Timeline


and

07/10/08:

23:10 Changeset [35121] by mrowe@apple.com

Define WEBKIT_VERSION_MIN_REQUIRED=WEBKIT_VERSION_LATEST when building WebKit to ensure that no symbols end up with the weak_import attribute.

Reviewed by Sam Weinig.

23:10 Changeset [35120] by mrowe@apple.com

Fix the Tiger build by omitting annotations from methods declared in categories when using old versions of GCC.

Reviewed by Sam Weinig.

22:15 Changeset [35119] by mrowe@apple.com

Build fix.

22:02 Changeset [35118] by sfalken@apple.com

Build fix.

21:54 Changeset [35117] by dino@apple.com

Reviewed by hyatt.

Calculate computed style for -webkit-transform property
https://bugs.webkit.org/show_bug.cgi?id=19864

Test: css3/transform-computed-style-001.html

  • css/CSSComputedStyleDeclaration.cpp:

(WebCore::computedTransform):
(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):

handles -webkit-transform property now

  • manual-tests/computed-transform-value.html: Added.
18:05 Changeset [35116] by mrowe@apple.com

Versioning.

18:05 Changeset [35115] by mrowe@apple.com

Merge r35112.

18:03 Changeset [35114] by mrowe@apple.com

New tag.

17:46 Changeset [35113] by mitz@apple.com

Reviewed by Adam Roben.

  • fix a parse error in inspector.js
  • page/inspector/inspector.js: Added missing brace.
17:43 Changeset [35112] by aroben@apple.com

Build fix

  • WebCore.vcproj/QTMovieWin.vcproj: Add OSXCompatibilityHeadersGNUCompatibility to the include path.
17:28 Changeset [35111] by cfleizach@apple.com

<rdar://problem/6067408> AX: internal anchors broken

Fixes a broken accessibility test

17:01 Changeset [35110] by mrowe@apple.com

Build fix.

16:47 Changeset [35109] by weinig@apple.com

2008-07-10 Sam Weinig <sam@webkit.org>

Reviewed by Mark Rowe.

Remove no-op debug method.

  • css/CSSSelector.cpp:
  • css/CSSSelector.h:
16:21 Changeset [35108] by weinig@apple.com

WebCore:

2008-07-10 Sam Weinig <sam@webkit.org>

Reviewed by Dave Hyatt and Darin Adler.

Add support for calling querySelector and querySelectorAll on DocumentFragments

  • Fixes bug where nodes not in the document tree would not match based on ID due to over optimization.

Test: fast/dom/SelectorAPI/detached-element.html

  • GNUmakefile.am:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • WebCoreSources.bkl:
  • bindings/js/JSDocumentFragmentCustom.cpp: Added. (WebCore::JSDocumentFragment::querySelector): Add custom code matching JSElement and JSDocument that checks for a 2nd arguments and throws an exception indicating we do not currently support the optional NSResolver part of the Selectors API spec. (WebCore::JSDocumentFragment::querySelectorAll): Ditto.
  • dom/DocumentFragment.idl: Add querySelector and querySelectorAll declarations.
  • dom/Node.cpp: (WebCore::Node::querySelector): Make the CSS parser parse the selector according to the strictness of the document so that mixed case ID selectors match in quirks mode. Also, don't use the fast ID path if the root is not in the DOM tree, as it won't work. (WebCore::Node::querySelectorAll): Ditto.
  • dom/SelectorNodeList.cpp: (WebCore::createSelectorNodeList): Don't use the fast ID path if the root is not in the DOM tree, as it won't work.

LayoutTests:

2008-07-10 Sam Weinig <sam@webkit.org>

Reviewed by Dave Hyatt and Darin Adler,

Add support for calling querySelector and querySelectorAll on DocumentFragments

  • Add test for nodes not in the document tree.
  • fast/dom/SelectorAPI/detached-element-expected.txt: Added.
  • fast/dom/SelectorAPI/detached-element.html: Added.
  • fast/dom/SelectorAPI/dumpNodeList-expected.txt:
  • fast/dom/SelectorAPI/dumpNodeList.html:
  • fast/dom/SelectorAPI/not-supported-NSResolver-expected.txt:
  • fast/dom/SelectorAPI/not-supported-NSResolver.html:
  • fast/dom/SelectorAPI/resources/detached-element.js: Added.
  • fast/dom/Window/window-properties-expected.txt:
16:06 Changeset [35107] by pewtermoose@webkit.org

2008-07-10 Anthony Ricaud <rik24d@gmail.com>

Bug 19389: querySelectorAll exception while searching invalid CSS selector
<https://bugs.webkit.org/show_bug.cgi?id=19389>

Reviewed by Tim Hatcher.

  • page/inspector/inspector.js: Added a try/catch block.
16:02 Changeset [35106] by pewtermoose@webkit.org

2008-07-10 Brent Fulgham <bfulgham@gmail.com>

Correct a build regression due to an uninitialized variable.

<https://bugs.webkit.org/show_bug.cgi?id=19976>

Reviewed by Darin.

  • platform/graphics/cairo/FontCairo.cpp: (WebCore::Font::drawGlyphs):
15:59 Changeset [35105] by aroben@apple.com

Fix Bug 19580: REGRESSION (r34432): PGO-only crash in HTMLCollection::resetCollectionInfo (codegen issue?)

<https://bugs.webkit.org/show_bug.cgi?id=19580>
<rdar://6029794>

Reviewed by Cameron Zwarich.

  • WebCore.vcproj/WebCore.vcproj: Disable LTCG for HTMLFormElement.cpp, which was causing some bad codegen in HTMLFormElement::elements.
  • html/HTMLFormElement.cpp: Touched this file to force it to rebuild.
15:53 Changeset [35104] by andersca@apple.com

2008-07-10 Anders Carlsson <andersca@apple.com>

Reviewed by Jon.

<rdar://problem/6067135>
WebKit should respond true to a query for NPNVSupportsWindowless.


Handle NPNVSupportsWindowless and return true.


  • bridge/npapi.h:
  • plugins/win/PluginViewWin.cpp: (WebCore::PluginView::getValue):
15:38 Changeset [35103] by sfalken@apple.com

Build fix.

  • DumpRenderTree/win/DumpRenderTree.vcproj:
14:56 Memory Use edited by ggaren@apple.com
(diff)
13:50 Changeset [35102] by kmccullough@apple.com

JavaScriptCore:

2008-07-10 Kevin McCullough <kmccullough@apple.com>

Reviewed by Darin.

-Minor cleanup. Renamed callTree() to head() and no longer use m_head
directly but instead keep it private and access via a method().

  • profiler/HeavyProfile.cpp: (KJS::HeavyProfile::HeavyProfile): (KJS::HeavyProfile::generateHeavyStructure): (KJS::HeavyProfile::addNode):
  • profiler/Profile.h: (KJS::Profile::head):
  • profiler/ProfileGenerator.cpp: (KJS::ProfileGenerator::ProfileGenerator):

WebCore:

2008-07-10 Kevin McCullough <kmccullough@apple.com>

Reviewed by Darin.

-Minor cleanup. Renamed callTree() to head() and no longer use m_head
directly but instead keep it private and access via a method().

  • page/JavaScriptProfile.cpp: (WebCore::getHeadCallback):
13:38 Memory Use edited by ggaren@apple.com
(diff)
13:35 Memory Use edited by ggaren@apple.com
(diff)
12:31 Changeset [35101] by dino@apple.com

When a mask image changes, ensure that elements
that use that mask image are repainted correctly.

<https://bugs.webkit.org/show_bug.cgi?id=19954>

Reviewed by Dave Hyatt

  • manual-tests/canvas-mask-redraw.html Manual testcase
  • rendering/RenderBox.cpp: (WebCore::RenderBox::imageChanged): (WebCore::RenderBox::repaintLayerRectsForImage):
  • rendering/RenderBox.h: Loop through background layers and mask layers, and if this image is used in a layer, compute a repaint rect and repaint.
  • rendering/RenderImage.cpp: (WebCore::RenderImage::imageChanged): If this image has a mask, call the base class method.
  • rendering/RenderObject.cpp: Remove a bogus 'return'.
12:05 Changeset [35100] by dino@apple.com

Transforms create a containing block, so
RenderLayer::shouldBeOverflowOnly() needs to look for transforms.
<https://bugs.webkit.org/show_bug.cgi?id=18886>

Reviewed by Dave Hyatt

Test: fast/transforms/transform-overflow.html

  • rendering/RenderLayer.cpp:

(WebCore::RenderLayer::shouldBeOverflowOnly):

11:08 Changeset [35099] by aroben@apple.com

Windows build fix

  • WebKit.vcproj/resource.h: Revert some changes that VS made to this file in r35083.
10:59 Changeset [35098] by andersca@apple.com

2008-07-10 Anders Carlsson <andersca@apple.com>

Reviewed by Mark.


Add availability macros for the new WebPlugin methods.

  • Plugins/WebPlugin.h:
  • Plugins/WebPluginViewFactory.h:
10:55 Changeset [35097] by ap@webkit.org

Reviewed by Mark Rowe.

Eliminate CollectorHeapIntrospector.

CollectorHeapIntrospector was added primarily in the hopes to improve leaks tool output,
a result that it didn't deliver. Also, it helped by labeling JSC heap regions as reported by
vmmap tool, but at the same time, it made them mislabeled as malloc'd ones - the correct
way to label mapped regions is to use a VM tag.

So, it makes more sense to remove it completely than to make it work with multiple heaps.

  • JavaScriptCore.exp:
  • JavaScriptCore.xcodeproj/project.pbxproj:
  • kjs/AllInOneFile.cpp:
  • kjs/InitializeThreading.cpp: (KJS::initializeThreading):
  • kjs/collector.cpp:
  • kjs/collector.h:
  • kjs/CollectorHeapIntrospector.cpp: Removed.
  • kjs/CollectorHeapIntrospector.h: Removed.
10:33 Changeset [35096] by kmccullough@apple.com

2008-07-09 Kevin McCullough <kmccullough@apple.com>

Reviewed by Darin.

<rdar://problem/5951532> JSProfiler: Implement heavy (or bottom-up)
view (19228)

  • Implemented the time and call count portionof heavy. Now all that we need is some UI.
  • profiler/CallIdentifier.h: Removed an unused constructor.
  • profiler/HeavyProfile.cpp: (KJS::HeavyProfile::HeavyProfile): Set the initial time of the head node so that percentages work correctly. (KJS::HeavyProfile::mergeProfiles): Sum the times and call count of nodes being merged.
  • profiler/ProfileNode.cpp: Set the intital values of time and call count when copying ProfileNodes. (KJS::ProfileNode::ProfileNode):
03:01 Changeset [35095] by jmalonzo@webkit.org

2008-07-10 Jan Michael Alonzo <jmalonzo@webkit.org>

Gtk build fix.

  • GNUmakefile.am: Add HeavyProfile.cpp
00:12 Changeset [35094] by oliver@apple.com

2008-07-09 Alex Mathews <possessedpenguinbob@gmail.com>

Reviewed by Oliver Hunt.

bug 19835: WebKit needs cross-platform filter system
<https://bugs.webkit.org/show_bug.cgi?id=19835>

More class refactoring in preparation for cross-platform filter
implementation.

07/09/08:

22:52 Changeset [35093] by mrowe@apple.com

Versioning.

22:52 Changeset [35092] by mrowe@apple.com

Merge r35089.

22:40 Changeset [35091] by mrowe@apple.com

Versioning.

22:38 Changeset [35090] by mrowe@apple.com

New tag.

22:32 Changeset [35089] by mrowe@apple.com

Don't warn about deprecated functions in production builds.

Reviewed by Geoff Garen.

21:51 Changeset [35088] by beidson@apple.com

WebCore:

2008-07-09 Brady Eidson <beidson@apple.com>

Reviewed by Darin

<rdar://problem/5823684> - Crash manipulating frame tree of a new frame before the new frame
has been installed in a frame tree.

The root of this problem was that calling init() on a new frame could end up calling arbitrary
javascript that might end up removing the frame from the tree. This opened up a small can of worms
such as the frame not having yet been installed in its frame tree, and other assumed behavior while
destroying the frame.

Test: fast/loader/frame-creation-removal.html

  • loader/FrameLoader.cpp: (WebCore::FrameLoader::endIfNotLoadingMainResource): If the frame doesn't have a page, don't close up the document and parser because they don't exist, and this frame is on its way out. (WebCore::FrameLoader::finishedParsing): We can't rely on the refCount check to discover "am I being deleted?" because we no longer store refCounts of 0. The new check is "do I have a FrameView?" while will always be false if the Frame is being destroyed.

WebKit/mac:

2008-07-09 Brady Eidson <beidson@apple.com>

Reviewed by Darin

<rdar://problem/5823684> - Crash manipulating frame tree of a new frame before the new frame
has been installed in a frame tree.

The root of this problem was that calling init() on a new frame could end up calling arbitrary
javascript that might end up removing the frame from the tree. This opened up a small can of worms
such as the frame not having yet been installed in its frame tree, and other assumed behavior while
destroying the frame.

Note that each platforms WebKit API layer needs to make this new guarantee:
"The new Frame must be installed in its FrameTree before newCoreFrame->init() is called"
I am fixing Mac, and Windows and GTK already have this property. Wx currently has subframes disabled
but will need to add this guarantee when re-enabling, and Qt is currently vulnerable to this same bug.

Alternately, the way frames are created right now is roundabout and asinine, and this is a key
architectural improvement we can make in the future so the individual platform clients are no longer
vulnerable to this problem, which should really have been a WebCore issue.

  • WebCoreSupport/WebFrameLoaderClient.mm: (WebFrameLoaderClient::createFrame): Don't bother null checking the newCoreFrame - can't be NULL. Don't appendChild() the new frame here. Null-check the new frame's page before loading the URL into it, as it might already have been removed from the page.
  • WebView/WebFrame.mm: (+[WebFrame _createFrameWithPage:frameName:frameView:ownerElement:]): If there is an ownerElement, go ahead and install the new frame in the frame tree *before* calling init() on it.

LayoutTests:

2008-07-09 Brady Eidson <beidson@apple.com>

Reviewed by Darin

Test for <rdar://problem/5823684>

  • fast/loader/frame-creation-removal-expected.txt: Added.
  • fast/loader/frame-creation-removal.html: Added.
  • fast/loader/resources/frame-creation-removal-iframe.html: Added.
21:32 Changeset [35087] by darin@apple.com

2008-07-09 Darin Adler <darin@apple.com>

20:33 Memory Use edited by ggaren@apple.com
(diff)
20:31 WikiStart edited by ggaren@apple.com
(diff)
20:31 Memory Use edited by ggaren@apple.com
(diff)
20:30 Memory Use created by ggaren@apple.com
20:11 Changeset [35086] by dino@apple.com

Changed to use the correct license in header comment (via Darin)

  • css/WebKitCSSTransformValue.idl:
18:39 Changeset [35085] by kevino@webkit.org

wx biuld fix. Add HeavyProfile.cpp to build files.

18:02 Changeset [35084] by dino@apple.com

Add DOM interface for WebKitCSSTransformValue.
https://bugs.webkit.org/show_bug.cgi?id=19863

Reviewed by Hyatt.

  • bindings/objc/DOMInternal.h:
  • css/WebKitCSSTransformValue.idl: Added.
  • bindings/scripts/CodeGeneratorObjC.pm:

make sure new class inherits from CSSValue not Node

  • DerivedSources.make:
  • GNUmakefile.am:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • WebCoreSources.bkl:

Adding new generated files

17:22 Changeset [35083] by britto@apple.com

2008-07-09 Maxime Britto <britto@apple.com>

Reviewed by NOBODY (OOPS!).

http://bugs.webkit.org/show_bug.cgi?id=14227
Add the middle click panning feature to the windows release.
Details on almost each method below.
Manual test is included in the patch.

  • ChangeLog:
  • WebCore.vcproj/WebCore.vcproj:
  • manual-tests/panScroll.html: Added.
  • manual-tests/resources/big-page.html: Added.
  • page/EventHandler.cpp: Added the panScroll start/stop handlers and adapted the autoscroll for both to share some functions. (WebCore::EventHandler::EventHandler): Initialize the new class members for the panScroll (WebCore::EventHandler::handleMouseDraggedEvent): Prevent the autoscroll to trigger if the panScroll is in progress (WebCore::EventHandler::handleMouseReleaseEvent): Prevent the autoscroll to stop the panScroll on mouse release (WebCore::EventHandler::handleAutoscroll): Added the specific calls for the panScroll (save mouse position, draw the panScroll icon) to the shared code (WebCore::EventHandler::autoscrollTimerFired): Specify the calls to perform for each function (autoscroll and panScroll) (WebCore::EventHandler::stopAutoscrollTimer): Specify the calls to perform for each function (autoscroll and panScroll) (WebCore::EventHandler::handleMousePressEvent): Trigger for the panScroll. Test is the button is the middle button and try to find a renderer where the panScroll is possible. If it finds one, it calls the handleAutoscroll method.
  • page/EventHandler.h: (WebCore::EventHandler::panScrollInProgress): Getter used by the sub frames's EventHandler to notice the main frame EventHandler (WebCore::EventHandler::setPanScrollInProgress): Setter used by the sub frames's EventHandler to notice the main frame EventHandler
  • platform/ScrollView.h: Added a method used by the RenderObject to test if the view can be scrolled. Added the print/remove panScroll icon methods between the windows platform flags
  • platform/gtk/ScrollViewGtk.cpp: (WebCore::ScrollView::isScrollable): Not implemented : returns true
  • platform/mac/ScrollViewMac.mm: Implemented because it was useful for the autoscroll feature. (WebCore::ScrollView::isScrollable):
  • platform/qt/ScrollViewQt.cpp: (WebCore::ScrollView::isScrollable): Not implemented : returns true
  • platform/win/ScrollViewWin.cpp: (WebCore::ScrollView::ScrollViewPrivate::ScrollViewPrivate): Added a boolean and an IntPoint to draw the icon. (WebCore::ScrollView::ScrollViewPrivate::scrollBackingStore): If the icon must be drawn, define the rect to invalidate to erase the previous icon and to display the new one. (WebCore::ScrollView::updateContents): I split this method to be able to invalidate a rect with window coordinates instead of contents coordinates. This function now converts the contents coord in window coord and calls the new updateWindowRect() method (WebCore::ScrollView::updateWindowRect): New method which performs the work that was doing the end of the old updateContents : invalidate a rect with window coordinates (WebCore::ScrollView::isScrollable): checks is the view has scrollbars (WebCore::ScrollView::printPanScrollIcon): sets the boolean and the IntPoint for the icon. Invalidate his rect and ask for repaint. (WebCore::ScrollView::removePanScrollIcon): sets the boolean to false. Invalidate his rect and ask for repaint. (WebCore::ScrollView::paint): Modified to draw the icon if needed.
  • platform/wx/ScrollViewWx.cpp: (WebCore::ScrollView::isScrollable): Not implemented : returns true
  • rendering/RenderLayer.cpp: (WebCore::RenderLayer::panScrollFromPoint): Receives the Middle click location and retrieves the currentMouse position from the EventHandler. With these informations it computes the direction to scroll to and the speed then calls for the final scroll.
  • rendering/RenderLayer.h:
  • rendering/RenderListBox.cpp: Adapt the panScroll to the ListBox object to scroll by lines instead of pixels (WebCore::RenderListBox::panScroll): (WebCore::RenderListBox::scrollToward): (WebCore::RenderListBox::autoscroll):
  • rendering/RenderListBox.h: (WebCore::RenderListBox::shouldPanScroll):
  • rendering/RenderObject.cpp: (WebCore::RenderObject::shouldAutoscroll): Improved the verification to avoid triggering the autoscroll/panScroll when the root object can't scroll (WebCore::RenderObject::panScroll): Calls the RenderLayer's panScrollFromPoint() method
  • rendering/RenderObject.h: (WebCore::RenderObject::stopPanScroll):

2008-07-09 Maxime Britto <britto@apple.com>

Reviewed by NOBODY (OOPS!).


Added the panning icon in the ressources.
Its name is compass.png to make a difference with moveCursor.png (from the WebCore ressources)

  • ChangeLog:
  • WebKit.vcproj/WebKit.rc:
  • WebKit.vcproj/panIcon.png: Added.
  • WebKit.vcproj/resource.h:
  • WebKitDLL.cpp: Declared the new ressource with the name "panIcon" (loadResourceIntoBuffer):
17:17 Changeset [35082] by eric@webkit.org

Reviewed by bdash.

Fix copying of expected results to correct location

  • Scripts/run-webkit-tests:
17:12 Changeset [35081] by dino@apple.com

Rename CSSTransformValue to WebKitCSSTransformValue as it is non-standard for
the moment. Also, WebKitCSSTransformValue is a CSSValueList (comma sep)
https://bugs.webkit.org/show_bug.cgi?id=19861

Reviewed by Hyatt

  • css/CSSParser.cpp:
  • css/CSSStyleSelector.cpp:
  • css/CSSTransformValue.cpp: Removed.
  • css/CSSTransformValue.h: Removed.
  • css/CSSValueList.h:
  • css/WebKitCSSTransformValue.cpp: Added.
  • css/WebKitCSSTransformValue.h: Added.
  • GNUmakefile.am:
  • WebCore.order:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • WebCoreSources.bkl:

updated for new file names

16:37 Changeset [35080] by kmccullough@apple.com

2008-07-09 Kevin McCullough <kmccullough@apple.com>

  • Windows build fix.
16:24 Changeset [35079] by kmccullough@apple.com

2008-07-09 Kevin McCullough <kmccullough@apple.com>

  • Build fix.
  • profiler/HeavyProfile.cpp: (KJS::HeavyProfile::mergeProfiles):
16:17 Changeset [35078] by eric@webkit.org

Updated printing test patch

16:07 Changeset [35077] by kmccullough@apple.com

2008-07-09 Kevin McCullough <kmccullough@apple.com>

Reviewed by Geoff and Adam.

<rdar://problem/5951532> JSProfiler: Implement Bottom-Up view (19228)

  • This is the plumbing for bottom-up, but does not include calculating time, mostly because I'm still undclear about what the end result should look like.
  • This, obviously, does not include the UI to expose this in the inspector yet.
  • JavaScriptCore.xcodeproj/project.pbxproj:
  • profiler/CallIdentifier.h: (KJS::CallIdentifier::CallIdentifier): (WTF::): Added HashTraits for CallIdentifiers to be used by a HashMap.
  • profiler/HeavyProfile.cpp: Added. (KJS::HeavyProfile::HeavyProfile): (KJS::HeavyProfile::generateHeavyStructure): (KJS::HeavyProfile::addNode): (KJS::HeavyProfile::mergeProfiles): (KJS::HeavyProfile::addAncestorsAsChildren):
  • profiler/HeavyProfile.h: Added. (KJS::HeavyProfile::create): (KJS::HeavyProfile::heavyProfile): (KJS::HeavyProfile::treeProfile):
  • profiler/Profile.cpp: Removed old commented out includes.
  • profiler/Profile.h: The m_head is needed by the HeavyProfile so it is now protected as opposed to private.
  • profiler/ProfileNode.cpp: (KJS::ProfileNode::ProfileNode): Created a constructor to copy ProfileNodes. (KJS::ProfileNode::findChild): Added a null check to make HeavyProfile children finding easier and avoid a potential crasher.
  • profiler/ProfileNode.h: Mostly moved things around but also added some functionality needed by HeavyProfile. (KJS::ProfileNode::create): (KJS::ProfileNode::functionName): (KJS::ProfileNode::url): (KJS::ProfileNode::lineNumber): (KJS::ProfileNode::head): (KJS::ProfileNode::setHead): (KJS::ProfileNode::setNextSibling): (KJS::ProfileNode::actualTotalTime): (KJS::ProfileNode::actualSelfTime):
  • profiler/TreeProfile.cpp: Implemented the ability to get a HeavyProfile. (KJS::TreeProfile::heavyProfile):
  • profiler/TreeProfile.h:
15:42 Changeset [35076] by eric@webkit.org

Reviewed by weinig.

Add Content-Type support to DumpRenderTree and run-webkit-tests
and move expected.txt files to expected.webarchive

https://bugs.webkit.org/show_bug.cgi?id=15565

  • DumpRenderTree/cg/ImageDiffCG.cpp: (compareImages): (main):
  • DumpRenderTree/cg/PixelDumpSupportCG.cpp: (printPNG):
  • DumpRenderTree/mac/DumpRenderTree.mm: (dump):
  • Scripts/run-webkit-tests:
14:30 Changeset [35075] by adele@apple.com

Fixing my typo from the last checkin

14:11 Changeset [35074] by hyatt@apple.com

2008-07-09 David Hyatt <hyatt@apple.com>

Switch transitions back to a "destination" model as far as choosing which transitions should apply on a style
change. Preserve the behavior of allowing stale transitions (in the absence of property changes) to run to
completion.

Reviewed by Dean

  • manual-tests/transitions.html:
  • manual-tests/transitions2.html:
  • page/AnimationController.cpp: (WebCore::ImplicitAnimation::reset): (WebCore::CompositeImplicitAnimation::animate): (WebCore::AnimationControllerPrivate::get): (WebCore::AnimationController::updateImplicitAnimations):
13:46 Changeset [35073] by adele@apple.com

2008-07-09 Michelangelo De Simone <m.des@mac.com>

Reviewed by Adele.

Added the virtual function Node::isTextControl() in order to simplify text field and textarea checks.

  • css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::SelectorChecker::checkOneSelector): Clean up in PseudoReadOnly and PseudoReadWrite cases removing unnecessary checks and casts in favor of the sole isTextControl() check.
  • dom/Node.h: Added base isTextControl().
  • html/HTMLInputElement.h: Added isTextControl() which wraps HTMLInputElement::isTextField().
  • html/HTMLTextAreaElement.h: Added isTextControl().
13:36 Changeset [35072] by rwlbuis@webkit.org

Reviewed by Eric.

https://bugs.webkit.org/show_bug.cgi?id=15431
SVGRenderStyle should store pre-modified resource URIs

Store pre-modified resource URIs to decrease string
operations during layout/rendering.

12:55 Changeset [35071] by andersca@apple.com

2008-07-09 Anders Carlsson <andersca@apple.com>

Reviewed by Darin.

Make some WebPlugin and WebPluginFactory SPI public.


  • Plugins/WebPlugin.h:
  • Plugins/WebPluginPrivate.h:
  • Plugins/WebPluginViewFactory.h:
  • Plugins/WebPluginViewFactoryPrivate.h:
11:35 Changeset [35070] by andersca@apple.com

2008-07-09 Anders Carlsson <andersca@apple.com>

Reviewed by Mitz.

Remove an unused instance variable.

  • loader/DocumentLoader.h:
10:23 Changeset [35069] by mitz@apple.com

Reviewed by Anders Carlsson.

  • remove unused #includes
  • dom/XMLTokenizer.cpp:
  • html/PreloadScanner.cpp:
  • loader/CachedCSSStyleSheet.cpp:
  • loader/CachedScript.cpp:
  • loader/CachedXBLDocument.cpp:
  • loader/CachedXSLStyleSheet.cpp:
  • page/mac/FrameMac.mm:
  • xml/XSLTProcessor.cpp:

07/08/08:

23:28 Changeset [35068] by ggaren@apple.com

JavaScriptCore:

2008-07-08 Geoffrey Garen <ggaren@apple.com>

Reviewed by Oliver Hunt.


Added support for checking if an object has custom properties in its
property map. WebCore uses this to optimize marking DOM wrappers.

WebCore:

2008-07-08 Geoffrey Garen <ggaren@apple.com>

Reviewed by Oliver Hunt.


Only artificially mark JS DOM wrappers if they have custom properties.


21X speedup on http://nerget.com/jstests/dom-mandelbrot.html.


No, that is not a typo.

  • bindings/js/JSDOMBinding.cpp: (WebCore::ScriptInterpreter::markDOMNodesForDocument):
16:46 Changeset [35067] by jhoneycutt@apple.com

2008-07-08 Jon Honeycutt <jhoneycutt@apple.com>

Reviewed by Anders.

Port r34988 to Mac: don't call NPP_DestroyStream if NPP_NewStream was
unsuccessful.

  • Plugins/WebBaseNetscapePluginStream.h: Added new member, newStreamSuccessful.
  • Plugins/WebBaseNetscapePluginStream.mm: (-[WebBaseNetscapePluginStream initWithRequestURL:plugin:notifyData:sendNotification:]): Initialize new member. (-[WebBaseNetscapePluginStream startStreamResponseURL:expectedContentLength:lastModifiedDate:MIMEType:headers:]): If NPP_NewStream is successful, set newStreamSuccessful to YES. (-[WebBaseNetscapePluginStream _destroyStream]): Only call NPP_DestroyStream if newStreamSuccessful is true.

Port r34988 to Mac: allow tests to define JavaScript to execute when
NPP_DestroyStream or NPP_URLNotify is called.

  • DumpRenderTree/TestNetscapePlugIn.subproj/main.cpp: (NPP_New): Remove initialization that happens in pluginAllocate. Look for new arguments onStreamDestroy and onURLNotify. (NPP_Destroy): Free onStreamDestroy and onURLNotify. (executeScript): Code moved from onStreamLoad. (NPP_NewStream): Call executeScript. (NPP_DestroyStream): Same. (NPP_URLNotify): Same.
15:32 Changeset [35066] by kevino@webkit.org

Reviewed by Kevin Ollivier.

Fix to previous patch for handling mouse up events.

https://bugs.webkit.org/show_bug.cgi?id=18464

13:20 Changeset [35065] by mitz@apple.com

Reviewed by Brady Eidson.

  • initialize the Archives log channel's state
  • platform/mac/LoggingMac.mm: (WebCore::InitializeLoggingChannelsIfNecessary):
11:45 Changeset [35064] by dino@apple.com

2008-07-08 Simon Fraser <simon.fraser@apple.com>

Reviewed by Darin.

Dean forgot to add this file needed for the tests.

  • traversal/resources/node-iterator-009-data.xml: Added.
11:38 Changeset [35063] by kmccullough@apple.com

2008-07-08 Kevin McCullough <kmccullough@apple.com>

Added manual test for the new heavy view.

  • manual-tests/inspector/profiler-test-heavy-view.html: Added.
09:40 Changeset [35062] by darin@apple.com

2008-07-08 Darin Adler <darin@apple.com>

Reviewed by Mitz.

  • fix <rdar://problem/6043731> REGRESSION (3-4): RedEnvelope.com looks wrong due to OpenCube QuickMenu appVersion check
  • page/Navigator.cpp: (WebCore::shouldHideFourDot): Add dqm_loader.js as another filename.
09:24 Changeset [35061] by mitz@apple.com

WebCore:

Reviewed by John Sullivan.

  • WebCore part of <rdar://problem/6008409> Need a way to disable updates in offscreen views

Added a setting, updatesWhenOffscreen(), which controls whether an
offscreen web view gets marked as needing update when its contents
change (the existing behavior) or not (a new behavior), in which case it
will be marked as needing update just before it goes on screen. The
existing behavior (updating while offscreen) remains the default.

  • WebCore.base.exp: Added Settings::setUpdatesWhenOffscreen().
  • page/FrameView.cpp: (WebCore::FrameView::shouldUpdateWhenOffscreen): Added. Returns the value from settings.
  • page/FrameView.h:
  • page/Settings.cpp: (WebCore::Settings::setUpdatesWhenOffscreen): Added.
  • page/Settings.h: (WebCore::Settings::updatesWhenOffscreen): Added.
  • platform/ScrollView.h:
  • platform/mac/ScrollViewMac.mm: (WebCore::ScrollView::updateContents): Added code to return early and not call -setNeedsDisplayInRect: if the view is not in a visible window and the setting is not to update when offscreen.

WebKit/mac:

Reviewed by John Sullivan.

  • WebKit part of <rdar://problem/6008409> Need a way to disable updates in offscreen views
  • WebView/WebHTMLView.mm: (-[WebHTMLView addWindowObservers]): Added code to observe when the window goes onscreen. (-[WebHTMLView removeWindowObservers]): Added. (-[WebHTMLView windowWillOrderOnScreen:]): Added. If the view is set to not update when offscreen, calls -setNeedsDisplay: just before it comes onscreen.
  • WebView/WebPreferenceKeysPrivate.h: Added preference key.
  • WebView/WebPreferences.m: (+[WebPreferences initialize]): Made updates when offscreen on by default. (-[WebPreferences updatesWhenOffscreen]): Added. (-[WebPreferences setUpdatesWhenOffscreen:]): Added.
  • WebView/WebPreferencesPrivate.h:
  • WebView/WebView.mm: (-[WebView _preferencesChangedNotification:]): Added code to update the updatesWhenOffscreen setting in WebCore.

WebKitLibraries:

Reviewed by John Sullivan.

  • WebKitSystemInterface part of <rdar://problem/6008409> Need a way to disable updates in offscreen views
  • WebKitSystemInterface.h: Added WKWindowWillOrderOnScreenNotification.
  • libWebKitSystemInterfaceLeopard.a: Updated.
  • libWebKitSystemInterfaceTiger.a: Updated.
08:45 Changeset [35060] by ap@webkit.org

Reviewed by Darin.

<rdar://problem/6058368> REGRESSION: Crash occurs in KJS::Machine after launching
Apple Installer

The problem was that JSUserObject::getCallData() was setting call type to native, but
left callData.native.function uninitialized.

  • JSObject.cpp: (JSUserObject::getCallData): Initialize callData.native.function. (nativeCallFunction): Call UserObjectImp::callAsFunction().
01:29 Changeset [35059] by hausmann@webkit.org

Prospective Gtk/Wx build fixes, add ProfileGenerator.cpp to the build.

01:22 Changeset [35058] by hausmann@webkit.org

Fix the Qt build and the SVG filter enabled build in general.

07/07/08:

21:17 Changeset [35057] by weinig@apple.com

2008-07-07 Sam Weinig <sam@webkit.org>

Reviewed by Anders Carlsson.

Remove extraneous null check.

  • dom/Node.cpp: (WebCore::Node::querySelector): (WebCore::Node::querySelectorAll):
21:15 Changeset [35056] by weinig@apple.com

WebCore:

2008-07-07 Sam Weinig <sam@webkit.org>

Reviewed by Anders Carlsson.

Fix for https://bugs.webkit.org/show_bug.cgi?id=19928
Use of CSS3 Namespaces syntax in Selectors API should throw an exception

  • Throw an NAMESPACE_ERR exception if a non-null or "*" namespace is used in a selector passed to querySelector or querySelectorAll.

Test: fast/dom/SelectorAPI/not-supported-namespace-in-selector.html

  • dom/Node.cpp: (WebCore::selectorNeedsNamespaceResolution): (WebCore::Node::querySelector): (WebCore::Node::querySelectorAll):

LayoutTests:

2008-07-07 Sam Weinig <sam@webkit.org>

Reviewed by Anders Carlsson.

Fix for https://bugs.webkit.org/show_bug.cgi?id=19928
Use of CSS3 Namespaces syntax in Selectors API should throw an exception

  • fast/dom/SelectorAPI/not-supported-namespace-in-selector-expected.txt: Added.
  • fast/dom/SelectorAPI/not-supported-namespace-in-selector.html: Added.
20:30 Changeset [35055] by ddkilzer@apple.com

releaseFastMallocFreeMemory() should always be defined

Reviewed by Darin.

  • JavaScriptCore.exp: Changed to export C++ binding for WTF::releaseFastMallocFreeMemory() instead of C binding for releaseFastMallocFreeMemory().
  • wtf/FastMalloc.cpp: Moved definitions of releaseFastMallocFreeMemory() to be in the WTF namespace regardless whether FORCE_SYSTEM_MALLOC is defined.
  • wtf/FastMalloc.h: Moved releaseFastMallocFreeMemory() from extern "C" binding to WTF::releaseFastMallocFreeMemory().
20:19 Changeset [35054] by dino@apple.com

2008-07-07 Simon Fraser <simon.fraser@apple.com>

Reviewed by Darin.

Fix for https://bugs.webkit.org/show_bug.cgi?id=19933
nodeIterator with filter fails on documents not in a frame

Tests: traversal/node-iterator-009.html

traversal/tree-walker-006.html

  • bindings/js/JSNodeFilterCondition.cpp:
  • bindings/js/JSNodeFilterCondition.h:
  • bindings/js/JSNodeFilterCustom.cpp:
  • bindings/js/JSNodeIteratorCustom.cpp:
  • bindings/js/JSTreeWalkerCustom.cpp:
  • bindings/objc/DOM.mm:
  • dom/NodeFilter.cpp:
  • dom/NodeFilter.h:
  • dom/NodeFilterCondition.cpp:
  • dom/NodeFilterCondition.h:
  • dom/NodeIterator.cpp:
  • dom/NodeIterator.h:
  • dom/Traversal.cpp:
  • dom/Traversal.h:
  • dom/TreeWalker.cpp:
  • dom/TreeWalker.h:


18:25 Changeset [35053] by cwzwarich@webkit.org

2008-07-07 Cameron Zwarich <cwzwarich@uwaterloo.ca>

Reviewed by Geoff.

Bug 19926: URL causes crash within a minute
<https://bugs.webkit.org/show_bug.cgi?id=19926>

Add a check that lastGlobalObject is non-null in Machine::execute()
before copying its globals to the current register file.

In theory, it is possible to make a test case for this, but it will
take a while to get it right.

  • VM/Machine.cpp: (KJS::Machine::execute):
18:13 Changeset [35052] by adele@apple.com

WebCore:

2008-07-07 Adele Peterson <adele@apple.com>

Reviewed by Dan Bernstein.

Fix for https://bugs.webkit.org/show_bug.cgi?id=19924
<rdar://problem/6057160> Disabled file input element updates its displayed file info when a file is dropped from the Finder.

  • page/DragController.cpp: (WebCore::DragController::concludeDrag): Check that the file control is enabled before taking dropped files.

LayoutTests:

2008-07-07 Adele Peterson <adele@apple.com>

Reviewed by Dan Bernstein.

Test for https://bugs.webkit.org/show_bug.cgi?id=19924
<rdar://problem/6057160> Disabled file input element updates its displayed file info when a file is dropped from the Finder.

  • fast/forms/dragging-to-disabled-file-input-expected.txt: Added.
  • fast/forms/dragging-to-disabled-file-input.html: Added.
17:46 Changeset [35051] by adele@apple.com

WebCore:

2008-07-07 Adele Peterson <adele@apple.com>

Reviewed by Beth.

Fix for <rdar://problem/5860507> <AUDIO> playback noticeably more quiet than QuickTime

Change default volume setting to 1.0.

  • html/HTMLMediaElement.cpp: (WebCore::HTMLMediaElement::HTMLMediaElement):
  • platform/graphics/MediaPlayer.cpp: (WebCore::MediaPlayer::MediaPlayer):

LayoutTests:

2008-07-07 Adele Peterson <adele@apple.com>

Reviewed by Beth.

Test for <rdar://problem/5860507> <AUDIO> playback noticeably more quiet than QuickTime

  • media/video-volume-expected.txt:
  • media/video-volume.html:
17:29 Changeset [35050] by darin@apple.com

2008-07-07 Darin Adler <darin@apple.com>

Reviewed by Mark Rowe.

  • fix <rdar://problem/6020441> REGRESSION: Layers on NWA.com render ugly

The old version of the OpenCube QuickMenu library used on this site still has code
that detects Netscape 4 by checking appVersion to see if it has the substring "4."
in it. We decided to special-case the filename of the script and tweak the appVersion
for files with that name.

  • bindings/js/ScriptController.cpp: (WebCore::ScriptController::ScriptController): Replace m_processingInlineCode with m_sourceURL. Use false instead of 0 to initialize a boolean. (WebCore::ScriptController::evaluate): Call argument sourceURL, not filename. Store current sourceURL in m_sourceURL. This fixes a mistake in the code that maintained the value of m_processingInlineCode, since the old code set it to false rather than restoring it. Renamed a local variable named sourceURL to exceptionSourceURL for clarity. (WebCore::ScriptController::processingUserGesture): Code that formerly used m_processingInlineCode to detect that it was evaluating code with no URL now uses m_sourceURL to do the same check.
  • bindings/js/ScriptController.h: Renamed filename argument to sourceURL; it has always been a URL, not a file path. Added a public sourceURL function and m_sourceURL and removed m_processingInlineCode.
  • page/Navigator.cpp: (WebCore::shouldHideFourDot): Added. Returns true if the currently running script has a source URL ending in "/dqm_script.js" and if the settings say we should do site-specific quirks (really JavaScript-library-specific in this case). (WebCore::Navigator::appVersion): Replace "4." with "4_" if shouldHideFourDot is true.
16:51 Changeset [35049] by cwzwarich@webkit.org

2008-07-07 Cameron Zwarich <cwzwarich@uwaterloo.ca>

Reviewed by Geoff.

Bug 19907: REGRESSION(r34824-r34941): Reproducible crash trying to log in to MediaTemple.net Account Center
<https://bugs.webkit.org/show_bug.cgi?id=19907>

Clear exceptions set on ExecStates before returning from NPAPI
callbacks, in order to avoid causing problems for the next script that
executes.

While fixing this bug, the question was raised of whether we are
correctly propagating exception information back to the caller:

Bug 19936: Correctly propagate exception information from NPAPI callbacks
<https://bugs.webkit.org/show_bug.cgi?id=19936>

  • bridge/NP_jsobject.cpp: (_NPN_InvokeDefault): (_NPN_Invoke): (_NPN_Evaluate): (_NPN_GetProperty): (_NPN_SetProperty): (_NPN_RemoveProperty): (_NPN_HasProperty): (_NPN_HasMethod): (_NPN_Enumerate):
16:26 Changeset [35048] by bdakin@apple.com

WebKitTools:

2008-07-07 Beth Dakin <bdakin@apple.com>

Reviewed by Anders.

This patch extends DRT accessibility tests to add the ability to
query the intValue, minValue, and maxValue of the focused element.

  • DumpRenderTree/AccessibilityController.cpp: (intValueOfFocusedElementCallback): (minValueOfFocusedElementCallback): (maxValueOfFocusedElementCallback): (AccessibilityController::staticFunctions):
  • DumpRenderTree/AccessibilityController.h:
  • DumpRenderTree/mac/AccessibilityControllerMac.mm: (AccessibilityController::intValueOfFocusedElement): (AccessibilityController::minValueOfFocusedElement): (AccessibilityController::maxValueOfFocusedElement):

LayoutTests:

2008-07-07 Beth Dakin <bdakin@apple.com>

Reviewed by Anders.

New test that uses the ability to query int value, min value, and
max value in the accessibility tests.

  • accessibility/aria-range-value-expected.txt: Added.
  • accessibility/aria-range-value.html: Added.
16:07 Changeset [35047] by mitz@apple.com

Reviewed by Darin Adler.

  • fix <rdar://problem/6057650> REGRESSION (r35025): Crash beneath FontCache::invalidate() when activating Safari with no windows open
  • css/CSSFontFaceSource.cpp: (WebCore::CSSFontFaceSource::getFontData): Added a null check because docLoader() can return 0 now.
  • css/CSSFontSelector.cpp: (WebCore::CSSFontSelector::docLoader): Added a null check of m_document. (WebCore::CSSFontSelector::addFontFaceRule): Ditto. (WebCore::CSSFontSelector::fontLoaded): Ditto. (WebCore::CSSFontSelector::fontCacheInvalidated): Ditto. (WebCore::fontDataForGenericFamily): Added a null check of document.
  • css/CSSFontSelector.h: (WebCore::CSSFontSelector::clearDocument): Added.
  • css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::~CSSStyleSelector): Added a call to CSSFontSelector::clearDocument(). When the style selector is destroyed, there is no guarantee that the document will continue to exist.
15:39 Changeset [35046] by darin@apple.com

2008-07-07 Darin Adler <darin@apple.com>

Rubber stamped by Adele.

  • VM/Machine.cpp: (KJS::Machine::privateExecute): Fix a typo in a comment.
15:38 Changeset [35045] by jchaffraix@webkit.org

WebCore:

2008-07-07 Julien Chaffraix <jchaffraix@webkit.org>

Not reviewed.

No svg build fix.

  • DerivedSources.make: Add JSSVGElementWrapperFactory.cpp target.
  • svg/animation/SMILTimeContainer.cpp: Add ENABLE(SVG) guard.
15:33 Changeset [35044] by sfalken@apple.com

JavaScriptCore:

2008-07-07 Steve Falkenburg <sfalken@apple.com>

Build fixes.

WebKitTools:

2008-07-07 Steve Falkenburg <sfalken@apple.com>

Fix build.

  • DumpRenderTree/win/ImageDiff.vcproj:
15:02 Changeset [35043] by andersca@apple.com

2008-07-07 Anders Carlsson <andersca@apple.com>

Reviewed by Darin.

Speculative fix for
<rdar://problem/5839800>
CrashTracer: [USER] 5802 crashes in Safari at com.apple.WebKit: -[WebHTMLView(WebPrivate) _updateMouseoverWithFakeEvent] + 389


Set _private->closed to YES before calling -[WebPluginController destroyAllPlugins].


My theory is that the plug-in destruction callbacks could end up rescheduling timers or re-adding notifications.
This is usually protected by _private->closed, but in this case it might still be false.

  • WebView/WebHTMLView.mm: (-[WebHTMLView close]):
14:34 Changeset [35042] by adele@apple.com

WebCore:

2008-07-07 Michelangelo De Simone <m.des@mac.com>

Reviewed by Adele.

Support for CSS3 :read-only and :read-write pseudoclasses for readonly text controls.

URL: http://www.w3.org/TR/web-forms-2/#relation

Tests: fast/css/readonly-pseudoclass-opera-001.html

fast/css/readonly-pseudoclass-opera-002.html
fast/css/readonly-pseudoclass-opera-003.html
fast/css/readonly-pseudoclass-opera-004.html
fast/css/readonly-pseudoclass-opera-005.html

  • css/CSSSelector.cpp: (WebCore::CSSSelector::extractPseudoType): Added "read-only" and "read-write" values for PseudoType extraction.
  • css/CSSSelector.h: (WebCore::CSSSelector::): Added PseudoReadOnly and PseudoReadWrite entries in PseudoType enum.
  • css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::SelectorChecker::checkOneSelector): Checks to determine whether to match :read-only and :read-write pseudoclasses on text controls.

LayoutTests:

2008-07-07 Michelangelo De Simone <m.des@mac.com>

Reviewed by Adele.

Test suite for :read-only and :read-write pseudoclasses, imported from
Opera.

  • fast/css/readonly-pseudoclass-opera-001-expected.txt: Added.
  • fast/css/readonly-pseudoclass-opera-001.html: Added.
  • fast/css/readonly-pseudoclass-opera-002-expected.txt: Added.
  • fast/css/readonly-pseudoclass-opera-002.html: Added.
  • fast/css/readonly-pseudoclass-opera-003-expected.txt: Added.
  • fast/css/readonly-pseudoclass-opera-003.html: Added.
  • fast/css/readonly-pseudoclass-opera-004-expected.txt: Added.
  • fast/css/readonly-pseudoclass-opera-004.html: Added.
  • fast/css/readonly-pseudoclass-opera-005-expected.txt: Added.
  • fast/css/readonly-pseudoclass-opera-005.html: Added.
14:29 Changeset [35041] by weinig@apple.com

WebCore:

2008-07-07 Sam Weinig <sam@webkit.org>

Reviewed by Geoffrey Garen.

Fix for https://bugs.webkit.org/show_bug.cgi?id=19928
querySelectorAll should throw an exception if a NSResolver is passed in.

  • Throw an NOT_SUPPORTED_ERR if a non-null or undefined parameter is passed as the second argument to querySelector or querySelectorAll.

Test: fast/dom/SelectorAPI/not-supported-NSResolver.html

  • bindings/js/JSDocumentCustom.cpp: (WebCore::JSDocument::querySelector): (WebCore::JSDocument::querySelectorAll):
  • bindings/js/JSElementCustom.cpp: (WebCore::JSElement::querySelector): (WebCore::JSElement::querySelectorAll):
  • dom/Document.idl:
  • dom/Element.idl:

LayoutTests:

2008-07-07 Sam Weinig <sam@webkit.org>

Reviewed by Geoffrey Garen.

Fix for https://bugs.webkit.org/show_bug.cgi?id=19928
querySelectorAll should throw an exception if a NSResolver is passed in.

  • fast/dom/SelectorAPI/not-supported-NSResolver-expected.txt: Added.
  • fast/dom/SelectorAPI/not-supported-NSResolver.html: Added.
13:35 Changeset [35040] by beidson@apple.com

WebCore:

2008-07-07 Brady Eidson <beidson@apple.com>

Reviewed by Mitz and Geoff

Test: fast/loader/empty-ref-versus-no-ref.html

Fix for https://bugs.webkit.org/show_bug.cgi?id=18951 and <rdar://problem/5921760>
"all DOM operations stop working when location.hash set to '#'"

The underlying problem is that KURL didn't really know the difference between "empty ref"
and "no ref at all" when changing the ref. So changing to the empty ref in JS ended up
affecting removal of the ref, which ended up causing an infinite load load, also killing
javascript.

  • platform/KURL.cpp: (WebCore::KURL::setRef): Changed "isEmpty()" to "isNull()", since NULL has the special meaning of "no ref at all" while empty means "empty ref"

LayoutTests:

2008-07-07 Brady Eidson <beidson@apple.com>

Reviewed by Mitz and Geoff

Test the fix for https://bugs.webkit.org/show_bug.cgi?id=18951 and <rdar://problem/5921760>

  • fast/loader/empty-ref-versus-no-ref-expected.txt: Added.
  • fast/loader/empty-ref-versus-no-ref.html: Added.
11:12 Changeset [35039] by kmccullough@apple.com

JavaScriptCore:

2008-07-07 Kevin McCullough <kmccullough@apple.com>

Reviewed by Darin.

When the profiler is running it gathers information and creates a
Profile. After it finishes the Profile can be sorted and have other
data refinements run over it. Both of these were done in the same class
before. Now I split the gathering operations into a new class called
ProfileGenerator.

  • JavaScriptCore.xcodeproj/project.pbxproj:
  • profiler/Profile.cpp: Removed code related to the gather stage of a Profile's creation. (KJS::Profile::create): (KJS::Profile::Profile):
  • profiler/Profile.h: Ditto. (KJS::Profile::title): (KJS::Profile::callTree): (KJS::Profile::setHead):
  • profiler/ProfileGenerator.cpp: Added. This is the class that will handle the stage of creating a Profile. Once the Profile is finished being created, this class goes away. (KJS::ProfileGenerator::create): (KJS::ProfileGenerator::ProfileGenerator): (KJS::ProfileGenerator::title): (KJS::ProfileGenerator::willExecute): (KJS::ProfileGenerator::didExecute): (KJS::ProfileGenerator::stopProfiling): (KJS::ProfileGenerator::didFinishAllExecution): (KJS::ProfileGenerator::removeProfileStart): (KJS::ProfileGenerator::removeProfileEnd):
  • profiler/ProfileGenerator.h: Added. (KJS::ProfileGenerator::profile): (KJS::ProfileGenerator::originatingGlobalExec): (KJS::ProfileGenerator::pageGroupIdentifier): (KJS::ProfileGenerator::client): (KJS::ProfileGenerator::stoppedProfiling):
  • profiler/Profiler.cpp: Now operates with the ProfileGenerator instead of the Profile. (KJS::Profiler::startProfiling): (KJS::Profiler::stopProfiling): (KJS::Profiler::didFinishAllExecution): It is here that the Profile is handed off to its client and the Profile Generator is no longer needed. (KJS::dispatchFunctionToProfiles): (KJS::Profiler::willExecute): (KJS::Profiler::didExecute):
  • profiler/Profiler.h: Cleaned up the includes and subsequently the forward declarations. Also use the new ProfileGenerator. (KJS::ProfilerClient::~ProfilerClient): (KJS::Profiler::currentProfiles):
  • profiler/TreeProfile.cpp: Use Profile's new interface. (KJS::TreeProfile::create): (KJS::TreeProfile::TreeProfile):
  • profiler/TreeProfile.h:

WebCore:

2008-07-07 Kevin McCullough <kmccullough@apple.com>

Reviewed by Darin.

Because profiler.h no longer #includes profile.h we need to explicitly
include it in console.cpp.

  • page/Console.cpp:
10:59 Changeset [35038] by zimmermann@webkit.org

Not reviewed. Build fix.
Fix --svg-filters build (affects Qt build, mac/win don't have it on by default)

09:01 Changeset [35037] by weinig@apple.com

2008-07-07 Sam Weinig <sam@webkit.org>

Reviewed by Cameron Zwarich.

Third step in broad cleanup effort.

[ File list elided ]

07:28 Changeset [35036] by mrowe@apple.com

Fix references to WebRenderNode.h and WebRenderNode.mm in Xcode project so that they don't have an unnecessary ".." in the path.

07:12 Changeset [35035] by zimmermann@webkit.org

Reviewed by Antti.

Fixes: http://bugs.webkit.org/show_bug.cgi?id=17779 (SVG 1.1 Errata demands "SVG JavaScript Liveness" support)

Revised version of the patch, not using pointer-to-member function callbacks anymore - after discussion with Adam Roben.
Remove unneccessary parameters of all macros. Hopefully build on Windows works as expected now, watching build bot...

05:55 Changeset [35034] by zimmermann@webkit.org

Reviewed by Oliver.

Working on: http://bugs.webkit.org/show_bug.cgi?id=17779 (SVG 1.1 Errata demands "SVG JavaScript Liveness" support)

Splitting up large patch in small chunk: Land valueAsString() implementation on it's own.

Added valueAsString() conversion for all SVG primitive types, needed by the SVG<->XML synchronization layer.
Unify "null value for a SVG animated type" concept, which was implemented in SVGDocumentExtensions before, in SVGAnimatedTemplate.

02:33 Changeset [35033] by jchaffraix@webkit.org

WebCore:

2008-07-07 Julien Chaffraix <jchaffraix@webkit.org>

Reviewed by Eric.

Disable renderer creation for XBLElement and
XBLDocument.

  • xbl/XBLDocument.h: (WebCore::XBLDocument::XBLDocument):
  • xbl/XBLElement.h: (WebCore::XBLElement::createRendererIfNeeded):
01:14 Changeset [35032] by jchaffraix@webkit.org

WebCore:

2008-07-07 Julien Chaffraix <jchaffraix@webkit.org>

Reviewed by Eric.

Bug 19918: [XBL] Add XBLDocument.

  • WebCore.xcodeproj/project.pbxproj:
  • dom/DOMImplementation.cpp: (WebCore::DOMImplementation::createDocument): Added XBLDocument creation.
  • dom/Document.h: (WebCore::Document::isXBLDocument):
  • xbl/XBLDocument.h: Added. (WebCore::XBLDocument::create): (WebCore::XBLDocument::isXBLDocument): (WebCore::XBLDocument::XBLDocument):

07/06/08:

23:14 Changeset [35031] by weinig@apple.com

WebCore:

2008-07-06 Sam Weinig <sam@webkit.org>

Reviewed by Dan Bernstein.

Rename "Access-Contol-Origin" to "Origin" to match the latest
Access Control for Cross-Site Request draft. (Editor's Draft 25 May 2008)

  • xml/XMLHttpRequest.cpp: (WebCore::XMLHttpRequest::makeSimpleCrossSiteAccessRequest): (WebCore::XMLHttpRequest::makeCrossSiteAccessRequestWithPreflight): (WebCore::XMLHttpRequest::handleAsynchronousPreflightResult):

LayoutTests:

2008-07-06 Sam Weinig <sam@webkit.org>

Reviewed by Dan Bernstein.

Rename "Access-Contol-Origin" to "Origin" to match the latest
Access Control for Cross-Site Request draft. (Editor's Draft 25 May 2008)

  • http/tests/xmlhttprequest/access-control-basic-allow-access-control-origin-header-data-url-expected.txt:
  • http/tests/xmlhttprequest/access-control-basic-allow-access-control-origin-header-expected.txt:
  • http/tests/xmlhttprequest/resources/access-control-basic-allow-access-control-origin-header.cgi:
21:17 Changeset [35030] by mitz@apple.com
  • try to fix the Windows build
  • WebCore.vcproj/WebCore.vcproj:
20:10 Changeset [35029] by mitz@apple.com
  • add missing brace
  • platform/graphics/qt/FontCacheQt.cpp:
20:08 Changeset [35028] by mitz@apple.com
  • try to fix the Qt build
  • platform/graphics/qt/FontCacheQt.cpp: (WebCore::FontCache::addClient): (WebCore::FontCache::removeClient):
19:49 Changeset [35027] by weinig@apple.com

JavaScriptCore:

2008-07-06 Sam Weinig <sam@webkit.org>

Reviewed by Cameron Zwarich.

Second step in broad cleanup effort.

[ File list elided ]

WebCore:

2008-07-06 Sam Weinig <sam@webkit.org>

Reviewed by Cameron Zwarich.

Add #include for kjs/protect.h.

  • xml/XMLHttpRequest.cpp: (WebCore::XMLHttpRequest::loadRequestAsynchronously):
19:40 Changeset [35026] by ddkilzer@apple.com

Save a few more bits in RenderStyle

Reviewed by Darin.

No test cases added since there is no change in behavior.

This changes WebCore::StyleMarqueeData from using 6 bits to 5 bits,
WebCore::RenderStyle::inherited_flags from using 36 bits to 35 bits,
and WebCore::RenderStyle::noninherited_flags from using 49 bits to
47 bits.

  • rendering/style/RenderStyle.h: (WebCore::StyleMarqueeData::behavior): Reduce from 3 bits to 2 bits since EMarqueeBehavior has 4 items. (WebCore::RenderStyle::inherited_flags._text_align): Reduce from 4 bits to 3 bits since ETextAlign has 8 items. (WebCore::RenderStyle::noninherited_flags._overflowX): Reduce from 4 bits to 3 bits since EOverflow has 6 items. (WebCore::RenderStyle::noninherited_flags._overflowY): Ditto.
17:00 Changeset [35025] by mitz@apple.com

Reviewed by Darin Adler.

  • fix <rdar://problem/5735163> WebCore caching makes text look wrong after font changes; need to respond appropriately instead
  • css/CSSFontSelector.cpp: (WebCore::CSSFontSelector::CSSFontSelector): Added a call to FontCache::addClient(). (WebCore::CSSFontSelector::~CSSFontSelector): Added a call to FontCache::removeClient(). (WebCore::CSSFontSelector::fontCacheInvalidated): Added. Called by the font cache when it is invalidated, and ensures that the document is updated.
  • css/CSSFontSelector.h:
  • platform/graphics/Font.cpp: (WebCore::Font::operator==): Added code to compare the font fallback lists' font cache generations.
  • platform/graphics/FontCache.cpp: (WebCore::FontCache::addClient): Added. (WebCore::FontCache::removeClient): Added. (WebCore::FontCache::generation): Added. (WebCore::FontCache::invalidate): Added. Deletes the platform data cache, increments the generation counter, and notifies all clients.
  • platform/graphics/FontCache.h:
  • platform/graphics/FontFallbackList.cpp: (WebCore::FontFallbackList::FontFallbackList): Added initialization of m_generation. (WebCore::FontFallbackList::invalidate): Added code to reset m_generation. (WebCore::FontFallbackList::fontDataAt): Added an assertion. (WebCore::FontFallbackList::fontDataForCharacters): Ditto. (WebCore::FontFallbackList::setPlatformFont): Ditto.
  • platform/graphics/FontFallbackList.h: (WebCore::FontFallbackList::generation): Added. Returns the font cache generation used to populate the list.
  • platform/graphics/FontSelector.h: (WebCore::FontSelector::fontCacheInvalidated): Added.
  • platform/graphics/mac/FontCacheMac.mm: (WebCore::fontCacheATSNotificationCallback): Added. Calls FontCache::invalidate(). (WebCore::FontCache::platformInit): Added code to register for ATS notifications.
  • svg/SVGFontFaceElement.cpp: Removed unneeded #inlcudes.
Note: See TracTimeline for information about the timeline view.