Move WebKitSite to Websites

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74395 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Move BugsSite to Websites

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74394 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Move PlanetWebKit to Websites

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74393 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add empty directory to hold web sites

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74392 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Yuta Kitamura  <yutak@chromium.org>

        Unreviewed.

        Update test expectations of http/tests/inspector/console-websocket-error.html
        on Mac, and add the test to GTK's Skipped list.

        * platform/gtk/Skipped:
        * platform/mac-leopard/http/tests/inspector/console-websocket-error-expected.txt: Copied from LayoutTests/platform/mac/http/tests/inspector/console-websocket-error-expected.txt.
        * platform/mac/http/tests/inspector/console-websocket-error-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74391 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Yuta Kitamura  <yutak@chromium.org>

        Reviewed by Alexey Proskuryakov.

        WebSocket errors should be logged to console
        https://bugs.webkit.org/show_bug.cgi?id=40945

        * http/tests/inspector/console-websocket-error.html: Added.
        * platform/mac/http/tests/inspector/console-websocket-error-expected.txt: Added.
        * platform/win/Skipped: Added console-websocket-error.html
        because HTTPS server is not supported on Windows yet.
2010-12-20  Yuta Kitamura  <yutak@chromium.org>

        Reviewed by Alexey Proskuryakov.

        WebSocket errors should be logged to console
        https://bugs.webkit.org/show_bug.cgi?id=40945

        Test: http/tests/inspector/console-websocket-error.html

        * platform/network/SocketStreamErrorBase.cpp:
        (WebCore::SocketStreamErrorBase::compare):
        * platform/network/SocketStreamErrorBase.h:
        (WebCore::SocketStreamErrorBase::failingURL):
        (WebCore::SocketStreamErrorBase::localizedDescription):
        (WebCore::SocketStreamErrorBase::SocketStreamErrorBase):
        * platform/network/cf/SocketStreamError.h:
        (WebCore::SocketStreamError::SocketStreamError):
        * platform/network/cf/SocketStreamHandle.h:
        * platform/network/cf/SocketStreamHandleCFNet.cpp:
        (WebCore::SocketStreamHandle::readStreamCallback):
        (WebCore::SocketStreamHandle::writeStreamCallback):
        (WebCore::SocketStreamHandle::reportErrorToClient):
        * websockets/WebSocketChannel.cpp:
        (WebCore::WebSocketChannel::didFail):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74390 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r74349.
        http://trac.webkit.org/changeset/74349
        https://bugs.webkit.org/show_bug.cgi?id=51372

        We are going to do better than reverting the code. (Requested
        by tonikitoo on #webkit).

        * fast/repaint/no-caret-repaint-in-non-content-editable-element.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74389 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Tony Gentilcore  <tonyg@chromium.org>

        Unreviewed build fix.

        Fix line breaks in fast/gradients/gradient-after-transparent-border expectations
        https://bugs.webkit.org/show_bug.cgi?id=51375

        * fast/gradients/gradient-after-transparent-border-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74388 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Tony Gentilcore  <tonyg@chromium.org>

        Unreviewed build fix.

        [chromium] Rebaseline inspector/console-dir-global and debugger-expand-scope
        https://bugs.webkit.org/show_bug.cgi?id=51373

        * platform/chromium-win/inspector/console-dir-global-expected.txt:
        * platform/chromium-win/inspector/debugger-expand-scope-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74387 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  MORITA Hajime  <morrita@google.com>

        Reviewed by James Robinson.

        https://bugs.webkit.org/show_bug.cgi?id=51240
        [Chromium][Skia] Border with a color with alpha != 1 breaks webkit gradient on skia.

        * fast/gradients/gradient-after-transparent-border-expected.txt: Added.
        * fast/gradients/gradient-after-transparent-border.html: Added.
        * platform/chromium-win/fast/gradients/gradient-after-transparent-border-expected.checksum: Added.
        * platform/chromium-win/fast/gradients/gradient-after-transparent-border-expected.png: Added.
2010-12-17  MORITA Hajime  <morrita@google.com>

        Reviewed by James Robinson.

        https://bugs.webkit.org/show_bug.cgi?id=51240
        [Chromium][Skia] Border with a color with alpha != 1 breaks webkit gradient on skia.

        Even after a shader was set, old color value had been remaining and passed to Skia.
        Then Skia refered an alpha component of that color to make pixels transparent.
        This change invalidates existing color value when a shader is given,
        which means the renderer will paint shapes with gradients or patterns.

        Test: fast/gradients/gradient-after-transparent-border.html

        * platform/graphics/skia/PlatformContextSkia.cpp:
        (WebCore::PlatformContextSkia::setStrokeShader):
        (WebCore::PlatformContextSkia::setFillShader):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74386 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Xiaomei Ji  <xji@chromium.org>

        Reviewed by Dan Bernstein.

        Fix move left/right by character hang.
        https://bugs.webkit.org/show_bug.cgi?id=51344

        Test: editing/selection/51344.html

        * editing/VisiblePosition.cpp:
        (WebCore::VisiblePosition::leftVisuallyDistinctCandidate):
        (WebCore::VisiblePosition::rightVisuallyDistinctCandidate):
2010-12-20  Xiaomei Ji  <xji@chromium.org>

        Reviewed by Dan Bernstein.

        Fix move left/right by character hang.
        https://bugs.webkit.org/show_bug.cgi?id=51344

        * editing/selection/51344-expected.txt: Added.
        * editing/selection/51344.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74385 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebKit2: Add resource load client API
https://bugs.webkit.org/show_bug.cgi?id=51370

Reviewed by Anders Carlsson.

This just pipes through the basic resource load client data.

* UIProcess/API/C/WKPage.cpp:
(WKPageSetPageResourceLoadClient):
* UIProcess/API/C/WKPage.h:
* UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::initializeResourceLoadClient):
(WebKit::WebPageProxy::didInitiateLoadForResource):
(WebKit::WebPageProxy::didSendRequestForResource):
(WebKit::WebPageProxy::didReceiveResponseForResource):
(WebKit::WebPageProxy::didReceiveContentLengthForResource):
(WebKit::WebPageProxy::didFinishLoadForResource):
(WebKit::WebPageProxy::didFailLoadForResource):
* UIProcess/WebPageProxy.h:
* UIProcess/WebPageProxy.messages.in:
* UIProcess/WebResourceLoadClient.cpp: Added.
(WebKit::WebResourceLoadClient::didInitiateLoadForResource):
(WebKit::WebResourceLoadClient::didSendRequestForResource):
(WebKit::WebResourceLoadClient::didReceiveResponseForResource):
(WebKit::WebResourceLoadClient::didReceiveContentLengthForResource):
(WebKit::WebResourceLoadClient::didFinishLoadForResource):
(WebKit::WebResourceLoadClient::didFailLoadForResource):
* UIProcess/WebResourceLoadClient.h: Added.
* WebKit2.pro:
* WebKit2.xcodeproj/project.pbxproj:
* WebProcess/WebCoreSupport/WebFrameLoaderClient.cpp:
(WebKit::WebFrameLoaderClient::assignIdentifierToInitialRequest):
(WebKit::WebFrameLoaderClient::dispatchWillSendRequest):
(WebKit::WebFrameLoaderClient::dispatchDidReceiveResponse):
(WebKit::WebFrameLoaderClient::dispatchDidReceiveContentLength):
(WebKit::WebFrameLoaderClient::dispatchDidFinishLoading):
(WebKit::WebFrameLoaderClient::dispatchDidFailLoading):
* win/WebKit2.vcproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74384 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Tony Gentilcore  <tonyg@chromium.org>

        Unreviewed build fix.

        [chromium] Add baseline for table-before-child-add.html
        https://bugs.webkit.org/show_bug.cgi?id=51369

        * platform/chromium-linux/fast/css-generated-content/table-before-child-add-expected.checksum: Added.
        * platform/chromium-linux/fast/css-generated-content/table-before-child-add-expected.png: Added.
        * platform/chromium-mac/fast/css-generated-content/table-before-child-add-expected.checksum: Added.
        * platform/chromium-mac/fast/css-generated-content/table-before-child-add-expected.png: Added.
        * platform/chromium-win/fast/css-generated-content/table-before-child-add-expected.checksum: Added.
        * platform/chromium-win/fast/css-generated-content/table-before-child-add-expected.png: Added.
        * platform/chromium-win/fast/css-generated-content/table-before-child-add-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74383 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Committing for Gavin since he needed to leave.

PPC build fix; stop using std::swap on PageAllocation/PageReservation,
this was failing on some compilers since the lack of default construction
for the m_executable/m_writable fields meant the value being swapped may
not have been fully initialized.

Patch by Gavin Barraclough <barraclough@apple.com> on 2010-12-20
* wtf/PageAllocation.h:
(WTF::PageAllocation::deallocate):
* wtf/PageBlock.h:
* wtf/PageReservation.h:
(WTF::PageReservation::deallocate):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74382 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Steve Lacey  <sjl@chromium.org>

        Reviewed by David Levin.

        Add missing default styles in mediaControlsChromium.css to prevent
        user overrides.
        https://bugs.webkit.org/show_bug.cgi?id=51360

        No new tests needed as functionality is covered by existing layout
        tests.

        * css/mediaControlsChromium.css:
        (audio::-webkit-media-controls-current-time-display, video::-webkit-media-controls-current-time-display):
        (audio::-webkit-media-controls-time-remaining-display, video::-webkit-media-controls-time-remaining-display):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74381 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add lastModifiedDate attribute to File interface per File API spec
https://bugs.webkit.org/show_bug.cgi?id=49895

Reviewed by David Levin.

WebCore:

lastModifiedDate is not supported for GObject bindings because Date is
not support in code generation.

Test: http/tests/local/fileapi/file-last-modified.html

* fileapi/File.cpp:
(WebCore::File::lastModifiedDate):
* fileapi/File.h:
* fileapi/File.idl:

LayoutTests:

* http/tests/local/fileapi/file-last-modified-expected.txt: Added.
* http/tests/local/fileapi/file-last-modified.html: Added.
* http/tests/local/fileapi/script-tests/file-last-modified.js: Added.
* http/tests/resources/file-last-modified.php: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74380 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Adrienne Walker  <enne@google.com>

        Unreviewed. Adding myself to the list of committers.

        * webkitpy/common/config/committers.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74379 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Joseph Pecoraro  <joepeck@webkit.org>

        Reviewed by Darin Adler.

        Update .gitattributes for WebKitTools -> Tools change
        https://bugs.webkit.org/show_bug.cgi?id=51365

        * .gitattributes: WebKitTools -> Tools rename. Removed no longer existing file references.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74378 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Oliver Hunt  <oliver@apple.com>

        Reviewed by Geoffrey Garen.

        |delete name| in strict mode code should be an early error
        https://bugs.webkit.org/show_bug.cgi?id=50431

        Disallow the |delete IDENTIFIER| production in strict mode, and removed
        a bunch of now unnecessary code.

        * parser/JSParser.cpp:
        (JSC::JSParser::Scope::collectFreeVariables):
        (JSC::jsParse):
        (JSC::JSParser::parseProgram):
        (JSC::JSParser::parseUnaryExpression):
        * parser/JSParser.h:
        * parser/Parser.cpp:
        (JSC::Parser::parse):
        * parser/Parser.h:
        (JSC::Parser::parse):
2010-12-20  Oliver Hunt  <oliver@apple.com>

        Reviewed by Geoffrey Garen.

        |delete name| in strict mode code should be an early error
        https://bugs.webkit.org/show_bug.cgi?id=50431

        Add a test to ensure that |delete IDENTIFIER| is disallowed, even if
        IDENTIFIER resolves to a deletable property.

        * fast/js/basic-strict-mode-expected.txt:
        * fast/js/script-tests/basic-strict-mode.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74377 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Anders Carlsson  <andersca@apple.com>

        Reviewed by Sam Weinig.

        Add a TextChecker class, hook up spelling and grammar toggling
        https://bugs.webkit.org/show_bug.cgi?id=51363

        * UIProcess/API/mac/WKView.mm:
        (-[WKView validateUserInterfaceItem:]):
        (-[WKView toggleContinuousSpellChecking:]):
        (-[WKView toggleGrammarChecking:]):
        * UIProcess/TextChecker.h: Added.
        * UIProcess/WebPageProxy.cpp:
        (WebKit::WebPageProxy::unmarkAllMisspellings):
        (WebKit::WebPageProxy::unmarkAllBadGrammar):
        * UIProcess/WebPageProxy.h:
        * UIProcess/mac/TextCheckerMac.mm: Added.
        (WebKit::TextChecker::isContinuousSpellCheckingAllowed):
        (WebKit::TextChecker::isContinuousSpellCheckingEnabled):
        (WebKit::TextChecker::setContinuousSpellCheckingEnabled):
        (WebKit::TextChecker::isGrammarCheckingEnabled):
        (WebKit::TextChecker::setGrammarCheckingEnabled):
        * UIProcess/qt/TextCheckerQt.cpp: Added.
        (WebKit::TextChecker::isContinuousSpellCheckingAllowed):
        (WebKit::TextChecker::isContinuousSpellCheckingEnabled):
        (WebKit::TextChecker::setContinuousSpellCheckingEnabled):
        (WebKit::TextChecker::isGrammarCheckingEnabled):
        (WebKit::TextChecker::setGrammarCheckingEnabled):
        * UIProcess/win/TextCheckerWin.cpp: Added.
        (WebKit::TextChecker::isContinuousSpellCheckingAllowed):
        (WebKit::TextChecker::isContinuousSpellCheckingEnabled):
        (WebKit::TextChecker::setContinuousSpellCheckingEnabled):
        (WebKit::TextChecker::isGrammarCheckingEnabled):
        (WebKit::TextChecker::setGrammarCheckingEnabled):
        * WebKit2.pro:
        * WebKit2.xcodeproj/project.pbxproj:
        * WebProcess/WebPage/WebPage.cpp:
        (WebKit::WebPage::unmarkAllMisspellings):
        (WebKit::WebPage::unmarkAllBadGrammar):
        * WebProcess/WebPage/WebPage.h:
        * WebProcess/WebPage/WebPage.messages.in:
        * win/WebKit2.vcproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74376 268f45cc-cd09-0410-ab3c-d52691b4dbfc

<rdar://problem/8757601> Don't install header files in to the Resources directory.

Reviewed by Dan Bernstein.

Tools:

* Scripts/check-for-inappropriate-files-in-framework: Don't allow .h files in the
Resources directory.

WebCore:

* WebCore.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74375 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Bug 51358 - Should check stack depth rather than using recursion limits in byte compilation

Reviewed by Olver Hunt.

The current implementation of recursion limit checking is not safe on smaller stacks.
Switch to using a common mechanism, shared with the parser, to check recursion limits.

Make bytecompiler use StackBounds. Empirical testing shows emitStrcat to have the largest
footprint on the stack, at just under 1k on x86-64.  Given this, the default recursion
check (requiring 4k of available space to recurse) seems reasonable.

* bytecompiler/BytecodeGenerator.cpp:
(JSC::BytecodeGenerator::BytecodeGenerator):
* bytecompiler/BytecodeGenerator.h:
(JSC::BytecodeGenerator::emitNode):
(JSC::BytecodeGenerator::emitNodeInConditionContext):
* bytecompiler/NodesCodegen.cpp:
(JSC::BinaryOpNode::emitStrcat):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74374 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        commit-queue should include bot id when attaching failure diffs
        https://bugs.webkit.org/show_bug.cgi?id=51280

        This is a tiny change to include the bot id in the name of the attachment.
        Most of this diff is just changing the unit test expectations
        now that I hid the comment printing when the comment is None.

        * Scripts/webkitpy/tool/bot/flakytestreporter.py:
        * Scripts/webkitpy/tool/bot/flakytestreporter_unittest.py:
        * Scripts/webkitpy/tool/commands/upload_unittest.py:
        * Scripts/webkitpy/tool/mocktool.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74373 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Andrei Popescu  <andreip@google.com>

        Reviewed by Jeremy Orlow.

        [Chromium] v8/SerializedScriptValue::readUint32 treats incoming raw values as signed instead of unsigned
        https://bugs.webkit.org/show_bug.cgi?id=51338

        * fast/dom/Window/window-postmessage-args-expected.txt:
        * fast/dom/Window/window-postmessage-args.html:
        * platform/chromium-mac/fast/dom/Window/window-postmessage-args-expected.txt:
        * platform/chromium-win/fast/dom/Window/window-postmessage-args-expected.txt:
2010-12-20  Andrei Popescu  <andreip@google.com>

        Reviewed by Jeremy Orlow.

        [Chromium] v8/SerializedScriptValue::readUint32 treats incoming raw values as signed instead of unsigned
        https://bugs.webkit.org/show_bug.cgi?id=51338

        * bindings/v8/SerializedScriptValue.cpp:
        (WebCore::ZigZag::Reader::readUint32):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74372 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Tony Gentilcore  <tonyg@chromium.org>

        Unreviewed build fix.

        Include pthread to fix chromium mac build (broken by r74360)
        https://bugs.webkit.org/show_bug.cgi?id=51356

        * wtf/StackBounds.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74371 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore: Fix for https://bugs.webkit.org/show_bug.cgi?id=51355 Crash in
RenderMathMLSubSup::layout() because of null base
-and corresponding-
<rdar://problem/8712200>

Reviewed by Ada Chan.

Only use base inside if(base) block.
* mathml/RenderMathMLSubSup.cpp:
(WebCore::RenderMathMLSubSup::layout):

LayoutTests: Test for https://bugs.webkit.org/show_bug.cgi?id=51355 Crash in
RenderMathMLSubSup::layout() because of null base
-and corresponding-
<rdar://problem/8712200>

Reviewed by Ada Chan.

* mathml/empty-msubsup-crash-expected.txt: Added.
* mathml/empty-msubsup-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74370 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Crash at WebCore::SplitElementCommand::SplitElementCommand
https://bugs.webkit.org/show_bug.cgi?id=51278
<rdar://problem/7347139>

Reviewed by Darin Adler.

WebCore:

Executing outdent command on an inline blockquote that is
not the first element in a block causes a crash.
The fix consists in chaging the was the start of block is
calculated in case of the inline blockquote.

Test: editing/execCommand/outdent-inline-blockquote.html
      editing/execCommand/outdent-inline-list.html

* editing/IndentOutdentCommand.cpp:
(WebCore::IndentOutdentCommand::outdentParagraph):

LayoutTests:

* editing/execCommand/outdent-inline-blockquote-expected.txt: Added.
* editing/execCommand/outdent-inline-blockquote.html: Added.
* editing/execCommand/outdent-inline-list-expected.txt: Added.
* editing/execCommand/outdent-inline-list.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74369 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Oliver Hunt  <oliver@apple.com>

        Reviewed by Darin Adler.

        Need to support serialisation of cyclic graphs in the internal structured cloning algorithm
        https://bugs.webkit.org/show_bug.cgi?id=51353

        Update test to cover correct behaviour, and extend to test for actual graph construction.

        * fast/dom/Window/window-postmessage-clone-expected.txt:
        * fast/dom/Window/window-postmessage-clone.html:
2010-12-20  Oliver Hunt  <oliver@apple.com>

        Reviewed by Darin Adler.

        Need to support serialisation of cyclic graphs in the internal structured cloning algorithm
        https://bugs.webkit.org/show_bug.cgi?id=51353

        The Internal Structured Clone algorithm has been changed to allow (and
        correctly clone) cyclic graphs.  This patch updates our implementation
        to provide that functionality.

        I've bumped the serialization version number, and added ObjectReferenceTag
        to represent references to objects that have already been seen.

        * bindings/js/SerializedScriptValue.cpp:
        (WebCore::CloneSerializer::startObjectInternal):
          Now that we have something a bit more complex than cycle checking
          I've replaced the duplicate code in startObject and startArray with
          a shared function that implements that logic to plant an object
          reference
        (WebCore::CloneSerializer::startObject):
        (WebCore::CloneSerializer::startArray):
          Lift out duplicate code
        (WebCore::CloneSerializer::endObject):
          Can't remove objects from the gcbuffer now as they need to remain live
          so we can identify graphs
        (WebCore::CloneSerializer::writeStringIndex):
        (WebCore::CloneSerializer::writeObjectIndex):
        (WebCore::CloneSerializer::writeConstantPoolIndex):
        (WebCore::CloneSerializer::write):
        (WebCore::CloneSerializer::serialize):
        (WebCore::CloneDeserializer::readStringIndex):
        (WebCore::CloneDeserializer::readConstantPoolIndex):
        (WebCore::CloneDeserializer::readTerminal):
        (WebCore::CloneDeserializer::deserialize):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74368 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Reviewed by Xan Lopez.

        Add WebPlugin, WebPluginDatabase, and ViewportAttributes to the
        documentation control files, in preparation for 1.3.8. Also take
        the opportunity to add descriptions to WebPlugin and
        WebPluginDatabase.

        * docs/GNUmakefile.am:
        * docs/webkitgtk-docs.sgml:
        * docs/webkitgtk-sections.txt:
        * webkit/webkitwebplugin.cpp:
        * webkit/webkitwebplugindatabase.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74367 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        * configure.ac: bump version for 1.3.8 release.

WebCore:

2010-12-20  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        * GNUmakefile.am: add missing files.

WebKit/gtk:

2010-12-20  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        * GNUmakefile.am: add missing files.
        * NEWS: update for 1.3.8 release.

JavaScriptCore:

2010-12-20  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        * GNUmakefile.am: add missing files.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74366 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Tools/vcbin/midl.exe needs to be rebuilt
https://bugs.webkit.org/show_bug.cgi?id=51347

Patch by Jeff Miller <jeffm@apple.com> on 2010-12-20
Reviewed by Dan Bernstein.

* vcbin/midl.exe: Rebuilt.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74365 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Abhishek Arya  <inferno@chromium.org>

        Reviewed by James Robinson.

        Pass the correct beforeChild value when adding a child to an anonymous table part.
        https://bugs.webkit.org/show_bug.cgi?id=50932

        When a "before" content child was added to an anonymous table part, it was added
        after the last child. This patch fixes the behavior by adding it properly before
        the first child.

        Test: fast/css-generated-content/table-before-child-add.html

        * rendering/RenderTableRow.cpp:
        (WebCore::RenderTableRow::addChild):
2010-12-20  Abhishek Arya  <inferno@chromium.org>

        Reviewed by James Robinson.

        Tests that beforeChild is added to table correctly.
        https://bugs.webkit.org/show_bug.cgi?id=50932

        * fast/css-generated-content/table-before-child-add.html: Added.
        * platform/mac/fast/css-generated-content/table-before-child-add-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74364 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  David Levin  <levin@chromium.org>

        Reviewed by Ariya Hidayat.

        .gitignore and webkit-tools-completion.sh have references to WebKitTools that should be Tools.
        https://bugs.webkit.org/show_bug.cgi?id=51343

        Follow up from the WebKitTools -> Tools rename.

        * .gitignore:
2010-12-20  David Levin  <levin@chromium.org>

        Reviewed by Ariya Hidayat.

        .gitignore and webkit-tools-completion.sh have references to WebKitTools that should be Tools.
        https://bugs.webkit.org/show_bug.cgi?id=51343

        Follow up from the WebKitTools -> Tools rename.

        * Scripts/webkit-tools-completion.sh:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74363 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Cosmin Truta  <ctruta@chromium.org>

        Reviewed by James Robinson.

        new-run-webkit-tests ignores trailing EOL differences in text tests
        https://bugs.webkit.org/show_bug.cgi?id=36983

        Changed the handling of new-line characters within new-run-webkit-tests
        to match old-run-webkit-tests.  Differences in leading and trailing empty
        lines in text expectation files are no longer ignored.

        * Scripts/webkitpy/layout_tests/port/base.py:
        * Scripts/webkitpy/layout_tests/port/test.py:
        Added unit tests.  Removed old duplicate unit test entries.
        * Scripts/webkitpy/layout_tests/test_types/text_diff.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74362 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Cosmin Truta  <ctruta@chromium.org>

        Reviewed by James Robinson.

        Even more rebaselining needed for new-run-webkit-tests update
        https://bugs.webkit.org/show_bug.cgi?id=51340

        * http/tests/misc/webtiming-ssl-expected.txt:
        * storage/indexeddb/cursor-delete-expected.txt:
        * storage/indexeddb/cursor-index-delete-expected.txt:
        * storage/indexeddb/transaction-crash-on-abort-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74361 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Bug 26276 - Need a mechanism to determine stack extent

Reviewed by Oliver Hunt.

JavaScriptCore:

This patch adds a class 'StackBounds', to hold information about the machine stack.
The implementation of this class broadly adheres to the current implmentation of
stack limit checking, and as such does not solve the problem of determining stack
extent, but gives us a common place to do so.

Currently two mechanism are provided to determine the stack origin (the point the
stack is growing away from). currentThreadStackBase() in Collector provides a
more accurate determination of the stack origin, so use this to calculate
StackBounds::m_origin; WTFThreadData::approximatedStackStart is less accurate, and
as such can be removed.  Cache the StackBounds on WTFThreadData such that they
need only be determined once per thread, and for non-API contexts cache this
information in JSGlobalData, to save a thread-specific access.

For the time being retain the estimate of stack size used by JSC's parser
(128 * sizeof(void*) * 1024), with a view to replacing this with something more
accurate in the near future.

* parser/JSParser.cpp:
(JSC::JSParser::canRecurse):
(JSC::JSParser::JSParser):
    Change to use StackBounds.
* runtime/Collector.cpp:
(JSC::Heap::registerThread):
(JSC::Heap::markCurrentThreadConservativelyInternal):
    Change to use StackBounds, cached on JSGlobalData.
* runtime/JSGlobalData.cpp:
(JSC::JSGlobalData::JSGlobalData):
* runtime/JSGlobalData.h:
(JSC::JSGlobalData::stack):
    Add a cached copy of StackBounds.
* wtf/StackBounds.cpp: Copied from JavaScriptCore/runtime/Collector.cpp.
(WTF::estimateStackBound):
(WTF::StackBounds::initialize):
(WTF::getStackMax):
    Copy code from Collector.cpp to determine stack origin.
* wtf/StackBounds.h: Added.
(WTF::StackBounds::StackBounds):
    No argument constructor; returns a null StackBounds.
(WTF::StackBounds::currentThreadStackBounds):
    Returns a StackBounds object representing the stack limits
    of the current thread.
(WTF::StackBounds::origin):
    Returns to stack origin (the point the stack is growing away
    from; the highest extent of the stack on machines where the
    stack grows downwards.
(WTF::StackBounds::recursionLimit):
    Returns a limit value that is 'a comfortable distance from
    the end of the stack'. Our concept of this is currently 1 page
    away from the end, however the default value may be tuned in
    the future, and clients may override passing a larger delta;
    should only be called on StackBounds object representing the
    stack of the thread this method is called on (checked by
    checkConsistency).
(WTF::StackBounds::recursionCheck):
    Checks whether we are currently 'a comfortable distance from
    the end of the stack'. Our concept of this is currently 1 page
    away from the end, however the default value may be tuned in
    the future, and clients may override passing a larger delta
    to apply when checking, if they wish to do so. This method
    should only be called on StackBounds object representing the
    stack of the thread this method is called on (checked by
    checkConsistency).
(WTF::StackBounds::current):
    Approximate current stack position. On machines where the stack
    is growing downwards this is the lowest address that might need
    conservative collection.
(WTF::StackBounds::isGrowingDownward):
    True for all platforms other than WINCE, which has to check.
(WTF::StackBounds::checkConsistency):
    This is called in methods that shoulds only be operating on a
    valid set of bounds; as such we expect m_origin != m_bounds
    (i.e. stack size != zero) - we're really testing that this
    object is not null (the constructor initializes both fields
    to zero).  Also checks that current() is within the stack's
    bounds.
* wtf/WTFThreadData.cpp:
(WTF::WTFThreadData::WTFThreadData):
* wtf/WTFThreadData.h:
(WTF::WTFThreadData::stack):
    Add the StackBounds member variable.

JavaScriptGlue:

Add forwarding header for StackBounds.h.

* ForwardingHeaders/wtf/StackBounds.h: Added.

WebCore:

Add forwarding header for StackBounds.h.

* ForwardingHeaders/wtf/StackBounds.h: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74360 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Anders Carlsson  <andersca@apple.com>

        Reviewed by Sam Weinig.

        Flash plug-in does not work on YouTube video page when accelerated compositing is disabled
        https://bugs.webkit.org/show_bug.cgi?id=51348
        <rdar://problem/8787667>

        * Platform/CoreIPC/HandleMessage.h:
        (CoreIPC::callMemberFunction):
        Add new overload.

        * PluginProcess/PluginControllerProxy.cpp:
        (WebKit::PluginControllerProxy::create):
        (WebKit::PluginControllerProxy::PluginControllerProxy):
        These now take an additional isAcceleratedCompositingEnabled parameter.

        (WebKit::PluginControllerProxy::isAcceleratedCompositingEnabled):
        Return m_isAcceleratedCompositingEnabled.

        * PluginProcess/PluginControllerProxy.h:
        * PluginProcess/WebProcessConnection.cpp:
        (WebKit::WebProcessConnection::createPlugin):
        This now takes an additional isAcceleratedCompositingEnabled parameter.

        * PluginProcess/WebProcessConnection.messages.in:
        Add isAcceleratedCompositingEnabled parameter to the CreatePlugin message.

        * WebProcess/Plugins/PluginProxy.cpp:
        (WebKit::PluginProxy::initialize):
        Pass isAcceleratedCompositingEnabled when sending the CreatePlugin message.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74359 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  James Robinson  <jamesr@chromium.org>

        Unreviewed, rolling out r74278.
        http://trac.webkit.org/changeset/74278
        https://bugs.webkit.org/show_bug.cgi?id=50833

        [chromium] Causes many layout tests to crash

        * platform/graphics/chromium/LayerRendererChromium.cpp:
        (WebCore::LayerRendererChromium::LayerRendererChromium):
        (WebCore::LayerRendererChromium::prepareToDrawLayers):
        (WebCore::LayerRendererChromium::drawLayers):
        (WebCore::LayerRendererChromium::useRenderSurface):
        (WebCore::LayerRendererChromium::setScissorToRect):
        * platform/graphics/chromium/LayerRendererChromium.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74358 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Sam Weinig.

        Factored common page set management into a new PageBlock base class
        https://bugs.webkit.org/show_bug.cgi?id=51285

        * Android.mk:
        * CMakeLists.txt:
        * GNUmakefile.am:
        * JavaScriptCore.gypi:
        * JavaScriptCore.vcproj/WTF/WTF.vcproj:
        * JavaScriptCore.xcodeproj/project.pbxproj:
        * interpreter/RegisterFile.h:
        (JSC::RegisterFile::RegisterFile):
        * jit/ExecutableAllocator.cpp:
        (JSC::ExecutableAllocator::intializePageSize):
        * wtf/PageAllocation.cpp: Removed.
        * wtf/PageAllocation.h:
        (WTF::PageAllocation::deallocate):
        (WTF::PageAllocation::PageAllocation):
        * wtf/PageReservation.h:
        (WTF::PageReservation::commit):
        (WTF::PageReservation::decommit):
        (WTF::PageReservation::deallocate):
        (WTF::PageReservation::PageReservation):
        * wtf/wtf.pri:
2010-12-20  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Sam Weinig.

        Factored common page set management into a new PageBlock base class
        https://bugs.webkit.org/show_bug.cgi?id=51285

        * ForwardingHeaders/wtf/PageBlock.h: Added.
2010-12-20  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Sam Weinig.

        Factored common page set management into a new PageBlock base class
        https://bugs.webkit.org/show_bug.cgi?id=51285

        * ForwardingHeaders/wtf/PageBlock.h: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74357 268f45cc-cd09-0410-ab3c-d52691b4dbfc

check-webkit-style should detect function declarations (and trivial functions).
https://bugs.webkit.org/show_bug.cgi?id=51303

Reviewed by Shinichiro Hamaji.

* Scripts/webkitpy/style/checkers/cpp.py:
(_FunctionState.begin): Add is_declaration and changed the line count
start to begin at -1 (which will keep the results consistent, since
the starting line number passed in is one less in this change).
(detect_functions): changed function detection to now catch trivial
functions and declarations.
(check_pass_ptr_usage): Don't check for Pass*Ptr on the first line
of the function as this may look at return values (when processing
a declaration).
* Scripts/webkitpy/style/checkers/cpp_unittest.py:
(FunctionDetectionTest.perform_function_detection): Basic mechanics
of testing the function detection.
(FunctionDetectionTest.test_basic_function_detection): Test a simple
function.
(FunctionDetectionTest.test_function_declaration_detection): Test a
declaration.
(FunctionDetectionTest.test_non_functions): A test case for a case
that caused the code to fail due to the { being in quotes.
(PassPtrTest.test_pass_ref_ptr_return_value): Added some more test
cases to help catch false alarms for return values.
(PassPtrTest.test_pass_ref_ptr_member_variable): Ensure that
we don't get false alarms for member variables either.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74356 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Anders Carlsson  <andersca@apple.com>

        Unreviewed, rolling out r74345.
        http://trac.webkit.org/changeset/74345
        https://bugs.webkit.org/show_bug.cgi?id=51115

        Not the correct fix.

        * Platform/CoreIPC/Connection.h:
        * Platform/CoreIPC/qt/ConnectionQt.cpp:
        (CoreIPC::Connection::open):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74355 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Steve Block  <steveblock@google.com>

        Reviewed by Jeremy Orlow.

        GeolocationPositionCache should do database access on background thread
        https://bugs.webkit.org/show_bug.cgi?id=50825

        Covered by existing Geolocation tests.

        * page/Coordinates.h:
        (WebCore::Coordinates::threadSafeCopy):
        * page/GeolocationPositionCache.cpp:
        (WebCore::GeolocationPositionCache::GeolocationPositionCache):
        (WebCore::GeolocationPositionCache::addUser):
        (WebCore::GeolocationPositionCache::removeUser):
        (WebCore::GeolocationPositionCache::setDatabasePath):
        (WebCore::GeolocationPositionCache::setCachedPosition):
        (WebCore::GeolocationPositionCache::cachedPosition):
        (WebCore::GeolocationPositionCache::startBackgroundThread):
        (WebCore::GeolocationPositionCache::threadEntryPoint):
        (WebCore::GeolocationPositionCache::threadEntryPointImpl):
        (WebCore::GeolocationPositionCache::triggerReadFromDatabase):
        (WebCore::GeolocationPositionCache::readFromDatabase):
        (WebCore::GeolocationPositionCache::readFromDatabaseImpl):
        (WebCore::GeolocationPositionCache::triggerWriteToDatabase):
        (WebCore::GeolocationPositionCache::writeToDatabase):
        (WebCore::GeolocationPositionCache::writeToDatabaseImpl):
        * page/GeolocationPositionCache.h:
        * page/Geoposition.h:
        (WebCore::Geoposition::threadSafeCopy):
        (WebCore::Geoposition::Geoposition):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74354 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Andras Becsi  <abecsi@webkit.org>

        Unreviewed build fix.

        [Qt][V8] Add  missing sources to the pro file and remove duplications to fix linking.

        No new tests needed.

        * WebCore.pro:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74353 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Alexander Pavlov  <apavlov@chromium.org>

        Reviewed by Yury Semikhatsky.

        Web Inspector: Make CSSCompletions a real class rather than singleton
        https://bugs.webkit.org/show_bug.cgi?id=51332

Drive-by: make use of keywords from re2js in the CSS tokenizer
        (apparently, various authors have presumed that re2js is processed at build time).

        No new tests, as this is a refactoring.

        * inspector/front-end/CSSCompletions.js:
        (WebInspector.CSSCompletions):
        (WebInspector.CSSCompletions.prototype.startsWith):
        (WebInspector.CSSCompletions.prototype.firstStartsWith):
        (WebInspector.CSSCompletions.prototype._firstIndexOfPrefix):
        (WebInspector.CSSCompletions.prototype.keySet):
        (WebInspector.CSSCompletions.prototype.next):
        (WebInspector.CSSCompletions.prototype.previous):
        (WebInspector.CSSCompletions.prototype._closest):
        * inspector/front-end/SourceCSSTokenizer.js:
        (WebInspector.SourceCSSTokenizer):
        * inspector/front-end/SourceCSSTokenizer.re2js:
        * inspector/front-end/StylesSidebarPane.js:
        (WebInspector.StylePropertyTreeElement.prototype):
        * inspector/front-end/inspector.js:
        (WebInspector.doLoadedDone.propertyNamesCallback):
        (WebInspector.doLoadedDone):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74352 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Andrey Kosyakov  <caseq@chromium.org>

        Unreviewed. Rebased platform-specific test expectations following default expectations change in r74234.

        * platform/mac-leopard/inspector/extensions-resources-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74351 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-10  Pavel Podivilov  <podivilov@chromium.org>

        Reviewed by Yury Semikhatsky.

        Web Inspector: "Add conditional breakpoint" - create breakpoint only after condition is provided.
        https://bugs.webkit.org/show_bug.cgi?id=50813

        * inspector/front-end/SourceFrame.js:
        (WebInspector.SourceFrame.prototype._addBreakpoint):
        (WebInspector.SourceFrame.prototype._breakpointRemoved):
        (WebInspector.SourceFrame.prototype._breakpointChanged):
        (WebInspector.SourceFrame.prototype._setBreakpointDecoration):
        (WebInspector.SourceFrame.prototype._removeBreakpointDecoration):
        (WebInspector.SourceFrame.prototype._contextMenu.addConditionalBreakpoint.didEditBreakpointCondition):
        (WebInspector.SourceFrame.prototype._contextMenu.addConditionalBreakpoint):
        (WebInspector.SourceFrame.prototype._contextMenu.else.editBreakpointCondition.didEditBreakpointCondition):
        (WebInspector.SourceFrame.prototype._contextMenu.else.editBreakpointCondition):
        (WebInspector.SourceFrame.prototype._contextMenu):
        (WebInspector.SourceFrame.prototype._editBreakpointCondition.finishEditing):
        (WebInspector.SourceFrame.prototype._editBreakpointCondition):
        (WebInspector.SourceFrame.prototype._continueToLine):
        (WebInspector.SourceFrame.prototype._setBreakpoint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74350 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-19  Antonio Gomes  <agomes@rim.com>

        Reviewed by Xan Lopez.

        no-caret-repaint-in-non-content-editable-element.html does not need to disable CaretBrowsing
        https://bugs.webkit.org/show_bug.cgi?id=51308

        Patch removes the useless statement to disable caret browsing from
        no-caret-repaint-in-non-content-editable-element.html.

        Caret browsing gets toggled OFF before the execution of each test
        on the DRTs that support testing it (GTK+, Qt and Chromium). On Mac or
        Windows's DRT testing this feature is unsupport, so caret
        browsing is also disabled by default.

        * fast/repaint/no-caret-repaint-in-non-content-editable-element.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74349 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-18  Antonio Gomes  <agomes@rim.com>

        Reviewed by Xan Lopez.

        [GTK+] Simplify spatial navigation handling in EditorClientGtk.cpp
        https://bugs.webkit.org/show_bug.cgi?id=51302

        Patch changes:
        1) The editor command for LEFT arrow key press without modifier
           from MoveBackward to MoveLeft;
        2) And similarly changes the editor command for RIGHT arrow key
          press without modifier from MoveForward to MoveRight.

        It has no implications since SelectionController::modifyMovingRight()
        handles "MoveRight" equaly to how SelectionController::modifyMovingForward()
        handles "MoveForward" when the granularity is 'CharacterGranularity'.
        The former falls back to the later in the case of other editor commands.

        Analogously, the same applies to how SelectionController::modifyMovingLeft() and
        SelectionController::modifyMovingForward() handle "MoveLeft" and "MoveBackward"
        editor commands, respectively.

        With this change we can remove the spatial navigation hack from moveCursorCallback(),
        in EditorClientGtk.cpp.

        * WebCoreSupport/EditorClientGtk.cpp:
        (WebKit::moveCursorCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74348 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Yury Semikhatsky  <yurys@chromium.org>

        Unreviewed. Removed custom expectation for a passing test.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74347 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Alejandro G. Castro  <alex@igalia.com>

        Unreviewed, added the proper expected result file.

        * platform/gtk/fast/text/emphasis-overlap-expected.txt: Added.
        * platform/gtk/fast/text/emphasis-overlap.html: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74346 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Balazs Kelemen  <kbalazs@webkit.org>

        Reviewed by Kenneth Rohde Christiansen.

        [Qt][WK2] Crash in web process after the connection had been closed
        https://bugs.webkit.org/show_bug.cgi?id=51115

        Do not call connectionDidClose with a connection that has been explicitly
        invalidated on the listener thread.
        * Platform/CoreIPC/Connection.h:
        * Platform/CoreIPC/qt/ConnectionQt.cpp:
        (CoreIPC::Connection::disconnectHandler): Added.
        (CoreIPC::Connection::open):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74345 268f45cc-cd09-0410-ab3c-d52691b4dbfc

[Qt] Improve the readability of FrameLoaderClientQt::createPlugin()
https://bugs.webkit.org/show_bug.cgi?id=51330

Reviewed by Andreas Kling.

Redefine the scopes to improve readability. Objects are now created in
the scope they are used in.

* WebCoreSupport/FrameLoaderClientQt.cpp:
(WebCore::FrameLoaderClientQt::createPlugin):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74344 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Andrei Popescu  <andreip@google.com>

        Reviewed by Jeremy Orlow.

        IDBTransactionBackedImpl instances can be accidentally deleted during calls to abort/commit.
        https://bugs.webkit.org/show_bug.cgi?id=51112

        * storage/indexeddb/transaction-crash-on-abort-expected.txt: Added.
        * storage/indexeddb/transaction-crash-on-abort.html: Added.

2010-12-20  Andrei Popescu  <andreip@google.com>

        Reviewed by Jeremy Orlow.

        IDBTransactionBackedImpl instances can be accidentally deleted during calls to abort/commit.
        https://bugs.webkit.org/show_bug.cgi?id=51112

        Makes IDBTransactionBackendImpl objects take a self reference while executing commit and abort.

        Test: storage/indexeddb/transaction-crash-on-abort.html

        * storage/IDBTransactionBackendImpl.cpp:
        (WebCore::IDBTransactionBackendImpl::abort):
        (WebCore::IDBTransactionBackendImpl::commit):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74343 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Andrei Popescu  <andreip@google.com>

        Reviewed by Jeremy Orlow.

        IDBCursor::delete is not implemented.
        https://bugs.webkit.org/show_bug.cgi?id=51110

        * storage/indexeddb/cursor-delete-expected.txt: Added.
        * storage/indexeddb/cursor-delete.html: Added.
        * storage/indexeddb/cursor-index-delete-expected.txt: Added.
        * storage/indexeddb/cursor-index-delete.html: Added.
2010-12-20  Andrei Popescu  <andreip@google.com>

        Reviewed by Jeremy Orlow.

        IDBCursor::delete is not implemented.
        https://bugs.webkit.org/show_bug.cgi?id=51110

        Implement IDBCursor::delete.
        See http://dvcs.w3.org/hg/IndexedDB/raw-file/tip/Overview.html#widl-IDBCursor-delete

        Tests: storage/indexeddb/cursor-delete.html
               storage/indexeddb/cursor-index-delete.html

        * storage/IDBCursor.cpp:
        (WebCore::IDBCursor::deleteFunction):
        * storage/IDBCursor.h:
        * storage/IDBCursor.idl:
        * storage/IDBCursorBackendImpl.cpp:
        (WebCore::IDBCursorBackendImpl::IDBCursorBackendImpl):
        (WebCore::IDBCursorBackendImpl::deleteFunction):
        (WebCore::IDBCursorBackendImpl::loadCurrentRow):
        * storage/IDBCursorBackendImpl.h:
        (WebCore::IDBCursorBackendImpl::create):
        * storage/IDBCursorBackendInterface.h:
        * storage/IDBIndexBackendImpl.cpp:
        (WebCore::IDBIndexBackendImpl::openCursorInternal):
        * storage/IDBKey.cpp:
        (WebCore::IDBKey::fromQuery):
        * storage/IDBObjectStoreBackendImpl.cpp:
        (WebCore::IDBObjectStoreBackendImpl::deleteInternal):
        (WebCore::IDBObjectStoreBackendImpl::openCursorInternal):
2010-12-20  Andrei Popescu  <andreip@google.com>

        Reviewed by Jeremy Orlow.

        IDBCursor::delete is not implemented.
        https://bugs.webkit.org/show_bug.cgi?id=51110

        * public/WebIDBCursor.h:
        (WebKit::WebIDBCursor::remove):
        (WebKit::WebIDBCursor::deleteFunction):
        * src/IDBCursorBackendProxy.cpp:
        (WebCore::IDBCursorBackendProxy::deleteFunction):
        * src/IDBCursorBackendProxy.h:
        * src/WebIDBCursorImpl.cpp:
        (WebKit::WebIDBCursorImpl::deleteFunction):
        * src/WebIDBCursorImpl.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74342 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Yury Semikhatsky  <yurys@chromium.org>

        Unreviewed. Update Chromium test expectations.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74341 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Alejandro G. Castro  <alex@igalia.com>

        Unreviewed expectation results update. Added also the tests
        failing due to the pixman rounding color issue to the same section
        of the skipped file: bug 49964.

        * platform/gtk/Skipped:
        * platform/gtk/fast/repaint/text-emphasis-h-expected.txt: Added.
        * platform/gtk/fast/repaint/text-emphasis-v-expected.txt: Added.
        * platform/gtk/fast/text/emphasis-overlap.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74340 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Xan Lopez  <xlopez@igalia.com>

        [GTK] editing/selection/extend-by-character-003.html is flakey
        https://bugs.webkit.org/show_bug.cgi?id=51325

        Skip flakey test.

        * platform/gtk/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74339 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Yury Semikhatsky  <yurys@chromium.org>

        Unreviewed. Update Chromium test expectations.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74338 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed.

[Qt] fast/text/emphasis-overlap.html fails
https://bugs.webkit.org/show_bug.cgi?id=51324

* platform/qt/Skipped: fast/text/emphasis-overlap.html added until fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74337 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed.

[Qt] Add platform specific expected result for new test introduced in r74321.

* platform/qt/fast/parser/xml-colon-entity-expected.txt: Copied from LayoutTests/fast/parser/xml-colon-entity-expected.txt.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74336 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed.

Add new tests to the Skipped list, because of missing layoutTestController.testRepaint()

* platform/qt/Skipped:
 - fast/repaint/text-emphasis-h.html added
 - fast/repaint/text-emphasis-v.html added

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74335 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Ryuan Choi  <ryuan.choi@samsung.com>

        Reviewed by Antonio Gomes.

        [CMAKE] Rename WEBKITTOOLS_DIR to TOOLS_DIR
        https://bugs.webkit.org/show_bug.cgi?id=51319

        Change WEBKITTOOLS_DIR variable's name and value to TOOLS_DIR and tools/.

        * CMakeLists.txt:
2010-12-20  Ryuan Choi  <ryuan.choi@samsung.com>

        Reviewed by Antonio Gomes.

        [CMAKE] Rename WEBKITTOOLS_DIR to TOOLS_DIR
        https://bugs.webkit.org/show_bug.cgi?id=51319

        * CMakeListsEfl.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74334 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-20  Yury Semikhatsky  <yurys@chromium.org>

        Reviewed by Pavel Feldman.

        Web Inspector: After a multiline command execution Console does not show the command prompt
        https://bugs.webkit.org/show_bug.cgi?id=51257

        * inspector/front-end/TextPrompt.js:
        (WebInspector.TextPrompt.prototype._moveBackInHistory):
        (WebInspector.TextPrompt.prototype._moveForwardInHistory):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74333 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Move WebKitExamplePlugins to Examples
https://bugs.webkit.org/show_bug.cgi?id=51291

Reviewed by Sam Weinig.

.:

* Examples: Copied from WebKitExamplePlugins.
* WebKitExamplePlugins: Removed.
* WebKitExamplePlugins/NetscapeCocoaPlugin: Removed.
* WebKitExamplePlugins/NetscapeCocoaPlugin/English.lproj: Removed.
* WebKitExamplePlugins/NetscapeCocoaPlugin/English.lproj/InfoPlist.strings: Removed.
* WebKitExamplePlugins/NetscapeCocoaPlugin/Info.plist: Removed.
* WebKitExamplePlugins/NetscapeCocoaPlugin/MenuHandler.h: Removed.
* WebKitExamplePlugins/NetscapeCocoaPlugin/MenuHandler.m: Removed.
* WebKitExamplePlugins/NetscapeCocoaPlugin/NetscapeCocoaPlugin.xcodeproj: Removed.
* WebKitExamplePlugins/NetscapeCocoaPlugin/NetscapeCocoaPlugin.xcodeproj/project.pbxproj: Removed.
* WebKitExamplePlugins/NetscapeCocoaPlugin/main.m: Removed.
* WebKitExamplePlugins/NetscapeCocoaPlugin/test.html: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/English.lproj: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/English.lproj/InfoPlist.strings: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/Info.plist: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/MovieControllerLayer.h: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/MovieControllerLayer.m: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/NetscapeCoreAnimationMoviePlugin.xcodeproj: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/NetscapeCoreAnimationMoviePlugin.xcodeproj/project.pbxproj: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/Pause.tiff: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/Play.tiff: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/SliderTrackCenter.tiff: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/SliderTrackLeft.tiff: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/SliderTrackRight.tiff: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/Thumb.tiff: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/main.m: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationMoviePlugin/test.html: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationPlugin: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationPlugin/Composition.qtz: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationPlugin/English.lproj: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationPlugin/English.lproj/InfoPlist.strings: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationPlugin/Info.plist: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationPlugin/NetscapeCoreAnimationPlugin.xcodeproj: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationPlugin/NetscapeCoreAnimationPlugin.xcodeproj/project.pbxproj: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationPlugin/main.m: Removed.
* WebKitExamplePlugins/NetscapeCoreAnimationPlugin/test.html: Removed.
* WebKitExamplePlugins/NetscapeInputMethodPlugin: Removed.
* WebKitExamplePlugins/NetscapeInputMethodPlugin/English.lproj: Removed.
* WebKitExamplePlugins/NetscapeInputMethodPlugin/English.lproj/InfoPlist.strings: Removed.
* WebKitExamplePlugins/NetscapeInputMethodPlugin/Info.plist: Removed.
* WebKitExamplePlugins/NetscapeInputMethodPlugin/NetscapeInputMethodPlugin.xcodeproj: Removed.
* WebKitExamplePlugins/NetscapeInputMethodPlugin/NetscapeInputMethodPlugin.xcodeproj/project.pbxproj: Removed.
* WebKitExamplePlugins/NetscapeInputMethodPlugin/main.m: Removed.
* WebKitExamplePlugins/NetscapeInputMethodPlugin/test.html: Removed.
* cmake/WebKitPackaging.cmake:

BugsSite:

* PrettyPatch/PrettyPatch.rb:

Tools:

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74332 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-19  Benjamin Kalman  <kalman@chromium.org>

        Reviewed by Antonio Gomes.

        [chromium] editing/selection/extend-selection-home-end.html should have one expected result
        https://bugs.webkit.org/show_bug.cgi?id=49873

        Change 100px to 5em in editing/selection/extend-selection-home-end.html so that chromium
        results are the same as the non-chromium results.

        * editing/selection/resources/extend-selection.js: Change 100px to 5em.
        * platform/chromium-win/editing/selection/extend-selection-home-end-expected.txt: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74331 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Improve the fix for <rdar://problem/8769525>.

Reviewed by Sam Weinig.

* UIProcess/API/mac/WKView.mm:
(-[WKView performKeyEquivalent:]): Do not retain the view. Doing so is necessary in
-[WebHTMLView performKeyEquivalent:], which this method was based upon, but not here, because
there is no way for the web process to cause the WKView to be released while it is handling
the event.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74330 268f45cc-cd09-0410-ab3c-d52691b4dbfc

<rdar://problem/8769525> WKView leak when performing key equivalents
https://bugs.webkit.org/show_bug.cgi?id=51315

Reviewed by Sam Weinig.

* UIProcess/API/mac/WKView.mm:
(-[WKView performKeyEquivalent:]): Use a RetainPtr to ensure that the view is released in the
early return code path.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74329 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-19  Helder Correia  <helder@sencha.com>

        Reviewed by Ariya Hidayat.

        [Qt] Blur distance should not be affected by transformations
        https://bugs.webkit.org/show_bug.cgi?id=51161

        From the spec at http://dev.w3.org/html5/2dcontext/#dom-context-2d-shadowblur:
        "The shadowBlur attribute specifies the level of the blurring effect.
        (The units do not map to coordinate space units, and are not affected
        by the current transformation matrix.)"

        Add test to ensure that the blur radius is not affected by the CTM.

        * fast/canvas/canvas-scale-shadowBlur-expected.txt: Added.
        * fast/canvas/canvas-scale-shadowBlur.html: Added.
        * fast/canvas/canvas-scale-strokePath-shadow-expected.txt: Increase
          shadowBlur and test tolerance to ease testing and support the
          different blurring algorithms.
        * fast/canvas/script-tests/canvas-scale-shadowBlur.js: Added.
        * fast/canvas/script-tests/canvas-scale-strokePath-shadow.js:
2010-12-19  Helder Correia  <helder@sencha.com>

        Reviewed by Ariya Hidayat.

        [Qt] Blur distance should not be affected by transformations
        https://bugs.webkit.org/show_bug.cgi?id=51161

        From the spec at http://dev.w3.org/html5/2dcontext/#dom-context-2d-shadowblur:
        "The shadowBlur attribute specifies the level of the blurring effect.
        (The units do not map to coordinate space units, and are not affected
        by the current transformation matrix.)"

        Test: fast/canvas/canvas-scale-shadowBlur.html

        * platform/graphics/ContextShadow.cpp:
        (WebCore::ContextShadow::adjustBlurDistance):
        * platform/graphics/ContextShadow.h:
        * platform/graphics/cairo/ContextShadowCairo.cpp:
        (WebCore::ContextShadow::beginShadowLayer):
        * platform/graphics/qt/ContextShadowQt.cpp:
        (WebCore::ContextShadow::beginShadowLayer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74328 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-19  Benjamin Poulain  <benjamin.poulain@nokia.com>

        Reviewed by Kenneth Rohde Christiansen.

        [Qt] [WK2] QWKPagePrivate::paint() does not need to save the graphics context
        https://bugs.webkit.org/show_bug.cgi?id=51300

        Remove the unnecessary save() and restore() on the QPainter.

        * UIProcess/API/qt/qwkpage.cpp:
        (QWKPagePrivate::paint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74327 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Text emphasis marks are not repainted correctly in flipped writing modes
https://bugs.webkit.org/show_bug.cgi?id=51307

Reviewed by Cameron Zwarich.

WebCore:

Tests: fast/repaint/text-emphasis-h.html
       fast/repaint/text-emphasis-v.html

* rendering/InlineFlowBox.cpp:
(WebCore::InlineFlowBox::addTextBoxVisualOverflow): Added flippedness check.

LayoutTests:

* fast/repaint/text-emphasis-h-expected.checksum: Added.
* fast/repaint/text-emphasis-h-expected.png: Added.
* fast/repaint/text-emphasis-h-expected.txt: Added.
* fast/repaint/text-emphasis-h.html: Added.
* fast/repaint/text-emphasis-v-expected.checksum: Added.
* fast/repaint/text-emphasis-v-expected.png: Added.
* fast/repaint/text-emphasis-v-expected.txt: Added.
* fast/repaint/text-emphasis-v.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74326 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-19  Rob Buis  <rwlbuis@gmail.com>

        Reviewed by Andreas Kling.

        Remove m_drawsContents from RenderSVGContainer.h
        https://bugs.webkit.org/show_bug.cgi?id=51305

        Remove m_drawsContents, its getter and setter and the call
        sites, since it has no effect anymore.

        * WebCore.order:
        * rendering/RenderSVGContainer.cpp:
        (WebCore::RenderSVGContainer::RenderSVGContainer):
        (WebCore::RenderSVGContainer::paint):
        * rendering/RenderSVGContainer.h:
        * rendering/RenderSVGResourceContainer.h:
        * rendering/SVGImageBufferTools.cpp:
        (WebCore::SVGImageBufferTools::renderSubtreeToImageBuffer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74325 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-19  Kenneth Rohde Christiansen  <kenneth@webkit.org>

        Reviewed by Andreas Kling.

        [Qt] [WK2] Client are expected to render their own background for WebPages
        https://bugs.webkit.org/show_bug.cgi?id=51296

        Remove the rendering of the checkerboard from the tiled DrawingAreaProxy

        * UIProcess/TiledDrawingAreaProxy.cpp:
        (WebKit::TiledDrawingAreaProxy::paint):
        * UIProcess/TiledDrawingAreaTile.h:
        * UIProcess/qt/TiledDrawingAreaTileQt.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74324 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-19  Benjamin Poulain  <benjamin.poulain@nokia.com>

        Reviewed by Kenneth Rohde Christiansen.

        WebKit does not build with worker but without the javascript debugger
        https://bugs.webkit.org/show_bug.cgi?id=51293

        Add missing guard for calls depending on the debugger.

        * inspector/InspectorController.cpp:
        (WebCore::InspectorController::populateScriptObjects):
        (WebCore::InspectorController::postWorkerNotificationToFrontend):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74323 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-19  Benjamin Poulain  <benjamin.poulain@nokia.com>

        Reviewed by Kenneth Rohde Christiansen.

        [Qt] [WK2] Client are expected to render their own background for WebPages
        https://bugs.webkit.org/show_bug.cgi?id=51296

        Do not render a white background before rendering the page with DrawingAreaProxy.

        * UIProcess/API/qt/qwkpage.cpp:
        (QWKPagePrivate::paint):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74322 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-18  Tom Sepez  <tsepez@chromium.org>

        Reviewed by David Levin.

        https://bugs.webkit.org/show_bug.cgi?id=51055
        Fix a null de-reference when an XML file contains a malformed entity
        of the form "&:;".

        * fast/parser/resources/xml-colon-entity.xml: Added.
        * fast/parser/xml-colon-entity-expected.txt: Added.
        * fast/parser/xml-colon-entity.html: Added.
2010-12-18  Tom Sepez  <tsepez@chromium.org>

        Reviewed by David Levin.

        https://bugs.webkit.org/show_bug.cgi?id=51055
        Fix a null de-reference when an XML file contains a malformed entity
        of the form "&:;".

        Test: fast/parser/xml-colon-entity.html

        * html/parser/HTMLEntitySearch.cpp:
        (WebCore::HTMLEntitySearch::advance):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74321 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-18  Noel Gordon  <noel.gordon@gmail.com>

        Reviewed by David Levin.

        [chromium] Simplify the PNG encoder.
        https://bugs.webkit.org/show_bug.cgi?id=50905

        Remove PNGEncoderState class - it was a proxy for the encoder output, and
        that is a Vector<>, so just write into it directly using append() to make
        use of Vector<>'s resize() smarts. Simplify the preMultipliedBGRAtoRGBA()
        and encode() routines, replace the OwnArrayPtr<> with a Vector<>.

        canvas.toDataURL("image/png") is covered by existing tests.

        * platform/image-encoders/skia/PNGImageEncoder.cpp:
        (WebCore::writeOutput): Use a Vector<>, append() the encoded data to it.
        (WebCore::preMultipliedBGRAtoRGBA): simplified.
        (WebCore::PNGImageEncoder::encode): ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74320 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Text emphasis marks are not always included in the selection highlight
https://bugs.webkit.org/show_bug.cgi?id=51299

Reviewed by Cameron Zwarich.

WebCore:

Affects pixel results of fast/text/emphasis.html

Account for annotations in selectionTop() and selectionBottom().

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::selectionTop):
(WebCore::RootInlineBox::selectionBottom):

LayoutTests:

* platform/mac/fast/text/emphasis-expected.checksum:
* platform/mac/fast/text/emphasis-expected.png:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74319 268f45cc-cd09-0410-ab3c-d52691b4dbfc

In flipped writing modes, opposite text emphasis marks from adjacent lines overlap
https://bugs.webkit.org/show_bug.cgi?id=51298

Reviewed by Cameron Zwarich.

WebCore:

Test: fast/text/emphasis-overlap.html

* rendering/RootInlineBox.cpp:
(WebCore::RootInlineBox::beforeAnnotationsAdjustment): Changed + to -.

LayoutTests:

* fast/text/emphasis-overlap-expected.checksum: Added.
* fast/text/emphasis-overlap-expected.png: Added.
* fast/text/emphasis-overlap-expected.txt: Added.
* fast/text/emphasis-overlap.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74318 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-18  Helder Correia  <helder@sencha.com> and Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Ariya Hidayat.

        [Qt] Canvas shadow offset should not be affected by any transformation
        https://bugs.webkit.org/show_bug.cgi?id=50422

        On a canvas context, shadows are currently affected by all
        transformations except scaling. According to the spec:
        http://www.whatwg.org/specs/web-apps/current-work/multipage/the-canvas-element.html#shadows

        "The shadowOffsetX and shadowOffsetY attributes specify the distance
        that the shadow will be offset in the positive horizontal and positive
        vertical distance respectively. Their values are in coordinate space
        units. They are not affected by the current transformation matrix."

        NOTE: this applies only to canvas, not to box shadows.

        Add new test to ensure that shadows are correctly transformed keeping
        the relative offset to the shape.

        * fast/canvas/canvas-scale-strokePath-shadow-expected.txt: See strokePath below.
        * fast/canvas/canvas-transforms-fillRect-shadow-expected.txt: Added.
        * fast/canvas/canvas-transforms-fillRect-shadow.html: Added.
        * fast/canvas/script-tests/canvas-scale-fillPath-shadow.js:
        * fast/canvas/script-tests/canvas-scale-fillRect-shadow.js:
        * fast/canvas/script-tests/canvas-scale-strokePath-shadow.js: Now using
          a lineWidth > 1 to make it easier to test and more fair among all
          ports, since there can be different transformation smoothness or
          aliasing settings.
        * fast/canvas/script-tests/canvas-transforms-fillRect-shadow.js: Added.
        * platform/gtk/Skipped: GTK port needs to implement this feature.
        * platform/gtk/fast/canvas/canvas-scale-fillPath-shadow-expected.txt: Removed.
        * platform/gtk/fast/canvas/canvas-scale-fillRect-shadow-expected.txt: Removed.
        * platform/gtk/fast/canvas/canvas-scale-strokePath-shadow-expected.txt: Removed.
2010-12-18  Helder Correia  <helder@sencha.com> and Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Ariya Hidayat.

        [Qt] Canvas shadow offset should not be affected by any transformation
        https://bugs.webkit.org/show_bug.cgi?id=50422

        On a canvas context, shadows are currently affected by all
        transformations except scaling. According to the spec:
        http://www.whatwg.org/specs/web-apps/current-work/multipage/the-canvas-element.html#shadows

        "The shadowOffsetX and shadowOffsetY attributes specify the distance
        that the shadow will be offset in the positive horizontal and positive
        vertical distance respectively. Their values are in coordinate space
        units. They are not affected by the current transformation matrix."

        NOTE: this applies only to canvas, not to box shadows.

        Test: fast/canvas/canvas-transforms-fillRect-shadow.html

        * platform/graphics/ContextShadow.cpp:
        (WebCore::ContextShadow::ContextShadow):
        (WebCore::ContextShadow::calculateLayerBoundingRect):
        * platform/graphics/ContextShadow.h:
        (WebCore::ContextShadow::setShadowsIgnoreTransforms):
        (WebCore::ContextShadow::shadowsIgnoreTransforms):
        (WebCore::ContextShadow::offset):
        * platform/graphics/qt/ContextShadowQt.cpp:
        (WebCore::ContextShadow::beginShadowLayer):
        (WebCore::ContextShadow::endShadowLayer):
        * platform/graphics/qt/GraphicsContextQt.cpp:
        (WebCore::mustUseContextShadow):
        (WebCore::GraphicsContext::fillPath):
        (WebCore::GraphicsContext::strokePath):
        (WebCore::GraphicsContext::fillRect):
        (WebCore::GraphicsContext::fillRoundedRect):
        (WebCore::GraphicsContext::setPlatformShadow):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74317 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-18  Ariya Hidayat  <ariya@sencha.com>

        Reviewed by Andreas Kling.

        [Qt] TransparencyLayer should inherit its container render hints
        https://bugs.webkit.org/show_bug.cgi?id=51283

        * platform/graphics/qt/TransparencyLayer.h:
        (WebCore::TransparencyLayer::TransparencyLayer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74316 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-18  Tony Gentilcore  <tonyg@chromium.org>

        Reviewed by Laszlo Gombos.

        [Web Timing] Rename domContentLoaded{Start,End}->domContentLoadedEvent{Start,End}
        https://bugs.webkit.org/show_bug.cgi?id=50943

        * fast/dom/Window/window-properties-performance-expected.txt:
        * fast/dom/script-tests/webtiming.js:
        (checkTimingBeforeLoad):
        (checkTimingWhileDeferred):
        (checkWebTimingOnDOMContentLoaded):
        (checkWebTimingWhileAsync):
        (checkWebTimingOnLoad):
        (checkWebTimingAfterLoad):
        * fast/dom/webtiming-document-open-expected.txt:
        * fast/dom/webtiming-expected.txt:
        * fast/dom/webtiming-navigate-within-document-expected.txt:
2010-12-18  Tony Gentilcore  <tonyg@chromium.org>

        Reviewed by Laszlo Gombos.

        [Web Timing] Rename domContentLoaded{Start,End}->domContentLoadedEvent{Start,End}
        https://bugs.webkit.org/show_bug.cgi?id=50943

        See: http://test.w3.org/webperf/specs/NavigationTiming/#nt-dom-content-event-start

        * dom/Document.cpp:
        (WebCore::Document::finishedParsing):
        * dom/DocumentTiming.h:
        (WebCore::DocumentTiming::DocumentTiming):
        * page/PerformanceTiming.cpp:
        (WebCore::PerformanceTiming::domContentLoadedEventStart):
        (WebCore::PerformanceTiming::domContentLoadedEventEnd):
        * page/PerformanceTiming.h:
        * page/PerformanceTiming.idl:
2010-12-18  Tony Gentilcore  <tonyg@chromium.org>

        Reviewed by Laszlo Gombos.

        [Web Timing] Rename domContentLoaded{Start,End}->domContentLoadedEvent{Start,End}
        https://bugs.webkit.org/show_bug.cgi?id=50943

        Exposes all dom* times to the chromium port. I'm particularly interested in
        domContentLoadedEventEnd as it compares to the FinishDoc metric.

        * public/WebPerformance.h:
        * src/WebPerformance.cpp:
        (WebKit::WebPerformance::domLoading):
        (WebKit::WebPerformance::domInteractive):
        (WebKit::WebPerformance::domContentLoadedEventStart):
        (WebKit::WebPerformance::domContentLoadedEventEnd):
        (WebKit::WebPerformance::domComplete):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74315 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Tony Gentilcore  <tonyg@chromium.org>

        Unreviewed.

        [chromium] Mark fast/text/emphasis* failing
        https://bugs.webkit.org/show_bug.cgi?id=51288

        They have square characters that appear broken.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74314 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Tony Gentilcore  <tonyg@chromium.org>

        Unreviewed fix.

        Fix JS syntax to avoid V8 error message
        https://bugs.webkit.org/show_bug.cgi?id=51287

        * transitions/interrupted-accelerated-transition.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74313 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Justin Schuh  <jschuh@chromium.org>

        Reviewed by Andreas Kling.

        Style fix for r73927
        https://bugs.webkit.org/show_bug.cgi?id=51194

        No functionality change. No new tests needed.

        * html/HTMLCanvasElement.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74312 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Carlos Garcia Campos  <cgarcia@igalia.com>

        Reviewed by Martin Robinson.

        [GTK] Buttons rendering was broken in r74129
        https://bugs.webkit.org/show_bug.cgi?id=51173

        * platform/gtk/RenderThemeGtk.cpp:
        (WebCore::RenderThemeGtk::paintRenderObject):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74311 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Zoltan Herczeg  <zherczeg@webkit.org>

        Reviewed by Dirk Schulze.

        Improve 'arithmetic' operator on feComposite
        https://bugs.webkit.org/show_bug.cgi?id=51105

        Fixing style errors and bug description for r74177.

        * platform/graphics/filters/FEComposite.cpp:
        (WebCore::arithmetic):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74310 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Michael Saboff  <msaboff@apple.com>

        Reviewed by Oliver Hunt.

        RegExp Jit'ed expression crashes clicking link on yelp.com
        https://bugs.webkit.org/show_bug.cgi?id=51284

        When transitioning between an non-repeating beginning of line
        anchored expression and the remaining refactored repeating
        expression, we should not clear any residual datalabel in
        state's m_backtrack.  It will be resolved and cleared in subsequent
        code when linkAlternativeBacktracks() is called for the repeating
        alternative(s).

        * yarr/RegexJIT.cpp:
        (JSC::Yarr::RegexGenerator::BacktrackDestination::clear):
        (JSC::Yarr::RegexGenerator::TermGenerationState::clearBacktrack):
2010-12-17  Michael Saboff  <msaboff@apple.com>

        Reviewed by Oliver Hunt.

        RegExp Jit'ed expression crashes clicking link on yelp.com
        https://bugs.webkit.org/show_bug.cgi?id=51284

        New tests to validate changes made in regex JIT.

        * fast/regex/parentheses-expected.txt:
        * fast/regex/script-tests/parentheses.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74309 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Updated for the renaming of WebKitTools to Tools

Rubber-stamped by Mark Rowe.

JavaScriptCore:

* JavaScriptCore.vcproj/JavaScriptCore/build-generated-files.sh:

WebCore:

* WebCore.vcproj/build-generated-files.sh:
* WebCore.vcproj/migrate-scripts.sh:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74308 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Tony Gentilcore  <tonyg@chromium.org>

        Unreviewed, build fix.

        Add WebKitTools -> Tools rename that got missed.

        * WebKit.gyp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74307 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Darin Adler  <darin@apple.com>

        Reviewed by Dan Bernstein.

        Window title for image documents shows corrupted characters instead of multiplication sign when using WebKit2
        https://bugs.webkit.org/show_bug.cgi?id=43505

        * WebProcess/WebCoreSupport/WebPlatformStrategies.cpp: Changed UI_STRING macros to return
        WebCore::String objects instead of C strings.
        (WebKit::formatLocalizedString): Added. Uses the Core Foundation string formatting function for
        reasons explained in the comment.
        (WebKit::WebPlatformStrategies::multipleFileUploadText): Use formatLocalizedString instead of
        String::format.
        (WebKit::WebPlatformStrategies::imageTitle): Ditto.
        (WebKit::WebPlatformStrategies::localizedMediaTimeDescription): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74306 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Add expected results for Windows XP to get the bot green for
http://trac.webkit.org/changeset/74281.
Unreviewed.

* platform/win-xp/fast/text/emphasis-expected.txt: Added.
* platform/win-xp/fast/text/emphasis-vertical-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74305 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Dirk Pranke  <dpranke@chromium.org>

        Unreviewed, build fix.

        Add WebKitTools -> Tools rename that got missed.

        * WebKit.gyp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74304 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Anders Carlsson  <andersca@apple.com>

        Reviewed by Sam Weinig.

        Resizing a WKView while loading a page can leave the page at a size that doesn't match the window
        https://bugs.webkit.org/show_bug.cgi?id=51282
        <rdar://problem/8133142>

        Fix a race condition in waitForMessage. If we time out on the wait condition, we would keep the
        m_waitForMessageMutex mutex unlocked for a brief period of time before taking the lock again and
        then removing the messageID/destinationID pair from the hash map. Under some circumstances, the
        connection queue would update the hash map right before we removed it, leading to a lost message.

        * Platform/CoreIPC/Connection.cpp:
        (CoreIPC::Connection::waitForMessage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74303 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed Qt buildfix after r74301.

Rename WebKitTools to Tools
https://bugs.webkit.org/show_bug.cgi?id=49861

* MiniBrowser/DerivedSources.pro:
* MiniBrowser/qt/MiniBrowser.pro:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74302 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Rename WebKitTools to Tools
https://bugs.webkit.org/show_bug.cgi?id=49861

Reviewed by Simon Fraser.

.:

* DerivedSources.pro:
* GNUmakefile.am:
* MakeFile:
* MakeFile.shared:
* Tools: Renamed from WebKitTools.
* WebKit.pro:
* cmake/WebKitPackaging.cmake:
* wscript:

BugsSite:

* PrettyPatch/PrettyPatch.rb:
* committers-autocomplete.js:

Tools:

* BuildSlaveSupport/build.webkit.org-config/master.cfg:
* CodeCoverage/README:
* DumpRenderTree/qt/DumpRenderTreeQt.cpp:
(WebCore::DumpRenderTree::initializeFonts):
* EWSTools/start-commit-queue.sh:
* EWSTools/start-queue.sh:
* GNUmakefile.am:
* MIDLWrapper/MIDLWrapper.cpp:
(wmain):
* MiniBrowser/qt/MiniBrowser.pro:
* Scripts/build-api-tests:
* Scripts/build-dumprendertree:
* Scripts/build-webkit:
* Scripts/build-webkittestrunner:
* Scripts/generate-coverage-data:
* Scripts/old-run-webkit-tests:
* Scripts/run-api-tests:
* Scripts/run-iexploder-tests:
* Scripts/run-javascriptcore-tests:
* Scripts/run-mangleme-tests:
* Scripts/run-sunspider:
* Scripts/run-webkit-websocketserver:
* Scripts/sunspider-compare-results:
* Scripts/test-webkitperl:
* Scripts/test-webkitpy:
* Scripts/update-iexploder-cssproperties:
* Scripts/update-webkit:
* Scripts/update-webkit-localizable-strings:
* Scripts/webkitdirs.pm:
* Scripts/webkitpy/common/checkout/scm.py:
* Scripts/webkitpy/common/checkout/scm_unittest.py:
* Scripts/webkitpy/common/config/committervalidator.py:
* Scripts/webkitpy/common/config/committervalidator_unittest.py:
* Scripts/webkitpy/common/config/ports.py:
* Scripts/webkitpy/common/config/ports_unittest.py:
* Scripts/webkitpy/common/system/logutils_unittest.py:
* Scripts/webkitpy/common/system/ospath_unittest.py:
* Scripts/webkitpy/layout_tests/port/chromium.py:
* Scripts/webkitpy/layout_tests/port/config.py:
* Scripts/webkitpy/layout_tests/port/config_unittest.py:
* Scripts/webkitpy/layout_tests/port/http_server.py:
* Scripts/webkitpy/layout_tests/port/test.py:
* Scripts/webkitpy/style/checker.py:
* Scripts/webkitpy/style/checker_unittest.py:
* Scripts/webkitpy/tool/bot/feeders_unittest.py:
* Scripts/webkitpy/tool/commands/queues_unittest.py:
* Scripts/webkitpy/tool/steps/steps_unittest.py:
* WebKitTestRunner/DerivedSources.pro:
* WebKitTestRunner/InjectedBundle/qt/ActivateFontsQt.cpp:
(WTR::activateFonts):
* WebKitTestRunner/InjectedBundle/qt/InjectedBundle.pro:
* wx/build/build_utils.py:

WebKitSite:

* asking_questions.html:
* building/tools.html:
* coding/scripts.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74301 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Clean up ChunkedUpdateDrawingAreaProxy
https://bugs.webkit.org/show_bug.cgi?id=51277

Reviewed by Simon Fraser.

Get rid of m_lastSetViewSize since it was always equal to m_size.

* UIProcess/ChunkedUpdateDrawingAreaProxy.cpp:
(WebKit::ChunkedUpdateDrawingAreaProxy::sizeDidChange):
Call sendSetSize().

(WebKit::ChunkedUpdateDrawingAreaProxy::didSetSize):
Call sendSetSize if the size differs.

(WebKit::ChunkedUpdateDrawingAreaProxy::sendSetSize):
Move code from sizeDidChange over here.

* UIProcess/ChunkedUpdateDrawingAreaProxy.h:
* UIProcess/DrawingAreaProxy.cpp:
(WebKit::DrawingAreaProxy::setSize):
Add back the early return if the sizes are equal.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74300 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Ariya Hidayat  <ariya@sencha.com>

        Reviewed by Oliver Hunt.

        [JSC] parseAssignmentExpression should use TreeBuilder::CreatesAST
        https://bugs.webkit.org/show_bug.cgi?id=51268

        * parser/JSParser.cpp:
        (JSC::JSParser::parseAssignmentExpression):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74299 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Unreviewed build fix.

* WebProcess/WebPage/win/WebPageWin.cpp:
(WebKit::WebPage::canHandleRequest): Provide alternative for
non-CFNetwork implementation.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74298 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2010-12-17  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Oliver Hunt.

        Removed RChunk from PageAllocation/PageReservation, since it's now unused.
        https://bugs.webkit.org/show_bug.cgi?id=51276

        * wtf/PageAllocation.h:
        (WTF::PageAllocation::PageAllocation):
        * wtf/PageReservation.h:
        (WTF::PageReservation::PageReservation):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74297 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Anders Carlsson.

        Implement WKView::setInitialFocus().
        https://bugs.webkit.org/show_bug.cgi?id=51274

        * UIProcess/API/C/win/WKView.cpp:
        (WKViewSetInitialFocus):
        * UIProcess/API/C/win/WKView.h:
        * UIProcess/WebPageProxy.cpp:
        (WebKit::WebPageProxy::setInitialFocus): Send message to the Web Process to set
        initial focus.
        * UIProcess/WebPageProxy.h:
        * UIProcess/win/WebView.cpp:
        (WebKit::WebView::setInitialFocus):
        * UIProcess/win/WebView.h:
        * WebProcess/WebPage/WebPage.cpp:
        (WebKit::WebPage::setInitialFocus): Set the focused node to null initially.  Then
        ask FocusController to set the initial focus based on the focus direction.
        * WebProcess/WebPage/WebPage.h:
        * WebProcess/WebPage/WebPage.messages.in: Add a new message for SetInitialFocus.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@74296 268f45cc-cd09-0410-ab3c-d52691b4dbfc