Timeline


and

04/25/10:

23:49 Changeset [58235] by yuzo@google.com

2010-04-25 Yuzo Fujishima <yuzo@google.com>

Unreviewed, Chromium test expectation change.

Rebaseline test expectation for Chromium.
The following tests pass for Chromium.

https://bugs.webkit.org/show_bug.cgi?id=38083

  • platform/chromium/fast/js/ToNumber-expected.txt: Added.
  • platform/chromium/fast/js/parseFloat-expected.txt: Added.
23:45 Changeset [58234] by tony@chromium.org

2010-04-25 Tony Chang <tony@chromium.org>

Not reviewed, trying to get the chromium bots green.
Adding image diffs to pixel tests.

REGRESSION(r58191): Caret moved a pixel, pixel tests failing
https://bugs.webkit.org/show_bug.cgi?id=38104

  • platform/chromium/test_expectations.txt:
23:12 Changeset [58233] by tony@chromium.org

2010-04-25 Tony Chang <tony@chromium.org>

Not reviewed, trying to get the chromium bots
green.

https://bugs.webkit.org/show_bug.cgi?id=38099
Rebaseline a new test and add another new test to the
expected fail list on windows and linux.

  • platform/chromium-linux/fast/css-generated-content/after-duplicated-after-split-expected.checksum: Added.
  • platform/chromium-linux/fast/css-generated-content/after-duplicated-after-split-expected.png: Added.
  • platform/chromium-win/fast/css-generated-content/after-duplicated-after-split-expected.checksum: Added.
  • platform/chromium-win/fast/css-generated-content/after-duplicated-after-split-expected.png: Added.
  • platform/chromium-win/fast/css-generated-content/after-duplicated-after-split-expected.txt: Added.
  • platform/chromium/test_expectations.txt:
22:56 Changeset [58232] by yuzo@google.com

2010-04-25 Yuzo Fujishima <yuzo@google.com>

Unreviewed, Chromium test expectation change.

editing/inserting/caret-position.html is expected to fail for Chromium.
This is because TextInputController::firstRectForCharacterRange is unimplemented for Chromium.

https://bugs.webkit.org/show_bug.cgi?id=38100

  • platform/chromium/test_expectations.txt:
15:03 Changeset [58231] by kevino@webkit.org

Reviewed by Kevin Ollivier.

Update focus handling code to match current approaches used by other platforms,
and fix focus handling for corner cases such as when a mouse down pops up a dialog.

https://bugs.webkit.org/show_bug.cgi?id=38086

09:16 Changeset [58230] by hamaji@chromium.org

2010-04-25 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by Simon Hausmann.

[Qt] layoutTestController.counterValueForElementById crashes for a nonexistent ID
https://bugs.webkit.org/show_bug.cgi?id=34573

  • WebCoreSupport/DumpRenderTreeSupportQt.cpp: (DumpRenderTreeSupportQt::counterValueForElementById):
08:46 Changeset [58229] by abarth@webkit.org

2010-04-25 Andrey Kosyakov <caseq@chromium.org>

Reviewed by Adam Barth.

Added tests for pages event handlers being called with
page context when an event is dispatched from an isolated
world.
https://bugs.webkit.org/show_bug.cgi?id=37947

  • http/tests/security/isolatedWorld/dispatchEvent-expected.txt: Added.
  • http/tests/security/isolatedWorld/dispatchEvent.html: Added.

2010-04-25 Andrey Kosyakov <caseq@chromium.org>

Reviewed by Adam Barth.

Get context properly when event handler was created on
the main world.
https://bugs.webkit.org/show_bug.cgi?id=37947

Test: http/tests/security/isolatedWorld/dispatchEvent.html

  • bindings/v8/WorldContextHandle.cpp: (WebCore::WorldContextHandle::adjustedContext):
06:52 Changeset [58228] by eric@webkit.org

2010-04-25 yael aharon <yael.aharon@nokia.com>

Reviewed by Adele Peterson.

Allow styling of HTMLProgressElement.
https://bugs.webkit.org/show_bug.cgi?id=37901

Added a test for changing the -webkit-appearance statically and dynamically.

  • fast/dom/HTMLProgressElement/progress-bar-value-pseudo-element.html: Added.
  • platform/chromium/test_expectations.txt:
  • platform/gtk/Skipped:
  • platform/mac/fast/dom/HTMLProgressElement: Added.
  • platform/mac/fast/dom/HTMLProgressElement/progress-bar-value-pseudo-element-expected.checksum: Added.
  • platform/mac/fast/dom/HTMLProgressElement/progress-bar-value-pseudo-element-expected.png: Added.
  • platform/mac/fast/dom/HTMLProgressElement/progress-bar-value-pseudo-element-expected.txt: Added.
  • platform/qt/Skipped:
  • platform/win/Skipped:

2010-04-25 yael aharon <yael.aharon@nokia.com>

Reviewed by Adele Peterson.

Allow styling of HTMLProgressElement.
https://bugs.webkit.org/show_bug.cgi?id=37901

Added a new pseudo element to represent the value portion of the progress element.
Web developers can style this pseudo element separate from the rest of the progress element.

Test: fast/dom/HTMLProgressElement/progress-bar-value-pseudo-element.html

  • css/CSSPrimitiveValueMappings.h: (WebCore::CSSPrimitiveValue::CSSPrimitiveValue):
  • css/CSSSelector.cpp: (WebCore::CSSSelector::pseudoId): (WebCore::nameToPseudoTypeMap): (WebCore::CSSSelector::extractPseudoType):
  • css/CSSSelector.h:
  • css/CSSValueKeywords.in:
  • platform/ThemeTypes.h:
  • rendering/RenderProgress.cpp: (WebCore::ProgressValueElement::isShadowNode): (WebCore::ProgressValueElement::shadowParentNode): (WebCore::ProgressValueElement::ProgressValueElement): (WebCore::RenderProgress::RenderProgress): (WebCore::RenderProgress::~RenderProgress): (WebCore::RenderProgress::layout): (WebCore::RenderProgress::styleDidChange): (WebCore::RenderProgress::updateFromElement): (WebCore::RenderProgress::updateValuePartState): (WebCore::RenderProgress::createStyleForValuePart): (WebCore::RenderProgress::updateAnimationState):
  • rendering/RenderTheme.cpp: (WebCore::RenderTheme::isControlStyled):
  • rendering/RenderProgress.h:
  • rendering/style/RenderStyleConstants.h:
04:37 Changeset [58227] by eric@webkit.org

2010-04-25 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Web Inspector: inspector client shouldn't check if it can be opened
docked if it is already in that state.

https://bugs.webkit.org/show_bug.cgi?id=37946

  • WebCoreSupport/WebInspectorClient.cpp: (WebInspectorFrontendClient::WebInspectorFrontendClient): (WebInspectorFrontendClient::showWindowWithoutNotifications):
  • WebCoreSupport/WebInspectorClient.h:

04/24/10:

23:44 Changeset [58226] by sfalken@apple.com

2010-04-24 Steve Falkenburg <sfalken@apple.com>

Reviewed by Sam Weinig.

Typo in Geolocation code causes crashes when updates are stopped
https://bugs.webkit.org/show_bug.cgi?id=38089
<rdar://problem/7904104> Crash closing geolocation tab after allowing to use geolocation

  • WebCoreSupport/WebGeolocationControllerClient.cpp: (WebGeolocationControllerClient::stopUpdating): Call unregister instead of register.
22:06 Changeset [58225] by darin@apple.com

Fix log message.

22:00 Changeset [58224] by darin@apple.com

2010-04-24 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

REGRESSION (r56560): Crash in parseFloat if passed invalid UTF-16 data
https://bugs.webkit.org/show_bug.cgi?id=38083
rdar://problem/7901044

Tests: fast/js/ToNumber.html

fast/js/parseFloat.html

  • runtime/JSGlobalObjectFunctions.cpp: (JSC::parseInt): Added a FIXME comment about a problem I noticed. (JSC::parseFloat): Added a FIXME comment about a problem I noticed; covered by test cases in the test I added.
  • runtime/UString.cpp: (JSC::UString::toDouble): Added FIXME comments about two problem I noticed; covered by test cases in the tests I added. Added a return statement so we don't crash when illegal UTF-16 sequences are present.

2010-04-24 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

REGRESSION (r56560): Crash in parseFloat if passed invalid UTF-16 data
https://bugs.webkit.org/show_bug.cgi?id=38083
rdar://problem/7901044

  • fast/js/parseFloat-expected.txt: Added.
  • fast/js/parseFloat.html: Added.
  • fast/js/script-tests/parseFloat.js: Added.
  • fast/js/ToNumber-expected.txt: Added.
  • fast/js/ToNumber.html: Added.
  • fast/js/script-tests/ToNumber.js: Added.
17:47 Changeset [58223] by mitz@apple.com

Add platform-specific results for Mac after r58222.

  • platform/mac/compositing/iframes: Added.
  • platform/mac/compositing/iframes/composited-iframe-expected.txt: Added.
16:59 Changeset [58222] by mitz@apple.com

Don't propagate compositing out of iframes on Mac
https://bugs.webkit.org/show_bug.cgi?id=38072

Patch by Simon Fraser <simon.fraser@apple.com> on 2010-04-24
Reviewed by Dan Bernstein.

Propagating compositing out of iframes caused too many regressions on Mac,
so only do it for other platforms that may need to hook compositing layers
together across iframe boundaries.

  • rendering/RenderLayerCompositor.cpp:

(WebCore::RenderLayerCompositor::requiresCompositingForIFrame):

14:54 Changeset [58221] by mitz@apple.com

<rdar://problem/7903728> REGRESSION (r51617): WebView fails to load plug-in MIME types
https://bugs.webkit.org/show_bug.cgi?id=38085

Reviewed by Darin Adler.

WebView was calling -_webView in a few places where it should have just used itself. It never
makes sense for WebView to call -_webView on itself, and these calls look like they were copied
from similar code in WebHTMLView, WebFrameView and WebDataSource, where -_webView has a different,
useful meaning.

  • WebView/WebView.mm:

(-[WebView drawSingleRect:]): Replaced [self _webView] with self.
(-[WebView _viewClass:andRepresentationClass:forMIMEType:]): Replaced [[self _webView] preferences]
with _private->preferences.
(-[WebView _canShowMIMEType:]): Ditto.

10:53 Changeset [58220] by eric@webkit.org

2010-04-24 Anton Muhin <antonm@chromium.org>

Reviewed by Darin Adler.

Allow to construct HashTraits<WebCore::QualifiedName>::constructDeletedValue

Former implementation attempted to use AtomicString(HashTableDeletedValue)
however those values cannot be used that way: one cannot construct
QualifiedNameImpl out of such AtomicString as we'll try to lookup this string
in the table, for example.
https://bugs.webkit.org/show_bug.cgi?id=37722

  • wtf/RefPtr.h: expose hash table deleted value

2010-04-24 Anton Muhin <antonm@chromium.org>

Reviewed by Darin Adler.

Allow to construct HashTraits<WebCore::QualifiedName>::constructDeletedValue

Former implementation attempted to use AtomicString(HashTableDeletedValue)
however those values cannot be used that way: one cannot construct
QualifiedNameImpl out of such AtomicString as we'll try to lookup this string
in the table, for example.
https://bugs.webkit.org/show_bug.cgi?id=37722

  • dom/QualifiedName.cpp: (WebCore::QualifiedName::deref): check that hash table deleted values never derefed
  • dom/QualifiedName.h: (WebCore::QualifiedName::QualifiedName): add a constructor to create hash table deleted values (WebCore::QualifiedName::isHashTableDeletedValue): add a check if given instance is hash table deleted value (WTF::):
09:40 Changeset [58219] by jchaffraix@webkit.org

protocolHostAndPortEquals host check makes a wrong assumption
https://bugs.webkit.org/show_bug.cgi?id=37777

Reviewed by Alexey Proskuryakov.

WebCore:

The host check assumed that both host started at the same position. This is true
if both URL are the same but sometimes one has credential and the other does not.
In this case, the method would compare invalid positions.

Test: http/tests/appcache/credential-url.html

  • platform/KURL.cpp:

(WebCore::protocolHostAndPortAreEqual):

  • platform/KURLGoogle.cpp:

(WebCore::protocolHostAndPortAreEqual):
Fix the host check to take both URL's credential into account.

LayoutTests:

Add a check that the URL with credential matches the current URL.

  • http/tests/appcache/credential-url-expected.txt: Added.
  • http/tests/appcache/credential-url.html: Added.
07:43 Changeset [58218] by robert@webkit.org

[Qt] Add more support for textInputController

2010-04-20 Robert Hogan <robert@webkit.org>

Reviewed by Simon Hausmann.

[Qt] Add more support for textInputController

Add support for selectedRange(), setMarkedText(), insertText(),
and firstRectForCharacterRange().

Unskip tests:

fast/forms/input-maxlength-ime-preedit.html
fast/forms/input-maxlength-ime-completed.html
fast/text/international/thai-cursor-position.html
fast/events/ime-composition-events-001.html
editing/selection/5825350-1.html
editing/selection/5825350-2.html
editing/selection/mixed-editability-10.html

https://bugs.webkit.org/show_bug.cgi?id=35702

  • platform/qt/Skipped:

2010-04-20 Robert Hogan <robert@webkit.org>

Reviewed by Simon Hausmann.

[Qt] Add more support for textInputController

Add support for selectedRange(), setMarkedText(), insertText(),
and firstRectForCharacterRange().

Unskip tests:

fast/forms/input-maxlength-ime-preedit.html
fast/forms/input-maxlength-ime-completed.html
fast/text/international/thai-cursor-position.html
fast/events/ime-composition-events-001.html
editing/selection/5825350-1.html
editing/selection/5825350-2.html
editing/selection/mixed-editability-10.html

https://bugs.webkit.org/show_bug.cgi?id=35702

  • DumpRenderTree/qt/TextInputControllerQt.cpp: (TextInputController::setMarkedText): (TextInputController::insertText): (TextInputController::selectedRange): (TextInputController::firstRectForCharacterRange):
  • DumpRenderTree/qt/TextInputControllerQt.h:

2010-04-20 Robert Hogan <robert@webkit.org>

Reviewed by Simon Hausmann.

[Qt] Add more support for textInputController

Add support for selectedRange(), setMarkedText(), insertText(),
and firstRectForCharacterRange().

https://bugs.webkit.org/show_bug.cgi?id=35702

  • Api/qwebpage.cpp: (QWebPagePrivate::inputMethodEvent):
  • WebCoreSupport/DumpRenderTreeSupportQt.cpp: (DumpRenderTreeSupportQt::selectedRange): (DumpRenderTreeSupportQt::firstRectForCharacterRange):
  • WebCoreSupport/DumpRenderTreeSupportQt.h:
  • tests/qwebpage/tst_qwebpage.cpp: (tst_QWebPage::inputMethods):
06:48 Changeset [58217] by zimmermann@webkit.org

2010-04-24 Nikolas Zimmermann <nzimmermann@rim.com>

Not reviewed.


Attempt to turn the last core builder green again (not related anymore to my previous SVG checkins :-)
Skip http/tests/media/video-play-stall-seek.html on Windows as well (just like it has been done for Mac in r58107).
Bug https://bugs.webkit.org/show_bug.cgi?id=38000 describes the problem, and says it's ok to skip the test.

  • platform/win/Skipped:
06:37 Changeset [58216] by zimmermann@webkit.org

2010-04-24 Nikolas Zimmermann <nzimmermann@rim.com>

Not reviewed. Add new results for all SVG tests that were missing on Gtk/Qt (since a while, just took the opportunity to add them now)

  • platform/gtk/svg/custom/broken-internal-references-expected.txt: Added.
  • platform/gtk/svg/custom/clip-path-id-changes-expected.txt: Added.
  • platform/qt/svg/custom/broken-internal-references-expected.txt: Added.
  • platform/qt/svg/custom/clip-path-id-changes-expected.txt: Added.
  • platform/qt/svg/custom/mask-on-multiple-objects-expected.txt: Added.
  • platform/qt/svg/custom/resource-invalidate-on-target-update-expected.txt: Added.
  • platform/qt/svg/filters/feColorMatrix-values-expected.txt: Added.
  • platform/qt/svg/filters/filter-clip-expected.txt: Added.
  • platform/qt/svg/zoom/page/zoom-coords-viewattr-01-b-expected.txt: Added.
  • platform/qt/svg/zoom/text/zoom-coords-viewattr-01-b-expected.txt: Added.
06:01 Changeset [58215] by zimmermann@webkit.org

2010-04-24 Nikolas Zimmermann <nzimmermann@rim.com>

Not reviewed. Update Qt SVG results, and a weird fast/xpath change (newline added).

  • fast/xpath/namespace-vs-predicate-expected.txt:
  • platform/qt/svg/css/circle-in-mask-with-shadow-expected.txt:
  • platform/qt/svg/css/mask-with-shadow-expected.txt:
  • platform/qt/svg/custom/gradient-userSpaceOnUse-with-percentage-expected.txt:
  • platform/qt/svg/custom/pattern-excessive-malloc-expected.txt:
  • platform/qt/svg/custom/pattern-size-bigger-than-target-size-expected.txt:
  • platform/qt/svg/custom/pattern-skew-transformed-expected.txt:
  • platform/qt/svg/custom/radial-gradient-with-outstanding-focalPoint-expected.txt:
  • platform/qt/svg/custom/use-on-symbol-inside-pattern-expected.txt:
  • platform/qt/svg/filters/feTile-expected.txt:
  • platform/qt/svg/text/selection-background-color-expected.txt:
  • platform/qt/svg/text/selection-styles-expected.txt:
05:47 Changeset [58214] by zimmermann@webkit.org

2010-04-24 Nikolas Zimmermann <nzimmermann@rim.com>

Not reviewed. Fix linkage on several bots (no idea, why it worked in a from-scratch rebuild on Leopard on my machine!) -> include NodeRenderStyle.h
Having to include NodeRenderStyle.h just to pull in the renderStyle() is a crazy concept, but I see what it tries to avoid (including RenderObject.h in Node.h)

  • rendering/SVGRenderSupport.cpp:
05:35 Changeset [58213] by zimmermann@webkit.org

2010-04-24 Nikolas Zimmermann <nzimmermann@rim.com>

Not reviewed. Fix release builds, wrap resourceMode assertion in NDEBUG blocks, use UNUSED_PARAM() otherwhise.

  • rendering/RenderSVGResourceClipper.cpp: (WebCore::RenderSVGResourceClipper::applyResource):
  • rendering/RenderSVGResourceFilter.cpp: (WebCore::RenderSVGResourceFilter::applyResource): (WebCore::RenderSVGResourceFilter::postApplyResource):
  • rendering/RenderSVGResourceMasker.cpp: (WebCore::RenderSVGResourceMasker::applyResource):
04:42 Changeset [58212] by zimmermann@webkit.org

2010-04-24 Nikolas Zimmermann <nzimmermann@rim.com>

Reviewed by Dirk Schulze.

SVGPaintServer needs to be converted to the new RenderSVGResource* system
https://bugs.webkit.org/show_bug.cgi?id=37986

Huge speedup for SVG painting using paint servers (gradient/patterns). Cache underlying
WebCore::Gradient/Pattern datastructures, do not rebuild them on every painting. This marks
the finish of the SVGResource -> RenderSVGResource transition.

Outline of same key changes:

  • RenderSVGResource is an abstract base class now, and the previous class is now named RenderSVGResourceContainer All resources except RenderSVGResourceSolidColor now inherit from RenderSVGResourceContainer, as they are all associated with a SVG*Element class. RenderSVGResourceSolidColor inherits from RenderSVGResource, and is not associated with any SVG*Element class. RenderSVGResourceSolidColor is not a render tree object, despite its name. The reason for that is consistency with all other painting resources.
  • RenderSVGResourceSolidColor does not live in the render tree, and exists only as static object, which is shared and always used when filling/stroking with solid colors - just like the old SVGPaintServerSolid.
  • RenderSVGResourceGradient/RenderSVGResourcePattern now store the underlying WebCore::Gradient/Pattern object instead of rebuilding it everytime we're asked to paint -> this is the main difference with the old concept, leading to much faster speed.
  • SVGResource has vanished. All resources (clipper/filter/marker/masker/gradient/pattern) now share the same codepaths to handle updates and client invalidation - which is a huge benefit, and makes the code easier to understand.
  • Android.mk: Remove svg/graphics/SVGResource*, svg/graphics/SVGPaintServer.h from build.
  • GNUmakefile.am: Ditto.
  • WebCore.gypi: Ditto.
  • WebCore.pro: Ditto.
  • WebCore.vcproj/WebCore.vcproj: Ditto.
  • WebCore.xcodeproj/project.pbxproj: Ditto.
  • platform/graphics/Gradient.cpp: Add addColorStop() method taking a ColorStop const-reference. (WebCore::Gradient::addColorStop):
  • platform/graphics/Gradient.h: Ditto.
  • platform/graphics/TextRun.h: Rename SVGPaintServer to RenderSVGResource (WebCore::TextRun::TextRun): (WebCore::TextRun::activePaintingResource): Renamed from activePaintServer. (WebCore::TextRun::setActivePaintingResource): Renamed from setActivePaintServer.
  • platform/graphics/win/GraphicsContextWin.cpp: Remove references to SVGResourceImage, which was removed a long time ago.
  • rendering/RenderObject.cpp: (WebCore::RenderObject::toRenderSVGResourceContainer): Renamed from toRenderSVGResource().
  • rendering/RenderObject.h: (WebCore::RenderObject::isSVGGradientStop): Added. (WebCore::RenderObject::isSVGResourceContainer): Renamed from isSVGResource().
  • rendering/RenderPath.cpp: Utilize RenderSVGResource::fill/strokePaintingResource to request paint servers, instead of SVGPaintServer. (WebCore::RenderPath::fillContains): Adapt to new concept. (WebCore::RenderPath::strokeContains): Ditto. (WebCore::fillAndStrokePath): Ditto. (WebCore::RenderPath::paint): Ditto.
  • rendering/RenderSVGGradientStop.cpp: (WebCore::RenderSVGGradientStop::styleDidChange): Ditto.
  • rendering/RenderSVGGradientStop.h: Fixed indention.
  • rendering/RenderSVGResource.cpp: Refactored old SVGPaintServer code to request fill/stroke painting resources, splitted up in several methods to aid readability. (WebCore::registerPendingResource): (WebCore::adjustColorForPseudoRules): (WebCore::RenderSVGResource::fillPaintingResource): (WebCore::RenderSVGResource::strokePaintingResource): (WebCore::RenderSVGResource::sharedSolidPaintingResource): (WebCore::RenderSVGResource::markForLayoutAndResourceInvalidation):
  • rendering/RenderSVGResource.h: Make RenderSVGResource an abstract base class, see above for the reasoning. (WebCore::RenderSVGResource::RenderSVGResource): (WebCore::RenderSVGResource::~RenderSVGResource): (WebCore::RenderSVGResource::postApplyResource):
  • rendering/RenderSVGResourceClipper.cpp: Inherit from RenderSVGResourceContainer instead of RenderSVGResource. (WebCore::RenderSVGResourceClipper::RenderSVGResourceClipper): (WebCore::RenderSVGResourceClipper::invalidateClient): (WebCore::RenderSVGResourceClipper::applyResource): Adapt to signature changes (add RenderStyle* and resource mode parameters) (WebCore::RenderSVGResourceClipper::applyClippingToContext):
  • rendering/RenderSVGResourceClipper.h:
  • rendering/RenderSVGResourceContainer.h: Added. Renamed from RenderSVGResource. Now inherits from RenderSVGHiddenContainer and RenderSVGResource. (WebCore::RenderSVGResourceContainer::RenderSVGResourceContainer): (WebCore::RenderSVGResourceContainer::~RenderSVGResourceContainer): (WebCore::RenderSVGResourceContainer::idChanged): (WebCore::RenderSVGResourceContainer::isSVGResourceContainer): (WebCore::RenderSVGResourceContainer::drawsContents): (WebCore::RenderSVGResourceContainer::toRenderSVGResourceContainer): (WebCore::getRenderSVGResourceContainerById): (WebCore::getRenderSVGResourceById):
  • rendering/RenderSVGResourceFilter.cpp: Inherit from RenderSVGResourceContainer instead of RenderSVGResource. (WebCore::RenderSVGResourceFilter::RenderSVGResourceFilter): (WebCore::RenderSVGResourceFilter::invalidateClient): (WebCore::RenderSVGResourceFilter::applyResource): Adapt to signature changes (add RenderStyle* and resource mode parameters) (WebCore::RenderSVGResourceFilter::postApplyResource):
  • rendering/RenderSVGResourceFilter.h:
  • rendering/RenderSVGResourceGradient.cpp: Moved from SVGPaintServerGradient, cleaned up and refactored. (WebCore::RenderSVGResourceGradient::RenderSVGResourceGradient): (WebCore::RenderSVGResourceGradient::~RenderSVGResourceGradient): (WebCore::RenderSVGResourceGradient::invalidateClients): (WebCore::RenderSVGResourceGradient::invalidateClient): (WebCore::createMaskAndSwapContextForTextGradient): (WebCore::clipToTextMask): (WebCore::RenderSVGResourceGradient::applyResource):
  • rendering/RenderSVGResourceGradient.h: (WebCore::RenderSVGResourceGradient::resourceBoundingBox):
  • rendering/RenderSVGResourceLinearGradient.cpp: Moved from SVGPaintServerLinearGradient, cleaned up and refactored. (WebCore::RenderSVGResourceLinearGradient::RenderSVGResourceLinearGradient): (WebCore::RenderSVGResourceLinearGradient::~RenderSVGResourceLinearGradient): (WebCore::RenderSVGResourceLinearGradient::buildGradient):
  • rendering/RenderSVGResourceLinearGradient.h: (WebCore::RenderSVGResourceLinearGradient::renderName): (WebCore::RenderSVGResourceLinearGradient::resourceType):
  • rendering/RenderSVGResourceMarker.cpp: Inherit from RenderSVGResourceContainer instead of RenderSVGResource. (WebCore::RenderSVGResourceMarker::RenderSVGResourceMarker): (WebCore::RenderSVGResourceMarker::invalidateClient):
  • rendering/RenderSVGResourceMarker.h: (WebCore::RenderSVGResourceMarker::applyResource):
  • rendering/RenderSVGResourceMasker.cpp: Inherit from RenderSVGResourceContainer instead of RenderSVGResource. (WebCore::RenderSVGResourceMasker::RenderSVGResourceMasker): (WebCore::RenderSVGResourceMasker::invalidateClient): (WebCore::RenderSVGResourceMasker::applyResource):
  • rendering/RenderSVGResourceMasker.h: (WebCore::MaskerData::MaskerData):
  • rendering/RenderSVGResourcePattern.cpp: Moved from SVGPaintServerPattern, cleaned up and refactored. (WebCore::RenderSVGResourcePattern::RenderSVGResourcePattern): (WebCore::RenderSVGResourcePattern::~RenderSVGResourcePattern): (WebCore::RenderSVGResourcePattern::invalidateClients): (WebCore::RenderSVGResourcePattern::invalidateClient): (WebCore::RenderSVGResourcePattern::applyResource): (WebCore::RenderSVGResourcePattern::postApplyResource): (WebCore::calculatePatternBoundaries): (WebCore::RenderSVGResourcePattern::calculatePatternBoundariesIncludingOverflow): (WebCore::RenderSVGResourcePattern::createTileImage): (WebCore::RenderSVGResourcePattern::buildPattern):
  • rendering/RenderSVGResourcePattern.h: (WebCore::RenderSVGResourcePattern::renderName): (WebCore::RenderSVGResourcePattern::resourceBoundingBox): (WebCore::RenderSVGResourcePattern::resourceType):
  • rendering/RenderSVGResourceRadialGradient.cpp: Moved from SVGPaintServerRadialGradient, cleaned up and refactored. (WebCore::RenderSVGResourceRadialGradient::RenderSVGResourceRadialGradient): (WebCore::RenderSVGResourceRadialGradient::~RenderSVGResourceRadialGradient): (WebCore::RenderSVGResourceRadialGradient::buildGradient):
  • rendering/RenderSVGResourceRadialGradient.h: (WebCore::RenderSVGResourceRadialGradient::renderName): (WebCore::RenderSVGResourceRadialGradient::resourceType):
  • rendering/RenderSVGResourceSolidColor.cpp: Moved from SVGPaintServerSolid, cleaned up and refactored. (WebCore::RenderSVGResourceSolidColor::RenderSVGResourceSolidColor): (WebCore::RenderSVGResourceSolidColor::~RenderSVGResourceSolidColor): (WebCore::RenderSVGResourceSolidColor::applyResource): (WebCore::RenderSVGResourceSolidColor::postApplyResource):
  • rendering/RenderSVGResourceSolidColor.h: (WebCore::RenderSVGResourceSolidColor::invalidateClients): (WebCore::RenderSVGResourceSolidColor::invalidateClient): (WebCore::RenderSVGResourceSolidColor::resourceBoundingBox): (WebCore::RenderSVGResourceSolidColor::resourceType): (WebCore::RenderSVGResourceSolidColor::color): (WebCore::RenderSVGResourceSolidColor::setColor):
  • rendering/RenderSVGText.cpp: (WebCore::RenderSVGText::paint): Early exit if painting is disabled.
  • rendering/RenderTreeAsText.cpp: (WebCore::write): Add RenderSVGGradientStop dumping. (WebCore::externalRepresentation): Remove SVGResource dumping.
  • rendering/SVGInlineTextBox.cpp: Adapt to SVGPaintServer changes (paint server request). (WebCore::SVGInlineTextBox::paintCharacters): (WebCore::SVGInlineTextBox::paintDecoration):
  • rendering/SVGInlineTextBox.h: Ditto. (WebCore::SVGTextPaintInfo::SVGTextPaintInfo):
  • rendering/SVGRenderSupport.cpp: Ditto. (WebCore::SVGRenderBase::prepareToRenderSVGContent): (WebCore::SVGRenderBase::finishRenderSVGContent): (WebCore::invalidatePaintingResource): (WebCore::deregisterFromResources): (WebCore::dashArrayFromRenderingStyle): Moved here from SVGPaintServer. (WebCore::applyStrokeStyleToContext): Ditto.
  • rendering/SVGRenderSupport.h:
  • rendering/SVGRenderTreeAsText.cpp: Add new gradient/pattern dumping code. Stops are now properly dumped as well! (WebCore::operator<<): (WebCore::writeSVGPaintingResource): (WebCore::writeStyle): (WebCore::boundingBoxModeString): (WebCore::writeCommonGradientProperties): (WebCore::writeSVGResourceContainer): (WebCore::writeSVGGradientStop):
  • rendering/SVGRenderTreeAsText.h:
  • rendering/SVGRootInlineBox.cpp: Adapt to SVGPaintServer changes (paint server request). (WebCore::SVGRootInlineBoxPaintWalker::SVGRootInlineBoxPaintWalker): (WebCore::SVGRootInlineBoxPaintWalker::~SVGRootInlineBoxPaintWalker): (WebCore::SVGRootInlineBoxPaintWalker::teardownFillPaintServer): (WebCore::SVGRootInlineBoxPaintWalker::teardownStrokePaintServer): (WebCore::SVGRootInlineBoxPaintWalker::setupBackground): (WebCore::SVGRootInlineBoxPaintWalker::setupFill): (WebCore::SVGRootInlineBoxPaintWalker::setupFillSelection): (WebCore::SVGRootInlineBoxPaintWalker::setupStroke): (WebCore::SVGRootInlineBoxPaintWalker::setupStrokeSelection): (WebCore::SVGRootInlineBoxPaintWalker::setupForeground): (WebCore::SVGRootInlineBoxPaintWalker::activePaintingResource): (WebCore::SVGRootInlineBoxPaintWalker::paintChunk):
  • svg/GradientAttributes.h: Remove SVGGradientStop, use Gradient::ColorStop vector. Cleanup. (WebCore::GradientAttributes::stops): (WebCore::GradientAttributes::setStops):
  • svg/LinearGradientAttributes.h: Move ENABLE(SVG) guard to correct location.
  • svg/PatternAttributes.h: Add missing includes and class forwards.
  • svg/RadialGradientAttributes.h: Move ENABLE(SVG) guard to correct location.
  • svg/SVGClipPathElement.cpp: Rename invalidateCanvasResources() to invalidateResourceClients(). (WebCore::SVGClipPathElement::svgAttributeChanged): (WebCore::SVGClipPathElement::childrenChanged): Only call invalidateResourceClients() when !changedByParser is set.
  • svg/SVGClipPathElement.h:
  • svg/SVGDocumentExtensions.cpp: Change HashMaps to hash AtomicStrings instead of Strings. Rename RenderSVGResource to RenderSVGResourceContainer. (WebCore::SVGDocumentExtensions::addResource): (WebCore::SVGDocumentExtensions::removeResource): (WebCore::SVGDocumentExtensions::resourceById): (WebCore::SVGDocumentExtensions::addPendingResource):
  • svg/SVGDocumentExtensions.h:
  • svg/SVGElement.cpp: (WebCore::SVGElement::insertedIntoDocument): Remove reference to SVGResource::invalidateClients(), handled differently now. (WebCore::SVGElement::updateAnimatedSVGAttribute): Change assertion, that is incorrect now.
  • svg/SVGElement.h: Make updateAnimatedSVGAttribute() public.
  • svg/SVGFont.cpp: Adapt to SVGPaintServer changes (paint server request). (WebCore::Font::drawTextUsingSVGFont):
  • svg/SVGGElement.cpp: If style()->display() is NONE, create a RenderSVGHiddenContainer, removing hacks in SVGStopElement for pservers-grad-19-b.svg (yes, SVG is crazy.) (WebCore::SVGGElement::createRenderer):
  • svg/SVGGElement.h: Remove childrenChanged() method which _always_ called renderer->setNeedsLayout(true), which is completly wrong and unnecessary. (WebCore::SVGGElement::rendererIsNeeded): Always return true.
  • svg/SVGGradientElement.cpp: Adapt to new RenderSVGResourceGradient code. (WebCore::SVGGradientElement::svgAttributeChanged): (WebCore::SVGGradientElement::childrenChanged): Only call invalidateResourceClients() when !changedByParser is set. (WebCore::SVGGradientElement::buildStops):
  • svg/SVGGradientElement.h:
  • svg/SVGLinearGradientElement.cpp: Adapt to new RenderSVGResourceLinearGradient code. (WebCore::SVGLinearGradientElement::svgAttributeChanged): (WebCore::SVGLinearGradientElement::createRenderer): (WebCore::SVGLinearGradientElement::collectGradientProperties): (WebCore::SVGLinearGradientElement::calculateStartEndPoints):
  • svg/SVGLinearGradientElement.h:
  • svg/SVGMarkerElement.cpp: Rename invalidateCanvasResources() to invalidateResourceClients(). (WebCore::SVGMarkerElement::svgAttributeChanged): (WebCore::SVGMarkerElement::childrenChanged): Only call invalidateResourceClients() when !changedByParser is set. (WebCore::SVGMarkerElement::setOrientToAuto): (WebCore::SVGMarkerElement::setOrientToAngle):
  • svg/SVGMaskElement.cpp: Rename invalidateCanvasResources() to invalidateResourceClients(). (WebCore::SVGMaskElement::svgAttributeChanged): (WebCore::SVGMaskElement::childrenChanged): Only call invalidateResourceClients() when !changedByParser is set.
  • svg/SVGPatternElement.cpp: Rename invalidateCanvasResources() to invalidateResourceClients(). (WebCore::SVGPatternElement::svgAttributeChanged): (WebCore::SVGPatternElement::childrenChanged): Only call invalidateResourceClients() when !changedByParser is set. (WebCore::SVGPatternElement::createRenderer):
  • svg/SVGPatternElement.h:
  • svg/SVGPolyElement.h: Remove unneded rendererIsNeeded() override.
  • svg/SVGRadialGradientElement.cpp: Adapt to new RenderSVGResourceRadialGradient code. (WebCore::SVGRadialGradientElement::svgAttributeChanged): (WebCore::SVGRadialGradientElement::createRenderer): (WebCore::SVGRadialGradientElement::collectGradientProperties): (WebCore::SVGRadialGradientElement::calculateFocalCenterPointsAndRadius):
  • svg/SVGRadialGradientElement.h:
  • svg/SVGStopElement.cpp: Clean up code, moved stop color calculation from SVGGradientElement to here, where it belongs. (WebCore::SVGStopElement::parseMappedAttribute): (WebCore::SVGStopElement::stopColorIncludingOpacity):
  • svg/SVGStopElement.h:
  • svg/SVGStyledElement.cpp: (WebCore::SVGStyledElement::~SVGStyledElement): (WebCore::SVGStyledElement::svgAttributeChanged): Only invalidateResourcesInAncestorChain()/deregisterFromResources() when we're not parsing. (WebCore::SVGStyledElement::invalidateResourcesInAncestorChain): Early exit if we find a parent resources, there is no resource nesting. (WebCore::SVGStyledElement::invalidateResourceClients): Renamed from invalidateCanvasResources() (WebCore::SVGStyledElement::childrenChanged): Only invalidate SVGElementInstances when !changedByParser is set.
  • svg/SVGStyledElement.h: Remove canvasResource() logic, remove detach() method.
  • svg/SVGTextPathElement.h:
  • svg/graphics/SVGPaintServer.h: Removed.
  • svg/graphics/SVGResource.cpp: Removed.
  • svg/graphics/SVGResource.h: Removed.

2010-04-24 Nikolas Zimmermann <nzimmermann@rim.com>

Reviewed by Dirk Schulze.

SVGPaintServer needs to be converted to the new RenderSVGResource* system
https://bugs.webkit.org/show_bug.cgi?id=37986

Rebaseline SVG tests, after the RenderSVGResource transition.

NOTE: This WILL affect _any_ platform with custom svg results for tests using gradients/patterns.

That means I have to wait for the bot results, and update them ASAP. Therefore landing out
of california business hours and on the weekend - please stay tuned I'll fix any breakages soon.
(Except for chromium whose test results are not available on build.webkit.org, sorry.)

  • platform/mac/svg/W3C-SVG-1.1/animate-elem-31-t-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/animate-elem-78-t-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/color-prop-01-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/coords-units-01-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/filters-color-01-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/filters-comptran-01-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/masking-mask-01-b-expected.checksum:
  • platform/mac/svg/W3C-SVG-1.1/masking-mask-01-b-expected.png:
  • platform/mac/svg/W3C-SVG-1.1/masking-mask-01-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/painting-render-01-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-01-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-02-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-03-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-04-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-05-b-expected.checksum:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-05-b-expected.png:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-05-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-06-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-07-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-08-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-09-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-10-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-11-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-12-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-13-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-14-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-15-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-16-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-17-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-18-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-19-b-expected.checksum:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-19-b-expected.png:
  • platform/mac/svg/W3C-SVG-1.1/pservers-grad-19-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/pservers-pattern-01-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/struct-group-03-t-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/struct-use-05-b-expected.txt:
  • platform/mac/svg/W3C-SVG-1.1/styling-inherit-01-b-expected.txt:
  • platform/mac/svg/batik/filters/feTile-expected.txt:
  • platform/mac/svg/batik/masking/maskRegions-expected.txt:
  • platform/mac/svg/batik/paints/gradientLimit-expected.txt:
  • platform/mac/svg/batik/paints/patternPreserveAspectRatioA-expected.txt:
  • platform/mac/svg/batik/paints/patternRegionA-expected.txt:
  • platform/mac/svg/batik/paints/patternRegions-expected.txt:
  • platform/mac/svg/batik/text/textEffect-expected.txt:
  • platform/mac/svg/batik/text/textEffect2-expected.txt:
  • platform/mac/svg/batik/text/textEffect3-expected.txt:
  • platform/mac/svg/batik/text/textProperties-expected.txt:
  • platform/mac/svg/carto.net/colourpicker-expected.txt:
  • platform/mac/svg/carto.net/tabgroup-expected.txt:
  • platform/mac/svg/css/circle-in-mask-with-shadow-expected.txt:
  • platform/mac/svg/css/mask-with-shadow-expected.txt:
  • platform/mac/svg/custom/broken-internal-references-expected.txt:
  • platform/mac/svg/custom/deep-dynamic-updates-expected.txt:
  • platform/mac/svg/custom/dominant-baseline-hanging-expected.checksum:
  • platform/mac/svg/custom/dominant-baseline-hanging-expected.png:
  • platform/mac/svg/custom/dominant-baseline-hanging-expected.txt:
  • platform/mac/svg/custom/feComponentTransfer-Discrete-expected.txt:
  • platform/mac/svg/custom/feComponentTransfer-Gamma-expected.txt:
  • platform/mac/svg/custom/feComponentTransfer-Linear-expected.txt:
  • platform/mac/svg/custom/feComponentTransfer-Table-expected.txt:
  • platform/mac/svg/custom/fill-fallback-expected.txt:
  • platform/mac/svg/custom/gradient-add-stops-expected.txt:
  • platform/mac/svg/custom/gradient-attr-update-expected.txt:
  • platform/mac/svg/custom/gradient-cycle-detection-expected.txt:
  • platform/mac/svg/custom/gradient-deep-referencing-expected.txt:
  • platform/mac/svg/custom/gradient-rotated-bbox-expected.txt:
  • platform/mac/svg/custom/gradient-stop-corner-cases-expected.txt:
  • platform/mac/svg/custom/gradient-stop-style-change-expected.txt:
  • platform/mac/svg/custom/gradient-stroke-width-expected.checksum:
  • platform/mac/svg/custom/gradient-stroke-width-expected.png:
  • platform/mac/svg/custom/gradient-stroke-width-expected.txt:
  • platform/mac/svg/custom/gradient-userSpaceOnUse-with-percentage-expected.txt:
  • platform/mac/svg/custom/grayscale-gradient-mask-expected.txt:
  • platform/mac/svg/custom/group-opacity-expected.txt:
  • platform/mac/svg/custom/inline-svg-in-xhtml-expected.txt:
  • platform/mac/svg/custom/invalid-css-expected.txt:
  • platform/mac/svg/custom/js-late-gradient-and-object-creation-expected.txt:
  • platform/mac/svg/custom/js-late-gradient-creation-expected.txt:
  • platform/mac/svg/custom/js-late-pattern-and-object-creation-expected.txt:
  • platform/mac/svg/custom/js-late-pattern-creation-expected.txt:
  • platform/mac/svg/custom/js-update-gradient-expected.checksum:
  • platform/mac/svg/custom/js-update-gradient-expected.png:
  • platform/mac/svg/custom/js-update-gradient-expected.txt:
  • platform/mac/svg/custom/js-update-pattern-child-expected.txt:
  • platform/mac/svg/custom/js-update-pattern-expected.txt:
  • platform/mac/svg/custom/js-update-stop-expected.txt:
  • platform/mac/svg/custom/large-bounding-box-percents-expected.txt:
  • platform/mac/svg/custom/pattern-cycle-detection-expected.txt:
  • platform/mac/svg/custom/pattern-deep-referencing-expected.txt:
  • platform/mac/svg/custom/pattern-excessive-malloc-expected.txt:
  • platform/mac/svg/custom/pattern-in-defs-expected.txt:
  • platform/mac/svg/custom/pattern-rotate-expected.txt:
  • platform/mac/svg/custom/pattern-size-bigger-than-target-size-expected.txt:
  • platform/mac/svg/custom/pattern-skew-transformed-expected.txt:
  • platform/mac/svg/custom/pattern-with-transformation-expected.txt:
  • platform/mac/svg/custom/pattern-y-offset-expected.txt:
  • platform/mac/svg/custom/resource-client-removal-expected.txt:
  • platform/mac/svg/custom/stroke-fallback-expected.txt:
  • platform/mac/svg/custom/stroked-pattern-expected.txt:
  • platform/mac/svg/custom/text-gradient-no-content-expected.txt:
  • platform/mac/svg/custom/use-on-symbol-inside-pattern-expected.txt:
  • platform/mac/svg/custom/use-recalcStyle-crash-expected.txt:
  • platform/mac/svg/filters/big-sized-filter-expected.txt:
  • platform/mac/svg/filters/feDisplacementMap-expected.txt:
  • platform/mac/svg/filters/feTile-expected.txt:
  • platform/mac/svg/filters/filterRes-expected.txt:
  • platform/mac/svg/hixie/error/003-expected.txt:
  • platform/mac/svg/hixie/perf/005-expected.txt:
  • platform/mac/svg/hixie/perf/006-expected.txt:
  • platform/mac/svg/text/selection-background-color-expected.txt:
  • platform/mac/svg/text/selection-styles-expected.txt:
  • platform/mac/svg/text/text-gradient-positioning-expected.txt:
  • platform/mac/svg/transforms/text-with-pattern-inside-transformed-html-expected.txt:
  • platform/mac/svg/transforms/text-with-pattern-with-svg-transform-expected.txt:
  • platform/win/svg/custom/broken-internal-references-expected.txt:
  • svg/custom/gradient-with-1d-boundingbox-expected.txt:
  • svg/custom/radial-gradient-with-outstanding-focalPoint-expected.txt:
03:29 Changeset [58211] by zimmermann@webkit.org

2010-04-24 Nikolas Zimmermann <nzimmermann@rim.com>

Not reviewed. Add win specific result for svg/custom/broken-internal-references.svg
It's a bad idea that textPath dumps FLT_MAX when it's not shown - that should be fixed, but it's not related to Moritas testcase.

  • platform/win/svg/custom/broken-internal-references-expected.txt: Added.
00:03 Changeset [58210] by eric@webkit.org

2010-04-23 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

check-webkit-style complains about non-utf8 data in layout test result
https://bugs.webkit.org/show_bug.cgi?id=38027

The problem was we were assuming patch files/diff output as utf-8.
Turns out they're not. We have to treat them as binary data because
a single patch may have multiple text files in it with conflicting encodings!

  • Scripts/webkitpy/common/checkout/api.py:
    • contents_at_revision returns a byte array, so decode it to unicode before passing it to parse_latest_entry_from_file
  • Scripts/webkitpy/common/checkout/api_unittest.py:
    • Update our mock mock_contents_at_revision to match the encoding semantics of the real one.
  • Scripts/webkitpy/common/checkout/scm.py:
    • Be careful not to decode output which may contain file contents (like diff, cat or show) as the encoding for that content is unknown.
  • Scripts/webkitpy/common/checkout/scm_unittest.py:
    • Update our tests to use both latin1 and utf-8 encoded data.
  • Scripts/webkitpy/common/net/bugzilla.py:
    • _fill_attachment_form should not assume unicode data. Callers may wish to attach other types of files to bugs.
  • Scripts/webkitpy/common/prettypatch.py:
    • Diffs are byte arrays, deal with them as such.
  • Scripts/webkitpy/common/prettypatch_unittest.py:
    • Test to make sure we handle diffs with multiple conflicting encodings.
  • Scripts/webkitpy/common/system/executive_unittest.py:
    • Make sure that our unicode support does not break our byte array input support for run_command.

04/23/10:

22:17 Changeset [58209] by eric@webkit.org

2010-04-23 Zhenyao Mo <zmo@google.com>

Reviewed by Darin Fisher.

Add isGLES2Compliant to GraphicsContext3D: make the method const.
https://bugs.webkit.org/show_bug.cgi?id=37872

  • platform/graphics/GraphicsContext3D.h: Make isGLES2Compliant() const.
  • platform/graphics/mac/GraphicsContext3DMac.cpp: Ditto. (WebCore::GraphicsContext3D::isGLES2Compliant):

2010-04-23 Zhenyao Mo <zmo@google.com>

Reviewed by Darin Fisher.

Add isGLES2Compliant to GraphicsContext3D: make the method const.
https://bugs.webkit.org/show_bug.cgi?id=37872

  • src/GraphicsContext3D.cpp: Make isGLES2Compliant() const. (WebCore::GraphicsContext3DInternal::isGLES2Compliant): (WebCore::GraphicsContext3D::isGLES2Compliant):
20:36 Changeset [58208] by eric@webkit.org

2010-04-23 Ben Murdoch <benm@google.com>

Reviewed by Jeremy Orlow.

[Android] The touchmove delta used in basic-single-touch-events.html is too small for Android
https://bugs.webkit.org/show_bug.cgi?id=38056

  • fast/events/touch/basic-single-touch-events-expected.txt:
    • update the test expected output
  • fast/events/touch/script-tests/basic-single-touch-events.js: (verifyTouch): update test expectation with new coordinates. (singleTouchSequence): increase the touch move delta. (touchTargets): Start the touch at a different point to

singleTouchSequence to make examining behavior easier.

20:27 Changeset [58207] by eric@webkit.org

2010-04-23 Qi Zhang <qi.2.zhang@nokia.com>

Reviewed by Laszlo Gombos.

[Qt] LayoutTests/fast/canvas/pointInPath.html passed, actually it failed
https://bugs.webkit.org/show_bug.cgi?id=37276

pointInPath.html passed because the expected.txt expect it fail, it is wrong.

  • platform/qt/fast/canvas/pointInPath-expected.txt: Removed.

2010-04-23 Qi Zhang <qi.2.zhang@nokia.com>

Reviewed by Laszlo Gombos.

[Qt] LayoutTests/fast/canvas/pointInPath.html passed, actually it failed
https://bugs.webkit.org/show_bug.cgi?id=37276

QPainterPath::contains doesn't count the point on the bound.

  • platform/graphics/qt/PathQt.cpp: (WebCore::isPointOnPathBorder): (WebCore::Path::contains):
18:07 Changeset [58206] by weinig@apple.com

Fix for https://bugs.webkit.org/show_bug.cgi?id=38060
Split up Threading.h

Reviewed by David Levin.

JavaScriptCore:

Move bits for of Threading.h into 3 new files.

  • Atomics.h for atomic operations.
  • ThreadSafeShared.h for the ThreadSafeShared class.
  • ThreadingPrimitives.h for the primitives and platform types.

Basic threading operations (creation, etc.) remain in Threading.h.

  • GNUmakefile.am:
  • JavaScriptCore.gypi:
  • JavaScriptCore.vcproj/WTF/WTF.vcproj:
  • JavaScriptCore.xcodeproj/project.pbxproj:
  • wtf/Atomics.h: Copied from wtf/Threading.h.
  • wtf/ThreadSafeShared.h: Copied from wtf/Threading.h.
  • wtf/Threading.h:
  • wtf/ThreadingPrimitives.h: Copied from wtf/Threading.h.

JavaScriptGlue:

Add necessary forwarding headers.

  • ForwardingHeaders/wtf/Atomics.h: Added.
  • ForwardingHeaders/wtf/ThreadSafeShared.h: Added.
  • ForwardingHeaders/wtf/ThreadingPrimitives.h: Added.

WebCore:

Add necessary forwarding headers.

  • ForwardingHeaders/wtf/Atomics.h: Added.
  • ForwardingHeaders/wtf/ThreadSafeShared.h: Added.
  • ForwardingHeaders/wtf/ThreadingPrimitives.h: Added.

WebKitTools:

Add necessary forwarding headers.

  • DumpRenderTree/ForwardingHeaders/wtf/Atomics.h: Added.
  • DumpRenderTree/ForwardingHeaders/wtf/ThreadSafeShared.h: Added.
  • DumpRenderTree/ForwardingHeaders/wtf/ThreadingPrimitives.h: Added.
18:05 Changeset [58205] by eric@webkit.org

2010-04-23 Rafael Weinstein <rafaelw@grafaelw.sfo.corp.google.com>

Reviewed by Darin Fisher.

This patch adds a vector of additionalFeatures to WindowFeatures
and populates it with any feature strings that evaluate to "yes"
and aren't directly observed in WindowFeatures. This allows
clients to capture experimental window features.

https://bugs.webkit.org/show_bug.cgi?id=38013

  • page/WindowFeatures.cpp: (WebCore::WindowFeatures::setWindowFeature):
  • page/WindowFeatures.h:

2010-04-23 Rafael Weinstein <rafaelw@grafaelw.sfo.corp.google.com>

Reviewed by Darin Fisher.

This patch adds a WebWindowFeatures binding struct for the
chromium client in anticipation of a follow-on patch which will
pass it to WebViewClient::createWindow.

https://bugs.webkit.org/show_bug.cgi?id=38013

  • public/WebViewClient.h: (WebKit::WebViewClient::createView):
  • public/WebWindowFeatures.h: Added. (WebKit::WebWindowFeatures::WebWindowFeatures):
  • src/ChromeClientImpl.cpp: (WebKit::ChromeClientImpl::createWindow):
17:54 Changeset [58204] by xji@chromium.org

2010-04-23 Xiaomei Ji <xji@chromium.org>

No need to review.

Add xji as committer.

  • Scripts/webkitpy/common/config/committers.py:
17:48 Changeset [58203] by simon.fraser@apple.com

2010-04-23 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein.

ASSERT(willBeComposited == needsToBeComposited(layer)); on gmail.com
https://bugs.webkit.org/show_bug.cgi?id=38066

When a layer goes into compositing mode because it has a negative z-index child that is going
into compositing mode, then set willBeComposited to true. Also add another assertion to catch
any other cases where the state of willBeComposited is mismanaged.

Test: compositing/composited-negative-zindex-child.html

  • rendering/RenderLayerCompositor.cpp: (WebCore::RenderLayerCompositor::computeCompositingRequirements):
17:48 Changeset [58202] by simon.fraser@apple.com

2010-04-23 Simon Fraser <simon.fraser@apple.com>

Reviewed by Anders Carlsson.

<rdar://problem/7894489> When printing Flash, send a drawRect event, rather than NPPrint

When printing Flash plug-ins in 32-bit, send a drawRect event with a CGContextRef, rather than calling
NPPrint with a GWorldPtr, since Flash prefers the CGContext path.

  • Plugins/WebNetscapePluginView.mm: (-[WebNetscapePluginView drawRect:]):
17:16 Changeset [58201] by mitz@apple.com

<rdar://problem/7898436> :after content is duplicated

Reviewed by Simon Fraser.

WebCore:

Test: fast/css-generated-content/after-duplicated-after-split.html

  • rendering/RenderInline.cpp:

(WebCore::RenderInline::splitInlines): Pass the correct owner of the child list.

LayoutTests:

  • fast/css-generated-content/after-duplicated-after-split.html: Added.
  • platform/mac/fast/css-generated-content/after-duplicated-after-split-expected.checksum: Added.
  • platform/mac/fast/css-generated-content/after-duplicated-after-split-expected.png: Added.
  • platform/mac/fast/css-generated-content/after-duplicated-after-split-expected.txt: Added.
17:12 Changeset [58200] by kevino@webkit.org

[wx] Build fix, add platform stub for new LayoutTestController method.

17:02 Changeset [58199] by weinig@apple.com

https://bugs.webkit.org/show_bug.cgi?id=38065
Merge mac and win DrawingAreaProxyUpdateChunk implementations.

Reviewed by Anders Carlsson.

  • UIProcess/API/mac/WKView.mm:

(-[WKView drawRect:]):

  • UIProcess/DrawingAreaProxy.cpp: Copied from UIProcess/mac/DrawingAreaProxy.mm.
  • UIProcess/DrawingAreaProxy.h: Copied from UIProcess/mac/DrawingAreaProxy.h.
  • UIProcess/DrawingAreaProxyUpdateChunk.cpp: Copied from UIProcess/mac/DrawingAreaProxyUpdateChunk.mm.

(WebKit::DrawingAreaProxyUpdateChunk::DrawingAreaProxyUpdateChunk):
(WebKit::DrawingAreaProxyUpdateChunk::paint):
(WebKit::DrawingAreaProxyUpdateChunk::setSize):
(WebKit::DrawingAreaProxyUpdateChunk::didSetSize):
(WebKit::DrawingAreaProxyUpdateChunk::update):
(WebKit::DrawingAreaProxyUpdateChunk::didReceiveMessage):

  • UIProcess/DrawingAreaProxyUpdateChunk.h: Copied from UIProcess/mac/DrawingAreaProxyUpdateChunk.h.
  • UIProcess/mac/DrawingAreaProxy.h: Removed.
  • UIProcess/mac/DrawingAreaProxy.mm: Removed.
  • UIProcess/mac/DrawingAreaProxyUpdateChunk.h: Removed.
  • UIProcess/mac/DrawingAreaProxyUpdateChunk.mm: Removed.
  • UIProcess/mac/DrawingAreaProxyUpdateChunkMac.mm: Copied from UIProcess/mac/DrawingAreaProxyUpdateChunk.mm.

(WebKit::DrawingAreaProxyUpdateChunk::page):
(WebKit::DrawingAreaProxyUpdateChunk::invalidateBackingStore):
(WebKit::DrawingAreaProxyUpdateChunk::platformPaint):
(WebKit::DrawingAreaProxyUpdateChunk::drawUpdateChunkIntoBackingStore):

  • UIProcess/win/DrawingAreaProxy.cpp: Removed.
  • UIProcess/win/DrawingAreaProxy.h: Removed.
  • UIProcess/win/DrawingAreaProxyUpdateChunkWin.cpp: Copied from UIProcess/win/DrawingAreaProxy.cpp.

(WebKit::DrawingAreaProxyUpdateChunk::page):
(WebKit::DrawingAreaProxyUpdateChunk::ensureBackingStore):
(WebKit::DrawingAreaProxyUpdateChunk::invalidateBackingStore):
(WebKit::DrawingAreaProxyUpdateChunk::platformPaint):
(WebKit::DrawingAreaProxyUpdateChunk::drawUpdateChunkIntoBackingStore):

  • UIProcess/win/WebView.cpp:

(WebKit::WebView::WebView):
(WebKit::WebView::onPaintEvent):

  • WebKit2.xcodeproj/project.pbxproj:
  • win/WebKit2.vcproj:
16:50 Changeset [58198] by xji@chromium.org

2010-04-23 Xiaomei Ji <xji@chromium.org>

Reviewed by Eric Seidel

QT does not support textInputController.firstRectForCharacterRange(),
so, skip editing/inserting/caret-position.html for QT.
https://bugs.webkit.org/show_bug.cgi?id=25319

  • platform/qt/Skipped:
16:21 Changeset [58197] by adele@apple.com

Fixing the Tiger build for real.

  • platform/graphics/mac/SimpleFontDataMac.mm:
16:08 Changeset [58196] by adele@apple.com

Fixing the Tiger build.

  • platform/graphics/mac/SimpleFontDataMac.mm:

(WebCore::SimpleFontData::platformBoundsForGlyph):

15:31 Changeset [58195] by morrita@google.com

2010-04-23 MORITA Hajime <morrita@google.com>

Not reviewed. Added missing expectation files for
https://bugs.webkit.org/show_bug.cgi?id=37187

  • platform/mac/svg/custom/broken-internal-references-expected.checksum: Added.
  • platform/mac/svg/custom/broken-internal-references-expected.png: Added.
15:20 Changeset [58194] by jianli@chromium.org

Add FileError for File API.
https://bugs.webkit.org/show_bug.cgi?id=37840

Reviewed by Dmitry Titov.

The test will be added when implementing FileReader and FilerWriter.

  • Android.derived.jscbindings.mk:
  • Android.derived.v8bindings.mk:
  • DerivedSources.cpp:
  • DerivedSources.make:
  • GNUmakefile.am:
  • WebCore.gypi:
  • WebCore.pri:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • dom/ExceptionCode.h:

(WebCore::):

  • html/FileError.h: Added.
  • html/FileError.idl: Added.
  • page/DOMWindow.idl:
15:20 Changeset [58193] by jianli@chromium.org

Improve code generator scripts to handle OR ('|') condition in the
extended attributes.
https://bugs.webkit.org/show_bug.cgi?id=37998

Reviewed by Adam Barth.

  • bindings/scripts/CodeGeneratorJS.pm:
  • bindings/scripts/CodeGeneratorObjC.pm:
  • bindings/scripts/CodeGeneratorV8.pm:
  • bindings/scripts/IDLStructure.pm:
15:16 Changeset [58192] by adele@apple.com

Fix for <rdar://problem/7855777> REGRESSION: Memory usage increase caused by storing glyph bounds in GlyphMetricsMap
https://bugs.webkit.org/show_bug.cgi?id=37936

Reviewed by Dan Bernstein.

This change breaks the GlyphMetricsMap into two maps - one for width and one for bounds, so that we don't store
a FloatRect for the glyph bounds unless we need to.

Covered by existing tests. This should not cause any change in functionality.

Updated for removal of GlyphMetricsMap.cpp

  • Android.mk:
  • GNUmakefile.am:
  • WebCore.gypi:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:

Made GlyphMetricsMap a template, so it can be used for separate width and bounds maps.

  • platform/graphics/GlyphMetricsMap.cpp: Removed.
  • platform/graphics/GlyphMetricsMap.h:

(WebCore::GlyphMetricsMap::metricsForGlyph):
(WebCore::GlyphMetricsMap::setMetricsForGlyph):
(WebCore::GlyphMetricsMap::GlyphMetricsPage::metricsForGlyph):
(WebCore::GlyphMetricsMap::GlyphMetricsPage::setMetricsForGlyph):
(WebCore::GlyphMetricsMap::GlyphMetricsPage::setMetricsForIndex):
(WebCore::::unknownMetrics):
(WebCore::::locatePageSlowCase):

  • platform/graphics/mac/ComplexTextController.cpp: (WebCore::ComplexTextController::adjustGlyphsAndAdvances):

Call boundsForGlyph instead of metricsForGlyph.

  • platform/graphics/win/UniscribeController.cpp: (WebCore::UniscribeController::shapeAndPlaceItem):

ditto.

  • platform/graphics/SimpleFontData.cpp: (WebCore::SimpleFontData::platformGlyphInit):

Call setMetricsForGlyph for both glyph maps.

Break getters and setters for metricsForGlyph into widthForGlyph and boundsForGlyph, maintaining present behavior.

  • platform/graphics/SimpleFontData.h:

(WebCore::):
(WebCore::SimpleFontData::boundsForGlyph):
(WebCore::SimpleFontData::widthForGlyph):

  • platform/graphics/cairo/SimpleFontDataCairo.cpp:

(WebCore::SimpleFontData::platformBoundsForGlyph):
(WebCore::SimpleFontData::platformWidthForGlyph):

  • platform/graphics/chromium/SimpleFontDataChromiumWin.cpp:

(WebCore::SimpleFontData::platformBoundsForGlyph):
(WebCore::SimpleFontData::platformWidthForGlyph):

  • platform/graphics/chromium/SimpleFontDataLinux.cpp:

(WebCore::SimpleFontData::platformBoundsForGlyph):
(WebCore::SimpleFontData::platformWidthForGlyph):

  • platform/graphics/gtk/SimpleFontDataPango.cpp:

(WebCore::SimpleFontData::platformBoundsForGlyph):
(WebCore::SimpleFontData::platformWidthForGlyph):

  • platform/graphics/mac/SimpleFontDataMac.mm:

(WebCore::SimpleFontData::platformBoundsForGlyph):
(WebCore::SimpleFontData::platformWidthForGlyph):

  • platform/graphics/win/SimpleFontDataCGWin.cpp:

(WebCore::SimpleFontData::platformBoundsForGlyph):
(WebCore::SimpleFontData::platformWidthForGlyph):

  • platform/graphics/win/SimpleFontDataCairoWin.cpp:

(WebCore::SimpleFontData::platformBoundsForGlyph):

  • platform/graphics/win/SimpleFontDataWin.cpp:

(WebCore::SimpleFontData::boundsForGDIGlyph):
(WebCore::SimpleFontData::widthForGDIGlyph):

  • platform/graphics/wince/SimpleFontDataWince.cpp:

(WebCore::SimpleFontData::platformBoundsForGlyph):

  • platform/graphics/wx/SimpleFontDataWx.cpp:

(WebCore::SimpleFontData::platformBoundsForGlyph):
(WebCore::SimpleFontData::platformWidthForGlyph):

15:07 Changeset [58191] by xji@chromium.org

2010-04-23 Xiaomei Ji <xji@chromium.org>

Reviewed by Dan Bernstein and Darin Adler

Fix issue "caret does not paint after type in characters in right
aligned div or after delete all characters in RTL div or
0px right padding RTL textarea"
https://bugs.webkit.org/show_bug.cgi?id=25319

Test: editing/inserting/caret-position.html

  • rendering/RenderText.cpp: (WebCore::RenderText::localCaretRect):

2010-04-23 Xiaomei Ji <xji@chromium.org>

Reviewed by Dan Bernstein and Darin Adler

Fix issue "caret does not paint after type in characters in right
aligned div or after delete all characters in RTL div or
0px right padding RTL textarea"
https://bugs.webkit.org/show_bug.cgi?id=25319

  • editing/inserting/caret-position-expected.txt: Added.
  • editing/inserting/caret-position.html: Added.
  • platform/mac/editing/input/caret-primary-bidi-expected.txt:
15:02 Changeset [58190] by eric@webkit.org

2010-04-23 No'am Rosenthal <noam.rosenthal@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] [Performance] GraphicsLayer: constructing the layers takes too long
https://bugs.webkit.org/show_bug.cgi?id=36365

The issue came from using QGraphicsView's cache as is. The problem is that
several code-paths require re-rendering of the item, but not re-rendering
of the web content into the cache.

The way to solve it is by having GraphicsLayerQt manage the cache directly
via QPixmapCache, instead of using QGraphicsItem cache modes.

FPS measurement shows significant improvement (20FPS before, 40FPS after)
on several use-cases, including blog-files/leaves on a desktop environment.

  • platform/graphics/qt/GraphicsLayerQt.cpp: (WebCore::GraphicsLayerQtImpl::GraphicsLayerQtImpl): (WebCore::GraphicsLayerQtImpl::recache): (WebCore::GraphicsLayerQtImpl::paint): (WebCore::GraphicsLayerQtImpl::flushChanges):
14:58 Changeset [58189] by weinig@apple.com

Fix for https://bugs.webkit.org/show_bug.cgi?id=38059
Merge mac and win DrawingAreaUpdateChunk implementation.

Reviewed by Anders Carlsson.

  • UIProcess/win/DrawingAreaProxy.cpp:

(WebKit::DrawingAreaProxy::didSetSize):
(WebKit::DrawingAreaProxy::didReceiveMessage):

  • UIProcess/win/DrawingAreaProxy.h:
  • WebKit2.xcodeproj/project.pbxproj:
  • WebProcess/WebPage/DrawingAreaUpdateChunk.cpp: Copied from WebProcess/WebPage/mac/DrawingAreaUpdateChunk.cpp.

(WebKit::DrawingAreaUpdateChunk::setSize):

  • WebProcess/WebPage/DrawingAreaUpdateChunk.h: Copied from WebProcess/WebPage/mac/DrawingAreaUpdateChunk.h.
  • WebProcess/WebPage/mac/DrawingAreaUpdateChunk.cpp: Removed.
  • WebProcess/WebPage/mac/DrawingAreaUpdateChunk.h: Removed.
  • WebProcess/WebPage/mac/DrawingAreaUpdateChunkMac.cpp: Copied from WebProcess/WebPage/mac/DrawingAreaUpdateChunk.cpp.
  • WebProcess/WebPage/win/DrawingAreaUpdateChunk.cpp: Removed.
  • WebProcess/WebPage/win/DrawingAreaUpdateChunk.h: Removed.
  • WebProcess/WebPage/win/DrawingAreaUpdateChunkWin.cpp: Copied from WebProcess/WebPage/win/DrawingAreaUpdateChunk.cpp.

(WebKit::DrawingAreaUpdateChunk::paintIntoUpdateChunk):

  • win/WebKit2.vcproj:
14:13 Changeset [58188] by jamesr@google.com

2010-04-23 James Robinson <jamesr@chromium.org>

Reviewed by Dimitri Glazkov.

Avoid marking symbols as dllexport in chromium win builds
https://bugs.webkit.org/show_bug.cgi?id=38058

No new tests.

  • config.h:
13:39 Changeset [58187] by morrita@google.com

2010-04-23 MORITA Hajime <morrita@google.com>

Reviewed by Nikolas Zimmermann.

https://bugs.webkit.org/show_bug.cgi?id=37187
SVG <use href="foo"> is interpreted as <use href="#foo">

getTarget() did return url parameter as is if doesn't have
fragment identifier. So fixed to return empty string in such case
because we need to distinguish "yyy.html" from "xxx.svg#yyy.html".


Test: svg/custom/broken-internal-references.svg

  • svg/SVGElement.cpp: (WebCore::SVGElement::insertedIntoDocument):
  • svg/SVGURIReference.cpp: (WebCore::SVGURIReference::getTarget):

2010-04-23 MORITA Hajime <morrita@google.com>

Reviewed by Nikolas Zimmermann.

https://bugs.webkit.org/show_bug.cgi?id=37187
SVG <use href="foo"> is interpreted as <use href="#foo">

  • platform/mac/svg/custom/broken-internal-references-expected.checksum: Added.
  • platform/mac/svg/custom/broken-internal-references-expected.png: Added.
  • platform/mac/svg/custom/broken-internal-references-expected.txt: Added.
  • svg/custom/broken-internal-references.svg: Added.
  • svg/dynamic-updates/SVGTRefElement-dom-href-attr-expected.txt:
  • svg/dynamic-updates/script-tests/SVGTRefElement-dom-href-attr.js: (executeTest): Updated to follow this fix. href value should be an URL, not an id.
13:22 Changeset [58186] by simon.fraser@apple.com

2010-04-23 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein.

When webkitAnimationEnd event fires, on-screen rendering should show the last frame of animation
https://bugs.webkit.org/show_bug.cgi?id=37955

After r37484, animation and transition end events were fired on a timer, after the animation/transition
ended. This opened up one event loop cycle in which the non-animating state of the element was visible
before the event fired, resulting in flashes in content that reset style from the event handler.

Fix by firing these events in the same event cycle as the animation end, once all animations
have been updated. This also required moving the place that start animations are fixed until
a later state in the state machine, so that animations have their start time set at the point
the event is fired (to avoid an assertion when using the freeze API in DRT).

Not testable, since the flash is very transitory and cannot be captured in a pixel test.

  • page/animation/AnimationBase.cpp: (WebCore::AnimationBase::updateStateMachine): Move the firing of 'start' events into the state where the start time is known.
  • page/animation/AnimationControllerPrivate.h:
  • page/animation/AnimationController.cpp: (WebCore::AnimationControllerPrivate::updateStyleIfNeededDispatcherFired): Call the new fireEventsAndUpdateStyle() method. (WebCore::AnimationControllerPrivate::fireEventsAndUpdateStyle): New method to share code that used to be in updateStyleIfNeededDispatcherFired(). (WebCore::AnimationControllerPrivate::animationTimerFired): Call fireEventsAndUpdateStyle() once we've processed all animations.
13:22 Changeset [58185] by simon.fraser@apple.com

2010-04-23 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein.

CompositeAnimation::updateKeyframeAnimations() can make a KeyframeAnimation for animation "none"
https://bugs.webkit.org/show_bug.cgi?id=38017

Check to see if the animation name is "none", and don't fire off a keyframe animation in
that case.

  • page/animation/CompositeAnimation.cpp: (WebCore::CompositeAnimation::updateKeyframeAnimations):
13:10 Changeset [58184] by andersca@apple.com

Remove an assert. (It's not valid when resizing).

Reviewed by Sam Weinig.

  • WebProcess/WebPage/win/DrawingAreaUpdateChunk.cpp:

(WebKit::DrawingAreaUpdateChunk::didUpdate):

12:43 Changeset [58183] by dpranke@chromium.org

2010-04-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Adam Barth.

new-run-webkit-tests: add a "--retry-failures" flag and a
"--no-retry-failures" flag (the former is the default). Also, rename
"--print-unexpected-results" and "--retry-unexpected-results" to
"--print-last-failures" and "--retry-last-failures" because the
retry flag was confusing. The new flag names aren't great, but
hopefully they're less confusing.

https://bugs.webkit.org/show_bug.cgi?id=37838

  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
12:35 Changeset [58182] by agl@chromium.org

2010-04-23 Adam Langley <agl@chromium.org>

Unreviewed, expectations update.

Merging Skia r526 lead to a number of pixels changing color by 1 in
layout tests. This is the rebaseline for Windows and Linux. (Mac isn't
using Skia for this.)

  • platform/chromium-linux/css2.1/t0803-c5505-mrgn-00-b-ag-expected.checksum:
  • platform/chromium-linux/css2.1/t0803-c5505-mrgn-00-b-ag-expected.png:
  • platform/chromium-linux/css2.1/t0804-c5509-padn-l-03-f-g-expected.checksum:
  • platform/chromium-linux/css2.1/t0804-c5509-padn-l-03-f-g-expected.png:
  • platform/chromium-linux/fast/dom/Window/btoa-pnglet-expected.checksum:
  • platform/chromium-linux/fast/dom/Window/btoa-pnglet-expected.png:
  • platform/chromium-linux/fast/replaced/image-solid-color-with-alpha-expected.checksum:
  • platform/chromium-linux/fast/replaced/image-solid-color-with-alpha-expected.png:
  • platform/chromium-linux/tables/mozilla/bugs/bug10565-expected.checksum:
  • platform/chromium-linux/tables/mozilla/bugs/bug10565-expected.png:
  • platform/chromium-linux/tables/mozilla/bugs/bug10633-expected.checksum:
  • platform/chromium-linux/tables/mozilla/bugs/bug10633-expected.png:
  • platform/chromium-linux/tables/mozilla/bugs/bug14323-expected.checksum:
  • platform/chromium-linux/tables/mozilla/bugs/bug14323-expected.png:
  • platform/chromium-linux/tables/mozilla/bugs/bug16012-expected.checksum:
  • platform/chromium-linux/tables/mozilla/bugs/bug16012-expected.png:
  • platform/chromium-linux/tables/mozilla/bugs/bug20804-expected.checksum:
  • platform/chromium-linux/tables/mozilla/bugs/bug20804-expected.png:
  • platform/chromium-linux/tables/mozilla/bugs/bug4427-expected.checksum:
  • platform/chromium-linux/tables/mozilla/bugs/bug4427-expected.png:
  • platform/chromium-linux/tables/mozilla/bugs/bug4803-expected.checksum:
  • platform/chromium-linux/tables/mozilla/bugs/bug4803-expected.png:
  • platform/chromium-win/css2.1/t0803-c5505-mrgn-00-b-ag-expected.checksum:
  • platform/chromium-win/css2.1/t0803-c5505-mrgn-00-b-ag-expected.png:
  • platform/chromium-win/css2.1/t0804-c5509-padn-l-03-f-g-expected.checksum:
  • platform/chromium-win/css2.1/t0804-c5509-padn-l-03-f-g-expected.png:
  • platform/chromium-win/fast/dom/Window/btoa-pnglet-expected.checksum:
  • platform/chromium-win/fast/dom/Window/btoa-pnglet-expected.png:
  • platform/chromium-win/fast/replaced/image-solid-color-with-alpha-expected.checksum:
  • platform/chromium-win/fast/replaced/image-solid-color-with-alpha-expected.png:
  • platform/chromium-win/tables/mozilla/bugs/bug10565-expected.checksum:
  • platform/chromium-win/tables/mozilla/bugs/bug10565-expected.png:
  • platform/chromium-win/tables/mozilla/bugs/bug10633-expected.checksum:
  • platform/chromium-win/tables/mozilla/bugs/bug10633-expected.png:
  • platform/chromium-win/tables/mozilla/bugs/bug14323-expected.checksum:
  • platform/chromium-win/tables/mozilla/bugs/bug14323-expected.png:
  • platform/chromium-win/tables/mozilla/bugs/bug16012-expected.checksum:
  • platform/chromium-win/tables/mozilla/bugs/bug16012-expected.png:
  • platform/chromium-win/tables/mozilla/bugs/bug20804-expected.checksum:
  • platform/chromium-win/tables/mozilla/bugs/bug20804-expected.png:
  • platform/chromium-win/tables/mozilla/bugs/bug4427-expected.checksum:
  • platform/chromium-win/tables/mozilla/bugs/bug4427-expected.png:
  • platform/chromium-win/tables/mozilla/bugs/bug4803-expected.checksum:
  • platform/chromium-win/tables/mozilla/bugs/bug4803-expected.png:
12:11 Changeset [58181] by weinig@apple.com

Fix Qt build.

  • wtf/qt/MainThreadQt.cpp: #include <QThread>
12:01 Changeset [58180] by andersca@apple.com

Fix build.

  • UIProcess/win/DrawingAreaProxy.cpp:

(WebKit::DrawingAreaProxy::paint):
(WebKit::DrawingAreaProxy::setSize):
(WebKit::DrawingAreaProxy::didReceiveMessage):

  • WebProcess/WebPage/win/DrawingAreaUpdateChunk.cpp:

(WebKit::DrawingAreaUpdateChunk::setSize):
(WebKit::DrawingAreaUpdateChunk::didReceiveMessage):

11:59 Changeset [58179] by weinig@apple.com

Fix for https://bugs.webkit.org/show_bug.cgi?id=38022
Move isMainThread predicate function to MainThread.h

Reviewed by Anders Carlsson.

JavaScriptCore:

  • wtf/MainThread.cpp:

(WTF::initializeMainThread):
(WTF::isMainThread):

  • wtf/MainThread.h:
  • wtf/Threading.h:
  • wtf/ThreadingPthreads.cpp:

(WTF::initializeThreading):

  • wtf/ThreadingWin.cpp:

(WTF::initializeThreading):

  • wtf/gtk/ThreadingGtk.cpp:

(WTF::initializeThreading):

  • wtf/mac/MainThreadMac.mm:

(WTF::initializeMainThreadPlatform):
(WTF::isMainThread):

  • wtf/qt/MainThreadQt.cpp:

(WTF::isMainThread):

  • wtf/qt/ThreadingQt.cpp:

(WTF::initializeThreading):

  • wtf/text/AtomicString.cpp:

JavaScriptGlue:

Added forwarding header for MainThread.h

  • ForwardingHeaders/wtf/MainThread.h: Added.

WebKitTools:

Added forwarding header for MainThread.h

  • DumpRenderTree/ForwardingHeaders/wtf/MainThread.h: Added.
11:31 Changeset [58178] by eric@webkit.org

2010-04-23 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Disable Netscape plugin support for minimal configuration
https://bugs.webkit.org/show_bug.cgi?id=38026

No new tests, as there is no new functionality.

  • WebCore.pri:

2010-04-23 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Disable Netscape plugin support for minimal configuration
https://bugs.webkit.org/show_bug.cgi?id=38026

Pass the minimal configuration option to Qt build system
as part of the CONFIG variable.

  • Scripts/build-webkit:
10:33 Changeset [58177] by hamaji@chromium.org

2010-04-23 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by Darin Adler.

Need borderAndPadding(Width|Height)
https://bugs.webkit.org/show_bug.cgi?id=38046

Refactoring only, so no new tests.

  • rendering/InlineBox.cpp: (WebCore::InlineBox::height):
  • rendering/RenderApplet.cpp: (WebCore::RenderApplet::createWidgetIfNecessary):
  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::isSelfCollapsingBlock): (WebCore::RenderBlock::calcPrefWidths):
  • rendering/RenderBox.cpp: (WebCore::RenderBox::calcBorderBoxWidth): (WebCore::RenderBox::calcBorderBoxHeight): (WebCore::RenderBox::calcContentBoxWidth): (WebCore::RenderBox::calcContentBoxHeight): (WebCore::RenderBox::calcWidth): (WebCore::RenderBox::calcHeight): (WebCore::RenderBox::calcPercentageHeight): (WebCore::RenderBox::calcReplacedHeightUsing): (WebCore::RenderBox::availableHeightUsing): (WebCore::RenderBox::calcAbsoluteHorizontal): (WebCore::RenderBox::calcAbsoluteVertical): (WebCore::RenderBox::calcAbsoluteHorizontalReplaced): (WebCore::RenderBox::calcAbsoluteVerticalReplaced): (WebCore::RenderBox::positionForPoint):
  • rendering/RenderBoxModelObject.h: (WebCore::RenderBoxModelObject::borderAndPaddingHeight): (WebCore::RenderBoxModelObject::borderAndPaddingWidth):
  • rendering/RenderDataGrid.cpp: (WebCore::RenderDataGrid::calcPrefWidths):
  • rendering/RenderFieldset.cpp: (WebCore::RenderFieldset::calcPrefWidths):
  • rendering/RenderFileUploadControl.cpp: (WebCore::RenderFileUploadControl::calcPrefWidths):
  • rendering/RenderFlexibleBox.cpp: (WebCore::RenderFlexibleBox::calcPrefWidths): (WebCore::RenderFlexibleBox::allowedChildFlex):
  • rendering/RenderImage.cpp: (WebCore::RenderImage::calcPrefWidths):
  • rendering/RenderLayer.cpp: (WebCore::RenderLayer::resize):
  • rendering/RenderListBox.cpp: (WebCore::RenderListBox::calcPrefWidths): (WebCore::RenderListBox::calcHeight):
  • rendering/RenderMenuList.cpp: (WebCore::RenderMenuList::calcPrefWidths):
  • rendering/RenderReplaced.cpp: (WebCore::RenderReplaced::calcPrefWidths):
  • rendering/RenderSVGRoot.cpp: (WebCore::RenderSVGRoot::calcPrefWidths):
  • rendering/RenderSlider.cpp: (WebCore::RenderSlider::calcPrefWidths): (WebCore::RenderSlider::layout):
  • rendering/RenderTableCell.cpp: (WebCore::RenderTableCell::styleOrColWidth):
  • rendering/RenderTextControl.cpp: (WebCore::RenderTextControl::textBlockHeight): (WebCore::RenderTextControl::textBlockWidth): (WebCore::RenderTextControl::calcHeight): (WebCore::RenderTextControl::calcPrefWidths):
  • rendering/RenderTextControlSingleLine.cpp: (WebCore::RenderTextControlSingleLine::layout):
  • rendering/RenderWidget.cpp: (WebCore::RenderWidget::updateWidgetPosition):
10:18 Changeset [58176] by ddkilzer@apple.com

Size mismatch between format string and argument in dumpTextEncodingNameMap()

Reviewed by Darin Adler.

See Bug 38030 and r58157.

  • platform/text/TextEncodingRegistry.cpp:

(WebCore::dumpTextEncodingNameMap): Assigned
textEncodingNameMap->size() to an unsigned variable before using
it in the fprintf() statement.

10:00 Changeset [58175] by ddkilzer@apple.com

BUILD FIX: Attempt to fix Windows build after Bug 36187 landed in r58168

  • DerivedSources.cpp: Changed include of JSMedia.cpp to

JSStyleMedia.cpp.

09:47 Changeset [58174] by ddkilzer@apple.com

BUILD FIX: Fix Mac build after Bug 36187 landed in r58168

  • WebCore.xcodeproj/project.pbxproj: Renamed DOMMedia.h,

DOMMedia.mm, DOMMediaInternal.h to DOMStyleMedia.h,
DOMStyleMedia.mm, DOMStyleMediaInternal.h.

09:33 Changeset [58173] by kenneth@webkit.org

Unreviewed build fix.

Change Media to StyleMedia

  • DerivedSources.make:
09:28 Changeset [58172] by kenneth@webkit.org

Unreviewed layout fix.

Change [object Media] to [object StyleMedia]

  • platform/qt/fast/dom/Window/window-properties-expected.txt:
09:24 Changeset [58171] by kenneth@webkit.org

Touch WebKit.idl to force a rebuild.

09:15 Changeset [58170] by kenneth@webkit.org

Unreviewed layout fix.

Change [object Media] to [object StyleMedia]

  • fast/dom/Window/window-properties-expected.txt:
09:08 Changeset [58169] by xan@webkit.org

2010-04-23 Xan Lopez <xlopez@igalia.com>

Try to fix the GTK+ bots.

  • GNUmakefile.am:
08:42 Changeset [58168] by kenneth@webkit.org

WebCore: Rename window.media to window.styleMedia
https://bugs.webkit.org/show_bug.cgi?id=36187

Reviewed by Laszlo Gombos.

Rename the interface Media to StyleMedia as required by the
new CSSOM View spec.

  • Android.derived.jscbindings.mk:
  • Android.derived.v8bindings.mk:
  • GNUmakefile.am:
  • WebCore.gypi:
  • WebCore.pri:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • css/Media.cpp: Removed.
  • css/Media.h: Removed.
  • css/Media.idl: Removed.
  • css/StyleMedia.cpp: Added.

(WebCore::StyleMedia::StyleMedia):
(WebCore::StyleMedia::type):
(WebCore::StyleMedia::matchMedium):

  • css/StyleMedia.h: Added.

(WebCore::StyleMedia::create):
(WebCore::StyleMedia::disconnectFrame):

  • css/StyleMedia.idl: Added.
  • page/DOMWindow.cpp:

(WebCore::DOMWindow::styleMedia):

  • page/DOMWindow.h:

(WebCore::DOMWindow::optionalMedia):

  • page/DOMWindow.idl:

LayoutTests: Updated test results as the interface Media was renamed
to StyleMedia.

Reviewed by NOBODY (OOPS!).

  • fast/dom/prototype-inheritance-2-expected.txt:
  • platform/win/fast/dom/prototype-inheritance-2-expected.txt:
08:38 Changeset [58167] by eric@webkit.org

2010-04-23 Gabor Rapcsanyi <rgabor@inf.u-szeged.hu>

Reviewed by Laszlo Gombos.

Add RVCT support for Linux

Developed in cooperation with Gabor Loki.

  • API/JSStringRef.h:
  • jit/ExecutableAllocator.h:
  • jit/ExecutableAllocatorPosix.cpp: (JSC::ExecutableAllocator::cacheFlush):
  • jit/JITStubs.cpp:
  • wtf/MathExtras.h:
  • wtf/unicode/qt4/UnicodeQt4.h:
08:02 Changeset [58166] by yurys@chromium.org

2010-04-23 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Web Inspector: provide JSC implementation for scopeType method on
call frame and use same jsvascript code for JSC and v8 when collecting
scope chain data.

https://bugs.webkit.org/show_bug.cgi?id=37663

  • bindings/js/JSInjectedScriptHostCustom.cpp: (WebCore::JSInjectedScriptHost::currentCallFrame):
  • bindings/js/JSJavaScriptCallFrameCustom.cpp: (WebCore::JSJavaScriptCallFrame::scopeType):
  • bindings/v8/custom/V8InjectedScriptHostCustom.cpp: (WebCore::V8InjectedScriptHost::currentCallFrameCallback):
  • inspector/InjectedScriptHost.idl:
  • inspector/front-end/InjectedScript.js: (injectedScriptConstructor.):
07:59 Changeset [58165] by eric@webkit.org

2010-04-23 Patrick Gansterer <paroga@paroga.com>

Reviewed by Laszlo Gombos.

[WIN] Move OwnPtrWin.cpp into platform specific folder.
https://bugs.webkit.org/show_bug.cgi?id=38042

07:45 Changeset [58164] by eric@webkit.org

2010-04-23 Patrick Gansterer <paroga@paroga.com>

Reviewed by Laszlo Gombos.

[BREWMP] Move OwnPtrBrew.cpp into platform specific folder.
https://bugs.webkit.org/show_bug.cgi?id=38042

  • wtf/OwnPtrBrew.cpp: Removed.
  • wtf/brew/OwnPtrBrew.cpp: Copied from JavaScriptCore/wtf/OwnPtrBrew.cpp.
06:05 Changeset [58163] by eric@webkit.org

2010-04-23 Alexander Pavlov <apavlov@chromium.org>

Reviewed by Pavel Feldman.

Web Inspector: Audits (Image Dimensions): full image URLs are prefixed with the hosting page name
https://bugs.webkit.org/show_bug.cgi?id=37988

  • inspector/front-end/AuditRules.js: (WebInspector.AuditRules.ImageDimensionsRule.prototype.doRun):
02:51 Changeset [58162] by eric@webkit.org

2010-04-23 Ilya Tikhonovsky <loislo@chromium.org>

Reviewed by Yury Semikhatsky.

WebInspector: Flaky Inspector tests.
https://bugs.webkit.org/show_bug.cgi?id=36217

  • inspector/InspectorController.cpp: (WebCore::InspectorController::didCommitLoad):
02:42 Changeset [58161] by eric@webkit.org

2010-04-23 Yaar Schnitman <yaar@chromium.org>

Reviewed by Adam Barth.

Auto-generate WebGLRenderingContext overloads in V8
https://bugs.webkit.org/show_bug.cgi?id=37818

  • fast/canvas/webgl/script-tests/texImageTest.js: Test was wrong: Passing 0 value for a WebGLArray argument in texImage2D should throw the same exception as when passed to textSubImage2D.
  • platform/chromium-mac/fast/canvas/webgl/texImageTest-expected.txt:
  • fast/canvas/webgl/texImageTest-expected.txt: Changes to reflect test.

2010-04-23 Yaar Schnitman <yaar@chromium.org>

Reviewed by Adam Barth.

Auto-generate WebGLRenderingContext overloads in V8
https://bugs.webkit.org/show_bug.cgi?id=37818

  • bindings/v8/custom/V8WebGLRenderingContextCustom.cpp:
  • html/canvas/WebGLRenderingContext.cpp: Added missing overloads for the following:. (WebCore::WebGLRenderingContext::texImage2D) (WebCore::WebGLRenderingContext::texSubImage2D)
  • html/canvas/WebGLRenderingContext.h: Added missing overloads.
  • html/canvas/WebGLRenderingContext.idl: IDL definition of overloads.
02:31 Changeset [58160] by eric@webkit.org

2010-04-23 Jeff Schiller <codedread@gmail.com>

Reviewed by Nikolas Zimmermann.

Display tooltips when hovering over SVG elements, Bug 16854
https://bugs.webkit.org/show_bug.cgi?id=16854

Manual test added for verifying tooltips.

  • manual-tests/svg-tooltip.svg: Added.
  • svg/SVGAElement.cpp: (WebCore::SVGAElement::title): xlink:title takes precedence, otherwise SVGStyledElement::title() is used
  • svg/SVGStyledElement.cpp: (WebCore::SVGStyledElement::title): checks for a shadow parent and uses that title, otherwise uses the content's title
  • svg/SVGStyledElement.h: add title() method declaration
01:43 Changeset [58159] by yuzo@google.com

2010-04-23 Yuzo Fujishima <yuzo@google.com>

Unreviewed, Chromium test expectation change.

Change expectation of fast/url/file-http-base.html for Chromium to TEXT.
https://bugs.webkit.org/show_bug.cgi?id=38038

  • platform/chromium/test_expectations.txt:
01:23 Changeset [58158] by ddkilzer@apple.com

<http://webkit.org/b/38032> No need to content sniff 304 responses
<rdar://problem/7891726>

Reviewed by Brady Eidson.

  • platform/network/mac/ResourceHandleMac.mm:

(-[WebCoreResourceHandleAsDelegate connection:didReceiveResponse:]):
No need to adjust the MIME type on 304 responses since they're
only used to determine if the resource needs to be refetched.

01:05 Changeset [58157] by ddkilzer@apple.com

<http://webkit.org/b/38030> Add WebCore::dumpTextEncodingNameMap() to dump text encoding map on debug builds

Reviewed by Alexey Proskuryakov.

  • platform/text/TextEncodingRegistry.cpp:

(WebCore::dumpTextEncodingNameMap): Added.

  • platform/text/TextEncodingRegistry.h:

(WebCore::dumpTextEncodingNameMap): Added declaration.

00:36 Changeset [58156] by ddkilzer@apple.com

<http://webkit.org/b/38029> +[WebTextIteratorPrivate initialize] is missing call to JSC::initializeThreading()

Reviewed by Timothy Hatcher.

Every other Objective-C class that calls
WebCoreObjCFinalizeOnMainThread(self) in +initialize also calls
JSC::initializeThreading(). The WebTextIteratorPrivate class
was the only one missing this call.

  • WebView/WebTextIterator.mm:

(+[WebTextIteratorPrivate initialize]): Added call to
JSC::initializeThreading().

00:18 Changeset [58155] by aestes@apple.com

Touch WebKit.idl to force a rebuild.

00:14 Changeset [58154] by aestes@apple.com

Roll out http://trac.webkit.org/changeset/55385.

Rubber stamped by Steve Falkenburg.

<rdar://problem/7884444>

  • Interfaces/IWebUIDelegatePrivate.idl:
  • Interfaces/WebKit.idl:
  • WebCoreSupport/WebFrameLoaderClient.cpp:

(WebFrameLoaderClient::createPlugin):

00:04 Changeset [58153] by tony@chromium.org

2010-04-22 Tony Chang <tony@chromium.org>

Reviewed by Dan Bernstein.

Crash in WebCore::TextIterator::handleTextNode() encountered in Google rich-text products
https://bugs.webkit.org/show_bug.cgi?id=37950

  • editing/text-iterator/rtl-selection-crash-expected.txt: Added.
  • editing/text-iterator/rtl-selection-crash.html: Added.

2010-04-22 Tony Chang <tony@chromium.org>

Reviewed by Dan Bernstein.

Crash in WebCore::TextIterator::handleTextNode() encountered in Google rich-text products
https://bugs.webkit.org/show_bug.cgi?id=37950

Test: editing/text-iterator/rtl-selection-crash.html

  • editing/TextIterator.cpp: (WebCore::TextIterator::handleTextNode):
00:00 Changeset [58152] by mjs@apple.com

2010-04-22 Maciej Stachowiak <mjs@apple.com>

Reviewed by Adam Barth.

Add a test for file URL canonicalization with an http base
https://bugs.webkit.org/show_bug.cgi?id=38035

  • fast/url/file-http-base-expected.txt: Added.
  • fast/url/file-http-base.html: Added.
  • fast/url/script-tests/file-http-base.js: Added.

04/22/10:

22:59 Changeset [58151] by mjs@apple.com

Fix up my immediately previous ChangeLogs.

22:47 Changeset [58150] by mjs@apple.com

2010-04-22 Maciej Stachowiak <mjs@apple.com>

Reviewed by Dan Bernstein.

Links around blocks (e.g. divs) results in too many VoiceOver call outs
https://bugs.webkit.org/show_bug.cgi?id=37079

The basic change is to modify the AccessibilityRenderObject tree
traversal methods to account for inline continuations in the
render tree and make the accessibility tree look as if
continuations didn't exist - the same as if CSS blocks could just
sit in CSS inlines. This is slightly tricky code but creates a
much saner accessibility tree.


Tests: accessibility/image-link-inline-cont.html

accessibility/image-link.html
accessibility/inline-continuations.html

  • accessibility/AccessibilityRenderObject.cpp: (WebCore::isInlineWithContinuation): Helper function for traversal functions to use in accounting for continuations. (WebCore::firstChildInContinuation): ditto (WebCore::firstChildConsideringContinuation): ditto (WebCore::lastChildConsideringContinuation): ditto (WebCore::startOfContinuations): ditto (WebCore::endOfContinuations): ditto (WebCore::childBeforeConsideringContinuations): ditto (WebCore::firstChildIsInlineContinuation): ditto (WebCore::lastChildHasContinuation): ditto (WebCore::AccessibilityRenderObject::firstChild): Account for inline continuations. (WebCore::AccessibilityRenderObject::lastChild): ditto (WebCore::AccessibilityRenderObject::previousSibling): Account for inline continuations and their anonymous block parents. (WebCore::AccessibilityRenderObject::nextSibling): ditto (WebCore::AccessibilityRenderObject::parentObjectIfExists): Account for inline continuations. (WebCore::AccessibilityRenderObject::parentObject): Account for inline continuations.
  • rendering/RenderInline.h: Make RenderInline::inlineContinuation public.

2010-04-22 Maciej Stachowiak <mjs@apple.com>

Reviewed by Dan Bernstein.

Links around blocks (e.g. divs) results in too many VoiceOver call outs
https://bugs.webkit.org/show_bug.cgi?id=37079


The new test cases verify the accessibility tree created by an image inside a link, and verify
that adding a div with role=presentation now has no effect on the accessibility tree (as expected).

  • accessibility/image-link-inline-cont-expected.txt: Added.
  • accessibility/image-link-inline-cont.html: Added.
  • accessibility/image-link.html: Added.
  • platform/gtk/Skipped:
  • platform/mac/accessibility/image-link-expected.txt: Added.
  • platform/win/Skipped:


Test to check that accessibility tree doesn't get duplicate content in the presence
of inline continuations (this was a bug in an earlier version of this patch).

  • accessibility/inline-continuations-expected.txt: Added.
  • accessibility/inline-continuations.html: Added.
21:51 Changeset [58149] by hamaji@chromium.org

2010-04-22 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by Darin Adler.

REGRESSION: TextIterator may use freed memory
https://bugs.webkit.org/show_bug.cgi?id=37973

  • platform/chromium/test_expectations.txt:

2010-04-22 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by Darin Adler.

REGRESSION: TextIterator may use freed memory
https://bugs.webkit.org/show_bug.cgi?id=37973

Added TextIterator::m_text to hold the returned String.

No new tests because there are no behavior change, but
copy-backslash-with-euc.html on chromium-win should test this.

  • editing/TextIterator.cpp: (WebCore::TextIterator::emitText):
  • editing/TextIterator.h:
21:48 Changeset [58148] by eric@webkit.org

2010-04-22 Adam Barth <abarth@webkit.org>

Reviewed by Darin Adler.

The new review tools shouldn't clear the comment box automatically
https://bugs.webkit.org/show_bug.cgi?id=38004

Instead, we now have a button for doing that explicitly. If people
like this tool, we can look at fancier solutions.

Also, tried to make the tool work on the edit page as well as the review page.

  • PrettyPatch/PrettyPatch.rb:
21:32 Changeset [58147] by ukai@chromium.org

Unreviewed fix.

  • Scripts/webkitpy/layout_tests/port/websocket_server.py: fix NameError: global name 'f' is not defined.
21:29 Changeset [58146] by eric@webkit.org

2010-04-22 James Robinson <jamesr@chromium.org>

Reviewed by Adam Barth.

Revert 58077 and follow-ups. It's broken.
https://bugs.webkit.org/show_bug.cgi?id=37664

  • Scripts/new-run-webkit-httpd:
  • Scripts/new-run-webkit-websocketserver:
  • Scripts/webkitpy/layout_tests/port/chromium_win.py:
  • Scripts/webkitpy/layout_tests/port/http_server.py:
  • Scripts/webkitpy/layout_tests/port/websocket_server.py:
21:20 Changeset [58145] by eric@webkit.org

2010-04-22 Michael Forney <mforney@mforney.org>

Reviewed by Adam Barth.

https://bugs.webkit.org/show_bug.cgi?id=37761

Only include JSDatabase.h and JSDatabaseCallback.h when database is
enabled. This partially fixes building with --disable-database.

  • bindings/js/JSDOMWindowCustom.cpp: Add a check for database around database-related headers.
20:39 Changeset [58144] by eric@webkit.org

2010-04-22 Ray Rischpater <Raymond.Rischpater@Nokia.com>

Reviewed by Darin Adler.

In HTMLInputElement.cpp, shouldUseInputMethod does not return true for
some text input types (TELEPHONE, NUMBER, URL, and EMAIL). Addressed
this by changing shouldUseInputMethod to use internal methods to
check that the field is a text field that isn't a password field.

No new tests.

Fixes <https://bugs.webkit.org/show_bug.cgi?id=37719>

  • html/HTMLInputElement.cpp: (WebCore::HTMLInputElement::shouldUseInputMethod):
20:30 Changeset [58143] by eric@webkit.org

2010-04-22 Yaar Schnitman <yaar@chromium.org>

Reviewed by Adam Barth.

Integrate v8 testing utility with webkit tests
https://bugs.webkit.org/show_bug.cgi?id=37731

  • bindings/v8/test/run_tests.py: Removed.

2010-04-22 Yaar Schnitman <yaar@chromium.org>

Reviewed by Adam Barth.

Integrate v8 testing utility with webkit tests
https://bugs.webkit.org/show_bug.cgi?id=37731

  • Scripts/run-bindings-tests: Added.
20:10 Changeset [58142] by eric@webkit.org

2010-04-22 Steve Block <steveblock@google.com>

Reviewed by Adam Barth.

Update Android to use isfinite, isinf, isnan and signbit from namespace std.
https://bugs.webkit.org/show_bug.cgi?id=37948

  • wtf/MathExtras.h:
19:52 Changeset [58141] by eric@webkit.org

2010-04-22 John Pavan <john.pavan@nokia.com>

Reviewed by Laszlo Gombos.

[Qt] inputMethodQuery returns coordinates in web page coordinates rather than in item coordinates.
https://bugs.webkit.org/show_bug.cgi?id=37163

QWebPage::inputMethodQuery is modified so that it
returns coordinates in the widget's coordinate system.
Tests are added for QGraphicsWebView and QWebView
to verify that this behavior is correct after the webpage
has been scrolled.

  • Api/qwebpage.cpp: (QWebPage::inputMethodQuery):
  • tests/qgraphicswebview/tst_qgraphicswebview.cpp: (tst_QGraphicsWebView::microFocusCoordinates):
  • tests/qwebview/tst_qwebview.cpp: (tst_QWebView::microFocusCoordinates):
19:07 Changeset [58140] by eric@webkit.org

2010-04-22 Zhenyao Mo <zmo@google.com>

Reviewed by Darin Fisher.

Add isGLES2Compliant to GraphicsContext3D
https://bugs.webkit.org/show_bug.cgi?id=37872

  • html/canvas/WebGLRenderingContext.cpp: Add isGLES2Compliant(). (WebCore::WebGLRenderingContext::isGLES2Compliant):
  • html/canvas/WebGLRenderingContext.h: Ditto.
  • platform/graphics/GraphicsContext3D.h: Ditto.
  • platform/graphics/mac/GraphicsContext3DMac.cpp: Ditto. (WebCore::GraphicsContext3D::isGLES2Compliant):

2010-04-22 Zhenyao Mo <zmo@google.com>

Reviewed by Darin Fisher.

Add isGLES2Compliant to GraphicsContext3D
https://bugs.webkit.org/show_bug.cgi?id=37872

  • public/WebGraphicsContext3D.h: Add isGLES2Compliant().
  • src/GraphicsContext3D.cpp: Ditto.
  • src/WebGraphicsContext3DDefaultImpl.cpp: Ditto. (WebKit::WebGraphicsContext3DDefaultImpl::isGLES2Compliant):
  • src/WebGraphicsContext3DDefaultImpl.h: Ditto.
18:55 Changeset [58139] by eric@webkit.org

2010-04-22 Eric Seidel <eric@webkit.org>

Reviewed by Dimitri Glazkov.

new-run-webkit-tests --verbose shows ever-increasing #EOF lines
https://bugs.webkit.org/show_bug.cgi?id=37794

  • Scripts/webkitpy/layout_tests/port/webkit.py:
    • Remove the assert() since our ServerProcess code does not always seem to be reading the full stderr output (or we're not waiting for it to).
18:32 Changeset [58138] by ukai@chromium.org

Unreviewed build fix.

  • wtf/MD5.cpp: fix include path of CString.h
18:31 Changeset [58137] by yuzo@google.com

2010-04-22 Yuzo Fujishima <yuzo@google.com>

Unreviewed. Chromium test expectation change.

Mark fast/events/touch/send-oncancel-event.html as failing (TEXT).
https://bugs.webkit.org/show_bug.cgi?id=38020

  • platform/chromium/test_expectations.txt:
18:13 Changeset [58136] by ukai@chromium.org

2010-04-22 Fumitoshi Ukai <ukai@chromium.org>

Reviewed by Adam Barth.

MD5 is required for WebSocket new protocol implementation
https://bugs.webkit.org/show_bug.cgi?id=37913

2010-04-22 Fumitoshi Ukai <ukai@chromium.org>

Reviewed by Adam Barth.

MD5 is required for WebSocket new protocol implementation
https://bugs.webkit.org/show_bug.cgi?id=37913

  • ForwardingHeaders/wtf/MD5.h: Added.
17:44 Changeset [58135] by andersca@apple.com

Remove an assert. (It's not valid when resizing).

Reviewed by Sam Weinig.

  • WebProcess/WebPage/mac/DrawingAreaUpdateChunk.cpp:

(WebKit::DrawingAreaUpdateChunk::didUpdate):

17:16 Changeset [58134] by barraclough@apple.com

Errk, accidentally committed commenting out two ASSERTs! reverting.

Reviewed by NOBODY (build fix).

  • runtime/Collector.cpp:

(JSC::Heap::protect):
(JSC::Heap::unprotect):

17:11 Changeset [58133] by barraclough@apple.com

https://bugs.webkit.org/show_bug.cgi?id=38006
Change lifetime of JSC::IdentifierTables used by WebCores to match AtomicStringTable

Reviewed by Geoff Garen.

Presently JSC's IdentifierTables are owned by the JSGlobalData. For
JSGlobalData objects created via the API this should continue to be the case,
but for the JSGlobalData objects used by WebCore (the main thread's common
global data, and those for workers) use a IdentifierTable provided (and owned)
by wtfThreadData. This allow the lifetime of these IdentifierTable to match
those of the corresponding AtomicStringTables.

  • API/APIShims.h:

(JSC::APIEntryShim::APIEntryShim):

  • API/JSContextRef.cpp:

(JSContextGroupCreate):

  • runtime/Collector.cpp:

(JSC::Heap::protect):
(JSC::Heap::unprotect):
(JSC::Heap::markRoots):

  • runtime/JSGlobalData.cpp:

(JSC::JSGlobalData::JSGlobalData):
(JSC::JSGlobalData::~JSGlobalData):
(JSC::JSGlobalData::createContextGroup):
(JSC::JSGlobalData::create):
(JSC::JSGlobalData::sharedInstance):

  • runtime/JSGlobalData.h:

(JSC::JSGlobalData::):
(JSC::JSGlobalData::isSharedInstance):

  • runtime/JSLock.cpp:

(JSC::JSLock::JSLock):
(JSC::JSLock::lock):
(JSC::JSLock::unlock):
(JSC::JSLock::DropAllLocks::DropAllLocks):

  • wtf/WTFThreadData.cpp:

(WTF::WTFThreadData::WTFThreadData):
(WTF::WTFThreadData::~WTFThreadData):

17:08 Changeset [58132] by ap@apple.com

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=38012
REGRESSION: Ctrl- and Option- shortcuts get wrong keyCode on non-QWERTY keyboard

  • platform/mac/KeyEventMac.mm: (WebCore::windowsKeyCodeForKeyEvent): Also look at unmodified characters, to avoid falling through to virtual key code lookup for Roman characters.
17:07 Changeset [58131] by barraclough@apple.com

Do not get the globaldata for the APIShim via the
global object; global object construction creates identifiers
which results in the ASSERTs checking the thread data's
identifierTable has been set failing.

Rubber stamped by Sam Weinig

  • JSUtils.cpp:

(unprotectGlobalObject):
(getThreadGlobalData):
(getThreadGlobalObject):
(JSGlueAPIEntry::JSGlueAPIEntry):
(JSGlueAPICallback::~JSGlueAPICallback):

17:02 Changeset [58130] by mrowe@apple.com

Versioning.

17:00 Changeset [58129] by mrowe@apple.com

New tag.

16:46 Changeset [58128] by eric@webkit.org

2010-04-22 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Add code to help debug new-run-webkit-test hangs on the Chromium bots
https://bugs.webkit.org/show_bug.cgi?id=38011

I can see no reasonable way to test this change.
Stubbing out sys._current_frames() and traceback.extract_stack
seems folly. Dumping real data would have line number
(and possibly other call stack) variance between runs.

  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
    • Add _dump_thread_states and _dump_thread_states_if_necessary to have our main thread dump the states of all threads every 60 seconds when running in verbose mode.
    • Better document what is going on in our main loop.
16:40 Changeset [58127] by jamesr@google.com

2010-04-22 James Robinson <jamesr@chromium.org>

Unreviewed, expectations only change.
Land chromium platform-specific results for fast/text/format-control.html

  • platform/chromium-linux/fast/text/format-control-expected.checksum: Added.
  • platform/chromium-linux/fast/text/format-control-expected.png: Added.
  • platform/chromium-mac/fast/text/format-control-expected.checksum: Added.
  • platform/chromium-win/fast/text/format-control-expected.checksum: Added.
  • platform/chromium-win/fast/text/format-control-expected.png: Added.
  • platform/chromium-win/fast/text/format-control-expected.txt: Added.
16:37 Changeset [58126] by eric@webkit.org

2010-04-22 Jocelyn Turcotte <jocelyn.turcotte@nokia.com>

Reviewed by Darin Adler.

Fix the MSVC 64bit build.
https://bugs.webkit.org/show_bug.cgi?id=37980

  • platform/text/TextStream.cpp:
  • platform/text/TextStream.h:
  • plugins/win/PluginViewWin.cpp: (WebCore::PluginView::platformStart):
16:35 Changeset [58125] by weinig@apple.com

Merge the prefix headers.

Reviewed by Anders Carlsson.

  • Configurations/BaseTarget.xcconfig:
  • WebKit2.xcodeproj/project.pbxproj:
  • WebKit2Prefix.h:
  • WebKit2_Prefix.pch: Removed.
16:00 Changeset [58124] by eric@webkit.org

2010-04-22 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Gustavo Noronha Silva.

Implement findThemePath to locate the correct theme file instead of
relying on a hardcoded location.
http://webkit.org/b/37996

  • efl/EWebLauncher/main.c: (findThemePath): Fix to locate the correct theme files. (main): Use findThemePath() instead of a hardcoded one.
15:58 Changeset [58123] by andersca@apple.com

Don't pass the new size when calling didSetSize, it's possible to get
the size from the update chunk.

Reviewed by Sam Weinig.

  • UIProcess/mac/DrawingAreaProxyUpdateChunk.h:
  • UIProcess/mac/DrawingAreaProxyUpdateChunk.mm:

(WebKit::DrawingAreaProxyUpdateChunk::didSetSize):
(WebKit::DrawingAreaProxyUpdateChunk::didReceiveMessage):

  • WebProcess/WebPage/mac/DrawingAreaUpdateChunk.cpp:

(WebKit::DrawingAreaUpdateChunk::setSize):

15:51 Changeset [58122] by eric@webkit.org

2010-04-22 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Gustavo Noronha Silva.

Update efl/ewebkit.pc.in to match the variable substitution syntax
of CMake.
http://webkit.org/b/37999

  • efl/ewebkit.pc.in: Updated.
15:41 Changeset [58121] by eric@webkit.org

2010-04-22 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Gustavo Noronha Silva.

Update EWebKit (EFL port) to match recent API changes.
http://webkit.org/b/37997

  • efl/ewk/ewk_frame.cpp: (ewk_frame_zoom_set): Change to use WebCore::ZoomMode. (ewk_frame_zoom_text_only_set): Change to use WebCore::ZoomMode. (_ewk_frame_handle_key_scrolling): s/WebCore::VK_/VK_/g (ewk_frame_plugin_create): Disable temporarily PluginView-related code until a proper plugin implementation is made.
  • efl/ewk/ewk_view.cpp: Fix typo in _parent_sc declaration. (ewk_view_selection_get): Fix a reference to WebCore::CString to WTF::CString.
15:33 Changeset [58120] by ap@apple.com

Rubber-stamped by Mark Rowe.

<rdar://problem/7805969> REGRESSION: iTunes unable to play trailers

Undo the changes made for https://bugs.webkit.org/show_bug.cgi?id=35215 (<rdar://problem/7673157>)
for now. Clients rely on the old behavior, so a fix that changes it will need to account for
those.

  • WebCoreSupport/WebFrameLoaderClient.mm: (WebFrameLoaderClient::createPlugin): Pass pluginURL instead of baseURL again.
15:09 Changeset [58119] by andersca@apple.com

Rename SetFrame and DidSetFrame to SetSize and DidSetSize.

Reviewed by Sam Weinig.

  • Shared/CoreIPCSupport/DrawingAreaMessageKinds.h:

(DrawingAreaMessage::):

  • Shared/CoreIPCSupport/DrawingAreaProxyMessageKinds.h:

(DrawingAreaProxyMessage::):

  • UIProcess/mac/DrawingAreaProxyUpdateChunk.mm:

(WebKit::DrawingAreaProxyUpdateChunk::drawRectIntoContext):
(WebKit::DrawingAreaProxyUpdateChunk::setSize):
(WebKit::DrawingAreaProxyUpdateChunk::didReceiveMessage):

  • WebProcess/WebPage/mac/DrawingAreaUpdateChunk.cpp:

(WebKit::DrawingAreaUpdateChunk::setSize):
(WebKit::DrawingAreaUpdateChunk::didReceiveMessage):

14:44 Changeset [58118] by eric@webkit.org

2010-04-22 Sheriff Bot <webkit.review.bot@gmail.com>

Unreviewed, rolling out r58110.
http://trac.webkit.org/changeset/58110
https://bugs.webkit.org/show_bug.cgi?id=38007

Caused an internal compiler error on Qt (Requested by abarth
on #webkit).

  • wtf/MathExtras.h:
13:56 Changeset [58117] by weinig@apple.com

Fix for https://bugs.webkit.org/show_bug.cgi?id=38002
Add rudimentary statistics gathering for WebKit2

Reviewed by Anders Carlsson.

WebKit2:

  • UIProcess/API/C/WKContext.cpp:

(WKContextGetStatistics):

  • UIProcess/API/C/WKContextPrivate.h: Copied from WebKit2/UIProcess/API/C/WKContext.h.
  • UIProcess/API/C/WKPageNamespace.cpp:

(WKPageNamespaceGetContext):

  • UIProcess/API/C/WKPageNamespace.h:
  • UIProcess/WebContext.cpp:

(WebKit::WebContext::getStatistics):

  • UIProcess/WebContext.h:
  • UIProcess/WebPageNamespace.cpp:

(WebKit::WebPageNamespace::getStatistics):

  • UIProcess/WebPageNamespace.h:
  • UIProcess/WebPageProxy.cpp:

(WebKit::WebPageProxy::getStatistics):

  • UIProcess/WebPageProxy.h:
  • UIProcess/WebProcessProxy.cpp:

(WebKit::WebProcessProxy::numberOfPages):

  • UIProcess/WebProcessProxy.h:
  • WebKit2.xcodeproj/project.pbxproj:
  • win/WebKit2.vcproj:

WebKitTools:

  • MiniBrowser/MiniBrowser.xcodeproj/project.pbxproj:
  • MiniBrowser/mac/AppDelegate.h:
  • MiniBrowser/mac/AppDelegate.m:

(-[BrowserAppDelegate init]):
(-[BrowserAppDelegate newWindow:]):
(-[BrowserAppDelegate getCurrentPageNamespace]):
(-[BrowserAppDelegate _setProcessModel:]):
(-[BrowserAppDelegate showStatisticsWindow:]):
(-[BrowserAppDelegate applicationWillTerminate:]):

  • MiniBrowser/mac/BrowserStatisticsWindow.xib: Added.
  • MiniBrowser/mac/BrowserStatisticsWindowController.h: Added.
  • MiniBrowser/mac/BrowserStatisticsWindowController.m: Added.

(-[BrowserStatisticsWindowController initWithThreadedWKContextRef:processWKContextRef:]):
(-[BrowserStatisticsWindowController dealloc]):
(-[BrowserStatisticsWindowController windowDidLoad]):
(-[BrowserStatisticsWindowController refreshStatistics:]):

  • MiniBrowser/mac/BrowserWindowController.h:
  • MiniBrowser/mac/English.lproj/MainMenu.xib:
  • MiniBrowser/mac/MiniBrowser_Prefix.pch:
13:50 Changeset [58116] by barraclough@apple.com
  • API/JSWeakObjectMapRefPrivate.cpp:

Reviewed by NOBODY (windows build fix).

13:42 Changeset [58115] by barraclough@apple.com
  • API/JSBase.cpp:
  • API/JSCallbackObject.cpp:

Reviewed by NOBODY (windows build fix).

13:31 Changeset [58114] by barraclough@apple.com

https://bugs.webkit.org/show_bug.cgi?id=37978
Unify JSC::IdentifierTable and WebCore::AtomicStringTable implementations.

Reviewed by Geoff Garen.

These two classes both implement a HashSet of uniqued StringImpls, with
translator classes to avoid unnecessary object creation. The only difference
between the classes is which flag (isIdentifier or inTable) is set.
Combine the two classes using a template predicated on which flag to use.

New class AtomicStringTable created, containing all the goodness from
IdentifierTable & AtomicStringTable, expect for Identifier's literalTable,
which has been moved onto JSGlobalData. Removed duplicate string translator
classes. Renamed StringImpl's inTable flag to more explicit 'isAtomic',
and set this on the empty string (which matches Identifier behaviour, and
removes a redundant check for zero-length).

(JSC::createLiteralTable):
(JSC::deleteLiteralTable):
(JSC::Identifier::add):
(JSC::Identifier::addSlowCase):

  • runtime/Identifier.h:
  • runtime/JSGlobalData.cpp:

(JSC::JSGlobalData::JSGlobalData):
(JSC::JSGlobalData::~JSGlobalData):

  • runtime/JSGlobalData.h:
  • wtf/WTFThreadData.cpp:

(WTF::WTFThreadData::WTFThreadData):
(WTF::WTFThreadData::~WTFThreadData):

  • wtf/WTFThreadData.h:

(WTF::WTFThreadData::atomicStringTable):

  • wtf/text/AtomicString.cpp:

(WebCore::table):
(WebCore::operator==):
(WebCore::AtomicString::add):
(WebCore::AtomicString::find):
(WebCore::AtomicString::remove):

  • wtf/text/AtomicStringTable.h: Added.

(WTF::CStringTranslator::hash):
(WTF::CStringTranslator::equal):
(WTF::CStringTranslator::translate):
(WTF::UCharBufferTranslator::hash):
(WTF::UCharBufferTranslator::equal):
(WTF::UCharBufferTranslator::translate):
(WTF::HashAndCharactersTranslator::hash):
(WTF::HashAndCharactersTranslator::equal):
(WTF::HashAndCharactersTranslator::translate):
(WTF::IdentifierOrAtomicStringTable::remove):
(WTF::::~IdentifierOrAtomicStringTable):
(WTF::::add):
(WTF::::find):

  • wtf/text/StringImpl.cpp:

(WebCore::StringImpl::~StringImpl):

  • wtf/text/StringImpl.h:

(WebCore::StringImpl::isAtomic):
(WebCore::StringImpl::setIsAtomic):
(WebCore::equal):

  • wtf/text/StringImplBase.h:

(WTF::StringImplBase::StringImplBase):

13:17 Changeset [58113] by morrita@google.com

2010-04-22 MORITA Hajime <morrita@google.com>

Not reviewed.

Picked a review feedback left on:
https://bugs.webkit.org/show_bug.cgi?id=37056
LayoutTests/editing/selection/move-left-right.html is slow


  • editing/selection/move-left-right.html:
13:12 Changeset [58112] by robert@webkit.org

[Qt] Fix createPlugin() tests in tst_qwebpage to match behaviour of Qt plugins

when PluginsEnabled is false.

2010-04-22 Robert Hogan <robert@webkit.org>

Reviewed by Simon Hausmann.

[Qt] Fix createPlugin() tests in tst_qwebpage to match behaviour of Qt plugins

when PluginsEnabled is false.

tst_qwebpage should have been updated as part of r56662.
(See https://bugs.webkit.org/show_bug.cgi?id=32196)

Updated documentation of QWebPage::createPlugin and QWebSetting::pluginsEnabled

to match the new behaviour.

  • Api/qwebpage.cpp: Update docs.
  • Api/qwebsettings.cpp: Update docs.
  • tests/qwebpage/tst_qwebpage.cpp: (createPlugin): (tst_QWebPage::createPluginWithPluginsEnabled): (tst_QWebPage::createPluginWithPluginsDisabled):
13:03 Changeset [58111] by dglazkov@chromium.org

2010-04-22 Dave Moore <davemoore@chromium.org>

Reviewed by Dimitri Glazkov.

Add test for icon change notifications.

https://bugs.webkit.org/show_bug.cgi?id=33812

  • fast/dom/icon-url-property-expected.txt: Added.
  • fast/dom/icon-url-property.html: Added.
  • Skipped on all platforms but Win

2010-04-22 Dave Moore <davemoore@chromium.org>

Reviewed by Dimitri Glazkov.

Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

https://bugs.webkit.org/show_bug.cgi?id=33812

Test: fast/dom/icon-url-property.html

  • dom/Document.cpp: (WebCore::Document::setIconURL):
  • loader/DocumentLoader.cpp: (WebCore::DocumentLoader::setIconURL):
  • loader/DocumentLoader.h: (WebCore::DocumentLoader::iconURL):
  • loader/EmptyClients.h: (WebCore::EmptyFrameLoaderClient::dispatchDidChangeIcons):
  • loader/FrameLoader.cpp: (WebCore::FrameLoader::setIconURL): (WebCore::FrameLoader::didChangeIcons):
  • loader/FrameLoader.h:
  • loader/FrameLoaderClient.h:

2010-04-22 Dave Moore <davemoore@chromium.org>

Reviewed by Dimitri Glazkov.

Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

https://bugs.webkit.org/show_bug.cgi?id=33812

  • public/WebFrameClient.h: (WebKit::WebFrameClient::didChangeIcons):
  • src/FrameLoaderClientImpl.cpp: (WebKit::FrameLoaderClientImpl::dispatchDidChangeIcons):
  • src/FrameLoaderClientImpl.h:

2010-04-22 Dave Moore <davemoore@chromium.org>

Reviewed by Dimitri Glazkov.

Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

  • WebCoreSupport/FrameLoaderClientGtk.cpp: (WebKit::FrameLoaderClient::dispatchDidChangeIcons):
  • WebCoreSupport/FrameLoaderClientGtk.h:

2010-04-22 Dave Moore <davemoore@chromium.org>

Reviewed by Dimitri Glazkov.

Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

https://bugs.webkit.org/show_bug.cgi?id=33812

  • WebCoreSupport/WebFrameLoaderClient.h:
  • WebCoreSupport/WebFrameLoaderClient.mm: (WebFrameLoaderClient::dispatchDidChangeIcons):

2010-04-22 Dave Moore <davemoore@chromium.org>

Reviewed by Dimitri Glazkov.

Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

https://bugs.webkit.org/show_bug.cgi?id=33812

  • WebCoreSupport/FrameLoaderClientQt.cpp: (WebCore::FrameLoaderClientQt::dispatchDidChangeIcons): (WebCore::FrameLoaderClientQt::didChangeTitle):
  • WebCoreSupport/FrameLoaderClientQt.h:

2010-04-22 Dave Moore <davemoore@chromium.org>

Reviewed by Dimitri Glazkov.

Added notification when the favicons for a page are changed
from a script.
The Document object will notify the frame loader, which will
notify the client. Implementations of FrameLoaderClient will
have to add one method; dispatchDidChangeIcons().

https://bugs.webkit.org/show_bug.cgi?id=33812

  • Interfaces/IWebFrameLoadDelegatePrivate2.idl:
  • WebCoreSupport/WebFrameLoaderClient.cpp: (WebFrameLoaderClient::dispatchDidChangeIcons):
  • WebCoreSupport/WebFrameLoaderClient.h:
  • WebFrame.cpp: (WebFrame::didChangeIcons):
  • WebFrame.h:

2010-04-22 Dave Moore <davemoore@chromium.org>

Reviewed by Dimitri Glazkov.

Add test support for icon changes.

https://bugs.webkit.org/show_bug.cgi?id=33812

  • DumpRenderTree/LayoutTestController.cpp: (LayoutTestController::LayoutTestController): (dumpIconChangesCallback): (LayoutTestController::staticFunctions):
  • DumpRenderTree/LayoutTestController.h: (LayoutTestController::dumpIconChanges): (LayoutTestController::setDumpIconChanges):
  • DumpRenderTree/win/FrameLoadDelegate.cpp: (FrameLoadDelegate::didChangeIcons):
  • DumpRenderTree/win/FrameLoadDelegate.h:
  • WinLauncher/WinLauncher.h: (WinLauncherWebHost::didChangeIcons):
12:55 Changeset [58110] by steveblock@google.com

2010-04-22 Steve Block <steveblock@google.com>

Reviewed by Adam Barth.

Update Android to use isfinite, isinf, isnan and signbit from namespace std.
https://bugs.webkit.org/show_bug.cgi?id=37948

  • wtf/MathExtras.h:
12:46 Changeset [58109] by eric.carlson@apple.com

2010-04-22 Eric Carlson <eric.carlson@apple.com>

Test fix, no review.

https://bugs.webkit.org/show_bug.cgi?id=38000
r58107 causes video-play-stall and video-play-stall-seek to fail

Skip http/tests/media/video-play-stall.html and http/tests/media/video-play-stall-seek.html
until we can figure out how to work around QuickTime bug.

  • platform/mac-leopard/Skipped:
  • platform/mac-tiger/Skipped:
12:39 Changeset [58108] by xan@webkit.org

WebCore:

2010-04-22 Xan Lopez <xlopez@igalia.com>

Reviewed by Gustavo Noronha.

[GTK] GObject DOM bindings
https://bugs.webkit.org/show_bug.cgi?id=33590

Extend the bindings to cover most of Document.idl

  • GNUmakefile.am:
  • bindings/gobject/WebKitDOMBinding.cpp: (WebKit::createWrapper): (WebKit::kit):
  • bindings/gobject/WebKitDOMBinding.h:
  • bindings/gobject/WebKitHTMLElementWrapperFactory.cpp: Added. (WebKit::createAnchorWrapper): (WebKit::createAppletWrapper): (WebKit::createAreaWrapper): (WebKit::createBaseWrapper): (WebKit::createBaseFontWrapper): (WebKit::createBlockquoteWrapper): (WebKit::createBodyWrapper): (WebKit::createBRWrapper): (WebKit::createButtonWrapper): (WebKit::createCanvasWrapper): (WebKit::createTableCaptionWrapper): (WebKit::createTableColWrapper): (WebKit::createModWrapper): (WebKit::createDirectoryWrapper): (WebKit::createDivWrapper): (WebKit::createDListWrapper): (WebKit::createEmbedWrapper): (WebKit::createFieldSetWrapper): (WebKit::createFontWrapper): (WebKit::createFormWrapper): (WebKit::createFrameWrapper): (WebKit::createFrameSetWrapper): (WebKit::createHeadingWrapper): (WebKit::createHeadWrapper): (WebKit::createHRWrapper): (WebKit::createHtmlWrapper): (WebKit::createIFrameWrapper): (WebKit::createImageWrapper): (WebKit::createInputWrapper): (WebKit::createIsIndexWrapper): (WebKit::createLabelWrapper): (WebKit::createLegendWrapper): (WebKit::createLIWrapper): (WebKit::createLinkWrapper): (WebKit::createMapWrapper): (WebKit::createMarqueeWrapper): (WebKit::createMenuWrapper): (WebKit::createMetaWrapper): (WebKit::createObjectWrapper): (WebKit::createOListWrapper): (WebKit::createOptGroupWrapper): (WebKit::createOptionWrapper): (WebKit::createParagraphWrapper): (WebKit::createParamWrapper): (WebKit::createPreWrapper): (WebKit::createQuoteWrapper): (WebKit::createScriptWrapper): (WebKit::createSelectWrapper): (WebKit::createStyleWrapper): (WebKit::createTableWrapper): (WebKit::createTableSectionWrapper): (WebKit::createTableCellWrapper): (WebKit::createTextAreaWrapper): (WebKit::createTitleWrapper): (WebKit::createTableRowWrapper): (WebKit::createUListWrapper): (WebKit::createHTMLElementWrapper):
  • bindings/gobject/WebKitHTMLElementWrapperFactory.h: Added.
  • bindings/scripts/CodeGeneratorGObject.pm:
  • dom/Node.idl:

WebKit/gtk:

2010-04-22 Xan Lopez <xlopez@igalia.com>

Reviewed by Gustavo Noronha.

[GTK] GObject DOM bindings
https://bugs.webkit.org/show_bug.cgi?id=33590

Add webkit_web_view_get_document to retrieve the WebKitDOMDocument
from a WebKitWebView, and a very simple test for it.

  • tests/testwebview.c:
  • webkit/webkitwebview.cpp: (webkit_web_view_get_document):
  • webkit/webkitwebview.h:
11:06 Changeset [58107] by eric.carlson@apple.com

2010-04-22 Eric Carlson <eric.carlson@apple.com>

Reviewed by Simon Fraser.

Do not pause movie when readyState drops below HAVE_FUTURE_DATA
https://bugs.webkit.org/show_bug.cgi?id=37991
<rdar://problem/7893937>

No new tests, we don't have infrastructure in DRT to test with streamed
movies.

  • html/HTMLMediaElement.cpp: (WebCore::HTMLMediaElement::HTMLMediaElement): Initialize m_readyStateMaximum. (WebCore::HTMLMediaElement::prepareForLoad): Reset m_readyStateMaximum. (WebCore::HTMLMediaElement::setReadyState): Maintain m_readyStateMaximum. (WebCore::HTMLMediaElement::potentiallyPlaying): Also return true if the readyState was previously >= HAVE_FUTURE_DATA.
  • html/HTMLMediaElement.h:
10:57 Changeset [58106] by eric@webkit.org

2010-04-22 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Gustavo Noronha Silva.

Update EFL port files to match recent API changes.
http://webkit.org/b/37876

  • efl/WebCoreSupport/ChromeClientEfl.cpp: (kit): Added. (WebCore::ChromeClientEfl::ChromeClientEfl): Coding style fix. (WebCore::ChromeClientEfl::runBeforeUnloadConfirmPanel): Coding style fix. (WebCore::ChromeClientEfl::mouseDidMoveOverElement): Coding style fix. (WebCore::ChromeClientEfl::runOpenPanel): Implemented. (WebCore::ChromeClientEfl::cancelGeolocationPermissionRequestForFrame): Stubbed. (WebCore::ChromeClientEfl::cancelGeolocationPermissionForFrame): Stubbed. (WebCore::ChromeClientEfl::invalidateContents): Stubbed. (WebCore::ChromeClientEfl::invalidateWindow): Stubbed. (WebCore::ChromeClientEfl::invalidateContentsAndWindow): Implemented. (WebCore::ChromeClientEfl::invalidateContentsForSlowScroll): Implemented. (WebCore::ChromeClientEfl::scroll): Updated. (WebCore::ChromeClientEfl::iconForFiles): Stubbed. (WebCore::ChromeClientEfl::chooseIconForFiles): Stubbed.
  • efl/WebCoreSupport/ChromeClientEfl.h: Add new method prototypes.
  • efl/WebCoreSupport/FrameLoaderClientEfl.cpp: (WebCore::FrameLoaderClientEfl::committedLoad): Call setEncoding() from FrameLoader::writer. (WebCore::FrameLoaderClientEfl::finishedLoading): Ditto. (WebCore::FrameLoaderClientEfl::dispatchDidFailLoading): Ditto. (WebCore::FrameLoaderClientEfl::setMainDocumentError): Ditto.
10:32 Changeset [58105] by eric@webkit.org

2010-04-22 Jocelyn Turcotte <jocelyn.turcotte@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Remove translatable strings from the hybridPixmap test.
https://bugs.webkit.org/show_bug.cgi?id=37867

  • tests/hybridPixmap/widget.ui:
10:13 Changeset [58104] by eric@webkit.org

2010-04-22 Jocelyn Turcotte <jocelyn.turcotte@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Remove dependency of JSC to QtGui
https://bugs.webkit.org/show_bug.cgi?id=37867

The patch also make sure that hal.h is in the include path on Symbian.
The dependency to QtGui took care of that before.

Patch by Thiago Macieira <thiago.macieira@nokia.com>

and Rohan McGovern <rohan.mcgovern@nokia.com>

10:04 Changeset [58103] by eric@webkit.org

2010-04-22 Zhenyao Mo <zmo@google.com>

Reviewed by Dimitri Glazkov.

Regression: framebufferRenderbuffer crashes with null renderBuffer
https://bugs.webkit.org/show_bug.cgi?id=37963

  • fast/canvas/webgl/framebuffer-object-attachment-expected.txt: Adding test cases for null renderbuffer input.
  • fast/canvas/webgl/framebuffer-object-attachment.html: Ditto.

2010-04-22 Zhenyao Mo <zmo@google.com>

Reviewed by Dimitri Glazkov.

Regression: framebufferRenderbuffer crashes with null renderBuffer
https://bugs.webkit.org/show_bug.cgi?id=37963

  • html/canvas/WebGLRenderingContext.cpp: (WebCore::WebGLRenderingContext::framebufferRenderbuffer): Dealing with null renderbuffer input.
09:54 Changeset [58102] by eric@webkit.org

2010-04-22 Zhenyao Mo <zmo@google.com>

Reviewed by Dimitri Glazkov.

Emulate GL_IMPLEMENTATION_COLOR_READ_FORMAT/TYPE for glGet
https://bugs.webkit.org/show_bug.cgi?id=37281

  • fast/canvas/webgl/read-pixels-expected.txt: Add test case for the other supported format/type in readPixels (besides RGBA/UNSIGNED_BYTE); also, changed to a simpler test framework.
  • fast/canvas/webgl/read-pixels.html: Ditto.

2010-04-22 Zhenyao Mo <zmo@google.com>

Reviewed by Dimitri Glazkov.

Emulate GL_IMPLEMENTATION_COLOR_READ_FORMAT/TYPE for glGet
https://bugs.webkit.org/show_bug.cgi?id=37281

  • html/canvas/WebGLRenderingContext.cpp: (WebCore::WebGLRenderingContext::WebGLRenderingContext): Remove error check because two enums are supported now. (WebCore::WebGLRenderingContext::getParameter): Add two enums. (WebCore::WebGLRenderingContext::readPixels): Fix a tiny bug.
  • platform/graphics/mac/GraphicsContext3DMac.cpp: (WebCore::GraphicsContext3D::getIntegerv): Emulate two enums.

2010-04-22 Zhenyao Mo <zmo@google.com>

Reviewed by Dimitri Glazkov.

Emulate GL_IMPLEMENTATION_COLOR_READ_FORMAT/TYPE for glGet
https://bugs.webkit.org/show_bug.cgi?id=37281

  • src/WebGraphicsContext3DDefaultImpl.cpp: (WebKit::WebGraphicsContext3DDefaultImpl::getIntegerv): Emulate two enums.
09:50 Changeset [58101] by abarth@webkit.org

2010-04-22 Adam Barth <abarth@webkit.org>

Unreviewed. Turns out we need to do more escaping because Ruby was
explanding our newlines, creating a syntax error.

  • PrettyPatch/PrettyPatch.rb:
09:38 Changeset [58100] by eric@webkit.org

2010-04-22 Jocelyn Turcotte <jocelyn.turcotte@nokia.com>

Reviewed by Tor Arne Vestbø.

[Qt] Corrects symbols visibility for JavaScriptCore.

https://bugs.webkit.org/show_bug.cgi?id=37867

09:28 Changeset [58099] by eric@webkit.org

2010-04-22 Marcus Bulach <bulach@chromium.org>

Reviewed by Dimitri Glazkov.

Enable LayoutTests for EventSource on chromium
https://bugs.webkit.org/show_bug.cgi?id=37863

  • platform/chromium/fast/dom/prototype-inheritance-expected.txt:
  • platform/chromium/test_expectations.txt:
08:46 Changeset [58098] by eric@webkit.org

2010-04-22 Diego Escalante Urrelo <descalante@igalia.com>

Reviewed by Xan Lopez.

[Gtk] Evaluate and create tests for all the AtkRole's implemented by
WebKitGtk
https://bugs.webkit.org/show_bug.cgi?id=34449

Expand testatkroles to test ATK_ROLE_COMBO_BOX.

  • tests/testatkroles.c: (test_webkit_atk_get_role_combobox): (main):
08:36 Changeset [58097] by eric@webkit.org

2010-04-22 Diego Escalante Urrelo <descalante@igalia.com>

Reviewed by Xan Lopez.

[GTK] Mute/unmute button on <video> elements are backwards
https://bugs.webkit.org/show_bug.cgi?id=33967

Fix mute/unmute buttons icons-action relation and explain why their
variable names and corresponding icons seem to be misleading but are
actually right. Original patch by Mike Hommey.

  • platform/gtk/RenderThemeGtk.cpp: (WebCore::RenderThemeGtk::initMediaStyling):
07:56 Changeset [58096] by eric@webkit.org

2010-04-22 Jocelyn Turcotte <jocelyn.turcotte@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Fix autotests .qrc file paths when built in Qt.

Compiling auto-tests from qt/tests/auto/qweb* produced
failing tests since these .pro files include the
ones in WebKit/qt/tests and the .qrc file was
not added to RESOURCES

  • tests/benchmarks/loading/loading.pro:
  • tests/benchmarks/painting/painting.pro:
  • tests/qgraphicswebview/qgraphicswebview.pro:
  • tests/qwebelement/qwebelement.pro:
  • tests/qwebframe/qwebframe.pro:
  • tests/qwebhistory/qwebhistory.pro:
  • tests/qwebhistoryinterface/qwebhistoryinterface.pro:
  • tests/qwebinspector/qwebinspector.pro:
  • tests/qwebpage/qwebpage.pro:
  • tests/qwebplugindatabase/qwebplugindatabase.pro:
  • tests/qwebview/qwebview.pro:
  • tests/tests.pri:
07:46 Changeset [58095] by eric@webkit.org

2010-04-22 Gustavo Sverzut Barbieri <barbieri@profusion.mobi>

Reviewed by Eric Seidel.

Fix build if NPAPI support is disabled
https://bugs.webkit.org/show_bug.cgi?id=36621

No new tests, this is a build fix.
Re-submit r58043 with fix for EFL.

  • plugins/PluginViewNone.cpp:
07:36 Changeset [58094] by eric@webkit.org

2010-04-22 Diego Escalante Urrelo <descalante@igalia.com>

Reviewed by Xan Lopez.

[Gtk] Evaluate and create tests for all the AtkRole's implemented by
WebKitGtk
https://bugs.webkit.org/show_bug.cgi?id=34449

Implement ATK_ROLE_COMBO_BOX.

  • accessibility/gtk/AccessibilityObjectAtk.cpp: (WebCore::AccessibilityObject::accessibilityPlatformIncludesObject):
  • accessibility/gtk/AccessibilityObjectWrapperAtk.cpp: (atkRole):
06:50 Changeset [58093] by zimmermann@webkit.org

2010-04-22 Nikolas Zimmermann <nzimmermann@rim.com>

Reviewed by Dirk Schulze.

SVGPaintServer needs to be converted to the new RenderSVGResource* system
https://bugs.webkit.org/show_bug.cgi?id=37986

No functional changes, just move the SVGPaintServer* classes from svg/graphics/ to rendering/, the new location for the RenderSVGResource* classes.
This is a preparation for the real patch which follows soon.

  • Android.mk: Rename files and move to the right location.
  • GNUmakefile.am: Ditto.
  • WebCore.gypi: Ditto.
  • WebCore.pro: Ditto.
  • WebCore.vcproj/WebCore.vcproj: Ditto.
  • WebCore.xcodeproj/project.pbxproj: Ditto.
  • rendering/RenderSVGResource.cpp: Copied from svg/graphics/SVGPaintServer.cpp.
  • rendering/RenderSVGResourceGradient.cpp: Copied from svg/graphics/SVGPaintServerGradient.cpp.
  • rendering/RenderSVGResourceGradient.h: Copied from svg/graphics/SVGPaintServerGradient.h.
  • rendering/RenderSVGResourceLinearGradient.cpp: Copied from svg/graphics/SVGPaintServerLinearGradient.cpp.
  • rendering/RenderSVGResourceLinearGradient.h: Copied from svg/graphics/SVGPaintServerLinearGradient.h.
  • rendering/RenderSVGResourcePattern.cpp: Copied from svg/graphics/SVGPaintServerPattern.cpp.
  • rendering/RenderSVGResourcePattern.h: Copied from svg/graphics/SVGPaintServerPattern.h.
  • rendering/RenderSVGResourceRadialGradient.cpp: Copied from svg/graphics/SVGPaintServerRadialGradient.cpp.
  • rendering/RenderSVGResourceRadialGradient.h: Copied from svg/graphics/SVGPaintServerRadialGradient.h.
  • rendering/RenderSVGResourceSolidColor.cpp: Copied from svg/graphics/SVGPaintServerSolid.cpp.
  • rendering/RenderSVGResourceSolidColor.h: Copied from svg/graphics/SVGPaintServerSolid.h.
  • rendering/SVGRenderTreeAsText.cpp: Change include file names.
  • svg/SVGFont.cpp: Ditto.
  • svg/SVGGradientElement.cpp: Ditto.
  • svg/SVGGradientElement.h: Ditto.
  • svg/SVGLinearGradientElement.cpp: Ditto.
  • svg/SVGPatternElement.cpp: Ditto.
  • svg/SVGPatternElement.h: Ditto.
  • svg/SVGRadialGradientElement.cpp: Ditto.
  • svg/graphics/SVGPaintServer.cpp: Removed.
  • svg/graphics/SVGPaintServerGradient.cpp: Removed.
  • svg/graphics/SVGPaintServerGradient.h: Removed.
  • svg/graphics/SVGPaintServerLinearGradient.cpp: Removed.
  • svg/graphics/SVGPaintServerLinearGradient.h: Removed.
  • svg/graphics/SVGPaintServerPattern.cpp: Removed.
  • svg/graphics/SVGPaintServerPattern.h: Removed.
  • svg/graphics/SVGPaintServerRadialGradient.cpp: Removed.
  • svg/graphics/SVGPaintServerRadialGradient.h: Removed.
  • svg/graphics/SVGPaintServerSolid.cpp: Removed.
  • svg/graphics/SVGPaintServerSolid.h: Removed.
  • svg/graphics/SVGResourceListener.h: Removed.
06:34 Changeset [58092] by eric@webkit.org

2010-04-22 Jocelyn Turcotte <jocelyn.turcotte@nokia.com>

Reviewed by Simon Hausmann.

[Qt] Hide the OpenGL and QtMultimedia dependencies from syncqt.

This prevents the dependent headers from being included by
qt/include/QtWebKit/QtWebKit

  • WebCore.pro:
06:24 Changeset [58091] by eric@webkit.org

2010-04-22 Gabor Loki <loki@webkit.org>

Reviewed by Gavin Barraclough.

Use BLX and BX to keep happy the return stack predictor above ARMv4
https://bugs.webkit.org/show_bug.cgi?id=37862

Inspired by Jacob Bramley's patch from JaegerMonkey

  • assembler/ARMAssembler.cpp: (JSC::ARMAssembler::executableCopy):
  • assembler/ARMAssembler.h: (JSC::ARMAssembler::): (JSC::ARMAssembler::bx): (JSC::ARMAssembler::blx): (JSC::ARMAssembler::loadBranchTarget): (JSC::ARMAssembler::jmp): (JSC::ARMAssembler::getLdrImmAddress):
  • assembler/MacroAssemblerARM.h: (JSC::MacroAssemblerARM::jump): (JSC::MacroAssemblerARM::nearCall): (JSC::MacroAssemblerARM::call): (JSC::MacroAssemblerARM::ret): (JSC::MacroAssemblerARM::prepareCall): (JSC::MacroAssemblerARM::call32):
05:28 Changeset [58090] by eric@webkit.org

2010-04-22 Zhenyao Mo <zmo@google.com>

Reviewed by Eric Seidel.

Turn on antialias and stencil as default canvas attributes.
https://bugs.webkit.org/show_bug.cgi?id=36972

  • fast/canvas/webgl/resources/webgl-test.js: (create3DContext): Turn on antialias and stencil as default canvas attributes.
05:09 Changeset [58089] by eric@webkit.org

2010-04-22 Gustavo Sverzut Barbieri <barbieri@profusion.mobi>

Reviewed by Adam Roben.

EFL does not support PluginDatabase yet.
http://webkit.org/b/37854

No behavior changes, so no new tests were added.

  • loader/FrameLoader.cpp: (WebCore::FrameLoader::defaultObjectContentType):
05:08 Changeset [58088] by eric@webkit.org

2010-04-22 Eric Seidel <eric@webkit.org>

Unreviewed. Script fix. Will ask dpranke to look tomorrow.

new-run-webkit-tests --verbose shows ever-increasing #EOF lines
https://bugs.webkit.org/show_bug.cgi?id=37794

The bots are seeing cases where .error is sometimes empty.
Lets make the code not crash in that case for now.

  • Scripts/webkitpy/layout_tests/port/webkit.py:
04:21 Changeset [58087] by eric@webkit.org

2010-04-22 Ilya Tikhonovsky <loislo@chromium.org>

Reviewed by Yury Semikhatsky.

Web Inspector: Timeline scrolling speed is slow if it has more than 1k timeline marks.
https://bugs.webkit.org/show_bug.cgi?id=37924

  • inspector/front-end/TimelineGrid.js: (WebInspector.TimelineGrid.prototype.updateDividers): (WebInspector.TimelineGrid.prototype.addEventDividers):
  • inspector/front-end/TimelineOverviewPane.js: (WebInspector.TimelineOverviewPane.prototype.updateEventDividers):
  • inspector/front-end/TimelinePanel.js: (WebInspector.TimelinePanel.prototype._updateEventDividers): (WebInspector.TimelinePanel.prototype._refresh):
04:15 Changeset [58086] by ukai@chromium.org

Unreviewed fix.

  • Scripts/webkitpy/layout_tests/port/websocket_server.py: fix NameError: global name 'f' is not defined.
04:11 Changeset [58085] by eric@webkit.org

2010-04-22 Kenneth Rohde Christiansen <kenneth@webkit.org>

Unreviewed expected results update.

Rename window.media to window.styleMedia
https://bugs.webkit.org/show_bug.cgi?id=36187

  • fast/dom/Window/window-properties-expected.txt:
  • fast/dom/Window/window-property-descriptors-expected.txt:
  • fast/dom/prototype-inheritance-2-expected.txt:
  • fast/dom/prototype-inheritance-expected.txt:
  • fast/media/print-restores-previous-mediatype-expected.txt:
  • platform/chromium/fast/dom/prototype-inheritance-expected.txt:
  • platform/gtk/fast/dom/Window/window-properties-expected.txt:
  • platform/gtk/fast/dom/Window/window-property-descriptors-expected.txt:
  • platform/gtk/fast/dom/prototype-inheritance-expected.txt:
  • platform/qt/fast/dom/Window/window-properties-expected.txt:
  • platform/qt/fast/dom/Window/window-property-descriptors-expected.txt:
  • platform/qt/fast/dom/prototype-inheritance-expected.txt:
  • platform/win/fast/dom/Window/window-property-descriptors-expected.txt:
  • platform/win/fast/dom/prototype-inheritance-2-expected.txt:
  • platform/win/fast/dom/prototype-inheritance-expected.txt:

2010-04-22 Kenneth Rohde Christiansen <kenneth@webkit.org>

Reviewed by Simon Fraser.

Rename window.media to window.styleMedia
https://bugs.webkit.org/show_bug.cgi?id=36187

It has been defined that the AbstractView media extension
defined in the CSSOM View spec should be renamed to styleMedia.
This patch does that and updates the current layout tests
making use of it.

  • page/AbstractView.idl:
  • page/DOMWindow.cpp: (WebCore::DOMWindow::styleMedia):
  • page/DOMWindow.h:
  • page/DOMWindow.idl:
03:41 Changeset [58084] by eric@webkit.org

2010-04-22 Anton Muhin <antonm@chromium.org>

Reviewed by Adam Barth.

[v8] Do not pass empty handle into SetHiddenValue which would crash.
https://bugs.webkit.org/show_bug.cgi?id=37801

  • bindings/v8/V8AbstractEventListener.cpp: (WebCore::V8AbstractEventListener::handleEvent): add couple of asserts to check for unexpected paths (WebCore::V8AbstractEventListener::invokeEventHandler): bail out of jsEvent is empty handle
03:31 Changeset [58083] by eric@webkit.org

2010-04-22 Diego Escalante Urrelo <descalante@igalia.com>

Reviewed by Xan Lopez.

[Gtk] Evaluate and create tests for all the AtkRole's implemented by
WebKitGtk
https://bugs.webkit.org/show_bug.cgi?id=34449

Expand testatkroles to test ATK_ROLE_SEPARATOR.

  • tests/testatkroles.c: (test_webkit_atk_get_role_separator): (main):
03:19 Changeset [58082] by ukai@chromium.org

Unreviewed fix.

  • Scripts/new-run-webkit-websocketserver: PyWebSocket no longer takes keyword argument 'register_cygwin'.
03:00 Changeset [58081] by ukai@chromium.org

Unreviewed fix.

  • Scripts/webkitpy/layout_tests/port/http_server.py: fix NameError: global name 'env' is not defined.
02:59 Changeset [58080] by eric@webkit.org

2010-04-22 Eric Seidel <eric@webkit.org>

Unreviewed. Restore the Leopard bots to using
old-run-webkit-tests for now.

  • Scripts/run-webkit-tests:
02:48 Changeset [58079] by ukai@chromium.org

Unreviewed fix.

  • Scripts/webkitpy/layout_tests/port/websocket_server.py: fix NameError: global name 'env' is not defined.
02:38 Changeset [58078] by eric@webkit.org

2010-04-22 Stephan Aßmus <superstippi@gmx.de>

Reviewed by David Levin.

[Haiku] Implement ImageBuffer support
https://bugs.webkit.org/show_bug.cgi?id=35288

Covered by existing tests.

Complete implementation of ImageBuffer for Haiku. Uses StillImage
class to export a WebCore::Image and associated GraphicsContext
to perform arbitrary drawing in the offscreen BBitmap buffer.

  • platform/graphics/haiku/ImageBufferData.h:
  • platform/graphics/haiku/ImageBufferHaiku.cpp, (WebCore::ImageBufferData::ImageBufferData), (WebCore::ImageBufferData::~ImageBufferData), (WebCore::ImageBuffer::ImageBuffer), (WebCore::ImageBuffer::~ImageBuffer), (WebCore::ImageBuffer::context), (WebCore::ImageBuffer::image):

Implementation uses offscreen BBitmap and BView, wraps StillImage
around those to provide WebCore::Image interface.

(WebCore::ImageBuffer::platformTransformColorSpace):
(WebCore::convertFromData):

Method just performs BGRA <-> RGBA conversion.

(WebCore::convertFromInternalData):

Method just performs BGRA <-> RGBA conversion and handles
pre-multiplying the color values if requested.

(WebCore::convertToInternalData):

Method just performs BGRA <-> RGBA conversion and handles
de.multiplying the color values if requested.

(WebCore::getImageData):

Common code for the next two methods.

(WebCore::ImageBuffer::getUnmultipliedImageData),
(WebCore::ImageBuffer::getPremultipliedImageData):

Implemented.

(WebCore::putImageData):

Common code for the next two methods.

(WebCore::ImageBuffer::putUnmultipliedImageData),
(WebCore::ImageBuffer::putPremultipliedImageData):

Implemented.

(WebCore::ImageBuffer::toDataURL):

Uses Haiku "Translation Kit" to convert image data to data
of the requested mime type.

02:34 Changeset [58077] by ukai@chromium.org

2010-04-22 Fumitoshi Ukai <ukai@chromium.org>

Reviewed by Adam Barth.

Chromium: Add --chromium option to new-run-webkit-websocketserver
https://bugs.webkit.org/show_bug.cgi?id=37664

Move cygwin setup logic in chromium_win.py.

  • Scripts/new-run-webkit-httpd: remove register_cygwin parameter to pass http_server.
  • Scripts/new-run-webkit-websocketserver: add --chromium flag
  • Scripts/webkitpy/layout_tests/port/chromium_win.py: setup for cygwin
  • Scripts/webkitpy/layout_tests/port/http_server.py: remove cygwin setup logic
  • Scripts/webkitpy/layout_tests/port/websocket_server.py: remove cygwin setup logic
02:20 Changeset [58076] by eric@webkit.org

2010-04-22 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

new-run-webkit-tests --verbose shows ever-increasing #EOF lines
https://bugs.webkit.org/show_bug.cgi?id=37794

  • Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py:
    • Fix the log message to explain that this is stderr output, not test output.
  • Scripts/webkitpy/layout_tests/port/webkit.py:
    • Be sure to reset the server_process.error after reading (seems like the wrong place to do this, but at least this fixes the bug and dpranke and I can talk about better designs later).
    • Also remove the #EOF from the stderr output before returning it.
02:12 Changeset [58075] by abarth@webkit.org

2010-04-22 Adam Barth <abarth@webkit.org>

Reviewed by Darin Adler.

Hack up PrettyDiff to allow line-by-line comments
https://bugs.webkit.org/show_bug.cgi?id=37886

Admittedly a bit of a hack, this is a basic line-by-line editor we can
play with thanks to Andrew Scherkus. It's meant to integrate with the
"review" page at bugs.webkit.org.

I changed a few things from Andrew's original version:
1) Trigger text boxes on single click to make iPhone/iPad reviewing
easier.
2) Clear the main text box on load.
3) Reference a version of prototype.js on bugs.webkit.org.

  • PrettyPatch/PrettyPatch.rb:
  • prototype.js: Added.
01:29 Changeset [58074] by eric@webkit.org

2010-04-22 Eric Seidel <eric@webkit.org>

Unreviewed. Fixing new-run-webkit-tests on the bots.

  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
    • Fixing _compat_shim_option to take variable keyword args.
    • Turns out that nargs + callback is ignored unless type is specified. So I added the type so that the option was properly consumed. This was why new-run-webkit-tests couldn't find any tests, it was looking for them under "20" because that was the argument which should have been consumed by --exit-after-n-failures.
01:03 Changeset [58073] by tony@chromium.org

2010-04-22 Tony Chang <tony@chromium.org>

Not reviewed, build fix for chromium Windows.

[chromium] fix ImageDiff compile on windows
https://bugs.webkit.org/show_bug.cgi?id=37979

  • DumpRenderTree/chromium/ImageDiff.cpp:
00:50 Changeset [58072] by abarth@webkit.org

2010-04-22 Adam Barth <abarth@webkit.org>

Unreviewed, rolling out r58069.
http://trac.webkit.org/changeset/58069
https://bugs.webkit.org/show_bug.cgi?id=27751

Broke compile on Windows.

  • editing/execCommand/clipboard-access-expected.txt: Removed.
  • editing/execCommand/clipboard-access.html: Removed.
  • editing/execCommand/script-tests/clipboard-access.js: Removed.

2010-04-22 Adam Barth <abarth@webkit.org>

Unreviewed, rolling out r58069.
http://trac.webkit.org/changeset/58069
https://bugs.webkit.org/show_bug.cgi?id=27751

Broke compile on Windows.

  • WebCore.base.exp:
  • editing/EditorCommand.cpp: (WebCore::supportedPaste): (WebCore::createCommandMap):
  • page/Settings.cpp: (WebCore::Settings::Settings):
  • page/Settings.h:

2010-04-22 Adam Barth <abarth@webkit.org>

Unreviewed, rolling out r58069.
http://trac.webkit.org/changeset/58069
https://bugs.webkit.org/show_bug.cgi?id=27751

Broke compile on Windows.

  • public/WebSettings.h:
  • src/WebSettingsImpl.cpp:
  • src/WebSettingsImpl.h:

2010-04-22 Adam Barth <abarth@webkit.org>

Unreviewed, rolling out r58069.
http://trac.webkit.org/changeset/58069
https://bugs.webkit.org/show_bug.cgi?id=27751

Broke compile on Windows.

  • webkit/webkitwebsettings.cpp: (webkit_web_settings_class_init): (webkit_web_settings_set_property): (webkit_web_settings_get_property): (webkit_web_settings_copy):
  • webkit/webkitwebview.cpp: (webkit_web_view_update_settings): (webkit_web_view_settings_notify):

2010-04-22 Adam Barth <abarth@webkit.org>

Unreviewed, rolling out r58069.
http://trac.webkit.org/changeset/58069
https://bugs.webkit.org/show_bug.cgi?id=27751

Broke compile on Windows.

  • WebView/WebPreferenceKeysPrivate.h:
  • WebView/WebPreferences.mm: (+[WebPreferences initialize]):
  • WebView/WebPreferencesPrivate.h:
  • WebView/WebView.mm: (-[WebView _preferencesChangedNotification:]):

2010-04-22 Adam Barth <abarth@webkit.org>

Unreviewed, rolling out r58069.
http://trac.webkit.org/changeset/58069
https://bugs.webkit.org/show_bug.cgi?id=27751

Broke compile on Windows.

  • Api/qwebsettings.cpp: (QWebSettingsPrivate::apply):
  • Api/qwebsettings.h:

2010-04-22 Adam Barth <abarth@webkit.org>

Unreviewed, rolling out r58069.
http://trac.webkit.org/changeset/58069
https://bugs.webkit.org/show_bug.cgi?id=27751

Broke compile on Windows.

  • Interfaces/IWebPreferencesPrivate.idl:
  • WebPreferenceKeysPrivate.h:
  • WebPreferences.cpp: (WebPreferences::initializeDefaultSettings):
  • WebPreferences.h:
  • WebView.cpp: (WebView::notifyPreferencesChanged):

2010-04-22 Adam Barth <abarth@webkit.org>

Unreviewed, rolling out r58069.
http://trac.webkit.org/changeset/58069
https://bugs.webkit.org/show_bug.cgi?id=27751

Broke compile on Windows.

  • DumpRenderTree/LayoutTestController.cpp: (LayoutTestController::staticFunctions):
  • DumpRenderTree/LayoutTestController.h:
  • DumpRenderTree/chromium/LayoutTestController.cpp: (LayoutTestController::LayoutTestController): (LayoutTestController::overridePreference):
  • DumpRenderTree/chromium/LayoutTestController.h:
  • DumpRenderTree/chromium/TestShell.cpp: (TestShell::resetWebSettings):
  • DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:
  • DumpRenderTree/mac/DumpRenderTree.mm: (resetDefaultsToConsistentValues):
  • DumpRenderTree/mac/LayoutTestControllerMac.mm:
  • DumpRenderTree/qt/DumpRenderTreeQt.cpp: (WebCore::WebPage::WebPage): (WebCore::WebPage::resetSettings):
  • DumpRenderTree/qt/LayoutTestControllerQt.cpp:
  • DumpRenderTree/qt/LayoutTestControllerQt.h:
  • DumpRenderTree/win/DumpRenderTree.cpp: (resetDefaultsToConsistentValues):
  • DumpRenderTree/win/LayoutTestControllerWin.cpp:
  • DumpRenderTree/wx/LayoutTestControllerWx.cpp:
00:40 Changeset [58071] by eric@webkit.org

2010-04-22 Eric Seidel <eric@webkit.org>

Unreviewed. Just adding logging.

Adding logging to help debug why the Leopard Bot
can't find any tests to run.

  • Scripts/webkitpy/layout_tests/layout_package/test_files.py:
  • Scripts/webkitpy/layout_tests/port/base.py:
  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
    • Make it a fatal error to have no tests to run.
00:40 Changeset [58070] by tony@chromium.org

2010-04-22 Tony Chang <tony@chromium.org>

Reviewed by Shinichiro Hamaji.

[chromium] roll chromium deps to pick up some windows build fixes
https://bugs.webkit.org/show_bug.cgi?id=37972

  • DEPS:
00:37 Changeset [58069] by abarth@webkit.org

2010-04-22 Abhishek Arya <inferno@chromium.org>

Reviewed by Adam Barth.

Manual test to verify that javascript cannot access clipboard.
https://bugs.webkit.org/show_bug.cgi?id=27751

  • editing/execCommand/clipboard-access-expected.txt: Added.
  • editing/execCommand/clipboard-access.html: Added.
  • editing/execCommand/script-tests/clipboard-access.js: Added. (enabled): (whenEnabled): (runTest):

2010-04-22 Abhishek Arya <inferno@chromium.org>

Reviewed by Adam Barth.

Add support for controlling clipboard access from javascript.
Clipboard access from javascript is disabled by default.
https://bugs.webkit.org/show_bug.cgi?id=27751

Test: editing/execCommand/clipboard-access.html

  • WebCore.base.exp:
  • editing/EditorCommand.cpp: (WebCore::supportedCopyCut): (WebCore::supportedPaste): (WebCore::createCommandMap):
  • page/Settings.cpp: (WebCore::Settings::Settings): (WebCore::Settings::setJavaScriptCanAccessClipboard):
  • page/Settings.h: (WebCore::Settings::javaScriptCanAccessClipboard):

2010-04-22 Abhishek Arya <inferno@chromium.org>

Reviewed by Adam Barth.

Add support for controlling clipboard access from javascript.
Clipboard access from javascript is disabled by default.
https://bugs.webkit.org/show_bug.cgi?id=27751

  • public/WebSettings.h:
  • src/WebSettingsImpl.cpp: (WebKit::WebSettingsImpl::setJavaScriptCanAccessClipboard):
  • src/WebSettingsImpl.h:

2010-04-22 Abhishek Arya <inferno@chromium.org>

Reviewed by Adam Barth.

Add support for controlling clipboard access from javascript.
Clipboard access from javascript is disabled by default.
https://bugs.webkit.org/show_bug.cgi?id=27751

  • webkit/webkitwebsettings.cpp: (webkit_web_settings_class_init): (webkit_web_settings_set_property): (webkit_web_settings_get_property): (webkit_web_settings_copy):
  • webkit/webkitwebview.cpp: (webkit_web_view_update_settings): (webkit_web_view_settings_notify):

2010-04-22 Abhishek Arya <inferno@chromium.org>

Reviewed by Adam Barth.

Add support for controlling clipboard access from javascript.
Clipboard access from javascript is disabled by default.
https://bugs.webkit.org/show_bug.cgi?id=27751

  • WebView/WebPreferenceKeysPrivate.h:
  • WebView/WebPreferences.mm: (+[WebPreferences initialize]): (-[WebPreferences javaScriptCanAccessClipboard]): (-[WebPreferences setJavaScriptCanAccessClipboard:]):
  • WebView/WebPreferencesPrivate.h:
  • WebView/WebView.mm: (-[WebView _preferencesChangedNotification:]):

2010-04-22 Abhishek Arya <inferno@chromium.org>

Reviewed by Adam Barth.

Add support for controlling clipboard access from javascript.
Clipboard access from javascript is disabled by default.
https://bugs.webkit.org/show_bug.cgi?id=27751

  • Api/qwebsettings.cpp: (QWebSettingsPrivate::apply):
  • Api/qwebsettings.h:

2010-04-22 Abhishek Arya <inferno@chromium.org>

Reviewed by Adam Barth.

Add support for controlling clipboard access from javascript.
Clipboard access from javascript is disabled by default.
https://bugs.webkit.org/show_bug.cgi?id=27751

  • Interfaces/IWebPreferencesPrivate.idl:
  • WebPreferenceKeysPrivate.h:
  • WebPreferences.cpp: (WebPreferences::initializeDefaultSettings): (WebPreferences::javaScriptCanAccessClipboard): (WebPreferences::setJavaScriptCanAccessClipboard):
  • WebPreferences.h:
  • WebView.cpp: (WebView::notifyPreferencesChanged):

2010-04-22 Abhishek Arya <inferno@chromium.org>

Reviewed by Adam Barth.

Add support for controlling clipboard access from javascript.
Clipboard access from javascript is enabled in test framework.
https://bugs.webkit.org/show_bug.cgi?id=27751

  • DumpRenderTree/LayoutTestController.cpp: (setJavaScriptCanAccessClipboardCallback): (LayoutTestController::staticFunctions):
  • DumpRenderTree/LayoutTestController.h:
  • DumpRenderTree/chromium/LayoutTestController.cpp: (LayoutTestController::LayoutTestController): (LayoutTestController::setJavaScriptCanAccessClipboard): (LayoutTestController::overridePreference):
  • DumpRenderTree/chromium/LayoutTestController.h:
  • DumpRenderTree/chromium/TestShell.cpp: (TestShell::resetWebSettings):
  • DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: (LayoutTestController::setJavaScriptCanAccessClipboard):
  • DumpRenderTree/mac/DumpRenderTree.mm: (resetDefaultsToConsistentValues):
  • DumpRenderTree/mac/LayoutTestControllerMac.mm: (LayoutTestController::setJavaScriptCanAccessClipboard):
  • DumpRenderTree/qt/DumpRenderTreeQt.cpp: (WebCore::WebPage::WebPage): (WebCore::WebPage::resetSettings):
  • DumpRenderTree/qt/LayoutTestControllerQt.cpp: (LayoutTestController::setJavaScriptCanAccessClipboard):
  • DumpRenderTree/qt/LayoutTestControllerQt.h:
  • DumpRenderTree/win/DumpRenderTree.cpp: (resetDefaultsToConsistentValues):
  • DumpRenderTree/win/LayoutTestControllerWin.cpp: (LayoutTestController::setJavaScriptCanAccessClipboard):
  • DumpRenderTree/wx/LayoutTestControllerWx.cpp: (LayoutTestController::setJavaScriptCanAccessClipboard):
00:29 Changeset [58068] by eric@webkit.org

2010-04-22 Stephan Aßmus <superstippi@gmx.de>

Reviewed by David Levin.

[Haiku] Improve SimpleFontDataHaiku
https://bugs.webkit.org/show_bug.cgi?id=37411

Covered by existing tests.

  • platform/graphics/haiku/SimpleFontDataHaiku.cpp:
    • Cleaned up includes
    • Removed prototype "charUnicodeToUTF8HACK"

(WebCore::SimpleFontData::platformInit):

  • Use const BFont pointer

(WebCore::SimpleFontData::smallCapsFontData):

  • "fontPlatformData" was leaked. It is only used as the key for the font cache lookup.

(WebCore::SimpleFontData::platformMetricsForGlyph):

  • Use existing WebCore encoding infrastructure instead of adding a hack for Haiku.
00:11 Changeset [58067] by aestes@apple.com

Reference count WebView's backing store bitmap to prevent
deleteBackingStore() from freeing the bitmap while it is still being
referenced by Core Animation.

Reviewed by Maciej Stachowiak.

https://bugs.webkit.org/show_bug.cgi?id=37954

  • WebView.cpp:

(WebView::ensureBackingStore):
(WebView::addToDirtyRegion):
(WebView::scrollBackingStore):
(WebView::updateBackingStore):
(WebView::paint):
(WebView::backingStore):
(releaseBackingStoreCallback): deref m_backingStoreBitmap once Core
Animation has dropeed its reference to the memory.
(WebView::updateRootLayerContents): ref m_backingStoreBitmap before
passing the memory to Core Animation to prevent deleteBackingStore()
from freeing it while it is still referenced by CA.

  • WebView.h: Make m_backingStoreBitmap a RefCountedGDIHandle<HBITMAP>,

and make m_backingStoreDirtyRegion a RefCountedGDIHandle<HRGN>.

04/21/10:

23:59 Changeset [58066] by abarth@webkit.org

2010-04-21 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

webkit-patch can't fetch attachments on security bugs
https://bugs.webkit.org/show_bug.cgi?id=37975

Instead of calling CURL, we just need to use our Mechanize object,
which understand bugs.webkit.org authentication.

  • Scripts/webkitpy/common/checkout/api.py:
  • Scripts/webkitpy/common/net/bugzilla.py:
23:50 Changeset [58065] by tony@chromium.org

2010-04-21 Tony Chang <tony@chromium.org>

Reviewed by Shinichiro Hamaji.

[chromium] gyp_webkit should be executable
https://bugs.webkit.org/show_bug.cgi?id=37971

  • gyp_webkit: Added property svn:executable.
23:49 Changeset [58064] by abarth@webkit.org

2010-04-21 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

URL LayoutTests don't run on Firefox
https://bugs.webkit.org/show_bug.cgi?id=37974

It turns out Firefox doesn't handle base URLs properly if we go through
the DOM to create the <base> and <a> elements. If we switch to
document.write, we can run the tests in Firefox (even though
document.write is lame).

  • fast/url/resources/utilities.js: (canonicalize): (setBaseURL):
23:48 Changeset [58063] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed. Temporary commit, will roll-out before morning.

Turning on new-run-webkit-tests for the Leopard build bot
for testing of the harness.
Users should not noctice (except for the results.html difference).

  • Scripts/run-webkit-tests:
23:38 Changeset [58062] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed. Fix typo in my previous fix attempt.

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbø
https://bugs.webkit.org/show_bug.cgi?id=37765

More on-the-bot debugging, sigh. I wish I had a local build.

  • Scripts/webkitpy/layout_tests/port/base.py:
23:32 Changeset [58061] by eric@webkit.org

2010-04-21 Steve Block <steveblock@google.com>

Reviewed by Kenneth Rohde Christiansen.

Replace Geolocation::suspend()/resume() which are required by Android.
https://bugs.webkit.org/show_bug.cgi?id=37942

These methods were removed in Bug 36255.

Build fix only, no new tests.

  • page/Geolocation.cpp: (WebCore::Geolocation::suspend): (WebCore::Geolocation::resume):
  • page/Geolocation.h:
23:29 Changeset [58060] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed. Fix typo in my previous fix attempt.

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbø
https://bugs.webkit.org/show_bug.cgi?id=37765

  • Scripts/webkitpy/layout_tests/port/base.py:
    • Add self. to class variable access.
23:18 Changeset [58059] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed. Another attempt to fix NRWT for chromium.

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbø
https://bugs.webkit.org/show_bug.cgi?id=37765

  • Scripts/webkitpy/layout_tests/port/base.py:
    • wdiff_text was returning a byte array instead of a unicode string. The simple fix was to just decode the result. However, seeing so much duplicated code with Executive made me cry, so I re-wrote the function to be more like pretty_patch_text and use run_command (which already knows how to handle unicode).
23:02 Changeset [58058] by eric@webkit.org

2010-04-21 Ben Murdoch <benm@google.com>

Reviewed by Jeremy Orlow.

send-oncancel-event touch test should use js-test-post-function.js
https://bugs.webkit.org/show_bug.cgi?id=37830

Restore reference to js-test-post-function removed erroneously in copy
when this test was move in http://trac.webkit.org/changeset/54253
and the corresponding call to the isSuccessfullyParsed function.

  • fast/events/touch/script-tests/send-oncancel-event.js: (touchcancelHandler): Call isSuccessfullyParsed.
  • fast/events/touch/send-oncancel-event.html: Add js-test-post-function.js back.
22:49 Changeset [58057] by eric@webkit.org

2010-04-21 Ray Rischpater <Raymond.Rischpater@Nokia.com>

In HTMLInputElement.cpp there are numerous style
violations.

This patch includes style changes to fix existing style
deviations in this file.

Rubber stamped by Darin Adler.

Fixes <https://bugs.webkit.org/show_bug.cgi?id=37881>

  • html/HTMLInputElement.cpp:
22:49 Changeset [58056] by abarth@webkit.org

2010-04-21 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Windows tests buildbots are too slow to be core builders
https://bugs.webkit.org/show_bug.cgi?id=37970

It's 10:45. The Windows test bots are still hours behind. They're too
slow to be core builders. When they get fast, we can add them back.

  • Scripts/webkitpy/common/net/buildbot.py:
  • Scripts/webkitpy/common/net/buildbot_unittest.py:
22:49 Changeset [58055] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed. Attempt one more time to fix NRWT for chromium.

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbø
https://bugs.webkit.org/show_bug.cgi?id=37765

I'm debugging in the blind because I don't have a chromium
build on this laptop.

  • Scripts/webkitpy/layout_tests/port/chromium.py:
    • Apply the previous fix to a second caller.
22:40 Changeset [58054] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed. Attempt one more time to fix NRWT for chromium.

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbø
https://bugs.webkit.org/show_bug.cgi?id=37765

  • Scripts/webkitpy/layout_tests/port/chromium.py:
    • Fix handling of test_shell output so that we always decode it as utf-8. Unlike DumpRenderTree test_shell does not ever return pixel data. It spits out the pixel dumps in a separate file, thus all output over stdout is utf-8 text.
22:39 Changeset [58053] by eric@webkit.org

2010-04-21 Gustavo Sverzut Barbieri <barbieri@profusion.mobi>

Reviewed by Adam Roben.

Add missing EFL WebCore file.
http://webkit.org/b/37854

No behavior changes, so no new tests were added.

  • bindings/js/ScriptControllerEfl.cpp: Added. (WebCore::ScriptController::createScriptInstanceForWidget):
21:45 Changeset [58052] by ap@apple.com

Reviewed by Shinichiro Hamaji.

https://bugs.webkit.org/show_bug.cgi?id=37964
Fix a typo in comments - Korean encoding name is windows-949, not windows-939

  • WebView/WebPreferences.mm: (+[WebPreferences _setInitialDefaultTextEncodingToSystemEncoding]):
21:38 Changeset [58051] by simon.fraser@apple.com

2010-04-21 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein.

Printing Core Animation-based plug-ins is broken
https://bugs.webkit.org/show_bug.cgi?id=37967

When we do a "flattening" paint because of printing, we need to stash
the flattening flag in the FrameView's m_paintBehavior, because
WebBaseNetscapePluginView needs to find it there to decide whether
to grab a bitmap for printing, for CA plug-ins.

  • page/FrameView.cpp: (WebCore::FrameView::paintContents):
21:30 Changeset [58050] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed. Attempt one more time to fix NRWT for chromium.

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbø
https://bugs.webkit.org/show_bug.cgi?id=37765

  • Scripts/webkitpy/layout_tests/test_types/image_diff.py:
    • Pass encoding to _save_baseline_data and write_output_files
  • Scripts/webkitpy/layout_tests/test_types/test_type_base.py:
    • Make _save_baseline_data and write_output_files take an encoding.
  • Scripts/webkitpy/layout_tests/test_types/text_diff.py:
  • Pass encoding to _save_baseline_data and write_output_files
20:47 Changeset [58049] by eric@webkit.org

2010-04-21 François Sausset <sausset@gmail.com>

Reviewed by Kenneth Rohde Christiansen.

Fix to take into account a change made in "StringImpl".
https://bugs.webkit.org/show_bug.cgi?id=37763

  • mathml/RenderMathMLOperator.cpp: (WebCore::RenderMathMLOperator::updateFromElement):
20:40 Changeset [58048] by morrita@google.com

2010-04-21 MORITA Hajime <morrita@google.com>

Reviewed by Shinichiro Hamaji.

https://bugs.webkit.org/show_bug.cgi?id=37056
LayoutTests/editing/selection/move-left-right.html is slow


Reduced style change that would trigger relayout. Note that You
might find that expectation looks storange. It is due to Bug
37016.

  • editing/selection/move-left-right.html:
  • platform/mac/editing/selection/move-left-right-expected.txt:
20:31 Changeset [58047] by eric@webkit.org

2010-04-21 No'am Rosenthal <noam.rosenthal@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] QtLauncher: make FPS measurement accurate
https://bugs.webkit.org/show_bug.cgi?id=37934

Instead of counting paints, which are not interchangeable with frames that
the user sees, we now set an arbitrary timer for FPS measurements. The idea is
that if the main thread is delayed for any reason, that timer would be delayed
as well.

  • QtLauncher/QtLauncher.pro:
  • QtLauncher/fpstimer.cpp: Added. (FpsTimer::FpsTimer): (FpsTimer::numFrames): (FpsTimer::start): (FpsTimer::stop): (FpsTimer::timerEvent):
  • QtLauncher/fpstimer.h: Added.
  • QtLauncher/webview.cpp: (WebViewGraphicsBased::setFrameRateMeasurementEnabled): (WebViewGraphicsBased::updateFrameRate): (WebViewGraphicsBased::paintEvent):
  • QtLauncher/webview.h:
20:22 Changeset [58046] by eric@webkit.org

2010-04-21 Jay Civelli <jcivelli@chromium.org>

Reviewed by Darin Fisher.

[Chromium] Don't notify the WebView that a popup was
opened when the popup is handled externally, as the popup
is in that case transparent to the WebView.
The notification was causing an ASSERT on Mac (where
the popup is handled externally) as there would be no
notification that the popup was closed.
https://bugs.webkit.org/show_bug.cgi?id=37825

  • src/ChromeClientImpl.cpp: (WebKit::ChromeClientImpl::popupOpened):
20:18 Changeset [58045] by aestes@apple.com

JavaScriptCore: Export WTF::deleteOwnedPtr(HFONT).

Rubber stamped by Mark Rowe.

WebCore: Create a template for creating reference-counted Windows GDI handles.
Refactor RefCountedHFONT to use this template.

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=37952

No change in behavior.

  • WebCore.vcproj/WebCore.vcproj: Add RefCountedHFONT.h
  • platform/graphics/cairo/FontPlatformData.h:

(WebCore::FontPlatformData::hfont): Refactor to return m_hfont->handle() instead of m_hfont->hfont().

  • platform/graphics/cg/FontPlatformData.h:

(WebCore::FontPlatformData::hfont): Same.

  • platform/graphics/win/FontPlatformDataCGWin.cpp:

(WebCore::FontPlatformData::FontPlatformData): Instantiate m_hfont with a RefCountedGDIHandle<HFONT>.

  • platform/graphics/win/FontPlatformDataWin.cpp:

(WebCore::FontPlatformData::FontPlatformData): Same.

  • platform/graphics/win/RefCountedGDIHandle.h: Added. Interface is identical to RefCountedHFONT with exception of renaming hfont() to handle().

(WebCore::RefCountedGDIHandle::create):
(WebCore::RefCountedGDIHandle::createDeleted):
(WebCore::RefCountedGDIHandle::~RefCountedGDIHandle):
(WebCore::RefCountedGDIHandle::handle): Return the GDI handle.
(WebCore::RefCountedGDIHandle::hash):
(WebCore::RefCountedGDIHandle::RefCountedGDIHandle):

  • platform/graphics/win/RefCountedHFONT.h: Removed.
20:16 Changeset [58044] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed, just adding missing ":" in python file.

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbø
https://bugs.webkit.org/show_bug.cgi?id=37765

new-run-webkit-tests --chromium was borked.

  • Scripts/webkitpy/layout_tests/port/chromium.py:
19:51 Changeset [58043] by laszlo.1.gombos@nokia.com

2010-04-21 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Darin Adler.

Fix build if NPAPI support is disabled
https://bugs.webkit.org/show_bug.cgi?id=36621

No new tests, this is a build fix.
Re-submit r56585 with fix for Chromium.

  • plugins/PluginView.cpp: Guard getValueStatic() with NETSCAPE_PLUGIN_API (WebCore::PluginView::getValue):
  • plugins/PluginView.h: Guard getValue() with NETSCAPE_PLUGIN_API
  • plugins/PluginViewNone.cpp: Guard platformGetValue() and platformGetValueStatic with NETSCAPE_PLUGIN_API; Guard privateBrowsingStateChanged() and setJavaScriptPaused() with PLATFORM(MAC) or PLATFORM(CHROMIUM)
19:48 Changeset [58042] by joepeck@webkit.org

2010-04-21 David Yonge-Mallo <davinci@chromium.org>

Reviewed by Dan Bernstein.

ZWNJ - Display non-printing, invisible character
https://bugs.webkit.org/show_bug.cgi?id=16131

Fix the (non)display of glyphs for ZWJ and ZWNJ in simple font code path.

Tests: fast/text/format-control.html

fast/text/zero-width-characters.html

  • platform/graphics/Font.h: (WebCore::Font::operator!=): (WebCore::Font::treatAsZeroWidthSpace): treat ZWNJ and ZWJ as ZWSP.
  • platform/graphics/GlyphPageTreeNode.cpp: (WebCore::GlyphPageTreeNode::initializePage): added ZWNJ and ZWJ.
  • platform/text/CharacterNames.h: added ZWNJ and ZWJ.

2010-04-21 David Yonge-Mallo <davinci@chromium.org>

Reviewed by Dan Bernstein.

ZWNJ - Display non-printing, invisible character
https://bugs.webkit.org/show_bug.cgi?id=16131

Added tests for the handling of ZWJ and ZWNJ characters in simple font
path. These characters have zero width but may change the widths of text
around them. Furthermore, their glyphs should not be displayed in
simple static text.

  • fast/text/format-control.html: Added.
  • fast/text/zero-width-characters.html: modified to test ZWSP, ZWJ, and ZWNJ.
  • platform/mac/fast/text/format-control-expected.checksum: Added.
  • platform/mac/fast/text/format-control-expected.png: Added.
  • platform/mac/fast/text/format-control-expected.txt: Added.
19:41 Changeset [58041] by eric@webkit.org

2010-04-21 Charles Wei <charles.wei@torchmobile.com.cn>

Reviewed by George Staikos.

Fix webkit build problem when xhtmlmp enabled, which is introduced by revision 57927, for
bug fix of 37175 , which seperates DocumentWriter from FrameLoader
https://bugs.webkit.org/show_bug.cgi?id=37915

No new tests since this only fixes the build problem.

  • dom/Document.cpp: (WebCore::Document::isXHTMLMPDocument):
19:17 Changeset [58040] by hamaji@chromium.org

2010-04-21 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by Darin Adler.

Call sites of TextIterator constructor are difficult to read
https://bugs.webkit.org/show_bug.cgi?id=37909

Now we use enum parameters instead of boolean parameters and
boolean version of constructors are eliminated.

This change also changes the names of boolean members so they are
now third person singular.

No new tests because this is just a refactoring.

  • WebCore.base.exp:
  • editing/TextIterator.cpp: (WebCore::TextIterator::TextIterator): (WebCore::TextIterator::advance): (WebCore::TextIterator::handleReplacedElement): (WebCore::TextIterator::shouldRepresentNodeOffsetZero): (WebCore::TextIterator::shouldEmitSpaceBeforeAndAfterNode): (WebCore::TextIterator::handleNonTextNode): (WebCore::TextIterator::exitNode): (WebCore::TextIterator::emitCharacter): (WebCore::TextIterator::emitText): (WebCore::CharacterIterator::CharacterIterator): (WebCore::TextIterator::rangeLength): (WebCore::TextIterator::rangeFromLocationAndLength): (WebCore::findPlainText):
  • editing/TextIterator.h: (WebCore::):
  • editing/VisibleSelection.cpp: (WebCore::VisibleSelection::appendTrailingWhitespace):
  • editing/visible_units.cpp: (WebCore::nextBoundary):
18:58 Changeset [58039] by laszlo.1.gombos@nokia.com

2010-04-21 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Unreviewed.

[Qt] Enable notifications layout tests that are passing.

Qt port has a partial implementation of Notifications,
see r57408.

  • platform/qt/Skipped:
18:56 Changeset [58038] by eric@webkit.org

2010-04-21 Jesus Sanchez-Palencia <jesus@webkit.org>

Reviewed by Kenneth Rohde Christiansen.

Add PageClientQt files.

[Qt] PageClientQt specific implementation for QWidget
https://bugs.webkit.org/show_bug.cgi?id=37858

  • WebCore.pro:

2010-04-21 Jesus Sanchez-Palencia <jesus@webkit.org>

Reviewed by Kenneth Rohde Christiansen.

Add PageClientQWidget implementation based on the old
QWebPageWidgetClient to PageClientQt files. Also fix
QWebPage::setView() to use PageClientQWidget.

[Qt] PageClientQt specific implementation for QWidget
https://bugs.webkit.org/show_bug.cgi?id=37858

  • Api/qwebpage.cpp: (QWebPage::setView):
  • WebCoreSupport/PageClientQt.cpp: Added. (WebCore::PageClientQWidget::scroll): (WebCore::PageClientQWidget::update): (WebCore::PageClientQWidget::setInputMethodEnabled): (WebCore::PageClientQWidget::inputMethodEnabled): (WebCore::PageClientQWidget::setInputMethodHint): (WebCore::PageClientQWidget::cursor): (WebCore::PageClientQWidget::updateCursor): (WebCore::PageClientQWidget::palette): (WebCore::PageClientQWidget::screenNumber): (WebCore::PageClientQWidget::ownerWidget): (WebCore::PageClientQWidget::geometryRelativeToOwnerWidget): (WebCore::PageClientQWidget::pluginParent): (WebCore::PageClientQWidget::style):
  • WebCoreSupport/PageClientQt.h: Added. (WebCore::PageClientQWidget::PageClientQWidget): (WebCore::PageClientQWidget::isQWidgetClient):
18:46 Changeset [58037] by eric@webkit.org

2010-04-21 Diego Escalante Urrelo <descalante@igalia.com>

Reviewed by Xan Lopez.

[Gtk] Evaluate and create tests for all the AtkRole's implemented by
WebKitGtk
https://bugs.webkit.org/show_bug.cgi?id=34449

Implement ATK_ROLE_SEPARATOR.

  • accessibility/AccessibilityRenderObject.cpp: (WebCore::AccessibilityRenderObject::determineAccessibilityRole):
  • accessibility/gtk/AccessibilityObjectAtk.cpp: (WebCore::AccessibilityObject::accessibilityPlatformIncludesObject):
18:34 Changeset [58036] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbø
https://bugs.webkit.org/show_bug.cgi?id=37765

Third time is the charm. I've fixed all the
new-run-webkit-tests regressions from previous attempts.

I fixed the queue to not ignore Tor as a reviwer in r57531,
but instead it throws an exception every time his name is in a patch.

This fixes our Executive.run_command code to work around a Popen
bug http://bugs.python.org/issue5290 whereby python versions before 2.6
do not correctly handle unicode objects as input or output to
Popen.communicate.

Following the advice of:
http://farmdev.com/talks/unicode/
I have changed all of webkitpy to use unicode() objects as strings
instead of str objects (which in Python 3 are renamed "bytes").

String literals were left as "foo" instead of converting to u"foo"
as u"foo" is only required if the string has a non-ascii code point.
Python is smart about comparing str() and unicode() values and will
log an error to the console if the comparison is ever invalid.

Executive.run* now correctly accept and return unicode() objects.
I attempted to fix all the places that we call .write() to make sure we
encode any unicode() objects into utf-8.

I removed several uses of StringIO. StringIO should generally always be
passed a unicode() value.

Likewise I replaced most calls to open() with codecs.open().
codecs.open() matches Python 3 open semantics in requiring an encoding
parameter. Files opened with codecs.open() with a unicode-compatible
encoding will vend unicode() objects from their read() calls, like how
StringIO created with a unicode() object will do.

I also deployed "with" statements wider (now that the project has
settled on Python 2.5) to close a bunch of file descriptor leaks.

  • Scripts/webkitpy/common/checkout/api_unittest.py:
    • Read/write utf-8 files instead of ascii.
    • Update the tests to use test for proper unicode() handling.
  • Scripts/webkitpy/common/checkout/changelog.py:
    • Document that parse_latest_entry_from_file expects file-like objects which return unicode strings.
  • Scripts/webkitpy/common/checkout/changelog_unittest.py:
    • Use unicode() strings instead of str() byte arrays.
    • Deploy "with" to close file descriptor leaks.
  • Scripts/webkitpy/common/checkout/commitinfo.py:
    • Remove unneeded import.
  • Scripts/webkitpy/common/checkout/scm.py:
    • Remove use of str().
  • Scripts/webkitpy/common/checkout/scm_unittest.py:
    • Read/write utf-8 files and use unicode() strings in testing.
  • Scripts/webkitpy/common/config/committers.py:
    • Use \u instead of \x to make slightly clearer what we're doing.
  • Scripts/webkitpy/common/net/bugzilla.py:
    • Add a new _string_contents() method and explain why

we have to call unicode() on the result of soup.string
and why it's safe to do so w/o needing to pass a codec name.

  • Remove the (unused) support for passing a file object to add_patch_to_bug().
  • Scripts/webkitpy/common/net/buildbot.py:
    • Use unicode() instead of str() when needing to coax a

NavigableString object into a unicode() object.

  • Scripts/webkitpy/common/net/buildbot_unittest.py:
    • Add a test which contains a unicode builder name.
  • Scripts/webkitpy/common/net/statusserver.py:
    • Remove use of str()
  • Scripts/webkitpy/common/prettypatch.py:
    • Write out the patch file as utf-8.
  • Scripts/webkitpy/common/system/autoinstall.py:
    • Write out files with a explicit encodings.
    • Deploy "with" to close file descriptor leaks.
  • Scripts/webkitpy/common/system/deprecated_logging.py:
    • Write out log files as utf-8.
  • Scripts/webkitpy/common/system/executive.py:
    • Make run* properly take and return unicode() objects.
    • Cleaned up input handling in run_command a little by adding a _compute_input() method which can return early instead of having such a long/cluttered if-block.
  • Scripts/webkitpy/common/system/executive_unittest.py:
    • Added a unit test to make sure we don't break Tor again!
  • Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py:
    • Write out the test list as utf-8.
  • Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
    • Write out json files as utf-8.
  • Scripts/webkitpy/layout_tests/port/apache_http_server.py:
    • Deploy "with" to close file descriptor leaks.
  • Scripts/webkitpy/layout_tests/port/chromium.py: Add Executive.py FIXME.
  • Scripts/webkitpy/layout_tests/port/chromium_linux.py: ditto.
  • Scripts/webkitpy/layout_tests/port/chromium_mac.py: ditto.
  • Scripts/webkitpy/layout_tests/port/gtk.py: ditto.
  • Scripts/webkitpy/layout_tests/port/mac.py: ditto.
  • Scripts/webkitpy/layout_tests/port/mac_unittest.py:
    • Make the skipped file parsing test unicode.
  • Scripts/webkitpy/layout_tests/port/qt.py: Add Executive.py FIXME.
  • Scripts/webkitpy/layout_tests/port/server_process.py: ditto.
  • Scripts/webkitpy/layout_tests/port/webkit.py:
    • Deploy "with" to close file descriptor leaks.
  • Scripts/webkitpy/layout_tests/port/websocket_server.py:
    • Make explicit the encodings of log files and pid files.
  • Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
    • Make encodings explicit and deploy "with".
  • Scripts/webkitpy/layout_tests/run_webkit_tests.py: ditto.
  • Scripts/webkitpy/layout_tests/test_types/image_diff.py: ditto.
  • Scripts/webkitpy/layout_tests/test_types/test_type_base.py: ditto.
  • Scripts/webkitpy/layout_tests/test_types/text_diff.py: ditto.
  • Scripts/webkitpy/style/filereader_unittest.py: ditto.
  • Scripts/webkitpy/thirdparty/init.py: ditto.
  • Scripts/webkitpy/tool/commands/earlywarningsystem.py:
    • Removed extra import.
  • Scripts/webkitpy/tool/commands/queues.py:
    • No need to map args to strings now that run_command does.
    • Update test results to match args changes.
    • Document our global argument hacks.
  • Scripts/webkitpy/tool/commands/upload.py:
    • Pass the diff directly to add_patch_to_bug instead of creating a StringIO file wrapper.
  • Scripts/webkitpy/tool/mocktool.py:
    • Rename add_patch_to_bug argument to match bugzilla.py
  • Scripts/webkitpy/tool/steps/abstractstep.py:
    • Executive.run_* now require lists instead of strings. The lack of this change was what broke webkit-patch for svn users the first time this was landed.
  • Scripts/webkitpy/tool/steps/postdiff.py:
    • Pass the diff directly to add_patch_to_bug instead of creating a StringIO file wrapper.
  • Scripts/webkitpy/tool/steps/postdiffforcommit.py: ditto
  • Scripts/webkitpy/tool/steps/postdiffforrevert.py: ditto
  • Scripts/webkitpy/tool/steps/steps_unittest.py:
    • Fixed spurious logging seen when running test-webkitpy
18:30 Changeset [58035] by eric@webkit.org

2010-04-21 David Leong <david.leong@nokia.com>

Reviewed by Simon Hausmann.

[Qt] Symbian apps crash on exit due to a bad qObject_cast.

https://bugs.webkit.org/show_bug.cgi?id=37303

Added check for NULL to avoid the crash. Reworked to fix memory leak

  • plugins/symbian/PluginViewSymbian.cpp: (WebCore::PluginView::platformDestroy):
18:03 Changeset [58034] by eric@webkit.org

2010-04-21 Gustavo Sverzut Barbieri <barbieri@profusion.mobi>

Reviewed by Nikolas Zimmermann.

Add missing EFL WebKit/efl theme files.
http://webkit.org/b/37854

  • efl/DefaultTheme/default.edc: Added.
  • efl/DefaultTheme/widget/button/button.edc: Added.
  • efl/DefaultTheme/widget/check/check.edc: Added.
  • efl/DefaultTheme/widget/combo/combo.edc: Added.
  • efl/DefaultTheme/widget/entry/entry.edc: Added.
  • efl/DefaultTheme/widget/file/file.edc: Added.
  • efl/DefaultTheme/widget/radio/radio.edc: Added.
  • efl/DefaultTheme/widget/scrollbar/scrollbar.edc: Added.
  • efl/DefaultTheme/widget/search/cancel/search_cancel.edc: Added.
  • efl/DefaultTheme/widget/search/decoration/search_decoration.edc: Added.
  • efl/DefaultTheme/widget/search/field/search_field.edc: Added.
17:55 Changeset [58033] by jamesr@google.com

2010-04-21 James Robinson <jamesr@chromium.org>

Unreviewed. Updating chromium test expecations.

  • platform/chromium/test_expectations.txt:
17:27 Changeset [58032] by tony@chromium.org

2010-04-21 Sheriff Bot <webkit.review.bot@gmail.com>

Unreviewed, rolling out r58028.
http://trac.webkit.org/changeset/58028
https://bugs.webkit.org/show_bug.cgi?id=37962

broke the chromium build (Requested by tonywork on #webkit).

  • DEPS:
17:25 Changeset [58031] by cfleizach@apple.com

aria-liveregion-notifications.html fails on leopard release bot
https://bugs.webkit.org/show_bug.cgi?id=37112

Reviewed by Alexey Proskuryakov.

WebCore:

Change the method that DRT uses to monitor AX notifications so that its robust
by just sending out NSNotification that can be listened to by anyone, instead
of keeping a static function pointer around.

This change is aimed to avoid flakiness seen in DRT when the notification handlers
are not being called at the appropriate time.

Tests: platform/mac/accessibility/aria-liveregions-addedelement.html

platform/mac/accessibility/aria-liveregions-changedalt.html
platform/mac/accessibility/aria-liveregions-changedtext.html
platform/mac/accessibility/aria-liveregions-removedelement.html

  • accessibility/mac/AccessibilityObjectWrapper.h:
  • accessibility/mac/AccessibilityObjectWrapper.mm:

(-[AccessibilityObjectWrapper accessibilitySetShouldRepostNotifications:]):
(-[AccessibilityObjectWrapper accessibilityPostedNotification:]):

WebKitTools:

Change the way that notifications are listened for by forcing clients
to call a remove listener as well to match the add listener. DRT will
assert if those are not done in the correct order.

  • DumpRenderTree/AccessibilityUIElement.cpp:

(removeNotificationListenerCallback):
(AccessibilityUIElement::getJSClass):

  • DumpRenderTree/AccessibilityUIElement.h:
  • DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:

(AccessibilityUIElement::removeNotificationListener):

  • DumpRenderTree/mac/AccessibilityUIElementMac.mm:

(-[AccessibilityNotificationHandler initWithPlatformElement:]):
(-[AccessibilityNotificationHandler dealloc]):
(-[AccessibilityNotificationHandler _notificationReceived:]):
(-[AccessibilityNotificationHandler setCallback:]):
(AccessibilityUIElement::AccessibilityUIElement):
(AccessibilityUIElement::~AccessibilityUIElement):
(AccessibilityUIElement::addNotificationListener):
(AccessibilityUIElement::removeNotificationListener):

  • DumpRenderTree/win/AccessibilityUIElementWin.cpp:

(AccessibilityUIElement::removeNotificationListener):

LayoutTests:

Add four more tests that test specific live region functionality. Having these
four tests run in order seems to stress the notification mechanism and will expose issues
more frequently.
Modify tests that use notification listeners to also "removeNotificationListener".

  • platform/mac-leopard/Skipped:
  • platform/mac/accessibility/aria-listbox-selectedchildren-change.html:
  • platform/mac/accessibility/aria-liveregion-on-image.html:
  • platform/mac/accessibility/aria-liveregions-addedelement-expected.txt: Added.
  • platform/mac/accessibility/aria-liveregions-addedelement.html: Added.
  • platform/mac/accessibility/aria-liveregions-changedalt-expected.txt: Added.
  • platform/mac/accessibility/aria-liveregions-changedalt.html: Added.
  • platform/mac/accessibility/aria-liveregions-changedtext-expected.txt: Added.
  • platform/mac/accessibility/aria-liveregions-changedtext.html: Added.
  • platform/mac/accessibility/aria-liveregions-notifications.html:
  • platform/mac/accessibility/aria-liveregions-removedelement-expected.txt: Added.
  • platform/mac/accessibility/aria-liveregions-removedelement.html: Added.
  • platform/mac/accessibility/change-notification-on-scroll.html:
17:18 Changeset [58030] by johnnyg@google.com

2010-04-21 John Gregg <johnnyg@google.com>

Reviewed by David Levin.

Fix up notifications request permission test.
https://bugs.webkit.org/show_bug.cgi?id=37959

  • fast/notifications/notifications-request-permission-expected.txt:
  • fast/notifications/notifications-request-permission.html:
17:16 Changeset [58029] by barraclough@apple.com

Remove old exports.

Reviewed by NOBODY (Leopard build fix).

  • WebCore.base.exp:
17:15 Changeset [58028] by evan@chromium.org

2010-04-21 Evan Martin <evan@chromium.org>

Unreviewed, just a dependency change.

[chromium] roll skia forward to r538
https://bugs.webkit.org/show_bug.cgi?id=37960

This will cause a bunch of pixel tests to fail due to bulleted
lists rendering slightly differently. I will grab new baselines
from the bots and check them in in a followup.

  • DEPS:
16:59 Changeset [58027] by barraclough@apple.com

Bug 37949 - Do no copy strings into a shared buffer when converting UStrings to Strings
UString and String now have the same internal representation; Just re-wrap the internal impl.

Reviewed by Darin Adler.

  • bindings/js/JSDOMBinding.cpp:

(WebCore::jsStringSlowCase):

  • bindings/js/JSDOMBinding.h:

(WebCore::ustringToString):
(WebCore::stringToUString):
(WebCore::identifierToString):
(WebCore::ustringToAtomicString):
(WebCore::identifierToAtomicString):

16:59 Changeset [58026] by barraclough@apple.com

Rubber Stamped by Oliver Hunt.

Make SmallStrings store an array of RefPtr<StringImpl>,
instead of a direct array of StringImpls. This allows
us to remove a friend (and a layering violation) from
WTF::StringImpl, and makes it so that all StringImpls
are individually heap allocated.

  • runtime/SmallStrings.cpp:

(JSC::SmallStringsStorage::rep):
(JSC::SmallStringsStorage::SmallStringsStorage):

  • wtf/text/StringImpl.h:
16:54 Changeset [58025] by ap@apple.com

Not reviewed.

https://bugs.webkit.org/show_bug.cgi?id=37933
<rdar://problem/7719540> XMLHttpRequest.withCredentials should be better enforced.

Tweaked the test to pass on Tiger.

  • http/tests/xmlhttprequest/cross-origin-authorization-with-embedder.html: Tiger synchronous loader wrongfully terats 401 response as a failure. While this causes incorrect behavior for XMLHttpRequest, it's acceptable for the purposes of this test.
16:39 Changeset [58024] by andersca@apple.com

2010-04-21 Anders Carlsson <andersca@apple.com>

Reviewed by Sam Weinig.

Add debug menu items to show/hide the Web View.
https://bugs.webkit.org/show_bug.cgi?id=37958

  • MiniBrowser/mac/BrowserWindowController.h:
  • MiniBrowser/mac/BrowserWindowController.m: (-[BrowserWindowController showHideWebView:]): (-[BrowserWindowController removeReinsertWebView:]): (-[BrowserWindowController validateMenuItem:]):
  • MiniBrowser/mac/English.lproj/MainMenu.xib:
16:22 Changeset [58023] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed. Rolling out unicode() changes as they broke NRWT for chromium.
Rolling out:
http://trac.webkit.org/changeset/58014
http://trac.webkit.org/changeset/58016
http://trac.webkit.org/changeset/58020

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbo
https://bugs.webkit.org/show_bug.cgi?id=37765

  • Scripts/webkitpy/common/checkout/api_unittest.py:
  • Scripts/webkitpy/common/checkout/changelog.py:
  • Scripts/webkitpy/common/checkout/changelog_unittest.py:
  • Scripts/webkitpy/common/checkout/commitinfo.py:
  • Scripts/webkitpy/common/checkout/scm.py:
  • Scripts/webkitpy/common/checkout/scm_unittest.py:
  • Scripts/webkitpy/common/config/committers.py:
  • Scripts/webkitpy/common/net/bugzilla.py:
  • Scripts/webkitpy/common/net/buildbot.py:
  • Scripts/webkitpy/common/net/buildbot_unittest.py:
  • Scripts/webkitpy/common/net/statusserver.py:
  • Scripts/webkitpy/common/prettypatch.py:
  • Scripts/webkitpy/common/system/autoinstall.py:
  • Scripts/webkitpy/common/system/deprecated_logging.py:
  • Scripts/webkitpy/common/system/executive.py:
  • Scripts/webkitpy/common/system/executive_unittest.py:
  • Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py:
  • Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
  • Scripts/webkitpy/layout_tests/port/apache_http_server.py:
  • Scripts/webkitpy/layout_tests/port/chromium.py:
  • Scripts/webkitpy/layout_tests/port/chromium_linux.py:
  • Scripts/webkitpy/layout_tests/port/chromium_mac.py:
  • Scripts/webkitpy/layout_tests/port/gtk.py:
  • Scripts/webkitpy/layout_tests/port/mac.py:
  • Scripts/webkitpy/layout_tests/port/mac_unittest.py:
  • Scripts/webkitpy/layout_tests/port/qt.py:
  • Scripts/webkitpy/layout_tests/port/server_process.py:
  • Scripts/webkitpy/layout_tests/port/webkit.py:
  • Scripts/webkitpy/layout_tests/port/websocket_server.py:
  • Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
  • Scripts/webkitpy/layout_tests/test_types/image_diff.py:
  • Scripts/webkitpy/layout_tests/test_types/test_type_base.py:
  • Scripts/webkitpy/layout_tests/test_types/text_diff.py:
  • Scripts/webkitpy/style/filereader_unittest.py:
  • Scripts/webkitpy/thirdparty/init.py:
  • Scripts/webkitpy/tool/commands/earlywarningsystem.py:
  • Scripts/webkitpy/tool/commands/queues.py:
  • Scripts/webkitpy/tool/commands/queues_unittest.py:
  • Scripts/webkitpy/tool/commands/upload.py:
  • Scripts/webkitpy/tool/mocktool.py:
  • Scripts/webkitpy/tool/steps/abstractstep.py:
  • Scripts/webkitpy/tool/steps/postdiff.py:
  • Scripts/webkitpy/tool/steps/postdiffforcommit.py:
  • Scripts/webkitpy/tool/steps/postdiffforrevert.py:
  • Scripts/webkitpy/tool/steps/steps_unittest.py:
16:20 Changeset [58022] by mrowe@apple.com

Tiger build fix.

  • Plugins/WebPluginController.mm: Add an #import that is necessary on Tiger.
16:16 Changeset [58021] by johnnyg@google.com

2010-04-15 John Gregg <johnnyg@google.com>

Reviewed by David Levin.

Enable the notifications layout tests, but skip them on the ports that don't
implement notifications.
https://bugs.webkit.org/show_bug.cgi?id=37667

  • fast/notifications/notifications-check-permission.html: Added.
  • fast/notifications/notifications-check-permission.html-disabled: Removed.
  • fast/notifications/notifications-display-close-events.html: Added.
  • fast/notifications/notifications-display-close-events.html-disabled: Removed.
  • fast/notifications/notifications-double-show.html: Added.
  • fast/notifications/notifications-double-show.html-disabled: Removed.
  • fast/notifications/notifications-no-icon.html: Added.
  • fast/notifications/notifications-no-icon.html-disabled: Removed.
  • fast/notifications/notifications-request-permission.html: Added.
  • fast/notifications/notifications-request-permission.html-disabled: Removed.
  • fast/notifications/notifications-with-permission.html: Added.
  • fast/notifications/notifications-with-permission.html-disabled: Removed.
  • fast/notifications/notifications-without-permission.html: Added.
  • fast/notifications/notifications-without-permission.html-disabled: Removed.
  • platform/gtk/Skipped:
  • platform/mac/Skipped:
  • platform/qt/Skipped:
  • platform/win/Skipped:
16:12 Changeset [58020] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed, fixing NRWT for real this time.

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbø
https://bugs.webkit.org/show_bug.cgi?id=37765

  • Scripts/webkitpy/layout_tests/test_types/test_type_base.py:
    • Add a hack to fix new-run-webkit-tests my understanding of codecs.open(encoding=None) must have been wrong.
16:11 Changeset [58019] by jamesr@google.com

2010-04-21 Evan Stade <estade@chromium.org>

Reviewed by Darin Fisher.

[chromium] Web page serializer saves tag names in upper case
https://bugs.webkit.org/show_bug.cgi?id=37940

Simply convert all tags to lower case.

This is tested by chrome tests:

  • save page browser tests
  • encoding ui tests
  • web page serializer test shell tests
  • src/WebPageSerializer.cpp: (WebKit::WebPageSerializer::generateMetaCharsetDeclaration): (WebKit::WebPageSerializer::generateBaseTagDeclaration):
  • src/WebPageSerializerImpl.cpp: (WebKit::WebPageSerializerImpl::openTagToString): (WebKit::WebPageSerializerImpl::endTagToString):
16:07 Changeset [58018] by mrowe@apple.com

<rdar://problem/7313430> Many crashes in Safari inside Flip4Mac below -[NSAlert didEndAlert:returnCode:contextInfo:]

Reviewed by Eric Carlson.

WebCore:

A manual test case is required here as the Flip4Mac plug-in displays an alert, and some manual tweaking of the plug-in's
preference file on disk is often required to reproduce the bug.

  • manual-tests/plugins/flip4mac-update-alert-over-navigation.html: Added.

WebKit/mac:

Existing versions of the Flip4Mac WebKit plug-in have an object lifetime bug related to an NSAlert that is
used to notify the user about updates to the plug-in. This bug can result in Safari crashing if the page
containing the plug-in navigates while the alert is displayed (<rdar://problem/7313430>).

The gist of the bug is thus: Flip4Mac sets an instance of the TSUpdateCheck class as the modal delegate of the
NSAlert instance. This TSUpdateCheck instance itself has a delegate. The delegate is set to the WmvPlugin
instance which is the NSView subclass that is exposed to WebKit as the plug-in view. Since this relationship
is that of delegates the TSUpdateCheck does not retain the WmvPlugin. This leads to a bug if the WmvPlugin
instance is destroyed before the TSUpdateCheck instance as the TSUpdateCheck instance will be left with a
pointer to a stale object. This will happen if a page containing the Flip4Mac plug-in triggers a navigation
while the update sheet is visible as the WmvPlugin instance is removed from the view hierarchy and there are
no other references to keep the object alive.

We work around this bug by patching the following two messages:

1) -[NSAlert beginSheetModalForWindow:modalDelegate:didEndSelector:contextInfo:]
2) -[TSUpdateCheck alertDidEnd:returnCode:contextInfo:]

Our override of 1) detects whether it is Flip4Mac's update sheet triggering the alert by checking whether the
modal delegate is an instance of TSUpdateCheck. If it is, it retains the modal delegate's delegate.

Our override of 2) then autoreleases the delegate, balancing the retain we added in 1).

These two overrides have the effect of ensuring that the WmvPlugin instance will always outlive the TSUpdateCheck
instance, preventing the TSUpdateCheck instance from accessing a stale delegate pointer and crashing the application.

  • Plugins/WebPluginController.mm:

(-[WebPluginController addPlugin:]): Check whether the plug-in being instantiated is the Flip4Mac plug-in and
install our workaround if it is.
(isKindOfClass): Helper function that checks whether the given object is an instance of the named class.
(WebKit_TSUpdateCheck_alertDidEnd_returnCode_contextInfo_): Autorelease the delegate.
(WebKit_NSAlert_beginSheetModalForWindow_modalDelegate_didEndSelector_contextInfo_): Retain the modal delegate's
delegate if this NSAlert belongs to the Flip4Mac plug-in.
(installFlip4MacPlugInWorkaroundIfNecessary): Swizzle the necessary methods. We swizzle the TSUpdateCheck methods
first since it is possible that in some versions of Flip4Mac the TSUpdateCheck class may not exist or may not have
the method we're interested in. In that case we want to bail out before patching any methods.

16:07 Changeset [58017] by mrowe@apple.com

<rdar://problem/7856151> REGRESSION: NPP_Destroy is not called when page navigates when plug-in is displaying modal dialog

Reviewed by Maciej Stachowiak.

WebCore:

This is a manual test case as I was not able to construct an automated test that reproduced the same issue without displaying
a modal dialog on-screen.

  • manual-tests/plugins/timeout-dialog-displayed-over-navigation.html: Added.
  • manual-tests/plugins/timeout-dialog-displayed-over-navigation.swf: Added.

WebKit/mac:

An interaction between the plug-in host and WebKit was resulting in WKPCSetModal being called while
NetscapePluginInstanceProxy was waiting on a reply to the GetScriptableNPObject message. This resulted
in calls to stop the plug-in being deferred due to the presence of plug-in code up the stack. This
could lead to crashes as it was possible for the plug-in view to be deallocated during the modal runloop.

  • Plugins/Hosted/NetscapePluginHostProxy.mm:

(WKPCInvalidateRect):
(WKPCSetModal): Defer the handling of setModal until the next runloop iteration if the host proxy
is already processing requests. This ensures that there will be no plug-in code on the stack when
the modal runloop is entered, which allows the plug-in to be stopped when the page is navigated while
a modal dialog is displayed.

15:51 Changeset [58016] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed, just fixing exception seen on builders.

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbø
https://bugs.webkit.org/show_bug.cgi?id=37765

  • Scripts/webkitpy/layout_tests/test_types/test_type_base.py:
    • Pass and encoding to _write_into_file_at_path
15:06 Changeset [58015] by simon.fraser@apple.com

2010-04-21 Simon Fraser <simon.fraser@apple.com>

Unreviewed; fix assertion in LayoutTests.

https://bugs.webkit.org/show_bug.cgi?id=37932
Assertion in transitions/transition-end-event-destroy-iframe.html

Now that composited iframes cause their parent document to enter compositing mode,
iframe painting via NSView does't send the required -viewWillDraw, which causes
this assertion. Disable the test until we figure out a solution.

  • platform/mac-snowleopard/Skipped:
15:04 Changeset [58014] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

REGRESSION(57531): the commit-queue still hates Tor Arne Vestbø
https://bugs.webkit.org/show_bug.cgi?id=37765

I fixed the queue to not ignore Tor as a reviwer in r57531,
but instead it throws an exception every time his name is in a patch.

This fixes our Executive.run_command code to work around a Popen
bug http://bugs.python.org/issue5290 whereby python versions before 2.6
do not correctly handle unicode objects as input or output to
Popen.communicate.

Following the advice of:
http://farmdev.com/talks/unicode/
I have changed all of webkitpy to use unicode() objects as strings
instead of str objects (which in Python 3 are renamed "bytes").

String literals were left as "foo" instead of converting to u"foo"
as u"foo" is only required if the string has a non-ascii code point.
Python is smart about comparing str() and unicode() values and will
log an error to the console if the comparison is ever invalid.

Executive.run* now correctly accept and return unicode() objects.
I attempted to fix all the places that we call .write() to make sure we
encode any unicode() objects into utf-8.

I removed several uses of StringIO. StringIO should generally always be
passed a unicode() value.

Likewise I replaced most calls to open() with codecs.open().
codecs.open() matches Python 3 open semantics in requiring an encoding
parameter. Files opened with codecs.open() with a unicode-compatible
encoding will vend unicode() objects from their read() calls, like how
StringIO created with a unicode() object will do.

I also deployed "with" statements wider (now that the project has
settled on Python 2.5) to close a bunch of file descriptor leaks.

  • Scripts/webkitpy/common/checkout/api_unittest.py:
    • Read/write utf-8 files instead of ascii.
    • Update the tests to use test for proper unicode() handling.
  • Scripts/webkitpy/common/checkout/changelog.py:
    • Document that parse_latest_entry_from_file expects file-like objects which return unicode strings.
  • Scripts/webkitpy/common/checkout/changelog_unittest.py:
    • Use unicode() strings instead of str() byte arrays.
    • Deploy "with" to close file descriptor leaks.
  • Scripts/webkitpy/common/checkout/commitinfo.py:
    • Remove unneeded import.
  • Scripts/webkitpy/common/checkout/scm.py:
    • Remove use of str().
  • Scripts/webkitpy/common/checkout/scm_unittest.py:
    • Read/write utf-8 files and use unicode() strings in testing.
  • Scripts/webkitpy/common/config/committers.py:
    • Use \u instead of \x to make slightly clearer what we're doing.
  • Scripts/webkitpy/common/net/bugzilla.py:
    • Add a new _string_contents() method and explain why

we have to call unicode() on the result of soup.string
and why it's safe to do so w/o needing to pass a codec name.

  • Remove the (unused) support for passing a file object to add_patch_to_bug().
  • Scripts/webkitpy/common/net/buildbot.py:
    • Use unicode() instead of str() when needing to coax a

NavigableString object into a unicode() object.

  • Scripts/webkitpy/common/net/buildbot_unittest.py:
    • Add a test which contains a unicode builder name.
  • Scripts/webkitpy/common/net/statusserver.py:
    • Remove use of str()
  • Scripts/webkitpy/common/prettypatch.py:
    • Write out the patch file as utf-8.
  • Scripts/webkitpy/common/system/autoinstall.py:
    • Write out files with a explicit encodings.
    • Deploy "with" to close file descriptor leaks.
  • Scripts/webkitpy/common/system/deprecated_logging.py:
    • Write out log files as utf-8.
  • Scripts/webkitpy/common/system/executive.py:
    • Make run* properly take and return unicode() objects.
    • Cleaned up input handling in run_command a little by adding a _compute_input() method which can return early instead of having such a long/cluttered if-block.
  • Scripts/webkitpy/common/system/executive_unittest.py:
    • Added a unit test to make sure we don't break Tor again!
  • Scripts/webkitpy/layout_tests/layout_package/dump_render_tree_thread.py:
    • Write out the test list as utf-8.
  • Scripts/webkitpy/layout_tests/layout_package/json_results_generator.py:
    • Write out json files as utf-8.
  • Scripts/webkitpy/layout_tests/port/apache_http_server.py:
    • Deploy "with" to close file descriptor leaks.
  • Scripts/webkitpy/layout_tests/port/chromium.py: Add Executive.py FIXME.
  • Scripts/webkitpy/layout_tests/port/chromium_linux.py: ditto.
  • Scripts/webkitpy/layout_tests/port/chromium_mac.py: ditto.
  • Scripts/webkitpy/layout_tests/port/gtk.py: ditto.
  • Scripts/webkitpy/layout_tests/port/mac.py: ditto.
  • Scripts/webkitpy/layout_tests/port/mac_unittest.py:
    • Make the skipped file parsing test unicode.
  • Scripts/webkitpy/layout_tests/port/qt.py: Add Executive.py FIXME.
  • Scripts/webkitpy/layout_tests/port/server_process.py: ditto.
  • Scripts/webkitpy/layout_tests/port/webkit.py:
    • Deploy "with" to close file descriptor leaks.
  • Scripts/webkitpy/layout_tests/port/websocket_server.py:
    • Make explicit the encodings of log files and pid files.
  • Scripts/webkitpy/layout_tests/rebaseline_chromium_webkit_tests.py:
    • Make encodings explicit and deploy "with".
  • Scripts/webkitpy/layout_tests/run_webkit_tests.py: ditto.
  • Scripts/webkitpy/layout_tests/test_types/image_diff.py: ditto.
  • Scripts/webkitpy/layout_tests/test_types/test_type_base.py: ditto.
  • Scripts/webkitpy/layout_tests/test_types/text_diff.py: ditto.
  • Scripts/webkitpy/style/filereader_unittest.py: ditto.
  • Scripts/webkitpy/thirdparty/init.py: ditto.
  • Scripts/webkitpy/tool/commands/earlywarningsystem.py:
    • Removed extra import.
  • Scripts/webkitpy/tool/commands/queues.py:
    • No need to map args to strings now that run_command does.
    • Update test results to match args changes.
    • Document our global argument hacks.
  • Scripts/webkitpy/tool/commands/upload.py:
    • Pass the diff directly to add_patch_to_bug instead of creating a StringIO file wrapper.
  • Scripts/webkitpy/tool/mocktool.py:
    • Rename add_patch_to_bug argument to match bugzilla.py
  • Scripts/webkitpy/tool/steps/abstractstep.py:
    • Executive.run_* now require lists instead of strings. The lack of this change was what broke webkit-patch for svn users the first time this was landed.
  • Scripts/webkitpy/tool/steps/postdiff.py:
    • Pass the diff directly to add_patch_to_bug instead of creating a StringIO file wrapper.
  • Scripts/webkitpy/tool/steps/postdiffforcommit.py: ditto
  • Scripts/webkitpy/tool/steps/postdiffforrevert.py: ditto
  • Scripts/webkitpy/tool/steps/steps_unittest.py:
    • Fixed spurious logging seen when running test-webkitpy
14:17 Changeset [58013] by jamesr@google.com

2010-04-21 James Robinson <jamesr@chromium.org>

Unreviewed. Updating chromium test expecations.

  • platform/chromium/test_expectations.txt:
13:59 Changeset [58012] by weinig@apple.com

Fix for https://bugs.webkit.org/show_bug.cgi?id=37937
Wean JavaScriptCore off calls to isMainThread()

Reviewed by Geoffrey Garen.

JavaScriptCore:

  • Replace use of isMainThread() for interpreter reentry checks with a stored value on the JSGlobalData.
  • Replace use of isMainThread() for useMainThread only check in the collector with a stored exclusive thread.
  • API/JSContextRef.cpp:

(JSContextGroupCreate):
Always default to a small stack type for uses of the JSC API. It is
unlikely that the interpreter reentry required on the web will be as
important for other uses of JavaScriptCore.

Update exports.

  • interpreter/Interpreter.cpp:

(JSC::Interpreter::execute):
(JSC::Interpreter::prepareForRepeatCall):
Use new stored JSGlobalData::maxReentryDepth instead of isMainThread().

  • interpreter/Interpreter.h:

Rename MaxMainThreadReentryDepth to MaxLargeThreadReentryDepth and
MaxSecondaryThreadReentryDepth to MaxSmallThreadReentryDepth.

  • jsc.cpp:

(main): Use the a large stack for jsc since it is always using the
main thread.

  • runtime/ArrayPrototype.cpp:

(JSC::arrayProtoFuncToString):
(JSC::arrayProtoFuncToLocaleString):
(JSC::arrayProtoFuncJoin):
Use new stored JSGlobalData::maxReentryDepth instead of isMainThread().

  • runtime/Collector.cpp:

(JSC::Heap::registerThread):
Use the concept of making JSC run on an exclusiveThread instead of
forcing a mainThreadOnly assertion.

  • runtime/JSGlobalData.cpp:

(JSC::JSGlobalData::JSGlobalData):
(JSC::JSGlobalData::createNonDefault):
(JSC::JSGlobalData::create):
(JSC::JSGlobalData::createLeaked):
(JSC::JSGlobalData::sharedInstance):

  • runtime/JSGlobalData.h:

Add ThreadStackType argument to JSGlobalData constructors and set
maxReentryDepth based on it.

WebCore:

No change in behavior.

  • bindings/js/JSDOMWindowBase.cpp:

(WebCore::JSDOMWindowBase::commonJSGlobalData):
Explicitly set a large stack type for the common JSGlobalData and
set the currently running thread as the exclusive thread for its
execution.

  • bindings/js/WorkerScriptController.cpp:

(WebCore::WorkerScriptController::WorkerScriptController):
Explicitly set a small stack type for the workers JSGlobalData.

13:28 Changeset [58011] by kinuko@chromium.org

2010-04-21 Kinuko Yasuda <kinuko@chromium.org>

Unreviewed.

Add myself in committers.py.

  • Scripts/webkitpy/common/config/committers.py:
13:08 Changeset [58010] by barraclough@apple.com

Reviewed by NOBODY (windows build fix pt. 3).

13:05 Changeset [58009] by ggaren@apple.com

<rdar://problem/7879485> Leopard: Crash after opening Software Update dialog

Reviewed by Sam Weinig.

  • JSUtils.cpp:

(unprotectGlobalObject): Use the version of JSGlueAPIEntry that doesn't
call pthread_getspecific, since it's not safe to call pthread_getspecific
from a thread-specific data destructor. (<rdar://problem/7889842>
pthread_getspecific returns 0 when called from thread-specific data
destructor function)

(getThreadGlobalObject): Make sure to set the currentIdentifierTable
when first constructing our JSGlobalObject, since JSGlueAPIEntry has not
yet had an opportunity to do so.

(JSGlueAPIEntry::JSGlueAPIEntry):

  • JSUtils.h: New version of JSGlueAPIEntry that doesn't call

pthread_getspecific.

12:56 Changeset [58008] by barraclough@apple.com

Errk! tree on fire, add fwd!

Reviewed by NOBODY (build fix).

  • ForwardingHeaders/wtf/text/StringHash.h: Added.
12:52 Changeset [58007] by barraclough@apple.com

Reviewed by NOBODY (windows build fix pt. 2).

12:49 Changeset [58006] by barraclough@apple.com

JavaScriptCore: * JavaScriptCore.gypi:

  • JavaScriptCore.pro:
  • wtf/qt/StringQt.cpp: Copied from WebCore/platform/text/qt/StringQt.cpp.

Reviewed by NOBODY (Qt build fix).

WebCore: * WebCore.gypi:

  • WebCore.pro:
  • platform/text/qt/StringQt.cpp: Removed.

Reviewed by NOBODY (Qt build fix).

12:49 Changeset [58005] by dpranke@chromium.org

2010-04-21 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

new-run-webkit-tests: fix a bug in the Chromium port where we would
try to talk to a crashed test_shell and raise exceptions that weren't
being caught.

https://bugs.webkit.org/show_bug.cgi?id=37941

  • Scripts/webkitpy/layout_tests/port/chromium.py:
12:41 Changeset [58004] by jamesr@google.com

2010-04-21 James Robinson <jamesr@chromium.org>

Updating chromium expectations for tests that need new layoutTestController
functionality.

  • platform/chromium/test_expectations.txt:
12:40 Changeset [58003] by barraclough@apple.com
  • API/JSValueRef.cpp:
  • JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.vcproj:
  • runtime/Identifier.cpp:

(JSC::IdentifierTable::~IdentifierTable):
(JSC::IdentifierTable::add):

  • runtime/Identifier.h:
  • wtf/WTFThreadData.h:

(JSC::IdentifierTable::remove):
(JSC::IdentifierTable::literalTable):

  • wtf/text/StringImpl.cpp:

(WebCore::StringImpl::~StringImpl):

Reviewed by NOBODY (windows build fix).

12:19 Changeset [58002] by ajwong@chromium.org

Not reviewed. Sheriffing cleanup.

[chromium] Add bug numbers for expectations created during sheriffing.

  • platform/chromium/test_expectations.txt:
12:13 Changeset [58001] by barraclough@apple.com

JavaScriptCore: Bug 37906 - Remove JSC::UStringImpl; unify with StringImpl.

Reviewed by Oliver Hunt, Darin Adler.

JSC::UStringImpl and WebCore::StringImpl (soon to be renamed to
WTF::StringImpl) are almost identical. Remove duplication of code by unifying
the two, move missing features from UStringImpl into StringImpl & delete the
class UStringImpl.

  • API/JSClassRef.cpp:
  • API/JSContextRef.cpp:
  • GNUmakefile.am:
  • JavaScriptCore.exp:
  • JavaScriptCore.pro:
  • JavaScriptCore.xcodeproj/project.pbxproj:
  • bytecode/EvalCodeCache.h:
  • bytecode/JumpTable.cpp:
  • profiler/ProfileNode.cpp:
  • runtime/Identifier.cpp:

(JSC::Identifier::add):

  • runtime/Identifier.h:

(JSC::Identifier::equal):

  • runtime/UString.cpp:
  • runtime/UString.h:

(WTF::):

  • runtime/UStringImpl.cpp: Removed.
  • runtime/UStringImpl.h:
  • wtf/text/StringHash.h:

(WebCore::StringHash::equal):
(WebCore::CaseFoldingHash::equal):

  • wtf/text/StringImpl.cpp:

(WebCore::StringImpl::~StringImpl):
(WebCore::StringImpl::empty):
(WebCore::StringImpl::sharedBuffer):
(WebCore::equal):

  • wtf/text/StringImpl.h:

(WebCore::StringImpl::StringImpl):
(WebCore::StringImpl::create):
(WebCore::StringImpl::tryCreateUninitialized):
(WebCore::StringImpl::cost):
(WebCore::StringImpl::isIdentifier):
(WebCore::StringImpl::setIsIdentifier):
(WebCore::StringImpl::computeHash):
(WebCore::StringImpl::copyChars):
(WebCore::StringImpl::):

JavaScriptGlue: Bug 37906 - Remove JSC::UStringImpl; unify with StringImpl.
Add forwarding header.

Reviewed by Oliver Hunt, Darin Adler.

  • ForwardingHeaders/wtf/ASCIICType.h: Added.
  • ForwardingHeaders/wtf/text/StringImpl.h: Added.

WebCore: Bug 37906 - Remove JSC::UStringImpl; unify with StringImpl.
Add include for StringHash.h.

Reviewed by Oliver Hunt, Darin Adler.

  • WebCore.xcodeproj/project.pbxproj:
  • bridge/c/c_class.cpp:
12:08 Changeset [58000] by laszlo.1.gombos@nokia.com

2010-04-21 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Unreviewed.

[Qt] [Symbian] Build fix.

Work around a Qt quirk. Some versions of Symbian port Qt
QFontDatabase::removeAllApplicationFonts symbol is not available.

  • DumpRenderTree/qt/DumpRenderTreeQt.cpp: (WebCore::DumpRenderTree::open):
11:32 Changeset [57999] by ap@apple.com

Unreviewed.

https://bugs.webkit.org/show_bug.cgi?id=37933
<rdar://problem/7719540> XMLHttpRequest.withCredentials should be better enforced.

Adding stub implementation of authenticateSession(). Depending on platform loader behavior,
a real implementation may or may not be necessary for the one test that currently uses it
to pass.

  • DumpRenderTree/qt/LayoutTestControllerQt.cpp: (LayoutTestController::authenticateSession):
  • DumpRenderTree/qt/LayoutTestControllerQt.h:
11:29 Changeset [57998] by mitz@apple.com

Fixed bug URL

11:14 Changeset [57997] by ap@apple.com

https://bugs.webkit.org/show_bug.cgi?id=37933
<rdar://problem/7719540> XMLHttpRequest.withCredentials should be better enforced.

  • platform/qt/Skipped: Skip the new test on Qt. There is nothing special about it, existing related tests are skipped already.
11:11 Changeset [57996] by ap@apple.com

Tiger build fix.

  • platform/network/mac/ResourceHandleMac.mm: (WebCore::createNSURLConnection): Fixed a typo, named an argument. (WebCore::ResourceHandle::start): Moved shouldUseCredentialStorage out of #if, since it's now passed to createNSURLConneciton() on all platforms (and then ignored on Tiger).
11:07 Changeset [57995] by ap@apple.com

Reviewed by Adam Roben.

Windows build fix.

  • platform/network/cf/ResourceHandleCFNet.cpp: Declare CFURLConnectionCreateWithProperties for now, as it's mistakenly missing from WebKitSupportLibrary headers.
10:34 Changeset [57994] by dglazkov@chromium.org

2010-04-21 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Darin Adler.

REGRESSION(r57292): Safari/Win and Chromium/Win no longer pass the acid3 test.
https://bugs.webkit.org/show_bug.cgi?id=37902

The issue is due to MSVC creating enums as signed. The fix is to store the value
as unsigned.

Test: http://acid3.acidtests.org/

  • rendering/style/RenderStyle.h: (WebCore::InheritedFlags): Changed type of _insideLink to unsigned.
10:33 Changeset [57993] by ap@apple.com

Reviewed by Brady Eidson.

https://bugs.webkit.org/show_bug.cgi?id=37933
<rdar://problem/7719540> XMLHttpRequest.withCredentials should be better enforced.

Test: http/tests/xmlhttprequest/cross-origin-authorization-with-embedder.html

This improves integration between ResourceHandle and Apple networking libraries.

  • platform/network/cf/ResourceHandleCFNet.cpp: (WebCore::createConnectionProperties): A new helper for creating connection properties dictionary. (WebCore::ResourceHandle::start): Pass connection properties. (WebCore::WebCoreSynchronousLoader::load): Ditto.
  • platform/network/mac/ResourceHandleMac.mm: (WebCore::createNSURLConnection): Factor out OS version dependent code for creating NSURLConnection. Tell NSURLConnection about credential policy upfront. (WebCore::ResourceHandle::start): Use the new function. (+[WebCoreSynchronousLoader loadRequest:allowStoredCredentials:returningResponse:error:]): Ditto.
10:30 Changeset [57992] by xji@chromium.org

2010-04-21 Xiaomei Ji <xji@chromium.org>

Reviewed by Dimitri Glazkov

This patch fixes [chromium] RTL <select> dropdown box expands to right
instead of left.
https://bugs.webkit.org/show_bug.cgi?id=37232

No automatic test is possible.

  • manual-tests/select_dropdown_box_alignment.html: Added.
  • platform/chromium/PopupMenuChromium.cpp: (WebCore::PopupContainer::layout): Adjust the x-axis of dropdown box for RTL.
10:21 Changeset [57991] by eric@webkit.org

2010-04-21 anton muhin <antonm@google.com>

Reviewed by Adam Barth.

[v8] Bail out if fetching of Object.prototype fails.
https://bugs.webkit.org/show_bug.cgi?id=37661

If for any reason we failed to fetch Object.prototype, context cannot
be properly initialized and we bail out.

  • bindings/v8/V8DOMWindowShell.cpp: (WebCore::V8DOMWindowShell::initContextIfNeeded): bail out if installHiddenObjectPrototype failed (WebCore::V8DOMWindowShell::installHiddenObjectPrototype): bail out if failed to fetch Object.prototype
  • bindings/v8/V8DOMWindowShell.h: return false if installHiddenObjectPrototype failed
09:40 Changeset [57990] by timothy@apple.com

Make UserContentURLPattern correctly check for subdomains and the URL
has the same suffix as the pattern. Also improve the parsing of the host.

https://bugs.webkit.org/show_bug.cgi?id=37357

Reviewed by Darin Adler.

  • page/UserContentURLPattern.cpp:

(WebCore::UserContentURLPattern::parse): Simplify the subdomain pattern parsing to
simply check for "*" only or a "*." prefix and then do a substring.
(WebCore::UserContentURLPattern::matchesHost): Check that the host has a "." in the
position before the suffix to ensure it a subdomain and not just a suffix match.

09:02 Changeset [57989] by jwieczorek@webkit.org

2010-04-21 Jakub Wieczorek <jwieczorek@webkit.org>

Unreviewed.

Skip the new fast/lists tests on Mac and Windows until these ports
implement LayoutTestController::markerTextForListItem().

  • platform/mac/Skipped:
  • platform/win/Skipped:
08:57 Changeset [57988] by tkent@chromium.org

2010-04-21 Kent Tamura <tkent@chromium.org>

Reviewed by Dimitri Glazkov.

[DRT/Chromium] Make DRT compilable in Chromium tree
https://bugs.webkit.org/show_bug.cgi?id=37923

We need to use different GYPs in a case of WebKit-only checkout
and a case of whole Chromium checkout because the relative paths
from webkit/ to WebKit/chromium/features.gypi are different in
these cases and we can't use 'conditions' for 'includes' in GYPs.

  • DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp:
08:43 Changeset [57987] by xan@webkit.org

2010-04-21 Xan Lopez <xlopez@igalia.com>

Try to fix compilation on GTK+ debug bots.

  • bindings/gobject/WebKitDOMBinding.cpp: (WebKit::createWrapper):
08:41 Changeset [57986] by jwieczorek@webkit.org

List item markers are not always updated after changes in the DOM.
https://bugs.webkit.org/show_bug.cgi?id=37060

Reviewed by Darin Adler.

WebCore:

In particular, they would not be updated when adding/removing an item
that is not a direct child of the list element.

Tests: fast/lists/ol-nested-items-dynamic-insert.html

fast/lists/ol-nested-items-dynamic-remove.html
fast/lists/ol-nested-items.html
fast/lists/ol-nested-list-dynamic-insert.html
fast/lists/ol-nested-list-dynamic-remove.html
fast/lists/ol-nested-list.html

  • rendering/RenderListItem.cpp:

(WebCore::updateListMarkerNumbers): Change it to traverse the whole subtree of a list, not only siblings of an item.

  • rendering/RenderListItem.h:
  • rendering/RenderObject.cpp: Move the updateListMarkerNumbers function to RenderListItem.cpp.

(WebCore::RenderObject::addChild): Move the code updating list markers to RenderObjectChildList for consistency.

  • rendering/RenderObjectChildList.cpp: Move the updateListMarkerNumbers function to RenderListItem.cpp.

(WebCore::RenderObjectChildList::removeChildNode): Pass the actual node being removed, not the next sibling.
(WebCore::RenderObjectChildList::appendChildNode): Pass the actual node being added, not the next sibling.
(WebCore::RenderObjectChildList::insertChildNode): Pass the actual node being added, not the next sibling.

  • rendering/RenderTreeAsText.cpp:

(WebCore::markerTextForListItem):

WebKit/gtk:

  • webkit/webkitprivate.h:
  • webkit/webkitwebframe.cpp:

(webkit_web_frame_marker_text_for_list_item): Add a private API to get the marker text for a list item.

WebKit/qt:

  • Api/qwebelement.h: Make DumpRenderTreeSupportQt a friend class.
  • WebCoreSupport/DumpRenderTreeSupportQt.cpp:

(DumpRenderTreeSupportQt::markerTextForListItem): Add a private API to get the marker text for a list item.

WebKitTools:

  • DumpRenderTree/LayoutTestController.cpp:

(markerTextForListItemCallback): A function that returns the marker text for a given list item.
(LayoutTestController::staticFunctions):

  • DumpRenderTree/LayoutTestController.h:
  • DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:

(LayoutTestController::markerTextForListItem): Implement it in the GTK port.

  • DumpRenderTree/mac/LayoutTestControllerMac.mm:

(LayoutTestController::markerTextForListItem): Add a stub.

  • DumpRenderTree/qt/LayoutTestControllerQt.cpp:

(LayoutTestController::LayoutTestController):
(LayoutTestController::markerTextForListItem): Implement it in the Qt port.

  • DumpRenderTree/qt/LayoutTestControllerQt.h:
  • DumpRenderTree/win/LayoutTestControllerWin.cpp:

(LayoutTestController::markerTextForListItem): Add a stub.

LayoutTests:

  • fast/lists/dump-list.js:

(indent):
(dumpListHelper): A recursive function to dump nested lists.
(dumpList): A function that dumps a readable plain text representation of a list.

  • fast/lists/ol-nested-items-dynamic-insert-expected.txt: Added.
  • fast/lists/ol-nested-items-dynamic-insert.html: Added.
  • fast/lists/ol-nested-items-dynamic-remove-expected.txt: Added.
  • fast/lists/ol-nested-items-dynamic-remove.html: Added.
  • fast/lists/ol-nested-items-expected.txt: Added.
  • fast/lists/ol-nested-items.html: Added.
  • fast/lists/ol-nested-list-dynamic-insert-expected.txt: Added.
  • fast/lists/ol-nested-list-dynamic-insert.html: Added.
  • fast/lists/ol-nested-list-dynamic-remove-expected.txt: Added.
  • fast/lists/ol-nested-list-dynamic-remove.html: Added.
  • fast/lists/ol-nested-list-expected.txt: Added.
  • fast/lists/ol-nested-list.html: Added.
  • platform/mac/Skipped: Skip the tests until this port implements the new LayoutTestController function.
  • platform/win/Skipped: Skip the tests until this port implements the new LayoutTestController function.
08:24 Changeset [57985] by xan@webkit.org

2010-04-21 Xan Lopez <xlopez@igalia.com>

Reviewed by Adam Barth.

[GTK] GObject DOM bindings
https://bugs.webkit.org/show_bug.cgi?id=33590

Add build bits for the GObject DOM bindings.

  • GNUmakefile.am:

2010-04-21 Xan Lopez <xlopez@igalia.com>

Reviewed by Adam Barth.

[GTK] GObject DOM bindings
https://bugs.webkit.org/show_bug.cgi?id=33590

Initial version of the GObject DOM bindings.

Only bindings for Node.idl and a few of its dependencies are
provided, without public API to access them at the
moment. References to the Document interfaces and to
EventListeners in Node.idl are ignored for GObject to make the
initial patch as small as possible, but will be enabled in a
follow-up patch.

  • GNUmakefile.am:
  • bindings/gobject/ConvertToUTF8String.cpp: Added. (convertToUTF8String):
  • bindings/gobject/ConvertToUTF8String.h: Added.
  • bindings/gobject/WebKitDOMBinding.cpp: Added. (WebKit::domObjects): (WebKit::DOMObjectCache::get): (WebKit::DOMObjectCache::put): (WebKit::DOMObjectCache::forget): (WebKit::createWrapper): (WebKit::kit):
  • bindings/gobject/WebKitDOMBinding.h: Added.
  • bindings/gobject/WebKitDOMObject.cpp: Added. (webkit_dom_object_init): (webkit_dom_object_class_init):
  • bindings/gobject/WebKitDOMObject.h: Added.
  • bindings/scripts/CodeGeneratorGObject.pm: Added.
  • bindings/scripts/gobject-generate-headers.pl: Added.
  • dom/Node.idl:
08:18 Changeset [57984] by eric@webkit.org

2010-04-21 Gustavo Sverzut Barbieri <barbieri@profusion.mobi>

Reviewed by Xan Lopez.

Wrong header being included in FontPlatformDataCairo.cpp
https://bugs.webkit.org/show_bug.cgi?id=37829

No behavior changes, so no new tests were added.

  • platform/graphics/cairo/FontPlatformDataCairo.cpp:
08:08 Changeset [57983] by aroben@apple.com

Exclude leaks in CGGradientCreateWithColorComponents from run-webkit-tests leaks output

The leak in CG is covered by <rdar://problem/7888492>.

Fixes <http://webkit.org/b/37927>.

Reviewed by Eric Carlson.

  • Scripts/old-run-webkit-tests:

(sub countAndPrintLeaks): Exclude leaks in
CGGradientCreateWithColorComponents on certain OSs.

08:08 Changeset [57982] by eric@webkit.org

2010-04-21 Yi Shen <yi.4.shen@nokia.com>

Reviewed by Simon Hausmann.

[Qt] Check the request empty or not in ChromeClientQt::createWindow()
https://bugs.webkit.org/show_bug.cgi?id=37821

  • WebCoreSupport/ChromeClientQt.cpp: (WebCore::ChromeClientQt::createWindow):
07:43 Changeset [57981] by tkent@chromium.org

2010-04-21 Kent Tamura <tkent@chromium.org>

Reviewed by Shinichiro Hamaji.

[DRT/Chromium] Import MockSpellCheck from Chromium
https://bugs.webkit.org/show_bug.cgi?id=37910

Import webkit/tools/test_shell/mock_spellcheck.{cc,h} rev.37241 of Chromium.

  • DumpRenderTree/DumpRenderTree.gyp/DumpRenderTree.gyp: Add ICU explicitly because WTFString.h includes ICU headers. Add MockSpellCheck.{cpp,h}.
  • DumpRenderTree/chromium/MockSpellCheck.cpp: Added.
  • DumpRenderTree/chromium/MockSpellCheck.h: Added.
  • DumpRenderTree/chromium/WebViewHost.cpp: (WebViewHost::spellCheck):
  • DumpRenderTree/chromium/WebViewHost.h:
  • DumpRenderTree/chromium/config.h: Define JS_EXPORTDATA, which is used by wtf/text/AtomicString.h included by wtf/text/WTFString.h.
07:42 Changeset [57980] by aroben@apple.com

Fix leaks of FilterData/SVGFilterBuilder in RenderSVGResourceFilter

Fixes <http://webkit.org/b/37922>.

Reviewed by Dave Hyatt.

  • rendering/RenderSVGResourceFilter.cpp:

(WebCore::RenderSVGResourceFilter::applyResource): Use an OwnPtr to
hold the heap-allocated FilterData object, so that we won't leak it
when we bail out of this function on error.

07:34 Changeset [57979] by eric@webkit.org

2010-04-21 Zoltan Herczeg <zherczeg@webkit.org>

Reviewed by Kenneth Rohde Christiansen.

[Qt] startAnimation() is not needed to preceede nativeImageForCurrentFrame()
https://bugs.webkit.org/show_bug.cgi?id=37844

nativeImageForCurrentFrame() resets the m_decoder parameter under Qt,
which is required by startAnimation() to detect frame and repetition counts.
Hence, Image::drawTiled cannot start animations under Qt:
<html><body background="animated.gif"></body></html> does not work

  • platform/graphics/qt/ImageDecoderQt.cpp: (WebCore::ImageDecoderQt::internalHandleCurrentImage):
07:09 Changeset [57978] by eric@webkit.org

2010-04-21 Patrick Gansterer <paroga@paroga.com>

Reviewed by Darin Adler.

Added missing #include "Lookup.h" in LUT source files.
https://bugs.webkit.org/show_bug.cgi?id=XXXXX

  • runtime/ArrayPrototype.cpp:
  • runtime/DatePrototype.cpp:
  • runtime/JSONObject.cpp:
  • runtime/MathObject.cpp:
  • runtime/NumberConstructor.cpp:
  • runtime/RegExpConstructor.cpp:
  • runtime/RegExpObject.cpp:
  • runtime/StringPrototype.cpp:
06:52 Changeset [57977] by eric@webkit.org

2010-04-21 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Add webkit-patch pretty-diff
https://bugs.webkit.org/show_bug.cgi?id=37892

This is slightly lame because it asks you whether the diff is correct,
but it's a starting point.

  • Scripts/webkitpy/tool/commands/init.py:
  • Scripts/webkitpy/tool/commands/prettydiff.py: Added.
  • Scripts/webkitpy/tool/main.py:
06:42 Changeset [57976] by ossy@webkit.org

[Qt] media/video-source-add-src.html was removed from tree
by r57820, so it's time to remove it from Qt Skipped list.

  • platform/qt/Skipped: media/video-source-add-src.html removed.
06:42 Changeset [57975] by eric@webkit.org

2010-04-21 Adam Barth <abarth@webkit.org>

Reviewed by Darin Adler.

Remove mention of non-existant --no-build option
https://bugs.webkit.org/show_bug.cgi?id=37893

The option doesn't exist!

  • Scripts/webkitpy/tool/commands/stepsequence.py:
06:32 Changeset [57974] by eric@webkit.org

2010-04-21 Balazs Kelemen <kb@inf.u-szeged.hu>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Interrupting JavaScript is cumbersome when you use QtLaucher for testing or profiling.
https://bugs.webkit.org/show_bug.cgi?id=37198

  • QtLauncher/main.cpp: (LauncherWindow::toggleInterruptingJavaScriptEnabled): (LauncherWindow::newWindow): (LauncherWindow::cloneWindow): (LauncherWindow::createChrome): (main):
  • QtLauncher/webpage.cpp: (WebPage::WebPage): (WebPage::shouldInterruptJavaScript):
  • QtLauncher/webpage.h: (WebPage::setInterruptingJavaScriptEnabled):
06:17 Changeset [57973] by eric@webkit.org

2010-04-21 Diego Escalante Urrelo <descalante@igalia.com>

Reviewed by Xan Lopez.

[Gtk] Evaluate and create tests for all the AtkRole's implemented by
WebKitGtk
https://bugs.webkit.org/show_bug.cgi?id=34449

Expand testatkroles to test ATK_ROLE_FORM.

  • tests/testatkroles.c: (test_webkit_atk_get_role_form): (main):
06:08 Changeset [57972] by eric@webkit.org

2010-04-21 Shu Chang <chang.shu@nokia.com>

Reviewed by Simon Hausmann.

[Qt] Fix Symbian build where QT_NO_SYSTEMTRAYICON is defined.
https://bugs.webkit.org/show_bug.cgi?id=37442

  • WebCoreSupport/NotificationPresenterClientQt.cpp: (NotificationPresenterClientQt::show):
  • WebCoreSupport/NotificationPresenterClientQt.h:
05:57 Changeset [57971] by eric@webkit.org

2010-04-21 Benjamin Poulain <ikipou@gmail.com>

Reviewed by Simon Fraser.

Update of fixed elements is not made correctly when the page has been scrolled
https://bugs.webkit.org/show_bug.cgi?id=36783

  • fast/repaint/fixed-child-move-after-scroll.html: Added.
  • fast/repaint/fixed-child-of-fixed-move-after-scroll.html: Added.
  • fast/repaint/fixed-child-of-transformed-move-after-scroll.html: Added.
  • fast/repaint/fixed-move-after-scroll.html: Added.
  • platform/mac/fast/repaint/fixed-child-move-after-scroll-expected.checksum: Added.
  • platform/mac/fast/repaint/fixed-child-move-after-scroll-expected.png: Added.
  • platform/mac/fast/repaint/fixed-child-move-after-scroll-expected.txt: Added.
  • platform/mac/fast/repaint/fixed-child-of-fixed-move-after-scroll-expected.checksum: Added.
  • platform/mac/fast/repaint/fixed-child-of-fixed-move-after-scroll-expected.png: Added.
  • platform/mac/fast/repaint/fixed-child-of-fixed-move-after-scroll-expected.txt: Added.
  • platform/mac/fast/repaint/fixed-child-of-transformed-move-after-scroll-expected.checksum: Added.
  • platform/mac/fast/repaint/fixed-child-of-transformed-move-after-scroll-expected.png: Added.
  • platform/mac/fast/repaint/fixed-child-of-transformed-move-after-scroll-expected.txt: Added.
  • platform/mac/fast/repaint/fixed-move-after-scroll-expected.checksum: Added.
  • platform/mac/fast/repaint/fixed-move-after-scroll-expected.png: Added.
  • platform/mac/fast/repaint/fixed-move-after-scroll-expected.txt: Added.

2010-04-21 Benjamin Poulain <ikipou@gmail.com>

Reviewed by Simon Fraser.

Update of fixed elements is not made correctly when the page has been scrolled
https://bugs.webkit.org/show_bug.cgi?id=36783

When a fixed element was updated, the old geometry was not repainted correctly
because the repaint rect was cached during the layout and not updated when
scrolling.

The rect is now updated while scrolling so the region updated correspond to the
region of the element on the screen.

The method RenderLayer::updateRepaintRectsAfterScroll() updates
the repaint rect of all fixed tree after scroll.

Tests: fast/repaint/fixed-child-move-after-scroll.html

fast/repaint/fixed-child-of-fixed-move-after-scroll.html
fast/repaint/fixed-child-of-transformed-move-after-scroll.html
fast/repaint/fixed-move-after-scroll.html

  • page/FrameView.cpp: (WebCore::FrameView::scrollPositionChanged):
  • rendering/RenderLayer.cpp: (WebCore::RenderLayer::updateRepaintRectsAfterScroll):
  • rendering/RenderLayer.h:
05:32 Changeset [57970] by eric@webkit.org

2010-04-21 Gustavo Sverzut Barbieri <barbieri@profusion.mobi>

Reviewed by Adam Roben.

Update EFL port to match recent API changes.
http://webkit.org/b/37853

No behavior changes, so no new tests were added.

  • platform/efl/FileSystemEfl.cpp:
  • platform/efl/LocalizedStringsEfl.cpp: (WebCore::missingPluginText): (WebCore::crashedPluginText):
  • platform/efl/MIMETypeRegistryEfl.cpp: (WebCore::MIMETypeRegistry::isApplicationPluginMIMEType):
  • platform/graphics/efl/ImageEfl.cpp:

2010-04-21 Gustavo Sverzut Barbieri <barbieri@profusion.mobi>

Reviewed by Adam Roben.

Update EFL port to match recent API changes.
http://webkit.org/b/37853

  • efl/WebCoreSupport/EditorClientEfl.cpp:
  • efl/WebCoreSupport/FrameLoaderClientEfl.cpp: (WebCore::FrameLoaderClientEfl::didTransferChildFrameToNewDocument): (WebCore::FrameLoaderClientEfl::objectContentType): (WebCore::FrameLoaderClientEfl::dispatchDidChangeIcons): (WebCore::FrameLoaderClientEfl::canShowMIMEType):
  • efl/WebCoreSupport/FrameLoaderClientEfl.h:
05:22 Changeset [57969] by eric@webkit.org

2010-04-21 Avi Drissman <avi@chromium.org>

Reviewed by Simon Fraser.

JPG images fail to print in Chromium
https://bugs.webkit.org/show_bug.cgi?id=37796

Image sources of JPG data with final=false fail to draw into PDF contexts even if later updated (<rdar://problem/7874035>). Therefore, destroy and re-create the image source when the final data arrives.

  • platform/graphics/cg/ImageSourceCG.cpp: (WebCore::ImageSource::setData):
05:11 Changeset [57968] by eric@webkit.org

2010-04-21 Gustavo Sverzut Barbieri <barbieri@profusion.mobi>

Reviewed by Nikolas Zimmermann.

Add missing EFL JavaScriptCore file.
http://webkit.org/b/37854

  • wtf/efl: Added.
  • wtf/efl/MainThreadEfl.cpp: Added. (WTF::initializeMainThreadPlatform): (WTF::timeoutFired): (WTF::scheduleDispatchFunctionsOnMainThread):
05:00 Changeset [57967] by eric@webkit.org

2010-04-21 Marcus Bulach <bulach@chromium.org>

Reviewed by Nate Chapin.

EventSource needs to be marked as an ActiveDomType.
https://bugs.webkit.org/show_bug.cgi?id=37857
Existing layout tests (fast/eventsource and http/tests/eventsource/) should pass when compiling with eventsource enabled.

  • bindings/scripts/CodeGeneratorV8.pm:
  • bindings/v8/V8DOMWrapper.cpp:
  • bindings/v8/custom/V8EventSourceConstructor.cpp: (WebCore::V8EventSource::constructorCallback):
04:14 Changeset [57966] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed, rolling out r57963.
http://trac.webkit.org/changeset/57963
https://bugs.webkit.org/show_bug.cgi?id=37759

Three tests started crashing on the Qt bot.

  • platform/qt/Skipped:

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed, rolling out r57963.
http://trac.webkit.org/changeset/57963
https://bugs.webkit.org/show_bug.cgi?id=37759

Three tests started crashing on the Qt bot.

  • WebCoreSupport/DumpRenderTreeSupportQt.cpp: (DumpRenderTreeSupportQt::isCommandEnabled):
  • WebCoreSupport/DumpRenderTreeSupportQt.h:

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed, rolling out r57963.
http://trac.webkit.org/changeset/57963
https://bugs.webkit.org/show_bug.cgi?id=37759

Three tests started crashing on the Qt bot.

  • DumpRenderTree/qt/LayoutTestControllerQt.cpp:
  • DumpRenderTree/qt/LayoutTestControllerQt.h:
04:02 Changeset [57965] by eric@webkit.org

2010-04-21 Tor Arne Vestbø <tor.arne.vestbo@nokia.com>

Reviewed by Eric Seidel.

Make new-run-webkit-tests work for the Qt port
https://bugs.webkit.org/show_bug.cgi?id=37588

  • Scripts/webkitpy/layout_tests/port/qt.py:
03:57 Changeset [57964] by eric@webkit.org

2010-04-21 Tor Arne Vestbø <tor.arne.vestbo@nokia.com>

Reviewed by Eric Seidel.

new-run-webkit-tests: try to detect alternate apache path
https://bugs.webkit.org/show_bug.cgi?id=37587

_check_port_build() also needs to return true in the
base implementation to not fail the check_build step.

  • Scripts/webkitpy/layout_tests/port/webkit.py:
03:54 Changeset [57963] by eric@webkit.org

2010-04-21 Yi Shen <yi.4.shen@nokia.com>

Reviewed by Simon Hausmann.

[Qt] Unskip a bunch of now passing tests, which was skipped because of
missing LayoutTestController::computedStyleIncludingVisitedInfo
https://bugs.webkit.org/show_bug.cgi?id=37759

  • platform/qt/Skipped:

2010-04-21 Yi Shen <yi.4.shen@nokia.com>

Reviewed by Simon Hausmann.

[Qt] Add LayoutTestController interface: computedStyleIncludingVisitedInfo
https://bugs.webkit.org/show_bug.cgi?id=37759

  • WebCoreSupport/DumpRenderTreeSupportQt.cpp: (DumpRenderTreeSupportQt::computedStyleIncludingVisitedInfo):
  • WebCoreSupport/DumpRenderTreeSupportQt.h:

2010-04-21 Yi Shen <yi.4.shen@nokia.com>

Reviewed by Simon Hausmann.

[Qt] Add LayoutTestController interface: computedStyleIncludingVisitedInfo
https://bugs.webkit.org/show_bug.cgi?id=37759

  • DumpRenderTree/qt/LayoutTestControllerQt.cpp: (LayoutTestController::computedStyleIncludingVisitedInfo):
  • DumpRenderTree/qt/LayoutTestControllerQt.h:
03:44 Changeset [57962] by eric@webkit.org

2010-04-21 Gustavo Sverzut Barbieri <barbieri@profusion.mobi>

Reviewed by Nikolas Zimmermann.

Add missing includes to platform/posix/FileSystemPOSIX.cpp
https://bugs.webkit.org/show_bug.cgi?id=37861

No behavior changes, so no new tests were added.

  • platform/posix/FileSystemPOSIX.cpp:
03:09 Changeset [57961] by eric@webkit.org

2010-04-21 No'am Rosenthal <noam.rosenthal@nokia.com>

Reviewed by Simon Fraser.

[Qt] Fix or remove the runtime flag for accelerated compositing.

This adds a way for a chrome client to disallow layers from becoming composited,
even if the settings enable accelerated compositing. This is necessary for platforms
where different views can be applied with the same settings to the same page.

We enable an API through ChromeClient to ask the chrome-client whether or not
it can render composited layers, which is taken into account when the compositor
decides whether or not to start compositing.

https://bugs.webkit.org/show_bug.cgi?id=37313

Pages under LayoutTests/compositing now work under QWebView, even when
QWebSettings::AcceleratedCompositingEnabled is on.

  • page/ChromeClient.h: (WebCore::ChromeClient::allowsAcceleratedCompositing):
  • platform/qt/QWebPageClient.h: (QWebPageClient::allowsAcceleratedCompositing):
  • rendering/RenderLayerCompositor.cpp: (WebCore::RenderLayerCompositor::canBeComposited):

2010-04-21 No'am Rosenthal <noam.rosenthal@nokia.com>

Reviewed by Simon Fraser.

[Qt] Fix or remove the runtime flag for accelerated compositing.
https://bugs.webkit.org/show_bug.cgi?id=37313

This lets the QWebPageClient "veto" the settings value for accelerated compositing.
In this case we allow accelerated compositing only on QGraphicsWebView.

  • Api/qgraphicswebview.cpp: (QGraphicsWebViewPrivate::allowsAcceleratedCompositing):
  • WebCoreSupport/ChromeClientQt.cpp: (WebCore::ChromeClientQt::allowsAcceleratedCompositing):
  • WebCoreSupport/ChromeClientQt.h:
02:48 Changeset [57960] by eric@webkit.org

2010-04-21 Eric Seidel <eric@webkit.org>

Unreviewed, test fix only.

new-run-webkit-tests: implement a --log trace message to be able to display detailed output of an individual test run
https://bugs.webkit.org/show_bug.cgi?id=37726

This change seems to have broken a test.
Attempting to handle the case where we don't have any
timing information. Dirk may have to correct this change.

  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
01:24 Changeset [57959] by rniwa@webkit.org

No review. Spurious whitespace was removed from project file.

  • WebCore.xcodeproj/project.pbxproj:
00:57 Changeset [57958] by alex@webkit.org

2010-04-21 Alejandro G. Castro <alex@igalia.com>

Rubber-stamped by Xan Lopez.

Skipping fast/frames/frame-crash-with-page-cache.html
https://bugs.webkit.org/show_bug.cgi?id=37914

  • platform/gtk/Skipped:
00:46 Changeset [57957] by rniwa@webkit.org

[Qt] Build fix: warning on L933 of CompositeEditCommand.cpp
https://bugs.webkit.org/show_bug.cgi?id=37912

Reviewed by Eric Seidel.

Replaced the ternary operator by an if statement because GCC was confused by the use of
ternary operator and producing warnings on Qt builds.

  • WebCore.xcodeproj/project.pbxproj:
  • editing/CompositeEditCommand.cpp:

(WebCore::CompositeEditCommand::moveParagraphs):

Note: See TracTimeline for information about the timeline view.