Back up partial change in r48072 in order to fix chromium reliability build break.

Reviewed by NOBODY (Chromium reliability build break).

* bindings/v8/V8AbstractEventListener.cpp:
(WebCore::V8AbstractEventListener::invokeEventHandler):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48096 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-05  Nicolas Weber  <thakis@chromium.org>

        Reviewed by Darin Fisher.

        https://bugs.webkit.org/show_bug.cgi?id=28975
        Fix file upload on chrome/mac. Images that have "Hide Extension" set,
        such as screenshots by default, can now be uploaded to sites such as
        imgur.com.

        * platform/FileSystem.h:
        Add pathGetDisplayFileName().
        * platform/chromium/FileChooserChromium.cpp:
        (WebCore::FileChooser::basenameForWidth):
        Call pathGetDisplayFileName() instead of pathGetFileName().
        * platform/chromium/FileSystemChromiumLinux.cpp:
        (WebCore::pathGetDisplayFileName):
        Implement pathGetDisplayFileName() by delegating to pathGetFileName().
        * platform/chromium/FileSystemChromiumMac.mm:
        (WebCore::pathGetFileName):
        Now returns a real filename instead of a presentational one, which
        fixes the problem addressed by this change.
        (WebCore::pathGetDisplayFileName):
        New function that returns a presentational filename; does what
        pathGetDisplayFileName() did prior to this patch.
        * platform/chromium/FileSystemChromiumWin.cpp:
        (WebCore::pathGetDisplayFileName):
        Implement pathGetDisplayFileName() by delegating to pathGetFileName().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48095 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-05  Joel Stanley  <joel@jms.id.au>

        Reviewed by David Levin.

        Expose functions to set the caret blink interval for Linux Chromium.
        https://bugs.webkit.org/show_bug.cgi?id=28931

        * rendering/RenderThemeChromiumLinux.cpp:
        (WebCore::RenderThemeChromiumLinux::setCaretBlinkInterval):
        (WebCore::RenderThemeChromiumLinux::caretBlinkInterval):
        * rendering/RenderThemeChromiumLinux.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48094 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-05  Eric Seidel  <eric@webkit.org>

        No review, results update only.

        Update results after r48043.

        * fast/js/JSON-stringify-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48093 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Prevent UI freeze in case when too many timers are in the timer queue.
https://bugs.webkit.org/show_bug.cgi?id=23865

Reviewed by David Levin.

The fix measures the elapsed time while executing timers. If we have too many
timers and it takes significant time to fire, quit the loop and reschedule.
This lets the run loop process user input (close the window for example).

* platform/ThreadTimers.h:
* platform/ThreadTimers.cpp:
(WebCore::ThreadTimers::sharedTimerFiredInternal):
(WebCore::ThreadTimers::fireTimersInNestedEventLoop):
Removed functions and members used before to grab a list of all expired timers.
Now we fetch one by one, measuring time and quitting the loop if it takes too long.
* platform/Timer.cpp:
(WebCore::TimerBase::isActive):
(WebCore::TimerBase::setNextFireTime):
Since timers are now fired one by one, there is no need to keep track of updated timers.
* manual-tests/input-starved-by-timers.html: Manual test that attempts to freeze browser by
creating of enough timers. It then keeps UI frozen and after 10 seconds starts to drain the timers
so the browser un-freezes. This is instead of a regular layout test because we don't have a way
to inject an input events via regular run loop in DRT - instead they are directly dispatched by
eventSender, which does not reproduce the UI freeze that happens with real input.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48086 268f45cc-cd09-0410-ab3c-d52691b4dbfc

<rdar://problem/7201063>

Reviewed by Adele Peterson.

Detach any custom scrollbars before detaching the document.

* page/Frame.cpp:
(WebCore::Frame::setView):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48085 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Windows build fix

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48084 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[[ToString]] conversion should use the actual toString function for String objects.

Reviewed by Gavin Barraclough.

Remove incorrect specialisations of toString conversions on StringObject.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48083 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed, build fix.

Remove remaining references to UserStyleSheetLoader after r48075.

* WebCore.gypi:
* WebCore.xcodeproj/project.pbxproj:
* page/qt/FrameQt.cpp:
* page/wince/FrameWince.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48082 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Michelangelo De Simone  <micdesim@gmail.com>

        Reviewed by David Levin.

        https://bugs.webkit.org/show_bug.cgi?id=28988
        Minor fix: "novalidate" attribute renamed in "noValidate", according to
        HTML5 specs:
        http://www.whatwg.org/specs/web-apps/current-work/#dom-fs-novalidate

        * fast/forms/novalidate-attribute-expected.txt:
        * fast/forms/resources/novalidate-attribute.js:
2009-09-04  Michelangelo De Simone  <micdesim@gmail.com>

        Reviewed by David Levin.

        https://bugs.webkit.org/show_bug.cgi?id=28988
        Minor fix: "novalidate" attribute renamed in "noValidate", according to
        HTML5 specs:
        http://www.whatwg.org/specs/web-apps/current-work/#dom-fs-novalidate

        * html/HTMLFormElement.cpp:
        (WebCore::HTMLFormElement::noValidate):
        (WebCore::HTMLFormElement::setNoValidate):
        * html/HTMLFormElement.h:
        * html/HTMLFormElement.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48081 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove unused line of code from WorkerContext
https://bugs.webkit.org/show_bug.cgi?id=28990

Reviewed by David Levin.

* workers/WorkerContext.cpp:
(WebCore::WorkerContext::importScripts):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48080 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix assertion seen when running buildbot.

Patch by Darin Adler <darin@apple.com> on 2009-09-04
Reviewed by Oliver Hunt.

* bindings/js/JSDOMBinding.h:
(WebCore::DOMObjectWithGlobalPointer::createStructure): Create the structure without
the HasDefaultMark bit since this has a custom mark function.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48079 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Fix expected test output.

        * fast/dom/DOMImplementation/detached-doctype-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48078 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Windows build fix.

* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def: Add new export.
* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore_debug.def: Add new export.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48077 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Windows build fix.

* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def: Remove unneeded export.
* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore_debug.def: Remove unneeded export.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48076 268f45cc-cd09-0410-ab3c-d52691b4dbfc

https://bugs.webkit.org/show_bug.cgi?id=28984, remove the remote stylesheet loader on Mac/Qt.
Make sure the other code path (now used by everybody) can handle data URLs.

Reviewed by Adam Roben.

As per the post on webkit-dev, the ability to work remotely isn't working properly anyway (it doesn't
hold up the pending sheet count, so you'd just FOUC on every Web page), and it's extremely unlikely
that anybody is relying on this support.

Tor Arne agreed it was ok to disable (he enabled it on the Qt side), and only Mac/Qt were using this
code.

* WebCore.pro:
* WebCore.xcodeproj/project.pbxproj:
* dom/Document.cpp:
(WebCore::Document::userStyleSheet):
* dom/Document.h:
* loader/FrameLoader.cpp:
(WebCore::FrameLoader::begin):
* loader/UserStyleSheetLoader.cpp: Removed.
* loader/UserStyleSheetLoader.h: Removed.
* page/Frame.cpp:
(WebCore::Frame::reapplyStyles):
* page/Frame.h:
* page/Page.cpp:
(WebCore::Page::userStyleSheetLocationChanged):
(WebCore::Page::userStyleSheet):
* page/mac/FrameMac.mm:
* page/qt/FrameQt.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48075 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Mikhail Naganov  <mnaganov@chromium.org>

        Reviewed by Timothy Hatcher.

        Add "sampling profiler" preference, simplify data display when it's set.

        https://bugs.webkit.org/show_bug.cgi?id=28969

        * inspector/front-end/ProfileDataGridTree.js:
        (WebInspector.ProfileDataGridNode.prototype.get data.formatMilliseconds):
        (WebInspector.ProfileDataGridNode.prototype.get data):
        * inspector/front-end/ProfileView.js:
        (WebInspector.ProfileView):
        * inspector/front-end/inspector.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48074 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove bogus assert that is affecting layout tests.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48073 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[V8] Run-time exception in onmessage handler is not forwarded to the worker object.
https://bugs.webkit.org/show_bug.cgi?id=28980

Reviewed by Adam Barth.

Tested by worker-close.html.

* bindings/v8/ScriptFunctionCall.cpp:
* bindings/v8/V8AbstractEventListener.cpp:
(WebCore::V8AbstractEventListener::invokeEventHandler):
* bindings/v8/V8Utilities.cpp:
(WebCore::getScriptExecutionContext):
(WebCore::reportException):
* bindings/v8/V8Utilities.h:
(WebCore::getScriptExecutionContext):
* bindings/v8/custom/V8MessageChannelConstructor.cpp:
(WebCore::CALLBACK_FUNC_DECL):
* bindings/v8/custom/V8SharedWorkerCustom.cpp:
(WebCore::CALLBACK_FUNC_DECL):
* bindings/v8/custom/V8WorkerCustom.cpp:
(WebCore::CALLBACK_FUNC_DECL):
* bindings/v8/custom/V8XMLHttpRequestConstructor.cpp:
(WebCore::CALLBACK_FUNC_DECL):
* bindings/v8/custom/V8XMLHttpRequestCustom.cpp:
(WebCore::CALLBACK_FUNC_DECL):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48072 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove unused m_sheet variable and setCSSStyleSheet methods.

Reviewed by Anders Carlsson.

* dom/Document.cpp:
* dom/Document.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48071 268f45cc-cd09-0410-ab3c-d52691b4dbfc

(Based on a patch from Adele).

Reviewed by David Hyatt and Adele Peterson.

Fix <rdar://problem/7185875>.

* page/Frame.cpp:
(WebCore::Frame::setView):
Detach any custom scroll bars from the old view.

* page/FrameView.cpp:
(WebCore::FrameView::detachCustomScrollbars):
Simplify the check (and make it work when body is null) by just checking that the scroll bar owner is
not a RenderPart object.

* rendering/RenderObject.cpp:
(WebCore::RenderObject::destroy):
ASSERT that frame() is not null here and also add a null check so that release builds won't crash.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48070 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Make m_userStyleSheetLoader in Frame an OwnPtr.

Reviewed by Adam Roben.

* page/Frame.cpp:
(WebCore::Frame::Frame):
(WebCore::Frame::~Frame):
* page/Frame.h:
* page/mac/FrameMac.mm:
(WebCore::Frame::setUserStyleSheetLocation):
(WebCore::Frame::setUserStyleSheet):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48069 268f45cc-cd09-0410-ab3c-d52691b4dbfc

DateInstance object collected on ARM JIT (JSValue: WTF_USE_JSVALUE32)
https://bugs.webkit.org/show_bug.cgi?id=28909

Patch by Darin Adler <darin@apple.com> on 2009-09-04
Reviewed by Geoff Garen.

Part two.

Make some improvements to garbage collection code:

    1) Create a runtime assertion that catches any classes that
       override markChildren but have the HasDefaultMark bit set.
    2) Remove checks of the mark bit outside the MarkStack::append
       function; they are redundant.
    3) Improve the efficiency of the asObject and asArray functions
       when called on JSCell* to avoid a round trip to JSValue.
    4) Make more callers use the checked asCell and asObject
       casting functions rather than unchecked casts.
    5) Removed the JSCell::marked function and other GC-related
       functions because these operations are no longer things that
       code other than the core GC code needs to do directly. Fixed
       callers that were calling them.

* runtime/Collector.cpp:
(JSC::Heap::markConservatively): Removed unneeded call to MarkStack::drain.
(JSC::Heap::markProtectedObjects): Removed unneeded check of the mark
bit and call to MarkStack::drain.
(JSC::Heap::collect): Removed unneeded checks of the mark bit and also
changed call to SmallStrings::mark to call markChildren instead to match
the rest of the objects.
(JSC::typeName): Removed unneeded cast to JSObject*.

* runtime/JSArray.h:
(JSC::asArray): Added an overload for JSCell* and changed the JSValue
version to call it. Removed some unneeded casts.
(JSC::JSArray::markChildrenDirect): Marked this function inline. It's in
a header, and if not marked inline this could lead to linking problems.
(JSC::MarkStack::markChildren): Added. This helper function is used by
the drain function to avoid repating code. Also added the code here to
check fro default mark violations in debug code. If a markChildren
function adds something to the mark stack, but the type info claimed
hasDefaultMark was true, then we will get an assertion now. Also fixed
the assertion about the mark bit to use the Heap function directly
because we don't have a JSCell::marked function any more.
(JSC::MarkStack::drain): Changed a local variable from "v" to "value",
and from "currentCell" to "cell". Changed to call markChildren in two
places instead of repeating a chain of if statements twice. Changed
code that reads and writes the mark bit to use Heap::isCellMarked and
Heap::markCell so we can eliminate the JSCell::marked and
JSCell::markCellDirect functions.

* runtime/JSCell.h: Removed JSCell's markCellDirect and marked member
functions. Added a comment explaining that asCell should be deprecated
in favor of the JSValue asCell member function.
(JSC::MarkStack::append): Added the assertion that catches callers
that have set the HasDefaultMark bit incorrectly. Changed
code that reads and writes the mark bit to use Heap::isCellMarked and
Heap::markCell so we can eliminate the JSCell::marked and
JSCell::markCellDirect functions. Moved the overload of
MarkStack::append for JSValue here so it can call through to the cell
version. The old version had a copy of all the code instead, but that
repeated the conversion from JSValue to JSCell* and the check for
whether a value is a cell multiple times.
(JSC::Structure::markAggregate): Moved this function here to avoid
dependencies for Structure.h, since this calls MarkStack::append.

* runtime/JSObject.cpp:
(JSC::JSObject::markChildren): Added code to clear
m_isCheckingForDefaultMarkViolation so the marking done by JSObject
doesn't trigger the assertion.

* runtime/JSValue.h: Moved some stray includes that were outside the
header guard inside it. Not sure how that happened! Removed the
GC-related member functions markChildren, hasChildren, marked, and
markDirect.

* runtime/JSWrapperObject.h: Made markChildren private.
(JSC::JSWrapperObject::createStructure): Added. Fixes a bug where the
HasDefaultMark bit was set.

* runtime/MarkStack.h: Added m_isCheckingForDefaultMarkViolation and
initialized it to false. Moved the append function body from here to
JSCell.h. Added a declaration of a private markChildren function used
inside the drain function.

* runtime/SmallStrings.cpp:
(JSC::SmallStrings::markChildren): Changed the name and style of this
function to match other functions. This allows us to share the normal
mark stack code path.

* runtime/SmallStrings.h: Changed the name and interface of mark to
the more-normal markChildren style.

* runtime/Structure.h: Moved the body of markAggregate into the
JSCell.h to avoid a circular dependency with JSCell.h.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48068 268f45cc-cd09-0410-ab3c-d52691b4dbfc

JavaScriptCore: DateInstance object collected on ARM JIT (JSValue: WTF_USE_JSVALUE32)
https://bugs.webkit.org/show_bug.cgi?id=28909

Patch by Darin Adler <darin@apple.com> on 2009-09-04
Reviewed by Geoff Garen.

Part one.

Make some improvements to garbage collection code:

    1) Fix the two classes that had the default mark bit set but
       should not.
    2) Remove checks of the mark bit outside the MarkStack::append
       function; they are redundant.
    3) Make more callers use the checked asCell and asObject
       casting functions rather than unchecked casts.
    4) Removed some GC-related functions because these operations are
       no longer things that code other than the core GC code needs
       to do directly. Fixed callers that were calling them.

* bytecode/CodeBlock.cpp:
(JSC::CodeBlock::markAggregate): Removed unneeded check of the mark
bit before calling MarkStack::append.

* interpreter/Register.h: Removed unneeded marked and markChildren
functions.

* jit/JITStubs.cpp:
(op_eq): Removed unneeded assertions, instead using checked casting
functions such as asObject.

* runtime/ArgList.h: Added now-needed forward declaration of MarkStack.

* runtime/GetterSetter.cpp:
(JSC::GetterSetter::markChildren): Remmoved unneeded check of the mark bit.

* runtime/GlobalEvalFunction.h:
(JSC::GlobalEvalFunction::createStructure): Added. Fixes a bug where the
HasDefaultMark bit was set.

* runtime/JSCell.cpp:
(JSC::JSCell::getObject): Use asObject to avoid a direct static_cast.

* runtime/JSObject.h:
(JSC::asObject): Added an overload for JSCell* and changed the JSValue
version to call it.
(JSC::JSValue::get): Use asObject to avoid a direct static_cast.

* runtime/JSValue.h: Moved some stray includes that were outside the
header guard inside it. Not sure how that happened! Removed the
GC-related member functions markChildren, hasChildren, marked, and
markDirect.

* runtime/JSWrapperObject.h: Made markChildren private.
(JSC::JSWrapperObject::createStructure): Added. Fixes a bug where the
HasDefaultMark bit was set. Later we may want to optimize this for
wrapper types that never have cells in their internal values, but there
is no measured performance regression in SunSpider or V8 doing this
all the time.

* runtime/MarkStack.cpp: Tweaked formatting.

JavaScriptGlue: * JSValueWrapper.cpp:
(JSValueWrapper::JSObjectMark): Removed a check of the mark
bit. It's OK to do more work in this case, and there is no
longer a public function to access the mark bit.

Reviewed by Geoff Garen.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48067 268f45cc-cd09-0410-ab3c-d52691b4dbfc

wx build fix. Switch USE_ defines over to the compiler so that they can be
checked by files not including config.h (like WebCorePrefix.h).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48066 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Show color swatches in the Web Inspector before the color text so clicking them will not
shift the swatch. Also makes multiple swatches in the single property toggle the color format.

https://bugs.webkit.org/show_bug.cgi?id=28978

Reviewed by Darin Adler.

* inspector/front-end/StylesSidebarPane.js:
(WebInspector.StylePropertyTreeElement.prototype.updateTitle): Some refactoring to consolidate
the processing we do on the property value. This eliminated the old nickname code, since the new
WebInspector.Color class handles this. We could also simplify the color regex since more
is handled by the Color class. Also no longer uses innerHTML for the linkify code.
(WebInspector.StylePropertyTreeElement.prototype.updateTitle.processValue): Helper function to
process a value given a regex and processor functions.
(WebInspector.StylePropertyTreeElement.prototype.updateTitle.linkifyURL): Make an anchor for the
URL with "url()" syntax surrounding it.
(WebInspector.StylePropertyTreeElement.prototype.updateTitle.processColor): Makes a color, if
there was an exception just return a text node.
(WebInspector.StylePropertyTreeElement.prototype.updateTitle.processColor.changeColorDisplay.changeTo):
Moved from later in the file.
(WebInspector.StylePropertyTreeElement.prototype.updateTitle.processColor.changeColorDisplay): Ditto.
* inspector/front-end/inspector.css:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48065 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

2009-09-04  Mark Mentovai  <mark@chromium.org>

        Reviewed by Dave Hyatt.

        https://bugs.webkit.org/show_bug.cgi?id=28614

        Account for scrollbar state changes that occur during layout.

        * platform/ScrollView.cpp:
        (WebCore::ScrollView::updateScrollbars):

        Perform a layout prior to checking whether the scrollbar modes are
        off, on, or automatic.  The modes may change during layout.

        * WebCore.base.exp:
        * WebCore.order:
        * page/Frame.cpp:
        (WebCore::Frame::createView):
        * page/FrameView.cpp:
        (WebCore::FrameView::FrameView):
        (WebCore::FrameView::resetScrollbars):
        (WebCore::FrameView::layout):
        * page/FrameView.h:
        * platform/ScrollView.h:

        Eliminate duplicated (and incorrect) scrollbar mode tracking between
        FrameView and ScrollView.

WebKit/mac:

2009-09-04  Mark Mentovai  <mark@chromium.org>

        Reviewed by Dave Hyatt.

        https://bugs.webkit.org/show_bug.cgi?id=28614

        Account for scrollbar state changes that occur during layout.

        * WebView/WebDynamicScrollBarsView.mm:
        (-[WebDynamicScrollBarsView updateScrollers]):

        Perform a layout prior to checking whether the scrollbar modes are
        off, on, or automatic.  The modes may change during layout.

        * WebView/WebFrameView.mm:
        (-[WebFrameView _install]):

        Eliminate duplicated (and incorrect) scrollbar mode tracking between
        FrameView and ScrollView.

WebKit/qt:

2009-09-04  Mark Mentovai  <mark@chromium.org>

        Reviewed by Dave Hyatt.

        https://bugs.webkit.org/show_bug.cgi?id=28614

        Account for scrollbar state changes that occur during layout.

        * Api/qwebframe.cpp:
        (QWebFrame::setScrollBarPolicy):

        Eliminate duplicated (and incorrect) scrollbar mode tracking between
        FrameView and ScrollView.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48064 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore: hostname and host are mixed up when manipulating anchor elements.
https://bugs.webkit.org/show_bug.cgi?id=28954

Patch by Yael Aharon <yael.aharon@nokia.com> on 2009-09-04
Reviewed by Darin Adler.

Swapped the implementation of host and hostname, and made sure not to return
the port number if it is default for the given protocol.
FireFox also avoids returning the protocol number if it is default.

Test: fast/dom/Element/hostname-host.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48063 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix <rdar://problem/7192902>.

Reviewed by Darin Adler and Steve Falkenburg.

Add the necessary plumbing for having QTMovieWin use WebCore timers.

* platform/graphics/win/MediaPlayerPrivateQuickTimeWin.cpp:
(WebCore::TaskTimer::initialize):
Create the timer and call QTMovieWin::setTaskTimerFuncs.

(WebCore::TaskTimer::setTaskTimerDelay):
Start the timer.

(WebCore::TaskTimer::stopTaskTimer):
Stop the timer.

(WebCore::TaskTimer::fired):
Call QTMovieWin::taskTimerFired().

(WebCore::MediaPlayerPrivate::load):
Initialize the Task timer.

* platform/graphics/win/QTMovieWin.cpp:
(updateTaskTimer):
Call the global stopTaskTimer function and setTaskTimerDelay.

(QTMovieWin::taskTimerFired):
Make this a member function.

(QTMovieWin::setTaskTimerFuncs):
Set the global function pointers.

(QTMovieWin::initializeQuickTime):
(DllMain):
No need to use the QT shared timer.
* platform/graphics/win/QTMovieWin.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48062 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix Chromium Win compile.  Missing return statement in
WebCore::RenderThemeChromiumSkia::paintSliderThumb.
https://bugs.webkit.org/show_bug.cgi?id=28974

Patch by Tony Chang <tony@chromium.org> on 2009-09-04
Reviewed by David Levin.

No new tests, build fix.

* rendering/RenderThemeChromiumSkia.cpp:
(WebCore::RenderThemeChromiumSkia::paintSliderThumb):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48061 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Make WebView function without a parent window

When WebView has no parent window, it becomes a message-only window,
which hides it from the screen and prevents it from appearing on the
taskbar.

Fixes <http://webkit.org/b/28971> <rdar://problem/4976998>
WebView::initWithFrame fails if there's no host window

Reviewed by Dave Hyatt.

* WebView.cpp:
(WebView::initWithFrame): m_hostWindow is null, use HWND_MESSAGE as
our parent window, which will make us a message-only window.
(WebView::setHostWindow): If he new host window is null, use
HWND_MESSAGE as our parent window, which will make us a message-only
window.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48060 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Dimitri Glazkov  <dglazkov@chromium.org>

        Unreviewed, build fix.

        [Chromium] Rolling out http://trac.webkit.org/changeset/48035, because it broke
        the build.
        https://bugs.webkit.org/show_bug.cgi?id=28696

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48059 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix vcproj file.

* WebCore.vcproj/WebCore.vcproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48058 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore: https://bugs.webkit.org/show_bug.cgi?id=28890, make simple user script injection work.

Reviewed by Adam Roben.

This patch adds new API for adding and removing user scripts from page groups.  User scripts
are bundled together in isolated worlds (you can have multiple scripts together in the same
world).

Added userscripts/ directory for holding new tests (along with a simple test of script injection).

* WebCore.base.exp:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* bindings/js/ScriptController.cpp:
(WebCore::ScriptController::evaluateInIsolatedWorld):
* bindings/js/ScriptController.h:
* bindings/v8/ScriptController.cpp:
(WebCore::ScriptController::evaluateInIsolatedWorld):
* bindings/v8/ScriptController.h:
* loader/FrameLoader.cpp:
(WebCore::FrameLoader::finishedParsing):
(WebCore::FrameLoader::dispatchDocumentElementAvailable):
* page/Frame.cpp:
(WebCore::Frame::injectUserScripts):
(WebCore::Frame::injectUserScriptsForWorld):
* page/Frame.h:
* page/PageGroup.cpp:
(WebCore::PageGroup::~PageGroup):
(WebCore::PageGroup::addUserScript):
(WebCore::PageGroup::removeUserContentForWorld):
(WebCore::PageGroup::removeAllUserContent):
* page/PageGroup.h:
(WebCore::PageGroup::userScripts):
* page/UserScript.h: Added.
(WebCore::UserScript::UserScript):
(WebCore::UserScript::source):
(WebCore::UserScript::url):
(WebCore::UserScript::patterns):
(WebCore::UserScript::worldID):
(WebCore::UserScript::injectionTime):
* page/UserScriptTypes.h: Added.
(WebCore::):

WebKit/mac: https://bugs.webkit.org/show_bug.cgi?id=28890, make simple user script injection work.

Reviewed by Adam Roben.

This patch adds new API for adding and removing user scripts from page groups.  User scripts
are bundled together in isolated worlds (you can have multiple scripts together in the same
world).

Added userscripts/ directory for holding new tests (along with a simple test of script injection).

* WebView/WebView.mm:
(+[WebView _addUserScriptToGroup:source:url:worldID:patterns:injectionTime:]):
(+[WebView _removeUserContentFromGroup:worldID:]):
(+[WebView _removeAllUserContentFromGroup:]):
* WebView/WebViewPrivate.h:

LayoutTests: https://bugs.webkit.org/show_bug.cgi?id=28890, make simple user script injection work.

Reviewed by Adam Roben.

This patch adds new API for adding and removing user scripts from page groups.  User scripts
are bundled together in isolated worlds (you can have multiple scripts together in the same
world).

Added userscripts/ directory for holding new tests (along with a simple test of script injection).

* platform/mac/userscripts: Added.
* platform/mac/userscripts/script-run-at-end-expected.txt: Added.
* userscripts: Added.
* userscripts/script-run-at-end.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48057 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Added missing ;

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48056 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=26402
        Crashes when using a detached DocumentType node

        Test: fast/dom/DOMImplementation/detached-doctype.html

        DocumentType nodes are the only one that can have null document(), and they can never have
        children, so they are a degenerate case for NodeLists.

        * dom/Node.cpp:
        (WebCore::Node::childNodes): Don't try register node list with document if there's no
        document. Since there are no changes possible for a DocumentType's node list, this is not
        necessary.
        (WebCore::Node::registerDynamicNodeList): Ditto.
        (WebCore::Node::unregisterDynamicNodeList): Ditto. I couldn't find a scenario where this
        was invoked for a DocumentType node, but I also couldn't prove that it can't be (the
        de-registration code path that I saw taken was via Node destructor, and that already has
        a null check for document).

        * bindings/js/JSNodeCustom.cpp:
        (WebCore::JSNode::addEventListener): Bail out if there is no document (just as we already do
        if there is no window).
        (WebCore::JSNode::removeEventListener): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48055 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Adam Barth  <abarth@webkit.org>

        Unreviewed attempt at a build fix.

        Added missing include.

        * WebCoreSupport/WebFrameLoaderClient.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48054 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Adam Barth  <abarth@webkit.org>

        Unreviewed attempt at a build fix.

        Fix typo of IWebFrameLoadDelegatePrivate2 and make
        IWebFrameLoadDelegatePrivate inherit from IWebFrameLoadDelegatePrivate
        as requestd by Adam Roben.

        * Interfaces/IWebFrameLoadDelegatePrivate2.idl:
        * WebCoreSupport/WebFrameLoaderClient.cpp:
        (WebFrameLoaderClient::didDisplayInsecureContent):
        (WebFrameLoaderClient::didRunInsecureContent):
2009-09-04  Adam Barth  <abarth@webkit.org>

        Unreviewed build fix.

        Update declaration of FrameLoadDelegate to reflect that
        IWebFrameLoadDelegatePrivate2 inherits from
        IWebFrameLoadDelegatePrivate.

        * DumpRenderTree/win/FrameLoadDelegate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48053 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Fix brainfart in r48050, tests should have paths

Rubber-stamped by Simon Hausmann.

* platform/qt/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48052 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=24696

        Plumb mixed content notifications to WebFrameLoadDelegatePrivate.

        * WebCoreSupport/WebFrameLoaderClient.mm:
        (WebFrameLoaderClient::didDisplayInsecureContent):
        (WebFrameLoaderClient::didRunInsecureContent):
        * WebView/WebDelegateImplementationCaching.h:
        * WebView/WebFrameLoadDelegatePrivate.h: Added.
2009-09-04  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=24696

        Plumb mixed content notifications to IWebFrameLoadDelegatePrivate.

        * Interfaces/IWebFrameLoadDelegatePrivate2.idl: Added.
        * Interfaces/WebKit.idl:
        * WebCoreSupport/WebFrameLoaderClient.cpp:
        (WebFrameLoaderClient::didDisplayInsecureContent):
        (WebFrameLoaderClient::didRunInsecureContent):
        * WebKit.vcproj/Interfaces.vcproj:
2009-09-04  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=24696

        Add testing instrumentation for mixed content.

        * DumpRenderTree/mac/FrameLoadDelegate.mm:
        (-[FrameLoadDelegate webView:]):
        (-[FrameLoadDelegate webView:didRunInsecureContent:]):
        * DumpRenderTree/win/FrameLoadDelegate.cpp:
        (descriptionSuitableForTestResult):
        (FrameLoadDelegate::QueryInterface):
        (FrameLoadDelegate::didDisplayInsecureContent):
        (FrameLoadDelegate::didRunInsecureContent):
        * DumpRenderTree/win/FrameLoadDelegate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48051 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Skip tests failing due to missing DRT features

Rubber-stamped by Simon Hausmann.

* platform/qt/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48050 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        [GTK] loadFinished never emitted
        https://bugs.webkit.org/show_bug.cgi?id=28935

        Do not report errors right away from
        ResourceHandle::start. Instead return true and schedule an error
        to be reported in the handle later. Otherwise WebCore gets
        confused and thinks the resource never finished loading, making it
        not emit the finished status.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48049 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Yong Li  <yong.li@torchmobile.com>

        Reviewed by David Levin.

        Remove unnecessary dependency on unistd.h
        https://bugs.webkit.org/show_bug.cgi?id=28962

        * runtime/Completion.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48048 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Ben Murdoch  <benm@google.com>

        Reviewed by Eric Seidel.

        [V8] DOM Storage bindings: Event Handler should create StorageEvents
        https://bugs.webkit.org/show_bug.cgi?id=28942

        Tests:
        Causes the following tests that were failing with V8 to pass:
        * storage/domstorage/localstorage/onstorage-attribute-setwindow.html
        * storage/domstorage/localstorage/complex-keys.html
        * storage/domstorage/localstorage/onstorage-attribute-markup.html
        * storage/domstorage/localstorage/simple-events.html
        * storage/domstorage/localstorage/onstorage-attribute-set-attribute.html
        * storage/domstorage/localstorage/index-get-and-set.html
        * storage/domstorage/sessionstorage/onstorage-attribute-setwindow.html
        * storage/domstorage/sessionstorage/onstorage-attribute-markup.html
        * storage/domstorage/sessionstorage/simple-events.html
        * storage/domstorage/sessionstorage/onstorage-attribute-set-attribute.html
        * storage/domstorage/sessionstorage/index-get-and-set.html

        Code changes:
        * bindings/v8/V8DOMWrapper.cpp:
        (WebCore::V8DOMWrapper::convertEventToV8Object):  Create StorageEvents.
        * bindings/v8/custom/V8StorageCustom.cpp:
        (WebCore::storageGetter): Check if we are retrieving the length property and return the number of items in the object rather than a property named length when invoking the getItem() member of local/sessionStorage.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48047 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Timothy Hatcher.

        https://bugs.webkit.org/show_bug.cgi?id=28911

        * WebCore.gypi:
        * WebCore.vcproj/WebCore.vcproj:
        * inspector/InspectorBackend.cpp:
        (WebCore::InspectorBackend::dispatchOnInjectedScript):
        * inspector/InspectorBackend.h:
        * inspector/InspectorBackend.idl:
        * inspector/front-end/ConsoleView.js:
        (WebInspector.ConsoleView.prototype.doEvalInWindow):
        (WebInspector.ConsoleView.prototype._formatarray):
        (WebInspector.ConsoleView.prototype._formatnode):
        * inspector/front-end/DOMAgent.js:
        * inspector/front-end/ElementsPanel.js:
        (WebInspector.ElementsPanel.this.treeOutline.focusedNodeChanged):
        (WebInspector.ElementsPanel.prototype.searchCanceled):
        (WebInspector.ElementsPanel.prototype.performSearch):
        * inspector/front-end/InjectedScript.js:
        (InjectedScript.dispatch):
        * inspector/front-end/InjectedScriptAccess.js: Added.
        (InjectedScriptAccess._installHandler.InjectedScriptAccess.methodName):
        (InjectedScriptAccess._installHandler):
        * inspector/front-end/MetricsSidebarPane.js:
        (WebInspector.MetricsSidebarPane.prototype.update):
        (WebInspector.MetricsSidebarPane.prototype.editingCommitted):
        * inspector/front-end/ObjectPropertiesSection.js:
        (WebInspector.ObjectPropertiesSection.prototype.update):
        (WebInspector.ObjectPropertyTreeElement.prototype.onpopulate):
        (WebInspector.ObjectPropertyTreeElement.prototype.applyExpression):
        * inspector/front-end/PropertiesSidebarPane.js:
        (WebInspector.PropertiesSidebarPane.prototype.update):
        * inspector/front-end/ResourcesPanel.js:
        (WebInspector.ResourceSidebarTreeElement.prototype.ondblclick):
        * inspector/front-end/ScriptsPanel.js:
        (WebInspector.ScriptsPanel.prototype.doEvalInCallFrame):
        * inspector/front-end/StylesSidebarPane.js:
        (WebInspector.StylesSidebarPane.prototype.update):
        (WebInspector.StylePropertiesSection.prototype.editingSelectorCommitted):
        (WebInspector.BlankStylePropertiesSection.prototype.editingCommitted):
        (WebInspector.StylePropertyTreeElement.prototype):
        * inspector/front-end/inspector.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48046 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Tony Chang  <tony@chromium.org>

        Reviewed by Eric Seidel.

        Implement <input type="range"> for Chromium Linux by painting
        the slider and the thumb using skia.
        https://bugs.webkit.org/show_bug.cgi?id=28928

        Covered by existing tests (we're currently hitting an assert for them).

        * rendering/RenderThemeChromiumLinux.cpp:
        (WebCore::RenderThemeChromiumLinux::adjustSliderThumbSize):
        * rendering/RenderThemeChromiumLinux.h:
        * rendering/RenderThemeChromiumSkia.cpp:
        (WebCore::drawVertLine):
        (WebCore::drawHorizLine):
        (WebCore::drawBox):
        (WebCore::RenderThemeChromiumSkia::paintSliderTrack):
        (WebCore::RenderThemeChromiumSkia::paintSliderThumb):
        * rendering/RenderThemeChromiumSkia.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48045 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Albert Bachand  <albertb@google.com>

        Reviewed by Eric Seidel.

        Modify the condition regarding the key modifier for opening combo
        boxes under GTK from just PLATFORM(GTK) to also look for
        (PLATFORM(CHROMIUM) && PLATFORM(LINUX)).

        * WebCore/dom/SelectElement.cpp:
        * WebCore/manual-tests/select-popup-on-spacebar.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48044 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Kent Tamura  <tkent@chromium.org>

        Reviewed by Eric Seidel.

        Tests for <input type=color> validation.
        https://bugs.webkit.org/show_bug.cgi?id=28966

        * fast/forms/ValidityState-patternMismatch-unsupported-expected.txt: Added.
        * fast/forms/ValidityState-patternMismatch-unsupported.html: Added.
        * fast/forms/ValidityState-typeMismatch-color-expected.txt: Added.
        * fast/forms/ValidityState-typeMismatch-color.html: Added.
        * fast/forms/resources/ValidityState-patternMismatch-unsupported.js: Added.
        * fast/forms/resources/ValidityState-typeMismatch-color.js: Added.
2009-09-04  Kent Tamura  <tkent@chromium.org>

        Reviewed by Eric Seidel.

        - Implement <input type=color> as a text field.
        - ValidityState::typeMismatch supports for type=color.
        https://bugs.webkit.org/show_bug.cgi?id=28966

        Tests: fast/forms/ValidityState-patternMismatch-unsupported.html
               fast/forms/ValidityState-typeMismatch-color.html

        * html/HTMLInputElement.cpp:
        (WebCore::HTMLInputElement::valueMissing):
        (WebCore::HTMLInputElement::patternMismatch):
        (WebCore::HTMLInputElement::setInputType):
        (WebCore::HTMLInputElement::formControlType):
        (WebCore::HTMLInputElement::saveFormControlState):
        (WebCore::HTMLInputElement::restoreFormControlState):
        (WebCore::HTMLInputElement::accessKeyAction):
        (WebCore::HTMLInputElement::rendererIsNeeded):
        (WebCore::HTMLInputElement::createRenderer):
        (WebCore::HTMLInputElement::appendFormData):
        (WebCore::HTMLInputElement::valueWithDefault):
        (WebCore::HTMLInputElement::storesValueSeparateFromAttribute):
        (WebCore::HTMLInputElement::defaultEventHandler):
        (WebCore::HTMLInputElement::isRequiredFormControl):
        (WebCore::HTMLInputElement::dataList):
        * html/HTMLInputElement.h:
        (WebCore::HTMLInputElement::):
        (WebCore::HTMLInputElement::isTextField):
        * html/ValidityState.cpp:
        (WebCore::ValidityState::typeMismatch):
        (WebCore::ValidityState::isValidColorString):
        * html/ValidityState.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48043 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[V8] Crash when updating document cache
https://bugs.webkit.org/show_bug.cgi?id=28965

Patch by Mads Ager <ager@chromium.org> on 2009-09-04
Reviewed by David Levin.

Initialization of a V8 JavaScript context can fail.  When that
happens, do not attempt to update the document wrapper cache.

This code path is exercised by http/tests/security/javascriptURL/javascriptURL-in-new-iframe.html.

* bindings/v8/V8Proxy.cpp:
(WebCore::V8Proxy::updateDocument):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48042 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Laszlo Gombos  <laszlo.1.gombos@nokia.com>

        Reviewed by Eric Seidel.

        [Qt] Fix tst_QWebPage::database autotest failure
        https://bugs.webkit.org/show_bug.cgi?id=28961

        Make sure that the test case enables the feature before
        the feature gets tested

        * tests/qwebpage/tst_qwebpage.cpp:
        (tst_QWebPage::database):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48041 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Yaar Schnitman  <yaar@chromium.org>

        Reviewed by Eric Seidel.

        Two bugs:

        https://bugs.webkit.org/show_bug.cgi?id=25489
        CSS property background-position is now serialized in shorthand string as
        background-position-x background-position-y, as the specification requires
        (See http://www.w3.org/TR/2008/WD-css3-background-20080910/#the-background-position).

        https://bugs.webkit.org/show_bug.cgi?id=26541
        CSS background-color is now be serialized before other background properties,
        as CSS specification requires (See http://dev.w3.org/csswg/cssom/#css-value).

        An existing layout test that accepted wrong order and skipped position test
        had to be modified.

        * fast/dom/background-shorthand-csstext-expected.txt:
        * fast/dom/background-shorthand-csstext.html:
2009-09-04  Yaar Schnitman  <yaar@chromium.org>

        Reviewed by Eric Seidel.

        Two bugs:

        https://bugs.webkit.org/show_bug.cgi?id=25489
        CSS property background-position is now serialized in shorthand string as
        background-position-x background-position-y, as the specification requires
        (See http://www.w3.org/TR/2008/WD-css3-background-20080910/#the-background-position).

        https://bugs.webkit.org/show_bug.cgi?id=26541
        CSS background-color is now be serialized before other background properties,
        as CSS specification requires (See http://dev.w3.org/csswg/cssom/#css-value).

        An existing layout test that accepted wrong order and skipped position test
        had to be modified.

        * css/CSSMutableStyleDeclaration.cpp:
        (WebCore::CSSMutableStyleDeclaration::getPropertyValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48040 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-04  Benjamin C Meyer  <benjamin.meyer@torchmobile.com>

        Reviewed by Eric Seidel.

        Any QtWebKit application crashes on exit if the inspector is
        used during the session rather then exiting.

        If a QWebView has a QWebPage that is destroyed the QWebView
        does not update its pointer and will crash the next time
        it access the page pointers (such as in its destructor).

        InspectorClientView should not call deleteLater when the page
        is destroyed because it does not have a parent and is a top
        level widget.  close() needs to be called so that QApplication
        can exit the application if quitOnLastWindowClosed is set
        and a InspectorClientView is the last window, otherwise
        the application will never exit.

        * Api/qwebview.cpp:
        (QWebViewPrivate::_q_pageDestroyed):
        (QWebView::setPage):
        * Api/qwebview.h:
        * WebCoreSupport/InspectorClientQt.cpp:
        (WebCore::InspectorClientView::InspectorClientView):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48039 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Kent Tamura  <tkent@chromium.org>

        Reviewed by David Levin.

        Test for HTMLTextAreaElement.textLength
        https://bugs.webkit.org/show_bug.cgi?id=28929

        * fast/forms/resources/textarea-textlength.js: Added.
        * fast/forms/textarea-textlength-expected.txt: Added.
        * fast/forms/textarea-textlength.html: Added.
2009-09-03  Kent Tamura  <tkent@chromium.org>

        Reviewed by David Levin.

        Add support for HTMLTextAreaElement.textLength
        https://bugs.webkit.org/show_bug.cgi?id=28929

        Test: fast/forms/textarea-textlength.html

        * html/HTMLTextAreaElement.h:
        (WebCore::HTMLTextAreaElement::textLength):
        * html/HTMLTextAreaElement.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48038 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin Adler.

        <rdar://problem/7180197>, https://bugs.webkit.org/show_bug.cgi?id=28822
        REGRESSION(r31231): Creating document with current document's DOCTYPE causes crashes

        Test: fast/dom/DOMImplementation/createDocument-with-used-doctype.html

        * dom/DOMImplementation.cpp: (WebCore::DOMImplementation::createDocument): Even though we
        cannot raise an exception immediately out of fear of breaking Acid3, we shouldn't use a
        DocumentType node in two documents.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48037 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore: Page Cache should support pages with Frames.
https://bugs.webkit.org/show_bug.cgi?id=13631

Reviewed by Sam Weinig.

Tests: fast/events/pageshow-pagehide-on-back-cached-with-frames.html
       fast/loader/frames-with-unload-handlers-in-page-cache.html

* loader/FrameLoader.cpp:
(WebCore::FrameLoader::canCachePageContainingThisFrame):
(WebCore::FrameLoader::logCanCacheFrameDecision):

LayoutTests: Page Cache should support pages with Frames.
https://bugs.webkit.org/show_bug.cgi?id=13631

Reviewed by Sam Weinig.

* fast/events/pageshow-pagehide-on-back-cached-with-frames-expected.txt: Added.
* fast/events/pageshow-pagehide-on-back-cached-with-frames.html: Added.
* fast/events/resources/pageshow-pagehide-subframe-cachable.html: Added.
* fast/events/resources/pageshow-pagehide-subsubframe-cachable.html: Added.
* fast/loader/frames-with-unload-handlers-in-page-cache-expected.txt: Added.
* fast/loader/frames-with-unload-handlers-in-page-cache.html: Added.
* fast/loader/resources/notify-done.html: Added.
* fast/loader/resources/subframe-with-subsubframe-with-unload-handler.html: Added.
* fast/loader/resources/subframe-with-unload-handler.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48036 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Steve VanDeBogart  <vandebo@chromium.org>

        Reviewed by Eric Seidel.

        Check that middle click triggers an onpaste event in Linux.
        https://bugs.webkit.org/show_bug.cgi?id=28696

        * platform/gtk/editing/pasteboard: Added.
        * platform/gtk/editing/pasteboard/middle-click-onpaste-expected.txt: Added.
        * platform/gtk/editing/pasteboard/resources: Added.
        * platform/gtk/editing/pasteboard/resources/TEMPLATE.html: Added.
        * platform/gtk/editing/pasteboard/resources/middle-click-onpaste.js: Added.
        (handlePaste):
2009-09-03  Steve VanDeBogart  <vandebo@chromium.org>

        Reviewed by Eric Seidel.

        Handle middle click in Chromium like QT
        https://bugs.webkit.org/show_bug.cgi?id=28696

        Tested by middle-click-onpaste.html.

        * platform/Pasteboard.h:
        * platform/chromium/ChromiumBridge.h:
        * platform/chromium/ClipboardChromium.cpp:
        (WebCore::ClipboardChromium::getData):
        * platform/chromium/PasteboardChromium.cpp:
        (WebCore::Pasteboard::Pasteboard):
        (WebCore::Pasteboard::isSelectionMode):
        (WebCore::Pasteboard::setSelectionMode):
        (WebCore::Pasteboard::canSmartReplace):
        (WebCore::Pasteboard::plainText):
        (WebCore::Pasteboard::documentFragment):
        * platform/chromium/PasteboardPrivate.h:
        (WebCore::PasteboardPrivate::):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48035 268f45cc-cd09-0410-ab3c-d52691b4dbfc

More partial work towards "Page Cache should support pages with Frames"
https://bugs.webkit.org/show_bug.cgi?id=13631

Reviewed by Sam Weinig.

Fix a bug where subframes weren't asked if they were cacheable.

No new tests, especially since page caching for frames isn't turned on yet.
A layout test will come in my next patch, which enables frames in the page cache.

* loader/FrameLoader.cpp:
(WebCore::FrameLoader::canCachePageContainingThisFrame):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48034 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Adam Barth  <abarth@webkit.org>

        Unreviewed build fix.

        Change notImplemented() to a FIXME.

        * WebCoreSupport/WebFrameLoaderClient.mm:
        (WebFrameLoaderClient::didDisplayInsecureContent):
        (WebFrameLoaderClient::didRunInsecureContent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48033 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Adam Barth  <abarth@webkit.org>

        Reviewed by eric@webkit.org.

        https://bugs.webkit.org/show_bug.cgi?id=24696

        Added mixed content methods to FrameLoaderClient.

        * loader/EmptyClients.h:
        (WebCore::EmptyFrameLoaderClient::didDisplayInsecureContent):
        (WebCore::EmptyFrameLoaderClient::didRunInsecureContent):
        * loader/FrameLoaderClient.h:
2009-09-03  Adam Barth  <abarth@webkit.org>

        Reviewed by eric@webkit.org.

        https://bugs.webkit.org/show_bug.cgi?id=24696

        Stub implementations of mixed content methods of FrameLoaderClient.

        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        (WebKit::FrameLoaderClient::didDisplayInsecureContent):
        (WebKit::FrameLoaderClient::didRunInsecureContent):
        * WebCoreSupport/FrameLoaderClientGtk.h:
2009-09-03  Adam Barth  <abarth@webkit.org>

        Reviewed by eric@webkit.org.

        https://bugs.webkit.org/show_bug.cgi?id=24696

        Stub implementations of mixed content methods of FrameLoaderClient.

        * WebCoreSupport/WebFrameLoaderClient.h:
        * WebCoreSupport/WebFrameLoaderClient.mm:
        (WebFrameLoaderClient::didDisplayInsecureContent):
        (WebFrameLoaderClient::didRunInsecureContent):
2009-09-03  Adam Barth  <abarth@webkit.org>

        Reviewed by eric@webkit.org.

        https://bugs.webkit.org/show_bug.cgi?id=24696

        Stub implementations of mixed content methods of FrameLoaderClient.

        * WebCoreSupport/FrameLoaderClientQt.cpp:
        (WebCore::FrameLoaderClientQt::didDisplayInsecureContent):
        (WebCore::FrameLoaderClientQt::didRunInsecureContent):
        * WebCoreSupport/FrameLoaderClientQt.h:
2009-09-03  Adam Barth  <abarth@webkit.org>

        Reviewed by eric@webkit.org.

        https://bugs.webkit.org/show_bug.cgi?id=24696

        Stub implementations of mixed content methods of FrameLoaderClient.

        * WebCoreSupport/WebFrameLoaderClient.cpp:
        (WebFrameLoaderClient::didDisplayInsecureContent):
        (WebFrameLoaderClient::didRunInsecureContent):
        * WebCoreSupport/WebFrameLoaderClient.h:
2009-09-03  Adam Barth  <abarth@webkit.org>

        Reviewed by eric@webkit.org.

        https://bugs.webkit.org/show_bug.cgi?id=24696

        Stub implementations of mixed content methods of FrameLoaderClient.

        * WebKitSupport/FrameLoaderClientWx.cpp:
        (WebCore::FrameLoaderClientWx::didDisplayInsecureContent):
        (WebCore::FrameLoaderClientWx::didRunInsecureContent):
        * WebKitSupport/FrameLoaderClientWx.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48032 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        Access private member variables directly instead of using the
        accessor.

        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        (WebKit::FrameLoaderClient::createFrame):
        (WebKit::FrameLoaderClient::windowObjectCleared):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48031 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Exception.toString() on Qt seems to print out a different value than other platforms.
https://bugs.webkit.org/show_bug.cgi?id=28952

Reviewed by Dave Levin.

* platform/qt/Skipped:
Disabled new worker tests that rely on the string format of DOM exceptions until bug #28952 is resolved.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48030 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Reviewed by Kevin Ollivier.

[wx] Frames support
https://bugs.webkit.org/show_bug.cgi?id=19041

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48029 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Page content is not displayed in case of HTTP status error
https://bugs.webkit.org/show_bug.cgi?id=28949

Patch by Yael Aharon <yael.aharon@nokia.com> on 2009-09-03
Reviewed by Simon Hausmann.

If the server sent content, show it to the user instead of an error,
even if the HTTP status code is not success.

* platform/network/qt/QNetworkReplyHandler.cpp:
* platform/network/qt/QNetworkReplyHandler.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48028 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    <webkit.org/b/28880> svn-apply --force doesn't actually work

        Reviewed by Eric Seidel.

        This fixes "svn-apply --force" and adds unit tests for the
        scm.apply_patch() method which uses this script.

        * Scripts/svn-apply: Created $globalExitCode variable that
        defaults to 0.  Exit with a value of $globalExitCode when the
        script is finished.
        (applyPatch): Ignore a non-zero $exitCode if $force is true, but
        set $globalExitCode to $exitCode so that svn-apply exits with a
        non-zero status if any patches did not apply cleanly.  Also
        print out the actual patch command if $force was not true.

        * Scripts/modules/scm.py:
        (scripts_directory): Added.  Extracted from script_path().
        (script_path): Extracted scripts_directory().
        * Scripts/modules/scm_unittest.py: Import urllib.
        (SVNTestRepository.setup): Save the original working directory
        in test_object since this represents the WebKit repository from
        where the unit tests are run.
        (SCMTest): Created new super class to hold utility methods.
        (SCMTest._create_patch): Creates a patch file on disk and a
        dictionary for use with scm.svn_apply().
        (SCMTest._setup_webkittools_scripts_symlink): Sets up a symlink
        back to WebKitTools/Scripts in the test repository so that
        scm.apply_patch() is able to find the svn-apply script.
        (SVNTest): Inherit from SCMTest instead of unittest.TestCase.
        (SVNTest.tearDown): Make sure to change directories back to the
        original_path before the next test.
        (SVNTest.test_apply_svn_patch): New test case for applying an
        svn patch with scm.apply_patch().
        (SVNTest.test_apply_svn_patch_force): New test case for applying
        an svn patch with scm.apply_patch() that conflicts.
        (GitTest): Inherit from SCMTest instead of unittest.TestCase.
        (GitTest.tearDown): Make sure to change directories back to the
        original_path before the next test.
        (GitTest.test_apply_git_patch): New test case for applying a git
        patch with scm.apply_patch().
        (GitTest.test_apply_git_patch_force): New test case for applying
        a git patch with scm.apply_patch() that conflicts.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48027 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Need to update v8 bindings to support passing multiple ports to postMessage()
https://bugs.webkit.org/show_bug.cgi?id=28839

Reviewed by Dimitri Glazkov.

Added support for passing MessagePortArray to v8 bindings.

New tests that now pass with V8:
       fast/events/message-port-multi.html
       fast/workers/worker-context-multi-port.html
       fast/workers/worker-multi-port.html

* WebCore.gypi:
Added V8MessageEventCustom.cpp and V8MessagePortCustom.h
* bindings/v8/custom/V8CustomBinding.h:
Added custom postMessage() handlers for various classes.
* bindings/v8/custom/V8DOMWindowCustom.cpp:
(WebCore::DOMWindowPostMessage):
Added custom handler that supports MessagePortArray.
* bindings/v8/custom/V8DedicatedWorkerContextCustom.cpp:
(WebCore::DedicatedWorkerContextPostMessage):
Added custom handler that supports MessagePortArray.
* bindings/v8/custom/V8MessageEventCustom.cpp: Added.
(WebCore::MessageEventPorts):
Added getter for ports attribute.
(WebCore::MessageEventInitMessageEvent):
Added custom handler that supports MessagePortArray.
* bindings/v8/custom/V8MessagePortCustom.cpp:
(WebCore::MessagePortPostMessage):
Added custom handler that supports MessagePortArray.
(WebCore::getMessagePortArray):
Added helper function that supports converting from a sequence-like object to a MessagePortArray.
* bindings/v8/custom/V8MessagePortCustom.h: Added.
Added declaration for getMessagePortArray().
* bindings/v8/custom/V8WorkerCustom.cpp:
(WebCore::WorkerPostMessage):
Added custom handler that supports MessagePortArray.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48026 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore: Need to update JS bindings and IDL files to support multiple message ports in postMessage()
https://bugs.webkit.org/show_bug.cgi?id=28460

Reviewed by Sam Weinig.

Added new toJSSequence() API which validates that a JSValue meets the WebIDL criteria for a sequence.

Tests: fast/events/message-port-multi.html
       fast/workers/worker-context-multi-port.html
       fast/workers/worker-multi-port.html

* GNUmakefile.am:
* WebCore.gypi:
* WebCore.pro:
* WebCore.vcproj/WebCore.vcproj:
* WebCore.xcodeproj/project.pbxproj:
* WebCoreSources.bkl:
Added JSMessagePortCustom.h and JSMessageEventCustom.cpp.
* bindings/js/JSDOMBinding.cpp:
(WebCore::toJSSequence):
Added toJSSequence() API to do validation on sequence-like objects per WebIDL.
* bindings/js/JSDOMBinding.h:
* bindings/js/JSDOMWindowCustom.cpp:
(WebCore::JSDOMWindow::postMessage):
Added custom handler for postMessage() that handles being passed MessagePortArrays.
* bindings/js/JSDedicatedWorkerContextCustom.cpp:
(WebCore::JSDedicatedWorkerContext::postMessage):
Added custom handler for postMessage() that handles being passed MessagePortArrays.
* bindings/js/JSMessageEventCustom.cpp: Added.
(WebCore::JSMessageEvent::ports):
Added custom ports() getter that converts from MessagePortArray to JSArray.
(WebCore::JSMessageEvent::initMessageEvent):
Added support for passing a MessagePortArray.
* bindings/js/JSMessagePortCustom.cpp:
(WebCore::JSMessagePort::postMessage):
Added custom handler for postMessage() that handles being passed MessagePortArray.
(WebCore::fillMessagePortArray):
New helper routine that validates/converts from a JS sequence o a WebCore::MessagePortArray.
* bindings/js/JSMessagePortCustom.h: Added.
* bindings/js/JSWorkerCustom.cpp:
(WebCore::JSWorker::postMessage):
Added custom handler for postMessage() that handles being passed MessagePortArray.
* dom/MessageEvent.cpp:
* dom/MessageEvent.h:
(WebCore::MessageEvent::ports):
Changed ports() to return a MessagePortArray* since this value can be null.
* dom/MessageEvent.idl:
Updated IDL to match HTML5 spec (now accepts MessagePortArrays instead of a solitary MessagePort).
* dom/MessagePort.cpp:
* dom/MessagePort.h:
* dom/MessagePort.idl:
Updated IDL to match HTML5 spec (postMessage() now accepts MessagePortArrays instead of a solitary MessagePort).
* page/DOMWindow.h:
* page/DOMWindow.idl:
Updated IDL to match HTML5 spec (postMessage() now accepts MessagePortArrays instead of a solitary MessagePort).
* workers/DedicatedWorkerContext.cpp:
* workers/DedicatedWorkerContext.h:
* workers/DedicatedWorkerContext.idl:
Updated IDL to match HTML5 spec (postMessage() now accepts MessagePortArrays instead of a solitary MessagePort).
* workers/Worker.cpp:
* workers/Worker.h:
* workers/Worker.idl:
Updated IDL to match HTML5 spec (postMessage() now accepts MessagePortArrays instead of a solitary MessagePort).

LayoutTests: Need to update JS bindings and IDL files to support multiple message ports in postMessage()
https://bugs.webkit.org/show_bug.cgi?id=28460

Reviewed by Sam Weinig.

Updated layout tests to match new postMessage() and MessageEvent APIs.

Added new layout tests to test sending/receiving multiple ports.

* fast/dom/Window/window-postmessage-args-expected.txt:
* fast/dom/Window/window-postmessage-args.html:
Updated to match new postMessage API.
* fast/events/init-events-expected.txt:
* fast/events/message-channel-gc-4.html:
Updated to match new postMessage API.
* fast/events/message-port-clone.html:
Updated to match new postMessage API.
* fast/events/message-port-deleted-document.html:
Updated to match new postMessage API.
* fast/events/message-port-deleted-frame.html:
Updated to match new postMessage API.
* fast/events/message-port-inactive-document.html:
Updated to match new postMessage API.
* fast/events/message-port-multi-expected.txt: Added.
* fast/events/message-port-multi.html: Added.
Added tests for various cases of sending multiple ports.
* fast/events/message-port.html:
Updated to match new postMessage API.
* fast/events/resources/init-events.js:
Added tests for sending ports to initMessageEvent()
* fast/events/resources/message-port-iframe.html:
* fast/events/resources/message-port-multi.js: Added.
Added tests for various cases of sending multiple ports.
* fast/workers/resources/shared-worker-common.js:
(onconnect):
Updated to reflect new MessageEvent API.
* fast/workers/resources/shared-worker-script-error.js:
(onconnect):
Updated to reflect new MessageEvent API.
* fast/workers/resources/worker-cloneport.js:
(onmessage):
Updated to reflect new MessageEvent API.
* fast/workers/resources/worker-context-multi-port.js: Added.
(worker.onmessage):
Added tests for various cases of sending multiple ports.
* fast/workers/resources/worker-context-thread-multi-port.js: Added.
Added tests for various cases of sending multiple ports.
* fast/workers/resources/worker-messageport.js:
(onmessage):
Updated to reflect new MessageEvent API.
* fast/workers/resources/worker-multi-port.js: Added.
(worker.onmessage):
Added tests for various cases of sending multiple ports.
* fast/workers/resources/worker-thread-multi-port.js: Added.
(onmessage):
Worker-thread side of multiple port layout test.
(testPassed):
(testFailed):
* fast/workers/worker-cloneport.html:
* fast/workers/worker-context-multi-port-expected.txt: Added.
* fast/workers/worker-context-multi-port.html: Added.
Added tests for sending multiple ports from the worker thread.
* fast/workers/worker-messageport-expected.txt:
* fast/workers/worker-messageport-gc.html:
* fast/workers/worker-messageport.html:
* fast/workers/worker-multi-port-expected.txt: Added.
* fast/workers/worker-multi-port.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48025 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Timothy Hatcher.

        WebInspector: Do not query non-function call frames for function name.

        https://bugs.webkit.org/show_bug.cgi?id=28799

        * inspector/front-end/InjectedScript.js:
        (InjectedScript.CallFrameProxy):
        (InjectedScript.CallFrameProxy.prototype._wrapScopeChain):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48024 268f45cc-cd09-0410-ab3c-d52691b4dbfc

* StringsNotToBeLocalized.txt:
Brought this file up to date.

Reviewed by Adam Roben

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48023 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Timothy Hatcher.

        Crash after typing "clear" and pressing return in inspector console.

        https://bugs.webkit.org/show_bug.cgi?id=28684

        * inspector/front-end/InjectedScript.js:
        (InjectedScript._clearConsoleMessages):
        (InjectedScript._inspectObject):
        (InjectedScript._ensureCommandLineAPIInstalled):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48022 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Skip a couple of failing layout-tests

Rubber-stamped by Ariya Hidayat.

We skip most of the D&D tests already, and the animation test
seems to be flakey on our platform since the commit that triggered
it only touched V8 bindings.

* platform/qt/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48021 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Vitaly Repeshko  <vitalyr@chromium.org>

        Reviewed by Adam Barth.

        [V8] Squeezed out some performance from string conversion.
        https://bugs.webkit.org/show_bug.cgi?id=28945

        * bindings/v8/V8Binding.cpp:
        (WebCore::WebCoreStringResource::WebCoreStringResource):
        (WebCore::WebCoreStringResource::~WebCoreStringResource):
        (WebCore::WebCoreStringResource::data):
        (WebCore::WebCoreStringResource::length):
        (WebCore::WebCoreStringResource::atomicString):
        (WebCore::WebCoreStringResource::string):
        Avoid calling virtual length method. Be more explicit with
        external memory computation. Avoid function calls in data and
        length.
        (WebCore::):
        (WebCore::v8StringToWebCoreString):
        (WebCore::v8StringToAtomicWebCoreString):
        (WebCore::v8NonStringValueToWebCoreString):
        (WebCore::v8NonStringValueToAtomicWebCoreString):
        Avoid extra calls of WebCoreStringResource::toString. When
        AtomicString is computed pass it immediately when constructing
        WebCoreStringResource. Use inline buffer for small strings.
        * bindings/v8/V8Binding.h:
        (WebCore::v8ValueToWebCoreString):
        (WebCore::v8ValueToAtomicWebCoreString):
        (WebCore::toString):
        Inline dispatch of string vs. non-string values.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48020 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Vitaly Repeshko  <vitalyr@chromium.org>

        Reviewed by Adam Barth.

        [V8] Inlined common case of V8IsolatedWorld::getEntered to speed
        up V8Proxy lookup.
        https://bugs.webkit.org/show_bug.cgi?id=28946

        * bindings/v8/V8IsolatedWorld.cpp:
        (WebCore::V8IsolatedWorld::getEnteredImpl):
        * bindings/v8/V8IsolatedWorld.h:
        (WebCore::V8IsolatedWorld::getEntered):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48019 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Reviewed by Xan Lopez.

        Remove bogus asserts that break Debug builds at runtime.

        * WebCoreSupport/DocumentLoaderGtk.cpp:
        (WebKit::DocumentLoader::detachFromFrame):
        (WebKit::DocumentLoader::decreaseLoadCount):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48018 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Renata Hodovan  <hodovan.renata@stud.u-szeged.hu>

        Reviewed by Ariya Hidayat.

        [Qt] Update outdated metrics in css2.1/t170602-bdr-conflct-w-* (50 files).
        Remove these files from Skipped.

        * platform/qt/Skipped:
        * platform/qt/css2.1/t170602-bdr-conflct-w-00-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-01-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-02-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-03-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-04-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-05-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-06-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-07-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-08-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-09-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-10-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-11-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-12-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-13-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-14-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-15-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-16-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-17-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-18-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-19-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-20-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-21-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-22-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-23-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-24-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-25-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-26-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-27-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-28-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-29-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-30-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-31-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-32-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-33-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-34-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-35-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-36-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-37-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-38-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-39-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-40-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-41-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-42-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-43-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-44-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-45-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-46-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-47-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-48-d-expected.txt:
        * platform/qt/css2.1/t170602-bdr-conflct-w-49-d-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48017 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Renata Hodovan  <hodovan.renata@stud.u-szeged.hu>

        Reviewed by Ariya Hidayat.

        [Qt] Update outdated metrics in css2.1/t1202-counters-* (17 files).
        Remove these files from Skipped.

        * platform/qt/Skipped:
        * platform/qt/css2.1/t1202-counters-00-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-01-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-02-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-03-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-05-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-06-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-07-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-08-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-09-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-11-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-12-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-13-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-14-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-15-b-expected.txt:
        * platform/qt/css2.1/t1202-counters-16-c-expected.txt:
        * platform/qt/css2.1/t1202-counters-17-d-expected.txt:
        * platform/qt/css2.1/t1202-counters-18-f-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48016 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Disable a few tests that require new controller features

Reviewed by Ariya Hidayat.

* platform/qt/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48015 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt/Linux] Add platform-spesific result for plugin-javascript-access

Rubber-stamped by Simon Hausmann..

* platform/qt-linux/plugins/plugin-javascript-access-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48014 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Renata Hodovan  <hodovan.renata@stud.u-szeged.hu>

        Reviewed by Ariya Hidayat.

        [Qt] Update outdated metrics in css2.1/t140201* (16 files).
        Remove these files from Skipped.

        * platform/qt/Skipped:
        * platform/qt/css2.1/t140201-c532-bgcolor-00-a-expected.txt:
        * platform/qt/css2.1/t140201-c532-bgcolor-01-b-expected.txt:
        * platform/qt/css2.1/t140201-c533-bgimage-00-a-expected.txt:
        * platform/qt/css2.1/t140201-c533-bgimage-01-b-g-expected.txt:
        * platform/qt/css2.1/t140201-c534-bgre-00-b-ag-expected.txt:
        * platform/qt/css2.1/t140201-c534-bgre-01-b-ag-expected.txt:
        * platform/qt/css2.1/t140201-c534-bgreps-00-c-ag-expected.txt:
        * platform/qt/css2.1/t140201-c534-bgreps-01-c-ag-expected.txt:
        * platform/qt/css2.1/t140201-c534-bgreps-02-c-ag-expected.txt:
        * platform/qt/css2.1/t140201-c534-bgreps-03-c-ag-expected.txt:
        * platform/qt/css2.1/t140201-c534-bgreps-04-c-ag-expected.txt:
        * platform/qt/css2.1/t140201-c534-bgreps-05-c-ag-expected.txt:
        * platform/qt/css2.1/t140201-c535-bg-fixd-00-b-g-expected.txt:
        * platform/qt/css2.1/t140201-c536-bgpos-00-b-ag-expected.txt:
        * platform/qt/css2.1/t140201-c536-bgpos-01-b-ag-expected.txt:
        * platform/qt/css2.1/t140201-c537-bgfxps-00-c-ag-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48013 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Fumitoshi Ukai  <ukai@chromium.org>

        Reviewed by Eric Seidel.

        Add strnstr for Linux and Windows in StringExtras.h
        https://bugs.webkit.org/show_bug.cgi?id=28901

        * wtf/StringExtras.h:
        (strnstr):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48012 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Kent Tamura  <tkent@chromium.org>

        Reviewed by Eric Seidel.

        Tests for maxLength and IME input.
        https://bugs.webkit.org/show_bug.cgi?id=25253

        * fast/forms/input-maxlength-ime-completed-expected.txt: Added.
        * fast/forms/input-maxlength-ime-completed.html: Added.
        * fast/forms/input-maxlength-ime-preedit-expected.txt: Added.
        * fast/forms/input-maxlength-ime-preedit.html: Added.
        * fast/forms/resources/input-maxlength-ime-completed.js: Added.
        * fast/forms/resources/input-maxlength-ime-preedit.js: Added.
        * platform/gtk/Skipped:
        * platform/win/Skipped:
2009-09-03  Kent Tamura  <tkent@chromium.org>

        Reviewed by Eric Seidel.

        Fix a regression bug that maxLength doesn't work for IME input.
        https://bugs.webkit.org/show_bug.cgi?id=25253

        Tests: fast/forms/input-maxlength-ime-completed.html
               fast/forms/input-maxlength-ime-preedit.html

        * dom/InputElement.cpp:
        (WebCore::InputElement::handleBeforeTextInsertedEvent):
        * dom/InputElement.h:
        * html/HTMLInputElement.cpp:
        (WebCore::HTMLInputElement::defaultEventHandler):
        * wml/WMLInputElement.cpp:
        (WebCore::WMLInputElement::defaultEventHandler):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48011 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Cameron McCormack  <cam@mcc.id.au>

        Reviewed by Eric Seidel.

        Make SVGStyleElement title/media/type setters do something
        https://bugs.webkit.org/show_bug.cgi?id=28828

        * svg/dom/resources/style-reflect.js: Added.
        * svg/dom/style-reflect-expected.txt: Added.
        * svg/dom/style-reflect.html: Added.
2009-09-03  Cameron McCormack  <cam@mcc.id.au>

        Reviewed by Eric Seidel.

        Make SVGStyleElement title/media/type setters do something
        https://bugs.webkit.org/show_bug.cgi?id=28828

        Test: svg/dom/style-reflect.html

        * svg/SVGStyleElement.cpp:
        (WebCore::SVGStyleElement::setXmlspace): Actually change the attribute.
        (WebCore::SVGStyleElement::setType): Ditto.
        (WebCore::SVGStyleElement::setMedia): Ditto.
        (WebCore::SVGStyleElement::setTitle): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48010 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Add support for platform-spesific layout-test results

Reviewed by Mark Rowe.

For the Qt port we use the qt-[mac|linux|win] directories and then fall
back to the generic qt directory for both test results and skipped list.

* Scripts/run-webkit-tests:
* Scripts/webkitdirs.pm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48009 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Ben Murdoch  <benm@google.com>

        https://bugs.webkit.org/show_bug.cgi?id=28872
        Test to verify statement error callbacks behave as expected when they return true or throw an exception.

        * storage/statement-error-callback-expected.txt: Added with properties svn:eol-style.
        * storage/statement-error-callback.html: Added with properties svn:eol-style.
2009-09-03  Ben Murdoch  <benm@google.com>

        https://bugs.webkit.org/show_bug.cgi?id=28872

        Fixes a bug where handleEvent() in the V8 Custom SQL Statement Error binding would not return the correct result to WebCore after invoking the callback.

        Test: storage/statement-error-callback.html

        * bindings/v8/custom/V8CustomSQLStatementErrorCallback.cpp:
        (WebCore::V8CustomSQLStatementErrorCallback::handleEvent): Return true from handleEvent() if the callback raised an exception or the callback itself returned true.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48008 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Laszlo Gombos  <laszlo.1.gombos@nokia.com>

        Reviewed by Eric Seidel.

        [Qt] Add a setting to turn SessionStorage on/off
        https://bugs.webkit.org/show_bug.cgi?id=28836

        * Api/qwebsettings.cpp:
        (QWebSettingsPrivate::apply):
        (QWebSettings::QWebSettings):
        * Api/qwebsettings.h:
        * tests/qwebpage/tst_qwebpage.cpp:
        (testFlag):
        (tst_QWebPage::testOptionalJSObjects):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48007 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Maxime Simon  <simon.maxime@gmail.com>

        Reviewed by Eric Seidel.

        [Haiku] Adding three font-specific files to WebCore:
        FontCustomPlatformData.cpp, FontCustomPlatformData.h,
        and FontPlatformData.h
        https://bugs.webkit.org/show_bug.cgi?id=28131

        * platform/graphics/haiku/FontCustomPlatformData.cpp: Added.
        (WebCore::FontCustomPlatformData::~FontCustomPlatformData):
        (WebCore::FontCustomPlatformData::fontPlatformData):
        (WebCore::createFontCustomPlatformData):
        * platform/graphics/haiku/FontCustomPlatformData.h: Added.
        (WebCore::FontCustomPlatformData::FontCustomPlatformData):
        * platform/graphics/haiku/FontPlatformData.h: Added.
        (WebCore::FontPlatformData::FontPlatformData):
        (WebCore::FontPlatformData::font):
        (WebCore::FontPlatformData::size):
        (WebCore::FontPlatformData::bold):
        (WebCore::FontPlatformData::oblique):
        (WebCore::FontPlatformData::hashTableDeletedFontValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48006 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-03  Zoltan Horvath  <hzoltan@inf.u-szeged.hu>

        Reviewed by Darin Adler.

        Allow custom memory allocation control for JavaScriptCore's HashEntry class
        https://bugs.webkit.org/show_bug.cgi?id=27830

        Inherits HashEntry class from FastAllocBase because it has been
        instantiated by 'new' JavaScriptCore/runtime/Lookup.cpp:32.

        * runtime/Lookup.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48005 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-02  Laurent Cerveau  <lcerveau@me.com>

        Reviewed by David Kilzer.

        <http://webkit.org/b/25517> build-webkit script should print build time at end

        * Scripts/build-webkit:
        Added startTime and endTime variable so that the build time is computed and printed as
        part of the build message; display formatting has been separated in a dedicated subroutine.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48004 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    <webkit.org/b/28881> svn-create-patch should check if the repo path is the same when trying to find the root

        Reviewed by Eric Seidel.

        * Scripts/VCSUtils.pm:
        (determineSvnRoot): Added back check for repository root that
        was removed in r46134 when this code lived in svn-create-patch.
        It's necessary to check both the repository root and the
        repository UUID in case two different working directories are
        checked out from the same repository.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48003 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Use new 512x512 icons for nightly builds.

Rubber-stamped by Mark Rowe.

* WebKitLauncher/webkit.icns:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48002 268f45cc-cd09-0410-ab3c-d52691b4dbfc

More partial work towards "Page Cache should support pages with Frames"
https://bugs.webkit.org/show_bug.cgi?id=13631

Reviewed by Darin Adler.

No new tests. (No change in behavior, current tests pass).

* loader/DocumentLoader.cpp:
(WebCore::DocumentLoader::stopLoading): Adopt the new enum form for FrameLoader::stopLoading().

* loader/FrameLoader.cpp:
(WebCore::FrameLoader::stopLoading): Take an enum to decide whether to dispatch no unload events,
  only unload, or unload and pagehide.
(WebCore::FrameLoader::closeURL):
(WebCore::FrameLoader::scheduleRedirection):
(WebCore::FrameLoader::cachePageForHistoryItem): Call pageHidden().
(WebCore::FrameLoader::pageHidden): Add to dispatch the pagehide event to all frames in the case
  where a page is added to the PageCache.
* loader/FrameLoader.h:

* loader/FrameLoaderTypes.h: Add an UnloadEventPolicy enum.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48001 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Should crash if JIT code buffer allocation fails.

Patch by Gavin Barraclough <barraclough@apple.com> on 2009-09-02
https://bugs.webkit.org/show_bug.cgi?id=28926
<rdar://problem/7031922>

* jit/ExecutableAllocatorPosix.cpp:
(JSC::ExecutablePool::systemAlloc):
* jit/ExecutableAllocatorWin.cpp:
(JSC::ExecutablePool::systemAlloc):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@48000 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore: More partial work towards "Page Cache should support pages with Frames"
https://bugs.webkit.org/show_bug.cgi?id=13631

Reviewed by Darin Adler.

- More CachedFrame destruction work.
- Including related Frame and FrameLoader changes and cleanup.

No new tests. (No change in behavior, current tests pass).

* history/CachedFrame.cpp:
(WebCore::CachedFrameBase::restore): Rebuild up the whole FrameTree when restoring.
(WebCore::CachedFrame::CachedFrame): Tear down the entire FrameTree when caching.
(WebCore::CachedFrame::destroy): For child frames that are being destroyed while in
  the PageCache, do lightweight cleanup to:
  A - Prevent referring to a stale Page object.
  B - Prevent World Leaks of WebKit objects.

* loader/FrameLoader.cpp:
(WebCore::FrameLoader::stopAllLoaders): stopAllLoaders() should never be called on Frames that are
  in the PageCache. ASSERT that fact.
(WebCore::FrameLoader::open): Remove a bogus release-build workaround, as the comment right above it
  explains it is not effective, and it'd be better to see the crash locally instead of downstream.
(WebCore::FrameLoader::closeAndRemoveChild): Make sure the owner element has a page pointer before
  referencing it.
(WebCore::FrameLoader::detachFromParent): Use the new detachViewsAndDocumentLoader() call
(WebCore::FrameLoader::detachViewsAndDocumentLoader): Does lightweight cleanup of client (WebKit) objects.
* loader/FrameLoader.h:

* page/Frame.cpp:
(WebCore::Frame::detachFromPage): Simply clear the Page pointer without doing any other work.
(WebCore::Frame::pageDestroyed): Use the new detachFromPage().
* page/Frame.h:

WebKit/mac: More partial work towards "Page Cache should support pages with Frames"
https://bugs.webkit.org/show_bug.cgi?id=13631

Reviewed by Darin Adler.

* WebView/WebHTMLView.mm:
(-[WebHTMLView _topHTMLView]): Rework the ASSERT in this method to reflect the
  reality of calling this method for pages currently in the PageCache.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47999 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-02  Jungshik Shin  <jshin@chromium.org>

        Reviewed by Dimitri Glazkov

        [Chromium] Add 'icu::' qualifier in preparation for ICU upgrade to 4.2
        in Chromium. This is to fix a new file added in r28742 since the
        last patch for this issue.

        https://bugs.webkit.org/show_bug.cgi?id=28410

        No change in the test result.

        * platform/graphics/chromium/FontLinux.cpp:
        (WebCore::TextRunWalker::getNormalizedTextRun):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47998 268f45cc-cd09-0410-ab3c-d52691b4dbfc

waf build fix. Remove local variable shadowing global.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47997 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-02  Dimitri Glazkov  <dglazkov@chromium.org>

        Unreviewed, build fix.

        [Chromium] Correct a typo to fix bustage, caused by r47991.
        https://bugs.webkit.org/show_bug.cgi?id=27933

        * platform/graphics/skia/ImageBufferSkia.cpp: Changed "Unpremultipled" to "Unmultiplied".

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47996 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-02  Dimitri Glazkov  <dglazkov@chromium.org>

        Unreviewed, build fix.

        [Chromium] Add a missing include to fix bustage, caused by r47991.
        https://bugs.webkit.org/show_bug.cgi?id=27933

        * platform/graphics/skia/ImageBufferSkia.cpp: Added SkColorPriv include.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47995 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Chromium] Implement file support for DragData.
https://bugs.webkit.org/show_bug.cgi?id=28896

Reviewed by Dimitri Glazkov.

* platform/chromium/DragDataChromium.cpp:
(WebCore::DragData::containsURL):
(WebCore::DragData::asURL):
(WebCore::DragData::containsCompatibleContent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47994 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Reviewed by Kevin Ollivier.

Protect libcurl shared data with Mutex objects via curl_set_share_opt.
https://bugs.webkit.org/show_bug.cgi?id=28920

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47993 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-02  Evan Stade  <estade@chromium.org>

        Reviewed by Eric Seidel.

        Expose functions to change the focus ring color for Linux Chromium

        https://bugs.webkit.org/show_bug.cgi?id=28887

        * rendering/RenderThemeChromiumLinux.cpp:
        (WebCore::RenderThemeChromiumLinux::RenderThemeChromiumLinux):
        (WebCore::RenderThemeChromiumLinux::setFocusRingColor):
        (WebCore::RenderThemeChromiumLinux::platformFocusRingColor):
        * rendering/RenderThemeChromiumLinux.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47992 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2009-09-02  Dirk Schulze  <krit@webkit.org>

        Reviewed by Eric Seidel.

        SVG Filter premultiplied color support for getImageDate/putImageData
        [https://bugs.webkit.org/show_bug.cgi?id=27933]

        Patch to get premultiplied color support for Skia on getImageDate/putImageData.

        * platform/graphics/skia/ImageBufferSkia.cpp:
        (WebCore::getImageData):
        (WebCore::putImageData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47991 268f45cc-cd09-0410-ab3c-d52691b4dbfc