WebKit-https.git
6 years ago[CMake] Allow user specified compiler flags to take precedence
commit-queue@webkit.org [Mon, 26 Nov 2012 08:31:18 +0000 (08:31 +0000)]
[CMake] Allow user specified compiler flags to take precedence
https://bugs.webkit.org/show_bug.cgi?id=103101

Patch by Laszlo Gombos <l.gombos@samsung.com> on 2012-11-26
Reviewed by Brent Fulgham.

Make sure that compiler and linker flags specified by the build system
are always prepended to the variables that can be specified by the
environment and the user as well.

* Source/cmake/OptionsCommon.cmake:
* Source/cmake/OptionsWindows.cmake:
* Source/cmake/WebKitHelpers.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135688 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[V8] Give isolated shells a lifecycle like that of main shells
commit-queue@webkit.org [Mon, 26 Nov 2012 08:10:35 +0000 (08:10 +0000)]
[V8] Give isolated shells a lifecycle like that of main shells
https://bugs.webkit.org/show_bug.cgi?id=96522

Patch by Dan Carney <dcarney@google.com> on 2012-11-26
Reviewed by Adam Barth.

Refactored the isolated shells in ScriptController
to be cleaned up the same way the main shell is.

No new tests. No change in functionality.

* bindings/v8/ScriptController.cpp:
(WebCore::ScriptController::~ScriptController):
(WebCore::ScriptController::clearForOutOfMemory):
(WebCore::ScriptController::clearForClose):
(WebCore::ScriptController::clearWindowShell):
* bindings/v8/ScriptController.h:
(ScriptController):
* bindings/v8/V8DOMWindowShell.cpp:
(WebCore::V8DOMWindowShell::destroyIsolatedShell):
(WebCore::V8DOMWindowShell::clearForClose):
* bindings/v8/V8DOMWindowShell.h:
(V8DOMWindowShell):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135687 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed gardening.
ryuan.choi@samsung.com [Mon, 26 Nov 2012 07:54:56 +0000 (07:54 +0000)]
Unreviewed gardening.
Unskip the fixed regressions from r135662

* platform/efl-wk1/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135686 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, rolling out r135656.
commit-queue@webkit.org [Mon, 26 Nov 2012 07:50:02 +0000 (07:50 +0000)]
Unreviewed, rolling out r135656.
http://trac.webkit.org/changeset/135656
https://bugs.webkit.org/show_bug.cgi?id=103218

Made a few SVG tests crash on all platforms (Requested by
apavlov on #webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-11-25

* css/CSSComputedStyleDeclaration.cpp:
(WebCore::valueForFamily):
* css/CSSValuePool.cpp:
(WebCore::CSSValuePool::createFontFamilyValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135685 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWebCore::RenderBlock::determineStartPosition crash
tasak@google.com [Mon, 26 Nov 2012 07:45:38 +0000 (07:45 +0000)]
WebCore::RenderBlock::determineStartPosition crash
https://bugs.webkit.org/show_bug.cgi?id=98993

Reviewed by Brent Fulgham.

Source/WebCore:

If we move some node and the node has some text,
InlineFlowBox::removeChild() is invoked. The method invokes
RootInlineBox::childRemoved(). childRemoved() checks whether the
removed inlinebox has the renderer of its parent's line break object.
If so, use setLineBreakInfo to make the parent's line break info to
be 0. However in RenderBlock::determineStartPosition(), the code
assume that all line break info is solved, i.e.
prevRootBox->lineBreakObj()->isText(). Since lineBreakObj() returns 0
because of removeChild(), determineStartPosition crash occurs.

Test: fast/inline/inline-box-append-child-crash.html

* rendering/RenderBlockLineLayout.cpp:
(WebCore::RenderBlock::determineStartPosition):
Checks whether lineBreakObj() is 0 or not before using lineBreakObj().

LayoutTests:

* fast/inline/inline-box-append-child-crash-expected.txt: Added.
* fast/inline/inline-box-append-child-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135684 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWeb Inspector: use native Element.prototype.normalize instead of custom defined Eleme...
commit-queue@webkit.org [Mon, 26 Nov 2012 06:35:57 +0000 (06:35 +0000)]
Web Inspector: use native Element.prototype.normalize instead of custom defined Element.prototype.pruneEmptyTextNodes
https://bugs.webkit.org/show_bug.cgi?id=103205

Patch by Nikita Vasilyev <me@elv1s.ru> on 2012-11-25
Reviewed by Pavel Feldman.

* inspector/front-end/DOMExtension.js:
* inspector/front-end/TextPrompt.js:
(WebInspector.TextPrompt.prototype.clearAutoComplete):
(WebInspector.TextPrompt.prototype._completionsReady):
(WebInspector.TextPrompt.prototype.applySuggestion):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135683 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWeb Inspector: remove unused DOM methods
commit-queue@webkit.org [Mon, 26 Nov 2012 06:33:08 +0000 (06:33 +0000)]
Web Inspector: remove unused DOM methods
https://bugs.webkit.org/show_bug.cgi?id=103204

Patch by Nikita Vasilyev <me@elv1s.ru> on 2012-11-25
Reviewed by Pavel Feldman.

* inspector/front-end/DOMExtension.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135682 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWeb Inspector: Remove unused Array.convert method
commit-queue@webkit.org [Mon, 26 Nov 2012 06:15:45 +0000 (06:15 +0000)]
Web Inspector: Remove unused Array.convert method
https://bugs.webkit.org/show_bug.cgi?id=103202

Patch by Nikita Vasilyev <me@elv1s.ru> on 2012-11-25
Reviewed by Daniel Bates.

* inspector/front-end/utilities.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135681 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAX: file upload input cannot be activated with VoiceOver
cfleizach@apple.com [Mon, 26 Nov 2012 06:07:47 +0000 (06:07 +0000)]
AX: file upload input cannot be activated with VoiceOver
https://bugs.webkit.org/show_bug.cgi?id=100343

Reviewed by Sam Weinig.

Source/WebCore:

Simulated events were not allowed to be processed in the file input type in the DOMActivate handler.
This was a problem for accessibility clients which rely on simulated events.

The solution is to mark the UserGestureIndicator as definitely processing an event.

Test: accessibility/file-upload-button-with-axpress.html

* accessibility/AccessibilityObject.cpp:
(WebCore::AccessibilityObject::press):

LayoutTests:

* accessibility/file-upload-button-with-axpress-expected.txt: Added.
* accessibility/file-upload-button-with-axpress.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135680 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWeb Inspector: dispatch an event upon heap snapshot filter change
commit-queue@webkit.org [Mon, 26 Nov 2012 05:58:39 +0000 (05:58 +0000)]
Web Inspector: dispatch an event upon heap snapshot filter change
https://bugs.webkit.org/show_bug.cgi?id=103201

Patch by Nikita Vasilyev <me@elv1s.ru> on 2012-11-25
Reviewed by Pavel Feldman.

Dispatch heapSnapshotFilterChanged event so it can be received by third parties.

* inspector/front-end/HeapSnapshotView.js:
(WebInspector.HeapSnapshotView.prototype._changeFilter):
* inspector/front-end/UserMetrics.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135679 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemove deprecated load-done signal
kaustubh@motorola.com [Mon, 26 Nov 2012 05:12:29 +0000 (05:12 +0000)]
Remove deprecated load-done signal
https://bugs.webkit.org/show_bug.cgi?id=72712

Reviewed by Brent Fulgham.

Remove deprecated load-done signal and migrate to load-status.

* WebCoreSupport/FrameLoaderClientGtk.cpp:
(WebKit):
(WebKit::FrameLoaderClient::dispatchDidFinishLoad):
* webkit/webkitwebframe.cpp:
(webkit_web_frame_class_init):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135678 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago2012-11-25 Kent Tamura <tkent@chromium.org>
tkent@chromium.org [Mon, 26 Nov 2012 01:54:54 +0000 (01:54 +0000)]
2012-11-25  Kent Tamura  <tkent@chromium.org>

        [Chromium] Fix svn:mime-type for images

        * platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance-expected.png: Added property svn:mime-type.
        * platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance-rtl-expected.png: Added property svn:mime-type.
        * platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance-with-scroll-bar-expected.png: Added property svn:mime-type.
        * platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance-expected.png: Added property svn:mime-type.
        * platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance-rtl-expected.png: Added property svn:mime-type.
        * platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance-with-scroll-bar-expected.png: Added property svn:mime-type.
        * platform/chromium-win-xp/editing/selection/vertical-lr-ltr-extend-line-backward-br-expected.png: Modified property svn:mime-type.
        * platform/chromium-win-xp/editing/selection/vertical-lr-ltr-extend-line-forward-br-expected.png: Modified property svn:mime-type.
        * platform/chromium-win-xp/editing/selection/vertical-rl-ltr-extend-line-backward-br-expected.png: Modified property svn:mime-type.
        * platform/chromium-win-xp/editing/selection/vertical-rl-ltr-extend-line-backward-p-expected.png: Modified property svn:mime-type.
        * platform/chromium-win-xp/editing/selection/vertical-rl-ltr-extend-line-forward-br-expected.png: Modified property svn:mime-type.
        * platform/chromium-win-xp/editing/selection/vertical-rl-ltr-extend-line-forward-p-expected.png: Modified property svn:mime-type.
        * platform/chromium-win-xp/fast/text/drawBidiText-expected.png: Modified property svn:mime-type.
        * platform/chromium-win-xp/fast/text/international/001-expected.png: Modified property svn:mime-type.
        * platform/chromium-win-xp/fast/text/international/002-expected.png: Modified property svn:mime-type.
        * platform/chromium-win-xp/fast/text/international/bidi-mirror-he-ar-expected.png: Modified property svn:mime-type.
        * platform/chromium-win-xp/fast/text/international/wrap-CJK-001-expected.png: Modified property svn:mime-type.
        * platform/chromium-win-xp/fast/text/justify-ideograph-leading-expansion-expected.png: Modified property svn:mime-type.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135677 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Chromium] Add tests for datetime/datetime-local suggestion picker
commit-queue@webkit.org [Mon, 26 Nov 2012 01:00:13 +0000 (01:00 +0000)]
[Chromium] Add tests for datetime/datetime-local suggestion picker
https://bugs.webkit.org/show_bug.cgi?id=102888

Patch by Kunihiko Sakamoto <ksakamoto@chromium.org> on 2012-11-25
Reviewed by Kent Tamura.

Adding tests for <input type=datetime> and <input type=datetime-local> with datalist.

* platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance-expected.png: Added.
* platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance-rtl-expected.png: Added.
* platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance-with-scroll-bar-expected.png: Added.
* platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance-expected.png: Added.
* platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance-rtl-expected.png: Added.
* platform/chromium-mac/platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance-with-scroll-bar-expected.png: Added.
* platform/chromium-win/platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-key-operations-expected.txt: Added.
* platform/chromium-win/platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-key-operations-expected.txt: Added.
* platform/chromium/TestExpectations:
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance-rtl-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance-rtl.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance-with-scroll-bar-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance-with-scroll-bar.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-appearance.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-key-operations-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-key-operations.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-min-max-attribute-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-min-max-attribute.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-mouse-operations-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-mouse-operations.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-reset-value-after-reload-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-reset-value-after-reload.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-step-attribute-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetime-suggestion-picker-step-attribute.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance-rtl-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance-rtl.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance-with-scroll-bar-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance-with-scroll-bar.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-appearance.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-key-operations-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-key-operations.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-min-max-attribute-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-min-max-attribute.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-mouse-operations-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-mouse-operations.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-reset-value-after-reload-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-reset-value-after-reload.html: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-step-attribute-expected.txt: Added.
* platform/chromium/fast/forms/suggestion-picker/datetimelocal-suggestion-picker-step-attribute.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135676 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRefactoring: Move the content of HTMLInputElement::subtreeHasChanged to TextFieldInpu...
tkent@chromium.org [Mon, 26 Nov 2012 00:38:30 +0000 (00:38 +0000)]
Refactoring: Move the content of HTMLInputElement::subtreeHasChanged to TextFieldInputType
https://bugs.webkit.org/show_bug.cgi?id=103195

Reviewed by Kentaro Hara.

HTMLInputElement::subtreeHasChanged is called only if the input is
a text field. The code should be moved to TextFieldInputType.

No new tests. This should not change any behavior.

* html/HTMLInputElement.cpp:
- Remove unnecessary NumberInputType.h inclusion.
- Remove convertFromVisibleValue. It was used only by subtreeHasChanged.
(WebCore::HTMLInputElement::subtreeHasChanged):
Move the code to TextFieldInputType::subtreeHasChanged except
calculateAndAdjustDirectionality, which is a protected member of
HTMLElement.
* html/HTMLInputElement.h:
(HTMLInputElement): Remove convertFromVisibleValue.

* html/InputType.cpp:
Move convertFromVisibleValue to TextFieldInputType.
(WebCore::InputType::subtreeHasChanged):
Add ASSERT_NOT_REACHED.
* html/InputType.h:
(InputType): Remove convertFromVisibleValue.

* html/TextFieldInputType.cpp:
(WebCore::TextFieldInputType::convertFromVisibleValue):
Moved from InputType.
(WebCore::TextFieldInputType::subtreeHasChanged):
Moved from HTMLInputElement. A latter part is moved to
didSetValueByUserEdit to be hooked by SearchInputType.
(WebCore::TextFieldInputType::didSetValueByUserEdit):
Moved from HTMLInputElement::subtreeHasChanged, and clean up the code.
* html/TextFieldInputType.h:
(TextFieldInputType):
- Move convertFromVisibleValue from InputType.
- Add didSetValueByUserEdit and subtreeHasChanged.

* html/SearchInputType.cpp:
(WebCore::SearchInputType::didSetValueByUserEdit):
Renamed from subtreeHasChanged, and calls TextFieldInputType::didSetValueByUserEdit.
* html/SearchInputType.h:
(SearchInputType): Rename subtreeHasChanged to didSetValueByUserEdit.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135675 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[V8] Move WorkerExecutionContextProxy::initializeIfNeeded() to V8Initializer
haraken@chromium.org [Mon, 26 Nov 2012 00:21:24 +0000 (00:21 +0000)]
[V8] Move WorkerExecutionContextProxy::initializeIfNeeded() to V8Initializer
https://bugs.webkit.org/show_bug.cgi?id=103061

Reviewed by Adam Barth.

This is an incremental step to remove WorkerExecutionContextProxy.
This patch moves WorkerExecutionContextProxy::initializeIfNeeded() to V8Initializer.
This patch also renames methods so that the names become consistent
between the main thread and workers.

No tests. No change in behavior.

* bindings/v8/V8Initializer.cpp:
(WebCore::reportFatalErrorInMainThread):
(WebCore::messageHandlerInMainThread):
(WebCore::failedAccessCheckCallbackInMainThread):
(WebCore::V8Initializer::initializeMainThreadIfNeeded):
(WebCore::reportFatalErrorInWorker):
(WebCore):
(WebCore::messageHandlerInWorker):
(WebCore::V8Initializer::initializeWorkerIfNeeded):
* bindings/v8/V8Initializer.h:
(V8Initializer):
* bindings/v8/WorkerContextExecutionProxy.cpp:
(WebCore::WorkerContextExecutionProxy::WorkerContextExecutionProxy):
* bindings/v8/WorkerContextExecutionProxy.h:
(WorkerContextExecutionProxy):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135674 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Refactor RenderThemeEfl::ThemePartCacheEntry::reuse()
commit-queue@webkit.org [Sun, 25 Nov 2012 19:19:08 +0000 (19:19 +0000)]
[EFL] Refactor RenderThemeEfl::ThemePartCacheEntry::reuse()
https://bugs.webkit.org/show_bug.cgi?id=103189

Patch by Christophe Dumez <christophe.dumez@intel.com> on 2012-11-25
Reviewed by Kenneth Rohde Christiansen.

Check if the entry size and type changed in
RenderThemeEfl::ThemePartCacheEntry::reuse() to avoid
useless processing if one of them did not change.

Remove useless call to cairo_surface_finish() since
we are using a smart pointer for the surface.

Resize the edge object *after* loading its content
from the theme file as it seems more logical this
way.

No new tests, no behavior change for layout tests.

* platform/efl/RenderThemeEfl.cpp:
(WebCore::RenderThemeEfl::ThemePartCacheEntry::reuse):
* platform/efl/RenderThemeEfl.h:
(ThemePartCacheEntry):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135673 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] Get rid of resetBitmapZoomScale()
zhajiang@rim.com [Sun, 25 Nov 2012 19:06:22 +0000 (19:06 +0000)]
[BlackBerry] Get rid of resetBitmapZoomScale()
https://bugs.webkit.org/show_bug.cgi?id=103200

Patch by Jacky Jiang  <zhajiang@rim.com>.
Reviewed by Rob Buis.
Internally reviewed by Gen Mak.

PR: 235707
Right now, resetBitmapZoomScale() takes an unused paramter and
dispatches a message to end the bitmap zooming. However, the
interaction has its own state machine to start bitmap zooming when it
is active and end bitmap zooming when it is inactive. This is
expecially for pinch zoom and scroll. From WebKit side, block zoom
which will call zoomBlock when the bitmap zoom finished doesn't even
need to dispatch the useless ending message to ZoomHandler.
resetBitmapZoomScale() is basically a noop for block zoom. And likewise
that is useless for zoomToInitialScaleOnLoad(). It can also potentially
cause problem if it is used somewhere else in the future and
incidentally breaks the interaction's own state machine. So just remove
this completely.

* Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::zoomToInitialScaleOnLoad):
(BlackBerry::WebKit::WebPagePrivate::zoomBlock):
* Api/WebPageClient.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135672 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRename DynamicNodeList to LiveNodeList
rniwa@webkit.org [Sun, 25 Nov 2012 18:51:36 +0000 (18:51 +0000)]
Rename DynamicNodeList to LiveNodeList
https://bugs.webkit.org/show_bug.cgi?id=103197

Reviewed by Ojan Vafai.

Rename DynamicNodeList to LiveNodeList to match the terminology used in DOM4 working draft:
http://www.w3.org/TR/2012/WD-dom-20120405/#concept-collection-live
"A collection (either NodeList or HTMLCollection) can be either live or static".

Also rename DynamicNodeListCacheBase to LiveNodeListBase, and merge DynamicSubtreeNodeList
into LiveNodeList (old DynamicNodeList) now that the only difference between those two classes
is the former calling registerNodeListCache and unregisterNodeListCache on Document.

This patch completes the series of simplification of NodeList/HTMLCollection classes.

* CMakeLists.txt:
* GNUmakefile.list.am:
* Target.pri:
* WebCore.gypi:
* WebCore.xcodeproj/project.pbxproj:
* bindings/js/JSHTMLOptionsCollectionCustom.cpp:
(WebCore::getNamedItems):
* bindings/js/JSNodeListCustom.cpp:
(WebCore::JSNodeListOwner::isReachableFromOpaqueRoots):
* bindings/v8/custom/V8NodeListCustom.cpp:
(WebCore::V8NodeList::opaqueRootForGC):
* dom/ChildNodeList.cpp:
(WebCore::ChildNodeList::ChildNodeList):
(WebCore::ChildNodeList::nodeMatches):
* dom/ChildNodeList.h:
* dom/ClassNodeList.cpp:
(WebCore::ClassNodeList::ClassNodeList):
* dom/ClassNodeList.h:
* dom/DOMAllInOne.cpp:
* dom/Document.cpp:
(WebCore::Document::registerNodeListCache):
(WebCore::Document::unregisterNodeListCache):
(WebCore):
* dom/Document.h:
(WebCore):
(Document):
* dom/DynamicNodeList.cpp: Removed.
* dom/DynamicNodeList.h: Removed.
* dom/LiveNodeList.cpp: Copied from Source/WebCore/dom/DynamicNodeList.cpp.
(WebCore::LiveNodeListBase::rootNode):
(WebCore::LiveNodeListBase::invalidateCache):
(WebCore::LiveNodeListBase::invalidateIdNameCacheMaps):
(WebCore::LiveNodeListBase::reportMemoryUsage):
(WebCore::LiveNodeList::namedItem):
* dom/LiveNodeList.h: Copied from Source/WebCore/dom/DynamicNodeList.h.
(WebCore::LiveNodeListBase::LiveNodeListBase):
(WebCore::LiveNodeListBase::shouldInvalidateTypeOnAttributeChange):
(WebCore::LiveNodeList::LiveNodeList):
(WebCore::LiveNodeList::~LiveNodeList):
(LiveNodeList):
* dom/MicroDataItemList.cpp:
(WebCore::MicroDataItemList::MicroDataItemList):
* dom/MicroDataItemList.h:
* dom/NameNodeList.cpp:
(WebCore::NameNodeList::NameNodeList):
* dom/NameNodeList.h:
* dom/Node.cpp:
(WebCore::shouldInvalidateNodeListCachesForAttr):
(WebCore::Document::invalidateNodeListCaches):
* dom/Node.h:
(WebCore):
* dom/NodeList.h:
(WebCore::NodeList::isLiveNodeList):
* dom/NodeRareData.h:
(NodeListsNodeData):
(WebCore::NodeListsNodeData::removeCacheWithAtomicName):
(WebCore::NodeListsNodeData::removeCacheWithName):
(WebCore::NodeListsNodeData::removeCacheWithQualifiedName):
(WebCore::NodeListsNodeData::adoptTreeScope):
* dom/PropertyNodeList.cpp:
(WebCore::PropertyNodeList::PropertyNodeList):
* dom/PropertyNodeList.h:
* dom/TagNodeList.cpp:
(WebCore::TagNodeList::TagNodeList):
* dom/TagNodeList.h:
* html/HTMLCollection.cpp:
(WebCore::HTMLCollection::HTMLCollection):
(WebCore::LiveNodeListBase::iterateForNextNode):
(WebCore::LiveNodeListBase::itemBeforeOrAfter):
(WebCore::LiveNodeListBase::itemBefore):
(WebCore::LiveNodeListBase::itemAfter):
(WebCore::LiveNodeListBase::isLastItemCloserThanLastOrCachedItem):
(WebCore::LiveNodeListBase::isFirstItemCloserThanCachedItem):
(WebCore::LiveNodeListBase::setItemCache):
(WebCore::LiveNodeListBase::length):
(WebCore::LiveNodeListBase::item):
(WebCore::LiveNodeListBase::itemBeforeOrAfterCachedItem):
* html/HTMLCollection.h:
* html/LabelsNodeList.cpp:
(WebCore::LabelsNodeList::LabelsNodeList):
* html/LabelsNodeList.h:
* html/RadioNodeList.cpp:
(WebCore::RadioNodeList::RadioNodeList):
* html/RadioNodeList.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135671 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoChanging position:relative to position:static results in mis-positioned div
robert@webkit.org [Sun, 25 Nov 2012 18:35:10 +0000 (18:35 +0000)]
Changing position:relative to position:static results in mis-positioned div
https://bugs.webkit.org/show_bug.cgi?id=26397

Reviewed by Ojan Vafai.

Source/WebCore:

When a block changes position from relative to static it is no longer the containing block for any
positioned objects it may have. If any of those positioned objects actually have a position specified
they are going to need a layout as their new containing block will likely have a different location they
need to offset from. Positioned objects without a specified position always get a layout anyway
in layoutPositionedObjects() so no need to worry about them in this situation.

Test: fast/block/abspos-child-container-changes-from-relative-to-static-expected.html

* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::styleWillChange):
(WebCore::RenderBlock::layoutPositionedObjects):
(WebCore::RenderBlock::removePositionedObjects):
* rendering/RenderBlock.h:
(RenderBlock):

LayoutTests:

* fast/block/abspos-child-container-changes-from-relative-to-static-expected.html: Added
* fast/block/abspos-child-container-changes-from-relative-to-static.html: Added

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135670 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[WK2] TiledBackingStore: page contents is scaled wrongly
commit-queue@webkit.org [Sun, 25 Nov 2012 18:30:14 +0000 (18:30 +0000)]
[WK2] TiledBackingStore: page contents is scaled wrongly
https://bugs.webkit.org/show_bug.cgi?id=103090

Patch by Mikhail Pozdnyakov <mikhail.pozdnyakov@intel.com> on 2012-11-25
Reviewed by Kenneth Rohde Christiansen.

Source/WebKit2:

Before this change the page contents scaling in PageViewportController was
defined all the times by m_rawAttributes.initialScale. If initial scale had not
been specified in the viewport meta tag it was set to m_minimumScaleToFit inside
PageViewportController::didChangeViewportAttributes().

The problem was that m_minimumScaleToFit could have wrong value as contents size
might have not be updated by the time PageViewportController::didChangeViewportAttributes()
was invoked.

The solution is to use m_minimumScaleToFit for contents scaling if initial scale
is not specified in the viewport meta tag, as it is updated all the time.
Also a flag m_initiallyFitToViewport is added to PageViewportController to detect
whether m_minimumScaleToFit should be used for scaling.

* UIProcess/PageViewportController.cpp:
(WebKit::PageViewportController::PageViewportController):
(WebKit::PageViewportController::didChangeContentsSize):
(WebKit::PageViewportController::pageTransitionViewportReady):
(WebKit::PageViewportController::didChangeViewportAttributes):
* UIProcess/PageViewportController.h:
(PageViewportController):

LayoutTests:

Added new layout test to verify that viewport 'width' descriptor does not
affect contents width of next loaded page.

* css3/device-adapt/resources/check-contents-width.html: Added.
* css3/device-adapt/viewport-width-not-affecting-next-page-expected.txt: Added.
* css3/device-adapt/viewport-width-not-affecting-next-page.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135669 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoMake renderer construction less generic
antti@apple.com [Sun, 25 Nov 2012 16:53:44 +0000 (16:53 +0000)]
Make renderer construction less generic
https://bugs.webkit.org/show_bug.cgi?id=103175

Reviewed by Ojan Vafai.

The renderer construction code currently operates on Nodes and is very generic. In reality
only Element and Text nodes can have renderers and the Text case is much simpler.

This patch separates the Text and Element renderer construction paths and makes it statically
known that other Node types can't have renderers. Less generic code is less branchy and enables
further optimizations.

* dom/CharacterData.cpp:
(WebCore::CharacterData::parserAppendData):
(WebCore::CharacterData::setDataAndUpdate):
(WebCore):
(WebCore::CharacterData::rendererIsNeeded):
(WebCore::CharacterData::createRenderer):

    Only Text subclass of CharacterData can have renderers.

* dom/CharacterData.h:
(CharacterData):
* dom/ContainerNode.h:
(WebCore::ContainerNode::childShouldCreateRenderer):

    Move childShouldCreateRenderer from Node to ContainerNode.

(ContainerNode):
* dom/Element.cpp:
(WebCore::Element::rendererIsNeeded):
(WebCore):
(WebCore::Element::attach):
(WebCore::Element::childShouldCreateRenderer):
* dom/Element.h:

    Move rendererIsNeeded and createRenderer from Node to Element.

(Element):
* dom/Node.cpp:
(WebCore::Node::attach):
(WebCore):
* dom/Node.h:
(Node):
* dom/NodeRenderingContext.cpp:
(WebCore::NodeRenderingContext::createRendererForElementIfNeeded):
(WebCore::NodeRenderingContext::createRendererForTextIfNeeded):

    Separate the Element and Text renderer creation paths. Both are less branchy.
    The Text path is much simpler and avoids a bunch of virtual calls.

(WebCore):
* dom/NodeRenderingContext.h:
(NodeRenderingContext):
* dom/Text.cpp:
(WebCore::Text::textRendererIsNeeded):
(WebCore::Text::createTextRendererIfNeeded):
(WebCore):
(WebCore::Text::createTextRenderer):
(WebCore::Text::attach):
(WebCore::Text::updateTextRenderer):
* dom/Text.h:

        Add non-virtual Text specific functions.

(WebCore):
(Text):
* rendering/RenderBlock.cpp:
(WebCore::RenderBlock::clone):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135668 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r135493): HTMLCollection and DynamicNodeList have two vtable pointers
rniwa@webkit.org [Sun, 25 Nov 2012 08:01:28 +0000 (08:01 +0000)]
REGRESSION(r135493): HTMLCollection and DynamicNodeList have two vtable pointers
https://bugs.webkit.org/show_bug.cgi?id=103096

Reviewed by Sam Weinig.

Co-authored by Ilya Tikhonovsky.

Made DynamicNodeListCacheBase inherit from NodeList. While HTMLCollection doesn't inherit
from NodeList in IDL, it makes a lot of sense for the C++ implementation to do so via
DynamicNodeListCacheBase since HTMLCollection and live NodeList share a lot of code in
DynamicNodeListCacheBase.

This lets remove proxies for item() and length() in DynamicNodeList and HTMLCollection
and directly implement them in DynamicNodeListCacheBase which used to provide itemCommon()
and lengthCommon().

Also renamed NodeList::itemWithName() to NodeList::namedItem() to match HTMLCollection's
naming convention.

Finally, removed reportMemoryUsage in DynamicNodeList::reportMemoryUsage since DynamicNodeList
now uses single inheritance.

* bindings/js/JSNodeListCustom.cpp:
(WebCore::JSNodeList::canGetItemsForName): Calls namedItem, which has been renamed from
itemWithName.
(WebCore::JSNodeList::nameGetter): Ditto.
* bindings/v8/custom/V8NodeListCustom.cpp:
(WebCore::V8NodeList::namedPropertyGetter): Ditto.
* bindings/v8/custom/V8NamedNodesCollection.cpp:
(WebCore::V8NamedNodesCollection::namedItem): Renamed from itemWithName.
* bindings/v8/custom/V8NamedNodesCollection.h:
* dom/ChildNodeList.cpp:
(WebCore::ChildNodeList::nodeMatches): Updated comment.
* dom/DynamicNodeList.cpp:
(WebCore::DynamicNodeList::namedItem): Renamed from itemWithName.
* dom/DynamicNodeList.h:
(DynamicNodeListCacheBase): Inhertis from NodeList and renamed lengthCommon and itemCommon
to virtual length and item respectively.
(DynamicNodeList): Now inherits from just DynamicNodeListCacheBase instead of NodeList
and DynamicNodeListCacheBase since the former now inhertis from NodeList. Also removed
length() and item() since they're implemented in DynamicNodeListCacheBase now and renamed
itemWithName() to namedItem() to match HTMLCollection's naming convention.
* dom/NodeList.h:
(NodeList::namedItem): Renamed from itemWithName. Note that this member function is not
exposed via IDL. It's merely used in the binding code.
* dom/StaticHashSetNodeList.cpp:
(WebCore::StaticHashSetNodeList::namedItem): Ditto.
* dom/StaticHashSetNodeList.h:
(StaticHashSetNodeList::namedItem): Ditto.
* dom/StaticNodeList.cpp:
(WebCore::StaticNodeList::namedItem): Ditto.
* dom/StaticNodeList.h:
(StaticNodeList::namedItem): Ditto.
* html/HTMLCollection.cpp:
(WebCore::DynamicNodeListCacheBase::length): Renamed from lengthCommon.
(WebCore::DynamicNodeListCacheBase::item): Renamed from itemCommon.
* html/HTMLCollection.h:
(HTMLCollection): Inherits from DynamicNodeListCacheBase since DynamicNodeListCacheBase
is already RefCount'ed and NodeList, from which DynamicNodeListCacheBase inherits,
inherits from ScriptWrappable.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135667 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoChromium should use TCMalloc on Mac to go fast
abarth@webkit.org [Sun, 25 Nov 2012 07:25:16 +0000 (07:25 +0000)]
Chromium should use TCMalloc on Mac to go fast
https://bugs.webkit.org/show_bug.cgi?id=102866

Reviewed by Eric Seidel.

This patch enables TCMalloc for some WebKit objects. It improves
dom-modify on Mac by 5-10%.

On non-Mac platforms, Chromium already uses TCMalloc throughout the
project. Unfortunately, we haven't yet figured out how to turn TCMalloc
on globally on Mac because the approach we use for other platforms makes
some OS X APIs sad.

The next best thing would be to enable TCMalloc for WebKit by
overriding the global new and delete operator, as is done on apple-mac
and other platforms. Unfortunately, we cannot use that approach either
because the Chromium WebKit API is not memory tight.

Fortunately, WebKit has a mechanism for selectively enabling TCMalloc
for a selection of objects by overriding the new and delete operators
on those objects. This patch opts chromium-mac into that scheme by
setting the appropriate preprocessor macros.

* WTF.gyp/WTF.gyp:
* wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135666 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoSubstitute "use" for "user" in sentence about naming convention for
dbates@webkit.org [Sun, 25 Nov 2012 04:49:14 +0000 (04:49 +0000)]
Substitute "use" for "user" in sentence about naming convention for
enum members so that it reads well.

* coding/coding-style.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135665 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] Find-on-page keeps stale pointer
commit-queue@webkit.org [Sun, 25 Nov 2012 04:13:29 +0000 (04:13 +0000)]
[BlackBerry] Find-on-page keeps stale pointer
https://bugs.webkit.org/show_bug.cgi?id=103190

Patch by Andy Chen <andchen@rim.com> on 2012-11-24
Reviewed by Rob Buis.

Find on page: pendingScopingEffort stores stale pointer when frames unload,
we need to cancel those scoping efforts.
PR 250807.

Internally reviewed by Yongxin Dai and Mike Fenton.

* WebKitSupport/InPageSearchManager.cpp:
(InPageSearchManager::DeferredScopeStringMatches):
(BlackBerry::WebKit::InPageSearchManager::frameUnloaded):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135664 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Refactor RenderThemeEfl::paintThemePart()
commit-queue@webkit.org [Sun, 25 Nov 2012 02:52:20 +0000 (02:52 +0000)]
[EFL] Refactor RenderThemeEfl::paintThemePart()
https://bugs.webkit.org/show_bug.cgi?id=103192

Patch by Christophe Dumez <christophe.dumez@intel.com> on 2012-11-24
Reviewed by Kenneth Rohde Christiansen.

Refactor RenderThemeEfl::paintThemePart() so that:
- Some C'ism is removed
- RTL related code is moved to applyEdjeRTLState()
- evas_render() is called instead of evas_render_updates()
  to avoid creating uselessly update rects.

No new tests, no behavior change for layout tests.

* platform/efl/RenderThemeEfl.cpp:
(WebCore::RenderThemeEfl::applyEdjeRTLState):
(WebCore):
(WebCore::RenderThemeEfl::paintThemePart):
* platform/efl/RenderThemeEfl.h:
(RenderThemeEfl):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135663 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Layout test comes to crash on WK1
commit-queue@webkit.org [Sun, 25 Nov 2012 02:38:09 +0000 (02:38 +0000)]
[EFL] Layout test comes to crash on WK1
https://bugs.webkit.org/show_bug.cgi?id=103170

Patch by Viatcheslav Ostapenko <v.ostapenko@samsung.com> on 2012-11-24
Reviewed by Kenneth Rohde Christiansen.

Switch off accelerated compositing for WK1 because GraphicsContext3D is
missing direct rendering implementation that is required for EFL WK1.

* ewk/ewk_view.cpp:
(_ewk_view_priv_new):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135662 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[chromium] Remove traces of MediaControlRootElement
commit-queue@webkit.org [Sun, 25 Nov 2012 02:22:20 +0000 (02:22 +0000)]
[chromium] Remove traces of MediaControlRootElement
https://bugs.webkit.org/show_bug.cgi?id=103174

Patch by Silvia Pfeiffer <silviapf@chromium.org> on 2012-11-24
Reviewed by Adam Barth.

Now that nothing is using MediaControlRootElement any more,
we can remove the FIXME and MediaControlRootElement enum left
over from bug 88871.

* public/WebContextMenuData.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135661 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL][WK2] Stop the mess with DECLARE_EWK_VIEW_CALLBACK arg type definition
commit-queue@webkit.org [Sun, 25 Nov 2012 02:18:08 +0000 (02:18 +0000)]
[EFL][WK2] Stop the mess with DECLARE_EWK_VIEW_CALLBACK arg type definition
https://bugs.webkit.org/show_bug.cgi?id=103187

Patch by Mikhail Pozdnyakov <mikhail.pozdnyakov@intel.com> on 2012-11-24
Reviewed by Kenneth Rohde Christiansen.

Now the client should pass to DECLARE_EWK_VIEW_CALLBACK macro exacly the
same arg type s/he wants to have in the Ewk_View callback.

* UIProcess/API/efl/EwkViewCallbacks.h:
(EvasObjectHolder):
(EwkViewCallbacks::EvasObjectHolder::EvasObjectHolder):
(EwkViewCallbacks):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135660 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[WinCairo] Avoid access violation when frame is NULL.
bfulgham@webkit.org [Sat, 24 Nov 2012 23:05:07 +0000 (23:05 +0000)]
[WinCairo] Avoid access violation when frame is NULL.
https://bugs.webkit.org/show_bug.cgi?id=68753

BitmapImage::drawFrameMatchingSourceSize causes an access violation
if BitmapImage::frameAtIndex returns NULL. (Found by David Delaune).

Reviewed by Simon Fraser.

* platform/graphics/win/ImageCairoWin.cpp:
(WebCore::BitmapImage::drawFrameMatchingSourceSize): Check for null
cairo_surface_t pointer and avoid dereferencing.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135659 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Refactor RenderThemeEfl::getThemePartFromCache()
commit-queue@webkit.org [Sat, 24 Nov 2012 22:25:11 +0000 (22:25 +0000)]
[EFL] Refactor RenderThemeEfl::getThemePartFromCache()
https://bugs.webkit.org/show_bug.cgi?id=103186

Patch by Christophe Dumez <christophe.dumez@intel.com> on 2012-11-24
Reviewed by Kenneth Rohde Christiansen.

Refactor getThemePartFromCache() so that a list is used
for the cache instead of a vector. The function moves
items to the head of the container which is much more
efficient with a list than a vector.

The list ordering is also slightly altered so that the
cached parts that were used last come first. Previously,
they were sorted by creation time which may cause an
item to be removed from the cache because it is older
even if is was used recently.

This patch also gets rid of some code duplication at the
end of the function.

No new tests, no behavior change.

* platform/efl/RenderThemeEfl.cpp:
(WebCore::RenderThemeEfl::ThemePartCacheEntry::create): Return
a smart pointer instead of a raw pointer.
(WebCore::RenderThemeEfl::getThemePartFromCache):
(WebCore::RenderThemeEfl::clearThemePartCache): Rename
flushThemePartCache to clearThemePartCache for clarity.
(WebCore::RenderThemeEfl::loadTheme):
(WebCore::RenderThemeEfl::RenderThemeEfl):
(WebCore::RenderThemeEfl::~RenderThemeEfl):
* platform/efl/RenderThemeEfl.h:
(ThemePartCacheEntry):
(RenderThemeEfl):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135658 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] RTSP should use the same MediaDocument as HTTP videos.
commit-queue@webkit.org [Sat, 24 Nov 2012 20:34:40 +0000 (20:34 +0000)]
[BlackBerry] RTSP should use the same MediaDocument as HTTP videos.
https://bugs.webkit.org/show_bug.cgi?id=103185
RIM PR 250114

Patch by Liam Quinn <lquinn@rim.com> on 2012-11-24
Reviewed by George Staikos.

Make our MediaPlayer implementation indicate that it supports RTSP. That way, DOMImplementation::createDocument will create the same MediaDocument as is used for HTTP videos.

* platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:
(WebCore::MediaPlayerPrivate::supportsType):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135657 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoPut computed style 'font-family' values in the CSSValuePool.
akling@apple.com [Sat, 24 Nov 2012 20:31:27 +0000 (20:31 +0000)]
Put computed style 'font-family' values in the CSSValuePool.
<http://webkit.org/b/103184>

Reviewed by Ojan Vafai.

Use CSSValuePool::createFontFamilyValue() when handing out 'font-family' values from CSSComputedStyleDeclaration.
This avoids creating extra CSSPrimitiveValue objects since we already have these in the pool from parsing.
Also added a max capacity to the font-family cache (using random eviction.)

* css/CSSComputedStyleDeclaration.cpp:
(WebCore::valueForFamily):
* css/CSSValuePool.cpp:
(WebCore::CSSValuePool::createFontFamilyValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135656 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] Text not selected when input field focused.
commit-queue@webkit.org [Sat, 24 Nov 2012 19:35:56 +0000 (19:35 +0000)]
[BlackBerry] Text not selected when input field focused.
https://bugs.webkit.org/show_bug.cgi?id=103182

Patch by Genevieve Mak <gmak@rim.com> on 2012-11-24
Reviewed by George Staikos.

Reviewed Internally by: Mike Fenton, Eli Fidler
PR #250163

* WebKitSupport/SelectionHandler.cpp:
(BlackBerry::WebKit::SelectionHandler::selectObject):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135655 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed gardening: updated expectations for failures in:
rjkroege@chromium.org [Sat, 24 Nov 2012 19:24:12 +0000 (19:24 +0000)]
Unreviewed gardening: updated expectations for failures in:
* http/tests/local/drag-over-remote-content.html
https://bugs.webkit.org/show_bug.cgi?id=103181
* media/video-seek-past-end-playing.html
https://bugs.webkit.org/show_bug.cgi?id=103183

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135654 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAdd the --order option to NRWT
zandobersek@gmail.com [Sat, 24 Nov 2012 17:15:11 +0000 (17:15 +0000)]
Add the --order option to NRWT
https://bugs.webkit.org/show_bug.cgi?id=102229

Reviewed by Dirk Pranke.

The --order option controls in what order the gathered tests will be run.
There are 3 options:
- natural, the tests are sorted in natural order,
- random, the tests are sorted in random order,
- none, the tests are kept in the order in which they were specified through
arguments or the test list.

The natural option is the default one and maintains the current behavior.

The none option is primarily intended to be used with the --test-list option,
with the test list file expected to list specific test names.
When a directory is specified as one of the paths from under which tests should
be gathered and the none option is used, that path will expand to all the tests
under it, with these tests sorted in the natural order.

Using --order=random duplicated the behavior of the --randomize-order option, so
the latter is removed.

This patch also incorporates the change with which makes it is possible to specify
a test via arguments or the test list multiple tests and then have the test run
that many times.

* Scripts/webkitpy/common/find_files.py:
(find): Accept an optional sorting key that is used to sort files found in the directories
that the given paths specify.
(_normalized_find): Both paths that are then searched for the tests and the tests that
were found are now kept in lists. This makes it possible to run a test multiple times
by specifying it multiple times through arguments or the test list file.
(_normalized_find.sort_by_directory_key):
* Scripts/webkitpy/layout_tests/controllers/layout_test_finder.py:
(LayoutTestFinder.find_tests): Keep the paths where we should search for tests in a list.
(LayoutTestFinder.skip_tests): Subtracting from a set of paths is necessary.
* Scripts/webkitpy/layout_tests/controllers/layout_test_runner.py:
(LayoutTestRunner.__init__): The Sharder class no longer needs a test path separator.
(Sharder.__init__): Ditto.
(Sharder._resize_shards):
* Scripts/webkitpy/layout_tests/controllers/layout_test_runner_unittest.py:
(SharderTests.get_shards): Modify the Shader initializing as required.
(SharderTests.test_multiple_locked_shards): The natural sorting and test name key unit tests
are now moved to the base Port unit tests.
* Scripts/webkitpy/layout_tests/controllers/manager.py:
(Manager._prepare_lists): Keep the test names in a list. Order that list as specified by
the new --order option.
* Scripts/webkitpy/layout_tests/port/base.py:
(Port.tests): This method should now return a list of all the tests found in the specified
paths.
(Port._real_tests): A list is required, so no conversion to a set is made.
(Port):
(Port.test_key): Place this method here for now, but it should really move into a separate
module.
(Port._natural_sort_key): Ditto.
(Port._natural_sort_key.tryint):
(Port._virtual_tests): A list is required, so the function now operates on that rather than
on a set.
* Scripts/webkitpy/layout_tests/port/base_unittest.py:
(PortTest.test_find_with_skipped_directories_2): Adjust the test case to check for an empty
list rather than an empty set.
(PortTest.test_dont_require_http_server):
(NaturalCompareTest): The natural sorting unit tests are placed here for now.
(NaturalCompareTest.assert_cmp):
(NaturalCompareTest.test_natural_compare):
(KeyCompareTest): The test name key unit tests are placed here for now.
(KeyCompareTest.assert_cmp):
(KeyCompareTest.test_test_key):
* Scripts/webkitpy/layout_tests/run_webkit_tests.py:
(parse_args): Remove the --randomize-order option, it's now possible to replicate its behavior
by using --order=random.
* Scripts/webkitpy/layout_tests/run_webkit_tests_integrationtest.py: Add unit tests covering
various aspects of this change.
(MainTest.test_natural_order):
(MainTest):
(MainTest.test_natural_order_test_specified_multiple_times):
(MainTest.test_random_order):
(MainTest.test_random_order_test_specified_multiple_times):
(MainTest.test_no_order):
(MainTest.test_no_order_test_specified_multiple_times):
(MainTest.test_no_order_with_directory_entries_in_natural_order):
* Scripts/webkitpy/tool/commands/queries.py:
(PrintExpectations.execute): Operate on a set of tests when printing out expectations.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135653 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL][WK2] Check errors in EwkView callbacks usage at compile time
commit-queue@webkit.org [Sat, 24 Nov 2012 16:46:03 +0000 (16:46 +0000)]
[EFL][WK2] Check errors in EwkView callbacks usage at compile time
https://bugs.webkit.org/show_bug.cgi?id=103177

Patch by Mikhail Pozdnyakov <mikhail.pozdnyakov@intel.com> on 2012-11-24
Reviewed by Kenneth Rohde Christiansen.

Errors in EwkView callbacks usage are now checked at compile time.

* UIProcess/API/efl/EwkViewCallbacks.h:
(CallBack):
(EwkViewCallbacks::CallBack::CallBack):
(EwkViewCallbacks::CallBack::call):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135652 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, rolling out r135648 and r135649.
zeno.albisser@digia.com [Sat, 24 Nov 2012 15:32:25 +0000 (15:32 +0000)]
Unreviewed, rolling out r135648 and r135649.
http://trac.webkit.org/changeset/135648
http://trac.webkit.org/changeset/135649
https://bugs.webkit.org/show_bug.cgi?id=103176

This change cannot be built with a qt-minimal configuration.
(Requested by zalbisser on #webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-11-24

* WebCoreSupport/PageClientQt.cpp:
(WebCore::PageClientQGraphicsWidget::setRootGraphicsLayer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135651 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoIt's confusing that return values of 'bool Node::dispatchEvent(...)' and 'bool Node...
hayato@chromium.org [Sat, 24 Nov 2012 15:11:55 +0000 (15:11 +0000)]
It's confusing that return values of 'bool Node::dispatchEvent(...)' and 'bool Node::dispatchMouseEvent(..)' have the opposite meanings.
https://bugs.webkit.org/show_bug.cgi?id=103058

Reviewed by Ojan Vafai.

Refactor MouseEventDispatchMediator::dispatchEvent() and
EventHandler::dispatchMouseEvent() so that they return bool
which has the same meaning of the return value of
Node::dispatchEvent().

No tests. No change in behavior.

* dom/MouseEvent.cpp:
(WebCore::MouseEventDispatchMediator::dispatchEvent):
* page/EventHandler.cpp:
(WebCore::EventHandler::handleMousePressEvent):
(WebCore::EventHandler::handleMouseDoubleClickEvent):
(WebCore::EventHandler::handleMouseMoveEvent):
(WebCore::EventHandler::handleMouseReleaseEvent):
(WebCore::EventHandler::dispatchMouseEvent):
(WebCore::EventHandler::sendContextMenuEvent):
(WebCore::EventHandler::sendContextMenuEventForKey):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135650 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed Qt build fix:
zeno.albisser@digia.com [Sat, 24 Nov 2012 15:11:27 +0000 (15:11 +0000)]
Unreviewed Qt build fix:

Make sure QGLWidget definition is available.

* WebCoreSupport/PageClientQt.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135649 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Qt] Make sure the QGLWidget context is current when creating the TextureMapper.
zeno.albisser@digia.com [Sat, 24 Nov 2012 14:41:09 +0000 (14:41 +0000)]
[Qt] Make sure the QGLWidget context is current when creating the TextureMapper.
https://bugs.webkit.org/show_bug.cgi?id=103142

When creating the TextureMapperGL for WK1 we have to make sure
that the GL context provided by the QGLWidget is current.
Otherwise the GraphicsContext3DQt created by TextureMapperGL will pick up
the wrong pointer by calling QOpenGLContext::currentContext().

Reviewed by Kenneth Rohde Christiansen.

* WebCoreSupport/PageClientQt.cpp:
(WebCore::PageClientQGraphicsWidget::setRootGraphicsLayer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135648 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed gardening.
zandobersek@gmail.com [Sat, 24 Nov 2012 11:28:31 +0000 (11:28 +0000)]
Unreviewed gardening.

* platform/efl/TestExpectations:
* platform/gtk/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135647 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, rolling out r135600.
commit-queue@webkit.org [Sat, 24 Nov 2012 10:50:44 +0000 (10:50 +0000)]
Unreviewed, rolling out r135600.
http://trac.webkit.org/changeset/135600
https://bugs.webkit.org/show_bug.cgi?id=103173

It made EFL API test is failed (Requested by gyuyoung_mbp on
#webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-11-24

* TestWebKitAPI/PlatformEfl.cmake:
* TestWebKitAPI/efl/PlatformUtilities.cpp:
(TestWebKitAPI::Util::sleep):
* TestWebKitAPI/efl/PlatformWebView.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135646 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed gardening. Too many tests come to crashe after build 7830 on
gyuyoung.kim@samsung.com [Sat, 24 Nov 2012 10:05:47 +0000 (10:05 +0000)]
Unreviewed gardening. Too many tests come to crashe after build 7830 on
EFL Linux 64-bit Release bot.

* platform/efl-wk1/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135645 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFix unused parameter compile warnings
commit-queue@webkit.org [Sat, 24 Nov 2012 06:40:29 +0000 (06:40 +0000)]
Fix unused parameter compile warnings
https://bugs.webkit.org/show_bug.cgi?id=103167

Patch by Kangil Han <kangil.han@samsung.com> on 2012-11-23
Reviewed by Gyuyoung Kim.

Remove compile warning messages by omitting parameter name and using UNUSED_PARAM.

* accessibility/atk/WebKitAccessibleHyperlink.cpp:
(webkitAccessibleHyperlinkIsSelectedLink):
* accessibility/atk/WebKitAccessibleInterfaceText.cpp:
(webkitAccessibleTextGetTextAfterOffset):
(webkitAccessibleTextGetTextAtOffset):
(webkitAccessibleTextGetTextBeforeOffset):
(webkitAccessibleTextGetOffsetAtPoint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135643 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, rolling out r135575.
ossy@webkit.org [Sat, 24 Nov 2012 06:13:20 +0000 (06:13 +0000)]
Unreviewed, rolling out r135575.
http://trac.webkit.org/changeset/135575
https://bugs.webkit.org/show_bug.cgi?id=103169

It made all tests assert (Requested by Ossy on #webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-11-23

Source/WebKit/qt:

* Api/qwebpage.cpp:
(QWebPagePrivate::QWebPagePrivate):
* WebCoreSupport/InitWebCoreQt.cpp:
(WebKit::initializeWebKitQt):

Source/WebKit2:

* qt/MainQt.cpp:
(WebKit):
(main):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135642 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agolistMarkerText() should create 8 bit strings when possible
msaboff@apple.com [Sat, 24 Nov 2012 05:18:31 +0000 (05:18 +0000)]
listMarkerText() should create 8 bit strings when possible
https://bugs.webkit.org/show_bug.cgi?id=103011

Reviewed by Filip Pizlo.

Made processing of marker text items use 8 bit strings where possible in order to reduce 16 bit strings.

Changes covered by existing tests.

* rendering/RenderListMarker.cpp:
(WebCore::toRoman): Changed character constants and buffer to build string to LChar's
(WebCore::toAlphabeticOrNumeric): Made a template based on character type.
(WebCore::toSymbolic): Made a template based on character type.
(WebCore::toAlphabetic): Made a template based on character type.
(WebCore::toNumeric): Added character type template parameter.
(WebCore::listMarkerText): Changed character constants to LChar where possible.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135641 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAny function that can log things should be able to easily log them to a memory buffer...
fpizlo@apple.com [Sat, 24 Nov 2012 03:16:47 +0000 (03:16 +0000)]
Any function that can log things should be able to easily log them to a memory buffer as well
https://bugs.webkit.org/show_bug.cgi?id=103000

Reviewed by Sam Weinig.

Source/JavaScriptCore:

Change all users of WTF::dataFile() to expect a PrintStream& rather than a FILE*.

* bytecode/Operands.h:
(JSC::OperandValueTraits::dump):
(JSC::dumpOperands):
(JSC):
* dfg/DFGAbstractState.cpp:
(JSC::DFG::AbstractState::dump):
* dfg/DFGAbstractState.h:
(AbstractState):
* dfg/DFGAbstractValue.h:
(JSC::DFG::AbstractValue::dump):
* dfg/DFGCommon.h:
(JSC::DFG::NodeIndexTraits::dump):
* dfg/DFGStructureAbstractValue.h:
(JSC::DFG::StructureAbstractValue::dump):
* dfg/DFGVariableEvent.cpp:
(JSC::DFG::VariableEvent::dump):
(JSC::DFG::VariableEvent::dumpFillInfo):
(JSC::DFG::VariableEvent::dumpSpillInfo):
* dfg/DFGVariableEvent.h:
(VariableEvent):
* disassembler/Disassembler.h:
(JSC):
(JSC::tryToDisassemble):
* disassembler/UDis86Disassembler.cpp:
(JSC::tryToDisassemble):

Source/WTF:

We have a number of places where we pass around a FILE* as a target to which to print
some logging information. But the purpose of passing FILE* instead of always assuming
that we should dump to stderr is that it may be sometimes useful to send the logging
information elsewhere. Unfortunately, FILE* isn't quite powerful enough: it's combersome
to use it to send logging to a string, for example.

We could get around this by using <iostream> and <sstream>, but so far this aspect of
C++ has not been part of the WebKit coding conventions. Personally I find <iostream>
awkward due to its abuse of operator overloading.

So this patch introduces the PrintStream abstract class, which offers printf-like
functionality while completely abstracting the destination and mechanism of the printing
output. It would be trivial to implement a StringPrintStream, for example. This will feed
into work on https://bugs.webkit.org/show_bug.cgi?id=102999.

This also sets us up for creating templatized print() and println() methods that will
allow us to say things like out.print("count = ", count, "\n"), but that is the topic
of https://bugs.webkit.org/show_bug.cgi?id=103009.

This patch also changes dataLog() to use FilePrintStream internally, and WTF::dataFile()
now returns a FilePrintStream&. Any previous users of WTF::dataFile() have been changed
to expect a PrintStream&.

* GNUmakefile.list.am:
* WTF.pro:
* WTF.vcproj/WTF.vcproj:
* WTF.xcodeproj/project.pbxproj:
* wtf/CMakeLists.txt:
* wtf/DataLog.cpp:
(WTF):
(WTF::initializeLogFileOnce):
(WTF::initializeLogFile):
(WTF::dataFile):
(WTF::dataLogV):
(WTF::dataLogString):
* wtf/DataLog.h:
(WTF):
* wtf/FilePrintStream.cpp: Added.
(WTF):
(WTF::FilePrintStream::FilePrintStream):
(WTF::FilePrintStream::~FilePrintStream):
(WTF::FilePrintStream::vprintf):
(WTF::FilePrintStream::flush):
* wtf/FilePrintStream.h: Added.
(WTF):
(FilePrintStream):
(WTF::FilePrintStream::file):
* wtf/PrintStream.cpp: Added.
(WTF):
(WTF::PrintStream::PrintStream):
(WTF::PrintStream::~PrintStream):
(WTF::PrintStream::printf):
(WTF::PrintStream::print):
(WTF::PrintStream::println):
(WTF::PrintStream::flush):
(WTF::print):
* wtf/PrintStream.h: Added.
(WTF):
(PrintStream):
(WTF::print):
(WTF::println):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135640 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] Define platform limitations in platform layer
commit-queue@webkit.org [Sat, 24 Nov 2012 01:30:21 +0000 (01:30 +0000)]
[BlackBerry] Define platform limitations in platform layer
https://bugs.webkit.org/show_bug.cgi?id=103160

Patch by Nima Ghanavatian <nghanavatian@rim.com> on 2012-11-23
Reviewed by Rob Buis.

Taking length as const unsigned in checkSpellingOfStringAsync.
This value is unchanging and should never be less than 0, even
in an error case.

Internally reviewed by Mike Fenton.

* Api/WebPageClient.h:
* WebKitSupport/InputHandler.cpp:
(BlackBerry::WebKit::InputHandler::requestCheckingOfString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135639 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemove unused ScrollByPixelVelocity
rjkroege@chromium.org [Sat, 24 Nov 2012 00:42:39 +0000 (00:42 +0000)]
Remove unused ScrollByPixelVelocity
https://bugs.webkit.org/show_bug.cgi?id=102840

Reviewed by Sam Weinig.

Source/WebCore:

https://bugs.webkit.org/show_bug.cgi?id=80311 wrongly introduced a new
kind of scroll: "ScrollByPixelVelocity". This code is not used on any platform
and will not be of use in the future so this patch removes it.

No new tests: code removal/cleanup.

* page/EventHandler.cpp:
(WebCore::EventHandler::defaultWheelEventHandler):
(WebCore::EventHandler::handleGestureEvent):
* platform/PlatformWheelEvent.h:
* platform/ScrollAnimator.cpp:
(WebCore::ScrollAnimator::handleWheelEvent):
* platform/ScrollAnimatorNone.cpp:
(WebCore::ScrollAnimatorNone::ScrollAnimatorNone):
(WebCore::ScrollAnimatorNone::scroll):
(WebCore::ScrollAnimatorNone::cancelAnimations):
(WebCore::ScrollAnimatorNone::animationTimerFired):
* platform/ScrollAnimatorNone.h:
(ScrollAnimatorNone):
* platform/ScrollTypes.h:
* platform/ScrollableArea.cpp:
(WebCore::ScrollableArea::scroll):

Source/WebKit/chromium:

Removed test for deleted code.

* tests/ScrollAnimatorNoneTest.cpp:

Source/WTF:

The GESTURE_ANIMATION feature turns on code in WebCore is unused.
Remove it.

No new tests: code removal/cleanup.

* wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135638 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed gardening: updated test expectations for failure in
rjkroege@chromium.org [Fri, 23 Nov 2012 23:20:20 +0000 (23:20 +0000)]
Unreviewed gardening: updated test expectations for failure in
* svg/batik/text/textPosition2.svg
https://bugs.webkit.org/show_bug.cgi?id=103148

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135637 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed gardening: text failures in:
rjkroege@chromium.org [Fri, 23 Nov 2012 23:02:26 +0000 (23:02 +0000)]
Unreviewed gardening: text failures in:
* inspector/console/command-line-api-inspect.html
* inspector/runtime/runtime-localStorage-getProperties.html
https://bugs.webkit.org/show_bug.cgi?id=103161

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135636 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed gardening: update test expectations for
rjkroege@chromium.org [Fri, 23 Nov 2012 22:39:58 +0000 (22:39 +0000)]
Unreviewed gardening: update test expectations for
svg/hixie/perf/007.xml

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135635 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS3 Backgrounds and Borders] Implement new CSS3 background-position parsing.
rjkroege@chromium.org [Fri, 23 Nov 2012 22:24:52 +0000 (22:24 +0000)]
[CSS3 Backgrounds and Borders] Implement new CSS3 background-position parsing.
https://bugs.webkit.org/show_bug.cgi?id=102104

Patch by Alexis Menard <alexis@webkit.org> on 2012-11-23
Reviewed by Julien Chaffraix.

Add new tests to cover the new parsing for the background position
property. For now the computed style checks are commented out because
the rendering part of the feature is not yet implemented (i.e.
FillLayer does not contain yet the feature and this is what
CSSComputedStyleDeclaration is using).

* fast/backgrounds/background-position-parsing-2-expected.txt:
* fast/backgrounds/background-position-parsing-2.html:
* platform/chromium/TestExpectations:
* platform/efl/TestExpectations:
* platform/gtk/TestExpectations:
* platform/mac/TestExpectations:
* platform/qt/TestExpectations:
* platform/win/TestExpectations:
* platform/wincairo/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135634 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemoving unnecessary RenderBox friend access in RenderObject
commit-queue@webkit.org [Fri, 23 Nov 2012 22:04:10 +0000 (22:04 +0000)]
Removing unnecessary RenderBox friend access in RenderObject
https://bugs.webkit.org/show_bug.cgi?id=103159

Patch by Adenilson Cavalcanti <cavalcantii@gmail.com> on 2012-11-23
Reviewed by Eric Seidel.

RenderObject has several classes marked as friend, between them,
there is RenderBox which is derived from it indirectly and is not accessing
its forefather private data making unnecessary the 'friend' access level.

No new tests, no changes in functionality.

* rendering/RenderObject.h:
(RenderObject):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135633 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CSS3 Backgrounds and Borders] Implement new CSS3 background-position parsing.
alexis@webkit.org [Fri, 23 Nov 2012 21:24:10 +0000 (21:24 +0000)]
[CSS3 Backgrounds and Borders] Implement new CSS3 background-position parsing.
https://bugs.webkit.org/show_bug.cgi?id=102104

Reviewed by Julien Chaffraix.

.:

Protect the new feature behind a feature flag.

* Source/cmake/WebKitFeatures.cmake:
* Source/cmakeconfig.h.cmake:
* configure.ac:

Source/JavaScriptCore:

Protect the new feature behind a feature flag.

* Configurations/FeatureDefines.xcconfig:

Source/WebCore:

Implement the parsing of the CSS3 background-position property which
allows to specify up to four values to position the background. It is
documented here :
http://www.w3.org/TR/css3-background/#the-background-position . I
decided to split up the patch is two pieces, the parsing and later I
will do the rendering. I kept the old code as much as possible (i.e.
the two values parsing) as it is used not only by background-position
but by few other properties. The new four values support is only valid
on the property itself but does not work on the shorthand.

Opera has already implemented this feature.

I extended backgrounds/background-position-parsing-2.html to cover the
new possible values as well as buggy values.

* Configurations/FeatureDefines.xcconfig:
* GNUmakefile.am:
* GNUmakefile.features.am:
* css/CSSParser.cpp:
(WebCore::CSSParser::parseFillPositionComponent): Modify this function
to allow creation of CSS primitive values with not only percentage or
length but with keywords also as the new <position> type allows to
specify keywords, for example top 30px right 20px.
(WebCore):
(WebCore::isValueConflictingWithCurrentEdge):
(WebCore::isBackgroundPositionKeyword):
(WebCore::CSSParser::parse3ValuesBackgroundPosition):
(WebCore::CSSParser::parse4ValuesBackgroundPosition):
(WebCore::CSSParser::parseFillBackgroundPosition): This function counts
the number of words for the property value and call dedicated parsing
functions for each case. Note that in case of two values we just call the old CSS 2.1
method.
(WebCore::CSSParser::parseFillPosition):
(WebCore::CSSParser::parseFillProperty): Only call the new parsing
function if the feature is activated otherwise fallback to the old code
path.
* css/CSSParser.h:

Source/WebKit/chromium:

Protect the new feature behind a feature flag.

* features.gypi:

Source/WebKit/mac:

Protect the new feature behind a feature flag.

* Configurations/FeatureDefines.xcconfig:

Source/WebKit2:

Protect the new feature behind a feature flag.

* Configurations/FeatureDefines.xcconfig:

Tools:

Protect the new feature behind a feature flag.

* Scripts/webkitperl/FeatureList.pm:
* qmake/mkspecs/features/features.pri:

WebKitLibraries:

Protect the new feature behind a feature flag.

* win/tools/vsprops/FeatureDefines.vsprops:
* win/tools/vsprops/FeatureDefinesCairo.vsprops:

LayoutTests:

Add new tests to cover the new parsing for the background position
property. For now the computed style checks are commented out because
the rendering part of the feature is not yet implemented (i.e.
FillLayer does not contain yet the feature and this is what
CSSComputedStyleDeclaration is using).

* fast/backgrounds/background-position-parsing-2-expected.txt:
* fast/backgrounds/background-position-parsing-2.html:
* platform/chromium/TestExpectations:
* platform/efl/TestExpectations:
* platform/gtk/TestExpectations:
* platform/mac/TestExpectations:
* platform/qt/TestExpectations:
* platform/win/TestExpectations:
* platform/wincairo/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135632 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed gardening.
zandobersek@gmail.com [Fri, 23 Nov 2012 19:32:52 +0000 (19:32 +0000)]
Unreviewed gardening.

Adding failure expectation for fast/dom/shadow/host-wrapper-reclaimed.html.

* platform/gtk/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135631 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed: updated test expecations -- additional failures in
rjkroege@chromium.org [Fri, 23 Nov 2012 19:17:50 +0000 (19:17 +0000)]
Unreviewed: updated test expecations -- additional failures in
* fast/text/large-text-composed-char.html
* fast/text/text-letter-spacing.html
* http/tests/inspector/indexeddb/database-data.html
and new failures:
* fast/frames/frame-unload-crash2.html

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135630 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoAdding occlusion detection to reduce overdraw in RenderBox background rendering
junov@google.com [Fri, 23 Nov 2012 18:43:51 +0000 (18:43 +0000)]
Adding occlusion detection to reduce overdraw in RenderBox background rendering
https://bugs.webkit.org/show_bug.cgi?id=102557

Reviewed by Stephen White.

Source/WebCore:

Adding hasAlpha method to all classes dervived from CSSImageValue and
StyleImage, in order to support occlusion testing.
This allows the FillLayer recursion to stop early when an opaque layer
is encountered while applying a CSS background style. This also allows
the elimination of background color application when the background is
completely covered by an image.  The optimization also skips the
clearing of the page's root element when the page background is
entirely covered by an image (e.g a tiled bitmap or a gradient)

Test: fast/backgrounds/background-opaque-images-over-color.html

* css/CSSCrossfadeValue.cpp:
(WebCore::subimageHasAlpha):
(WebCore):
(WebCore::CSSCrossfadeValue::isPending):
(WebCore::CSSCrossfadeValue::hasAlpha):
* css/CSSCrossfadeValue.h:
(CSSCrossfadeValue):
* css/CSSGradientValue.cpp:
(WebCore::CSSGradientValue::hasAlpha):
(WebCore):
* css/CSSGradientValue.h:
(CSSGradientValue):
* css/CSSImageGeneratorValue.cpp:
(WebCore::CSSImageGeneratorValue::hasAlpha):
(WebCore):
* css/CSSImageGeneratorValue.h:
(CSSImageGeneratorValue):
* css/CSSImageValue.cpp:
(WebCore::CSSImageValue::hasAlpha):
(WebCore):
* css/CSSImageValue.h:
(WebCore):
(CSSImageValue):
* loader/cache/CachedImage.cpp:
(WebCore::CachedImage::currentFrameHasAlpha):
Test whether a cached image has alpha, with side effect of ensuring
the image is in cache. (new method)
(WebCore):
* loader/cache/CachedImage.h:
(CachedImage):
* rendering/RenderBox.cpp:
(WebCore::RenderBox::paintFillLayers):
Added layer recursion early exit when an opaque layer is encountered.
* rendering/RenderBoxModelObject.cpp:
(WebCore::RenderBoxModelObject::paintFillLayerExtended):
Eliminated background color fill when the layer contains an opaque
image that covers the entire RenderBox.
* rendering/style/FillLayer.cpp:
(WebCore::FillLayer::hasOpaqueImage):
Returns true if the layer's image is known to be fully opaque.
(WebCore):
(WebCore::FillLayer::hasRepeatXY):
* rendering/style/FillLayer.h:
(FillLayer):
* rendering/style/StyleCachedImage.cpp:
(WebCore::StyleCachedImage::hasAlpha):
(WebCore):
* rendering/style/StyleCachedImage.h:
(StyleCachedImage):
* rendering/style/StyleCachedImageSet.cpp:
(WebCore::StyleCachedImageSet::hasAlpha):
(WebCore):
* rendering/style/StyleCachedImageSet.h:
(StyleCachedImageSet):
* rendering/style/StyleGeneratedImage.cpp:
(WebCore::StyleGeneratedImage::hasAlpha):
(WebCore):
* rendering/style/StyleGeneratedImage.h:
(StyleGeneratedImage):
* rendering/style/StyleImage.h:
(StyleImage):
* rendering/style/StylePendingImage.h:
(WebCore::StylePendingImage::hasAlpha):

LayoutTests:

New test exercises different CSS background style use cases that
are candidates for optimization by occlusion culling.

* fast/backgrounds/background-opaque-images-over-color-expected.txt: Added.
* fast/backgrounds/background-opaque-images-over-color.html: Added.
* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135629 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoGardening: updated expectations (pass/timeout) for
rjkroege@chromium.org [Fri, 23 Nov 2012 18:39:51 +0000 (18:39 +0000)]
Gardening: updated expectations (pass/timeout) for
week-multiple-fields-preserve-value-after-history-back.html
https://bugs.webkit.org/show_bug.cgi?id=103152

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135628 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Adapt accessibility feature to WebKitTestRunner.
commit-queue@webkit.org [Fri, 23 Nov 2012 18:17:00 +0000 (18:17 +0000)]
[EFL] Adapt accessibility feature to WebKitTestRunner.
https://bugs.webkit.org/show_bug.cgi?id=100757

Patch by Krzysztof Czech <k.czech@samsung.com> on 2012-11-23
Reviewed by Laszlo Gombos.

Skipping Accessibility tests for wk1/wk2.
Some of the implementation is still missing.

* platform/efl/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135627 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Update my e-mail address.
mario@webkit.org [Fri, 23 Nov 2012 17:15:31 +0000 (17:15 +0000)]
Unreviewed. Update my e-mail address.

* Scripts/webkitpy/common/config/committers.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135626 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoPage background color bleeds through inner edge of div border with rounded edges
junov@google.com [Fri, 23 Nov 2012 17:13:32 +0000 (17:13 +0000)]
Page background color bleeds through inner edge of div border with rounded edges
https://bugs.webkit.org/show_bug.cgi?id=103089

Reviewed by Stephen White.

Source/WebCore:

Test that verifies whether a RenderBoxEdge is more than 2 pixels wide
was applying the content scale factor backwards, so the
obscuresBackgroundEdge was falsely passing for downsized boxes.

Test: fast/backgrounds/gradient-background-leakage-2.html

* rendering/RenderBoxModelObject.cpp:
(WebCore::BorderEdge::obscuresBackgroundEdge):

LayoutTests:

New layout test to verify that there is no color bleeding between
a div element's border and background when there are rounded corners
an a scale that reduces a thick border down to a width of less than
two pixels.

* fast/backgrounds/gradient-background-leakage-2-expected.txt: Added.
* fast/backgrounds/gradient-background-leakage-2.html: Added.
* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135625 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoCoordinated Graphics: Match the behavior of selfOrAncestorHaveNonAffineTransforms...
commit-queue@webkit.org [Fri, 23 Nov 2012 17:10:03 +0000 (17:10 +0000)]
Coordinated Graphics: Match the behavior of selfOrAncestorHaveNonAffineTransforms() to its name.
https://bugs.webkit.org/show_bug.cgi?id=103097

Patch by Huang Dongsung <luxtella@company100.net> on 2012-11-23
Reviewed by Noam Rosenthal.

CoordinatedGraphicsLayer::selfOrAncestorHaveNonAffineTransforms() does
not check its ancestor. This patch makes this method check its ancestor
to match its name.

* WebProcess/WebPage/CoordinatedGraphics/CoordinatedGraphicsLayer.cpp:
(WebCore::CoordinatedGraphicsLayer::selfOrAncestorHaveNonAffineTransforms):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135624 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoGardening: update expectations for fast/events/stopPropagation-submit.html flaky
rjkroege@chromium.org [Fri, 23 Nov 2012 17:06:32 +0000 (17:06 +0000)]
Gardening: update expectations for fast/events/stopPropagation-submit.html flaky
https://bugs.webkit.org/show_bug.cgi?id=103149

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135623 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoCoordinated Graphics: LayerTreeRenderer::removeTile() does not remove tiles actually.
commit-queue@webkit.org [Fri, 23 Nov 2012 16:57:59 +0000 (16:57 +0000)]
Coordinated Graphics: LayerTreeRenderer::removeTile() does not remove tiles actually.
https://bugs.webkit.org/show_bug.cgi?id=103030

Patch by Huang Dongsung <luxtella@company100.net> on 2012-11-23
Reviewed by Noam Rosenthal.

LayerTreeRenderer::removeTile() must add a backing store into
m_backingStoresWithPendingBuffers. Otherwise, tiles are not removed.

* UIProcess/CoordinatedGraphics/LayerTreeRenderer.cpp:
(WebKit::LayerTreeRenderer::removeTile):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135622 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[CMake] Add support for winflexbison distribution
paroga@webkit.org [Fri, 23 Nov 2012 16:54:35 +0000 (16:54 +0000)]
[CMake] Add support for winflexbison distribution
https://bugs.webkit.org/show_bug.cgi?id=102551

Reviewed by Laszlo Gombos.

Since GnuWin32 does not provide recent versions of bision and flex supporting
the alternative winflexbison distribution is the prefered option.

* Source/cmake/WebKitMacros.cmake:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135621 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[TexMap] Remove dependency of TextureMapperLayer on GraphicsLayerTextureMapper.
commit-queue@webkit.org [Fri, 23 Nov 2012 16:51:15 +0000 (16:51 +0000)]
[TexMap] Remove dependency of TextureMapperLayer on GraphicsLayerTextureMapper.
https://bugs.webkit.org/show_bug.cgi?id=103046

Patch by Huang Dongsung <luxtella@company100.net> on 2012-11-23
Reviewed by Noam Rosenthal.

Source/WebCore:

It is a preparation patch for Threaded Coordinated Graphics on WK1.

We want to make TextureMapperLayer into an actor model. After making an actor
model, GraphicsLayerTextureMapper and LayerTreeRenderer will communicate with
TextureMapperLayer using message passing. In the first step, we want to hide
GraphicsLayerTextureMapper from TextureMapperLayer.

There are 2 kinds of dependencies of TextureMapperLayer in GraphicsLayerTextureMapper.
1. To query layer states
2. To paint layer contents

In this patch, we removed the second dependency. Currently, TextureMapperLayer
requests painting contents to GraphicsLayerTextureMapper in WK1 while
Coordinated Graphics prepares the contents in LayerTreeRenderer. This patch
makes WK1 prepare the contents in GraphicsLayerTextureMapper like
LayerTreeRenderer.

We will remove the first dependency in a follow-up patch.

Covered by existing layout tests.

* platform/graphics/texmap/GraphicsLayerTextureMapper.cpp:
(WebCore::GraphicsLayerTextureMapper::GraphicsLayerTextureMapper):
(WebCore::GraphicsLayerTextureMapper::flushCompositingStateForThisLayerOnly):
(WebCore::GraphicsLayerTextureMapper::flushCompositingState):
(WebCore):
(WebCore::GraphicsLayerTextureMapper::didFlushCompositingState):
(WebCore::GraphicsLayerTextureMapper::didFlushCompositingStateRecursive):
(WebCore::GraphicsLayerTextureMapper::updateBackingStore):
(WebCore::GraphicsLayerTextureMapper::prepareBackingStore):
(WebCore::GraphicsLayerTextureMapper::shouldHaveBackingStore):
(WebCore::GraphicsLayerTextureMapper::drawRepaintCounter):
(WebCore::GraphicsLayerTextureMapper::setDebugBorder):
* platform/graphics/texmap/GraphicsLayerTextureMapper.h:
(GraphicsLayerTextureMapper):
(WebCore::GraphicsLayerTextureMapper::setHasOwnBackingStore):
* platform/graphics/texmap/TextureMapperLayer.cpp:
(WebCore::TextureMapperLayer::rootLayer):
(WebCore::TextureMapperLayer::paintSelf):
(WebCore::TextureMapperLayer::textureMapper):
(WebCore):
(WebCore::TextureMapperLayer::flushCompositingState):
(WebCore::TextureMapperLayer::flushCompositingStateSelf):
* platform/graphics/texmap/TextureMapperLayer.h:
(WebCore):
(WebCore::TextureMapperLayer::TextureMapperLayer):
(TextureMapperLayer):
(WebCore::TextureMapperLayer::setBackingStore):
(State):
(WebCore::TextureMapperLayer::State::State):

Source/WebKit2:

This patch makes GraphicsLayerTextureMapper create its own backing
store. However, LayerTreeRenderer also creates its own backing store. So
when LayerTreeRenderer uses GraphicsLayerTextureMapper,
LayerTreeRenderer needs GraphicsLayerTextureMapper to not create its own
backing store.

* UIProcess/CoordinatedGraphics/LayerTreeRenderer.cpp:
(WebKit::LayerTreeRenderer::createLayer):
(WebKit::LayerTreeRenderer::getBackingStore):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135620 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL][GTK] Adapt accessibility feature to WebKitTestRunner.
commit-queue@webkit.org [Fri, 23 Nov 2012 16:47:57 +0000 (16:47 +0000)]
[EFL][GTK] Adapt accessibility feature to WebKitTestRunner.
https://bugs.webkit.org/show_bug.cgi?id=100757

Patch by Mateusz Leszko <m.leszko@samsung.com> on 2012-11-23
Reviewed by Laszlo Gombos.

Tools:

Allow to build accessibility feature in WebkitTestRunner for EFL Platform.
Updating CMake files due to changes from Bug 99011.

* WebKitTestRunner/InjectedBundle/AccessibilityController.cpp:
(WTR::AccessibilityController::AccessibilityController):
(WTR):
* WebKitTestRunner/InjectedBundle/AccessibilityController.h:
(AccessibilityController):
* WebKitTestRunner/InjectedBundle/AccessibilityUIElement.cpp:
(WTR):
* WebKitTestRunner/InjectedBundle/AccessibilityUIElement.h:
(AccessibilityUIElement):
* WebKitTestRunner/InjectedBundle/atk/AccessibilityUIElementAtk.cpp: - Removed gtk/gtk.h dependency.
* WebKitTestRunner/PlatformEfl.cmake: - Adding atk includes.

LayoutTests:

Unskipping EFL Accessiblity tests for wk2.

* platform/efl-wk1/TestExpectations:
* platform/efl/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135619 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[ENCHANT] Infinite loop when end word position is not found
carlosgc@webkit.org [Fri, 23 Nov 2012 16:40:33 +0000 (16:40 +0000)]
[ENCHANT] Infinite loop when end word position is not found
https://bugs.webkit.org/show_bug.cgi?id=103139

Reviewed by Martin Robinson.

Covered by existing tests in editing/spelling.

* editing/visible_units.cpp:
(WebCore::isLogicalStartOfWord): Make it private again.
(WebCore::islogicalEndOfWord): Ditto.
* editing/visible_units.h:
* platform/text/enchant/TextCheckerEnchant.cpp:
(WebCore::TextCheckerEnchant::checkSpellingOfWord): Helper funtion
to check a word inside a string using the loaded dictionaries.
(WebCore::TextCheckerEnchant::checkSpellingOfString): Optimize and
simply the code by using the word iterator to iterate the
words.
* platform/text/enchant/TextCheckerEnchant.h:
(TextCheckerEnchant):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135618 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed gardening: test expectations failures.
rjkroege@chromium.org [Fri, 23 Nov 2012 16:38:13 +0000 (16:38 +0000)]
Unreviewed gardening: test expectations failures.
https://bugs.webkit.org/show_bug.cgi?id=103148

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135617 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoREGRESSION(r135486): ASSERT(isMainThread()) hit in LayerTreeRenderer::setActive(bool)
commit-queue@webkit.org [Fri, 23 Nov 2012 16:37:40 +0000 (16:37 +0000)]
REGRESSION(r135486): ASSERT(isMainThread()) hit in LayerTreeRenderer::setActive(bool)
https://bugs.webkit.org/show_bug.cgi?id=103077

Patch by Huang Dongsung <luxtella@company100.net> on 2012-11-23
Reviewed by Noam Rosenthal.

Currently, the painting thread uses purgeGLResources() and setActive().
QQuickWebPage::updatePaintNode() indirectly calls both methods after
locking the main thread.

This patch changes two points.
1. Remove ASSERT(isMainThread()) in methods that the painting thread can
call.
2. Make purgeGLResources() and setActive() call methods of
LayerTreeCoordinatorProxy via dispatchOnMainThread().

* UIProcess/CoordinatedGraphics/LayerTreeRenderer.cpp:
(WebKit::LayerTreeRenderer::LayerTreeRenderer):
(WebKit::LayerTreeRenderer::renderNextFrame):
(WebKit::LayerTreeRenderer::purgeGLResources):
(WebKit):
(WebKit::LayerTreeRenderer::purgeBackingStores):
(WebKit::LayerTreeRenderer::setActive):
* UIProcess/CoordinatedGraphics/LayerTreeRenderer.h:
(LayerTreeRenderer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135616 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] Meta viewport can't be changed dynamically
zhajiang@rim.com [Fri, 23 Nov 2012 16:27:30 +0000 (16:27 +0000)]
[BlackBerry] Meta viewport can't be changed dynamically
https://bugs.webkit.org/show_bug.cgi?id=103098

Patch by Jacky Jiang  <zhajiang@rim.com>.
Reviewed by Rob Buis.
Internally reviewed by Arvid Nilsson.

PR: 237072
Reset m_userPerformedManualZoom and enable
m_shouldZoomToInitialScaleAfterLoadFinished so that we can relayout the
page and zoom it to fit the screen when we dynamically change the meta
viewport after the load is finished.

* Api/WebPage.cpp:
(BlackBerry::WebKit::WebPagePrivate::dispatchViewportPropertiesDidChange):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135615 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed gardening: update expectations for now crashing test
rjkroege@chromium.org [Fri, 23 Nov 2012 16:15:51 +0000 (16:15 +0000)]
Unreviewed gardening: update expectations for now crashing test
visibility-simple-canvas2d-layer.html
https://bugs.webkit.org/show_bug.cgi?id=98086

* platform/chromium/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135614 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] Drop Synchronous Mouse Events
commit-queue@webkit.org [Fri, 23 Nov 2012 16:11:39 +0000 (16:11 +0000)]
[BlackBerry] Drop Synchronous Mouse Events
https://bugs.webkit.org/show_bug.cgi?id=102980

Patch by Genevieve Mak <gmak@rim.com> on 2012-11-23
Reviewed by Rob Buis.

Reviewed Internally by George Staikos and Mike Lattanzio.
Don't send actual Touch Events to fullscreen plugins unless
its a Touch Cancel, DoubleTap or TouchHold.
Instead send mouse events as long as its a single finger.
PR #242781

* Api/WebPage.cpp:
(BlackBerry::WebKit::WebPage::touchPointAsMouseEvent):
(BlackBerry::WebKit::WebPagePrivate::dispatchTouchEventToFullScreenPlugin):
(WebKit):
(BlackBerry::WebKit::WebPagePrivate::dispatchTouchPointAsMouseEventToFullScreenPlugin):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135613 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Define WTF_PLATFORM_EFL in Platform.h
commit-queue@webkit.org [Fri, 23 Nov 2012 15:36:37 +0000 (15:36 +0000)]
[EFL] Define WTF_PLATFORM_EFL in Platform.h
https://bugs.webkit.org/show_bug.cgi?id=101482

Patch by Laszlo Gombos <l.gombos@samsung.com> on 2012-11-23
Reviewed by Kenneth Rohde Christiansen.

.:

Remove the definition of WTF_PLATFORM_EFL from the build system to
make the EFL port consistent with other ports.

* Source/cmake/OptionsEfl.cmake:

Source/WTF:

Define WTF_PLATFORM_EFL in Platform.h to be consistent with
other ports.

* wtf/Platform.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135612 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Qt] GraphicsContext3DQt makes wrong surface current.
zeno.albisser@digia.com [Fri, 23 Nov 2012 15:34:56 +0000 (15:34 +0000)]
[Qt] GraphicsContext3DQt makes wrong surface current.
https://bugs.webkit.org/show_bug.cgi?id=103136

blitMultisampleFramebufferAndRestoreContext restores
the wrong surface in case the currentContext is the same
as m_platformContext.
In this case we would reset the bound surface to NULL.
Instead we should just leave the context unchanged.

Reviewed by Kenneth Rohde Christiansen.

* platform/graphics/qt/GraphicsContext3DQt.cpp:
(WebCore::GraphicsContext3DPrivate::blitMultisampleFramebufferAndRestoreContext):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135611 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoFix the ARM traditional build after r135330
ossy@webkit.org [Fri, 23 Nov 2012 15:32:46 +0000 (15:32 +0000)]
Fix the ARM traditional build after r135330
https://bugs.webkit.org/show_bug.cgi?id=102871

Patch by Gabor Ballabas <gaborb@inf.u-szeged.hu> on 2012-11-23
Reviewed by Zoltan Herczeg.

Added missing functionality to traditional ARM architecture.

* assembler/ARMAssembler.h:
(JSC::ARMAssembler::revertJump):
(ARMAssembler):
* assembler/MacroAssemblerARM.h:
(JSC::MacroAssemblerARM::startOfPatchableBranchPtrWithPatchOnAddress):
(JSC::MacroAssemblerARM::startOfBranchPtrWithPatchOnRegister):
(MacroAssemblerARM):
(JSC::MacroAssemblerARM::revertJumpReplacementToBranchPtrWithPatch):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135610 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL][WK2][AC] EwkViewImpl::transformToScreen() does not retrieve the Ecore_X_Window...
commit-queue@webkit.org [Fri, 23 Nov 2012 15:15:58 +0000 (15:15 +0000)]
[EFL][WK2][AC] EwkViewImpl::transformToScreen() does not retrieve the Ecore_X_Window properly
https://bugs.webkit.org/show_bug.cgi?id=103133

Patch by Christophe Dumez <christophe.dumez@intel.com> on 2012-11-23
Reviewed by Kenneth Rohde Christiansen.

If accelerated compositing is enabled, EwkViewImpl::transformToScreen() should
attempt to retrieve the Ecore_X_Window using ecore_evas_gl_x11_window_get() first
and only fall back to ecore_evas_software_x11_window_get().

* UIProcess/API/efl/EwkViewImpl.cpp:
(EwkViewImpl::transformToScreen):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135609 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWeb Inspector: remove total item from NMI snapshot grid
commit-queue@webkit.org [Fri, 23 Nov 2012 15:13:38 +0000 (15:13 +0000)]
Web Inspector: remove total item from NMI snapshot grid
https://bugs.webkit.org/show_bug.cgi?id=101926

Patch by Alexei Filippov <alph@chromium.org> on 2012-11-23
Reviewed by Yury Semikhatsky.

* inspector/front-end/NativeMemorySnapshotView.js:
(WebInspector.NativeSnapshotDataGrid):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135608 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[Qt] GraphicsContext3DQt should not overwrite glClearColor of sharing context.
zeno.albisser@digia.com [Fri, 23 Nov 2012 15:11:56 +0000 (15:11 +0000)]
[Qt] GraphicsContext3DQt should not overwrite glClearColor of sharing context.
https://bugs.webkit.org/show_bug.cgi?id=103135

In case we use GraphicsContext3DQt to render into or
read from an existing context, we should not reset
the glClearColor on construction.
Otherwise we might overwrite an existing definition from
the sharing context.

Reviewed by Kenneth Rohde Christiansen.

* platform/graphics/qt/GraphicsContext3DQt.cpp:
(WebCore::GraphicsContext3D::GraphicsContext3D):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135607 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWeb Inspector: damaging last chunk does not remove its garbage from DOM
pfeldman@chromium.org [Fri, 23 Nov 2012 15:01:35 +0000 (15:01 +0000)]
Web Inspector: damaging last chunk does not remove its garbage from DOM
https://bugs.webkit.org/show_bug.cgi?id=103132

Reviewed by Vsevolod Vlasov.

Source/WebCore:

Removing everything between undamaged chunks, not registered chunk rows.

* inspector/front-end/DefaultTextEditor.js:
(WebInspector.TextEditorMainPanel.prototype._insertSpanBefore):
(WebInspector.TextEditorMainPanel.prototype._updateChunksForRanges):
(WebInspector.TextEditorMainChunk.prototype.updateCollapsedLineRow):
(WebInspector.TextEditorMainChunk.prototype.firstElement):
(WebInspector.TextEditorMainChunk.prototype.lastElement):

LayoutTests:

* inspector/editor/editor-test.js:
(initialize_EditorTests.InspectorTest.createTestEditor):
(initialize_EditorTests.InspectorTest.dumpEditorChunks):
(initialize_EditorTests.InspectorTest.dumpEditorModel):
(initialize_EditorTests.InspectorTest.dumpEditorDOM):
(initialize_EditorTests):
* inspector/editor/text-editor-remove-chunks-from-dom-expected.txt: Added.
* inspector/editor/text-editor-remove-chunks-from-dom.html: Added.
* inspector/editor/text-editor-selection-expected.txt:
* inspector/editor/text-editor-selection.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135606 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRenderLayer minor clean-up: replace raw pointers with OwnPtrs.
loislo@chromium.org [Fri, 23 Nov 2012 14:55:46 +0000 (14:55 +0000)]
RenderLayer minor clean-up: replace raw pointers with OwnPtrs.
https://bugs.webkit.org/show_bug.cgi?id=103134

Reviewed by Eric Seidel.

I found that RenderLayer still uses raw pointers to the objects which it owns.
Looks like these pointers could be replaced with OwnPtrs.

* rendering/RenderLayer.cpp:
(WebCore::RenderLayer::RenderLayer):
(WebCore::RenderLayer::~RenderLayer):
(WebCore::RenderLayer::paintLayerContents):
(WebCore::RenderLayer::hitTestLayer):
(WebCore::RenderLayer::updateNormalFlowList):
(WebCore::RenderLayer::collectLayers):
(WebCore::RenderLayer::styleChanged):
* rendering/RenderLayer.h:
(WebCore::RenderLayer::marquee):
(WebCore::RenderLayer::posZOrderList):
(WebCore::RenderLayer::negZOrderList):
(WebCore::RenderLayer::normalFlowList):
(RenderLayer):
(WebCore::RenderLayer::clearZOrderLists):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135605 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL][AC] Mouse cursor is not updated when accelerated compositing is enabled
commit-queue@webkit.org [Fri, 23 Nov 2012 14:22:55 +0000 (14:22 +0000)]
[EFL][AC] Mouse cursor is not updated when accelerated compositing is enabled
https://bugs.webkit.org/show_bug.cgi?id=103131

Patch by Christophe Dumez <christophe.dumez@intel.com> on 2012-11-23
Reviewed by Kenneth Rohde Christiansen.

Make sure we try to get the Ecore_X_Window using ecore_evas_gl_x11_window_get()
in applyFallbackCursor() if accelerated compositing is enabled instead of
using ecore_evas_software_x11_window_get(). This fixes mouse cursor update
when accelerated compositing is enabled.

No new tests, no behavior change for layout tests.

* platform/efl/EflScreenUtilities.cpp:
(WebCore::applyFallbackCursor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135604 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoMemory instrumentation: extract MemoryObjectInfo declaration into a separate file
yurys@chromium.org [Fri, 23 Nov 2012 14:08:00 +0000 (14:08 +0000)]
Memory instrumentation: extract MemoryObjectInfo declaration into a separate file
https://bugs.webkit.org/show_bug.cgi?id=102510

Reviewed by Pavel Feldman.

Source/JavaScriptCore:

Added new symbols for the methods that have moved into .../wtf/MemoryInstrumentation.cpp

* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:

Source/WebCore:

Added necessary includes where MemoryObjectInfo declaration is required.

* platform/audio/FFTFrame.cpp:
* rendering/style/RenderStyle.cpp:
* rendering/style/StyleRareInheritedData.cpp:
* rendering/style/StyleRareNonInheritedData.cpp:

Source/WTF:

Moved MemoryObjectInfo into separate header. Moved definition of MemoryInstrumentation
methods that depend on MemoryObjectInfo declaration into MemoryInstrumentation.cpp to
make MemoryInstrumentation require only forward declaration of MemoryObjectInfo.

* GNUmakefile.list.am:
* WTF.gypi:
* WTF.pro:
* WTF.vcproj/WTF.vcproj:
* wtf/CMakeLists.txt:
* wtf/MemoryInstrumentation.cpp: Added.
(WTF):
(WTF::MemoryInstrumentation::MemoryInstrumentation):
(WTF::MemoryInstrumentation::~MemoryInstrumentation):
(WTF::MemoryInstrumentation::getObjectType): this method allows to get object type without
pulling in MemoryObjectInfo.h and all its dependencies.
(WTF::MemoryInstrumentation::callReportObjectInfo):
(WTF::MemoryInstrumentation::InstrumentedPointerBase::InstrumentedPointerBase):
(WTF::MemoryInstrumentation::InstrumentedPointerBase::process):
(WTF::MemoryClassInfo::init):
(WTF::MemoryClassInfo::addRawBuffer):
(WTF::MemoryClassInfo::addPrivateBuffer):
* wtf/MemoryInstrumentation.h:
(MemoryInstrumentation):
(WTF::MemoryInstrumentation::addRootObject):
(InstrumentedPointerBase):
(WTF::MemoryInstrumentation::reportObjectMemoryUsage):
(InstrumentedPointer):
(WTF::MemoryInstrumentation::addObject): we now pass owner's MemoryObjectInfo in all places
where we report objects pointed by the owner.
(WTF::MemoryInstrumentation::OwningTraits::addObject):
(WTF::MemoryClassInfo::MemoryClassInfo):
(WTF::MemoryClassInfo::addMember):
(MemoryClassInfo):
(WTF::MemoryInstrumentation::addObjectImpl):
(WTF::::InstrumentedPointer):
(WTF::::callReportMemoryUsage):
* wtf/MemoryObjectInfo.h: Added.
(WTF):
(MemoryObjectInfo):
(WTF::MemoryObjectInfo::MemoryObjectInfo):
(WTF::MemoryObjectInfo::objectType):
(WTF::MemoryObjectInfo::objectSize):
(WTF::MemoryObjectInfo::reportedPointer):
(WTF::MemoryObjectInfo::memoryInstrumentation):
(WTF::MemoryObjectInfo::reportObjectInfo):

Tools:

Added missing includes.

* TestWebKitAPI/Tests/WTF/MemoryInstrumentationTest.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135603 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agotst_qwebpage fails after QMimeDatabase patch
allan.jensen@digia.com [Fri, 23 Nov 2012 14:05:11 +0000 (14:05 +0000)]
tst_qwebpage fails after QMimeDatabase patch
https://bugs.webkit.org/show_bug.cgi?id=103125

Reviewed by Simon Hausmann.

Replace the static extension->mimetype table with a QMimeDatabase lookup

* tests/qwebpage/tst_qwebpage.cpp:
(tst_QWebPage::findText):
(tst_QWebPage::supportedContentType):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135602 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemove V8DOMWindowShell::getEntered
commit-queue@webkit.org [Fri, 23 Nov 2012 13:44:29 +0000 (13:44 +0000)]
Remove V8DOMWindowShell::getEntered
https://bugs.webkit.org/show_bug.cgi?id=96637

Patch by Dan Carney <dcarney@google.com> on 2012-11-23
Reviewed by Adam Barth.

V8DOMWindowShell::getEntered was refactored so that the window shell
no longer has to be kept alive by a v8 context. Instead, only
the DOMWrapperWorld will be kept alive.

No new tests. No change in functionality.

* bindings/v8/DOMDataStore.cpp:
(WebCore::DOMDataStore::current):
* bindings/v8/DOMWrapperWorld.cpp:
(WebCore::isolatedWorldWeakCallback):
(WebCore):
(WebCore::DOMWrapperWorld::makeContextWeak):
(WebCore::DOMWrapperWorld::setIsolatedWorldField):
* bindings/v8/DOMWrapperWorld.h:
(DOMWrapperWorld):
(WebCore::DOMWrapperWorld::isolated):
* bindings/v8/ScriptController.cpp:
(WebCore::ScriptController::existingWindowShell):
(WebCore::ScriptController::windowShell):
(WebCore::ScriptController::evaluateInIsolatedWorld):
(WebCore::ScriptController::currentWorldContext):
(WebCore::ScriptController::collectIsolatedContexts):
* bindings/v8/ScriptController.h:
(ScriptController):
* bindings/v8/V8Binding.h:
(WebCore::worldForEnteredContextIfIsolated):
* bindings/v8/V8DOMWindowShell.cpp:
(WebCore::V8DOMWindowShell::destroyIsolatedShell):
(WebCore::V8DOMWindowShell::disposeContext):
(WebCore::V8DOMWindowShell::initializeIfNeeded):
* bindings/v8/V8DOMWindowShell.h:
(V8DOMWindowShell):
* bindings/v8/V8PerContextData.h:
* bindings/v8/WorldContextHandle.cpp:
(WebCore::WorldContextHandle::WorldContextHandle):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135601 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL][WK2] Add support for PlatformWebView::simulateSpacebarKeyPress() in TestWebKitAPI
commit-queue@webkit.org [Fri, 23 Nov 2012 13:38:04 +0000 (13:38 +0000)]
[EFL][WK2] Add support for PlatformWebView::simulateSpacebarKeyPress() in TestWebKitAPI
https://bugs.webkit.org/show_bug.cgi?id=103123

Patch by Sudarsana Nagineni <sudarsana.nagineni@intel.com> on 2012-11-23
Reviewed by Kenneth Rohde Christiansen.

Add support for PlatformWebView::simulateSpacebarKeyPress() and enable
WebKit2 API test ResponsivenessTimerDoesntFireEarly.

* TestWebKitAPI/PlatformEfl.cmake: Enable ResponsivenessTimerDoesntFireEarly
test.
* TestWebKitAPI/efl/PlatformUtilities.cpp:
(TestWebKitAPI::Util::sleep): Use usleep() instead of sleep since the
test case is passing 0.5 seconds to sleep.
* TestWebKitAPI/efl/PlatformWebView.cpp:
(TestWebKitAPI::PlatformWebView::simulateSpacebarKeyPress): Implemented.
(TestWebKitAPI):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135600 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWeb Inspector: Refine JsDoc in ProfilesPanel.js
commit-queue@webkit.org [Fri, 23 Nov 2012 13:25:18 +0000 (13:25 +0000)]
Web Inspector: Refine JsDoc in ProfilesPanel.js
https://bugs.webkit.org/show_bug.cgi?id=103042

Patch by Eugene Klyuchnikov <eustas.bug@gmail.com> on 2012-11-23
Reviewed by Yury Semikhatsky.

I'm going to make profiler output typed.

This will require some amount of refactoring.

Before refactoring I would like to make things
unambiguous by adding JsDoc where they are missing.

* inspector/front-end/ProfilesPanel.js:
(WebInspector.ProfilesPanel.prototype._onToggleProfiling):
Added to eliminate parameter type ambiguity.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135599 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoCorrect input[type=number] value sanitization for user-input
tkent@chromium.org [Fri, 23 Nov 2012 13:11:46 +0000 (13:11 +0000)]
Correct input[type=number] value sanitization for user-input
https://bugs.webkit.org/show_bug.cgi?id=103018

Reviewed by Kentaro Hara.

Source/WebCore:

If a number field has non-number string, HTMLInputElement::value is not
updated and returns the past valid value. It doesn't match to the value
sanitization algorithm defined by the HTML standard [1], and Opera's
behavior. We should sanitize non-number strings to "".

[1] http://www.whatwg.org/specs/web-apps/current-work/multipage/states-of-the-type-attribute.html#number-state-(type=number)
> The value sanitization algorithm is as follows: If the value of the
> element is not a valid floating-point number, then set it to the empty
> string instead.

No new tests. Update existing tests;
fast/forms/number/number-commit-valid-only.html and
fast/forms/number/number-unacceptable-style.html

* html/HTMLInputElement.cpp:
(WebCore::HTMLInputElement::subtreeHasChanged):
Remove isAcceptableValue check.

* html/HTMLInputElement.h:
(HTMLInputElement): isAcceptableValue is no longer needed.
* html/InputType.cpp: Remove isAcceptableValue.
* html/InputType.h: Ditto.
* html/NumberInputType.cpp: Ditto.
(WebCore::NumberInputType::hasUnacceptableValue):
Fold the isAcceptableValue content into this.
* html/NumberInputType.h:
(NumberInputType): Remove isAcceptableValue.

LayoutTests:

* fast/forms/number/number-commit-valid-only-expected.txt:
* fast/forms/number/number-commit-valid-only.html:
Update the expected HTMLInputElement::value results.
* fast/forms/number/number-unacceptable-style-expected.txt:
* fast/forms/number/number-unacceptable-style.html:
Add some checks for HTMLInputElement::value.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135598 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[sh4] JavaScriptCore JIT build is broken since r130839
commit-queue@webkit.org [Fri, 23 Nov 2012 13:11:28 +0000 (13:11 +0000)]
[sh4] JavaScriptCore JIT build is broken since r130839
Add missing implementation for sh4 arch.
https://bugs.webkit.org/show_bug.cgi?id=101479

Patch by Julien BRIANCEAU <jbrianceau@nds.com> on 2012-11-23
Reviewed by Filip Pizlo.

* assembler/MacroAssemblerSH4.h:
(JSC::MacroAssemblerSH4::load8Signed):
(MacroAssemblerSH4):
(JSC::MacroAssemblerSH4::load16Signed):
(JSC::MacroAssemblerSH4::store8):
(JSC::MacroAssemblerSH4::store16):
(JSC::MacroAssemblerSH4::moveDoubleToInts):
(JSC::MacroAssemblerSH4::moveIntsToDouble):
(JSC::MacroAssemblerSH4::loadFloat):
(JSC::MacroAssemblerSH4::loadDouble):
(JSC::MacroAssemblerSH4::storeFloat):
(JSC::MacroAssemblerSH4::storeDouble):
(JSC::MacroAssemblerSH4::addDouble):
(JSC::MacroAssemblerSH4::convertFloatToDouble):
(JSC::MacroAssemblerSH4::convertDoubleToFloat):
(JSC::MacroAssemblerSH4::urshift32):
* assembler/SH4Assembler.h:
(JSC::SH4Assembler::sublRegReg):
(JSC::SH4Assembler::subvlRegReg):
(JSC::SH4Assembler::floatfpulfrn):
(JSC::SH4Assembler::fldsfpul):
(JSC::SH4Assembler::fstsfpul):
(JSC::SH4Assembler::dcnvsd):
(SH4Assembler):
(JSC::SH4Assembler::movbRegMem):
(JSC::SH4Assembler::sizeOfConstantPool):
(JSC::SH4Assembler::linkJump):
(JSC::SH4Assembler::printInstr):
(JSC::SH4Assembler::printBlockInstr):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135597 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed, rolling out r135592.
yurys@chromium.org [Fri, 23 Nov 2012 12:57:46 +0000 (12:57 +0000)]
Unreviewed, rolling out r135592.
http://trac.webkit.org/changeset/135592
https://bugs.webkit.org/show_bug.cgi?id=103126

broke Chromium Mac compilation (Requested by yurys on
#webkit).

Patch by Sheriff Bot <webkit.review.bot@gmail.com> on 2012-11-23

* WebCore.gypi:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135596 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[WK2][WKTR] WebKitTestRunner needs eventSender.contextClick()
commit-queue@webkit.org [Fri, 23 Nov 2012 12:55:45 +0000 (12:55 +0000)]
[WK2][WKTR] WebKitTestRunner needs eventSender.contextClick()
https://bugs.webkit.org/show_bug.cgi?id=86881

Patch by Wojciech Bielawski <w.bielawski@samsung.com> on 2012-11-23
Reviewed by Kenneth Rohde Christiansen.

Source/WebKit2:

InjectedBundle API extended to retrieve contextMenu entries names.
WebContextMenu exteded with function to retrieve menu items.
This API is required by layout tests.
Common code in WebContextMenu.cpp moved to separate method menuItemsWithUserData().
Behaviour of dependend methods is unchanged.

* WebProcess/InjectedBundle/API/c/WKBundlePage.cpp:
(WKBundlePageGetContextMenuEntriesNames):
* WebProcess/InjectedBundle/API/c/WKBundlePage.h:
* WebProcess/WebPage/WebContextMenu.cpp:
(WebKit::WebContextMenu::items):
(WebKit):
* WebProcess/WebPage/WebContextMenu.h:
(WebContextMenu):

Tools:

eventSender object extended with contextClick() method in WebKitTestRunner.
contextClick() method returns an array of menu items objects with field 'title' containing title of single
context menu entry. According to some tests menu items objects shall also support 'click()' method, but it's not
delivered in this patch.

* WebKitTestRunner/InjectedBundle/Bindings/EventSendingController.idl:
* WebKitTestRunner/InjectedBundle/EventSendingController.cpp:
(WTR::EventSendingController::contextClick):
(WTR):
* WebKitTestRunner/InjectedBundle/EventSendingController.h:
(EventSendingController):

LayoutTests:

Added expected results for two tests for EFL platform. Test expectations changed
for EFL WK1/WK2 and GTK platforms

* platform/efl/TestExpectations:
* platform/efl-wk1/TestExpectations:
* platform/efl-wk2/TestExpectations:
* platform/gkt-wk2/TestExpectations:
* platform/gtk/editing/selection/5354455-2-expected.txt:
* platform/efl/editing/selection/5354455-2-expected.txt: Added.
* platform/efl/fast/events/context-no-deselect-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135595 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[BlackBerry] InputHandler::ensureFocusTextElementVisible incorrectly used y coordinat...
commit-queue@webkit.org [Fri, 23 Nov 2012 12:54:53 +0000 (12:54 +0000)]
[BlackBerry] InputHandler::ensureFocusTextElementVisible incorrectly used y coordinate as x coordinate
https://bugs.webkit.org/show_bug.cgi?id=103091

Patch by Andrew Lo <anlo@rim.com> on 2012-11-23
Reviewed by Yong Li.

Internally reviewed by Mike Fenton.
Internal PR249523
Incorrectly used y coordinate as x coordinate in
InputHandler::ensureFocusTextElementVisible.

* WebKitSupport/InputHandler.cpp:
(BlackBerry::WebKit::InputHandler::ensureFocusTextElementVisible):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135594 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed. Fix Chromium Win compilation after r135580.
yurys@chromium.org [Fri, 23 Nov 2012 12:48:34 +0000 (12:48 +0000)]
Unreviewed. Fix Chromium Win compilation after r135580.
https://bugs.webkit.org/show_bug.cgi?id=101772

* tests/WebFrameTest.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135593 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoRemove references to non-chromium entries from WebCore.gypi (part 2)
yurys@chromium.org [Fri, 23 Nov 2012 12:38:07 +0000 (12:38 +0000)]
Remove references to non-chromium entries from WebCore.gypi (part 2)
https://bugs.webkit.org/show_bug.cgi?id=103124

Reviewed by Pavel Feldman.

* WebCore.gypi: removed references to gtk, cocoa, posix, ios, cf, win and mac

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135592 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWeb Inspector: [Canvas] add startCapturing and stopCapturing methods to protocol
commit-queue@webkit.org [Fri, 23 Nov 2012 12:20:24 +0000 (12:20 +0000)]
Web Inspector: [Canvas] add startCapturing and stopCapturing methods to protocol
https://bugs.webkit.org/show_bug.cgi?id=103053

Patch by Andrey Adaikin <aandrey@chromium.org> on 2012-11-23
Reviewed by Pavel Feldman.

Add startCapturing and stopCapturing methods to the protocol to instrument several canvas frames in a row.
Drive-by: support start offset index in retrieving trace logs in the backend as well as "alive" flag.
* inspector/InjectedScriptCanvasModule.cpp:
(WebCore::InjectedScriptCanvasModule::captureFrame):
(WebCore):
(WebCore::InjectedScriptCanvasModule::startCapturing):
(WebCore::InjectedScriptCanvasModule::callStartCapturingFunction):
(WebCore::InjectedScriptCanvasModule::stopCapturing):
(WebCore::InjectedScriptCanvasModule::dropTraceLog):
(WebCore::InjectedScriptCanvasModule::callVoidFunctionWithTraceLogIdArgument):
(WebCore::InjectedScriptCanvasModule::traceLog):
* inspector/InjectedScriptCanvasModule.h:
(InjectedScriptCanvasModule):
* inspector/InjectedScriptCanvasModuleSource.js:
(.):
* inspector/Inspector.json:
* inspector/InspectorCanvasAgent.cpp:
(WebCore::InspectorCanvasAgent::startCapturing):
(WebCore):
(WebCore::InspectorCanvasAgent::stopCapturing):
(WebCore::InspectorCanvasAgent::getTraceLog):
* inspector/InspectorCanvasAgent.h:
(InspectorCanvasAgent):
* inspector/front-end/CanvasProfileView.js:
(WebInspector.CanvasProfileView):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135591 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years ago[EFL] Newly added tests in r135436 and r135456 are failng
commit-queue@webkit.org [Fri, 23 Nov 2012 12:15:59 +0000 (12:15 +0000)]
[EFL] Newly added tests in r135436 and r135456 are failng
https://bugs.webkit.org/show_bug.cgi?id=103112

Unreviewed EFL gardening.

Skip new failing tests introduced in r135436 and r135456.

Patch by Sudarsana Nagineni <sudarsana.nagineni@intel.com> on 2012-11-23

* platform/efl-wk2/TestExpectations:
* platform/efl/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135590 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoUnreviewed gardening, skip failing tests.
ossy@webkit.org [Fri, 23 Nov 2012 12:09:23 +0000 (12:09 +0000)]
Unreviewed gardening, skip failing tests.

* platform/qt-5.0-wk2/TestExpectations:
* platform/wk2/TestExpectations:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135589 268f45cc-cd09-0410-ab3c-d52691b4dbfc

6 years agoWeb Inspector: dispatch events upon certain user actions on inspector's notification...
caseq@chromium.org [Fri, 23 Nov 2012 12:08:42 +0000 (12:08 +0000)]
Web Inspector: dispatch events upon certain user actions on inspector's notification services
https://bugs.webkit.org/show_bug.cgi?id=103114

Reviewed by Pavel Feldman.

Add a bunch of WebInspector.UserMetrics.UserAction, so platform-specific code can wire to user actions
without breaking too often.

* inspector/front-end/ElementsPanel.js:
(WebInspector.ElementsPanel.prototype._setPseudoClassForNodeId):
* inspector/front-end/JavaScriptSourceFrame.js:
(WebInspector.JavaScriptSourceFrame.prototype._setBreakpoint):
* inspector/front-end/NetworkItemView.js:
(WebInspector.NetworkItemView.prototype._tabSelected.set WebInspector):
(WebInspector.NetworkItemView.prototype._tabSelected):
* inspector/front-end/NetworkPanel.js:
(WebInspector.NetworkLogView.prototype._sortItems):
(WebInspector.NetworkDataGridNode.prototype.select):
* inspector/front-end/ProfilesPanel.js:
(WebInspector.ProfilesPanel.prototype.addProfileHeader):
* inspector/front-end/ScriptsPanel.js:
(WebInspector.ScriptsPanel.prototype._showSourceLine):
(WebInspector.ScriptsPanel.prototype._toggleFormatSource):
* inspector/front-end/UISourceCode.js:
(WebInspector.UISourceCode.prototype.revertToOriginal):
(WebInspector.UISourceCode.prototype.revertAndClearHistory):
(WebInspector.UISourceCode.prototype.commitWorkingCopy):
* inspector/front-end/UserMetrics.js:
* inspector/front-end/externs.js:
* inspector/front-end/inspector.js:
(WebInspector._requestZoom.set var):
(WebInspector._requestZoom):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@135588 268f45cc-cd09-0410-ab3c-d52691b4dbfc