WebKit-https.git
12 years agoRemove the --64-bit argument from scripts in favor of them detecting when 64-bit...
mrowe@apple.com [Sat, 14 Mar 2009 06:17:16 +0000 (06:17 +0000)]
Remove the --64-bit argument from scripts in favor of them detecting when 64-bit should be preferred.

Reviewed by Timothy Hatcher.

The scripts will automatically target 64-bit if the system and hardware support it.  This can be
overridden by passing --32-bit to individual scripts, or using set-webkit-configuration --32-bit
to make the override persistent.

* Scripts/build-webkit: Remove architecture-related code.
* Scripts/gdb-safari: Remove architecture-related code, and clean up how the environment variables are passed to gdb.
* Scripts/run-javascriptcore-tests: Remove architecture-related code.
* Scripts/run-safari: Ditto.
* Scripts/run-webkit-tests: Ditto.
* Scripts/set-webkit-configuration: Handle the --32-bit and --64-bit arguments.  The --32-bit argument will set the
architecture preference to the 32-bit architecture of the machine.  The --64-bit argument will remove any architecture
override that is in effect so that 64-bit support will be automatically detected.
* Scripts/webkitdirs.pm: Add auto-detection of the best architecture for the machine, and the ability to override the
auto-detection.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41696 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoPrevent AllInOneFile.cpp and ProfileGenerator.cpp from rebuilding unnecessarily when
mrowe@apple.com [Sat, 14 Mar 2009 06:17:08 +0000 (06:17 +0000)]
Prevent AllInOneFile.cpp and ProfileGenerator.cpp from rebuilding unnecessarily when
switching between building in Xcode and via build-webkit.

Reviewed by David Kilzer.

build-webkit passes FEATURE_DEFINES to xcodebuild, resulting in it being present in the
Derived Sources build settings.  When building in Xcode, this setting isn't present so
Xcode reruns the script build phases.  This results in a new version of TracingDtrace.h
being generated, and the files that include it being rebuilt.

* JavaScriptCore.xcodeproj/project.pbxproj:  Don't regenerate TracingDtrace.h if it is
already newer than the input file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41695 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 John Abd-El-Malek <jam@google.com>
dimich@chromium.org [Sat, 14 Mar 2009 00:53:28 +0000 (00:53 +0000)]
2009-03-13  John Abd-El-Malek  <jam@google.com>

        Reviewed by Darin Fisher.

        https://bugs.webkit.org/show_bug.cgi?id=24593
        Added requestorID so we can track the request to its WebView without using frame (which was removed).

        * platform/network/chromium/ResourceRequest.h:
        (WebCore::ResourceRequest::ResourceRequest):
        (WebCore::ResourceRequest::requestorID):
        (WebCore::ResourceRequest::setRequestorID):
        (WebCore::ResourceRequest::setTargetType):
        (WebCore::ResourceRequest::policyURL):
        (WebCore::ResourceRequest::setPolicyURL):
        (WebCore::ResourceRequest::requestorPid):
        (WebCore::ResourceRequest::setRequestorPid):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41694 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Mike Belshe <mike@belshe.com>
dglazkov@chromium.org [Fri, 13 Mar 2009 23:07:24 +0000 (23:07 +0000)]
2009-03-13  Mike Belshe  <mike@belshe.com>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24561
        Add custom V8 bindings for HTMLElementCanvas, Location.

        * bindings/v8/custom/V8HTMLCanvasElementCustom.cpp: Added.
        * bindings/v8/custom/V8LocationCustom.cpp: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41693 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Stephen White <senorblanco@chromium.org>
eric@webkit.org [Fri, 13 Mar 2009 22:47:20 +0000 (22:47 +0000)]
2009-03-13  Stephen White  <senorblanco@chromium.org>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=24584

        Fix transparent text rendering on Chromium.  FontChromiumWin
        was calling beginTransparencyLayer()/endTransparencyLayer(), with
        a TransparencyWin inside to do GDI ClearType rendering over an
        opaque background.  TransparencyWin does its special sauce
        in the destructor, but it was being called too late to be used
        correctly in the layer.  Put the special sauce into a new function,
        composite(), and call that explicitly instead.

        * platform/graphics/chromium/FontChromiumWin.cpp:
        (WebCore::TransparencyAwareFontPainter::TransparencyAwareFontPainter::~TransparencyAwareFontPainter):
        * platform/graphics/chromium/TransparencyWin.cpp:
        (WebCore::TransparencyWin::~TransparencyWin):
        (WebCore::TransparencyWin::composite):
        (WebCore::TransparencyWin::init):
        * platform/graphics/chromium/TransparencyWin.h:
        * rendering/RenderThemeChromiumWin.cpp:
        (WebCore::):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41692 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Mike Belshe <mike@belshe.com>
dglazkov@chromium.org [Fri, 13 Mar 2009 22:46:44 +0000 (22:46 +0000)]
2009-03-13  Mike Belshe  <mike@belshe.com>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24562
        Add custom implementation for getCSSCanvasContext to V8 bindings.

        * bindings/v8/custom/V8DocumentCustom.cpp: Added new method.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41691 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Ojan Vafai <ojan@chromium.org>
eric@webkit.org [Fri, 13 Mar 2009 22:43:18 +0000 (22:43 +0000)]
2009-03-13  Ojan Vafai  <ojan@chromium.org>

        Reviewed by Eric Seidel.

        Add tests for various types of textareas. Tests in both
        quirks and standards modes. Although only the standards
        modes tests fit in the pixel results.
        https://bugs.webkit.org/show_bug.cgi?id=24370

        * fast/forms/basic-textareas.html: Added.
        * platform/mac/fast/forms/basic-textareas-expected.checksum: Added.
        * platform/mac/fast/forms/basic-textareas-expected.png: Added.
        * platform/mac/fast/forms/basic-textareas-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41690 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Mike Belshe <mike@belse.com>
dglazkov@chromium.org [Fri, 13 Mar 2009 22:42:29 +0000 (22:42 +0000)]
2009-03-13  Mike Belshe  <mike@belse.com>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24580
        Fix query() to match KURL behavior.

        * platform/KURLGoogle.cpp:
        (WebCore::KURL::query): remove extra logic around question mark.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41689 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Fri, 13 Mar 2009 22:32:47 +0000 (22:32 +0000)]
2009-03-13  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker and Geoffrey Garen.

        <rdar://problem/6590384>
        REGRESSION (Safari 3-4): Tiger-only Crash occurs at WebView hostWindow () after reloading a set of tabs then quitting

        When we're doing fast teardown, plug-in views can be destroyed from -[WebView dealloc]'s [super dealloc] call,
        and thus calling -[WebView hostWindow] will crash since _private is nil.

        * WebView/WebView.mm:
        (-[WebView hostWindow]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41688 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoBug 24474: AX: in multi-body tables, asking for a cell at a specific coordinate can...
cfleizach@apple.com [Fri, 13 Mar 2009 22:20:51 +0000 (22:20 +0000)]
Bug 24474: AX: in multi-body tables, asking for a cell at a specific coordinate can return nil
https://bugs.webkit.org/show_bug.cgi?id=24474

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41687 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Jian Li <jianli@chromium.org>
dimich@chromium.org [Fri, 13 Mar 2009 22:00:09 +0000 (22:00 +0000)]
2009-03-13  Jian Li  <jianli@chromium.org>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24589
        Upstream changes to V8 event listeners (Chromium r11133) in order to
        fix worker functionality break in Chromium.

        * bindings/v8/V8AbstractEventListener.cpp:
        (WebCore::V8AbstractEventListener::invokeEventHandler):
        (WebCore::V8AbstractEventListener::handleEvent):
        * bindings/v8/V8AbstractEventListener.h:
        * bindings/v8/V8WorkerContextEventListener.cpp:
        (WebCore::V8WorkerContextEventListener::handleEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41686 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Norbert Leser <norbert.leser@nokia.com>
dimich@chromium.org [Fri, 13 Mar 2009 21:52:19 +0000 (21:52 +0000)]
2009-03-13  Norbert Leser  <norbert.leser@nokia.com>

        Reviewed by Darin Adler.

        Resolved name conflict with globally defined tzname in Symbian.
        Replaced with different name instead of using namespace qualifier
        (appeared to be less clumsy).

        * runtime/DateMath.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41685 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Dimitri Glazkov <dglazkov@chromium.org>
dglazkov@chromium.org [Fri, 13 Mar 2009 21:31:02 +0000 (21:31 +0000)]
2009-03-13  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Timothy Hatcher.

        https://bugs.webkit.org/show_bug.cgi?id=24524
        Introduce ScriptObject and ScriptFunctionCall abstractions.

        * GNUmakefile.am: Added ScriptObject and ScriptFunctionCall to project.
        * WebCore.pro: Ditto.
        * WebCore.vcproj/WebCore.vcproj: Ditto.
        * WebCore.xcodeproj/project.pbxproj: Ditto.
        * WebCoreSources.bkl: Ditto.
        * bindings/js/ScriptFunctionCall.cpp: Added.
        * bindings/js/ScriptFunctionCall.h: Added.
        * bindings/js/ScriptObject.cpp: Added.
        * bindings/js/ScriptObject.h: Added.
        * bindings/js/ScriptObjectQuarantine.cpp: Added.
        * bindings/js/ScriptObjectQuarantine.h: Added.
        * inspector/InspectorController.cpp:
        (WebCore::InspectorController::populateScriptObjects): Changed to use InspectorDatabaseResource::bind.
        (WebCore::InspectorController::resetScriptObjects): Changed to use InspectorDatabaseResource::unbind.
        (WebCore::InspectorController::didOpenDatabase): Changed to use InspectorDatabaseResource::unbind.
        * inspector/InspectorController.h: Removed addScriptDatabaseResource and
        removeScriptDatabaseResource declarations.
        * inspector/InspectorDatabaseResource.cpp:
        (WebCore::InspectorDatabaseResource::InspectorDatabaseResource):
        (WebCore::InspectorDatabaseResource::bind): Added.
        (WebCore::InspectorDatabaseResource::unbind): Added.
        * inspector/InspectorDatabaseResource.h: Added bind and unbind declarations.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41684 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Peter Kasting <pkasting@google.com>
pkasting@chromium.org [Fri, 13 Mar 2009 21:26:19 +0000 (21:26 +0000)]
2009-03-13  Peter Kasting  <pkasting@google.com>

        Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=24467
        Make Skia drawRect() and stroke behavior match CG.

        This makes drawRect() ignore the stroke width (like CG does), and
        adds a warning comment about that to the appropriate header.

        It also eliminates some hacky code in Skia's stroke preparation,
        which tried to adjust odd-width strokes to fall on pixel boundaries.
        Not only did this not match CG, it wouldn't necessarily work right,
        because there could be other transforms (e.g. full-page zoom) that
        would affect the stroke before it reached the device pixel level.

        * platform/graphics/GraphicsContext.h:
        * platform/graphics/skia/GraphicsContextSkia.cpp:
        (WebCore::GraphicsContext::drawRect):
        (WebCore::GraphicsContext::fillRect):
        * platform/graphics/skia/PlatformContextSkia.cpp:
        (PlatformContextSkia::drawRect):
        (PlatformContextSkia::setupPaintForStroking):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41683 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Jian Li <jianli@chromium.org>
dimich@chromium.org [Fri, 13 Mar 2009 20:45:17 +0000 (20:45 +0000)]
2009-03-13  Jian Li  <jianli@chromium.org>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24583
        Need to change scope of constructor and destructor of V8ObjectEventListener from private
        to protected so that they could be used in its derived class V8WorkerContextEventListener.

        * bindings/v8/V8ObjectEventListener.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41682 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Fri, 13 Mar 2009 19:53:48 +0000 (19:53 +0000)]
2009-03-13  Anders Carlsson  <andersca@apple.com>

        And yet another attempt...

        * Plugins/WebNetscapePluginEventHandlerCocoa.h:
        (WebNetscapePluginEventHandlerCocoa::installKeyEventHandler):
        (WebNetscapePluginEventHandlerCocoa::removeKeyEventHandler):
        * Plugins/WebNetscapePluginEventHandlerCocoa.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41681 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Fri, 13 Mar 2009 19:52:10 +0000 (19:52 +0000)]
2009-03-13  Anders Carlsson  <andersca@apple.com>

        Another attempt at fixing the build.

        * Plugins/WebNetscapePluginEventHandlerCocoa.mm:
        (WebNetscapePluginEventHandlerCocoa::WebNetscapePluginEventHandlerCocoa):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41680 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Fri, 13 Mar 2009 19:39:13 +0000 (19:39 +0000)]
2009-03-13  Anders Carlsson  <andersca@apple.com>

        Try to fix the SL build.

        * Plugins/WebNetscapePluginEventHandlerCocoa.h:
        * Plugins/WebNetscapePluginEventHandlerCocoa.mm:
        (WebNetscapePluginEventHandlerCocoa::keyDown):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41679 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 David Levin <levin@chromium.org>
levin@chromium.org [Fri, 13 Mar 2009 19:13:16 +0000 (19:13 +0000)]
2009-03-13  David Levin  <levin@chromium.org>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24579
        Many V8*Event* files are misplaced.

        This code is infrastructure for supporting the bindings and thus should
        be in the bindings/v8 directory.

        * bindings/v8/V8AbstractEventListener.cpp: Renamed from WebCore/bindings/v8/custom/V8AbstractEventListener.cpp.
        * bindings/v8/V8AbstractEventListener.h: Renamed from WebCore/bindings/v8/custom/V8AbstractEventListener.h.
        * bindings/v8/V8LazyEventListener.cpp: Renamed from WebCore/bindings/v8/custom/V8LazyEventListener.cpp.
        * bindings/v8/V8LazyEventListener.h: Renamed from WebCore/bindings/v8/custom/V8LazyEventListener.h.
        * bindings/v8/V8ObjectEventListener.cpp: Renamed from WebCore/bindings/v8/custom/V8ObjectEventListener.cpp.
        * bindings/v8/V8ObjectEventListener.h: Renamed from WebCore/bindings/v8/custom/V8ObjectEventListener.h.
        * bindings/v8/V8WorkerContextEventListener.cpp: Renamed from WebCore/bindings/v8/custom/V8WorkerContextEventListener.cpp.
        * bindings/v8/V8WorkerContextEventListener.h: Renamed from WebCore/bindings/v8/custom/V8WorkerContextEventListener.h.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41678 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoWebCore:
adele@apple.com [Fri, 13 Mar 2009 19:06:46 +0000 (19:06 +0000)]
WebCore:

2009-03-13  Adele Peterson  <adele@apple.com>

        Reviewed by Justin Garcia.

        Fix for <rdar://problem/5089327> Color of quoted content is wrong when pasted inside other quoted content

        Test: editing/pasteboard/5089327.html

        Handle spans being pasted within a quoted region in the same way we handle "paste as quotation" content.

        * editing/ReplaceSelectionCommand.cpp:
        (WebCore::handleStyleSpansBeforeInsertion):
        (WebCore::ReplaceSelectionCommand::handleStyleSpans):

LayoutTests:

2009-03-13  Adele Peterson  <adele@apple.com>

        Reviewed by Justin Garcia.

        Tests for <rdar://problem/5089327> Color of quoted content is wrong when pasted inside other quoted content

        * editing/pasteboard/5089327.html: Added.
        * platform/mac/editing/pasteboard/5089327-expected.checksum: Added.
        * platform/mac/editing/pasteboard/5089327-expected.png: Added.
        * platform/mac/editing/pasteboard/5089327-expected.txt: Added.

        Updated colors for pasted content.

        * platform/mac/editing/pasteboard/paste-blockquote-into-blockquote-4-expected.checksum:
        * platform/mac/editing/pasteboard/paste-blockquote-into-blockquote-4-expected.png:
        * platform/mac/editing/pasteboard/paste-blockquote-into-blockquote-4-expected.txt:
        * platform/mac/editing/pasteboard/paste-blockquote-into-blockquote-expected.checksum:
        * platform/mac/editing/pasteboard/paste-blockquote-into-blockquote-expected.png:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41677 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Fri, 13 Mar 2009 18:57:36 +0000 (18:57 +0000)]
2009-03-13  Anders Carlsson  <andersca@apple.com>

        Reviewed by Kevin Decker.

        Update to match new API.

        * NetscapeInputMethodPlugin/main.m:
        (NPP_New):
        (NPP_Destroy):
        (NPP_SetWindow):
        (handleDraw):
        (invalidatePlugin):
        (handleFocusChanged):
        (handleMouseMoved):
        (handleMouseDown):
        (handleTextFieldKeyDown):
        (handleTextInput):
        (NPP_HandleEvent):
        (NPP_URLNotify):
        (NPP_GetValue):
        (NPP_SetValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41676 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Greg Bolsinga <bolsinga@apple.com>
bolsinga@apple.com [Fri, 13 Mar 2009 18:56:42 +0000 (18:56 +0000)]
2009-03-13  Greg Bolsinga  <bolsinga@apple.com>

        Reviewed by Simon Fraser.

        Update Geolocation perimission dialogs to be asynchronous.
        https://bugs.webkit.org/show_bug.cgi?id=24505

        WebGeolocation is a wrapper around WebCore::Geolocation. It mimics the
        coding style set by WebSecurityOrigin.

        WebChromeClient now calls the private UI delegate method
        -webView:frame:requestGeolocationPermission:securityOrigin:

        * WebCoreSupport/WebChromeClient.h:
        * WebCoreSupport/WebChromeClient.mm:
        (WebChromeClient::requestGeolocationPermissionForFrame):
        * WebCoreSupport/WebGeolocation.mm: Added.
        (WebCore::if):
        (-[WebGeolocation shouldClearCache]):
        (-[WebGeolocation setIsAllowed:]):
        (-[WebGeolocation dealloc]):
        * WebCoreSupport/WebGeolocationInternal.h: Added.
        * WebCoreSupport/WebGeolocationPrivate.h: Added.
        * WebView/WebUIDelegatePrivate.h:

2009-03-13  Greg Bolsinga  <bolsinga@apple.com>

        Reviewed by Simon Fraser.

        Update Geolocation perimission dialogs to be asynchronous.
        https://bugs.webkit.org/show_bug.cgi?id=24505

        Add WebGeolocation, a wrapper around WebCore::Geolocation. It mimics the
        coding style set by WebSecurityOrigin.

        * WebKit.xcodeproj/project.pbxproj:

2009-03-13  Greg Bolsinga  <bolsinga@apple.com>

        Reviewed by Simon Fraser.

        Update Geolocation perimission dialogs to be asynchronous.
        https://bugs.webkit.org/show_bug.cgi?id=24505

        Geolocation now requests permission from the Chrome asynchronously.
        The Chrome is passed the Geolocation object, and the Chrome sets the permission
        on the Geolocation. Geolocation also tracks if the Chrome should clear its cache
        of SecurityOrigins with geolocation permission. This is so that the GeolocationService
        can inform the Chrome of its request, and the Chrome is also free to implement its
        own policy.

        * WebCore.base.exp:
        * WebCore.xcodeproj/project.pbxproj:
        * page/Chrome.cpp:
        (WebCore::Chrome::requestGeolocationPermissionForFrame):
        * page/Chrome.h:
        * page/ChromeClient.h:
        (WebCore::ChromeClient::requestGeolocationPermissionForFrame):
        * page/Geolocation.cpp:
        (WebCore::Geolocation::Geolocation):
        (WebCore::Geolocation::getCurrentPosition):
        (WebCore::Geolocation::watchPosition):
        (WebCore::Geolocation::setIsAllowed):
        (WebCore::Geolocation::displayChallengeIfNecessary):
        (WebCore::Geolocation::geolocationServicePositionChanged):
        * page/Geolocation.h:
        (WebCore::Geolocation::isAllowed):
        (WebCore::Geolocation::setShouldClearCache):
        (WebCore::Geolocation::shouldClearCache):
        (WebCore::Geolocation::):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41675 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoWebCore:
andersca@apple.com [Fri, 13 Mar 2009 18:44:19 +0000 (18:44 +0000)]
WebCore:

2009-03-13  Anders Carlsson  <andersca@apple.com>

        Reviewed by Dan Bernstein.

        <rdar://problem/6610666> Revise the Cocoa event model text API

        Add a NPCocoaEventTextInput event type. Remove the text input variables.

        * bridge/npapi.h:
        (_NPCocoaEvent::):

WebKit:

2009-03-13  Anders Carlsson  <andersca@apple.com>

        Reviewed by Dan Bernstein.

        <rdar://problem/6610666> Revise the Cocoa event model text API

        Remove nptextinput.h

        * WebKit.xcodeproj/project.pbxproj:

WebKit/mac:

2009-03-13  Anders Carlsson  <andersca@apple.com>

        Reviewed by Dan Bernstein.

        <rdar://problem/6610666> Revise the Cocoa event model text API

        Replace the text input API with a simpler API that uses a separate text input window.

        * Plugins/WebNetscapePluginEventHandlerCocoa.h:
        * Plugins/WebNetscapePluginEventHandlerCocoa.mm:
        (WebNetscapePluginEventHandlerCocoa::WebNetscapePluginEventHandlerCocoa):
        Initialize m_keyEventHandler to 0.

        (WebNetscapePluginEventHandlerCocoa::keyDown):
        If the plug-in returns 0 from NPP_HandleEvent, pass the event to the TSM machinery.

        (WebNetscapePluginEventHandlerCocoa::focusChanged):
        Install/remove the key event handler as needed.

        (WebNetscapePluginEventHandlerCocoa::handleTSMEvent):
        Get the text and send a TextInput event.

        * Plugins/WebNetscapePluginView.h:
        * Plugins/WebNetscapePluginView.mm:
        Remove the old text input API.

        (-[WebNetscapePluginView inputContext]):
        Always return nil here.

        * Plugins/npapi.mm:
        * Plugins/nptextinput.h: Removed.

WebKitTools:

2009-03-13  Anders Carlsson  <andersca@apple.com>

        Reviewed by Dan Bernstein.

        <rdar://problem/6610666> Revise the Cocoa event model text API

        Add a case statement for NPCocoaEventTextInput.

        * DumpRenderTree/TestNetscapePlugIn.subproj/main.cpp:
        (handleEventCocoa):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41674 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Anders.
kdecker@apple.com [Fri, 13 Mar 2009 18:18:44 +0000 (18:18 +0000)]
    Reviewed by Anders.

        <rdar://problem/6630340> REGRESSION (39114-39115): Unity Web Player no longer works if Flip4Mac is also installed

        The code assumed if we have a plug-in that supports "application/x-oleobject" we should always prefer the object tag
        over of an embed tag. That assumption can cause the Mac platform to load the wrong plug-in, as Flip4Mac claims supports
        for x-oleobject.

        * rendering/RenderPartObject.cpp:
        (WebCore::shouldUseEmbedDescendant): Made the Mac platform always return true here.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41673 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Pamela Greene <pam@chromium.org>
pam@chromium.org [Fri, 13 Mar 2009 17:45:42 +0000 (17:45 +0000)]
2009-03-13  Pamela Greene  <pam@chromium.org>

        Reviewed by Darin Fisher.

        Add tests for window property lookup precedence in the presence of
        conflicting names in a child frame and in a named document element.

        https://bugs.webkit.org/show_bug.cgi?id=24170

        * fast/dom/Window/resources/window-properties.js: Added.
        * fast/dom/Window/window-lookup-precedence-expected.txt: Added.
        * fast/dom/Window/window-lookup-precedence.html: Added.
        * fast/js/resources/js-test-pre.js:
        (shouldEvaluateTo): Added. Compares expected values of various types.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41672 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Jian Li <jianli@chromium.org>
levin@chromium.org [Fri, 13 Mar 2009 17:32:19 +0000 (17:32 +0000)]
2009-03-13  Jian Li  <jianli@chromium.org>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24559
        Need to port JSC fix (r41565) to V8 in order to fix layout test onload-single-line-comment.html.

        * bindings/v8/custom/V8LazyEventListener.cpp:
        (WebCore::V8LazyEventListener::getListenerFunction):
        (WebCore::V8LazyEventListener::getWrappedListenerFunction):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41671 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Jian Li <jianli@chromium.org>
levin@chromium.org [Fri, 13 Mar 2009 17:32:12 +0000 (17:32 +0000)]
2009-03-13  Jian Li  <jianli@chromium.org>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24557
        This is to support running multiple workers in a single worker process in chromium.

        * bindings/v8/custom/V8WorkerContextEventListener.cpp:
        (WebCore::V8WorkerContextEventListener::handleEvent):
        Add locker to v8 event listener of worker context.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41670 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Pamela Greene <pam@chromium.org>
pam@chromium.org [Fri, 13 Mar 2009 17:09:52 +0000 (17:09 +0000)]
2009-03-13  Pamela Greene  <pam@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Add test cases to ensure that DOM constructors have higher lookup
        precedence than a document element with the same name.

        https://bugs.webkit.org/show_bug.cgi?id=24520

        * fast/dom/dom-constructors-expected.txt:
        * fast/dom/dom-constructors.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41669 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Pamela Greene <pam@chromium.org>
pam@chromium.org [Fri, 13 Mar 2009 16:20:43 +0000 (16:20 +0000)]
2009-03-13  Pamela Greene  <pam@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Add a test to check that the JS wrapper for the document in an
        inner frame is created in the correct context when it's accessed
        from the outer frame first.

        https://bugs.webkit.org/show_bug.cgi?id=24564

        * fast/dom/resources/wrapper-context-inner.html: Added.
        * fast/dom/wrapper-context-expected.txt: Added.
        * fast/dom/wrapper-context.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41668 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Sam Weinig.
ap@webkit.org [Fri, 13 Mar 2009 13:25:52 +0000 (13:25 +0000)]
    Reviewed by Sam Weinig.

        https://bugs.webkit.org/show_bug.cgi?id=24575
        XHR response is incorrect after a network error.

        Tests: http/tests/xmlhttprequest/cross-site-denied-response-sync-2.html
               http/tests/xmlhttprequest/cross-site-denied-response-sync.html
               http/tests/xmlhttprequest/cross-site-denied-response.html

        * xml/XMLHttpRequest.cpp:
        (WebCore::XMLHttpRequest::networkError): Call internalAbort() to fully reset the request.
        (WebCore::XMLHttpRequest::didFailRedirectCheck): No need to call internalAbort() here. Note
        that since internalAbort() can drop GC protection, it is not safe to use the object after
        this call.
        (WebCore::XMLHttpRequest::didReceiveData): Check that the request wasn't aborted, and
        return early if it was. This can happen during sync requests, as the loader does not know
        that it was aborted, and just synthesizes all callbacks.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41667 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Updating results after the fix for bug 3547.
ap@webkit.org [Fri, 13 Mar 2009 12:57:19 +0000 (12:57 +0000)]
    Updating results after the fix for bug 3547.

        * fast/loader/xmlhttprequest-missing-file-exception-expected.txt: The status text for
        a failed local file load is no longer "OK", which is good.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41666 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Adam Bergkvist <adam.bergkvist@ericsson.com>
ap@webkit.org [Fri, 13 Mar 2009 11:06:47 +0000 (11:06 +0000)]
2009-03-13  Adam Bergkvist <adam.bergkvist@ericsson.com>

        Reviewed by Alexey Proskuryakov.

        https://bugs.webkit.org/show_bug.cgi?id=3547
        XMLHttpRequest.statusText returns always "OK"

        Covered by existing tests (which now pass on all platforms but Mac).

        * xml/XMLHttpRequest.cpp:
        (WebCore::XMLHttpRequest::statusText): Return ResourceResponse status text. It is now up
        to each platform to correctly set the status text or set it to "OK" to retain current
        behavior.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41665 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-13 Adam Bergkvist <adam.bergkvist@ericsson.com>
ap@webkit.org [Fri, 13 Mar 2009 10:05:42 +0000 (10:05 +0000)]
2009-03-13  Adam Bergkvist <adam.bergkvist@ericsson.com>

        Reviewed by Alexey Proskuryakov.

        https://bugs.webkit.org/show_bug.cgi?id=24349
        [QT] HTTP status text is never set

        Set HTTP status text to the reason phrase attribute of QNetworkReply.

        * platform/network/qt/QNetworkReplyHandler.cpp:
        (WebCore::QNetworkReplyHandler::sendResponseIfNeeded):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41664 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago - Skipped file fix
mitz@apple.com [Fri, 13 Mar 2009 06:23:36 +0000 (06:23 +0000)]
    - Skipped file fix

        * platform/mac-tiger/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41663 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Fri, 13 Mar 2009 06:07:10 +0000 (06:07 +0000)]
2009-03-12  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Mark Rowe

        <rdar://problem/6622300>: Reproducible crash on
        <http://www.editgrid.com/explore/tnc/dave/FusionChart%3A_Candlestick>

        Prevent CSSStyleSheet::checkLoaded() writing to freed memory when it gets
        deleted from under itself. The sheetLoaded() notification can allow scripts
        to run via HTMLTokenizer::executeScriptsWaitingForStylesheets(),
        which can cause the last ref to the CSSStyleSheet to be released.

        * css/CSSStyleSheet.cpp:
        (WebCore::CSSStyleSheet::checkLoaded):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41662 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agowx build fix. Let DerivedSources.make know we want database APIs.
kevino@webkit.org [Fri, 13 Mar 2009 05:28:57 +0000 (05:28 +0000)]
wx build fix. Let DerivedSources.make know we want database APIs.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41661 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago<rdar://problem/6548446> TCMalloc_SystemRelease should use madvise rather than re...
mrowe@apple.com [Fri, 13 Mar 2009 03:33:27 +0000 (03:33 +0000)]
<rdar://problem/6548446> TCMalloc_SystemRelease should use madvise rather than re-mmaping span of pages

* wtf/FastMalloc.cpp:
(WTF::mergeDecommittedStates): If either of the spans has been released to the system, release the other
span as well so that the flag in the merged span is accurate.
* wtf/Platform.h:
* wtf/TCSystemAlloc.cpp: Track decommitted spans when using MADV_FREE_REUSABLE / MADV_FREE_REUSE.
(TCMalloc_SystemRelease): Use madvise with MADV_FREE_REUSABLE when it is available.
(TCMalloc_SystemCommit): Use madvise with MADV_FREE_REUSE when it is available.
* wtf/TCSystemAlloc.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41660 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Adam Treat <adam.treat@torchmobile.com>
treat@webkit.org [Fri, 13 Mar 2009 03:10:31 +0000 (03:10 +0000)]
2009-03-12  Adam Treat  <adam.treat@torchmobile.com>

        Reviewed by Oliver Hunt.

        https://bugs.webkit.org/show_bug.cgi?id=24498
        Fix the Qt port to use the same algorithm for drawing dashed and dotted
        borders as the other ports.  This makes the Qt port pixel-for-pixel perfect
        compared to border drawing with Apple's canonical mac port and much closer
        to konqueror and firefox behavior.

        * platform/graphics/qt/GraphicsContextQt.cpp:
        (WebCore::GraphicsContext::drawLine):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41659 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Anders Carlsson.
mitz@apple.com [Fri, 13 Mar 2009 00:15:36 +0000 (00:15 +0000)]
    Reviewed by Anders Carlsson.

        - skip editing/selection/word-thai.html on Tiger because it has an older
          version of Unicode.

        * platform/mac-tiger/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41658 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Justin Garcia.
eric@webkit.org [Thu, 12 Mar 2009 23:05:52 +0000 (23:05 +0000)]
    Reviewed by Justin Garcia.

        Clean up DOMSelection to use some helper functions
        making the code smaller, and less error-prone.
        https://bugs.webkit.org/show_bug.cgi?id=19221

        I tried moving DOMSelection off of rangeCompliantEquivalent
        but failed.  VisibleSelection holds positions like (table, 1) to mean
        "after the table".

        * page/DOMSelection.cpp:
        (WebCore::DOMSelection::visibleSelection):
        (WebCore::anchorPosition):
        (WebCore::focusPosition):
        (WebCore::basePosition):
        (WebCore::extentPosition):
        (WebCore::DOMSelection::anchorNode):
        (WebCore::DOMSelection::anchorOffset):
        (WebCore::DOMSelection::focusNode):
        (WebCore::DOMSelection::focusOffset):
        (WebCore::DOMSelection::baseNode):
        (WebCore::DOMSelection::baseOffset):
        (WebCore::DOMSelection::extentNode):
        (WebCore::DOMSelection::extentOffset):
        * page/DOMSelection.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41657 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Dmitry Titov <dimich@chromium.org>
dimich@chromium.org [Thu, 12 Mar 2009 22:47:32 +0000 (22:47 +0000)]
2009-03-12  Dmitry Titov  <dimich@chromium.org>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24563
        Change ResourceRequest to be a struct rather then a class to match other declarations.

        * platform/network/chromium/ResourceRequest.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41656 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoAdd Mark.
andersca@apple.com [Thu, 12 Mar 2009 22:15:21 +0000 (22:15 +0000)]
Add Mark.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41655 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Thu, 12 Mar 2009 22:13:39 +0000 (22:13 +0000)]
2009-03-12  Anders Carlsson  <andersca@apple.com>

        Reviewed by Mike Thole.

        Fix <rdar://problem/6624105>.

        Make sure to process incoming messages for the NSEventTrackingRunLoopMode as well.

        * Plugins/Hosted/NetscapePluginHostProxy.mm:
        (WebKit::NetscapePluginHostProxy::NetscapePluginHostProxy):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41654 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Aaron Boodman <aa@chromium.org>
levin@chromium.org [Thu, 12 Mar 2009 22:05:33 +0000 (22:05 +0000)]
2009-03-12  Aaron Boodman  <aa@chromium.org>

        Reviewed by Dave Hyatt.

        Bug 16348: @font-face does not affect the default style
        <https://bugs.webkit.org/show_bug.cgi?id=16348>

        * fast/css/font-face-default-font.html:
        Handle the default font being either 'Times' or 'Times New Roman'.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41653 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 David Levin <levin@chromium.org>
levin@chromium.org [Thu, 12 Mar 2009 22:05:25 +0000 (22:05 +0000)]
2009-03-12  David Levin  <levin@chromium.org>

        Reviewed by  Darin Fisher.

        Bug 24530: width100percent-searchfield.html should be fixed for chromium.
        <https://bugs.webkit.org/show_bug.cgi?id=24530>

        Compensate for r39924 in chromium, which broke LayoutTests/fast/replaced/width100percent-searchfield.html

        * css/themeChromiumWin.css: Added.
        Overrides the changes to start supporting input[type="search"] on Windows (since
        Chromium doesn't do anything special for this).

        * rendering/RenderThemeChromiumWin.cpp:
        (WebCore::supportsFocus):
        (WebCore::RenderThemeChromiumWin::extraDefaultStyleSheet):
        (WebCore::RenderThemeChromiumWin::determineState):
        (WebCore::RenderThemeChromiumWin::getThemeData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41652 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Pamela Greene <pam@chromium.org>
pam@chromium.org [Thu, 12 Mar 2009 22:01:49 +0000 (22:01 +0000)]
2009-03-12  Pamela Greene  <pam@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Add a test to check that an invalid second argument of
        window.postMessage is ignored. Note that this is probably not the
        correct behavior according to the HTML5 spec, so this test will
        need to be changed once we support that.

        https://bugs.webkit.org/show_bug.cgi?id=24523

        * fast/dom/Window/window-postmessage-args-expected.txt: Added.
        * fast/dom/Window/window-postmessage-args.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41651 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Greg Bolsinga <bolsinga@apple.com>
bolsinga@apple.com [Thu, 12 Mar 2009 21:21:21 +0000 (21:21 +0000)]
2009-03-12  Greg Bolsinga  <bolsinga@apple.com>

        Reviewed by Antti Koivisto.

        https://bugs.webkit.org/show_bug.cgi?id=24506
        Geolocation in Safari differs from the spec, no Coordinates attribute on Position

        Get up to date.

        * DerivedSources.make:
        * GNUmakefile.am:
        * WebCore.pro:
        * WebCore.scons:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * WebCoreSources.bkl:
        * bindings/js/JSGeolocationCustom.cpp:
        (WebCore::createPositionOptions):
        * page/Coordinates.cpp: Added.
        (WebCore::Coordinates::toString):
        * page/Coordinates.h: Added.
        (WebCore::Coordinates::create):
        (WebCore::Coordinates::latitude):
        (WebCore::Coordinates::longitude):
        (WebCore::Coordinates::altitude):
        (WebCore::Coordinates::accuracy):
        (WebCore::Coordinates::altitudeAccuracy):
        (WebCore::Coordinates::heading):
        (WebCore::Coordinates::speed):
        (WebCore::Coordinates::Coordinates):
        * page/Coordinates.idl: Added.
        * page/Geoposition.cpp:
        (WebCore::Geoposition::toString):
        * page/Geoposition.h:
        (WebCore::Geoposition::create):
        (WebCore::Geoposition::coords):
        (WebCore::Geoposition::Geoposition):
        * page/Geoposition.idl:
        * page/PositionOptions.h:
        (WebCore::PositionOptions::create):
        (WebCore::PositionOptions::maximumAge):
        (WebCore::PositionOptions::setMaximumAge):
        (WebCore::PositionOptions::PositionOptions):
        * platform/mac/GeolocationServiceMac.mm:
        (-[WebCoreCoreLocationObserver locationManager:didUpdateToLocation:fromLocation:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41650 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Dimitri Glazkov <dglazkov@chromium.org>
dglazkov@chromium.org [Thu, 12 Mar 2009 21:10:54 +0000 (21:10 +0000)]
2009-03-12  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Darin Fisher.

        https://bugs.webkit.org/show_bug.cgi?id=24496
        Fix console logging of non-string values by coercing the argument to
        a string, for V8 bindings.

        Test: fast/js/console-non-string-values.html

        * bindings/v8/ScriptValue.cpp: Added toString method.
        (WebCore::ScriptValue::toString):
        * bindings/v8/ScriptValue.h: Added PlatformString, ScriptState includes
            and toString declaration.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41649 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Darin Fisher <darin@chromium.org>
darin@chromium.org [Thu, 12 Mar 2009 21:08:51 +0000 (21:08 +0000)]
2009-03-12  Darin Fisher  <darin@chromium.org>

        Reviewed by Dimitri Glazkov.

        Eliminate ChromiumBridge::uiResourceProtocol
        https://bugs.webkit.org/show_bug.cgi?id=24558

        * platform/chromium/ChromiumBridge.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41648 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Greg Bolsinga <bolsinga@apple.com>
bolsinga@apple.com [Thu, 12 Mar 2009 20:44:12 +0000 (20:44 +0000)]
2009-03-12  Greg Bolsinga  <bolsinga@apple.com>

        Reviewed by Simon Fraser.

        Missing breaks in switch statement
        https://bugs.webkit.org/show_bug.cgi?id=24556

        * platform/mac/GeolocationServiceMac.mm:
        (-[WebCoreCoreLocationObserver locationManager:didFailWithError:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41647 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Darin Fisher <darin@chromium.org>
darin@chromium.org [Thu, 12 Mar 2009 20:39:01 +0000 (20:39 +0000)]
2009-03-12  Darin Fisher  <darin@chromium.org>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24554
        Remove some unused functions from ChromiumBridge.

        * platform/chromium/ChromiumBridge.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41646 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Darin Fisher and Justin Garcia.
eric@webkit.org [Thu, 12 Mar 2009 20:29:30 +0000 (20:29 +0000)]
    Reviewed by Darin Fisher and Justin Garcia.

        Safari crashes during drag and drop in Google presentations
        due to mutation event handlers removing DOM content during insertNode
        https://bugs.webkit.org/show_bug.cgi?id=22634

        Added a bunch of "null" checks to make sure nodes are still
        in the document before we operate on them.  This is an
        inelegant solution, but it's the best we have for now.

        Test: editing/selection/crash-on-drag-with-mutation-events.html

        * editing/CompositeEditCommand.cpp:
        (WebCore::CompositeEditCommand::insertNodeAt):
        * editing/ReplaceSelectionCommand.cpp:
        (WebCore::ReplaceSelectionCommand::doApply):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41645 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Thu, 12 Mar 2009 20:11:55 +0000 (20:11 +0000)]
2009-03-12  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Adele Peterson.

        https://bugs.webkit.org/show_bug.cgi?id=24508

        Fix updating of text field with placeholder text when value is set
        by ensuring that updatePlaceholderVisibility() is called after
        the value has been updated, not before.

        Test: fast/forms/placeholder-set-value.html

        * html/HTMLInputElement.cpp:
        (WebCore::HTMLInputElement::setValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41644 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Dimitri Glazkov <dglazkov@chromium.org>
dglazkov@chromium.org [Thu, 12 Mar 2009 19:52:54 +0000 (19:52 +0000)]
2009-03-12  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Adam Treat.

        https://bugs.webkit.org/show_bug.cgi?id=24525
        REGRESSION: Inspector window doesn't close when inspected page is
        destroyed. This is a revert of r41158, which became unnecessary when
        InspectorController became ref-counted in r41462.

        * inspector/InspectorController.cpp:
        (WebCore::InspectorController::inspectedPageDestroyed): Reset m_inspectedPage
            after calling close().
        (WebCore::InspectorController::stopUserInitiatedProfiling): Remove
            m_inspectedPage check guard around profile logic.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41643 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoWebCore:
pkasting@chromium.org [Thu, 12 Mar 2009 19:41:38 +0000 (19:41 +0000)]
WebCore:

2009-03-12  Peter Kasting  <pkasting@google.com>

        Reviewed by Darin Fisher.

        https://bugs.webkit.org/show_bug.cgi?id=24502
        Make horizontal scrolling on Windows always go the correct direction.

        * platform/PlatformWheelEvent.h:
        * platform/win/WheelEventWin.cpp:
        (WebCore::PlatformWheelEvent::PlatformWheelEvent):

WebKit/win:

2009-03-12  Peter Kasting  <pkasting@google.com>

        Reviewed by Darin Fisher.

        https://bugs.webkit.org/show_bug.cgi?id=24502
        Make horizontal scrolling on Windows always go the correct direction.

        * WebView.cpp:
        (WebView::mouseWheel):
        (WebViewWndProc):
        * WebView.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41642 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Thu, 12 Mar 2009 19:36:15 +0000 (19:36 +0000)]
2009-03-12  Anders Carlsson  <andersca@apple.com>

        Reviewed by Geoffrey Garen.

        WebKit side of <rdar://problem/6607801>

        * Plugins/Hosted/NetscapePluginInstanceProxy.mm:
        (WebKit::NetscapePluginInstanceProxy::destroy):
        Pass a requestID to _WKPCDestroyPluginInstance and wait until we get a reply back.

        * Plugins/Hosted/WebKitPluginHost.defs:
        Add requestID parameter.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41641 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoWebCore:
dglazkov@chromium.org [Thu, 12 Mar 2009 19:34:02 +0000 (19:34 +0000)]
WebCore:

2009-03-12  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Simon Fraser.

        https://bugs.webkit.org/show_bug.cgi?id=24496
        Fix console logging of non-string values by coercing the argument to
        a string.

        Test: fast/js/console-non-string-values.html

        * bindings/js/ScriptValue.cpp: Removed PlatformString include.
        * bindings/js/ScriptValue.h: Added toString method.
        (WebCore::ScriptValue::toString):
        * page/Console.cpp:
        (WebCore::getFirstArgumentAsString): Changed firstArgumentAsString method to use
            ScriptValue::toString.
        (WebCore::Console::addMessage): added extra ScriptState argument to callsite.
        (WebCore::Console::count): Ditto.

LayoutTests:

2009-03-12  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24496
        Test for logging non-string values.

        * fast/js/console-non-string-values-expected.txt: Added.
        * fast/js/console-non-string-values.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41640 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Thu, 12 Mar 2009 19:32:25 +0000 (19:32 +0000)]
2009-03-12  Gustavo Noronha Silva  <gns@gnome.org>

        Reviewed by Alexey  Proskuryakov.

        https://bugs.webkit.org/show_bug.cgi?id=24553
        [GTK] Improvements to WebKitDownload

        Rename 'state' to 'status' to match the naming in the frame
        loader rework that we plan to land soonish, and make it a
        property, for the same reason.

        * webkit/webkitdownload.cpp:
        (_WebKitDownloadPrivate::):
        (_WebKitDownloadPrivate::webkit_download_finalize):
        (_WebKitDownloadPrivate::webkit_download_get_property):
        (_WebKitDownloadPrivate::webkit_download_set_property):
        (_WebKitDownloadPrivate::webkit_download_class_init):
        (_WebKitDownloadPrivate::webkit_download_init):
        (_WebKitDownloadPrivate::webkit_download_start):
        (_WebKitDownloadPrivate::webkit_download_cancel):
        (_WebKitDownloadPrivate::webkit_download_set_destination_uri):
        (_WebKitDownloadPrivate::webkit_download_get_status):
        (_WebKitDownloadPrivate::webkit_download_set_status):
        (_WebKitDownloadPrivate::webkit_download_received_data):
        (_WebKitDownloadPrivate::webkit_download_finished_loading):
        (_WebKitDownloadPrivate::webkit_download_error):
        * webkit/webkitdownload.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41639 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Thu, 12 Mar 2009 19:32:06 +0000 (19:32 +0000)]
2009-03-12  Gustavo Noronha Silva  <gns@gnome.org>

        Reviewed by Alexey  Proskuryakov.

        https://bugs.webkit.org/show_bug.cgi?id=24553
        [GTK] Improvements to WebKitDownload

        Improved documentation for the download-requested signal, to make
        its usage clear.

        * webkit/webkitwebview.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41638 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Thu, 12 Mar 2009 19:27:41 +0000 (19:27 +0000)]
2009-03-12  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dave Hyatt

        https://bugs.webkit.org/show_bug.cgi?id=24552

        Hit testing involving normal flow, self-painting layers (e.g. reflections)
        was broken because it was not null-testing the result of hitTestLayer(),
        so bailing early with a nil hit layer.

        Test: fast/layers/normal-flow-hit-test.html

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::hitTestLayer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41637 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Pamela Greene <pam@chromium.org>
pam@chromium.org [Thu, 12 Mar 2009 18:33:07 +0000 (18:33 +0000)]
2009-03-12  Pamela Greene  <pam@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Add test cases to make sure window.location and window.navigator
        have higher lookup precedence than a DOM element with that ID.

        https://bugs.webkit.org/show_bug.cgi?id=24522

        * fast/dom/Window/window-special-properties-expected.txt:
        * fast/dom/Window/window-special-properties.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41636 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Anders Carlsson.
kdecker@apple.com [Thu, 12 Mar 2009 18:30:11 +0000 (18:30 +0000)]
    Reviewed by Anders Carlsson.

        * Plugins/WebNetscapePluginPackage.mm:
        (-[WebNetscapePluginPackage _unloadWithShutdown:]): Simply a small SUPPORT_CFM code block.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41635 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Anders Carlsson.
kdecker@apple.com [Thu, 12 Mar 2009 18:20:07 +0000 (18:20 +0000)]
    Reviewed by Anders Carlsson.

        Fixed: <rdar://problem/5815862> Opening a subclassed NSWindow from a Safari plugin causes Safari to crash on Quit

        This fix addresses crashes in both Silverlight and ChemDraw.  This type of crash would occur because AppKit still
        had a reference to open windows that the plugin created (which no longer exist).

        * Plugins/WebNetscapePluginPackage.mm:
        (-[WebNetscapePluginPackage _unloadWithShutdown:]): Do not unload the plug-in bundle on browser shutdown.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41634 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Dimitri Glazkov <dglazkov@chromium.org>
dglazkov@chromium.org [Thu, 12 Mar 2009 18:08:08 +0000 (18:08 +0000)]
2009-03-12  Dimitri Glazkov  <dglazkov@chromium.org>

        Reviewed by Cameron Zwarich.

        Removed zero-sized files, left over from improperly applied patch.

        * platform/graphics/chromium/ThemeHelperChromiumWin.cpp: Removed.
        * platform/graphics/chromium/ThemeHelperChromiumWin.h: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41633 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Adam Treat <adam.treat@torchmobile.com>
treat@webkit.org [Thu, 12 Mar 2009 18:04:05 +0000 (18:04 +0000)]
2009-03-12  Adam Treat  <adam.treat@torchmobile.com>

        Reviewed by NOBODY (Build fix).

        Include string.h for strlen usage.

        * wtf/Threading.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41632 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Pamela Greene <pam@chromium.org>
pam@chromium.org [Thu, 12 Mar 2009 17:24:55 +0000 (17:24 +0000)]
2009-03-12  Pamela Greene  <pam@chromium.org>

        Reviewed by Darin Adler.

        Add a test to make sure GC doesn't drop XHR wrappers too soon.
        https://bugs.webkit.org/show_bug.cgi?id=24528

        * fast/dom/xmlhttprequest-gc-expected.txt: Added.
        * fast/dom/xmlhttprequest-gc.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41631 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Adam Treat <adam.treat@torchmobile.com>
treat@webkit.org [Thu, 12 Mar 2009 16:35:29 +0000 (16:35 +0000)]
2009-03-12  Adam Treat  <adam.treat@torchmobile.com>

        Reviewed by George Staikos.

        Do not ignore alpha color in optimization.  Hard to believe, but there
        are websites that fill the entire page with a tiled image consisting of
        nothing but a 100% transparent 1x1 image. <cough>orbitz.com</cough>

        * platform/graphics/qt/ImageQt.cpp:
        (WebCore::BitmapImage::checkForSolidColor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41630 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Implement DOMHTMLDocument::title
aroben@apple.com [Thu, 12 Mar 2009 16:22:30 +0000 (16:22 +0000)]
    Implement DOMHTMLDocument::title

        Reviewed by Dan Bernstein.

        * DOMHTMLClasses.cpp:
        (DOMHTMLDocument::title): Call through to WebCore::Document (with
        obligatory COM error checking/type conversions).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41629 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoFixed ChangeLog date.
ddkilzer@apple.com [Thu, 12 Mar 2009 15:28:04 +0000 (15:28 +0000)]
Fixed ChangeLog date.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41628 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Add NO_RETURN attribute to runInteractive() when not using readline
ddkilzer@apple.com [Thu, 12 Mar 2009 15:26:20 +0000 (15:26 +0000)]
    Add NO_RETURN attribute to runInteractive() when not using readline

        Reviewed by Darin Adler.

        * jsc.cpp:
        (runInteractive): If the readline library is not used, this method
        will never return, thus the NO_RETURN attribute is needed to prevent
        a gcc warning.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41627 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Adopt setThreadNameInternal on Windows
aroben@apple.com [Thu, 12 Mar 2009 15:01:46 +0000 (15:01 +0000)]
    Adopt setThreadNameInternal on Windows

        Also changed a Windows-only assertion about thread name length to an
        all-platform log message.

        Reviewed by Adam Treat.

        * wtf/Threading.cpp:
        (WTF::createThread): Warn if the thread name is longer than 31
        characters, as Visual Studio will truncate names longer than that
        length.

        * wtf/ThreadingWin.cpp:
        (WTF::setThreadNameInternal): Renamed from setThreadName and changed
        to always operate on the current thread.
        (WTF::initializeThreading): Changed to use setThreadNameInternal.
        (WTF::createThreadInternal): Removed call to setThreadName. This is
        now handled by threadEntryPoint and setThreadNameInternal.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41626 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Thu, 12 Mar 2009 15:00:25 +0000 (15:00 +0000)]
2009-03-12  Gustavo Noronha Silva  <gns@gnome.org>

        Reviewed by Alp Toker.

        https://bugs.webkit.org/show_bug.cgi?id=24541
        Scrolling with home and end keys not always works

        Also make page up and page down keys be handled by the webview key
        event code, so that they also work in cases where the GTK+
        scrollbars don't handle them directly, like in the bugzill's patch
        review page.

        * webkit/webkitwebview.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41625 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Thu, 12 Mar 2009 15:00:11 +0000 (15:00 +0000)]
2009-03-12  Gustavo Noronha Silva  <gns@gnome.org>

        Reviewed by Alp Toker.

        https://bugs.webkit.org/show_bug.cgi?id=24541
        Scrolling with home and end keys not always works

        Make home and end keys behave more consistently for scrolling the
        view.

        * webkit/webkitwebview.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41624 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Bug 24378: resolve-ChangeLogs should use git status or svn status to find...
ddkilzer@apple.com [Thu, 12 Mar 2009 14:46:05 +0000 (14:46 +0000)]
    Bug 24378: resolve-ChangeLogs should use git status or svn status to find and fix unmerged ChangeLogs

        <https://bugs.webkit.org/show_bug.cgi?id=24378>

        Reviewed by Adam Roben.

        * Scripts/resolve-ChangeLogs: If -f|--fix-merged is not passed
        and no file or directory names are specified on the command-line
        then try to find unmerged ChangeLog files based on 'svn stat' or
        'git diff'.  Added global $isGit and $isSVN variables so that
        isGit() and isSVN() only have to be called once.
        (findUnmergedChangeLogs): Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41623 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Thu, 12 Mar 2009 13:18:42 +0000 (13:18 +0000)]
2009-03-12  Gustavo Noronha Silva  <gns@gnome.org>

        Reviewed by Alp Toker.

        https://bugs.webkit.org/show_bug.cgi?id=24254
        [GTK] spacebar doesn't scroll down

        Make spacebar and shift+spacebar scroll like page up and down
        respectively would.

        * webkit/webkitwebview.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41622 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Julien Chaffraix <jchaffraix@webkit.org>
jchaffraix@webkit.org [Thu, 12 Mar 2009 12:42:09 +0000 (12:42 +0000)]
2009-03-12  Julien Chaffraix  <jchaffraix@webkit.org>

        Reviewed by Darin Adler.

        Bug 24110: cloneNode should call cloneElement and not the reverse

        - Splitted the code from cloneNode into cloneElementWithChildren and cloneElementWithChildren.
          Now cloneNode calls one of the 2 previous methods.

        - Renamed cloneElement to cloneElementWithoutChildren as it was the previous behaviour.

        - Moved cloneNode to the Element private section so that WebCore callers cannot use it.

        - Removed Element::cloneNode usage through WebCore.

        * dom/Element.cpp:
        (WebCore::Element::cloneNode): Moved to Element's private section and it
        now calls the two next methods.
        (WebCore::Element::cloneElementWithChildren): Added.
        (WebCore::Element::cloneElementWithoutChildren): Renamed from cloneElement
        to avoid ambiguity.
        * dom/Element.h:

        * editing/ApplyStyleCommand.cpp:
        (WebCore::ApplyStyleCommand::addInlineStyleIfNeeded): Changed call to cloneElement
        to call to cloneElementWithoutChildren.
        * editing/BreakBlockquoteCommand.cpp:
        (WebCore::BreakBlockquoteCommand::doApply): Ditto.
        * editing/IndentOutdentCommand.cpp:
        (WebCore::IndentOutdentCommand::indentRegion): Ditto.
        * editing/InsertParagraphSeparatorCommand.cpp:
        (WebCore::InsertParagraphSeparatorCommand::doApply): Ditto.
        * editing/ModifySelectionListLevel.cpp:
        (WebCore::IncreaseSelectionListLevelCommand::doApply): Ditto.
        * editing/SplitElementCommand.cpp:
        (WebCore::SplitElementCommand::doApply): Ditto.
        * editing/markup.cpp:
        (WebCore::createFragmentFromText): Ditto.
        * svg/SVGUseElement.cpp:
        (WebCore::SVGUseElement::buildShadowTree): Ditto.
        (WebCore::SVGUseElement::expandUseElementsInShadowTree): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41621 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Dirk Schulze <krit@webkit.org>
krit@webkit.org [Thu, 12 Mar 2009 11:08:22 +0000 (11:08 +0000)]
2009-03-12  Dirk Schulze  <krit@webkit.org>

        Reviewed by Oliver Hunt.

        Fixed a crash on the 1x1 pixel check for background image in gtk.
        We need to load the image first and need to be sure that the image
        is not null before checking the type.

        * platform/graphics/cairo/ImageCairo.cpp:
        (WebCore::BitmapImage::checkForSolidColor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41620 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Thu, 12 Mar 2009 08:46:07 +0000 (08:46 +0000)]
2009-03-12  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Alexey Proskuryakov.

        https://bugs.webkit.org/show_bug.cgi?id=24544
        Rename setUrl to setURL

        Use proper case for setUrl method, rename it to setURL.

        * platform/network/ResourceResponseBase.cpp:
        (WebCore::ResourceResponseBase::adopt):
        * platform/network/ResourceResponseBase.h:
        * platform/network/curl/ResourceHandleManager.cpp:
        (WebCore::writeCallback):
        (WebCore::headerCallback):
        * platform/network/soup/ResourceHandleSoup.cpp:
        (WebCore::fillResponseFromMessage):
        (WebCore::queryInfoCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41619 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-12 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Thu, 12 Mar 2009 07:58:03 +0000 (07:58 +0000)]
2009-03-12  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Alexey Proskuryakov.

        https://bugs.webkit.org/show_bug.cgi?id=24519
        [GTK] Use two argument KURL ctor in ResourceHandleSoup

        Use two argument KURL ctor, the single argument ctor expects its
        input to be the output of a previous KURL::parse call, which is
        not the case here.

        * platform/network/soup/ResourceHandleSoup.cpp:
        (WebCore::fillResponseFromMessage):
        (WebCore::queryInfoCallback):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41618 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-11 Jon Honeycutt <jhoneycutt@apple.com>
jhoneycutt@apple.com [Thu, 12 Mar 2009 07:41:33 +0000 (07:41 +0000)]
2009-03-11  Jon Honeycutt  <jhoneycutt@apple.com>

        Fix for <rdar://6418681>
        https://bugs.webkit.org/show_bug.cgi?id=22644

        Reviewed by Steve Falkenburg.

        * plugins/win/PluginPackageWin.cpp:
        (WebCore::PluginPackage::isPluginBlacklisted): Add the Citrix ICA
        Client plug-in to the blacklist; it requires a Mozilla-based browser.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41617 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoBuild fix
oliver@apple.com [Thu, 12 Mar 2009 07:26:43 +0000 (07:26 +0000)]
Build fix

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41616 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago<rdar://problem/6673852> Widget "Simple World Clock" does not display with Safari 4
oliver@apple.com [Thu, 12 Mar 2009 06:39:38 +0000 (06:39 +0000)]
<rdar://problem/6673852> Widget "Simple World Clock" does not display with Safari 4

Reviewed by Adele Peterson.

A number of dashboard widgets rely on canvas failing silently when given invalid
colors for gradient stops, but both the spec and firefox throw exceptions here.
So we work around this by creating a dashboard quirk that will only suppress the
exception in dashboard compatibility mode.

Test: fast/canvas/canvas-gradient-addStop-error.html

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41615 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Remove duplicate header include
ddkilzer@apple.com [Thu, 12 Mar 2009 06:02:10 +0000 (06:02 +0000)]
    Remove duplicate header include

        Rubber-stamped by Mark Rowe.

        * WebView/WebView.mm: Remove duplicate #include <runtime/InitializeThreading.h>.
        Also realphabetized lowercase #include statements.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41614 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Clarify comments regarding order of FEATURE_DEFINES
ddkilzer@apple.com [Thu, 12 Mar 2009 05:14:55 +0000 (05:14 +0000)]
    Clarify comments regarding order of FEATURE_DEFINES

        Rubber-stamped by Mark Rowe.

JavaScriptCore:

        * Configurations/JavaScriptCore.xcconfig: Added warning about
        the consequences when FEATURE_DEFINES are not kept in sync.

WebCore:

        * Configurations/WebCore.xcconfig: Added warning about the
        consequences when FEATURE_DEFINES are not kept in sync.

WebKit/mac:

        * Configurations/WebKit.xcconfig: Added warning about the
        consequences when FEATURE_DEFINES are not kept in sync.

WebKitTools:

        * Scripts/build-webkit: Added warning about keeping
        FEATURE_DEFINES in order and the consequences when they are not.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41613 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-11 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Thu, 12 Mar 2009 04:27:42 +0000 (04:27 +0000)]
2009-03-11  Gustavo Noronha Silva  <gns@gnome.org>

        Reviewed by Holger Freyther.

        https://bugs.webkit.org/show_bug.cgi?id=24515
        [GTK] soup backend should use GMappedFile instead of mmap directly

        Rewrite mmap code used to upload files to use GMappedFile, so that
        it is platform-independent.

        * platform/network/soup/ResourceHandleSoup.cpp:
        (WebCore::ResourceHandle::startHttp):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41610 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoVersioning.
mrowe@apple.com [Thu, 12 Mar 2009 04:06:24 +0000 (04:06 +0000)]
Versioning.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41609 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoJavaScriptCore:
mitz@apple.com [Thu, 12 Mar 2009 03:39:14 +0000 (03:39 +0000)]
JavaScriptCore:

        Reviewed by Darin Adler.

        - WTF support for fixing <rdar://problem/3919124> Thai text selection
          in Safari is incorrect

        * wtf/unicode/icu/UnicodeIcu.h:
        (WTF::Unicode::hasLineBreakingPropertyComplexContext): Added. Returns
        whether the character has Unicode line breaking property value SA
        ("Complex Context").
        * wtf/unicode/qt4/UnicodeQt4.h:
        (WTF::Unicode::hasLineBreakingPropertyComplexContext): Added an
        implementation that always returns false.

WebCore:

        Reviewed by Darin Adler.

        - WebCore part of fixing <rdar://problem/3919124> Thai text selection
          in Safari is incorrect

        Test: platform/mac/editing/selection/word-thai.html

        Tested on Mac, but should work on all ICU platforms (or more precisely
        on all platforms that implement
        WTF::Unicode::hasLineBreakingPropertyComplexContext() correctly).

        * editing/TextIterator.cpp:
        (WebCore::BackwardsCharacterIterator::BackwardsCharacterIterator):
        Added.
        (WebCore::BackwardsCharacterIterator::range): Added.
        (WebCore::BackwardsCharacterIterator::advance): Added.

        * editing/TextIterator.h: Added BackwardsCharacterIterator with the
        minimal functionality required for this patch.
        (WebCore::BackwardsCharacterIterator::atEnd): Added.

        * editing/visible_units.cpp:
        (WebCore::firstNonComplexContextLineBreak): Added this helper function
        that returns the index of the first character in the string whose
        Unicode line breaking property value is not SA ("Complex Context"), or
        the length of the string if there is no such character.
        (WebCore::lastNonComplexContextLineBreak): Added this helper function
        that returns the index of the last character in the string whose
        Unicode line breaking property value is not SA ("Complex Context"), or
        -1 if there is no such character.
        (WebCore::previousBoundary): Changed the signature of the search
        function to include an offset parameter. Renamed the 'exception' local
        variable to 'ec' and changed its type to ExceptionCode. Extend the
        string forwards until the first character with Unicode line breaking
        property value other than SA. This gives the boundary search function
        enough context in the forward direction. Changed to use a
        BackwardsCharacterIterator for translating the backwards offset into
        a position.
        (WebCore::nextBoundary): Changed the signature of the search
        function to include an offset parameter. Extend the string backwards
        until the first character with Unicode line breaking property value
        other than SA. This gives the boundary search function enough context in
        the backwards direction. Restricted the workaround for <rdar://5192593>
        only to the line break case, because it was causing an extra character
        to be selected following a word that ended with a combining mark.
        (WebCore::startWordBoundary): Added an offset parameter. Check that
        going backwards from the given offset, there is a character with Unicode
        line breaking property value other than SA, and otherwise return 0 to
        request more context.
        (WebCore::endWordBoundary): Similar, but in reverse.
        (WebCore::previousWordPositionBoundary): Similar.
        (WebCore::nextWordPositionBoundary): Similar.
        (WebCore::startSentenceBoundary): Updated for the new search function
        signature.
        (WebCore::endSentenceBoundary): Ditto.
        (WebCore::previousSentencePositionBoundary): Ditto.
        (WebCore::nextSentencePositionBoundary): Ditto.

LayoutTests:

        Reviewed by Darin Adler.

        - test for <rdar://problem/3919124> Thai text selection in Safari is
          incorrect

        * platform/mac/editing/selection/word-thai-expected.txt: Added.
        * platform/mac/editing/selection/word-thai.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41607 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-11 Darin Adler <darin@apple.com>
darin@apple.com [Thu, 12 Mar 2009 01:29:07 +0000 (01:29 +0000)]
2009-03-11  Darin Adler  <darin@apple.com>

        Reviewed by Mark Rowe.

        Don't use "::" in thread names because it is ugly!

        * loader/icon/IconDatabase.cpp:
        (WebCore::IconDatabase::open): Use "WebCore: " instead of "WebCore::".
        * platform/network/cf/ResourceHandleCFNet.cpp:
        (WebCore::ResourceHandle::loaderRunLoop): Call this "WebCore: CFNetwork Loader" to make
        it clear this is a WebCore-created thread.
        * storage/DatabaseThread.cpp:
        (WebCore::DatabaseThread::start): Use "WebCore: " instead of "WebCore::".
        * storage/LocalStorageThread.cpp:
        (WebCore::LocalStorageThread::start): Use "WebCore: " instead of "WebCore::".
        * workers/WorkerThread.cpp:
        (WebCore::WorkerThread::start): Use "WebCore: " instead of "WebCore::".

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41606 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-11 Darin Adler <darin@apple.com>
darin@apple.com [Thu, 12 Mar 2009 01:26:55 +0000 (01:26 +0000)]
2009-03-11  Darin Adler  <darin@apple.com>

        Reviewed by Mark Rowe.

        Give threads names on platforms with pthread_setname_np.

        * wtf/Threading.cpp:
        (WTF::NewThreadContext::NewThreadContext): Initialize thread name.
        (WTF::threadEntryPoint): Call setThreadNameInternal.
        (WTF::createThread): Pass thread name.

        * wtf/Threading.h: Added new comments, setThreadNameInternal.

        * wtf/ThreadingGtk.cpp:
        (WTF::setThreadNameInternal): Added. Empty.
        * wtf/ThreadingNone.cpp:
        (WTF::setThreadNameInternal): Added. Empty.
        * wtf/ThreadingPthreads.cpp:
        (WTF::setThreadNameInternal): Call pthread_setname_np when available.
        * wtf/ThreadingQt.cpp:
        (WTF::setThreadNameInternal): Added. Empty.
        * wtf/ThreadingWin.cpp:
        (WTF::setThreadNameInternal): Added. Empty.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41605 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago<rdar://problem/6653286> WebCore's implementation of screenIsMonochrome is needlessly...
mrowe@apple.com [Thu, 12 Mar 2009 00:53:21 +0000 (00:53 +0000)]
<rdar://problem/6653286> WebCore's implementation of screenIsMonochrome is needlessly complex

Rubber-stamped by Darin Adler.

* platform/mac/PlatformScreenMac.mm:
(WebCore::screenIsMonochrome): Return false.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41604 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-11 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Thu, 12 Mar 2009 00:51:45 +0000 (00:51 +0000)]
2009-03-11  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Darin Adler

        https://bugs.webkit.org/show_bug.cgi?id=24531

        Make sure that AnimationControllerPrivate::getAnimatedStyleForRenderer() never
        returns a null style if it has a valid renderer, which could happen if a
        CompositeAnimation existed, but wasn't running any animations or transitions.

        * page/animation/AnimationController.cpp:
        (WebCore::AnimationControllerPrivate::getAnimatedStyleForRenderer):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41603 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoWebCore:
bdakin@apple.com [Wed, 11 Mar 2009 23:25:32 +0000 (23:25 +0000)]
WebCore:

2009-03-11  Beth Dakin  <bdakin@apple.com>

        Reviewed by Dan Bernstein.

        Fix for https://bugs.webkit.org/show_bug.cgi?id=14749 percentage
        top value on position:relative child not calculated correctly if
        parent has percentage height
        - and corresponding -
        <rdar://problem/6172925>

        * rendering/RenderBoxModelObject.cpp:
        (WebCore::RenderBoxModelObject::relativePositionOffsetY):

LayoutTests:

2009-03-11  Beth Dakin  <bdakin@apple.com>

        Reviewed by Dan Bernstein.

        Test for https://bugs.webkit.org/show_bug.cgi?id=14749 percentage
        top value on position:relative child not calculated correctly if
        parent has percentage height
        - and corresponding -
        <rdar://problem/6172925>

        * fast/css/percent-top-value-with-relative-position.html: Added.
        * platform/mac/fast/css/percent-top-value-with-relative-position-expected.checksum: Added.
        * platform/mac/fast/css/percent-top-value-with-relative-position-expected.png: Added.
        * platform/mac/fast/css/percent-top-value-with-relative-position-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41602 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-11 Jian Li <jianli@chromium.org>
levin@chromium.org [Wed, 11 Mar 2009 21:54:23 +0000 (21:54 +0000)]
2009-03-11  Jian Li  <jianli@chromium.org>

        Reviewed by Dimitri Glazkov.

        Implement ScriptValue::hasNoValue for V8.
        https://bugs.webkit.org/show_bug.cgi?id=24507

        * bindings/v8/ScriptValue.h:
        (WebCore::ScriptValue::hasNoValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41601 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-11 David Levin <levin@chromium.org>
levin@chromium.org [Wed, 11 Mar 2009 21:54:17 +0000 (21:54 +0000)]
2009-03-11  David Levin  <levin@chromium.org>

        Reviewed by Dimitri Glazkov.

        Bug 24459: Add v8 bindings for event.
        <https://bugs.webkit.org/show_bug.cgi?id=24459>

        Adding custom event bindings.

        * bindings/v8/custom/V8AbstractEventListener.cpp: Added.
        * bindings/v8/custom/V8AbstractEventListener.h: Added.
        * bindings/v8/custom/V8AttrCustom.cpp:
        Fixed messed up copyright header.
        * bindings/v8/custom/V8CustomEventListener.cpp: Added.
        * bindings/v8/custom/V8CustomEventListener.h:
        Replaced forwarding header with the real implementation.
        * bindings/v8/custom/V8ElementCustom.cpp:
        * bindings/v8/custom/V8LazyEventListener.cpp: Added.
        * bindings/v8/custom/V8LazyEventListener.h: Added.
        * bindings/v8/custom/V8ObjectEventListener.cpp: Added.
        * bindings/v8/custom/V8ObjectEventListener.h: Added.
        * bindings/v8/custom/V8WorkerContextEventListener.cpp: Added.
        * bindings/v8/custom/V8WorkerContextEventListener.h: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41600 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-11 Darin Fisher <darin@chromium.org>
darin@chromium.org [Wed, 11 Mar 2009 21:40:49 +0000 (21:40 +0000)]
2009-03-11  Darin Fisher  <darin@chromium.org>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=24521
        Add some missing includes for the non-JSC build.

        * platform/text/PlatformString.h: added OwnPtr.h
        * xml/XMLHttpRequest.cpp: added SecurityOrigin.h

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41599 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-11 Pamela Greene <pam@chromium.org>
pam@chromium.org [Wed, 11 Mar 2009 21:17:03 +0000 (21:17 +0000)]
2009-03-11  Pamela Greene  <pam@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Add test for potential crash in window.frames.length
        These files were accidentally left out of r41459.

        https://bugs.webkit.org/show_bug.cgi?id=24339

        * fast/dom/Window/window-collection-length-no-crash-expected.txt: Added.
        * fast/dom/Window/window-collection-length-no-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41598 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-11 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Wed, 11 Mar 2009 20:58:54 +0000 (20:58 +0000)]
2009-03-11  Anders Carlsson  <andersca@apple.com>

        Reviewed by Sam Weinig.

        WebKit side of <rdar://problem/6656147>.

        * Plugins/Hosted/NetscapePluginHostManager.mm:
        (WebKit::NetscapePluginHostManager::instantiatePlugin):
        Pass the requestID to _WKPHInstantiatePlugin.

        * Plugins/Hosted/NetscapePluginHostProxy.mm:
        Pass the requestID to setCurrentReply.

        * Plugins/Hosted/NetscapePluginInstanceProxy.h:
        (WebKit::NetscapePluginInstanceProxy::setCurrentReply):
        Store the reply in a map with the requestID as the key.

        (WebKit::NetscapePluginInstanceProxy::waitForReply):
        Wait for a reply that matches the given requestID.

        * Plugins/Hosted/NetscapePluginInstanceProxy.mm:
        (WebKit::NetscapePluginInstanceProxy::NetscapePluginInstanceProxy):
        Initialize member variables.

        (WebKit::NetscapePluginInstanceProxy::~NetscapePluginInstanceProxy):
        Delete all requests.

        (WebKit::NetscapePluginInstanceProxy::print):
        Pass the requestID to _WKPHPluginInstancePrint.

        (WebKit::NetscapePluginInstanceProxy::loadRequest):
        Rename m_currentRequestID to m_currentURLRequestID.

        (WebKit::NetscapePluginInstanceProxy::processRequestsAndWaitForReply):
        Process requests until we find a reply with the right requestID.

        (WebKit::NetscapePluginInstanceProxy::createBindingsInstance):
        Pass a requestID to the _WKPH function.

        (WebKit::NetscapePluginInstanceProxy::nextRequestID):
        Ditto.

        * Plugins/Hosted/ProxyInstance.mm:
        Pass a requestID to the _WKPH functions.

        * Plugins/Hosted/WebKitPluginClient.defs:
        * Plugins/Hosted/WebKitPluginHost.defs:
        Add requestID parameters.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41597 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-03-11 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Wed, 11 Mar 2009 20:41:29 +0000 (20:41 +0000)]
2009-03-11  Simon Fraser  <simon.fraser@apple.com>

        Fix Scons build. Patch from Alexander Guy.

        * WebCore.scons:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41596 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Export functions needed by ThreadSpecific from WebKit
aroben@apple.com [Wed, 11 Mar 2009 20:39:34 +0000 (20:39 +0000)]
    Export functions needed by ThreadSpecific from WebKit

        Reviewed by John Sullivan.

        * WebKit.vcproj/WebKit.def:
        * WebKit.vcproj/WebKit_debug.def:
        Export WTF::tlsKeyCount and WTF::tlsKeys.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41595 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Change the Windows implementation of ThreadSpecific to use functions instead...
aroben@apple.com [Wed, 11 Mar 2009 20:39:21 +0000 (20:39 +0000)]
    Change the Windows implementation of ThreadSpecific to use functions instead of extern globals

        This will make it easier to export ThreadSpecific from WebKit.

        Reviewed by John Sullivan.

        * API/JSBase.cpp:
        (JSEvaluateScript):
        Touched this file to force ThreadSpecific.h to be copied into
        $WebKitOutputDir.

        * wtf/ThreadSpecific.h: Replaced g_tls_key_count with tlsKeyCount()
        and g_tls_keys with tlsKeys().

        (WTF::::ThreadSpecific):
        (WTF::::~ThreadSpecific):
        (WTF::::get):
        (WTF::::set):
        (WTF::::destroy):
        Updated to use the new functions.

        * wtf/ThreadSpecificWin.cpp:
        (WTF::tlsKeyCount):
        (WTF::tlsKeys):
        Added.

        (WTF::ThreadSpecificThreadExit): Changed to use the new functions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@41594 268f45cc-cd09-0410-ab3c-d52691b4dbfc