Timeline

Jan 26, 2007:

11:53 PM Changeset in webkit [19184] by aroben

• 1 edit in trunk/WebCore/ChangeLog

Added forgotten Radar reference.

8:50 PM Changeset in webkit [19183] by ggaren

• 34 edits
  2 adds in trunk

JavaScriptCore:

Reviewed by Maciej Stachowiak.

Fixed <rdar://problem/4608404> WebScriptObject's _rootObject lack
of ownership policy causes crashes (e.g., in Dashcode)

The old model for RootObject ownership was either to (1) leak them or (2) assign
them to a single owner -- the WebCore::Frame -- which would destroy them
when it believed that all of its plug-ins had unloaded.

This model was broken because of (1) and also because plug-ins are not the only
RootObject clients. All Bindings clients are RootObjects clients, including
applications, which outlive any particular WebCore::Frame.

The new model for RootObject ownership is to reference-count them, with a
throw-back to the old model: The WebCore::Frame tracks the RootObjects
it creates, and invalidates them when it believes that all of its plug-ins
have unloaded.

We maintain this throw-back to avoid plug-in leaks, particularly from Java.
Java is completely broken when it comes to releasing JavaScript objects.
Comments in our code allege that Java does not always call finalize when
collecting objects. Moreoever, my own testing reveals that, when Java does
notify JavaScript of a finalize, the data it provides is totally bogus.

This setup is far from ideal, but I don't think we can do better without
completely rewriting the bindings code, and possibly part of the Java
plug-in / VM.

Layout tests pass. No additional leaks reported. WebCore/manual-tests/*liveconnect*
and a few LiveConnect demos on the web also run without a hitch.

const RootObject* => RootObject*, since we need to ref/deref

• bindings/NP_jsobject.cpp: (jsDeallocate): deref our RootObjects. Also unprotect or JSObject, instead of just relying on the RootObject to do it for us when it's invalidated. (_isSafeScript): Check RootObject validity. (_NPN_CreateScriptObject): ditto (_NPN_Invoke): ditto (_NPN_Evaluate): ditto (_NPN_GetProperty): ditto (_NPN_SetProperty): ditto (_NPN_RemoveProperty): ditto (_NPN_HasProperty): ditto (_NPN_HasMethod): ditto (_NPN_SetException): ditto

• bindings/runtime_root.cpp: Revived bit-rotted LIAR LIAR LIAR comment.

LOOK: Added support for invalidating RootObjects without deleting them,
which is the main goal of this patch.

Moved protect counting into the RootObject class, to emphasize that
the RootObject protects the JSObject, and unprotects it upon being invalidated.

addNativeReference => RootObject::gcProtect
removeNativeReference => RootObject::gcUnprotect
ProtectCountSet::contains => RootObject::gcIsProtected

I know we'll all be sad to see the word "native" go.

• bindings/runtime_root.h: Added ref-counting support to RootObject, with all the standard accoutrements.

• bindings/c/c_utility.cpp: (KJS::Bindings::convertValueToNPVariant): If we can't find a valid RootObject, return void instead of just leaking.

• bindings/jni/jni_instance.cpp: (JavaInstance::JavaInstance): Don't take a RootObject in our constructor; be like other Instances and require the caller to call setRootObject. This reduces the number of ownership code paths. (JavaInstance::invokeMethod): Check RootObject for validity.
• bindings/jni/jni_instance.h: Removed private no-arg constructor. Having an arg constructor accomplishes the same thing.

• bindings/jni/jni_jsobject.cpp: (JavaJSObject::invoke): No need to call findProtectCountSet, because finalize() checks for RootObject validity. (JavaJSObject::JavaJSObject): check RootObject for validity (JavaJSObject::call): ditto (JavaJSObject::eval): ditto (JavaJSObject::getMember): ditto (JavaJSObject::setMember): ditto (JavaJSObject::removeMember): ditto (JavaJSObject::getSlot): ditto (JavaJSObject::setSlot): ditto (JavaJSObject::toString): ditto (JavaJSObject::finalize): ditto (JavaJSObject::createNative): No need to tell the RootObject to protect the global object, since the RootObject already owns the interpreter.

• bindings/jni/jni_runtime.cpp: (JavaArray::JavaArray): Removed copy construcutor becaue it was unused. Dead code is dangerous code.

• bindings/objc/objc_runtime.mm: Added WebUndefined protocol. Previous use of WebScriptObject was bogus, because WebUndefined is not a subclass of WebScriptObject. (convertValueToObjcObject): If we can't find a valid RootObject, return nil instead of just leaking.

• bindings/objc/objc_utility.mm: (KJS::Bindings::convertValueToObjcValue): If we can't find a valid RootObject, return nil instead of just leaking.

LayoutTests:

Reviewed by Maciej Stachowiak.

Added test for <rdar://problem/4608404> WebScriptObject's _rootObject lack
of ownership policy causes crashes (e.g., in Dashcode)

No test for Java or NPP versions of this bug because there's no reliable way to
make Java and NPP objects outlive their RootObjects (although Java objects
sometimes do).

• plugins/root-object-premature-delete-crash-expected.txt: Added.
• plugins/root-object-premature-delete-crash.html: Added.

WebCore:

Reviewed by Maciej Stachowiak.

Fixed <rdar://problem/4608404> WebScriptObject's _executionContext lack
of ownership policy causes crashes (e.g., in Dashcode)

Added RootObject ref-counting goodness.

• page/mac/FrameMac.h:
• page/mac/FrameMac.mm: (WebCore::FrameMac::cleanupPluginObjects): Invalidate our RootObjects instead of detroying them. Track _bindingRootObject separately from the rest of our RootObjects, since it has its own variable.

• page/mac/WebCoreFrameBridge.mm: (createRootObject): Use the Frame's new, more encapsulated function to create a RootObject.

• bindings/objc/WebScriptObject.mm: Nixed rootObject setters, since they were unused and they complicated reference-counting.

WebKitTools:

Reviewed by Maciej Stachowiak.

Added support for test for <rdar://problem/4608404> WebScriptObject's
_rootObject lack of ownership policy causes crashes (e.g., in Dashcode)

• DumpRenderTree/DumpRenderTree.m: (+[LayoutTestController isSelectorExcludedFromWebScript:]): (+[LayoutTestController webScriptNameForSelector:]): (-[LayoutTestController storeWebScriptObject:]): (-[LayoutTestController accessStoredWebScriptObject]): (-[LayoutTestController dealloc]):

8:28 PM Changeset in webkit [19182] by ddkilzer

• 2 edits in trunk/WebKit

WebKit:

Reviewed by Timothy.

Fixes crash drawing avatar on mail.yahoo.com.

• Plugins/WebBaseNetscapePluginStream.m: Retain the object since destroyStreamWithError: might release the last reference to it.

8:10 PM Changeset in webkit [22858] by sfalken

• 2 edits in branches/WindowsMerge/WebKitWin

2007-01-26 Steve Falkenburg <​sfalken@apple.com>

B&I build fix

• WebKit.vcproj/WebKit.vcproj:

7:51 PM Changeset in webkit [19181] by ddkilzer

• 2 edits in trunk/WebCore

WebCore:

Reviewed by Eric.

Release build fix.

• ksvg2/svg/SVGAnimationElement.cpp: (WebCore::SVGAnimationElement::calculateValueIndexAndPercentagePast): Added static cast to unsigned when calculating flooredValueIndex.

6:38 PM Changeset in webkit [19180] by brmorris

• 2 edits in S60/branches/3.1m/WebCore

bujtas <​zbujtas@gmail.com>

Rs'd & merged to 3.1 branch by Brad

DESC: HTMLScriptElementImpl gets notified twice about 'insert new script' event. MLIO-6X9HKM

​http://bugs.webkit.org/show_bug.cgi?id=12428

fix: prevents double notification.

6:37 PM Changeset in webkit [19179] by brmorris

• 2 edits in S60/trunk/WebCore

bujtas <​zbujtas@gmail.com>

Rs'd by Brad.
DESC: HTMLScriptElementImpl gets notified twice about 'insert new script' event. MLIO-6X9HKM

​http://bugs.webkit.org/show_bug.cgi?id=12428

fix: prevents double notification.

6:31 PM Changeset in webkit [19178] by aliceli1

• 4 edits
  3 adds in trunk

JavaScriptCore:

Reviewed by Maciej.

Fix for Repeated string concatenation results in OOM crash
​http://bugs.webkit.org/show_bug.cgi?id=11131

• kjs/operations.cpp: (KJS::add): Throw exception if string addition result is null
• kjs/ustring.cpp: (KJS::UString::UString): Don't call memcpy when malloc failed

LayoutTests:

Reviewed by Maciej.

Test for "Repeated string concatenation results in OOM crash"
​http://bugs.webkit.org/show_bug.cgi?id=11131

• fast/js/resources/string-concatenate-outofmemory.js: Added.
• fast/js/string-concatenate-outofmemory-expected.txt: Added.
• fast/js/string-concatenate-outofmemory.html: Added.

6:28 PM Changeset in webkit [19177] by brmorris

• 8 edits in S60/branches/3.1m

bujtas <​zbujtas@gmail.com>

Rs'd by Brad.
DESC: display fallback content in case of unsupported plugin PKAI-6WGSHS
​http://bugs.webkit.org/show_bug.cgi?id=12426

fix: webkit should return null, if the plugin is not supported.

• bridge/WebCoreBridge.h:
• kwq/KWQKHTMLPart.cpp: (KWQKHTMLPart::createPart): (KWQKHTMLPart::createSoundstart):

6:21 PM Changeset in webkit [19176] by darin

• 4 edits in trunk/WebCore

• checked in slightly newer revision of the patch below

• css/CSSComputedStyleDeclaration.cpp: (WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue): Grouped all the unimplemented values at the bottom. Fixed the ifdef for CSS_PROPWEBKIT_DASHBOARD_REGION so that it compiles on all platforms.

• css/cssparser.cpp: (WebCore::CSSParser::parseValue): Removed default case and added type so that the compiler warns about missing properties.
• css/cssstyleselector.cpp: (WebCore::CSSStyleSelector::applyProperty): Ditto.

6:13 PM Changeset in webkit [19175] by weinig

• 12 edits in trunk

LayoutTests:

Reviewed by Beth.

• updated test results for change to computed style

• editing/pasteboard/paste-table-002-expected.txt: Updated for change to order of style properties.

• fast/css/computed-style-expected.txt: Updated for change to order of style properties and to add new ones.

WebCore:

Reviewed by Beth.

• fix ​http://bugs.webkit.org/show_bug.cgi?id=12412 "unimplemented propertyID: 46" seen at dojotoolkit.org

Covered by fast/css/computed-style.html

• css/CSSPropertyNames.in: Sorted the file into two separate sections, one for properties without the -webkit- prefix and another for properties with.

• css/makeprop: Generate a CSSPropertyID enum instead of a set of #defines. Removed unused CSS_PROP_MIN and CSS_PROP_MAX. Renamed CSS_PROP_TOTAL to numCSSProperties and changed it to a const int instead of #define. Added a declaration of getPropertyName in the header. Changed the return type of getPropertyName to const char* and the parameter type to CSSPropertyID.

• css/CSSComputedStyleDeclaration.cpp: Sort the lists of properties in an easier to maintain order -- properties without a WebKit prefix first. (WebCore::valueForLength): Changed return value to PassRefPtr. (WebCore::valueForMaxLength): Ditto. (WebCore::valueForBorderStyle): Ditto. (WebCore::valueForTextAlign): Ditto. (WebCore::valueForAppearance): Added. (WebCore::valueForMarginCollapse): Added. (WebCore::valueForShadow): Changed return value to PassRefPtr. (WebCore::getPositionOffsetValue): Ditto. (WebCore::currentColorOrValidColor): Ditto. (WebCore::CSSComputedStyleDeclaration::cssText): Added cast since getPropertyName now takes an enum value. (WebCore::primitiveValueFromLength): Changed return value to PassRefPtr. (WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue): Added cast to CSSPropertyID so that gcc will detect missing enums in the switch statement. Simplified the background-size and border-spacing implementations so they don't have local variables. Added lots of cases to the switch statement for unimplemented properties. Implemented the outline-width, box-sizing, -webkit-appearance, -webkit-rtl-ordering, -webkit-user-drag, and -webkit-user-select properties. (WebCore::CSSComputedStyleDeclaration::item): Added cast since getPropertyName now takes an enum value.

• css/CSSMutableStyleDeclaration.cpp: Removed declaration of getPropertyName. (WebCore::CSSMutableStyleDeclaration::item): Added cast since getPropertyName now takes an enum value.

• css/CSSProperty.cpp: Removed declaration of getPropertyName. (WebCore::CSSProperty::cssText): Added cast since getPropertyName now takes an enum value and returns a const char*.

• css/CSSStyleDeclaration.cpp: Removed declaration of getPropertyName. (WebCore::CSSStyleDeclaration::getPropertyShorthand): Added cast since getPropertyName now takes an enum value and returns a const char*.

WebKit:

Reviewed by Beth.

• WebInspector/webInspector/inspector.js: Updated for new computed style properties.

6:11 PM Changeset in webkit [19174] by aliceli1

• 2 edits in trunk/WebCore

Reviewed by Brady.

Fixed <rdar://problem/4838076> CrashTracer: 10 crashes in Mail at com.apple.WebCore: WTF::HashMap<WebCore::String

• loader/DocLoader.cpp: (WebCore::DocLoader::requestResource): Added nil check around resource from cache.

6:11 PM Changeset in webkit [22857] by aroben

• 2 edits in branches/WindowsMerge/WebCore

Reviewed by Oliver.

Fix frequent failed assertions when painting scroll bars.

6:05 PM Changeset in webkit [19173] by aroben

• 2 edits in trunk/WebCore

Reviewed by Oliver.

Build fix.

• ksvg2/svg/SVGAnimationElement.cpp: (WebCore::SVGAnimationElement::calculateValueIndexAndPercentagePast): Keep index variables as unsigned until we need to convert to float.

6:04 PM Changeset in webkit [19172] by aliceli1

• 7 edits
  4 adds in trunk

LayoutTests:

Reviewed by Ada.

Added tests for <rdar://problem/4919097> REGRESSION: Ctrl-Y fails to yank (11076)

• editing/pasteboard/emacs-ctrl-a-k-y-expected.checksum: Added.
• editing/pasteboard/emacs-ctrl-a-k-y-expected.png: Added.
• editing/pasteboard/emacs-ctrl-a-k-y-expected.txt: Added.
• editing/pasteboard/emacs-ctrl-a-k-y.html: Added. Test for <rdar://problem/4919097>

• editing/deleting/delete-by-word-001-expected.txt:
• editing/deleting/delete-by-word-002-expected.txt:
• editing/deleting/delete-to-end-of-paragraph-expected.txt:
• editing/pasteboard/emacs-cntl-y-001-expected.txt: The expected results of these tests were affected by the WebCore checkin of the same number. The results are the same, but there were more editing delegates that were called, as expected. I suspect these tests were created during the time that <rdar://problem/4919097> was broken

WebCore:

Reviewed by Ada.

Fixed <rdar://problem/4919097> REGRESSION: Ctrl-Y fails to yank (11076)

• editing/Editor.cpp: (WebCore::Editor::deleteWithDirection): If there's no selection, locally create a range that is to the end of the granularity unit and pass that to deleteRange().

6:01 PM Changeset in webkit [22856] by aroben

• 3 edits in branches/WindowsMerge/WebKitWin

WebKitWin:

Reviewed by Oliver.

Engine side of fix for <rdar://problem/4958371> Boomer leaks WebViews

The problem was that we were calling RevokeDragDrop after our view
window had been destroyed, RevokeDragDrop wouldn't call Release on us.

Fix <rdar://problem/4958382> WebViews leak their HWND if no host window is set

• WebView.cpp: (WebView::~WebView): Call DestroyWindow if our view window hasn't been destroyed yet. (WebViewWndProc): Added an ASSERT, and added call to revokeDragDrop when handling WM_DESTROY. (WebView::initWithFrame): Call our own registerDragDrop. (WebView::close): Removed call to RevokeDragDrop. At this point our window has been destroyed, so it's too late to do this. (WebView::registerDragDrop): Added. (WebView::revokeDragDrop): Added.
• WebView.h: Added declarations.

5:55 PM Changeset in webkit [19171] by weinig

• 2 edits in trunk/WebCore

Reviewed by Maciej.

Qt build fix.

• page/qt/EventHandlerQt.cpp:

5:44 PM Changeset in webkit [22855] by oliver

• 2 edits in branches/WindowsMerge/WebCore

2007-01-26 Oliver Hunt <​oliver@apple.com>

Build fix.

• WebCore.vcproj/WebCore.vcproj:

5:43 PM Changeset in webkit [19170] by weinig

• 2 edits in trunk/WebCore

Reviewed by Hyatt.

Fix so that we only relayout frameset children on calls to layout.

• page/FrameView.cpp: (WebCore::FrameView::layout):

5:35 PM Changeset in webkit [19169] by oliver

• 3 edits in trunk/WebCore

2007-01-26 Oliver Hunt <​oliver@apple.com>

Build Fix

• ksvg2/svg/ColorDistance.cpp:
• ksvg2/svg/SVGAnimationElement.cpp: (WebCore::SVGAnimationElement::handleTimerEvent):

4:59 PM Changeset in webkit [19168] by eseidel

• 20 edits
  14 adds in trunk/WebCore

2007-01-26 Eric Seidel <​eric@webkit.org>

Reviewed by olliej.

Add support for calcMode, keyTimes and (nearly) keySplines.
​http://bugs.webkit.org/show_bug.cgi?id=12350
Re-architect much of the animation system to remove huge sections of code.
Added new ColorDistance, SVGTransformDistance to simplify animateColor, animateTransform calculations.
I also added basic <animateMotion> support since it was so easy w/ the new architecture.

Adding manual animation tests until bug 12074 is fixed.

• WebCore.xcodeproj/project.pbxproj:
• ksvg2/misc/SVGTimer.cpp: (WebCore::SVGTimer::applyAnimations):
• ksvg2/svg/ColorDistance.cpp: Added. (WebCore::ColorDistance::ColorDistance): (WebCore::clampColorValue): (WebCore::ColorDistance::scaledDistance): (WebCore::ColorDistance::addColorsAndClamp): (WebCore::ColorDistance::addToColorAndClamp): (WebCore::ColorDistance::isZero): (WebCore::ColorDistance::distance):
• ksvg2/svg/ColorDistance.h: Added.
• ksvg2/svg/SVGAnimateColorElement.cpp: (WebCore::SVGAnimateColorElement::SVGAnimateColorElement): (WebCore::SVGAnimateColorElement::updateAnimationBaseValueFromElement): (WebCore::SVGAnimateColorElement::applyAnimatedValueToElement): (WebCore::SVGAnimateColorElement::updateAnimatedValue): (WebCore::SVGAnimateColorElement::calculateFromAndToValues):
• ksvg2/svg/SVGAnimateColorElement.h: (WebCore::SVGAnimateColorElement::contextElement):
• ksvg2/svg/SVGAnimateElement.h: (WebCore::SVGAnimateElement::updateAnimatedValue): (WebCore::SVGAnimateElement::calculateFromAndToValues):
• ksvg2/svg/SVGAnimateMotionElement.cpp: (WebCore::SVGAnimateMotionElement::hasValidTarget): (WebCore::SVGAnimateMotionElement::updateAnimatedValue): (WebCore::parsePoint): (WebCore::SVGAnimateMotionElement::calculateFromAndToValues): (WebCore::SVGAnimateMotionElement::updateAnimationBaseValueFromElement): (WebCore::SVGAnimateMotionElement::applyAnimatedValueToElement):
• ksvg2/svg/SVGAnimateMotionElement.h:
• ksvg2/svg/SVGAnimateTransformElement.cpp: (WebCore::SVGAnimateTransformElement::SVGAnimateTransformElement): (WebCore::SVGAnimateTransformElement::updateAnimatedValue): (WebCore::SVGAnimateTransformElement::updateAnimationBaseValueFromElement): (WebCore::SVGAnimateTransformElement::applyAnimatedValueToElement): (WebCore::SVGAnimateTransformElement::calculateFromAndToValues): (WebCore::SVGAnimateTransformElement::parseTransformValue):
• ksvg2/svg/SVGAnimateTransformElement.h: (WebCore::SVGAnimateTransformElement::contextElement):
• ksvg2/svg/SVGAnimationElement.cpp: (WebCore::SVGAnimationElement::parseKeyNumbers): (WebCore::parseValues): (WebCore::parseKeySplines): (WebCore::SVGAnimationElement::parseMappedAttribute): (WebCore::SVGAnimationElement::targetAttributeAnimatedValue): (WebCore::SVGAnimationElement::setTargetAttributeAnimatedValue): (WebCore::SVGAnimationElement::detectAnimationMode): (WebCore::adjustPercentagePastForKeySplines): (WebCore::SVGAnimationElement::valueIndexAndPercentagePastForDistance): (WebCore::SVGAnimationElement::calculateTotalDistance): (WebCore::caculateValueIndexForKeyTimes): (WebCore::SVGAnimationElement::isValidAnimation): (WebCore::SVGAnimationElement::calculateValueIndexAndPercentagePast): (WebCore::SVGAnimationElement::updateAnimationBaseValueFromElement): (WebCore::SVGAnimationElement::applyAnimatedValueToElement): (WebCore::SVGAnimationElement::handleTimerEvent): (WebCore::SVGAnimationElement::updateAnimatedValueForElapsedSeconds):
• ksvg2/svg/SVGAnimationElement.h:
• ksvg2/svg/SVGParserUtilities.cpp: (WebCore::SVGPolyParser::parsePoints):
• ksvg2/svg/SVGParserUtilities.h:
• ksvg2/svg/SVGSetElement.cpp:
• ksvg2/svg/SVGSetElement.h: (WebCore::SVGSetElement::updateAnimatedValue): (WebCore::SVGSetElement::calculateFromAndToValues):
• ksvg2/svg/SVGTransform.cpp: (SVGTransform::type): (SVGTransform::translate): (SVGTransform::scale): (SVGTransform::setSkewX): (SVGTransform::setSkewY):
• ksvg2/svg/SVGTransform.h: (WebCore::operator==): (WebCore::operator!=):
• ksvg2/svg/SVGTransformDistance.cpp: Added. (WebCore::SVGTransformDistance::SVGTransformDistance): (WebCore::SVGTransformDistance::scaledDistance): (WebCore::SVGTransformDistance::addSVGTransforms): (WebCore::SVGTransformDistance::addSVGTransform): (WebCore::SVGTransformDistance::addToSVGTransform): (WebCore::SVGTransformDistance::isZero): (WebCore::SVGTransformDistance::distance):
• ksvg2/svg/SVGTransformDistance.h: Added.
• ksvg2/svg/SVGTransformList.cpp: (SVGTransformList::concatenateForType):
• ksvg2/svg/SVGTransformList.h:
• manual-tests/animation/animateColor-by.svg: Added.
• manual-tests/animation/animateColor-from-by.svg: Added.
• manual-tests/animation/animateColor-repeat-indefinite.svg: Added.
• manual-tests/animation/animateColor-to-from.svg: Added.
• manual-tests/animation/animateColor-to.svg: Added.
• manual-tests/animation/animateColor-values-simple.svg: Added.
• manual-tests/animation/animateMotion-to.svg: Added.
• manual-tests/animation/animateTransform-keyTimes.svg: Added.
• manual-tests/animation/animateTransform-repeat-once.svg: Added.

4:52 PM Changeset in webkit [19167] by kdecker

• 4 edits in trunk/WebKit

Reviewed by andersca.

Fixed: <rdar://problem/4946922> WebBaseNetscapePluginView leaks memory
​http://bugs.webkit.org/show_bug.cgi?id=11523

• Plugins/WebBaseNetscapePluginStream.m:

(-[WebBaseNetscapePluginStream setPlugin:]): Calls -[WebBaseNetscapePluginView disconnectStream:]

• Plugins/WebBaseNetscapePluginView.h: Added disconnectStream: to header.
• Plugins/WebBaseNetscapePluginView.mm: (-[WebBaseNetscapePluginView stop]): Make a copy of the streams collection prior to calling stop all streams. This is necessary because calling stop has the side effect of removing the stream from this same collection. (-[WebBaseNetscapePluginView disconnectStream:]): Added. Removes the stream from the streams collection.

4:16 PM Changeset in webkit [22854] by bdakin

• 2 edits in branches/WindowsMerge/WebCore

Reviewed by Darin.

Due to <rdar://problem/4956565> REGRESSION: After scrolling frame,
hovering over link in this frame doesn't change cursor to pointing
hand

lastEventIsMouseUp() is a Mac-only issue. This patch makes it Mac-
only code, so we can remove the temporary link stub for it.

• platform/win/TemporaryLinkStubs.cpp:

4:03 PM Changeset in webkit [19166] by bdakin

• 4 edits in trunk/WebCore

Reviewed by Darin.

Fix for <rdar://problem/4956565> REGRESSION: After scrolling frame,
hovering over link in this frame doesn't change cursor to pointing
hand

The mouseMove event was not being propagated correctly after using
the mouse to scroll the frame because m_mousePressed was never
getting set to false.

• page/EventHandler.cpp: (WebCore::EventHandler::handleMousePressEvent): This code does not belong here.
• page/EventHandler.h: lastEventIsMouseUp() is only ever relevant in EventHandlerMac, so it can just be a static function there.
• page/mac/EventHandlerMac.mm: (WebCore::lastEventIsMouseUp): Make this static. (WebCore::EventHandler::passMouseDownEventToWidget): Here is where we need to set m_mousePressed to false if lastEventIsMouseUp() is true.

3:58 PM Changeset in webkit [22853] by andersca

• 2 edits in branches/WindowsMerge/WebKitWin

Reviewed by Adam.

• WebKitDLL.cpp: Add WebURLRequest entries to the big array.

3:36 PM Changeset in webkit [22852] by andersca

• 6 edits in branches/WindowsMerge/WebKitWin

WebKitWin:

Reviewed by Adam.

<rdar://problem/4923845>
Implement IWebURLRequest.

• Interfaces/WebKit.idl:
• WebKitClassFactory.cpp: (WebKitClassFactory::CreateInstance):
• WebKitDLL.cpp: (DllGetClassObject):
• WebMutableURLRequest.cpp: (WebMutableURLRequest::WebMutableURLRequest): (WebMutableURLRequest::createInstance): (WebMutableURLRequest::createImmutableInstance): (WebMutableURLRequest::QueryInterface):
• WebMutableURLRequest.h:

2:47 PM Changeset in webkit [19165] by hyatt

• 4 adds in trunk/LayoutTests/fast/css

Add layout test for orphaned CSS unit coalescing.

2:45 PM Changeset in webkit [19164] by hyatt

• 4 edits in trunk/WebCore

Fix for style regression caused by strictness checking of the number of properties. This caused code like:
"width: 20 px" to fail, because we interpreted it as two values and rejected it. Our old code allowed 20 to be
used like a pixel value and then ignored the orphaned unit.

This patch actually scans the list of values looking for orphaned units and reattaches them to the numeric
values that they should correspond to. This means rules like "width: 5 em" will now work in quirks mdoe and
the "em" unit type will be honored.

Reviewed by beth

• WebCore.xcodeproj/project.pbxproj:
• css/cssparser.cpp: (WebCore::unitFromString): (WebCore::CSSParser::checkForOrphanedUnits): (WebCore::CSSParser::parseValue):
• css/cssparser.h: (WebCore::ValueList::valueAt): (WebCore::ValueList::deleteValueAt):

2:19 PM Changeset in webkit [19163] by staikos

• 14 edits in trunk

What's left of my trashed work to add more Qt functionality

2:02 PM Changeset in webkit [19162] by beidson

• 2 edits in trunk/WebCore

Fixed build that use RetainPtr::adoptCFReference(CFTypeRef)

1:13 PM Changeset in webkit [19161] by lars

• 3 edits in trunk/WebKitQt

Make it compile again with Qt 4.2 and add the copyright
headers where forgotten in the last submit.

11:51 AM Changeset in webkit [22851] by sfalken

• 2 edits in branches/WindowsMerge/WebKitWin

2007-01-26 Steve Falkenburg <​sfalken@apple.com>

B&I build fix

• WebKit.vcproj/WebKit.vcproj:

11:50 AM Changeset in webkit [22850] by andersca

• 3 edits in branches/WindowsMerge/WebKitWin

WebKitWin:

Reviewed by Adam.

Don't pretend that we're an IWebHTTPURLResponse when we're not.

• WebURLResponse.cpp: (WebURLResponse::QueryInterface): (WebURLResponse::allHeaderFields): (WebURLResponse::localizedStringForStatusCode): (WebURLResponse::statusCode):

• WebView.cpp: (WebView::canShowMIMEType): Implement this.

11:31 AM Changeset in webkit [19160] by darin

• 5 edits in trunk/WebCore

Reviewed by Brady.

• fix <rdar://problem/4956688> move Mac-specific part of icon code into IconMac
• fix a GC-compatibility problem in the icon code I noticed by code inspection

• platform/FileChooser.cpp: (WebCore::FileChooser::chooseIcon): Removed Mac-specific path check -- that's in newIconForFile now.

• platform/graphics/Icon.h: Changed the image to be a RetainPtr. The old code was using -[NSObject retain] which is not good enough for a pointer in a C++ object.
• platform/graphics/mac/IconMac.mm: (WebCore::Icon::Icon): Got rid of unneeded code to do what RetainPtr does for us. (WebCore::Icon::~Icon): Ditto. (WebCore::Icon::newIconForFile): Added check for non-absolute file names. Updated for use of RetainPtr. Use constructor that takes NSImage. (WebCore::Icon::paint): Add get() calls needed now that we are using a RetainPtr.

• WebCore.xcodeproj/project.pbxproj: Let Xcode do its thing.

10:10 AM Changeset in webkit [19159] by ap

• 3 edits
  2 adds in trunk

Reviewed by Maciej.

​http://bugs.webkit.org/show_bug.cgi?id=12385
Assertion failure when loading subresources from "data:" frames

Test: fast/loader/opaque-base-url.html

• platform/KURL.cpp: (WebCore::KURL::init): Make the URL invalid if it's relative, but has an opaque or empty base.

10:01 AM Changeset in webkit [22849] by beidson

• 2 edits in branches/WindowsMerge/WebKitWin

I'm in ur WebDownloadz, null checkin ur pointerz

9:20 AM Changeset in webkit [19158] by darin

• 4 edits in trunk

LayoutTests:

• fast/events/objc-event-api-expected.txt: Updated test result now that we no longer invalidate clicks when the mouse moves a bit.

WebCore:

Reviewed by Darin.

• page/EventHandler.cpp: (WebCore::EventHandler::handleMouseMoveEvent): Removed a call to invalidateClick(), since we don't want to cancel a click if the mouse moves.

9:14 AM Changeset in webkit [19157] by darin

• 2 edits in trunk/WebCore

• css/MediaQueryEvaluator.cpp: (WebCore::MediaQueryEvaluator::mediaTypeMatchSpecific): Oops, forgot a !.

9:10 AM Changeset in webkit [19156] by darin

• 5 edits in trunk/WebCore

Patch by Nate Begeman.
Reviewed by Darin and Maciej.

• <rdar://problem/4946877> Avoid unnecessary allocations in CSS, patch attached (12401) ​http://bugs.webkit.org/show_bug.cgi?id=12401

• css/MediaQueryEvaluator.h: Add mediaTypeMatchSpecific, a function that takes a char* to avoid creating and destroying a String for simple media type queries.
• css/MediaQueryEvaluator.cpp: (WebCore::MediaQueryEvaluator::mediaTypeMatchSpecific): Added. Besides avoiding the conversion to String, this also doesn't have the special case code to handle "" and "all".

• css/cssstyleselector.cpp: (WebCore::CSSStyleSelector::matchUARules): Call mediaTypeMatchSpecific here when deciding whether to use the print version of the user agent style sheet.

• loader/Cache.cpp: (WebCore::Cache::requestResource): Don't convert to a string and back to a KURL when calling createResource. Avoids reparsing the URL string.

8:55 AM Changeset in webkit [19155] by brmorris

• 3 edits in S60/trunk/WebCore

yongjzha

Reviewed by <​zbujtas@gmail.com>
DESC: PNUN-6XJJLF When use css border property outset, and set border color to blue, the browser
shows black at the bottom edge and right edge.
fix: correct RGB to HSV and HSV to RGB conversion algorithms in KWQColor.cpp

​http://bugs.webkit.org/show_bug.cgi?id=12393

8:53 AM Changeset in webkit [19154] by brmorris

• 3 edits in S60/branches/3.1m/WebCore

yongjzha

Reviewed by <​zbujtas@gmail.com>
DESC: PNUN-6XJJLF When use css border property outset, and set border color to blue, the browser
shows black at the bottom edge and right edge.
fix: correct RGB to HSV and HSV to RGB conversion algorithms in KWQColor.cpp

​http://bugs.webkit.org/show_bug.cgi?id=12393

8:49 AM Changeset in webkit [19153] by brmorris

• 2 edits in S60/branches/3.1m/WebKit

2007-01-26 Deepa

Reviewed by Sachin.
DESC: Resolve FavIcon issue in cases of websites with the same sub domain name (AKKO-6XJDCQ)
The function ExtractDomainL( ) was changed to include the entire host name
instead of just the last two parts.

8:48 AM Changeset in webkit [19152] by brmorris

• 2 edits in S60/trunk/WebKit

2007-01-26 Deepa

Reviewed by Sachin.
DESC: Resolve FavIcon issue in cases of websites with the same sub domain name (AKKO-6XJDCQ)
The function ExtractDomainL( ) was changed to include the entire host name
instead of just the last two parts.

5:13 AM Changeset in webkit [19151] by zack

• 5 edits
  3 adds in trunk

History support on platform Qt. r=lars
We might switch to a model but for now this is ok.

5:01 AM Changeset in webkit [19150] by zack

• 4 edits
  1 add in trunk/WebCore

introduce platform scrollbars on Qt platform

3:40 AM Changeset in webkit [19149] by bdash

• 19 edits
  2 adds in trunk

2007-01-26 Mitz Pettel <​mitz@webkit.org>

Reviewed by Darin.

• fix ​http://bugs.webkit.org/show_bug.cgi?id=11109 REGRESSION (r15471): menus are offset on eink.com <rdar://problem/4824760>

Changed offset{Left,Top} to match Firefox. If the body element is statically
positioned, it is still returned as the offsetParent of elements whose containing
block is the initial containing block (the root), but their offset{Left,Top} are
relative to the root. If the body element is positioned, then offsets are relative
to it. The body's own offset{Left,Top} are always 0.

Test: fast/dom/Element/offsetLeft-offsetTop-body-quirk.html

• rendering/RenderObject.cpp: (WebCore::RenderObject::offsetLeft): (WebCore::RenderObject::offsetTop):

2007-01-26 Mitz Pettel <​mitz@webkit.org>

Reviewed by Darin.

• test and updated tests and results for ​http://bugs.webkit.org/show_bug.cgi?id=11109 REGRESSION (r15471): menus are offset on eink.com <rdar://problem/4824760>

• editing/execCommand/findString-2.html:
• editing/pasteboard/4947130.html:
• editing/pasteboard/drop-text-without-selection.html:
• editing/selection/anchor-focus1-expected.txt:
• editing/selection/click-before-and-after-table.html:
• editing/selection/drag-select-1.html:
• editing/selection/editable-links-expected.txt:
• editing/selection/mixed-editability-1.html:
• editing/selection/paragraph-granularity.html:
• editing/selection/selection-actions.html:
• editing/selection/word-granularity.html:
• fast/dom/Element/offsetLeft-offsetTop-body-quirk-expected.txt: Added.
• fast/dom/Element/offsetLeft-offsetTop-body-quirk.html: Added.
• fast/forms/drag-into-textarea.html:
• fast/forms/listbox-selection-2.html:
• fast/forms/listbox-selection.html:
• fast/table/click-near-anonymous-table-expected.txt:
• fast/table/click-near-anonymous-table.html:

3:35 AM Changeset in webkit [19148] by bdash

• 6 edits
  4 adds in trunk

2007-01-26 Mitz Pettel <​mitz@webkit.org>

Reviewed by Reviewed by Darin and Maciej.

• fix ​http://bugs.webkit.org/show_bug.cgi?id=10983 REGRESSION (r12290): Drop shadow of flickr photo note is positioned incorrectly the second time it's shown

Test: fast/dynamic/staticY.html

• rendering/RenderBlock.cpp: (WebCore::RenderBlock::layoutPositionedObjects): Avoid marking the object's parents, since they have already been laid out.
• rendering/RenderBox.cpp: (WebCore::RenderBox::setStaticY): Added a call to setChildNeedsLayout() to ensure relayout by this box's container (third and fourth cases in the layout test).
• rendering/RenderBox.h:
• rendering/RenderObject.cpp: (WebCore::RenderObject::markContainingBlocksForLayout): For positioned objects with static y, also mark the object's parent to ensure that it recomputes the static y value (first case in the layout test).

2007-01-26 Mitz Pettel <​mitz@webkit.org>

Reviewed by Darin and Maciej.

• test for ​http://bugs.webkit.org/show_bug.cgi?id=10983 REGRESSION (r12290): Drop shadow of flickr photo note is positioned incorrectly the second time it's shown

Tests three cases that the patch fixes and another one that was fixed in
r8284 but did not have a test.

• fast/dynamic/staticY-expected.checksum: Added.
• fast/dynamic/staticY-expected.png: Added.
• fast/dynamic/staticY-expected.txt: Added.
• fast/dynamic/staticY.html: Added.

3:16 AM Changeset in webkit [19147] by lars

• 10 edits
  2 deletes in trunk

Get rid of FrameQtClient. It's unused since we moved over
to the loader.

3:14 AM Changeset in webkit [19146] by zack

• 2 edits in trunk/WebKitQt

Correctly position child frames within the
toplevel frame (put them on the viewport). r=lars

2:25 AM Changeset in webkit [22848] by aroben

• 1 edit in branches/WindowsMerge/WebKitWin/ChangeLog

Updated ChangeLog with Radar reference.

1:57 AM Changeset in webkit [19145] by lars

• 1904 edits
  3 deletes in trunk/LayoutTestResults

Update the test cases after the last change to
DumpRenderTree. Hope this will not be needed again.

1:50 AM Changeset in webkit [19144] by zack

• 2 edits in trunk/WebCore

fix the compile

1:44 AM Changeset in webkit [19143] by aroben

• 2 edits in trunk/WebCore

Reviewed by Maciej.

Fixed an issue on tivofaq.com where you could never tab forward into
the left frame after tabbing out of the document.

• page/FocusController.cpp: (WebCore::FocusController::advanceFocus): Clear the focused frame before passing off focus to Chrome so that when we re-enter the document we start at the main frame instead of the last focused frame.

1:27 AM Changeset in webkit [22847] by oliver

• 2 edits in branches/WindowsMerge/WebCore

2007-01-26 Oliver Hunt <​oliver@apple.com>

Build fix -- adding SVGInlineTextBox.{cpp,h} to project

• WebCore.vcproj/WebCore.vcproj:

1:22 AM Changeset in webkit [19142] by lars

• 6 edits in trunk

Reviewed by Zack

WebKitQt:
Make sure we resize the HTML when resizing a
QWebFrame. Also ensure that the render tree
is up to date before dumping it.

DumpRenderTree:
Small hack to ensure that our top level frame actually has the
correct size. Unfortunately this means I'll have to regenerate
all test cases :/

1:20 AM Changeset in webkit [19141] by lars

• 2 edits in trunk/LayoutTestResults

Minor update, skip two test case that are hanging
forever.

1:03 AM Changeset in webkit [19140] by rwlbuis

• 8 edits
  2 adds in trunk/WebCore

Reviewed by olliej.

​http://bugs.webkit.org/show_bug.cgi?id=11911
SVG exhibits double-text-selection behavior

Fix selection height of the inline text boxes for svg.

1:00 AM Changeset in webkit [19139] by zack

• 2 edits in trunk/WebKitQt

Display pages inside a mainwindow instead of a custom qwidget
in the qtlauncher.

12:54 AM Changeset in webkit [19138] by zack

• 2 edits in trunk/WebCore

Fix drawing of tiled images. r=lars

12:25 AM Changeset in webkit [19137] by zack

• 9 edits in trunk/WebCore

Fix image loading and implement icon class

Jan 25, 2007:

11:53 PM Changeset in webkit [22846] by oliver

• 2 edits in branches/WindowsMerge/WebCore

2007-01-25 Oliver Hunt <​oliver@apple.com>

Reviewed by Maciej.

Fixed DragDataWin behaviour so that it will provide plain text
for a file dropped on the window.

• platform/win/DragDataWin.cpp: (WebCore::DragData::asPlainText):

11:32 PM Changeset in webkit [22845] by beidson

• 5 edits in branches/WindowsMerge/WebKitWin

Make WebDownloads no longer start themselves

11:30 PM Changeset in webkit [19136] by lars

• 2 edits in trunk/JavaScriptCore

Patch from Jan Kraemer.

Reviewed by Maciej

Fix for ​http://bugs.webkit.org/show_bug.cgi?id=12382

Fix crash on architectures with 32 bit ints and
64 bit longs (For example Linux on AMD64)

8:25 PM Changeset in webkit [22844] by beidson

• 1 edit in branches/WindowsMerge/WebCore/platform/network/cf/FormDataStreamCFNet.cpp

Fix the build for people who aren't me and aren't currently running ToT CF

7:18 PM Changeset in webkit [22843] by beidson

• 3 edits in branches/WindowsMerge/WebCore

Prepare for upcoming CF/CFNetwork api changes

7:13 PM Changeset in webkit [22842] by adele

• 3 edits in branches/WindowsMerge/WebCore

WebCoreWin:

Reviewed by Adam.

Added support for slider.

4:51 PM Changeset in webkit [22841] by adachan

• 3 edits in branches/WindowsMerge/WebKitWin

2007-01-25 Ada Chan <​adachan@apple.com>

WebDownload.cpp reviewed by Brady.
WebFrame.cpp reviewed by Anders.

• WebDownload.cpp: (WebDownload::cancel): added null check
• WebFrame.cpp: (WebFrame::dispatchDecidePolicyForMIMEType): fixed bug that we treated loading image urls as downloads.

4:35 PM Changeset in webkit [22840] by aroben

• 6 edits in branches/WindowsMerge/WebKitWin

WebKitWin:

Reviewed by Steve and Darin.

3:56 PM Changeset in webkit [22839] by sfalken

• 1 edit in branches/WindowsMerge/WebKitWin/WebKit.vcproj/VERSION

Bump version for submit

3:51 PM Changeset in webkit [19135] by kdecker

• 4 edits in trunk/WebKit

Backed out my last patch because it crashes espn.com. Stay tuned for a newer version..

• Plugins/WebBaseNetscapePluginStream.m: (-[WebBaseNetscapePluginStream setPlugin:]): Removed call to streamIsDead.
• Plugins/WebBaseNetscapePluginView.h: Removed streamIsDead.
• Plugins/WebBaseNetscapePluginView.mm: Ditto.

3:46 PM Changeset in webkit [19134] by sfalken

• 1 copy in tags/Safari-521.32.15b

New tag.

3:05 PM Changeset in webkit [19133] by justing

• 5 edits in trunk

LayoutTests:

Reviewed by oliver

<​http://bugs.webkit.org/show_bug.cgi?id=12409>
REGRESSION (r19039): Drag caret isn't cleared after a drop

• editing/pasteboard/4947130-expected.checksum:
• editing/pasteboard/4947130-expected.png:

WebCore:

Reviewed by oliver

<​http://bugs.webkit.org/show_bug.cgi?id=12409>
REGRESSION (r19039): Drag caret isn't cleared after a drop

No new layout tests because the regression was
caught by current layout tests.

• page/DragController.cpp: (WebCore::DragController::concludeDrag): Clear the drag caret.

3:01 PM Changeset in webkit [22838] by sfalken

• 2 edits in branches/WindowsMerge/WebCore

2007-01-25 Steve Falkenburg <​sfalken@apple.com>

B&I build fix

• WebCore.vcproj/WebCore.vcproj:

2:46 PM Changeset in webkit [19132] by darin

• 3 edits in trunk/WebKit

Reviewed by Beth.

• fix <rdar://problem/4952766> Safari has a top secret color picker that can be used to... uhh... I don't know

• Panels/English.lproj/WebAuthenticationPanel.nib/info.nib: Let Interface Builder have its way.
• Panels/English.lproj/WebAuthenticationPanel.nib/objects.nib: Remove the NSColorWell that was in here (for no good reason).

1:30 PM Changeset in webkit [19131] by kdecker

• 4 edits in trunk/WebKit

Reviewed by andersca.

A few tweaks with of a fix done by Steve Gehrman.

Fixed: <rdar://problem/4946922> WebBaseNetscapePluginView leaks memory

​http://bugs.webkit.org/show_bug.cgi?id=11523

• Plugins/WebBaseNetscapePluginStream.m: (-[WebBaseNetscapePluginStream setPlugin:]): Calls -[WebBaseNetscapePluginView streamIsDead:]
• Plugins/WebBaseNetscapePluginView.h: Added streamIsDead to header.
• Plugins/WebBaseNetscapePluginView.mm: (-[WebBaseNetscapePluginView streamIsDead:]): Added. Removes the stream from the streams collection.

12:21 PM Changeset in webkit [19130] by ggaren

• 4 edits
  4 adds in trunk

LayoutTests:

Reviewed by Geoffrey Garen.

Added tests for ​http://bugs.webkit.org/show_bug.cgi?id=12342
REGRESSION: destroying a frame from its own script causes various crashes

Thanks to Alexey Proskuryakov for the original versions of these tests.

• fast/dom/exception-no-frame-inline-script-crash-expected.txt: Added.
• fast/dom/exception-no-frame-inline-script-crash.html: Added.
• fast/dom/exception-no-frame-timeout-crash-expected.txt: Added.
• fast/dom/exception-no-frame-timeout-crash.html: Added.

WebCore:

Reviewed and landed by Geoffrey Garen.

• fix ​http://bugs.webkit.org/show_bug.cgi?id=12342 REGRESSION: destroying a frame from its own script causes various crashes

• bindings/js/kjs_window.cpp: (KJS::ScheduledAction::execute): Use RefPtr for the frame and the interpreter object so they don't get deleted out from underneath us.

• bindings/js/kjs_proxy.cpp: (WebCore::KJSProxy::evaluate): Add a check for a page of 0. This was the only call site for the addMessageToConsole function that did not have a check for a page of 0.

11:52 AM Changeset in webkit [19129] by lars

• 2 edits
  94 adds in trunk/LayoutTestResults

Add the next 100 test cases.

11:27 AM Changeset in webkit [19128] by lars

• 2 edits
  300 adds in trunk/LayoutTestResults

Add more test cases that pass on the Qt build.

10:42 AM Changeset in webkit [19127] by lars

• 2 edits
  157 adds in trunk/LayoutTestResults

The tests in dom/svg and dom/xhtml as well as
most of the remaining tests in tables/ pass now as
well, so lets run the test suite over them.

10:13 AM Changeset in webkit [19126] by bdakin

• 3 edits
  4 adds in trunk

WebCore:

Reviewed and merged by Beth.

Fix for <rdar://problem/4921692> not processing comments inside a
title element leads to XSS (for at least BlogSpot)

This is equivalent to commit 626791 in the KDE tree. Basically, the
patch allows comment parsing in titles to avoid capturing a
<script> tag by accident.

• html/HTMLTokenizer.cpp: (WebCore::HTMLTokenizer::parseSpecial): (WebCore::HTMLTokenizer::parseComment):

LayoutTests:

Test for <rdar://problem/4921692> not processing comments inside a
title element leads to XSS (for at least BlogSpot)

• fast/parser/parseCommentsInTitles-expected.checksum: Added.
• fast/parser/parseCommentsInTitles-expected.png: Added.
• fast/parser/parseCommentsInTitles-expected.txt: Added.
• fast/parser/parseCommentsInTitles.html: Added.

10:12 AM Changeset in webkit [19125] by brmorris

• 2 edits in S60/trunk/WebCore

bujtas <​zbujtas@gmail.com>

Reviewed by Sachin.
DESC: ISHG-6S8Q5X fix list marker offset problem.
markers get off when the marker image, specified in CSS, is 404.
fix: issue relayout, when the errorImage is set on the marker.

​http://bugs.webkit.org/show_bug.cgi?id=12404

10:12 AM Changeset in webkit [19124] by brmorris

• 2 edits in S60/branches/3.1m/WebCore

bujtas <​zbujtas@gmail.com>

Reviewed by Sachin.
DESC: ISHG-6S8Q5X fix list marker offset problem.
markers get off when the marker image, specified in CSS, is 404.
fix: issue relayout, when the errorImage is set on the marker.

​http://bugs.webkit.org/show_bug.cgi?id=12404

9:41 AM Changeset in webkit [22837] by aroben

• 6 edits in branches/WindowsMerge/WebKitWin

Reviewed by Anders.

Part of <rdar://problem/4855618> HTML-based load failure pages.

• Interfaces/IWebError.idl: Added failingURL method declaration.
• WebError.cpp: (WebError::failingURL): Implemented.
• WebError.h: Updated for interface changes.
• WebFrame.cpp: (WebFrame::loadHTMLString): New protected helper method called by IWebFrame::loadHTMLString and IWebFrame::loadAlternateHTMLString. (WebFrame::loadAlternateHTMLString): Calls loadHTMLString.
• WebFrame.h: Added new declaration.

8:09 AM Changeset in webkit [19123] by brmorris

• 2 edits in S60/branches/3.1m/WebKit

spadma

Reviewed by Zalan.
DESC: ResolveEmbedded link not called for plugins (ASAA-6X5DX8)
​http://bugzilla.opendarwin.org/show_bug.cgi?id=12400

• BrowserView/src/WebKitLoader.cpp: (CWebKitLoader::LoadPluginL):

7:39 AM Changeset in webkit [19122] by darin

• 2 edits in trunk/WebCore

• ksvg2/svg/SVGPreserveAspectRatio.cpp: Add a couple of comments.

7:33 AM Changeset in webkit [19121] by darin

• 3 edits
  3 adds in trunk

LayoutTests:

Reviewed by Maciej.

• test for bugs found by code inspection in the SVGPreserveAspectRatio parser

• svg/dom/preserve-aspect-ratio-parser-expected.txt: Added.
• svg/dom/preserve-aspect-ratio-parser.html: Added.

WebCore:

Reviewed by Maciej.

• fix ​http://bugs.webkit.org/show_bug.cgi?id=12365 Reproducible crash in WebCore::SVGPreserveAspectRatio::parsePreserveAspectRatio in svg/W3C-SVG-1.1/animate-elem-40-t.svg under guard malloc
• fix bug where the parser accepts "meet" or "slice" as an entire string
• fix bug where "slice" is ignored if the old alignment was "none" even if the newly-parsed alignment is something other than "none", and "slice" is not ignored if the old alignment was not "none", but the newly-parsed alignment is "none"
• optimize to only call notifyAttributeChange if the attribute actually changed

Test: svg/dom/preserve-aspect-ratio-parser-test.html

• ksvg2/svg/SVGPreserveAspectRatio.cpp: (WebCore::checkString): Added an overload for using char* constants which is much simpler than the old one, and almost as fast. (WebCore::SVGPreserveAspectRatio::parsePreserveAspectRatio): Fail if there's no more text after "defer". Fail if there's neither "none" nor an x/y specified. Check that there's more text before looking at the "m" and "s" for "meet" and "slice". Merged the failure and success cases. Only call notifyAttributeChange if something changed.

I think all the "optimized" code where we check the first character before
calling checkString isn't really helpful, but I decided to leave it as-is
for now.

6:54 AM Changeset in webkit [19120] by lars

• 3 edits in trunk/WebCore

Reviewed by Zack

Trivial implementation of getMIMETypeForExtension.
Use the above method in the ResourceHandleManager to set
a mimetype on Resources that don't have one set (which
includes everything loaded from local disk).

Should fix quite a few test cases :)

6:20 AM Changeset in webkit [19119] by sullivan

• 3 edits in trunk/WebKit

Reviewed by Kevin, Geoff, Brady, and Darin

• fixed <rdar://problem/4918446> Safari's temp files (PDF's) should be in a sub-folder when calling Preview

• WebView/WebPDFView.mm: (-[WebPDFView _path]): use _temporaryPDFDirectoryPath method instead of #defines for hardwiring strings; stop bad practice of modifying the const char* returned by fileSystemRepresentation (-[WebPDFView _temporaryPDFDirectoryPath]): new method, lazily creates and returns a secure temporary directory created with NSTemporaryDirectory() and mkdtemp

• English.lproj/StringsNotToBeLocalized.txt: Updated for these and other recent changes

6:03 AM Changeset in webkit [19118] by lars

• 2 edits
  808 adds in trunk/LayoutTestResults

Add more test results for the Qt build.

5:06 AM Changeset in webkit [19117] by zack

• 2 edits in trunk/WebCore

Path from Harald Fernengel to fix compile on platforms
where qreal is not a double.
r=me

4:48 AM Changeset in webkit [19116] by zack

• 7 edits
  1 add in trunk/WebCore

implement wheel event handling and use eventhandler explictly.
r=lars

3:07 AM BuildingOnWindows edited by ext-daniel.knauth@nokia.com

(diff)

2:54 AM Changeset in webkit [19115] by rwlbuis

• 3 edits
  4 adds in trunk

Reviewed by Maciej.

​http://bugs.webkit.org/show_bug.cgi?id=12391
SVG containers with viewports should respect overflow when hit testing
Take into account the overflow property when hit testing <svg>.

2:04 AM Changeset in webkit [19114] by lars

• 2 edits in trunk/WebKitQt

Reviewed by Zack

Emit the loadDone() signal from teh onload event callback
again. This hangs on one or two test cases, but removes
hundreds of failures elsewhere.

2:04 AM Changeset in webkit [19113] by lars

• 2 edits in trunk/LayoutTestResults

remove dom/html/level2/html from skipped tests.

12:56 AM Changeset in webkit [19112] by lars

• 611 adds in trunk/LayoutTestResults

Add css1 and css2.1 test results for the Qt build

12:54 AM Changeset in webkit [19111] by lars

• 4 edits
  1 delete in trunk/WebKitTools

Reviewed by Darin and Zack

Fix ​http://bugs.webkit.org/show_bug.cgi?id=12398

Move the test results for Qt into a directory of it's own
(WebKit/LayoutTestResults/qt). Leave the Mac results where
they are for now and share the text only results between
Mac and Qt.

Add support for a LayoutTestResults/platform/Skipped file
to run-webkit-tests to be able to ignore certain tests

Remove the old tests-skipped.txt from Qts DumpRenderTree
implementation.

12:47 AM Changeset in webkit [22836] by adachan

• 3 edits in branches/WindowsMerge/WebKitWin

2007-01-25 Ada Chan <​adachan@apple.com>

WebKitWin:

Reviewed by Maciej.

Added init methods instead of doing all the initialization in the constructor.
There was a lifetime issue with passing the WebDownload to delegate calls when
its ref count is still 0.

• WebDownload.cpp: (WebDownload::init): (WebDownload::createInstance):
• WebDownload.h:

12:38 AM Changeset in webkit [19110] by aliceli1

• 3 edits
  4 adds in trunk

LayoutTests:

Reviewed by Maciej.

added test for
<rdar://problem/4808039> REGRESSION: eBay ID check never completes
also known as
​http://bugs.webkit.org/show_bug.cgi?id=11230

• http/tests/misc/location-replace-crossdomain-expected.txt: Added.
• http/tests/misc/location-replace-crossdomain.html: Added.
• http/tests/misc/resources/failure.html: Added.
• http/tests/misc/resources/success.html: Added.

WebCore:

Reviewed by Maciej.

fixed <rdar://problem/4808039> REGRESSION: eBay ID check never completes
Also known as ​http://bugs.webkit.org/show_bug.cgi?id=11230

• bindings/js/kjs_window.cpp: It's not necessary to check safeScript when doing a location.replace, .reload, or .assign (KJS::Location::getOwnPropertySlot): (KJS::LocationFunc::callAsFunction):

12:34 AM Changeset in webkit [19109] by aliceli1

• 4 edits
  2 adds in trunk

LayoutTests:

Reviewed by Geoff and Adele

added test for <rdar://problem/4866655> REGRESSION: Copying out of Calculator & CIFIlterWidget is broken on 9A320

• editing/execCommand/copy-without-selection-expected.txt: Added.
• editing/execCommand/copy-without-selection.html: Added.

WebCore:

Reviewed by Geoff and Adele.

fixed <rdar://problem/4866655> REGRESSION: Copying out of Calculator & CIFIlterWidget is broken on 9A320

• editing/Editor.cpp: (WebCore::Editor::canDHTMLCut): (WebCore::Editor::canDHTMLCopy): (WebCore::Editor::tryDHTMLCopy): (WebCore::Editor::tryDHTMLCut): Reverting some mistakes made in r17504: Anything in r17503 that used to call mayDHTML* still calls canDHTML* as it was changed in r17504, where canDHTML* now calls !selectionController::isInPasswordField instead of editor::canCopy Anything else in r17503 that used to call mayCopy now calls (!)selectionController::isInPasswordField.

• editing/JSEditor.cpp: copy via JS should be enabled as long as selection is not in a password field. a range selection shouldn't be required.