WebKit-https.git
7 years agoAX: Add a 'choose product' link for the new-ax-bug alias.
commit-queue@webkit.org [Wed, 17 Jul 2013 00:50:16 +0000 (00:50 +0000)]
AX: Add a 'choose product' link for the new-ax-bug alias.
https://bugs.webkit.org/show_bug.cgi?id=118697

Patch by James Craig <james@cookiecrook.com> on 2013-07-16
Reviewed by Chris Fleizach.

* template/en/custom/global/choose-product.html.tmpl: Added new link for accessibility bugs.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152751 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMissing break in WebGLRenderingContext::validateCompressedTexFuncData()
psolanki@apple.com [Wed, 17 Jul 2013 00:33:03 +0000 (00:33 +0000)]
Missing break in WebGLRenderingContext::validateCompressedTexFuncData()
https://bugs.webkit.org/show_bug.cgi?id=118742

Reviewed by Ryosuke Niwa.

* html/canvas/WebGLRenderingContext.cpp:
(WebCore::WebGLRenderingContext::validateCompressedTexFuncData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152750 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoWeb Inspector: InspectorFrontendHost is undefined
commit-queue@webkit.org [Wed, 17 Jul 2013 00:32:14 +0000 (00:32 +0000)]
Web Inspector: InspectorFrontendHost is undefined
https://bugs.webkit.org/show_bug.cgi?id=118715

Patch by Seokju Kwon <seokju.kwon@gmail.com> on 2013-07-16
Reviewed by Timothy Hatcher.

Add InspectorFrontendHostStub.js to New Inspector because of the InspectorFrontendHost initialization.

* UserInterface/InspectorFrontendHostStub.js: Added.
(.WebInspector.InspectorFrontendHostStub):
(.WebInspector.InspectorFrontendHostStub.prototype.platform):
(.WebInspector.InspectorFrontendHostStub.prototype.port):
(.WebInspector.InspectorFrontendHostStub.prototype.bringToFront):
(.WebInspector.InspectorFrontendHostStub.prototype.closeWindow):
(.WebInspector.InspectorFrontendHostStub.prototype.requestSetDockSide):
(.WebInspector.InspectorFrontendHostStub.prototype.setAttachedWindowHeight):
(.WebInspector.InspectorFrontendHostStub.prototype.setAttachedWindowWidth):
(.WebInspector.InspectorFrontendHostStub.prototype.setToolbarHeight):
(.WebInspector.InspectorFrontendHostStub.prototype.moveWindowBy):
(.WebInspector.InspectorFrontendHostStub.prototype.loaded):
(.WebInspector.InspectorFrontendHostStub.prototype.localizedStringsURL):
(.WebInspector.InspectorFrontendHostStub.prototype.inspectedURLChanged):
(.WebInspector.InspectorFrontendHostStub.prototype.copyText):
(.WebInspector.InspectorFrontendHostStub.prototype.openInNewTab):
(.WebInspector.InspectorFrontendHostStub.prototype.canSave):
(.WebInspector.InspectorFrontendHostStub.prototype.save):
(.WebInspector.InspectorFrontendHostStub.prototype.append):
(.WebInspector.InspectorFrontendHostStub.prototype.close.cleanup):
(.WebInspector.InspectorFrontendHostStub.prototype.close):
(.WebInspector.InspectorFrontendHostStub.prototype.sendMessageToBackend):
(.WebInspector.InspectorFrontendHostStub.prototype.loadResourceSynchronously):
* UserInterface/Main.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152749 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agocheck-webkit-style: "using namespace foo;" should be flagged as an error only in...
kseo@webkit.org [Wed, 17 Jul 2013 00:12:22 +0000 (00:12 +0000)]
check-webkit-style: "using namespace foo;" should be flagged as an error only in headers
https://bugs.webkit.org/show_bug.cgi?id=118755

Reviewed by Ryosuke Niwa.

Fix r152719 to check "using namespace foo;" only in headers.

* Scripts/webkitpy/style/checkers/cpp.py:
(check_using_namespace):
(check_style):
* Scripts/webkitpy/style/checkers/cpp_unittest.py:
(WebKitStyleTest.test_using_namespace):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152748 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoTypo in DFGInsertionSet.h header guard: "DFGInsectionSet_h" -> "DFGInsertionSet_h"
dfarler@apple.com [Tue, 16 Jul 2013 23:24:33 +0000 (23:24 +0000)]
Typo in DFGInsertionSet.h header guard: "DFGInsectionSet_h" -> "DFGInsertionSet_h"
https://bugs.webkit.org/show_bug.cgi?id=118753

Reviewed by Geoffrey Garen.

* dfg/DFGInsertionSet.h:
"DFGInsectionSet_h" -> "DFGInsertionSet_h"

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152747 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdded passing WebGL tests.
achristensen@apple.com [Tue, 16 Jul 2013 22:57:11 +0000 (22:57 +0000)]
Added passing WebGL tests.
https://bugs.webkit.org/show_bug.cgi?id=118696

Reviewed by Dean Jackson.

* platform/mac/TestExpectations: Replaced webgl directory with list of failing webgl tests.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152746 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAtomicize HTMLAnchorElement.hash before passing it to JS.
akling@apple.com [Tue, 16 Jul 2013 22:03:18 +0000 (22:03 +0000)]
Atomicize HTMLAnchorElement.hash before passing it to JS.
<http://webkit.org/b/118748>
<rdar://problem/14459780>

Reviewed by Sam Weinig.

Instead of generating a unique string from '#' + the URL hash every time HTMLAnchorElement.hash
is queried, turn it into an AtomicString. This prevents massive memory usage in scripts that
fetch .hash a lot.

Reduces memory consumption by 4.65 MB when viewing the full HTML5 spec at <http://whatwg.org/c>

* html/HTMLAnchorElement.cpp:
(WebCore::HTMLAnchorElement::hash):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152743 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMakeRope fixup shouldn't lead to an Identity without kids
fpizlo@apple.com [Tue, 16 Jul 2013 21:41:06 +0000 (21:41 +0000)]
MakeRope fixup shouldn't lead to an Identity without kids
https://bugs.webkit.org/show_bug.cgi?id=118745

Source/JavaScriptCore:

Reviewed by Mark Hahnenberg.

Make the empty string pruning part of fixupMakeRope() stop if it's on the last child.

Make Node::convertToIdentity release-assert that it has exactly one kid.

* dfg/DFGFixupPhase.cpp:
(JSC::DFG::FixupPhase::fixupMakeRope):
* dfg/DFGNode.h:
(JSC::DFG::Node::convertToIdentity):

LayoutTests:

Reviewed by Mark Hahnenberg.

* fast/js/dfg-make-rope-all-empty-strings.html: Added.
* fast/js/dfg-make-rope-all-empty-strings-expected.txt: Added.
* fast/js/jsc-test-list:
* fast/js/script-tests/dfg-make-rope-all-empty-strings.js: Added.
(foo):
(bar):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152742 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMake WebVTTParser return cue data instead of cue DOM objects
commit-queue@webkit.org [Tue, 16 Jul 2013 21:34:12 +0000 (21:34 +0000)]
Make WebVTTParser return cue data instead of cue DOM objects
https://bugs.webkit.org/show_bug.cgi?id=118687

Patch by Brendan Long <b.long@cablelabs.com> on 2013-07-16
Reviewed by Eric Carlson.

No new tests because this doesn't change functionality.

* html/track/InbandTextTrack.cpp:
Make TextTrackCueMap handle WebVTTCues instead of just generic cues.
m_dataToCueMap and m_cueToDataMap were renamed m_genericDataToCueMap
and m_genericCueToDataMap. The cue maps were turned into pointers so we
only allocate the ones we need.
(WebCore::TextTrackCueMap::TextTrackCueMap): Initialize maps to 0.
(WebCore::TextTrackCueMap::~TextTrackCueMap): Delete allocated maps.
(WebCore::TextTrackCueMap::add): Changed to allocate cue maps as needed, and added WebVTT version.
(WebCore::TextTrackCueMap::find): Checked to check if cue maps are allocated, and added WebVTT version.
(WebCore::TextTrackCueMap::findGenericData): Changed to accept TextTrackCue instead of TextTrackCueGeneric.
(WebCore::TextTrackCueMap::findWebVTTData): Same as findGenericData, except for WebVTTCueData.
(WebCore::TextTrackCueMap::remove): Accept TextTrackCue instead of TextTrackCueGeneric, and look in both maps.
(WebCore::InbandTextTrack::addWebVTTCue): Added, based on addGenericCue.
(WebCore::InbandTextTrack::removeWebVTTCue): Added, almost identical to removeGenericCue.
(WebCore::InbandTextTrack::removeCue): m_cueMap.remove() takes a TextTrackCue now so it can remove both types of cue.
(WebCore::InbandTextTrack::willRemoveTextTrackPrivate): Use ASSERT_UNUSED instead of UNUSED_PARAM + ASSERT.
* html/track/InbandTextTrack.h: Add new functions above, change maps to pointers and add maps for holding WebVTT cues.
* html/track/WebVTTParser.cpp:
(WebCore::WebVTTParser::getNewCues): Return WebVTTCueData instead of TextTrackCue.
(WebCore::WebVTTParser::createNewCue): Create WebVTTCueData instead of TextTrackCue.
* html/track/WebVTTParser.h: Add WebVTTCueData class, based on GenericCueData.
The following functions are just constructors, destructors, getters, or setters.
(WebCore::WebVTTCueData::create):
(WebCore::WebVTTCueData::~WebVTTCueData):
(WebCore::WebVTTCueData::startTime):
(WebCore::WebVTTCueData::setStartTime):
(WebCore::WebVTTCueData::endTime):
(WebCore::WebVTTCueData::setEndTime):
(WebCore::WebVTTCueData::id):
(WebCore::WebVTTCueData::setId):
(WebCore::WebVTTCueData::content):
(WebCore::WebVTTCueData::setContent):
(WebCore::WebVTTCueData::settings):
(WebCore::WebVTTCueData::setSettings):
(WebCore::WebVTTCueData::WebVTTCueData):
* loader/TextTrackLoader.cpp:
(WebCore::TextTrackLoader::getNewCues): Convert WebVTTCueData to TextTrackCue when we get them.
* platform/graphics/InbandTextTrackPrivateClient.h: Add addWebVTTCue and removeWebVTTCue functions to the interface.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152741 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago REGRESSION (r150291): Chinese predictive input pop-up disappears on twitter.com
ap@apple.com [Tue, 16 Jul 2013 21:14:58 +0000 (21:14 +0000)]
    REGRESSION (r150291): Chinese predictive input pop-up disappears on twitter.com
        https://bugs.webkit.org/show_bug.cgi?id=118739
        <rdar://problem/14300350>

        Reviewed by Daniel Bates.

        EditorState tracking on UI side is very fragile. This is a targeted fix just for
        this specific scenario.

        The issue was that we had a stale m_temporarilyClosedComposition flag, which was
        making UI process believe that a composition was closed from WebProcess side,
        and notify input method about that. It shouldn't have been a problem, because there
        is no composition at this point indeed, but this extra call is a problem for
        predictive input, which works in a somewhat unconventional way (<rdar://problem/14458297>).

        * UIProcess/mac/WebPageProxyMac.mm:
        (WebKit::WebPageProxy::insertText): Reset m_temporarilyClosedComposition flag when
        getting new EditorState in response to performing this action. There is certainly
        no composition after insertText, and all interested parties are on the same page
        already.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152740 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUse the correct cursor value for the unavailable plugin indicator
timothy_horton@apple.com [Tue, 16 Jul 2013 21:02:59 +0000 (21:02 +0000)]
Use the correct cursor value for the unavailable plugin indicator
https://bugs.webkit.org/show_bug.cgi?id=118747
<rdar://problem/14456498>

Reviewed by Dean Jackson.

CSS says "pointer" is the hand cursor, not "hand".

* css/unavailablePlugIns.css:
(object::-webkit-unavailable-plugin-content > div):
(object::-webkit-unavailable-plugin-content > div > span):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152739 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRemove reference to JSValueStructSupport.h from JSExport.h
mhahnenberg@apple.com [Tue, 16 Jul 2013 20:35:30 +0000 (20:35 +0000)]
Remove reference to JSValueStructSupport.h from JSExport.h
https://bugs.webkit.org/show_bug.cgi?id=118746

Reviewed by Filip Pizlo.

* API/JSExport.h: No such header exists, so it doesn't make sense to reference it.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152737 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed rebaseline test added in r152612 for Apple Windows port.
roger_fong@apple.com [Tue, 16 Jul 2013 20:32:13 +0000 (20:32 +0000)]
Unreviewed rebaseline test added in r152612 for Apple Windows port.

* platform/win/editing/caret: Added.
* platform/win/editing/caret/caret-color-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152736 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed rebaselines, AppleWin port.
roger_fong@apple.com [Tue, 16 Jul 2013 19:38:09 +0000 (19:38 +0000)]
Unreviewed rebaselines, AppleWin port.

* platform/win/fast/js/global-constructors-attributes-dedicated-worker-expected.txt:
* platform/win/fast/js/global-constructors-attributes-shared-worker-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152735 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed. Adjust test and update results for AppleWin port after r151868.
roger_fong@apple.com [Tue, 16 Jul 2013 18:51:20 +0000 (18:51 +0000)]
Unreviewed. Adjust test and update results for AppleWin port after r151868.

* platform/win/accessibility/multiple-select-element-role-expected.txt:
* platform/win/accessibility/multiple-select-element-role.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152734 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoSource/WebCore: Protect against the LayerFlushController being deleted inside its...
simon.fraser@apple.com [Tue, 16 Jul 2013 18:44:45 +0000 (18:44 +0000)]
Source/WebCore: Protect against the LayerFlushController being deleted inside its flushLayers() callback
https://bugs.webkit.org/show_bug.cgi?id=118741

Reviewed by Tim Horton.

It's possible (especially on iOS) for the LayerFlushController to be destroyed
inside its callback, via -[WebView _close]. Protect against this by making
it refcounted, and holding a ref across the callback.

Due to the odd relationship in which LayerFlushController owns its LayerFlushScheduler
by value, we achieve this by allowing subclasses of LayerFlushScheduler
to override runLoopObserverCallback(). WebViewLayerFlushScheduler uses the
override to protect the owner of the LayerFlushScheduler, which is the
LayerFlushController, when the callback is firing.

* WebCore.exp.in:
* platform/graphics/ca/LayerFlushScheduler.h: Make runLoopObserverCallback() and the dtor virtual.

Source/WebKit/mac: Protect against the LayerFlushController being deleted inside its flushLayers() callback
https://bugs.webkit.org/show_bug.cgi?id=118741
<rdar://problem/14402651>

Reviewed by Tim Horton.

It's possible (especially on iOS) for the LayerFlushController to be destroyed
inside its callback, via -[WebView _close]. Protect against this by making
it refcounted, and holding a ref across the callback.

Due to the odd relationship in which LayerFlushController owns its LayerFlushScheduler
by value, we achieve this by allowing subclasses of LayerFlushScheduler
to override runLoopObserverCallback(). WebViewLayerFlushScheduler uses the
override to protect the owner of the LayerFlushScheduler, which is the
LayerFlushController, when the callback is firing.

* WebView/WebView.mm:
(-[WebView _close]):
* WebView/WebViewData.h:
(WebViewLayerFlushScheduler::~WebViewLayerFlushScheduler):
(LayerFlushController::create):
* WebView/WebViewData.mm:
(LayerFlushController::invalidate):
(WebViewLayerFlushScheduler::WebViewLayerFlushScheduler):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152733 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed rebaselines, AppleWin port.
roger_fong@apple.com [Tue, 16 Jul 2013 18:38:09 +0000 (18:38 +0000)]
Unreviewed rebaselines, AppleWin port.

* platform/win/fast/js/global-constructors-attributes-dedicated-worker-expected.txt: Added.
* platform/win/fast/js/global-constructors-attributes-shared-worker-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152732 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed gardening. -webkit-user-select: all not supported on AppleWin port.
roger_fong@apple.com [Tue, 16 Jul 2013 18:16:38 +0000 (18:16 +0000)]
Unreviewed gardening. -webkit-user-select: all not supported on AppleWin port.
https://bugs.webkit.org/show_bug.cgi?id=118740

* platform/win/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152731 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoFix some NSDictionary misuse pointed out by the clang static analyzer
jberlin@webkit.org [Tue, 16 Jul 2013 18:04:15 +0000 (18:04 +0000)]
Fix some NSDictionary misuse pointed out by the clang static analyzer
https://bugs.webkit.org/show_bug.cgi?id=118736

Reviewed by Anders Carlsson.

Source/WebKit/mac:

* Plugins/Hosted/NetscapePluginHostManager.mm:
(WebKit::NetscapePluginHostManager::spawnPluginHost):
Explicitly cast the CFStringRef to an NSString *.
* WebCoreSupport/WebEditorClient.mm:
(WebEditorClient::documentFragmentFromAttributedString):
Do not put a key with a nil value pair into the NSDictionary.

Source/WebKit2:

* WebProcess/WebCoreSupport/mac/WebEditorClientMac.mm:
(WebKit::WebEditorClient::documentFragmentFromAttributedString):
Do not put a key with a nil value pair into the NSDictionary.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152730 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoGet rid of multiple inheritance support from the bindings generators
ch.dumez@sisa.samsung.com [Tue, 16 Jul 2013 16:24:45 +0000 (16:24 +0000)]
Get rid of multiple inheritance support from the bindings generators
https://bugs.webkit.org/show_bug.cgi?id=118353

Reviewed by Kentaro Hara.

Source/WebCore:

Remove multiple inheritance support from the bindings generators (except
the ObjC one which still needs it for now). Multiple inheritance is no
longer supported in Web IDL and it makes the scripts more complex than
they needs to be.

Multiple inheritance support is no longer needed now that our SVG
interfaces have been updated to use the simpler SVG2 inheritance model
and use Web IDL 'implements' statements.

Note that the IDL parser still support multiple inheritance for now in
order not to break the ObjC bindings generator.

No new tests, no behavior change.

* bindings/scripts/CodeGenerator.pm:
* bindings/scripts/CodeGeneratorCPP.pm:
(GetParentImplClassName):
(GetParent):
(GenerateImplementation):
* bindings/scripts/CodeGeneratorGObject.pm:
(GetParentClassName):
(GetParentImplClassName):
(GetParentGObjType):
* bindings/scripts/CodeGeneratorJS.pm:
(GetParentClassName):
(GenerateHeader):
(GenerateImplementation):
(GenerateConstructorDefinition):
* bindings/scripts/CodeGeneratorObjC.pm:
(AddMethodsConstantsAndAttributesFromParentInterfaces):
(GenerateImplementation):
* bindings/scripts/IDLParser.pm:
(parseInterface):
(parseException):
(parseInheritance):

Tools:

Update CodeGeneratorTestRunner.pm to use domInterface->parent instead of
deprecated domInterface->parents.

* WebKitTestRunner/InjectedBundle/Bindings/CodeGeneratorTestRunner.pm:
(_parentInterface):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152725 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[GTK] Unreviewed gardening. Update test expectations after r152701
commit-queue@webkit.org [Tue, 16 Jul 2013 16:08:31 +0000 (16:08 +0000)]
[GTK] Unreviewed gardening. Update test expectations after r152701
https://bugs.webkit.org/show_bug.cgi?id=118731

Unreviewed GTK gardening.

Patch by Simon Pena <simon.pena@samsung.com> on 2013-07-16

* platform/gtk/TestExpectations: Skip failing tests after r152701.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152722 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoHTMLMediaElement should not add cues for disabled text tracks
eric.carlson@apple.com [Tue, 16 Jul 2013 15:12:46 +0000 (15:12 +0000)]
HTMLMediaElement should not add cues for disabled text tracks
https://bugs.webkit.org/show_bug.cgi?id=118682

Source/WebCore:

Reviewed by Ryosuke Niwa.

Merge https://src.chromium.org/viewvc/blink?view=rev&revision=153810

Tests: media/track/track-disabled-addcue.html
       media/track/track-disabled.html

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::textTrackAddCues): Do nothing if the track is disabled.
(WebCore::HTMLMediaElement::textTrackAddCue): Ditto.

LayoutTests:

Merge https://src.chromium.org/viewvc/blink?view=rev&revision=153810

Reviewed by Ryosuke Niwa.

* media/track/track-disabled-addcue-expected.txt: Added.
* media/track/track-disabled-addcue.html: Added.
* media/track/track-disabled-expected.txt: Added.
* media/track/track-disabled.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152721 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Qt] Replace Nokia with Qt Project in service name.
michael.bruning@digia.com [Tue, 16 Jul 2013 13:51:32 +0000 (13:51 +0000)]
[Qt] Replace Nokia with Qt Project in service name.
https://bugs.webkit.org/show_bug.cgi?id=118729

Reviewed by Simon Hausmann.

* UIProcess/Launcher/qt/ProcessLauncherQt.cpp:
(WebKit::ProcessLauncher::launchProcess):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152720 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agocheck-webkit-style: "using namespace foo;" should be flagged as an error in headers
commit-queue@webkit.org [Tue, 16 Jul 2013 13:05:24 +0000 (13:05 +0000)]
check-webkit-style: "using namespace foo;" should be flagged as an error in headers
https://bugs.webkit.org/show_bug.cgi?id=57241

Patch by Brian Holt <brian.holt@samsung.com> on 2013-07-16
Reviewed by Ryosuke Niwa.

Added check for "using namespace foo" and unit test.

* Scripts/webkitpy/style/checkers/cpp.py:
(check_using_namespace):
(check_style):
(CppChecker):
* Scripts/webkitpy/style/checkers/cpp_unittest.py:
(WebKitStyleTest.test_using_namespace):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152719 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[GTK] fast/events/platform-wheelevent-with-delta-zero-crash.html is crashing
commit-queue@webkit.org [Tue, 16 Jul 2013 13:03:38 +0000 (13:03 +0000)]
[GTK] fast/events/platform-wheelevent-with-delta-zero-crash.html is crashing
https://bugs.webkit.org/show_bug.cgi?id=110222

Patch by Anton Obzhirov <a.obzhirov@samsung.com> on 2013-07-16
Reviewed by Gustavo Noronha Silva.

Tools:

Added handling of zero mouse scroll deltas in EventSender.

* DumpRenderTree/gtk/EventSender.cpp:
(mouseScrollByCallback):

LayoutTests:

* platform/gtk/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152718 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agowebkitpy: replace urllib with urllib2 for downloading additional modules
commit-queue@webkit.org [Tue, 16 Jul 2013 12:58:57 +0000 (12:58 +0000)]
webkitpy: replace urllib with urllib2 for downloading additional modules
https://bugs.webkit.org/show_bug.cgi?id=118671

Patch by Brian Holt <brian.holt@samsung.com> on 2013-07-16
Reviewed by Ryosuke Niwa.

Urllib2 is available whereever urllib is, but has better proxy support.

* Scripts/webkitpy/common/system/autoinstall.py:
(AutoInstaller._download_to_stream):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152717 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[ATK] Adds support for aria-haspopup property.
commit-queue@webkit.org [Tue, 16 Jul 2013 12:55:56 +0000 (12:55 +0000)]
[ATK] Adds support for aria-haspopup property.
https://bugs.webkit.org/show_bug.cgi?id=117834

Source/WebCore:

Patch by Krzysztof Czech <k.czech@ymail.com> on 2013-07-16
Reviewed by Christophe Dumez.

Adds support for aria-haspopup property.

Tests: accessibility/element-haspopup.html
       accessibility/popup-button-title.html

* accessibility/atk/WebKitAccessibleWrapperAtk.cpp:
(webkitAccessibleGetAttributes):

Tools:

Patch by Krzysztof Czech <k.czech@samsung.com> on 2013-07-16
Reviewed by Christophe Dumez.

Adds implementation for testing aria-haspopup property.

* DumpRenderTree/atk/AccessibilityUIElementAtk.cpp:
(getAttributeSetValueForId):
(AccessibilityUIElement::stringAttributeValue):
(AccessibilityUIElement::hasPopup):
* WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp:
(WTR::getAttributeSetValueForId):
(WTR::AccessibilityUIElement::stringAttributeValue):
(WTR::AccessibilityUIElement::hasPopup):

LayoutTests:

Patch by Krzysztof Czech <k.czech@ymail.com> on 2013-07-16
Reviewed by Christophe Dumez.

Moving accessibility tests related to aria-haspopup property.

* accessibility/element-haspopup-expected.txt: Renamed from LayoutTests/platform/mac/accessibility/element-haspopup-expected.txt.
* accessibility/element-haspopup.html: Renamed from LayoutTests/platform/mac/accessibility/element-haspopup.html.
* accessibility/popup-button-title-expected.txt: Renamed from LayoutTests/platform/mac/accessibility/popup-button-title-expected.txt.
* accessibility/popup-button-title.html: Renamed from LayoutTests/platform/mac/accessibility/popup-button-title.html.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152716 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Qt] Replace Nokia references.
michael.bruning@digia.com [Tue, 16 Jul 2013 12:22:20 +0000 (12:22 +0000)]
[Qt] Replace Nokia references.
https://bugs.webkit.org/show_bug.cgi?id=118728

Reviewed by Simon Hausmann.

Source/WebKit/qt:

* Api/qwebkitplatformplugin.h:
* examples/platformplugin/qwebkitplatformplugin.h:

Tools:

* MiniBrowser/qt/MiniBrowserApplication.cpp:
(MiniBrowserApplication::MiniBrowserApplication):
* QtTestBrowser/qttestbrowser.cpp:
(LauncherApplication::LauncherApplication):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152715 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Qt]Unreviewed gardening. New baseline and skip failing tests.
zarvai@inf.u-szeged.hu [Tue, 16 Jul 2013 11:57:08 +0000 (11:57 +0000)]
[Qt]Unreviewed gardening. New baseline and skip failing tests.

Patch by Gabor Abraham <abrhm@inf.u-szeged.hu> on 2013-07-16

* platform/qt-5.0-wk1/TestExpectations:
* platform/qt-5.0-wk1/http/tests/security/contentSecurityPolicy/object-src-none-allowed-expected.txt: Added.
* platform/qt/TestExpectations:
* platform/qt/editing/caret/caret-color-expected.png: Added.
* platform/qt/editing/caret/caret-color-expected.txt: Added.
* platform/qt/http/tests/security/contentSecurityPolicy/object-src-none-blocked-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152714 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoEWK2BackForwardListTest should be defined by inheriting from EWK2UnitTestBase.
commit-queue@webkit.org [Tue, 16 Jul 2013 11:10:43 +0000 (11:10 +0000)]
EWK2BackForwardListTest should be defined by inheriting from EWK2UnitTestBase.
https://bugs.webkit.org/show_bug.cgi?id=118717

Patch by Dong-Gwan Kim <donggwan.kim@samsung.com> on 2013-07-16
Reviewed by Christophe Dumez.

* UIProcess/API/efl/tests/test_ewk2_back_forward_list.cpp:
(EWK2BackForwardListTest::checkItem):
(EWK2BackForwardListTest::urlFromTitle):
(EWK2BackForwardListTest::freeEinaList):
(TEST_F):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152713 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[GTK] Remove compile warnings about GTK+ API deprecated after 3.6
carlosgc@webkit.org [Tue, 16 Jul 2013 10:47:17 +0000 (10:47 +0000)]
[GTK] Remove compile warnings about GTK+ API deprecated after 3.6
https://bugs.webkit.org/show_bug.cgi?id=118237

Reviewed by Philippe Normand.

We depend on GTK+3.6 so we are not interested in compile warnings
about deprecated API after 3.6

* Source/autotools/SetupAutoconfHeader.m4: Define
GDK_VERSION_MIN_REQUIRED in config.h.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152712 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[texmap][GStreamer] upload onto the texture only the buffer to be painted
commit-queue@webkit.org [Tue, 16 Jul 2013 10:16:46 +0000 (10:16 +0000)]
[texmap][GStreamer] upload onto the texture only the buffer to be painted
https://bugs.webkit.org/show_bug.cgi?id=118471

Patch by Víctor Manuel Jáquez Leal <vjaquez@igalia.com> on 2013-07-16
Reviewed by Philippe Normand.

Right now all the buffers are uploaded onto the texture. With this
logic pose situations where buffers that will not be painted are
uploaded.

This patch uploads only the buffers that are going to be shown.

With this approach, the buffers may arrive before a GraphicsLayer is
set, so we should be more cautious with GraphicsLayerTextureMapper's
client.

No new tests, covered by existing tests.

* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
(WebCore::MediaPlayerPrivateGStreamerBase::MediaPlayerPrivateGStreamerBase):
(WebCore::MediaPlayerPrivateGStreamerBase::updateTexture):
(WebCore::MediaPlayerPrivateGStreamerBase::triggerRepaint):
(WebCore::MediaPlayerPrivateGStreamerBase::paint):
(WebCore::MediaPlayerPrivateGStreamerBase::paintToTextureMapper):
* platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.h:
* platform/graphics/texmap/GraphicsLayerTextureMapper.cpp:
(WebCore::GraphicsLayerTextureMapper::~GraphicsLayerTextureMapper):
(WebCore::GraphicsLayerTextureMapper::setContentsToMedia):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152711 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[GStreamer] verify and log the received caps in the video sink
commit-queue@webkit.org [Tue, 16 Jul 2013 09:50:24 +0000 (09:50 +0000)]
[GStreamer] verify and log the received caps in the video sink
https://bugs.webkit.org/show_bug.cgi?id=118559

Patch by Víctor Manuel Jáquez Leal <vjaquez@igalia.com> on 2013-07-16
Reviewed by Philippe Normand.

Verify the correctness of the received caps in the video sink, and
also log them for debugging purposes.

No new tests, no behavior change.

* platform/graphics/gstreamer/VideoSinkGStreamer.cpp:
(webkitVideoSinkSetCaps):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152710 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUse toHTMLSelectElement and dismiss isHTMLSelectElement
kangil.han@samsung.com [Tue, 16 Jul 2013 09:20:04 +0000 (09:20 +0000)]
Use toHTMLSelectElement and dismiss isHTMLSelectElement
https://bugs.webkit.org/show_bug.cgi?id=118714

Reviewed by Ryosuke Niwa.

To avoid direct use of static_cast, this patch introduces toHTMLIFrameElement for code cleanup.
Additionally, this patch removes isHTMLSelectElement because not all element subclasses can be checked by a combination of tag names.

Source/WebCore:

* html/HTMLSelectElement.h:
(WebCore::toHTMLSelectElement):
* testing/Internals.cpp:
(WebCore::Internals::isSelectPopupVisible):

Source/WebKit/blackberry:

* WebKitSupport/InputHandler.cpp:
(BlackBerry::WebKit::InputHandler::didNodeOpenPopup):
(BlackBerry::WebKit::InputHandler::setPopupListIndex):
(BlackBerry::WebKit::InputHandler::setPopupListIndexes):

Source/WebKit/win:

* DOMHTMLClasses.cpp:
(DOMHTMLSelectElement::options):
(DOMHTMLSelectElement::activateItemAtIndex):

Source/WebKit2:

* WebProcess/Plugins/PDF/PDFPluginChoiceAnnotation.mm:
(WebKit::PDFPluginChoiceAnnotation::commit):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152709 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoReplace mouseClick with more meaningful method in test_ewk2_text_checker.cpp
commit-queue@webkit.org [Tue, 16 Jul 2013 07:35:49 +0000 (07:35 +0000)]
Replace mouseClick with more meaningful method in test_ewk2_text_checker.cpp
https://bugs.webkit.org/show_bug.cgi?id=118699

Patch by Dong-Gwan Kim <donggwan.kim@samsung.com> on 2013-07-16
Reviewed by Christophe Dumez.

Many mouseClick methods are used in test_ewk2_text_checker.cpp.
But it is difficult to understand what they mean.
So i would like to replace those with more meaningful method for readability.

* UIProcess/API/efl/tests/test_ewk2_text_checker.cpp:
(EWK2TextCheckerTest::clickSelectAllWordsWithSpellcheckButton):
(EWK2TextCheckerTest::clickSelectAllWordsWithoutSpellcheckButton):
(EWK2TextCheckerTest::clickSelectSubWordWithSpellcheckButton):
(EWK2TextCheckerTest::showContextMenuWithFirstLineText):
(EWK2TextCheckerTest::showContextMenuWithSecondLineText):
(EWK2TextCheckerTest::selectFirstWordInFirstLineText):
(EWK2TextCheckerTest::selectFirstWordInSecondLineText):
(TEST_F):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152708 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoSource/WebCore: Fix a crash in Range::processContents().
yutak@chromium.org [Tue, 16 Jul 2013 06:55:31 +0000 (06:55 +0000)]
Source/WebCore: Fix a crash in Range::processContents().

NULL ptr in WebCore::Range::processAncestorsAndTheirSiblings
https://bugs.webkit.org/show_bug.cgi?id=77614

Reviewed by Ryosuke Niwa.

This change is ported from Blink revision 153483:
https://src.chromium.org/viewvc/blink?revision=153483&view=revision

This crash can be initiated by calling Range.detach() while deleteContents()
is processing the same range. Range::processContents() should save the state
of the range since mutation events can change the state of the range.

Test: fast/dom/Range/detach-range-during-deletecontents.html

* dom/Range.cpp:
(WebCore::Range::processContents):
* dom/RangeBoundaryPoint.h:
(WebCore::RangeBoundaryPoint::RangeBoundaryPoint):

LayoutTests: NULL ptr in WebCore::Range::processAncestorsAndTheirSiblings
https://bugs.webkit.org/show_bug.cgi?id=77614

Reviewed by Ryosuke Niwa.

This change is ported from Blink revision 153483:
https://src.chromium.org/viewvc/blink?revision=153483&view=revision

* fast/dom/Range/detach-range-during-deletecontents-expected.txt: Added.
* fast/dom/Range/detach-range-during-deletecontents.html: Added.
* fast/dom/Range/resources/detach-range-during-deletecontents-iframe.xhtml: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152707 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd a threading assertion to AtomicString::remove
benjamin@webkit.org [Tue, 16 Jul 2013 06:32:43 +0000 (06:32 +0000)]
Add a threading assertion to AtomicString::remove
https://bugs.webkit.org/show_bug.cgi?id=118695

Reviewed by Sam Weinig.

* wtf/text/AtomicString.cpp:
(WTF::AtomicString::remove): The destructor of StringImpl remove
the string from the string table if it is atomic. This needs to be done
on the same thread as the one on which the string was added (otherwise the original
table would keep a dangling pointer to a dead string).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152706 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoInput element that becomes visible during a simulated click event from an associated...
rniwa@webkit.org [Tue, 16 Jul 2013 05:19:36 +0000 (05:19 +0000)]
Input element that becomes visible during a simulated click event from an associated label element doesn't get focused
https://bugs.webkit.org/show_bug.cgi?id=118700

Reviewed by Kent Tamura.

Source/WebCore:

Inspired by https://chromium.googlesource.com/chromium/blink/+/6435f70d53403b250e261a914f30d55142f81476.

The bug was caused by isMouseFocusable() check in HTMLLabelElement::defaultEventHandler not updating the layout even
though the call to dispatchSimulatedClick may have dirtied the render tree.

Explicitly update the layout so that we have up-to-date render tree.

Test: fast/forms/label/label-becomes-visible-while-clicking-on-label.html

* html/HTMLLabelElement.cpp:
(WebCore::HTMLLabelElement::defaultEventHandler):

LayoutTests:

Add a regression test.

* fast/forms/label/label-becomes-visible-while-clicking-on-label-expected.txt: Added.
* fast/forms/label/label-becomes-visible-while-clicking-on-label.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152705 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoIntroduce toHTMLIFrameElement
kangil.han@samsung.com [Tue, 16 Jul 2013 05:01:15 +0000 (05:01 +0000)]
Introduce toHTMLIFrameElement
https://bugs.webkit.org/show_bug.cgi?id=118672

Reviewed by Ryosuke Niwa.

To avoid direct use of static_cast, this patch introduces toHTMLIFrameElement for code cleanup.

Source/WebCore:

* bindings/js/JSHTMLDocumentCustom.cpp:
(WebCore::JSHTMLDocument::nameGetter):
* css/StyleResolver.cpp:
(WebCore::StyleResolver::adjustRenderStyle):
* dom/Document.cpp:
(WebCore::Document::seamlessParentIFrame):
* html/HTMLIFrameElement.h:
(WebCore::toHTMLIFrameElement):
* rendering/RenderEmbeddedObject.cpp:
(WebCore::RenderEmbeddedObject::viewCleared):
* rendering/RenderIFrame.cpp:
(WebCore::RenderIFrame::isSeamless):
(WebCore::RenderIFrame::flattenFrame):

Source/WebKit/win:

* DOMHTMLClasses.cpp:
(DOMHTMLIFrameElement::contentFrame):

Source/WebKit2:

* WebProcess/InjectedBundle/DOM/InjectedBundleNodeHandle.cpp:
(WebKit::InjectedBundleNodeHandle::htmlIFrameElementContentFrame):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152704 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed rebaselines after r152701.
timothy_horton@apple.com [Tue, 16 Jul 2013 04:56:29 +0000 (04:56 +0000)]
Unreviewed rebaselines after r152701.

* editing/pasteboard/paste-noplugin-expected.txt:
* http/tests/security/contentSecurityPolicy/object-src-url-blocked-expected.txt:
* platform/mac/accessibility/plugin-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152703 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd support for WaveShaperNode.oversample
crogers@google.com [Tue, 16 Jul 2013 04:55:32 +0000 (04:55 +0000)]
Add support for WaveShaperNode.oversample
https://bugs.webkit.org/show_bug.cgi?id=117435

Reviewed by Kenneth Russell.

Source/WebCore:

Adapted from Blink:
https://codereview.chromium.org/15619003/

Please see specification for details:
https://dvcs.w3.org/hg/audio/raw-file/tip/webaudio/specification.html#WaveShaperNode

.oversample specifies what type of oversampling (if any) should be used when applying the shaping curve.
The default value is "none", meaning the curve will be applied directly to the input samples.
A value of "2x" or "4x" can improve the quality of the processing by avoiding some aliasing,
with the "4x" value yielding the highest quality.

Tests: webaudio/waveshaper-oversample-2x.html
       webaudio/waveshaper-oversample-4x.html

* CMakeLists.txt:
* GNUmakefile.list.am:
* Modules/webaudio/WaveShaperDSPKernel.cpp:
(WebCore::WaveShaperDSPKernel::WaveShaperDSPKernel):
(WebCore::WaveShaperDSPKernel::lazyInitializeOversampling):
(WebCore::WaveShaperDSPKernel::process):
(WebCore::WaveShaperDSPKernel::processCurve):
(WebCore::WaveShaperDSPKernel::processCurve2x):
(WebCore::WaveShaperDSPKernel::processCurve4x):
(WebCore::WaveShaperDSPKernel::reset):
(WebCore::WaveShaperDSPKernel::latencyTime):
* Modules/webaudio/WaveShaperDSPKernel.h:
* Modules/webaudio/WaveShaperNode.cpp:
(WebCore::WaveShaperNode::WaveShaperNode):
(WebCore::WaveShaperNode::setOversample):
(WebCore::WaveShaperNode::oversample):
* Modules/webaudio/WaveShaperNode.h:
(WebCore::WaveShaperNode::latency):
* Modules/webaudio/WaveShaperNode.idl:
* Modules/webaudio/WaveShaperProcessor.cpp:
(WebCore::WaveShaperProcessor::WaveShaperProcessor):
(WebCore::WaveShaperProcessor::setOversample):
* Modules/webaudio/WaveShaperProcessor.h:
(WebCore::WaveShaperProcessor::oversample):
* WebCore.xcodeproj/project.pbxproj:
* platform/audio/DownSampler.cpp: Added.
(WebCore::DownSampler::DownSampler):
(WebCore::DownSampler::initializeKernel):
(WebCore::DownSampler::process):
(WebCore::DownSampler::reset):
(WebCore::DownSampler::latencyFrames):
* platform/audio/DownSampler.h: Added.
* platform/audio/UpSampler.cpp: Added.
(WebCore::UpSampler::UpSampler):
(WebCore::UpSampler::initializeKernel):
(WebCore::UpSampler::process):
(WebCore::UpSampler::reset):
(WebCore::UpSampler::latencyFrames):
* platform/audio/UpSampler.h: Added.

LayoutTests:

* webaudio/resources/waveshaper-testing.js: Added.
(T0):
(T1):
(T2):
(T3):
(T4):
(generateWaveShapingCurve):
(checkShapedCurve):
(createImpulseBuffer):
(runWaveShaperOversamplingTest):
* webaudio/waveshaper-oversample-2x-expected.txt: Added.
* webaudio/waveshaper-oversample-2x.html: Added.
* webaudio/waveshaper-oversample-4x-expected.txt: Added.
* webaudio/waveshaper-oversample-4x.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152702 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUpdate blocked/missing plug-in UI
timothy_horton@apple.com [Tue, 16 Jul 2013 04:01:36 +0000 (04:01 +0000)]
Update blocked/missing plug-in UI
https://bugs.webkit.org/show_bug.cgi?id=118347
<rdar://problem/14209318>

Reviewed by Sam Weinig.
Patch by Anders Carlsson, Antoine Quint, Sam Weinig, and myself.

* CMakeLists.txt:
* DerivedSources.make:
* DerivedSources.pri:
* GNUmakefile.am:
* WebCore.exp.in:
* WebCore.vcxproj/WebCore.vcxproj:
* WebCore.vcxproj/WebCore.vcxproj.filters:
* WebCore.xcodeproj/project.pbxproj:
* css/CSSDefaultStyleSheets.cpp:
* css/unavailablePlugIns.css: Added.
Add new Source/WebCore/css/unavailablePlugIns.css stylesheet.

* dom/EventListener.h:
New event listener type to support the creation of an event listener
for the label and icon for missing / blocked plug-in UI.

* html/HTMLPlugInElement.cpp:
(WebCore::HTMLPlugInElement::defaultEventHandler):
Remove explicit event handling since we're now using a DOM event handler
on the label and icon in the shadow root.

* html/HTMLPlugInImageElement.cpp:
(WebCore::HTMLPlugInImageElement::didAddUserAgentShadowRoot):
Don't populate the shadow root with plugin snapshotting elements
if we're not planning on showing a snapshot, because this means that
the shadow root was created for the unavailable plugin indicator instead.

* page/ChromeClient.h:
(WebCore::ChromeClient::shouldUnavailablePluginMessageIncludeButton):
* rendering/RenderEmbeddedObject.cpp:
(WebCore::RenderEmbeddedObject::RenderEmbeddedObject):
(WebCore::shouldUnavailablePluginMessageIncludeButton):
Rename from shouldUnavailablePluginMessageBeButton to
shouldUnavailablePluginMessageIncludeButton for accuracy.

(WebCore::RenderEmbeddedObject::setPluginUnavailabilityReasonDescription):
Set custom text to be displayed in the unavailability indicator.

(WebCore::RenderEmbeddedObject::setPluginUnavailabilityReason):
(WebCore::RenderEmbeddedObject::handleUnavailablePluginButtonClickEvent):
(WebCore::RenderEmbeddedObject::willBeDestroyed):
(WebCore::RenderEmbeddedObject::isReplacementObscured):
(WebCore::RenderEmbeddedObjectEventListener::handleEvent):
(WebCore::RenderEmbeddedObjectEventListener::operator==):
Remove the old code used to render the missing / blocked plug-in UI in C++ in favor
of populating a shadow root in setPluginUnavailabilityReason(). We add a DOM event
listener in the shadow root to eventually call into unavailablePluginButtonClicked()
on the ChromeClient.

(WebCore::RenderEmbeddedObject::paint):
(WebCore::RenderEmbeddedObject::layout):
Paint and lay out UA shadow root children.

(WebCore::RenderEmbeddedObject::canHaveChildren):
canHaveChildren can use the existence of a UA shadow root to decide
whether a RenderEmbeddedObject can have children.

* rendering/RenderEmbeddedObject.h:
(WebCore::RenderEmbeddedObjectEventListener::create):
(WebCore::RenderEmbeddedObjectEventListener::cast):
(WebCore::RenderEmbeddedObjectEventListener::RenderEmbeddedObjectEventListener):
Create the new event listener class necessary to handle events in the shadow root.

(WebCore::RenderEmbeddedObject::pluginUnavailabilityReasonDescription): Added.

* rendering/RenderWidget.h:
Promote willBeDestroyed() to be public so we may override it in the RenderEmbeddedObject subclass.

* WebCoreSupport/WebChromeClient.h:
* WebCoreSupport/WebChromeClient.mm:
(WebChromeClient::shouldUnavailablePluginMessageIncludeButton):
Rename from shouldUnavailablePluginMessageBeButton to
shouldUnavailablePluginMessageIncludeButton for accuracy.

* Platform/CoreIPC/HandleMessage.h:
(CoreIPC::callMemberFunction):
Add a 6 argument -> 4 reply message handler.
(It seems that FindPlugin is getting a little out of hand.)

* Shared/APIClientTraits.cpp:
* Shared/APIClientTraits.h:
* UIProcess/API/C/WKPage.h:
* UIProcess/WebLoaderClient.cpp:
(WebKit::WebLoaderClient::pluginLoadPolicy):
* UIProcess/WebLoaderClient.h:
Add an unavailability description out-argument to a new version of
pluginLoadPolicy, so clients can override the text of the unavailable
plugin indicator. Bump the WKPageLoaderClient version and update APIClientTraits.

* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::findPlugin):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
Plumb the unavailability description through to the WebProcess via
the FindPlugin message.

* WebProcess/Plugins/PluginProcessConnection.cpp:
(WebKit::PluginProcessConnection::didClose):
Make a strongly referencing copy of the list of PluginProxies, so we can
throw them away without the HashMap being mutated underneath us (see the comment).

* WebProcess/Plugins/PluginView.cpp:
(WebKit::PluginView::pluginProcessCrashed):
Invalidate the plug-in's widget before setting the plugin unavailability reason,
because doing so would cause the plug-in's renderer to be torn down, making
invalidate() crash.

* WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::shouldUnavailablePluginMessageIncludeButton):
* WebProcess/WebCoreSupport/WebChromeClient.h:
Rename from shouldUnavailablePluginMessageBeButton to
shouldUnavailablePluginMessageIncludeButton for accuracy.

* WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::createPlugin):
Acquire the unavailability description from the client
(via FindPlugin) and hand it to the RenderEmbeddedObject.
(WebKit::WebPage::canPluginHandleResponse):
Re-acquire the plugin's renderer, in case setPluginUnavailabilityReason destroyed it.

* MiniBrowser/mac/WK2BrowserWindowController.m:
(-[WK2BrowserWindowController awakeFromNib]):
* WebKitTestRunner/TestController.cpp:
(WTR::TestController::createWebViewWithOptions):
Match new WKPageLoaderClient entry.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152701 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed. Add Gwang Yoon Hwang to contributor list.
kseo@webkit.org [Tue, 16 Jul 2013 03:56:49 +0000 (03:56 +0000)]
Unreviewed. Add Gwang Yoon Hwang to contributor list.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152700 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMark svg/custom/xlink-prefix-in-attributes.html as failing on Mac
jpfau@apple.com [Tue, 16 Jul 2013 03:32:08 +0000 (03:32 +0000)]
Mark svg/custom/xlink-prefix-in-attributes.html as failing on Mac
https://bugs.webkit.org/show_bug.cgi?id=118701

Unreviewed, changing test expectations

* platform/mac/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152699 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoPluginProcess deny file-read-data /Library/Application Support/Macromedia/FlashPlayer...
commit-queue@webkit.org [Tue, 16 Jul 2013 02:36:28 +0000 (02:36 +0000)]
PluginProcess deny file-read-data /Library/Application Support/Macromedia/FlashPlayerTrust
https://bugs.webkit.org/show_bug.cgi?id=118690
<rdar://problem/14255963>

Patch by Simon Cooper <scooper@apple.com> on 2013-07-15
Reviewed by Dean Jackson.

Silently deny access to FlashPlayerTrust. This location contains
files containing lists of paths that Flash Player will
"allow" access to (without asking the user). Since the plugin
sandbox won't permit the access to the listed paths it is better
to silently block attempts to read these "whitelists". The
"whitelists" are created by other Adobe "installer" like
applications.

* Resources/PlugInSandboxProfiles/com.macromedia.Flash Player.plugin.sb:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152698 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUpdate James Craig's email address.
cfleizach@apple.com [Tue, 16 Jul 2013 01:29:30 +0000 (01:29 +0000)]
Update James Craig's email address.

* Scripts/webkitpy/common/config/contributors.json:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152691 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoPack WebCore::Font slightly better.
akling@apple.com [Tue, 16 Jul 2013 01:19:17 +0000 (01:19 +0000)]
Pack WebCore::Font slightly better.
<http://webkit.org/b/118681>
<rdar://problem/14446014>

Reviewed by Sam Weinig.

Knock 8 bytes off of Font by packing the members better. Reduces memory consumption by 429 kB when
viewing the full HTML5 spec at <http://whatwg.org/c>

* platform/graphics/Font.h:
(WebCore::Font::typesettingFeatures):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152689 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoXMLSerializer doesn't include namespaces on nodes in HTML documents
rwlbuis@webkit.org [Tue, 16 Jul 2013 00:37:02 +0000 (00:37 +0000)]
XMLSerializer doesn't include namespaces on nodes in HTML documents
https://bugs.webkit.org/show_bug.cgi?id=16496

Patch by Rob Buis <rwlbuis@webkit.org> on 2013-07-15
Reviewed by Ryosuke Niwa.

Source/WebCore:

Introduce a xml fragment serialization mode as indicated by
http://html5.org/specs/dom-parsing.html#xmlserializer (commit 00b84d2). In this mode
the XML fragment serialization algorithm is respected, the changes in this patch do the following:

- elements are self-closing if otherwise this would lead to invalid xml syntax.
- the resulting xml is XML namespace-well-formed (http://www.w3.org/TR/xml-names11/#Conformance).
This is achieved by always, when in xml fragment serialization mode, attempting to write out the element/attribute namespace,
preventing using the XML namespace as a default namespace and special casing the use of element's in XML namespace by using the xml prefix.

The chosen approach matches FireFox 25 behavior.

Test: fast/dom/dom-serialize-namespace.html

* WebCore.order: Adapt to changed createMarkup signature.
* editing/MarkupAccumulator.cpp:
(WebCore::MarkupAccumulator::MarkupAccumulator):
(WebCore::MarkupAccumulator::serializeNodesWithNamespaces): makes sure xml namespace/prefix is known so it is never used in namespace declarations.
(WebCore::MarkupAccumulator::appendNamespace): Avoid adding namespace declarations that do not differ from current default namespace.
(WebCore::MarkupAccumulator::appendOpenTag): Print xml prefix if the element's namespace is XML to avoid conflicts.
(WebCore::MarkupAccumulator::appendAttribute):
(WebCore::MarkupAccumulator::shouldAddNamespaceAttribute): Also take into account xmlns attributes with no namespace.
(WebCore::MarkupAccumulator::shouldSelfClose): Force self-closing to create well-formed XML elements.
* editing/MarkupAccumulator.h: Use EFragmentSerialization.
(WebCore::MarkupAccumulator::inXMLFragmentSerialization):
* editing/markup.cpp:
(WebCore::createMarkup):
* editing/markup.h: Add EFragmentSerialization enum.
* xml/XMLSerializer.cpp:
(WebCore::XMLSerializer::serializeToString):

LayoutTests:

The updated tests are progressions and match FF.

* fast/dom/Element/getAttribute-check-case-sensitivity-expected.txt:
* fast/dom/Element/script-tests/getAttribute-check-case-sensitivity.js:
* fast/dom/XMLSerializer-xml-namespace-expected.txt:
* fast/dom/dom-serialize-namespace-expected.txt: Added.
* fast/dom/dom-serialize-namespace.html: Added.
* fast/xsl/xslt-processor-expected.txt:
* inspector/elements/set-outer-html-expected.txt:
* svg/custom/xlink-prefix-in-attributes-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152685 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoWeb Inspector: Remove unused columns from select query in ImageUtilities.js
commit-queue@webkit.org [Tue, 16 Jul 2013 00:32:40 +0000 (00:32 +0000)]
Web Inspector: Remove unused columns from select query in ImageUtilities.js
https://bugs.webkit.org/show_bug.cgi?id=118694

Patch by Seokju Kwon <seokju.kwon@gmail.com> on 2013-07-15
Reviewed by Timothy Hatcher.

* UserInterface/ImageUtilities.js: Remove pixelRatio and formatVersion from select query.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152684 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[EFL] Add fallback for TestWebKitAPI if opengl is not available.
commit-queue@webkit.org [Mon, 15 Jul 2013 23:47:32 +0000 (23:47 +0000)]
[EFL] Add fallback for TestWebKitAPI if opengl is not available.
https://bugs.webkit.org/show_bug.cgi?id=118583

Patch by Dong-Gwan Kim <donggwan.kim@samsung.com> on 2013-07-15
Reviewed by Christophe Dumez.

There're many fails on efl webkit2 API test if opengl is not available.
So, add fallback to non-opengl evas engine.

* TestWebKitAPI/efl/PlatformWebView.cpp:
(TestWebKitAPI::initEcoreEvas):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152675 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMediaFragmentURIParser::parseFragments shouldn't upconvert 8-bit string
rniwa@webkit.org [Mon, 15 Jul 2013 23:42:36 +0000 (23:42 +0000)]
MediaFragmentURIParser::parseFragments shouldn't upconvert 8-bit string
https://bugs.webkit.org/show_bug.cgi?id=118692

Reviewed by Michael Saboff.

Merge https://chromium.googlesource.com/chromium/blink/+/2aa17cc56a807f7e05d386d6eb66cda25e6b0542.

* html/MediaFragmentURIParser.cpp:
(WebCore::MediaFragmentURIParser::parseFragments):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152673 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoNetworkStorageSession::createDefaultSession is only implemented by Soup
kseo@webkit.org [Mon, 15 Jul 2013 23:41:23 +0000 (23:41 +0000)]
NetworkStorageSession::createDefaultSession is only implemented by Soup
https://bugs.webkit.org/show_bug.cgi?id=118679

Reviewed by Alexey Proskuryakov.

Source/WebCore:

Remove NetworkStorageSession::createDefaultSession() because
NetworkStorageSession::defaultStorageSession() returns the default
session.

No new tests, no behavior change.

* platform/network/NetworkStorageSession.h:
* platform/network/soup/NetworkStorageSessionSoup.cpp:

Source/WebKit/efl:

Replace NetworkStorageSession::createDefaultSession() with
NetworkStorageSession::defaultStorageSession().

* ewk/ewk_view.cpp:
(_ewk_view_priv_new):
(EWKPrivate::storageSession):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152672 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUpdate CodeMirror to 3.14.1.
timothy@apple.com [Mon, 15 Jul 2013 23:31:12 +0000 (23:31 +0000)]
Update CodeMirror to 3.14.1.

https://bugs.webkit.org/show_bug.cgi?id=118605

Reviewed by Joseph Pecoraro.

* Scripts/update-codemirror-resources.rb: Updated the list to add closebrackets.js and sorted the files.
* Tools/PrettyPrinting/CodeMirrorFormatters.js:
* Tools/PrettyPrinting/Formatter.js:
* Tools/PrettyPrinting/FormatterContentBuilder.js:
* Tools/PrettyPrinting/codemirror.css:
* Tools/PrettyPrinting/codemirror.js:
* Tools/PrettyPrinting/css.js:
* Tools/PrettyPrinting/javascript.js:
* UserInterface/External/CodeMirror/closebrackets.js:
* UserInterface/External/CodeMirror/codemirror.css:
* UserInterface/External/CodeMirror/codemirror.js:
* UserInterface/External/CodeMirror/coffeescript.js:
* UserInterface/External/CodeMirror/comment.js:
* UserInterface/External/CodeMirror/css.js:
* UserInterface/External/CodeMirror/javascript.js:
* UserInterface/External/CodeMirror/matchbrackets.js:
* UserInterface/External/CodeMirror/sql.js:
* UserInterface/External/CodeMirror/xml.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152669 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agocompositionstart event should contain the text to be replaced
rniwa@webkit.org [Mon, 15 Jul 2013 23:28:50 +0000 (23:28 +0000)]
compositionstart event should contain the text to be replaced
https://bugs.webkit.org/show_bug.cgi?id=118684

Reviewed by Alexey Proskuryakov.

Source/WebCore:

Merge https://chromium.googlesource.com/chromium/blink/+/9e04f728a95ad7d4783b1d31c3cdc1412dd6cc4d.

DOM3 level 3 specifies compositionstart event's data to be the text to be replaced.
http://www.w3.org/TR/DOM-Level-3-Events/#event-type-compositionstart

According to the author of the Blink change, IE10 on Windows and Firefox22 on Linux confirms
to the specified behavior.

Test: fast/events/ime-composition-events-001.html

* editing/Editor.cpp:
(WebCore::Editor::setComposition):

LayoutTests:

Add a test case.

* fast/events/ime-composition-events-001-expected.txt:
* fast/events/ime-composition-events-001.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152668 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoxmlDocPtrForString shouldn't upconvert 8-bit string
rniwa@webkit.org [Mon, 15 Jul 2013 23:27:35 +0000 (23:27 +0000)]
xmlDocPtrForString shouldn't upconvert 8-bit string
https://bugs.webkit.org/show_bug.cgi?id=118693

Reviewed by Michael Saboff.

Merge https://chromium.googlesource.com/chromium/blink/+/5d9b240db74591c3689415c7b7c5180fadb379e9

Prior to this changeset, we were upconverting the entire XML document.
Let xmllib2 parse a document as iso-8859-1 when it's a 8-bit string.

* xml/parser/XMLDocumentParserLibxml2.cpp:
(WebCore::nativeEndianUTF16Encoding):
(WebCore::xmlDocPtrForString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152667 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoReplace WKCACFTypes include with d3d9 include.
roger_fong@apple.com [Mon, 15 Jul 2013 20:46:36 +0000 (20:46 +0000)]
Replace WKCACFTypes include with d3d9 include.

Rubberstamped by Eric Carlson.

* platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152655 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed. Remove unused vsprops files.
roger_fong@apple.com [Mon, 15 Jul 2013 20:44:26 +0000 (20:44 +0000)]
Unreviewed. Remove unused vsprops files.

* win/tools/vsprops/FeatureDefines.vsprops: Removed.
* win/tools/vsprops/FeatureDefinesCairo.vsprops: Removed.
* win/tools/vsprops/WinCairo.vsprops: Removed.
* win/tools/vsprops/cURL.vsprops: Removed.
* win/tools/vsprops/common.vsprops: Removed.
* win/tools/vsprops/debug.vsprops: Removed.
* win/tools/vsprops/debug_all.vsprops: Removed.
* win/tools/vsprops/debug_wincairo.vsprops: Removed.
* win/tools/vsprops/production.vsprops: Removed.
* win/tools/vsprops/release.vsprops: Removed.
* win/tools/vsprops/releaseproduction.vsprops: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152654 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoText with "text-overflow:ellipsis" and an SVG font do not render correctly
ddkilzer@apple.com [Mon, 15 Jul 2013 20:07:42 +0000 (20:07 +0000)]
Text with "text-overflow:ellipsis" and an SVG font do not render correctly
https://bugs.webkit.org/show_bug.cgi?id=118669

Patch by Antoine Quint <graouts@apple.com> on 2013-07-15
Reviewed by David Kilzer.

Source/WebCore:

Partial runs may be routed through the complex path in several Font functions. While incomplete,
we can check for a non-NULL run.renderingContext() and exempt runs that have it from being forced
down the complex path. A more thorough fix is already tracked by http://webkit.org/b/100050.

Adding a new testcase which is a copy of svg/text/text-overflow-ellipsis-svgfont.html with kerning
and ligatures turned on explicitly since DRT runs with those off by default and this option is required
to be on for the issue to reproduce.

Fix courtesy of Dan Bernstein.

Test: svg/text/text-overflow-ellipsis-svgfont-kerning-ligatures.html

* platform/graphics/Font.cpp:
(WebCore::Font::drawText):
(WebCore::Font::drawEmphasisMarks):
(WebCore::Font::selectionRectForText):

LayoutTests:

* platform/mac/svg/text/text-overflow-ellipsis-svgfont-kerning-ligatures-expected.png: Added.
* platform/mac/svg/text/text-overflow-ellipsis-svgfont-kerning-ligatures-expected.txt: Added.
* svg/text/text-overflow-ellipsis-svgfont-kerning-ligatures.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152653 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMake sure to add main resources for sub-frames to the Network Timeline.
timothy@apple.com [Mon, 15 Jul 2013 19:46:30 +0000 (19:46 +0000)]
Make sure to add main resources for sub-frames to the Network Timeline.

https://bugs.webkit.org/show_bug.cgi?id=118649

Reviewed by Joseph Pecoraro.

* UserInterface/TimelineManager.js:
(WebInspector.TimelineManager.prototype._mainResourceDidChange): Remove the
call to _clear() and early return for event.target.isMainFrame(). The _clear()
is already handled if needed by the call to _startAutoRecording(), so it was
redundant here and caused the erroneous addition of isMainFrame() early return.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152652 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoProperly handle style text changes while the color popover is open.
timothy@apple.com [Mon, 15 Jul 2013 18:53:27 +0000 (18:53 +0000)]
Properly handle style text changes while the color popover is open.

https://bugs.webkit.org/show_bug.cgi?id=118575

Reviewed by Joseph Pecoraro.

* UserInterface/CSSStyleDeclarationTextEditor.js:
(WebInspector.CSSStyleDeclarationTextEditor.prototype._colorSwatchClicked.updateCodeMirror):
The original text marker might have been cleared by a style update, in this case we need to
find the new color text marker so we know the right range for the new style color text.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152651 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Curl] Download does not replace existing file.
commit-queue@webkit.org [Mon, 15 Jul 2013 18:22:49 +0000 (18:22 +0000)]
[Curl] Download does not replace existing file.
https://bugs.webkit.org/show_bug.cgi?id=118356

Patch by peavo@outlook.com <peavo@outlook.com> on 2013-07-15
Reviewed by Brent Fulgham.

If the download target file already exists, it is not replaced after the download has finished.
At this point, the user would have already approved to replace the file, so it should be replaced.

* platform/network/curl/CurlDownload.cpp:
(CurlDownload::moveFileToDestination): Move file and replace it if it already exists.
If the new file is on a different volume, do a copy/delete operation.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152649 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[CoordinatedGraphics] Crash at CoordinatedGraphicsScene::updateImageBacking
luciano.wolf@openbossa.org [Mon, 15 Jul 2013 17:50:50 +0000 (17:50 +0000)]
[CoordinatedGraphics] Crash at CoordinatedGraphicsScene::updateImageBacking
https://bugs.webkit.org/show_bug.cgi?id=118613

The operations of creating/removing/updating/clearing weren't in the
right sequence. Changing the order solves the issue because when an image is
replaced the old one is deleted before the new one is created, and IDs are
not unique in cairo (they're based on the image pointer).

Reviewed by Noam Rosenthal.

* platform/graphics/texmap/coordinated/CoordinatedGraphicsScene.cpp:
(WebCore::CoordinatedGraphicsScene::syncImageBackings):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152647 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoSource/WebKit/mac: [WK2] Calling Notification.requestPermission() without a callback...
jonlee@apple.com [Mon, 15 Jul 2013 17:16:12 +0000 (17:16 +0000)]
Source/WebKit/mac: [WK2] Calling Notification.requestPermission() without a callback crashes
https://bugs.webkit.org/show_bug.cgi?id=118654
<rdar://problem/14293352>

Reviewed by Alexey Proskuryakov.

* WebCoreSupport/WebNotificationClient.mm:
(-[WebNotificationPolicyListener initWithCallback:]): The callback can be NULL, so remove the ASSERT.
(-[WebNotificationPolicyListener allow]): Check for the callback before asking it to handle the event.
(-[WebNotificationPolicyListener deny]): Ditto.

Source/WebKit2: Calling Notification.requestPermission() without a callback crashes
https://bugs.webkit.org/show_bug.cgi?id=118654
<rdar://problem/14293352>

Reviewed by Alexey Proskuryakov.

When requestPermission() is called, we short-circuit if we know the permission was already set.
The callback handler gets invoked. But now that callbacks are optional, we need to check to make
sure a callback was provided.

* WebProcess/Notifications/NotificationPermissionRequestManager.cpp:
(WebKit::NotificationPermissionRequestManager::startRequest): Check for the callback.

LayoutTests: [WK2] Calling Notification.requestPermission() without a callback crashes
https://bugs.webkit.org/show_bug.cgi?id=118654
<rdar://problem/14293352>

Reviewed by Alexey Proskuryakov.

* http/tests/notifications/notification-request-permission-expected.txt:
* http/tests/notifications/notification-request-permission.html: Add additional subtests that check
that the call succeeds even after granting or denying permission.

* http/tests/notifications/notification-request-permission-no-callback-expected.txt: Added.
* http/tests/notifications/notification-request-permission-no-callback.html: Added. Similar test to
notification-request-permission, without the callback function.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152642 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agocheck-webkit-style: Ignore false positive: Instance of 'Popen' has no 'pid' member
commit-queue@webkit.org [Mon, 15 Jul 2013 16:32:41 +0000 (16:32 +0000)]
check-webkit-style: Ignore false positive: Instance of 'Popen' has no 'pid' member
https://bugs.webkit.org/show_bug.cgi?id=118592

Patch by Brian Holt <brian.holt@samsung.com> on 2013-07-15
Reviewed by Martin Robinson.

Added a suppression for pylint false positives and a unit test.

* Scripts/webkitpy/style/checkers/python.py:
(Pylinter):
* Scripts/webkitpy/style/checkers/python_unittest.py:
(PythonCheckerTest.test_check):
(PythonCheckerTest):
(PythonCheckerTest.test_pylint_false_positives):
(PythonCheckerTest.test_pylint_false_positives._mock_handle_style_error):
* Scripts/webkitpy/style/checkers/python_unittest_falsepositives.py: Added.
(test_popen):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152639 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[GTK] Remote inspector server should send an error page if resources not found
commit-queue@webkit.org [Mon, 15 Jul 2013 15:36:56 +0000 (15:36 +0000)]
[GTK] Remote inspector server should send an error page if resources not found
https://bugs.webkit.org/show_bug.cgi?id=117137

Patch by Anton Obzhirov <a.obzhirov@samsung.com> on 2013-07-15
Reviewed by Gustavo Noronha Silva.

Sometimes it is difficult to understand why remote inspector fails to deliver the content.
This patch gives an option to show error page with some basic explanation.

* UIProcess/InspectorServer/gtk/WebInspectorServerGtk.cpp:
(WebKit::WebInspectorServer::platformResourceForPath):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152633 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Qt][WK2] Unreviewed cleanup for pixel tester bot.
zarvai@inf.u-szeged.hu [Mon, 15 Jul 2013 14:47:04 +0000 (14:47 +0000)]
[Qt][WK2] Unreviewed cleanup for pixel tester bot.

* platform/qt-5.0-wk2/TestExpectations:
* platform/qt-5.0-wk2/editing/deleting/delete-image-002-expected.png:
* platform/qt-5.0-wk2/editing/selection/move-by-character-004-expected.png:
* platform/qt-5.0-wk2/fast/canvas/setWidthResetAfterForcedRender-expected.png:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152632 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[ATK] Leak: Leak in WebKitAccessibleTextGetText
commit-queue@webkit.org [Mon, 15 Jul 2013 13:24:25 +0000 (13:24 +0000)]
[ATK] Leak: Leak in WebKitAccessibleTextGetText
https://bugs.webkit.org/show_bug.cgi?id=118596

Patch by Brian Holt <brian.holt@samsung.com> on 2013-07-15
Reviewed by Carlos Garcia Campos.

Fixed memory leak.

* accessibility/atk/WebKitAccessibleInterfaceText.cpp:
(getGailTextUtilForAtk):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152631 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Qt][WK2] Unreviewed gardening. Skipping bad pixel results.
zarvai@inf.u-szeged.hu [Mon, 15 Jul 2013 12:02:21 +0000 (12:02 +0000)]
[Qt][WK2] Unreviewed gardening. Skipping bad pixel results.

* platform/qt-5.0-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152630 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[GTK] Increase the JHBuild version of GStreamer to 1.0.7
commit-queue@webkit.org [Mon, 15 Jul 2013 10:01:33 +0000 (10:01 +0000)]
[GTK] Increase the JHBuild version of GStreamer to 1.0.7
https://bugs.webkit.org/show_bug.cgi?id=115166

Patch by Xabier Rodriguez Calvar <calvaris@igalia.com> on 2013-07-15
Reviewed by Philippe Normand.

* gtk/jhbuild.modules: Bumping GStreamer version up to 1.0.8.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152629 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Qt][WK2] Unreviewed gardening. Skipping bad pixel results.
zarvai@inf.u-szeged.hu [Mon, 15 Jul 2013 09:44:35 +0000 (09:44 +0000)]
[Qt][WK2] Unreviewed gardening. Skipping bad pixel results.

* platform/qt-5.0-wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152628 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[GTK] [WK2] Check if the keyEventQueue is empty in WebPageProxy::getEditorCommandsFor...
commit-queue@webkit.org [Mon, 15 Jul 2013 08:49:48 +0000 (08:49 +0000)]
[GTK] [WK2] Check if the keyEventQueue is empty in WebPageProxy::getEditorCommandsForKeyEvent
https://bugs.webkit.org/show_bug.cgi?id=118530

Patch by Simon Pena <simon.pena@samsung.com> on 2013-07-15
Reviewed by Carlos Garcia Campos.

When a key event is started in the WebProcess (e.g. in the inspector)
it doesn't have an associated GdkEvent queued. Adding a check in
WebPageProxy::getEditorCommandsForKeyEvent ensures there's always a queued
key event to process.

In GTK-WK2, inspector-protocol/input/dispatchKeyEvent.html no longer crashes once
this fix is in place.

* UIProcess/gtk/WebPageProxyGtk.cpp:
(WebKit::WebPageProxy::getEditorCommandsForKeyEvent): Ensure m_keyEventQueue is
not empty.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152627 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[WK2][GTK] SIGSEV in webkitWebViewBaseSizeAllocate
commit-queue@webkit.org [Mon, 15 Jul 2013 08:14:18 +0000 (08:14 +0000)]
[WK2][GTK] SIGSEV in webkitWebViewBaseSizeAllocate
https://bugs.webkit.org/show_bug.cgi?id=118052

Patch by Alberto Garcia <berto@igalia.com> on 2013-07-15
Reviewed by Carlos Garcia Campos.

We don't need to check whether the drawing area is empty or not in
order to set needsResizeOnMap.

It can also happen that the drawing area pointer is null (if
e.g. the web process crashes), which is the reason for this
SIGSEGV.

A test to prevent this was added in r88646 but was later moved to
resizeWebKitWebViewBaseFromAllocation().

* UIProcess/API/gtk/WebKitWebViewBase.cpp:
(webkitWebViewBaseSizeAllocate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152626 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMinor cleaning of AtomicString::addSlowCase
benjamin@webkit.org [Mon, 15 Jul 2013 08:08:51 +0000 (08:08 +0000)]
Minor cleaning of AtomicString::addSlowCase
https://bugs.webkit.org/show_bug.cgi?id=118615

Reviewed by Geoffrey Garen.

Give the variable a proper name.
Add assertions to clarify the in and out states.

* wtf/text/AtomicString.cpp:
(WTF::AtomicString::addSlowCase):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152625 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed, rolling out r152598.
commit-queue@webkit.org [Mon, 15 Jul 2013 06:41:06 +0000 (06:41 +0000)]
Unreviewed, rolling out r152598.
http://trac.webkit.org/changeset/152598
https://bugs.webkit.org/show_bug.cgi?id=118658

totally broke the web inspector (JS ERROR: SyntaxError:
Expected token '}') (Requested by thorton on #webkit).

* Tools/PrettyPrinting/CodeMirrorFormatters.js:
* Tools/PrettyPrinting/Formatter.js:
* Tools/PrettyPrinting/FormatterContentBuilder.js:
* Tools/PrettyPrinting/codemirror.css:
(.CodeMirror-scrollbar-filler):
(.CodeMirror-gutters):
(.CodeMirror-vscrollbar, .CodeMirror-hscrollbar, .CodeMirror-scrollbar-filler):
(.CodeMirror-widget):
* Tools/PrettyPrinting/codemirror.js:
(window.CodeMirror):
(window.CodeMirror.):
(.do):
(moveVisually):
* Tools/PrettyPrinting/css.js:
(.):
(CodeMirror):
* Tools/PrettyPrinting/javascript.js:
(.):
* UserInterface/External/CodeMirror/codemirror.css:
(.CodeMirror-widget):
* UserInterface/External/CodeMirror/codemirror.js:
(window.CodeMirror):
(window.CodeMirror.):
* UserInterface/External/CodeMirror/coffeescript.js:
* UserInterface/External/CodeMirror/comment.js:
(.):
* UserInterface/External/CodeMirror/css.js:
(.):
(CodeMirror):
* UserInterface/External/CodeMirror/javascript.js:
(.):
* UserInterface/External/CodeMirror/matchbrackets.js:
(.):
* UserInterface/External/CodeMirror/sql.js:
* UserInterface/External/CodeMirror/xml.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152624 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoOrigins with invalid ports are not marked as unique (118652)
jonlee@apple.com [Mon, 15 Jul 2013 06:11:14 +0000 (06:11 +0000)]
Origins with invalid ports are not marked as unique (118652)
https://bugs.webkit.org/show_bug.cgi?id=118652
<rdar://problem/14437753>

Reviewed by Sam Weinig.

* page/SecurityOrigin.cpp:
(WebCore::SecurityOrigin::create): Return a unique origin if the port is out of range.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152623 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAnother Qt build fix attempt after r152619.
rniwa@webkit.org [Mon, 15 Jul 2013 01:27:20 +0000 (01:27 +0000)]
Another Qt build fix attempt after r152619.

* dom/QualifiedName.cpp:
(WebCore::QualifiedName::toString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152622 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoQt build fix attempt after r152619.
rniwa@webkit.org [Mon, 15 Jul 2013 01:22:02 +0000 (01:22 +0000)]
Qt build fix attempt after r152619.

* dom/QualifiedName.cpp:
(WebCore::QualifiedName::toString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152621 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd Company100 to team.html
kseo@webkit.org [Mon, 15 Jul 2013 01:01:54 +0000 (01:01 +0000)]
Add Company100 to team.html
https://bugs.webkit.org/show_bug.cgi?id=118633

Reviewed by Andreas Kling.

* team.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152620 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoMac build fix attempt after r152615.
rniwa@webkit.org [Mon, 15 Jul 2013 00:47:53 +0000 (00:47 +0000)]
Mac build fix attempt after r152615.

* svg/SVGFilterElement.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152619 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoModernize QualifiedName by wrapping gNameCache in a function and using more early...
rniwa@webkit.org [Mon, 15 Jul 2013 00:45:12 +0000 (00:45 +0000)]
Modernize QualifiedName by wrapping gNameCache in a function and using more early exits
https://bugs.webkit.org/show_bug.cgi?id=118299

Reviewed by Andreas Kling.

Address Ben's review comment.

* dom/QualifiedName.cpp:
(WebCore::QualifiedName::toString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152618 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoapply-style-command-crash.html is flaky
rniwa@webkit.org [Mon, 15 Jul 2013 00:35:07 +0000 (00:35 +0000)]
apply-style-command-crash.html is flaky
https://bugs.webkit.org/show_bug.cgi?id=118634

Reviewed by Andreas Kling.

Merge https://chromium.googlesource.com/chromium/blink/+/a30dad44ded92d24ed56761434cf6041688659a0.

Also replace CRLF to LF and get rid of svn:executable.

* editing/execCommand/apply-style-command-crash.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152617 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoHTMLTextFormControlElement::valueWithHardLineBreaks shouldn't upconvert 8-bit string
rniwa@webkit.org [Sun, 14 Jul 2013 14:49:38 +0000 (14:49 +0000)]
HTMLTextFormControlElement::valueWithHardLineBreaks shouldn't upconvert 8-bit string
https://bugs.webkit.org/show_bug.cgi?id=118631

Reviewed by Andreas Kling.

Merge https://chromium.googlesource.com/chromium/blink/+/2c7c70a7bd3d61bfbf97de95c459ec9f40b46e31.

* html/HTMLTextFormControlElement.cpp:
(WebCore::HTMLTextFormControlElement::valueWithHardLineBreaks):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152616 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAdd toSVGFilterElement() and use it.
gyuyoung.kim@samsung.com [Sun, 14 Jul 2013 13:59:31 +0000 (13:59 +0000)]
Add toSVGFilterElement() and use it.
https://bugs.webkit.org/show_bug.cgi?id=118653

Reviewed by Ryosuke Niwa.

We can use toSVGFilterElement() instead of using static_cast<..> to improve readability.
Blink also uses it.

No new tests, no behavior change.

* rendering/svg/RenderSVGResourceFilter.cpp:
(WebCore::RenderSVGResourceFilter::buildPrimitives):
(WebCore::RenderSVGResourceFilter::applyResource):
(WebCore::RenderSVGResourceFilter::resourceBoundingBox):
* rendering/svg/RenderSVGResourceFilter.h:
(WebCore::RenderSVGResourceFilter::filterUnits):
(WebCore::RenderSVGResourceFilter::primitiveUnits):
* rendering/svg/SVGResources.cpp:
(WebCore::targetReferenceFromResource):
* svg/SVGFilterElement.h:
(WebCore::toSVGFilterElement):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152615 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoIntroduce toHTMLVideoElement
kangil.han@samsung.com [Sun, 14 Jul 2013 08:55:49 +0000 (08:55 +0000)]
Introduce toHTMLVideoElement
https://bugs.webkit.org/show_bug.cgi?id=118582

Reviewed by Ryosuke Niwa.

To avoid direct use of static_cast, this patch introduces toHTMLVideoElement.

Source/WebCore:

* html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::getPluginProxyParams):
* html/HTMLVideoElement.h:
(WebCore::toHTMLVideoElement):
* html/MediaDocument.cpp:
(WebCore::MediaDocumentParser::createDocumentStructure):
(WebCore::descendentVideoElement):
(WebCore::ancestorVideoElement):
* html/shadow/MediaControlElements.cpp:
(WebCore::MediaControlTextTrackContainerElement::updateDisplay):
* rendering/HitTestResult.cpp:
(WebCore::HitTestResult::enterFullscreenForVideo):
* rendering/RenderVideo.cpp:
(WebCore::RenderVideo::videoElement):

Source/WebKit/qt:

* WebCoreSupport/DumpRenderTreeSupportQt.cpp:
(DumpRenderTreeSupportQt::mediaContentUrlByElementId):
* WebCoreSupport/FullScreenVideoQt.cpp:
(WebCore::FullScreenVideoQt::enterFullScreenForNode):
(WebCore::FullScreenVideoQt::exitFullScreenForNode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152614 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed, rolling out r151978.
commit-queue@webkit.org [Sun, 14 Jul 2013 06:12:03 +0000 (06:12 +0000)]
Unreviewed, rolling out r151978.
http://trac.webkit.org/changeset/151978
https://bugs.webkit.org/show_bug.cgi?id=118651

Caused regressions at least 3 websites (Requested by rniwa on
#webkit).

Source/JavaScriptCore:

* runtime/JSCJSValue.h:
* runtime/JSString.h:

Source/WebCore:

* bindings/js/JSDOMBinding.cpp:
* bindings/js/JSDOMBinding.h:
* bindings/scripts/CodeGeneratorJS.pm:
(GetNativeTypeFromSignature):
(JSValueToNative):
* bindings/scripts/IDLAttributes.txt:
* bindings/scripts/test/JS/JSTestObj.cpp:
(WebCore::setJSTestObjReflectedStringAttr):
(WebCore::setJSTestObjReflectedURLAttr):
(WebCore::setJSTestObjReflectedCustomURLAttr):
* dom/Document.idl:
* dom/Element.idl:
* dom/Node.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152613 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoCaret should respect text background color
joone.hur@intel.com [Sun, 14 Jul 2013 06:03:05 +0000 (06:03 +0000)]
Caret should respect text background color
https://bugs.webkit.org/show_bug.cgi?id=117493

Reviewed by Ryosuke Niwa.

Source/WebCore:

This patch allows the caret to become visible in the black background
by getting the caret color from the color of the element containing
the text, not the parent element that has the contentEditable attribute.

Test: editing/caret/caret-color.html

* editing/FrameSelection.cpp:
(WebCore::CaretBase::paintCaret):

LayoutTests:

This tests to make sure that the caret is visible in the black background
by getting the caret color from the color of the element(span) containing
the text, not the parent element(div) that has the contentEditable
attribute.

* editing/caret/caret-color-expected.txt: Added.
* editing/caret/caret-color.html: Added.
* platform/efl/editing/caret/caret-color-expected.png: Added.
* platform/mac/TestExpectations: Skip this test case.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152612 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoAvoid upconverting strings in various places in WebCore
rniwa@webkit.org [Sun, 14 Jul 2013 01:11:43 +0000 (01:11 +0000)]
Avoid upconverting strings in various places in WebCore
https://bugs.webkit.org/show_bug.cgi?id=118632

Reviewed by Andreas Kling.

Avoid calling String::characters() in various places since it upconverts 8-bit strings to 16-bit strings.

Merge
https://chromium.googlesource.com/chromium/blink/+/a6162e2ad7c7870e22445d3c463d17e7ac871e80
https://chromium.googlesource.com/chromium/blink/+/18095209b3f467758b83894e7b14f813f6953f81
https://chromium.googlesource.com/chromium/blink/+/fefcf2b95d55f24c60fd2e95978cf4544f3c92ca
https://chromium.googlesource.com/chromium/blink/+/8e0527b0fb33998318aedfd74b3511025f7ff294
https://chromium.googlesource.com/chromium/blink/+/feaf798b04597b0849b4000fc305264895d3eac5
https://chromium.googlesource.com/chromium/blink/+/bd1a49103a6e07b1023d2c742d8217769efbffb4

* css/CSSParser.cpp:
(WebCore::CSSParser::setupParser):
(WebCore::CSSParser::parseImageSet):
(WebCore::CSSParser::rewriteSpecifiers):
* css/CSSParserValues.h:
(WebCore::CSSParserString::init):
(WebCore::CSSParserString::clear):
* dom/DatasetDOMStringMap.cpp:
(WebCore::isValidAttributeName):
(WebCore::convertAttributeNameToPropertyName):
(WebCore::propertyNameMatchesAttributeName):
(WebCore::isValidPropertyName):
(WebCore::convertPropertyNameToAttributeName):
* dom/Range.cpp:
(WebCore::Range::toString):
* dom/SecurityContext.cpp:
(WebCore::SecurityContext::parseSandboxPolicy):
* html/canvas/CanvasRenderingContext2D.cpp:
(WebCore::CanvasRenderingContext2D::measureText):
* html/track/TextTrackCue.cpp:
(WebCore::TextTrackCue::markFutureAndPastNodes):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152611 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoparseHTMLInteger shouldn't upconvert 8-bit string
rniwa@webkit.org [Sat, 13 Jul 2013 15:32:07 +0000 (15:32 +0000)]
parseHTMLInteger shouldn't upconvert 8-bit string
https://bugs.webkit.org/show_bug.cgi?id=118629

Reviewed by Sam Weinig.

Merge https://chromium.googlesource.com/chromium/blink/+/e0ebab23e0c16b49dc90c9c39b8bbbf4f243a01e.

* html/parser/HTMLParserIdioms.cpp:
(WebCore::parseHTMLInteger):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152610 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed. Add an email address and update my domain of experts.
kseo@webkit.org [Sat, 13 Jul 2013 03:44:16 +0000 (03:44 +0000)]
Unreviewed. Add an email address and update my domain of experts.

* Scripts/webkitpy/common/config/contributors.json:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152609 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoRemove some dead stores pointed out by the clang static analyzer.
jberlin@webkit.org [Fri, 12 Jul 2013 23:59:16 +0000 (23:59 +0000)]
Remove some dead stores pointed out by the clang static analyzer.
https://bugs.webkit.org/show_bug.cgi?id=118608

Reviewed by Anders Carlsson.

* Plugins/Hosted/NetscapePluginHostManager.mm:
(WebKit::NetscapePluginHostManager::instantiatePlugin):
kr is never read after this point, so there is no reason to assign to it.

* Plugins/WebBasePluginPackage.mm:
(-[WebBasePluginPackage getPluginInfoFromPLists]):
The extensions array is only used in the while loop. Move the declaration to the first place
it is used in the while loop. Since it is always reassigned at that location, and since it
is never read after the subsequent for loop, the assignment to an array containing the empty
string is never used. Remove it.
Similarly, the value of the description is always overwritten before used to assign to
mimeClassInfo.desc or to pluginInfo.desc, so the code to assign it to the empty string is
never used. Remove it.

* WebView/WebTextCompletionController.mm:
(-[WebTextCompletionController _placePopupWindow:]):
maxWidth is never read (and never was since it was added in r7311) after it is used on the
line above to set windowFrame.size.width.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152608 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[iOS] TextTrackRepresentation not cleared when captions are disabled
eric.carlson@apple.com [Fri, 12 Jul 2013 23:35:15 +0000 (23:35 +0000)]
[iOS] TextTrackRepresentation not cleared when captions are disabled
https://bugs.webkit.org/show_bug.cgi?id=118623

Reviewed by Dean Jackson.

* html/shadow/MediaControlElements.cpp:
(WebCore::MediaControlTextTrackContainerElement::updateDisplay): Do not return immediately
    when captions are disabled so the necessary cleanup happens.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152607 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoSource/JavaScriptCore: Optimize addStrackTraceIfNecessary to be faster in the case...
commit-queue@webkit.org [Fri, 12 Jul 2013 22:37:51 +0000 (22:37 +0000)]
Source/JavaScriptCore: Optimize addStrackTraceIfNecessary to be faster in the case when it's not necessary
https://bugs.webkit.org/show_bug.cgi?id=118328

Patch by Chris Curtis <chris_curtis@apple.com> on 2013-07-12
Reviewed by Geoffrey Garen.

Retrieving the stack is costly. We want to get it only once. By moving the check
for the .stack property above the code to retrieve the stack, we ensure this.

* interpreter/Interpreter.cpp:
(JSC::Interpreter::addStackTraceIfNecessary):

LayoutTests: By optimizing when the stack is added a two tests needed to be modifed to show correct results.
https://bugs.webkit.org/show_bug.cgi?id=118328

Patch by Chris Curtis <chris_curtis@apple.com> on 2013-07-12
Reviewed by Geoffrey Garen.

* inspector/console/console-exception-stack-traces.html: This test compares the console's currect
stack with the error object's stack. The test was failing on decodeURI() and eval() which create
a new frame on the stack to execute. The console's stack was unaware of these calls and the size
of the stacks would not match. I added a check to pass if it was the specific case with decodeURI
or eval.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152606 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[jhbuild] Avoid restarting from scratch failed initial checkouts
commit-queue@webkit.org [Fri, 12 Jul 2013 20:05:08 +0000 (20:05 +0000)]
[jhbuild] Avoid restarting from scratch failed initial checkouts
https://bugs.webkit.org/show_bug.cgi?id=118603

Patch by Emanuele Aina <emanuele.aina@collabora.com> on 2013-07-12
Reviewed by Martin Robinson.

If the initial checkout is interrupted midway (eg. because the internet
connection went down) config files do not get generated and that causes
the next checkout to start from scratch (the missing files cause the
script to delete the build root and the downloaded sources).

By generating the files before invoking jhbuild, subsequent runs are able
to start where the previous run left.

* Scripts/update-webkit-libs-jhbuild:
(saveJhbuildMd5): invoke it before runJhbuild().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152605 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Windows] Remove single-worker restriction on tests.
bfulgham@apple.com [Fri, 12 Jul 2013 19:51:38 +0000 (19:51 +0000)]
[Windows] Remove single-worker restriction on tests.

* Scripts/webkitpy/port/win.py: Remove restriction of one worker.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152604 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Windows] Attempt to satisfy the Open Source Windows bots.
bfulgham@apple.com [Fri, 12 Jul 2013 19:30:23 +0000 (19:30 +0000)]
[Windows] Attempt to satisfy the Open Source Windows bots.

* DerivedSources.make: Restore the HAVE_AVCF macro detection.
* config.h: Return to using HAVE_AVCF to activate AVFoundation code.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152603 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUnreviewed, rolling out r152599.
commit-queue@webkit.org [Fri, 12 Jul 2013 19:05:04 +0000 (19:05 +0000)]
Unreviewed, rolling out r152599.
http://trac.webkit.org/changeset/152599
https://bugs.webkit.org/show_bug.cgi?id=118612

This patch includes a failing unit test (Requested by
mrobinson on #webkit).

* Scripts/webkitpy/style/checkers/python.py:
(Pylinter):
* Scripts/webkitpy/style/checkers/python_unittest.py:
(PythonCheckerTest.test_check):
* Scripts/webkitpy/style/checkers/python_unittest_falsepositives.py: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152602 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Tools] Ignore missing GNUmakefile when detecting previous arch in build-webkit
commit-queue@webkit.org [Fri, 12 Jul 2013 18:57:35 +0000 (18:57 +0000)]
[Tools] Ignore missing GNUmakefile when detecting previous arch in build-webkit
https://bugs.webkit.org/show_bug.cgi?id=91223

Patch by Emanuele Aina <emanuele.aina@collabora.com> on 2013-07-12
Reviewed by Martin Robinson.

When run on a clean checkout, `build-webkit` complains that no
GNUmakefile exists when trying to detect the previously configured
architecture, but in this case there's no previous configuration so
the warning is pointless.

* Scripts/webkitdirs.pm:
(determineArchitecture): Redirect grep stderr to /dev/null.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152601 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years ago[Windows] Build correction after r152573/r152577.
bfulgham@apple.com [Fri, 12 Jul 2013 18:54:07 +0000 (18:54 +0000)]
[Windows] Build correction after r152573/r152577.
https://bugs.webkit.org/show_bug.cgi?id=118610

Reviewed by Oliver Hunt.

* jit/JITThunks.cpp:
(JSC::JITThunks::hostFunctionStub): Hand-feed MSVC++ the fact that we want the second
argument of the make_pair to be a function pointer.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152600 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agocheck-webkit-style: Ignore false positive: Instance of 'Popen' has no 'pid' member
commit-queue@webkit.org [Fri, 12 Jul 2013 18:36:25 +0000 (18:36 +0000)]
check-webkit-style: Ignore false positive: Instance of 'Popen' has no 'pid' member
https://bugs.webkit.org/show_bug.cgi?id=118592

Patch by Brian Holt <brian.holt@samsung.com> on 2013-07-12
Reviewed by Martin Robinson.

Added a suppression for pylint false positives and a unit test.

* Scripts/webkitpy/style/checkers/python.py:
(Pylinter):
* Scripts/webkitpy/style/checkers/python_unittest.py:
(PythonCheckerTest.test_check):
(PythonCheckerTest):
(PythonCheckerTest.test_pylint_false_positives):
(PythonCheckerTest.test_pylint_false_positives._mock_handle_pylint_false_positives):
* Scripts/webkitpy/style/checkers/python_unittest_falsepositives.py: Added.
(test_popen):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152599 268f45cc-cd09-0410-ab3c-d52691b4dbfc

7 years agoUpdate CodeMirror to 7f560afa1c (3.14+).
timothy@apple.com [Fri, 12 Jul 2013 18:31:29 +0000 (18:31 +0000)]
Update CodeMirror to 7f560afa1c (3.14+).

https://bugs.webkit.org/show_bug.cgi?id=118605

Reviewed by Sam Weinig.

* Tools/PrettyPrinting/CodeMirrorFormatters.js:
* Tools/PrettyPrinting/Formatter.js:
* Tools/PrettyPrinting/FormatterContentBuilder.js:
* Tools/PrettyPrinting/codemirror.css:
* Tools/PrettyPrinting/codemirror.js:
* Tools/PrettyPrinting/css.js:
* Tools/PrettyPrinting/javascript.js:
* UserInterface/External/CodeMirror/codemirror.css:
* UserInterface/External/CodeMirror/codemirror.js:
* UserInterface/External/CodeMirror/coffeescript.js:
* UserInterface/External/CodeMirror/comment.js:
* UserInterface/External/CodeMirror/css.js:
* UserInterface/External/CodeMirror/javascript.js:
* UserInterface/External/CodeMirror/matchbrackets.js:
* UserInterface/External/CodeMirror/sql.js:
* UserInterface/External/CodeMirror/xml.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@152598 268f45cc-cd09-0410-ab3c-d52691b4dbfc