-scroll-expected.txt: Added.
        * fast/css/zoom-body-scroll.html: Added.



git-svn-id: svn://svn.chromium.org/blink/trunk@42424 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        - fix <rdar://problem/6752340> Light blue-green background in content
          area in Mail

        * WebView/WebView.mm:
        (-[WebView _commonInitializationWithFrameName:groupName:usesDocumentViews:]):
        Use device white, rather than calibrated white, as the default
        background color.



git-svn-id: svn://svn.chromium.org/blink/trunk@42423 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        <rdar://problem/6781108> Having a breakpoint active prevents page reload

        Reviewed by Darin Adler.

        * inspector/InspectorController.cpp:
        (WebCore::InspectorController::didContinue): Call the resumedScript function.
        * inspector/InspectorController.h: Added didContinue.
        * inspector/JavaScriptDebugListener.h: Ditto.
        * inspector/JavaScriptDebugServer.cpp:
        (WebCore::JavaScriptDebugServer::pauseIfNeeded): Call didContinue on the listeners.
        * inspector/front-end/ScriptsPanel.js:
        (WebInspector.ScriptsPanel.prototype.debuggerResumed): Update the interface and state.
        * inspector/front-end/inspector.js:
        (WebInspector.resumedScript): Call ScriptsPanel.debuggerResumed.
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::continueLoadAfterNavigationPolicy): Call resumeDebugger.

git-svn-id: svn://svn.chromium.org/blink/trunk@42422 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        * platform/win/fast/text/international/complex-character-based-fallback-expected.txt:

git-svn-id: svn://svn.chromium.org/blink/trunk@42421 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Rubber stamped by Adam Roben.

        Disabled fast/dom/gc-8.html, which was using Timer in an unsafe way that now
        triggers an assertion.

        We'll reenable the test after fixing the problem. But this is a test-only problem.

        * fast/dom/gc-8.html: Removed.
        * fast/dom/gc-8.html-disabled: Copied from fast/dom/gc-8.html.



git-svn-id: svn://svn.chromium.org/blink/trunk@42420 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        
        r21436 changed the semantics of AXObjectCache::get() to not create
        objects on demand. Changing this get() call into a getOrCreate() call
        fixes the crash.
        
        Rubber stamped by Jon Honeycutt.

        * AccessibleDocument.cpp:
        (AccessibleDocument::AccessibleDocument):



git-svn-id: svn://svn.chromium.org/blink/trunk@42419 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Adele Peterson.

        Remove unused include for JSLock.

        * page/mac/FrameMac.mm:



git-svn-id: svn://svn.chromium.org/blink/trunk@42418 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Darin Adler.

        Add back the keepAlive timer for after the script execution to
        protect the callers of evaluate, as was part of the original intent
        of timer.

        * bindings/js/ScriptController.cpp:
        (WebCore::ScriptController::evaluate):
        * page/mac/FrameMac.mm:



git-svn-id: svn://svn.chromium.org/blink/trunk@42417 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        - test and expected result updates following r42379

        * fast/text/offsetForPosition-cluster-at-zero-expected.txt: Updated
        results.
        * fast/text/offsetForPosition-cluster-at-zero.html: Removed
        font-dependent alert that was accidently left in this test.
        * platform/mac-tiger/fast/text/international/thai-line-breaks-expected.checksum:
        * platform/mac-tiger/fast/text/international/thai-line-breaks-expected.png:
        * platform/mac-tiger/fast/text/international/thai-line-breaks-expected.txt:
        * platform/mac-tiger/fast/text/monospace-width-cache-expected.checksum: Added.
        * platform/mac-tiger/fast/text/monospace-width-cache-expected.png: Added.
        * platform/mac-tiger/fast/text/monospace-width-cache-expected.txt: Added.
        * platform/win/Skipped:



git-svn-id: svn://svn.chromium.org/blink/trunk@42416 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        If we reload, the animation controller sticks around and has
        a stale animation time. So the needed delay is computed wrong.
        I added a reset of the animation timer on load.

        I did not do a test case because this only happens on reload so I
        couldn't make an automated test. I tried using an iframe and using
        reload() but that doesn't tickle the bug.



git-svn-id: svn://svn.chromium.org/blink/trunk@42415 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Dan Bernstein.

        * DumpRenderTree/win/DumpRenderTree.cpp:
        (initialize): Added WebKit Layout Tests 2.ttf.

git-svn-id: svn://svn.chromium.org/blink/trunk@42414 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Darin Adler.

        Move WORKERS header files inside ENABLE(WORKERS).
        So if WORKERS is not enabled and the generated files are not generated, the build is still working.

        * bindings/js/JSDOMGlobalObject.cpp:
        * bindings/js/JSMessageChannelConstructor.cpp:
        * bindings/js/ScheduledAction.cpp:



git-svn-id: svn://svn.chromium.org/blink/trunk@42413 bbb929c8-8fbe-4397-9dbb-9b2b20218538

2009-04-10  Darin Adler  <darin@apple.com>

        Reviewed by Brady Eidson.

        <rdar://problem/6773515> crash in push_heap inside WebCore when printing

        The crash was due to manipulating a timer on a different thread than the one it was created on.
        This adds some diagnostics so we'll catch that kind of mistake immediately.

        * platform/Timer.cpp:
        (WebCore::TimerBase::TimerBase): Store the thread identifier in debug versions.
        (WebCore::TimerBase::start): Assert it's the same thread the timer was created on.
        (WebCore::TimerBase::stop): Ditto.
        (WebCore::TimerBase::isActive): Ditto.
        (WebCore::TimerBase::setNextFireTime): Ditto.

        * platform/Timer.h: Added the thread identifier data member.

WebKit/mac:

2009-04-10  Darin Adler  <darin@apple.com>

        Reviewed by Brady Eidson.

        <rdar://problem/6773515> crash in push_heap inside WebCore when printing

        The crash was due to manipulating a timer on a different thread than the one it was created on.

        * History/WebHistoryItem.mm:
        (-[WebWindowWatcher windowWillClose:]): Call later on main thread, if called on non-main thread.
        * WebView/WebHTMLView.mm:
        (-[WebHTMLView windowDidBecomeKey:]): Ditto.
        (-[WebHTMLView windowDidResignKey:]): Ditto.
        (-[WebHTMLView windowWillClose:]): Ditto.
        (-[WebHTMLView _updateControlTints]): Added. Factored out the non-thread-safe part of
        our override of _windowChangedKeyState.
        (-[WebHTMLView _windowChangedKeyState]): Call _updateControlTints later on main thread, if
        called on non-main thread.
        * WebView/WebPreferences.mm:
        (-[WebPreferences _postPreferencesChangesNotification]): Call later on main thread, if called
        on non-main thread



git-svn-id: svn://svn.chromium.org/blink/trunk@42412 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        This allows the IDL files to be found in Solution Explorer.

        Rubber-stamped in advance by Steve Falkenburg.

        * WebKit.vcproj/Interfaces.vcproj:

git-svn-id: svn://svn.chromium.org/blink/trunk@42411 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Dan Bernstein

        https://bugs.webkit.org/show_bug.cgi?id=25108

        If you remove a class with a transition while that transition is running
        the animation timer will continue to fire after the transition is finished.
        This has no visual indication, but it does drain the processor. And in some
        cases it might even cause a glitch in future animations. Unfortunately there
        is no way to test this without putting in printfs.

        This happens because the animation logic is never traversed after a transition
        is removed, so we never get a chance to cleanup. So I added cleanup in the logic
        that fires the dispatch of the last style change when the animation finishes.

        Test: transitions/remove-transition-style.html

        * page/animation/AnimationController.cpp:
        (WebCore::AnimationControllerPrivate::updateStyleIfNeededDispatcherFired):
        * page/animation/CompositeAnimation.cpp:
        (WebCore::CompositeAnimation::updateTransitions):
        (WebCore::CompositeAnimation::animate):
        (WebCore::CompositeAnimation::cleanupFinishedAnimations):
        * page/animation/CompositeAnimation.h:

git-svn-id: svn://svn.chromium.org/blink/trunk@42410 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Geoffrey Garen.

        Fix for <rdar://problem/6648858>
        CrashTracer: [REGRESSION(Safari 4)] 60 crashes in Safari at com.apple.WebCore • WebCore::ScriptController::evaluate + 241

        The Frame (and therefore ScriptController) were being destroyed
        during JavaScript execution causing the JSDOMWindowShell to become
        null.

        * bindings/js/ScriptController.cpp:
        (WebCore::ScriptController::evaluate): Use a RefPtr to protect the Frame
        instead of the keep-alive timer, since a nested event loop used from with-in 
        JavaScript execution could cause the timer to fire before returning.



git-svn-id: svn://svn.chromium.org/blink/trunk@42409 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        Reviewed by Timothy Hatcher.

        https://bugs.webkit.org/show_bug.cgi?id=25133
        Fix an error in InspectorController's bindings, clean up a few header
        includes, make empty object construction in ScriptObject less ambiguous.

        * bindings/js/JSInspectorControllerCustom.cpp:
        (WebCore::JSInspectorController::setSetting): Send correct argument from
            the bindings.
        * bindings/js/ScriptObject.cpp:
        (WebCore::ScriptObject::createNew): Added.
        * bindings/js/ScriptObject.h: Removed constructor, added decl for createNew.
        * inspector/ConsoleMessage.cpp: Remove unused header.
        * inspector/ConsoleMessage.h: Ditto.
        * inspector/InspectorResource.cpp:
        (WebCore::createHeadersObject): Changed to use ScriptObject::createNew.

git-svn-id: svn://svn.chromium.org/blink/trunk@42408 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        - re-enable css2.1 tests that were disabled following r42379 and make
          them use a different fallback font for certain characters, to get
          the vertical metrics to match Mac

        * platform/win/Skipped:
        * platform/win/css2.1/resources/Mac-compatible-font-fallback.css:



git-svn-id: svn://svn.chromium.org/blink/trunk@42407 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        This both fixes a bug where I was not sending in the right time
        to continue the animation and was not recomputing current time
        when coming out of suspend.



git-svn-id: svn://svn.chromium.org/blink/trunk@42406 bbb929c8-8fbe-4397-9dbb-9b2b20218538

        WebKitSystemInterface changes for <rdar://problem/6646998>

        * libWebKitSystemInterfaceLeopard.a:
        * libWebKitSystemInterfaceTiger.a:



git-svn-id: svn://svn.chromium.org/blink/trunk@42405 bbb929c8-8fbe-4397-9dbb-9b2b20218538