Context Navigation

[Qt] Fix memory leak by deleting instance of ThreadPrivate
in function waitForThreadCompletion(), synchronously, or in
detachThread(), asynchronously.
https://bugs.webkit.org/show_bug.cgi?id=32614

wtf/qt/ThreadingQt.cpp: (WTF::waitForThreadCompletion): (WTF::detachThread):

6:12 AM Changeset in webkit [52549] by zoltan@webkit.org

2 edits in trunk/LayoutTests

Rubber-stamped by Simon Hausmann.

Patch by Andras Becsi <abecsi@inf.u-szeged.hu> on 2009-12-24
[Qt] Skip failing test until a solution for layoutTestController.notifyDone() is found.
https://bugs.webkit.org/show_bug.cgi?id=31626

platform/qt/Skipped: inspector/console-clear.html skipped

5:46 AM Changeset in webkit [52548] by jberlin@webkit.org

7 edits
7 adds in trunk

Fix for WebKit bug 29968 - Selecting text with text-overflow ellipsis
should not show cut off text
https://bugs.webkit.org/show_bug.cgi?id=29968

Reviewed by Dan Bernstein.

Also fixes the issue on both Mac OS X and Windows where the highlight
would extend only partway into the ellipsis based on the size of the
characters being truncated.

Test: editing/selection/select-text-overflow-ellipsis.html

WebCore:

rendering/EllipsisBox.cpp: (WebCore::EllipsisBox::paint): When the ellipsis have been selected, paints them with the color for selected text. (WebCore::EllipsisBox::selectionRect): Returns the selection rect for the ellipsis text. (WebCore::EllipsisBox::paintSelection): Paints the selection highlight around the ellipsis text.

rendering/EllipsisBox.h: (WebCore::EllipsisBox::EllipsisBox): Default the selection state of the EllipsisBox to SelectionNone. (WebCore::EllipsisBox::setSelectionState): Allow the selection state to be explicitly set by the InlineTextBox that it is associated with, since the selection state depends on the position of the selection within the truncated text. (WebCore::EllipsisBox::selectionState): Return the cached selection state.

rendering/InlineTextBox.cpp: (WebCore::InlineTextBox::selectionState): If the text is truncated and the selection extends into the truncation where there are ellipsis, set the selection state on the EllipsisBox. (WebCore::paintTextWithShadows): Allow for an upper bound to be specified on the maximum length of the text to be painted. (WebCore::InlineTextBox::paint): Make sure the text that is drawn gets appropriately truncated. This was not an issue before on Mac OS X because it wasn't necessary to draw the selected text separately (the selected text had the same color as the non-selected text). (WebCore::InlineTextBox::paintSelection): If the text is being truncated, let whatever is being painted in the space after the truncated text paint its own highlight.

rendering/InlineTextBox.h: (WebCore::InlineTextBox::truncation): Allow RenderText to access the truncation status.

rendering/RenderText.cpp: (WebCore::RenderText::selectionRectForRepaint): Make sure that any ellipsis box that falls within the selection also gets repainted.

Places the expected results under the mac and win platforms separately
because they are pixel test results and also the size of the
RenderBlocks are showing up as slightly different between mac and win.

LayoutTests:

editing/selection/select-text-overflow-ellipsis.html: Added.
platform/mac/editing/selection/select-text-overflow-ellipsis-expected.checksum: Added.
platform/mac/editing/selection/select-text-overflow-ellipsis-expected.png: Added.
platform/mac/editing/selection/select-text-overflow-ellipsis-expected.txt: Added.
platform/win/editing/selection/select-text-overflow-ellipsis-expected.checksum: Added.
platform/win/editing/selection/select-text-overflow-ellipsis-expected.png: Added.
platform/win/editing/selection/select-text-overflow-ellipsis-expected.txt: Added.

4:33 AM Changeset in webkit [52547] by pfeldman@chromium.org

24 edits
1 add
2 deletes in trunk

2009-12-24 Pavel Feldman <pfeldman@chromium.org>

Reviewed by Timothy Hatcher.

Web Inspector: add Settings.js class that would manage
frontend settings.

This change:

Splits preferences and settings
Pushes settings to the frontend on load
Stores settings on each change
Stores showInheritedStyles as setting.

https://bugs.webkit.org/show_bug.cgi?id=32832

WebCore.gypi:
WebCore.vcproj/WebCore.vcproj:
inspector/InspectorBackend.cpp: (WebCore::InspectorBackend::saveFrontendSettings):
inspector/InspectorBackend.h:
inspector/InspectorBackend.idl:
inspector/InspectorController.cpp: (WebCore::InspectorController::populateScriptObjects):
inspector/InspectorController.h:
inspector/InspectorFrontend.cpp: (WebCore::InspectorFrontend::populateFrontendSettings):
inspector/InspectorFrontend.h:
inspector/InspectorFrontendHost.cpp:
inspector/InspectorFrontendHost.h:
inspector/InspectorFrontendHost.idl:
inspector/front-end/EventListenersSidebarPane.js: (WebInspector.EventListenersSidebarPane.prototype._settingsLoaded): (WebInspector.EventListenersSidebarPane.prototype): ():
inspector/front-end/InspectorBackendStub.js: (.WebInspector.InspectorBackendStub): (.WebInspector.InspectorBackendStub.prototype.saveFrontendSettings): (.WebInspector.InspectorBackendStub.prototype.releaseWrapperObjectGroup):
inspector/front-end/InspectorFrontendHostStub.js: (.WebInspector.InspectorFrontendHostStub): (.WebInspector.InspectorFrontendHostStub.prototype.hiddenPanels):
inspector/front-end/ResourcesPanel.js: (WebInspector.ResourcesPanel.prototype._createStatusbarButtons): (WebInspector.ResourcesPanel.prototype._settingsLoaded): (WebInspector.ResourcesPanel.prototype._toggleLargerResources): (WebInspector.ResourcesPanel.prototype._setLargerResources):
inspector/front-end/ScriptsPanel.js: (WebInspector.ScriptsPanel.prototype._showScriptOrResource): (WebInspector.ScriptsPanel.prototype._addScriptToFilesMenu):
inspector/front-end/Settings.js: Added. (WebInspector.populateFrontendSettings): (WebInspector.Settings): (WebInspector.Settings.prototype._load): (WebInspector.Settings.prototype._installSetting): (WebInspector.Settings.prototype._get): (WebInspector.Settings.prototype._set):
inspector/front-end/StylesSidebarPane.js: (WebInspector.StylesSidebarPane.prototype._settingsLoaded): (WebInspector.StylesSidebarPane.prototype.update): (WebInspector.StylesSidebarPane.prototype._changeSetting): (WebInspector.StylesSidebarPane.prototype._changeColorFormat): (WebInspector.StylePropertiesSection.showInheritedToggleFunction): (WebInspector.StylePropertiesSection):
inspector/front-end/WatchExpressionsSidebarPane.js: (WebInspector.WatchExpressionsSidebarPane): (WebInspector.WatchExpressionsSidebarPane.prototype._settingsLoaded): (WebInspector.WatchExpressionsSidebarPane.prototype.refreshExpressions): (WebInspector.WatchExpressionsSection): (WebInspector.WatchExpressionsSection.prototype.saveExpressions):
inspector/front-end/WebKit.qrc:
inspector/front-end/inspector.html:
inspector/front-end/inspector.js: (WebInspector.loaded):

3:12 AM Changeset in webkit [52546] by pfeldman@chromium.org

1 edit
2 adds in trunk/LayoutTests

2009-12-24 Pavel Feldman <pfeldman@chromium.org>

Not reviewed: follow up to r52545, adding missing tests.

inspector/console-clear-expected.txt: Added.
inspector/console-clear.html: Added.

3:05 AM Changeset in webkit [52545] by pfeldman@chromium.org

15 edits in trunk

2009-12-23 Pavel Feldman <pfeldman@chromium.org>

Reviewed by Timothy Hatcher.

Web Inspector: [REGRESSION] console's clear command does not work.

https://bugs.webkit.org/show_bug.cgi?id=32897

Test: inspector/console-clear.html

WebCore:

inspector/InjectedScriptHost.cpp: (WebCore::InjectedScriptHost::clearConsoleMessages):
inspector/InjectedScriptHost.h:
inspector/InjectedScriptHost.idl:
inspector/InspectorBackend.cpp:
inspector/InspectorBackend.h:
inspector/InspectorBackend.idl:
inspector/InspectorController.cpp: (WebCore::InspectorController::clearConsoleMessages): (WebCore::InspectorController::didCommitLoad):
inspector/InspectorController.h:
inspector/front-end/ConsoleView.js: (WebInspector.ConsoleView): (WebInspector.ConsoleView.prototype.requestClearConsoleMessages): (WebInspector.ConsoleView.prototype.clearConsoleMessages): (WebInspector.ConsoleView.prototype._clearButtonClicked):
inspector/front-end/InjectedScript.js: (InjectedScript.clearConsoleMessages): (InjectedScript._ensureCommandLineAPIInstalled):
inspector/front-end/InjectedScriptAccess.js:
inspector/front-end/inspector.js: (WebInspector.clearConsoleMessages): (WebInspector.reset):

LayoutTests:

inspector/console-clear-expected.txt: Added.
inspector/console-clear.html: Added.
inspector/console-tests.js: (dumpConsoleMessages.callback): (dumpConsoleMessages):

2:38 AM Changeset in webkit [52544] by pfeldman@chromium.org

10 edits in trunk

2009-12-23 Pavel Feldman <pfeldman@chromium.org>

Reviewed by Timothy Hatcher.

Web Inspector: Console dumps strings in escaped form.

https://bugs.webkit.org/show_bug.cgi?id=32488

inspector/front-end/ConsoleView.js: (WebInspector.ConsoleMessage.prototype._formatIndividualValue):
inspector/console-format-expected.txt:

Dec 23, 2009:

11:18 PM Changeset in webkit [52543] by eric@webkit.org

4 edits in trunk/WebKitTools

2009-12-23 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

bugzilla-tool should accept global options anywhere
https://bugs.webkit.org/show_bug.cgi?id=26912

Scripts/bugzilla-tool:
- Use the new global_options class property.
- Add a handle_global_options callback to avoid needing callbacks for each global option.
Scripts/modules/multicommandtool.py:
- Make the code use one combined option parser. This allows us to accept global options anywhere and individual command options before commands.
- Add a handle_global_options callback to avoid needing callbacks for each global option.
- Make the Command hold the option parser, but allow the tool to override it.
- The default option parser is used for help printing and when Commands are run stand alone, but are otherwise not used.
- Add Command.main to codify the idea that Commands should support being run stand-alone.
- Change _split_args to _split_command_name_from_args now that args are unified.
Scripts/modules/multicommandtool_unittest.py:
- Test that "tool" and "tool help" show the same help.
- Test that args are accepted before commands

9:49 PM Changeset in webkit [52542] by ap@apple.com

3 edits
2 adds in trunk

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=32905
With Pinyin Simplified IM, a wrong character is deleted from google.com suggestion

Test: platform/mac/editing/input/selection-change-closes-typing.html

editing/Editor.cpp: (WebCore::Editor::confirmComposition): If this function wasn't allowed to change selection, it needs to close the current typing command - otherwise, its idea of selection would be used for future typing.

9:31 PM Changeset in webkit [52541] by hamaji@chromium.org

6 edits in trunk/WebKitTools

2009-12-20 Chris Jerdonek <chris.jerdonek@gmail.com>

Reviewed by David Levin.

Moved some sections of code in preparation to refactor
check-webkit-style's argument parser to avoid setting
global variables.

https://bugs.webkit.org/show_bug.cgi?id=32592

Scripts/check-webkit-style:
- Moved _USAGE string to style.py.
- Addressed FIXME by eliminating dependencies on cpp_style.py.

Scripts/modules/cpp_style.py:
- Moved default arguments and style categories to style.py.
- Moved exit_with_usage(), exit_with_categories(), and parse_arguments() to style.py.
- Removed references in _CppStyleState to the global variables now in style.py.

Scripts/modules/cpp_style_unittest.py:
- Moved parse_arguments() unit tests to style_unittest.py.

Scripts/modules/style.py:
- Added _USAGE string from check-webkit-style.
- Added default arguments and style categories from cpp_style.py.
- Added exit_with_usage(), exit_with_categories(), and parse_arguments() from cpp_sstyle.py.

Scripts/modules/style_unittest.py: Added.
- Added parse_arguments() unit tests from cpp_style_unittest.py.

Scripts/run-webkit-unittests:
- Added unit tests from style_unittest.py.

6:31 PM Changeset in webkit [52540] by eric@webkit.org

3 edits in trunk/WebKitTools

2009-12-23 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

commit-queue should mention bug 30084 when rejecting patches until it can be fixed
https://bugs.webkit.org/show_bug.cgi?id=32911

Scripts/modules/bugzilla.py: Make the message even more fancy.
Scripts/modules/bugzilla_unittest.py: Test our new fancy message.

6:23 PM Changeset in webkit [52539] by eric@webkit.org

2 edits in trunk/WebCore

2009-12-23 Yuta Kitamura <yutak@chromium.org>

Reviewed by Eric Seidel.

PlatformContextSkia::beginLayerClippedToImage() needs to clip outside of the image.

This method should limit the current painting window to the bounds of the given
image, but it did not. Due to this bug, unnecessary contents were rendered in
some (limited) situations.

[Skia] PlatformContextSkia::beginLayerClippedToImage does not clip outside of the image
https://bugs.webkit.org/show_bug.cgi?id=32817

No new tests, because this is a fix of an existing test which is currently failing in
Chromium.

platform/graphics/skia/PlatformContextSkia.cpp: (PlatformContextSkia::beginLayerClippedToImage): Add clipRect() call.

6:15 PM Changeset in webkit [52538] by eric@webkit.org

2 edits in trunk/JavaScriptCore

2009-12-23 Kwang Yul Seo <skyul@company100.net>

Reviewed by Laszlo Gombos.

Include stddef.h for ptrdiff_t
https://bugs.webkit.org/show_bug.cgi?id=32891

ptrdiff_t is typedef-ed in stddef.h.
Include stddef.h in jit/ExecutableAllocator.h.

jit/ExecutableAllocator.h:

5:45 PM Changeset in webkit [52537] by Nikolas Zimmermann

198 edits in trunk/WebCore

2009-12-23 Nikolas Zimmermann <nzimmermann@rim.com>

Rubber-stamped by Eric Seidel.

Remove all references to "GenerateConstructor", as this is the default behaviour now.

5:38 PM Changeset in webkit [52536] by dumi@chromium.org

2 edits in trunk/WebCore

A new database should be registered with the Document,
DatabaseTracker and DatabaseThread only after it was successfully
opened and its version was successfully verified.

Reviewed by Eric Seidel.

Fixes a regression introduced by
http://trac.webkit.org/changeset/52530.

LayoutTests/storage/open-database-while-transaction-in-progress.html
should pass again (or rather, it should pass again when run after
open-database-set-empty-version.html).

https://bugs.webkit.org/show_bug.cgi?id=32913

storage/Database.cpp:

(WebCore::Database::openDatabase):
(WebCore::Database::Database):
(WebCore::Database::performOpenAndVerify):

5:15 PM Changeset in webkit [52535] by levin@chromium.org

7 edits
2 deletes in trunk

2009-12-23 David Levin <levin@chromium.org>

REGRESSION (r52494): Assertion failure in Frame::caretBlinkTimerFired() (selection()->isCaret())
https://bugs.webkit.org/show_bug.cgi?id=32903

No review but ok'ed by Darin Adler. Rolling out r52494 due to above regression.

LayoutTests:

fast/forms/selection-layout-reentry-strange-case-expected.txt: Removed.
fast/forms/selection-layout-reentry-strange-case.html: Removed.
platform/mac/accessibility/frame-with-title-expected.txt:

WebCore:

editing/SelectionController.cpp: (WebCore::SelectionController::SelectionController): (WebCore::SelectionController::setSelection):
editing/SelectionController.h:
page/Frame.cpp: (WebCore::Frame::setCaretVisible): (WebCore::Frame::selectionLayoutChanged):
page/FrameView.cpp: (WebCore::FrameView::needsLayout):

4:55 PM Changeset in webkit [52534] by Nikolas Zimmermann

75 edits in trunk

2009-12-23 Nikolas Zimmermann <nzimmermann@rim.com>

Reviewed by Eric Seidel.

Reverse JS GenerateConstructor logic
https://bugs.webkit.org/show_bug.cgi?id=32910

Make 'GenerateConstructor' the default setting for all classes, defined in IDL files. The flag is now obsolete and
can be removed from all IDL files in a follow-up patch. Add new 'OmitConstructor' flag, that allows to restore the
old default behaviour: do not generate JSFoobarConstructor class.

No change in functionality, despite the fact that we're generating a lot more constructors now, as they have to
be exposed through DOMWindow.idl -- this can be done per affected class in follow-up patches. Especially the SVG
classes have to be exposed, most of them are missing JS constructors so far.

As side effect HTMLOptionsCollection is now correctly exposing its constructor, thus fixing a test in fast/dom/wrapper-classes.html

4:34 PM Changeset in webkit [52533] by bweinstein@apple.com

1 edit
1 add in trunk/LayoutTests

Rubber-stamped by Eric Seidel.

Added Windows specific results for a test that was committed in r52505.

platform/win/fast/forms/basic-textareas-quirks-expected.txt: Added.

3:22 PM Changeset in webkit [52532] by abarth@webkit.org

4 edits in trunk

2009-12-23 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

"Refused to execute a JavaScript script" error when embedding SWF with
a URL that is also a query parameter
https://bugs.webkit.org/show_bug.cgi?id=32908

Update expected results to show that we don't raise an alarm in this case.

http/tests/security/xssAuditor/object-src-inject-expected.txt:

2009-12-23 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

"Refused to execute a JavaScript script" error when embedding SWF with
a URL that is also a query parameter
https://bugs.webkit.org/show_bug.cgi?id=32908

Don't block direct injections into the object src attribute unless
there's an illegal character (like < or ") in the URL. This change
lets some very unusual vulnerabilities through the filter but removes a
false positive that we've seen several times.

page/XSSAuditor.cpp: (WebCore::XSSAuditor::canLoadObject):

2:34 PM Changeset in webkit [52531] by Laszlo Gombos

2 edits in trunk/WebKitTools

2009-12-23 Laszlo Gombos <Laszlo Gombos>

Not reviewed, adding myself to the reviewers list.

Scripts/modules/committers.py:

2:33 PM Changeset in webkit [52530] by dumi@chromium.org

2 edits in trunk/WebCore

Remove pointers to a database that failed to open from the
DatabaseTracker and Document.

Reviewed by Dmitry Titov.

https://bugs.webkit.org/show_bug.cgi?id=32885

storage/Database.cpp:

(WebCore::Database::openDatabase):

1:03 PM Changeset in webkit [52529] by eric@webkit.org

2 edits in trunk/JavaScriptCore

2009-12-23 Patrick Gansterer <paroga@paroga.com>

Reviewed by Eric Seidel.

Buildfix after r47092.

wtf/wince/MemoryManager.cpp: (WTF::tryFastMalloc): (WTF::tryFastZeroedMalloc): (WTF::tryFastCalloc): (WTF::tryFastRealloc):

12:40 PM Changeset in webkit [52528] by abarth@webkit.org

7 edits in trunk/WebKitTools

2009-12-23 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

[bzt] Create a prepare-diff command that creates a bug and a ChangeLog
https://bugs.webkit.org/show_bug.cgi?id=32895

The workflow Maciej and I discussed is as follows:

1) Write code.
2) bugzilla-tool prepare-diff
3) Edit ChangeLogs
4) bugzilla-tool post-diff

We might want to experimenting with combining 2-4 into a single
command, but that might be stressful to edit the ChangeLogs modally.

Removed submit-patch since it has the modal ChangeLog editing but none
oof the bug creating fun.

Scripts/modules/bugzilla.py:
Scripts/modules/buildsteps.py:
Scripts/modules/commands/upload.py:
Scripts/modules/commands/upload_unittest.py:

11:49 AM Changeset in webkit [52527] by mitz@apple.com

3 edits
2 adds in trunk

<rdar://problem/7487164> First line of text cannot be selected
https://bugs.webkit.org/show_bug.cgi?id=32749

Reviewed by Darin Adler.

WebCore:

Test: fast/text/remove-zero-length-run.html

rendering/RenderText.cpp:

(WebCore::RenderText::positionLineBox): Changed code that assumed that if a box was being

removed, it was the only box in the RenderText. Instead, correctly preserve the list of
text boxes.

(WebCore::RenderText::checkConsistency): Updated for earlier rename.

LayoutTests:

fast/text/remove-zero-length-run-expected.txt: Added.
fast/text/remove-zero-length-run.html: Added.

5:35 AM WebKit Team edited by loki@webkit.org

(diff)

5:32 AM Changeset in webkit [52526] by loki@webkit.org

2 edits in trunk/WebKitTools

Unreviewed; added myself to the committers list.

Scripts/modules/committers.py:

5:19 AM QtWebKitBackportingFixes edited by Laszlo Gombos

(diff)

4:52 AM Changeset in webkit [52525] by vestbo@webkit.org

8 edits
3 adds in trunk

Fix for newlines turning into missing-gyphs when using SVG fonts

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=32710

Font::drawText() has three code paths:

drawTextUsingSVGFont()
drawSimpleText()
drawComplexText()

The latter two both have logic to replace newlines and tabs, as well
as a few other control characters that are considered spaces, with
spaces, before actually drawing the text. Previously the SVG font
path did not have this kind of logic.

This change introduces two helper functions in Font to do this space
replacement, which are then used for the simple path, the SVG path,
and the Qt implementation of the complex path.

Test: svg/custom/svg-fonts-in-html-linebreaks.html

WebCore:

platform/graphics/Font.cpp:

(WebCore::Font::normalizeSpaces): Helper function

platform/graphics/Font.h:
platform/graphics/FontFastPath.cpp:

(WebCore::Font::glyphDataForCharacter): Use helper

platform/graphics/qt/FontQt.cpp:

(WebCore::fromRawDataWithoutRef): Rename from "qstring"
(WebCore::Font::drawComplexText): Use helper
(WebCore::Font::floatWidthForComplexText): Use helper
(WebCore::Font::offsetForPositionForComplexText): Use helper
(WebCore::Font::selectionRectForComplexText): Use helper

platform/text/CharacterNames.h: Add enum for 0x0020
svg/SVGFont.cpp: Use helper to fix bug

(WebCore::SVGTextRunWalker::walk):

LayoutTests:

svg/custom/resources/ABCFont.svg: Added. Font for test case
svg/custom/svg-fonts-in-html-linebreaks-expected.txt: Added.
svg/custom/svg-fonts-in-html-linebreaks.html: New test case

3:05 AM Changeset in webkit [52524] by tkent@chromium.org

10 edits
18 adds in trunk

2009-12-23 Kent Tamura <tkent@chromium.org>

Reviewed by Darin Adler.

HTMLInputElement::valueAsDate getter support.
https://bugs.webkit.org/show_bug.cgi?id=32876

To implement the valueAsDate getter,

Add a method to calculate milliseconds from 1970-01-01 to ISODateTime.
Introduce m_type field to ISODateTime.

Tests: fast/forms/input-valueasdate-date.html

fast/forms/input-valueasdate-datetime.html
fast/forms/input-valueasdate-datetimelocal.html
fast/forms/input-valueasdate-month.html
fast/forms/input-valueasdate-time.html
fast/forms/input-valueasdate-week.html

html/HTMLInputElement.cpp: (WebCore::HTMLInputElement::valueAsDate): Implement it with ISODateTime::millisecondsSinceEpoch(). (WebCore::HTMLInputElement::formStringToISODateTime): Early exit for a null String. This avoids assertion failures in ISODateTime::parse*().
html/ISODateTime.cpp: (WebCore::ISODateTime::parseMonth): Sets m_type. (WebCore::ISODateTime::parseDate): ditto. (WebCore::ISODateTime::parseWeek): ditto. (WebCore::ISODateTime::parseTime): ditto. (WebCore::ISODateTime::parseDateTimeLocal): ditto. (WebCore::ISODateTime::parseDateTime): ditto. (WebCore::ISODateTime::millisecondsSinceEpochForTime): Added. (WebCore::ISODateTime::millisecondsSinceEpoch): Added.
html/ISODateTime.h: (WebCore::ISODateTime::ISODateTime): (WebCore::ISODateTime::invalidMilliseconds):

2009-12-23 Kent Tamura <tkent@chromium.org>