11 years ago[Chromium] Implement file support for DragData.
jianli@chromium.org [Wed, 2 Sep 2009 21:30:08 +0000 (21:30 +0000)]
[Chromium] Implement file support for DragData.

Reviewed by Dimitri Glazkov.

* platform/chromium/DragDataChromium.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47994 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoReviewed by Kevin Ollivier.
kevino@webkit.org [Wed, 2 Sep 2009 21:07:36 +0000 (21:07 +0000)]
Reviewed by Kevin Ollivier.

Protect libcurl shared data with Mutex objects via curl_set_share_opt.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47993 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Evan Stade <estade@chromium.org>
eric@webkit.org [Wed, 2 Sep 2009 20:45:48 +0000 (20:45 +0000)]
2009-09-02  Evan Stade  <estade@chromium.org>

        Reviewed by Eric Seidel.

        Expose functions to change the focus ring color for Linux Chromium


        * rendering/RenderThemeChromiumLinux.cpp:
        * rendering/RenderThemeChromiumLinux.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47992 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Dirk Schulze <krit@webkit.org>
eric@webkit.org [Wed, 2 Sep 2009 20:38:52 +0000 (20:38 +0000)]
2009-09-02  Dirk Schulze  <krit@webkit.org>

        Reviewed by Eric Seidel.

        SVG Filter premultiplied color support for getImageDate/putImageData

        Patch to get premultiplied color support for Skia on getImageDate/putImageData.

        * platform/graphics/skia/ImageBufferSkia.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47991 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoForget to rename openInFrameloader() to open(), which was requested in Darin's review...
beidson@apple.com [Wed, 2 Sep 2009 20:23:49 +0000 (20:23 +0000)]
Forget to rename openInFrameloader() to open(), which was requested in Darin's review of my last patch.

* history/CachedFrame.cpp:
* history/CachedFrame.h:
* history/CachedPage.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47990 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoIn http://trac.webkit.org/changeset/47943 I added new header dependencies to FrameLoa...
beidson@apple.com [Wed, 2 Sep 2009 20:16:38 +0000 (20:16 +0000)]
In trac.webkit.org/changeset/47943 I added new header dependencies to FrameLoader.h.
Removing them is a little trouble, but worth it going forward.
The pattern of creating a "CachedFrameBase" as the interface CachedFrame provides to FrameLoader
was suggested by Darin Adler.

Reviewed by Darin Adler.

No new tests. (No change in behavior)

Make CachedFrame inherit privately from CachedFrameBase, which becomes the interface FrameLoader uses:
* history/CachedFrame.cpp:
* history/CachedFrame.h:

Call the new ::openInFrameLoader() method on the CachedFrame, as CachedFrameBase is now the only
thing capable of interacting with FrameLoader::open(CachedFrameBase):
* history/CachedPage.cpp:
* history/CachedPage.h:

Make ::open(CachedFrame) public, and change it to ::open(CachedFrameBase):
* loader/FrameLoader.cpp:
* loader/FrameLoader.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47989 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Zan Dobersek <zandobersek@gmail.com>
ddkilzer@apple.com [Wed, 2 Sep 2009 20:13:52 +0000 (20:13 +0000)]
2009-09-02  Zan Dobersek  <zandobersek@gmail.com>

        Reviewed by David Kilzer.

        Calls exitStatus function from the main package where it is also defined.

        * Scripts/VCSUtils.pm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47988 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore: Cannot pres Enter to escape from bulleted list when <ul> is a child of ...
rniwa@webkit.org [Wed, 2 Sep 2009 20:13:27 +0000 (20:13 +0000)]
WebCore: Cannot pres Enter to escape from bulleted list when <ul> is a child of <li>.

Patch by Ryosuke Niwa <rniwa@webkit.org> on 2009-09-02
Reviewed by Eric Seidel.

This patch modifies breakOutOfEmptyListItem to break out of a properly nested list
(a list to break out of is inside another list item).

When the empty list item appears at the end of another list item, WebKit breaks out of
the outer list item, and adds new item below the outer list item.
Otherwise, WebKit breaks out of the the inner list item and inserts new paragraph.

Test: editing/execCommand/break-out-of-empty-list-item.html

* editing/CompositeEditCommand.cpp:

LayoutTests: Cannot press Enter to escape from bulleted list when <ul> is a child of <li>.

Patch by Ryosuke Niwa <rniwa@webkit.org> on 2009-09-02
Reviewed by Eric Seidel.

This patch adds a test to make sure WebKit break out of an empty list item.

* editing/execCommand/break-out-of-empty-list-item-expected.txt: Added.
* editing/execCommand/break-out-of-empty-list-item.html: Added.
* editing/execCommand/resources/break-out-of-empty-list-item.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47987 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Brady Eidson.
ap@apple.com [Wed, 2 Sep 2009 19:55:00 +0000 (19:55 +0000)]
    Reviewed by Brady Eidson.

        Fix the test to pass on Tiger.

        Tiger's ICU decodes the invalid text differently. Moved it inside the comment - this also
        matches the original from the Web better.

        * fast/encoding/japanese-encoding-mix-expected.txt:
        * fast/encoding/japanese-encoding-mix.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47986 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoMore partial work towards "Page Cache should support pages with Frames"
beidson@apple.com [Wed, 2 Sep 2009 19:32:33 +0000 (19:32 +0000)]
More partial work towards "Page Cache should support pages with Frames"

Reviewed by Sam Weinig.

This patch is primarily about the "CachedFrame tree" and making slightly different decisions
based on whether a CachedFrame is the main frame or a subframe.

- Store whether a CachedFrame represents the main frame or a subframe.
- "Chop off the frame tree" for the main frame when creating its CachedFrame, as main frames
  are reused and need to start new page loads with an empty frame tree.
- Reattach the frame tree to the main frame when restoring it.
- open() the cached subframes in their respective loaders when restoring.
- Properly clear() and destroy() cached subframes.
- When committing to a new page load, and after caching the previous page, only clear the
  FrameView for the main frame.

Note that all of above will eventually be necessary as I continue to land this work in pieces,
but doesn't actually change any behavior right now because we still refuse to even *try* to
cache pages with frames.

No new tests. No change in behavior, and current tests continue to pass.

* history/CachedFrame.cpp:
* history/CachedFrame.h:

* loader/FrameLoader.cpp:
* loader/FrameLoader.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47985 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoNo review, build fix.
beidson@apple.com [Wed, 2 Sep 2009 18:32:36 +0000 (18:32 +0000)]
No review, build fix.

http://trac.webkit.org/changeset/47976 broke the build.
Let's try compiling before we land things, shall we?

* platform/graphics/SimpleFontData.cpp:
(WebCore::SimpleFontData::SimpleFontData): Order the initialization list correctly.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47984 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agowaf build fixes for Windows/MSVC and Mac/Snow Leopard.
kevino@webkit.org [Wed, 2 Sep 2009 18:28:30 +0000 (18:28 +0000)]
waf build fixes for Windows/MSVC and Mac/Snow Leopard.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47983 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoUnreviewed build fix for Chromium.
benm@google.com [Wed, 2 Sep 2009 18:14:16 +0000 (18:14 +0000)]
Unreviewed build fix for Chromium.

Missed #including V8Proxy.h in WebCore/bindings/v8/V8Binding.cpp.

* bindings/v8/V8Binding.cpp: add V8Proxy.h as an include

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47982 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoReviewed by Kevin Ollivier.
kevino@webkit.org [Wed, 2 Sep 2009 18:03:13 +0000 (18:03 +0000)]
Reviewed by Kevin Ollivier.

wx port: Call WTF::initializeThreading().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47981 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agowaf build fix, don't define symbols we now grab from other sources.
kevino@webkit.org [Wed, 2 Sep 2009 17:41:00 +0000 (17:41 +0000)]
waf build fix, don't define symbols we now grab from other sources.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47980 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoBuild fix for building on Windows.
kevino@webkit.org [Wed, 2 Sep 2009 17:39:55 +0000 (17:39 +0000)]
Build fix for building on Windows.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47979 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoBuild fixes for PluginViewNone.cpp and a wx build fix for PluginView.cpp.
kevino@webkit.org [Wed, 2 Sep 2009 17:34:49 +0000 (17:34 +0000)]
Build fixes for PluginViewNone.cpp and a wx build fix for PluginView.cpp.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47978 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agowx build fix. Use the 2 param KURL constructor.
kevino@webkit.org [Wed, 2 Sep 2009 17:32:35 +0000 (17:32 +0000)]
wx build fix. Use the 2 param KURL constructor.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47977 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Charles Wei <charles.wei@torchmobile.com.cn>
yong.li@torchmobile.com [Wed, 2 Sep 2009 17:10:24 +0000 (17:10 +0000)]
2009-09-02  Charles Wei  <charles.wei@torchmobile.com.cn>

        Reviewed by Eric Seidel.

        Fix the build failure of WebKit for Linux/Qt when WML is enabled

        No test cases needed since this just fixes the build problem.

        * wml/WMLElement.cpp:
        * wml/WMLTableElement.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47976 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Yong Li <yong.li@torchmobile.com>
yong.li@torchmobile.com [Wed, 2 Sep 2009 16:37:52 +0000 (16:37 +0000)]
2009-09-02 Yong Li <yong.li@torchmobile.com>

        Reviewed by Adam Barth.

        WINCE PORT: use shared buffer for custom font data

        Refactored by Joe Mason <joe.mason@torchmobile.com>

        * loader/CachedFont.cpp:
        add WINCE to platforms using cached custom data
        * platform/graphics/opentype/OpenTypeUtilities.cpp:
        (WebCore::renameFont): implement for WinCE
        * platform/graphics/opentype/OpenTypeUtilities.h:
        build fixes
        * platform/graphics/wince/FontCustomPlatformData.cpp:
        (WebCore::setCustomFontCache): add accessor
        (WebCore::createFontCustomPlatformData): change param to SharedBuffer
        * platform/graphics/wince/FontCustomPlatformData.h:
        update function signatures

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47975 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Yong Li <yong.li@torchmobile.com>
yong.li@torchmobile.com [Wed, 2 Sep 2009 16:32:55 +0000 (16:32 +0000)]
2009-09-02 Yong Li <yong.li@torchmobile.com>

        Reviewed by Adam Barth.

        WINCE PORT: font-related build fixes and minor bugfixes

        * platform/graphics/FontCache.h: declare functions added in https://bugs.webkit.org/show_bug.cgi?id=27509
        * platform/graphics/GlyphPageTreeNode.cpp:
        (WebCore::GlyphPageTreeNode::initializePage): fix ambiguous overload error in min/max on Windows
        * platform/graphics/SimpleFontData.cpp:
        (WebCore::SimpleFontData::SimpleFontData): initialize member vars
        * platform/graphics/SimpleFontData.h: disable unused members on Wince to save space

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47974 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Jeremy Orlow <jorlow@chromium.org>
abarth@webkit.org [Wed, 2 Sep 2009 15:54:01 +0000 (15:54 +0000)]
2009-09-02  Jeremy Orlow  <jorlow@chromium.org>

        Reviewed by Adam Barth.

        Another pass at releasing the storage mutex.

        Another pass at releasing the storage mutex (when applicable) when we finish
        executing JavaScript.  http://dev.w3.org/html5/spec/Overview.html#storage-mutex
        describes the required behavior.

        As it turns out, https://bugs.webkit.org/show_bug.cgi?id=28789 didn't solve
        the problem correctly.  First of all, I missed that events and timeouts go
        through a different call path.  Second of all, I didn't consider recursion
        correctly.  Third of all, my check to see if LocalStorage is alive actually
        instantiates it if it isn't yet.  Fourth, I forgot DOM_STORAGE guards around
        it.  So I'm reverting that change completely and doing it more cleanly.
        Unfortunately, the solution isn't as portable (and thus I've left out the JSC
        implementation for now).

        NOTE: This change now tracks recursion on function callbacks in V8.  I ran this
        by Mads Ager and Mike Belshe and they seemed to think adding the checks were
        fine.  Most callbacks are asynchronous and thus wouldn't be nested.  The few
        scenareos where you can have nested callbacks probably should have always been
        protected by a recursion guard.

        Unfortunately, this can't be tested without a hook added to the layout test
        controllers since the lock is (by design) not supposed to be observable by
        JavaScript.  https://bugs.webkit.org/show_bug.cgi?id=28906

        * bindings/js/ScriptController.cpp:
        (WebCore::ScriptController::evaluate): revert
        * bindings/v8/ScriptController.cpp:
        (WebCore::ScriptController::evaluate): revert
        * bindings/v8/V8Proxy.cpp:
            Add releaseStorageMutex call.  A small amount of clean up.
            Add releaseStorageMutex call.  A small amount of clean up.
            Check to see if localStorage has been instantiated.  If so, tell it to
            unlock all storage areas.
        * bindings/v8/V8Proxy.h:
        * page/PageGroup.h:
        (WebCore::PageGroup::hasLocalStorage):  Make hasLocalStorage public.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47973 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoReviewed by Eric Seidel.
kevino@webkit.org [Wed, 2 Sep 2009 15:48:53 +0000 (15:48 +0000)]
Reviewed by Eric Seidel.

Changes needed for build-webkit to support the waf build system for the wx port.


git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47972 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago <http://webkit.org/b/27168> With Subversion 1.6, update-webkit prompts on...
ddkilzer@apple.com [Wed, 2 Sep 2009 15:24:27 +0000 (15:24 +0000)]
    <webkit.org/b/27168> With Subversion 1.6, update-webkit prompts on conflicts

        Reviewed by Eric Seidel.

        * Scripts/update-webkit: Added "--accept postpone" to
        @svnOptions when running with svn-1.6 or newer.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47971 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Moved svn 1.6 version check into VCSUtils::isSVNVersion16OrNewer()
ddkilzer@apple.com [Wed, 2 Sep 2009 15:24:23 +0000 (15:24 +0000)]
    Moved svn 1.6 version check into VCSUtils::isSVNVersion16OrNewer()

        Reviewed by Eric Seidel.

        * Scripts/VCSUtils.pm:
        (@EXPORT): Added &isSVNVersion16OrNewer.
        (svnVersion): Added.  Internal method that gets the SVN version
        and caches it.
        (isSVNVersion16OrNewer): Added.  Method that does the SVN 1.6
        version check.
        * Scripts/prepare-ChangeLog: Switched to use new
        isSVNVersion16OrNewer() method.
        * Scripts/resolve-ChangeLogs: Ditto.
        * Scripts/svn-create-patch: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47970 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Clean up VCSUtils.pm
ddkilzer@apple.com [Wed, 2 Sep 2009 15:24:18 +0000 (15:24 +0000)]
    Clean up VCSUtils.pm

        Reviewed by Eric Seidel.

        * Scripts/VCSUtils.pm: Added proper package statement.  Fixed
        indentation of BEGIN block.  Listed each exported method on a
        line by itself.  Added methods to the export list after adding
        the package statement.  Sorted module variables.  Moved
        definiton of $gitRoot next to other module variables.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47969 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Yong Li <yong.li@torchmobile.com>
abarth@webkit.org [Wed, 2 Sep 2009 15:19:26 +0000 (15:19 +0000)]
2009-09-02  Yong Li  <yong.li@torchmobile.com>

        Reviewed by Adam Barth.

        WINCE PORT: Implement ImageBuffer for WINCE
        Implement Image interface with a light class BufferedImage
        for rendering ImageBuffer

        * platform/graphics/wince/ImageBufferData.h: Added.
        * platform/graphics/wince/ImageBufferWince.cpp: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47968 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoJoin the various PluginView destructors into one shared implementation
vestbo@webkit.org [Wed, 2 Sep 2009 14:24:31 +0000 (14:24 +0000)]
Join the various PluginView destructors into one shared implementation

Reviewed by Eric Seidel.


The Qt port used to delete the platform plugin widget as the last step.
Now this is done before cleaning up the script objects and unloading the
plugin package, similar to how the Win port does it.

* plugins/PluginView.cpp:
* plugins/PluginViewNone.cpp:
* plugins/gtk/PluginViewGtk.cpp:
* plugins/mac/PluginViewMac.cpp:
* plugins/qt/PluginViewQt.cpp:
* plugins/win/PluginViewWin.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47967 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoShare PluginView::init() between ports
vestbo@webkit.org [Wed, 2 Sep 2009 13:50:39 +0000 (13:50 +0000)]
Share PluginView::init() between ports

Reviewed by Eric Seidel.


The port-spesific bits are moved to platformStart(), which now returns
a bool based on the success of the platformStart.

m_hasPendingGeometryChange is now initialized in the initializer list
instead of in PluginViewQt's init()

PluginViewMac (currently used only by the Qt port) used to manually call
stop() in the case where an unsupported drawing or event model was detected.
This was wrong, as the m_status and m_isStarted fields should be used to
guard against doing operations on a plugin in this intermediate state.

* plugins/PluginView.cpp:
* plugins/PluginView.h:
* plugins/PluginViewNone.cpp:
* plugins/gtk/PluginViewGtk.cpp:
* plugins/mac/PluginViewMac.cpp:
* plugins/qt/PluginViewQt.cpp:
* plugins/win/PluginViewWin.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47966 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Wed, 2 Sep 2009 13:06:04 +0000 (13:06 +0000)]
2009-09-02  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Reviewed by Xan Lopez and Gustavo Noronha.

         [Gtk] Implement a WebDataSource for the gtk port

        Add data source API getters to WebKitWebFrame.
        Also add WebKitWebDataSource unit test.

        * tests/testwebdatasource.c: Added.
        * webkit/webkitwebframe.cpp:
        * webkit/webkitwebframe.h:

        Add WebKitWebDataSource unit test to the build script

        * GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47965 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Wed, 2 Sep 2009 12:58:59 +0000 (12:58 +0000)]
2009-09-02  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Reviewed by Xan Lopez and Gustavo Noronha.

        [Gtk] Implement a WebDataSource for the gtk port

        Subclass DocumentLoader for Gtk and implement it under the WebKit
        namespace. Add WebKitWebDataSource implementation that closely
        follows the mac implementation of the API.

        * WebCoreSupport/DocumentLoaderGtk.cpp: Added.
        * WebCoreSupport/DocumentLoaderGtk.h: Added.
        * webkit/webkitwebdatasource.cpp: Added.
        * webkit/webkitwebdatasource.h: Added.

        Add WebKitWebDataSource and DocumentLoaderGtk to the build script.

        * GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47964 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago[Qt] Make sure we relayout the page after evaluating JS
vestbo@webkit.org [Wed, 2 Sep 2009 12:40:18 +0000 (12:40 +0000)]
[Qt] Make sure we relayout the page after evaluating JS

Patch by Richard Moore <rich@kde.org> on 2009-09-02
Reviewed by Tor Arne Vestbø.


QtWebKit does not always seem to reflow the page when evaluating
javascript. This patch changes the way evaluateJavaScript works to
use the frameloader which ensures that this is done properly.

* Api/qwebframe.cpp:
* tests/qwebframe/tst_qwebframe.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47963 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore: https://bugs.webkit.org/show_bug.cgi?id=28737
benm@google.com [Wed, 2 Sep 2009 11:38:44 +0000 (11:38 +0000)]
WebCore: https://bugs.webkit.org/show_bug.cgi?id=28737

Patch by Ben Murdoch <benm@google.com> on 2009-09-02
Update JavaScript bindings in JSC and V8 for SQLTransaction to throw exceptions in the right places.

* bindings/js/JSSQLTransactionCustom.cpp:
(WebCore::JSSQLTransaction::executeSql): Throw an exception if no parameters are passed to executeSql().
* bindings/v8/V8Binding.cpp:
(WebCore::v8ValueToWebCoreString): Catch an exception if thrown from toString().
* bindings/v8/custom/V8SQLTransactionCustom.cpp:
(WebCore::CALLBACK_FUNC_DECL): Allow "arrayish objects" to be passed to executeSql.

LayoutTests: https://bugs.webkit.org/show_bug.cgi?id=28737

Patch by Ben Murdoch <benm@google.com> on 2009-09-02
Update the execute-sql-args.html to be agnostic to exception message formatting between JSC and V8 and throw an exception when no parameters are passed to the executeSql() function.

* storage/execute-sql-args-expected.txt:
* storage/execute-sql-args.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47962 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago[Qt] Fix layout-test plugins/plugin-javascript-access.html
vestbo@webkit.org [Wed, 2 Sep 2009 10:16:51 +0000 (10:16 +0000)]
[Qt] Fix layout-test plugins/plugin-javascript-access.html

Rubber-stamped by Simon Hausmann.

* DumpRenderTree/qt/TestNetscapePlugin/TestNetscapePlugin.pro:
* platform/qt/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47961 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Wed, 2 Sep 2009 09:59:45 +0000 (09:59 +0000)]
2009-09-01  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Timothy Hatcher.

        WebInspector: Migrate Databases tab to InjectedScript /
        serialized interaction.


        * inspector/InspectorDOMAgent.cpp:
        * inspector/front-end/Database.js:
        * inspector/front-end/DatabaseQueryView.js:
        * inspector/front-end/DatabaseTableView.js:
        * inspector/front-end/StoragePanel.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47960 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Norbert Leser <norbert.leser@nokia.com>
eric@webkit.org [Wed, 2 Sep 2009 09:49:31 +0000 (09:49 +0000)]
2009-09-02  Norbert Leser  <norbert.leser@nokia.com>

        Reviewed by Eric Seidel.

        Use fastMalloc when neither MMAP nor VIRTUALALLOC are enabled

        RegisterFile constructor currently throws #error when both
        MMAP and VIRTUALALLOC conditions fail.
        On any platform that does not provide these features
        (for instance, Symbian),
        the fallback should be regular malloc (or fastMalloc).
        It is functionally equivalent in this case, even though it may
        have certain drawbacks such as lack of dynamic pre-allocation.

        * interpreter/RegisterFile.cpp:
        * interpreter/RegisterFile.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47959 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Wed, 2 Sep 2009 09:46:21 +0000 (09:46 +0000)]
2009-09-02  Pavel Feldman  <pfeldman@chromium.org>

        Not reviewed (trivial followup fix).

        Web Inspector: Trivial follow up fix to r47944.


        * inspector/InspectorBackend.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47958 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Renata Hodovan <hodovan.renata@stud.u-szeged.hu>
zoltan@webkit.org [Wed, 2 Sep 2009 09:44:43 +0000 (09:44 +0000)]
2009-09-02  Renata Hodovan  <hodovan.renata@stud.u-szeged.hu>

        Reviewed by Ariya Hidayat.

        [Qt] Update css2.1/t1202-counter-* and css2.1/t1205* tests after new metrics.
        Remove these files from Skipped (and fixed Skipped according to Changeset 47815).

        * platform/qt/Skipped:
        * platform/qt/css2.1/t1202-counter-00-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-01-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-02-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-03-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-04-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-05-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-06-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-07-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-08-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-09-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-11-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-12-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-13-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-14-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-15-b-expected.txt:
        * platform/qt/css2.1/t1202-counter-16-f-expected.txt:
        * platform/qt/css2.1/t1205-c561-list-displ-00-b-expected.txt:
        * platform/qt/css2.1/t1205-c563-list-type-00-b-expected.txt:
        * platform/qt/css2.1/t1205-c563-list-type-01-b-expected.txt:
        * platform/qt/css2.1/t1205-c564-list-img-00-b-g-expected.txt:
        * platform/qt/css2.1/t1205-c565-list-pos-00-b-expected.txt:
        * platform/qt/css2.1/t1205-c566-list-stl-01-c-g-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47957 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Shinichiro Hamaji <hamaji@chromium.org>
eric@webkit.org [Wed, 2 Sep 2009 08:08:24 +0000 (08:08 +0000)]
2009-09-02  Shinichiro Hamaji  <hamaji@chromium.org>

        Reviewed by Eric Seidel.

        js tests should move into jstests subdirectory instead of resources/

        make-script-wrappers supports both resources and script-tests directories.
        run-webkit-tests ignores files in script-tests.
        Move script tests of animations to check the new script is working.

        * animations/animation-css-rule-types.html:
        * animations/resources/TEMPLATE.html: Removed.
        * animations/resources/animation-css-rule-types.js: Removed.
        * animations/script-tests: Added.
        * animations/script-tests/TEMPLATE.html: Copied from LayoutTests/animations/resources/TEMPLATE.html.
        * animations/script-tests/animation-css-rule-types.js: Copied from LayoutTests/animations/resources/animation-css-rule-types.js.
2009-09-02  Shinichiro Hamaji  <hamaji@chromium.org>

        Reviewed by Eric Seidel.

        js tests should move into jstests subdirectory instead of resources/

        make-script-wrappers supports both resources and script-tests directories.
        run-webkit-tests ignores files in script-tests.
        Move script tests of animations to check the new script is working.

        * Scripts/make-script-test-wrappers:
        * Scripts/run-webkit-tests:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47956 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-02 Szabo Carol <carol.szabo@nokia.com>
eric@webkit.org [Wed, 2 Sep 2009 08:01:23 +0000 (08:01 +0000)]
2009-09-02  Szabo Carol  <carol.szabo@nokia.com>

        Reviewed by David Levin.

        check-webkit-style uses python from /usr/bin instead of the PATH

        * Scripts/bugzilla-tool:
        * Scripts/check-webkit-style:
        * Scripts/run-webkit-unittests:
        * Scripts/update-sources-list.py:
          Changed the first line from
          #!/usr/bin/env python
          which causes python to be invoked from the path location returned
          by "which python" when any of these scripts are launched.
          these are currently all the python scripts in WebKitTools/Scripts.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47955 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Brian Weinstein <bweinstein@apple.com>
bweinstein@apple.com [Wed, 2 Sep 2009 05:03:56 +0000 (05:03 +0000)]
2009-09-01  Brian Weinstein  <bweinstein@apple.com>

        Reviewed by John Sullivan.

        Added a new case to aria-readonly.html to account for textareas instead of just input fields.

        * accessibility/aria-readonly-expected.txt:
        * accessibility/aria-readonly.html:

2009-09-01  Brian Weinstein  <bweinstein@apple.com>

        Reviewed by John Sullivan.

        Fixed a bug in handling of isReadOnly, where textareas were casted to HTMLInputElements.

        * accessibility/AccessibilityRenderObject.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47954 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago <http://webkit.org/b/28601> bugzilla-tool post-commits posts commits backwards
ddkilzer@apple.com [Wed, 2 Sep 2009 02:26:55 +0000 (02:26 +0000)]
    <webkit.org/b/28601> bugzilla-tool post-commits posts commits backwards

        Reviewed by Adam Barth.

        * Scripts/modules/scm.py:
        (Git.commit_ids_from_commitish_arguments): Reverse the list of
        commits returned from git-rev-list since we always want to post
        the oldest patches first to bugs.webkit.org.
        * Scripts/modules/scm_unittest.py:
        (run): Added return statement to return the output now that we
        want it sometimes.
        (SVNTestRepository._setup_test_commits): Added a fourth commit
        so the GitTest.test_commitish_order() test has more commits to
        work with.
        (GitTest.test_commitish_order): Added unit test for change to
        Git.commit_ids_from_commitish_arguments() in scm.py.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47953 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Patrick Mueller <Patrick_Mueller@us.ibm.com>
eric@webkit.org [Wed, 2 Sep 2009 00:38:08 +0000 (00:38 +0000)]
2009-09-01  Patrick Mueller  <Patrick_Mueller@us.ibm.com>

        Reviewed by Timothy Hatcher.

        add utility method to enable logging messages for inspector development

        manual test case added (see below)

        * inspector/front-end/ConsoleView.js:
        * inspector/front-end/inspector.css:
        * inspector/front-end/inspector.js:
        * manual-tests/inspector/webinspector-log.html: Added.
        * page/Console.cpp:
        * page/Console.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47952 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Eric Seidel <eric@webkit.org>
eric@webkit.org [Wed, 2 Sep 2009 00:15:33 +0000 (00:15 +0000)]
2009-09-01  Eric Seidel  <eric@webkit.org>

        Rubber stamped by Simon Fraser.

        REGRESSION: media/video-source-add-src.html (and other media tests?) crashing/timing-out intermittently

        Disabling this test on all platforms until
        a media expert can have a chance to look at it.
        See bug 28845 for further discussion.

        * media/video-source-add-src.html-disabled: Renamed from LayoutTests/media/video-source-add-src.html.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47951 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Darin Adler.
ap@apple.com [Wed, 2 Sep 2009 00:06:05 +0000 (00:06 +0000)]
    Reviewed by Darin Adler.

        When a rare EUC-JP character is present, explicitly (and correctly) labelled EUC-JP document
        is mistreated as Shift_JIS

        <rdar://problem/5828506> Safari should not try auto-detecting if charset information exists

        WebKit used to unconditionally auto-detect document encoding if it was one of Japanese ones.
        This didn't really match any browser, and did cause problems on some sites.

        Our detector is very quick to label an EUC-JP page as Shift_JIS if it notices a byte that
        cannot be part of EUC-JP source. However, such bytes do appear in Web pages sometimes,
        especially in comments. Also, some valid EUC-JP sequences are not known to it.

        Some pages that are fixed by this change:

        <rdar://problem/5934750> Encoding in incorrect at http://cocodeparis.blog45.fc2.com/. This
        page has incorrect encoding in title, so no browser displays it 100% fine. We now match
        Firefox and IE.

        <rdar://problem/6007713> The web page http://www.sf.us.emb-japan.go.jp/top.htm is garbled
        in Safari. There is a stray byte in comments that was makes charset detector think it can't
        be EUC-JP.

        <rdar://problem/6965711> A Japanese web page www.nippontv.studiolegend.tv is garbled in
        Safari. Many comments on this page are encoded as Shift_JIS - IE magically displays that
        fine in View Source mode (it doesn't seem to have an ability to switch between EUC-JP and
        Shift_JIS in normal HTML view though).

        Test: fast/encoding/japanese-encoding-mix.html

        * loader/TextResourceDecoder.cpp: (WebCore::TextResourceDecoder::decode): Only invoke auto-
        detection if there is no solid charset information yet.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47950 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago <http://webkit.org/b/28877> Implement bugzilla-tool mark-fixed
ddkilzer@apple.com [Tue, 1 Sep 2009 23:07:53 +0000 (23:07 +0000)]
    <webkit.org/b/28877> Implement bugzilla-tool mark-fixed

        Reviewed by David Levin.

        The mark-fixed subcommand is for those times when you don't use
        bugzilla-tool to commit a patch, but you want to use it to close
        the bug with a committed-revision message.

        * Scripts/bugzilla-tool:
        (bug_comment_from_svn_revision): Added.  Extracted from
        (bug_comment_from_commit_text): Extracted
        bug_comment_from_svn_revision() from this method.
        (MarkBugFixed.__init__): Added.
        (MarkBugFixed._fetch_commit_log): Added.  Retrieves the commit
        log from the last commit if no svn revision is specified, else
        the commit log for the specified svn revision.
        (MarkBugFixed._determine_bug_id_and_svn_revision): Added.
        Attempts to determine the bug id and svn revision if one or both
        were not defined on the command line.
        (MarkBugFixed.execute): Added.  Adds a comment about the
        revision that fixed the bug and closes the bug.
        (BugzillaTool.__init__): Added mark-fixed subcommand.
        * Scripts/modules/bugzilla.py:
        (Bugzilla.fetch_title_from_bug): Added.  Returns the title of a
        bug given a bug id.
        * Scripts/modules/scm.py:
        (SCM.strip_r_from_svn_revision): Added.  Utility method to strip
        the leading 'r' from an svn revision.
        (SCM.svn_commit_log): Added.  Subclasses must override.
        (SCM.last_svn_commit_log): Added.  Subclasses must override.
        (SVN.svn_commit_log): Added.  Returns svn log for a given
        (SVN.last_svn_commit_log): Added.  Uses svnversion to find the
        last commit in an svn working directory and then runs svn log.
        (Git.svn_commit_log): Added.  Returns svn log for a given
        (Git.last_svn_commit_log): Added.  Runs git-svn-log with a limit
        of one log message.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47949 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Victor Wang <victorw@chromium.org>
dglazkov@chromium.org [Tue, 1 Sep 2009 20:18:51 +0000 (20:18 +0000)]
2009-09-01  Victor Wang  <victorw@chromium.org>

        Reviewed by Darin Adler.


        Fix fast/dom/StyleSheet/ownerNode-lifetime-2.html
        The problem is WebKit <link> nodes will only have a
        style sheet created if they have href attribute.
        The src attribute to non-existant css doesn't work.
        Change src to href to make this test works in WebKit.

        Also update the expectation result.

        * fast/dom/StyleSheet/ownerNode-lifetime-2-expected.txt:
        * fast/dom/StyleSheet/ownerNode-lifetime-2.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47948 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago[V8] FileList cannot be accessed via index in Chromium.
jianli@chromium.org [Tue, 1 Sep 2009 20:16:57 +0000 (20:16 +0000)]
[V8] FileList cannot be accessed via index in Chromium.

Reviewed by Dimitri Glazkov.

Tested by clipboard-file-access.html.

* WebCore.gypi:
* bindings/v8/V8DOMWrapper.cpp:
* bindings/v8/custom/V8CustomBinding.h:
* bindings/v8/custom/V8FileListCustom.cpp: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47947 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoRemove the workaround added in r47316.
jhoneycutt@apple.com [Tue, 1 Sep 2009 19:52:30 +0000 (19:52 +0000)]
Remove the workaround added in r47316.

Reviewed by Sam Weinig.

* accessibility/AXObjectCache.cpp:

* accessibility/AXObjectCache.h:
Remove the Document argument to the AXObjectCache constructor and the
m_document member variable, and replace the no-parameter
handleFocusedUIElementChanged() with the two-parameter GTK function.

* accessibility/chromium/AXObjectCacheChromium.cpp:

* accessibility/gtk/AXObjectCacheAtk.cpp:

* accessibility/gtk/AccessibilityObjectWrapperAtk.cpp:
Don't pass a null Document when constructing the AXObjectCache.

* accessibility/mac/AXObjectCacheMac.mm:

* accessibility/win/AXObjectCacheWin.cpp:
Get the document from the RenderObject; m_document was removed.

* dom/Document.cpp:
Don't pass the Document when constructing the AXObjectCache.
Pass the old and new focused RenderObjects when calling

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47946 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago <http://webkit.org/b/28880> svn-apply --force doesn't actually work
ddkilzer@apple.com [Tue, 1 Sep 2009 18:42:29 +0000 (18:42 +0000)]
    <webkit.org/b/28880> svn-apply --force doesn't actually work

        Reviewed by Brady Eidson.

        * Scripts/svn-apply:
        (applyPatch): Add "--force" to $options arrayref if $force is

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47945 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Yury Semikhatsky <yurys@chromium.org>
pfeldman@chromium.org [Tue, 1 Sep 2009 18:17:09 +0000 (18:17 +0000)]
2009-09-01  Yury Semikhatsky  <yurys@chromium.org>

        Reviewed by Timothy Hatcher.

        Check that InspectorDOMAgent and InspectorFrontend are not null
        before accessing them in InspectorBackend. They may be null if
        InspectorController is closed before an asynchronous operation is


        * inspector/InspectorBackend.cpp:
        * inspector/InspectorBackend.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47944 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoPage Cache should support pages with frames
beidson@apple.com [Tue, 1 Sep 2009 18:13:58 +0000 (18:13 +0000)]
Page Cache should support pages with frames

Reviewed by Sam Weinig.

Things learned in the full task that this patch addresses:
- The notions of clearing a CachedFrame and destroying a CachedFrame were two concepts that
  needed to be managed separately.
- Once we start restoring pages with multiple CachedFrames, we'll have to be calling
  FrameLoader::open(CachedFrame&) multiple times with different CachedFrames, and this process
  will be guided by the CachedFrames themselves.

No new tests. (No change in behavior, current layout tests pass)

* history/CachedFrame.cpp:
(WebCore::CachedFrame::clear): Only clear pointers out.
(WebCore::CachedFrame::destroy): Perform the destructive cleanup work here - Things that shouldn't
  occur simply by navigating back to a page with a CachedFrame.
* history/CachedFrame.h:

* history/CachedPage.cpp:
(WebCore::CachedPage::restore): Call clear() after restoring.
* history/CachedPage.h:

* history/PageCache.cpp:
(WebCore::PageCache::releaseAutoreleasedPagesNow): destroy() CachedPages that are being pruned,
  instead of clearing them.

* loader/FrameLoader.cpp:
(WebCore::FrameLoader::commitProvisionalLoad): Let the CachedPage clear() itself after restoring.
(WebCore::FrameLoader::open): Don't ::open() the CachedFrame, as the CachedPage does that now.
* loader/FrameLoader.h:  Add the friending so CachedPage and CachedFrame can better guide the process.
  Even though CachedFrame::restore() isn't used for now, it will be soon.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47943 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoAdd a "lowMemoryNotification" method to the V8 ScriptController class.
benm@google.com [Tue, 1 Sep 2009 18:00:54 +0000 (18:00 +0000)]
Add a "lowMemoryNotification" method to the V8 ScriptController class.
This is needed in order to allow the host application (Android Web browser in this case)
to call the v8::V8::LowMemoryNotification() method added to V8 in

Patch by Andrei Popescu <andreip@google.com> on 2009-09-01
Reviewed by David Levin.


* bindings/v8/ScriptController.cpp:
* bindings/v8/ScriptController.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47942 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Fumitoshi Ukai <ukai@chromium.org>
abarth@webkit.org [Tue, 1 Sep 2009 16:39:12 +0000 (16:39 +0000)]
2009-09-01  Fumitoshi Ukai  <ukai@chromium.org>

        Reviewed by Adam Barth.

        V8 Bindings for WebSocket API.

        * bindings/v8/DOMObjectsInclude.h:
        * bindings/v8/DerivedSourcesAllInOne.cpp:
        * bindings/v8/V8DOMWrapper.cpp:
        * bindings/v8/V8DOMWrapper.h:
        * bindings/v8/V8Index.cpp:
        * bindings/v8/V8Index.h:
        * bindings/v8/WorkerContextExecutionProxy.cpp:
        * bindings/v8/custom/V8CustomBinding.h:
        * bindings/v8/custom/V8WebSocketCustom.cpp: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47941 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoPost a notification when a WebView is closing
aroben@apple.com [Tue, 1 Sep 2009 16:35:01 +0000 (16:35 +0000)]
Post a notification when a WebView is closing

This is useful for situations where some code wants to react to the
WebView closing but isn't in charge of the WebView and can't otherwise
be notified. This notification could replace
IWebUIDelegatePrivate::webViewClosing if we decide that the
notification system isn't too burdensome.


Reviewed by John Sullivan.

* Interfaces/IWebViewPrivate.idl: Added the
WebViewWillCloseNotification string constant.
* WebKit.vcproj/WebKit.vcproj: Link against comsuppw.lib in all
configurations so that we can use _bstr_t.
* WebView.cpp:
(WebView::close): Post the WebViewWillCloseNotification.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47940 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoUnreviewed (build fix).
pkasting@chromium.org [Tue, 1 Sep 2009 16:17:07 +0000 (16:17 +0000)]
Unreviewed (build fix).

Missed a "." -> "->" conversion.

* platform/graphics/qt/ImageDecoderQt.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47939 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoUnreviewed (build fix).
pkasting@chromium.org [Tue, 1 Sep 2009 16:12:23 +0000 (16:12 +0000)]
Unreviewed (build fix).

Try to fix Qt build, take 2.  The Qt code had what looks like a
broken virtual function override.

* platform/graphics/qt/ImageDecoderQt.cpp:
* platform/graphics/qt/ImageDecoderQt.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47938 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Tue, 1 Sep 2009 16:09:09 +0000 (16:09 +0000)]
2009-09-01  Adam Barth  <abarth@webkit.org>

        Reviewed by Sam Weinig.

        Uninitialized variable in XSSAuditor

        One ought to initialize variables before use even when it doesn't
        matter.  No behavior change.

        * page/XSSAuditor.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47937 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoUnreviewed (build fix).
pkasting@chromium.org [Tue, 1 Sep 2009 15:57:47 +0000 (15:57 +0000)]
Unreviewed (build fix).

Try to fix Qt build.

* platform/graphics/ImageSource.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47936 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agohttps://bugs.webkit.org/show_bug.cgi?id=27965
pkasting@chromium.org [Tue, 1 Sep 2009 15:50:53 +0000 (15:50 +0000)]
Eliminate most of ImageSourceQt.cpp in favor of ImageSource.cpp.

Reviewed by Eric Seidel.

* WebCore.pro:
* platform/graphics/ImageSource.cpp:
* platform/graphics/qt/ImageSourceQt.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47935 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Tue, 1 Sep 2009 14:58:52 +0000 (14:58 +0000)]
2009-09-01  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Timothy Hatcher.

        WebInspector: Make InjectedScript self-contained (move necessary
        methods from utilities.js into InjectedScript).


        * inspector/front-end/ElementsPanel.js:
        * inspector/front-end/InjectedScript.js:
        * inspector/front-end/utilities.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47934 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Tue, 1 Sep 2009 14:41:07 +0000 (14:41 +0000)]
2009-09-01  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Timothy Hatcher.

        WebInspector: console log message repeat count double.


        * inspector/front-end/InjectedScript.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47933 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago <http://webkit.org/b/28725> resolve-ChangeLogs: determineVCSRoot() returns...
ddkilzer@apple.com [Tue, 1 Sep 2009 14:11:34 +0000 (14:11 +0000)]
    <webkit.org/b/28725> resolve-ChangeLogs: determineVCSRoot() returns incorrect repository root during git filter-branch

        Reviewed by Adam Roben.

        When git-filter-branch has been invoked to rewrite ChangeLog
        files on series of git commits, it changes directories into
        .git-rewrite/t before re-running resolve-ChangeLogs.  This
        causes determineVCSRoot() in VCSUtils.pm to return
        ".git-rewrite/t", which causes that path to be prepended to all
        ChangeLog paths, which results in an error like this:

            error: pathspec '.git-rewrite/t/ChangeLog' did not match any file(s) known to git.
            Died at WebKitTools/Scripts/resolve-ChangeLogs line 376.

        The correct way to fix this is not to try to find the repository
        root when invoked by git-filter-branch.

        * Scripts/resolve-ChangeLogs: If isInGitFilterBranch() is true,
        set $relativePath to '.' instead of calling
        (isInGitFilterBranch): Added.  Checks for the existence of the
        MAPPED_PREVIOUS_COMMIT environment variable.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47932 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>
kov@webkit.org [Tue, 1 Sep 2009 13:06:58 +0000 (13:06 +0000)]
2009-09-01  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Reviewed by Xan Lopez.

        Avoid causing unexpected issues with window sizing when sites
        disable scrollbars completely.

        * WebCoreSupport/ChromeClientGtk.cpp:
        * tests/testwindow.c:
        * webkit/webkitwebframe.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47931 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoLayoutTests:
xan@webkit.org [Tue, 1 Sep 2009 12:50:41 +0000 (12:50 +0000)]

2009-09-01  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        Add support for Fedora distros in the http tests

        Add httpd conf file for Fedora.

        * http/conf/fedora-httpd.conf: Added.


2009-09-01  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        Add support for Fedora distros in the http tests

        Add detection code for Fedora distribution, and use the proper
        httpd conf file when needed.

        * Scripts/run-webkit-httpd:
        * Scripts/run-webkit-tests:
        * Scripts/webkitdirs.pm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47930 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Tue, 1 Sep 2009 12:19:28 +0000 (12:19 +0000)]
2009-09-01  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Reviewed by Xan Lopez and Gustavo Noronha.

        [Gtk] Implement a WebDataSource for the gtk port

        Implement WebKitWebResource for the resource-related API for

        * GNUmakefile.am:

        Add WebKitWebResource to the build script.

        * GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47929 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Tue, 1 Sep 2009 12:05:47 +0000 (12:05 +0000)]
2009-09-01  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Reviewed by Eric Seidel.

        [Gtk] DRT needs implementation of overridePreference

        Unskip tests that are now working.

        * platform/gtk/Skipped:

        Implement overridePreference.

        * DumpRenderTree/gtk/DumpRenderTree.cpp:
        * DumpRenderTree/gtk/LayoutTestControllerGtk.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47928 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Jan Michael Alonzo <jmalonzo@webkit.org>
jmalonzo@webkit.org [Tue, 1 Sep 2009 11:55:24 +0000 (11:55 +0000)]
2009-09-01  Jan Michael Alonzo  <jmalonzo@webkit.org>

        Reviewed by Mark Rowe.

        [Gtk] Fix DATALIST build

        Unskip now working fast/forms/input-list.html.

        * platform/gtk/Skipped:


        * GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47927 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Cameron McCormack <cam@mcc.id.au>
eric@webkit.org [Tue, 1 Sep 2009 11:11:47 +0000 (11:11 +0000)]
2009-09-01  Cameron McCormack  <cam@mcc.id.au>

        Reviewed by Eric Seidel.

        SVGSVGElement suspend methods argument/return types incorrect

        Also fix the style of the argument names.

        * svg/SVGSVGElement.cpp:
        * svg/SVGSVGElement.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47926 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Roland Steiner <rolandsteiner@google.com>
eric@webkit.org [Tue, 1 Sep 2009 11:01:30 +0000 (11:01 +0000)]
2009-09-01  Roland Steiner  <rolandsteiner@google.com>

        Reviewed by Eric Seidel.

        Fix bug 28808:  [Skia] Fix flaky layout test svg/dynamic-updates/SVGClipPathElement-dom-clipPathUnits-attr.html [DEBUG]

        The bug fix adds an explicit check for the validity of the path when it is
        transformed into local coordinates.
        This was the most suitable place I could find that didn't cause an inordinate
        amount of refactoring. Also, it seems pertinent and may also catch other cases
        that are not limited to clipping paths.

        TEST: existing svg/dynamic-updates/SVGClipPathElement-dom-clipPathUnits-attr.html

        * platform/graphics/skia/GraphicsContextSkia.cpp: make isPathSkiaSafe accessible
        * platform/graphics/skia/PlatformContextSkia.cpp:
        (PlatformContextSkia::currentPathInLocalCoordinates): check Skia safety of path

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47925 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Tue, 1 Sep 2009 10:54:48 +0000 (10:54 +0000)]
2009-09-01  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Jan Alonzo.

        [GTK] Do not emit extra FINISHED load-status signals

        Do not emit FINISHED load-status signals on our own, just rely on
        WebCore for that. Avoids emitting FINISHED several times when we
        fail to load a URI.

        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        * tests/testloading.c:
        * webkit/webkitwebframe.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47924 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Joseph Pecoraro <joepeck@webkit.org>
eric@webkit.org [Tue, 1 Sep 2009 10:24:36 +0000 (10:24 +0000)]
2009-09-01  Joseph Pecoraro  <joepeck@webkit.org>

        <http://webkit.org/b/28623> svn-[un]apply should change directories to the repository root before [un]applying

        Reviewed by Eric Seidel.

        Jump back and forth between the repository root directory (to apply) and the
        directory the script was run from (to find the patch).

        * Scripts/svn-apply:
        * Scripts/svn-unapply:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47923 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Yusuke Sato <yusukes@chromium.org>
eric@webkit.org [Tue, 1 Sep 2009 10:14:49 +0000 (10:14 +0000)]
2009-09-01  Yusuke Sato  <yusukes@chromium.org>

        Reviewed by Eric Seidel.

        [Chromium] Combining Diacritical Marks (U+0300..) are not handled correctly.

        Normalize (NFC) a TextRun when the run contains combining diacritical marks
        so that Skia can pick a correct glyph without relying on GSUB table in a font.
        This change is for Chromium Linux.

        * platform/graphics/chromium/FontLinux.cpp:
        (WebCore::TextRunWalker::getTextRun): Added.
        (WebCore::TextRunWalker::getNormalizedTextRun): Added.

        This function is called when the run contains U+0300..U+036F and converts characters in
        the run to the combined form (NFC) using ICU.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47922 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Vitaly Repeshko <vitalyr@chromium.org>
eric@webkit.org [Tue, 1 Sep 2009 10:05:33 +0000 (10:05 +0000)]
2009-09-01  Vitaly Repeshko  <vitalyr@chromium.org>

        Reviewed by Dimitri Glazkov.

        [V8] Inlined fast paths of V8DOMWrapper::convertNodeToV8Object and

        * bindings/v8/V8DOMWrapper.cpp:
        (WebCore::V8DOMWrapper::convertDocumentToV8Object): Special
        function for document wrapping.
        (WebCore::V8DOMWrapper::convertNewNodeToV8Object): Slow case of
        wrapping a new node.
        * bindings/v8/V8DOMWrapper.h:
        (WebCore::V8DOMWrapper::convertNodeToV8Object): Extracted fast
        path from V8DOMWrapper.cpp.
        * bindings/v8/V8Proxy.cpp:
        * bindings/v8/V8Proxy.h:
        (WebCore::V8Proxy::createWrapperFromCache): Extracted fast path
        from V8Proxy.cpp.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47921 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Eric Seidel <eric@webkit.org>
eric@webkit.org [Tue, 1 Sep 2009 09:34:54 +0000 (09:34 +0000)]
2009-09-01  Eric Seidel  <eric@webkit.org>

        No review, windows build fix only.  Clearly I should have read the patch closer. :(

        Allow excluding certain plugins from loading

        * plugins/PluginDatabase.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47920 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Alexander Pavlov <apavlov@chromium.org>
pfeldman@chromium.org [Tue, 1 Sep 2009 09:31:19 +0000 (09:31 +0000)]
2009-09-01  Alexander Pavlov  <apavlov@chromium.org>

        Reviewed by Timothy Hatcher.

        Implement conditional breakpoints in the Web Inspector backend
        and add frontend JavaScript stubs.

        * inspector/InspectorBackend.cpp:
        * inspector/InspectorBackend.h:
        * inspector/InspectorBackend.idl:
        * inspector/JavaScriptDebugServer.cpp:
        * inspector/JavaScriptDebugServer.h:
        * inspector/front-end/Breakpoint.js:
        (WebInspector.Breakpoint.prototype.get id):
        (WebInspector.Breakpoint.prototype.get condition):
        (WebInspector.Breakpoint.prototype.set condition):
        * inspector/front-end/BreakpointsSidebarPane.js:
        * inspector/front-end/ScriptsPanel.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47919 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-09-01 Marius Renn <damarvy@gmail.com>
eric@webkit.org [Tue, 1 Sep 2009 09:13:41 +0000 (09:13 +0000)]
2009-09-01  Marius Renn  <damarvy@gmail.com>

        Reviewed by Eric Seidel.

        Added delegate to PluginDatabase to disable loading certain plugins.
        This is useful when you want to load from the standard plugin
        directory, but want to exclude certain plugins. Plugins may be
        excluded early on by name, so that no plugin code is executed (which
        may have caused a crash).

        * plugins/PluginDatabase.cpp:
        * plugins/PluginDatabase.h:
        * plugins/PluginDatabaseClient.h: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47918 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoRubberstamped by Sam Weinig
beidson@apple.com [Tue, 1 Sep 2009 03:45:38 +0000 (03:45 +0000)]
Rubberstamped by Sam Weinig

* WebCore.base.exp: Removed an unneeded symbol export.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47917 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore
kov@webkit.org [Tue, 1 Sep 2009 01:34:33 +0000 (01:34 +0000)]

2009-08-31  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Reviewed by Alexey Proskuryakov.

        When downloading a file, two GET requests are sent to the HTTP server

        Can't think of a way to test this.

        * loader/ResourceLoader.cpp:
        (WebCore::ResourceLoader::releaseResources): only nullify the
        handle's client when it is the ResourceLoader, to support the case
        in which a download API uses a different client


2009-08-31  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Reviewed by Alexey Proskuryakov.

        When downloading a file, two GET requests are sent to the HTTP server

        Support reusing a connection that is already ongoing, instead of
        cancelling the connection and creating another.

        * WebCoreSupport/FrameLoaderClientGtk.cpp:
        * webkit/webkitdownload.cpp:
        * webkit/webkitprivate.h:
        * webkit/webkitwebview.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47916 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-08-31 Brian Weinstein <bweinstein@apple.com>
bweinstein@apple.com [Mon, 31 Aug 2009 23:44:33 +0000 (23:44 +0000)]
2009-08-31  Brian Weinstein  <bweinstein@apple.com>

        Reviewed by Darin Adler.

        Text Fields and Text Areas are reported as read-only by inspect32.exe.

        Added another case in AcccessibiltyRenderObject::isReadOnly to test if text fields
        and texts areas are read-only.

        * accessibility/AccessibilityRenderObject.cpp:

2009-08-31  Brian Weinstein  <bweinstein@apple.com>

        Reviewed by Darin Adler.

        Added a case to check for readonly and non-readonly set through the readonly attribute
        instead of through aria-readonly. This tests our new case in AccessibilityRenderObject::isReadOnly.

        * accessibility/aria-readonly-expected.txt:
        * accessibility/aria-readonly.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47915 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore: fast/workers/worker-lifecycle.html crashes intermittently on snowleopard
atwilson@chromium.org [Mon, 31 Aug 2009 23:40:08 +0000 (23:40 +0000)]
WebCore: fast/workers/worker-lifecycle.html crashes intermittently on snowleopard

Reviewed by Alexey Proskuryakov.

Changed WorkerContext::close() to no longer stop the worker thread.
It is now the parent's responsibility to shut down the worker thread when it is notified that the context is closed.

* workers/AbstractWorker.cpp:
Updated dispatchScriptErrorEvent to pass-through the "handled" value from dispatchEvent(), to allow event handlers added with addEventListener() to mark events as handled.
* workers/DefaultSharedWorkerRepository.cpp:
Now shuts down the worker thread when notified that the context is closed.
Now handles being invoked when the context is already in the process of shutting down.
* workers/WorkerContext.cpp:
No longer calls WorkerThread::stop() to avoid race conditions with shutting down the thread while the parent is still interacting with it.
* workers/WorkerContext.h:
isClosing() is now public so WorkerRunLoop can call it to determine whether to process tasks.
* workers/WorkerMessagingProxy.cpp:
Error events should still be delivered even if the worker thread is closing.
Also fixed problem where error events were not dispatched if error handlers were added via addEventListener().
New task that invokes terminateWorkerContext() on the parent thread when the context is closed.
Now fires off a WorkerTerminateTask to shut down the thread when the context is closed.
* workers/WorkerMessagingProxy.h:
WorkerMessagingProxy now overrides workerContextClosed().
* workers/WorkerRunLoop.cpp:
Now drops tasks on the floor if the WorkerContext is closing.

LayoutTests: fast/workers/worker-lifecycle.html crashes intermittently on snowleopard

Reviewed by Alexey Proskuryakov.

Removed worker-lifecycle.html from the list of Skipped tests since it no longer crashes.
Added test for case where we capture script errors via an event listener added via addEventListener().

* fast/workers/worker-script-error.html
* fast/workers/worker-script-error-expected.txt
Added testScriptErrorAddEventListener to test catching script errors via addEventListener().
* platform/mac-snowleopard/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47914 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoRubber-stamped by Mark Rowe.
bdakin@apple.com [Mon, 31 Aug 2009 23:36:39 +0000 (23:36 +0000)]
Rubber-stamped by Mark Rowe.

Meant to update these results with my earlier check-in, but I

* platform/mac/fast/inspector/style-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47913 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoNot reviewed (build fix)
ojan@chromium.org [Mon, 31 Aug 2009 23:30:23 +0000 (23:30 +0000)]
Not reviewed (build fix)

Build fix for Chromium to match r49707.

* bindings/v8/ScriptCallFrame.cpp:
* bindings/v8/ScriptSourceCode.h:
* platform/KURLGoogle.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47912 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Gtk build fix.
ap@apple.com [Mon, 31 Aug 2009 23:01:22 +0000 (23:01 +0000)]
    Gtk build fix.

        This is also a behavior fix, at least on debug builds - a caller of
        webkit_web_history_item_new_with_data() could pass an URL that could cause an assertion
        failure in KURL.

        * webkit/webkitwebhistoryitem.cpp: (webkit_web_history_item_new_with_data): Pass a null
        KURL as base to parse properly.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47911 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Darin Adler.
ap@apple.com [Mon, 31 Aug 2009 22:49:33 +0000 (22:49 +0000)]
    Reviewed by Darin Adler.

        Element.baseURI parses xml:base attribute incorrectly

        Test: fast/dom/base-attribute-parsing.xhtml

        * dom/Element.cpp: (WebCore::Element::baseURI): Avoid assertion failure (no change in
        release mode).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47910 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
dglazkov@chromium.org [Mon, 31 Aug 2009 22:27:01 +0000 (22:27 +0000)]

2009-08-31  Dimitri Glazkov  <dglazkov@chromium.org>

        Reverting http://trac.webkit.org/changeset/47904, because it caused
        layout test failure.

        Test: fast/dom/Window/new-window-opener.html


2009-08-31  Dimitri Glazkov  <dglazkov@chromium.org>

    Reverting http://trac.webkit.org/changeset/47904, because it caused
    layout test failure.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47909 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoRemove WebViewEditingContextMenu.nib and WebViewEditingContextMenuOld.nib.
mrowe@apple.com [Mon, 31 Aug 2009 22:09:24 +0000 (22:09 +0000)]
Remove WebViewEditingContextMenu.nib and WebViewEditingContextMenuOld.nib.

Rubber-stamped by Beth Dakin.

We stopped using these back in r18592 when we switched over to context menus
driven by WebCore.

* English.lproj/WebViewEditingContextMenu.nib/classes.nib: Removed.
* English.lproj/WebViewEditingContextMenu.nib/info.nib: Removed.
* English.lproj/WebViewEditingContextMenu.nib/objects.nib: Removed.
* English.lproj/WebViewEditingContextMenuOld.nib/classes.nib: Removed.
* English.lproj/WebViewEditingContextMenuOld.nib/info.nib: Removed.
* English.lproj/WebViewEditingContextMenuOld.nib/objects.nib: Removed.
* WebKit.xcodeproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47908 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago Reviewed by Darin Adler.
ap@apple.com [Mon, 31 Aug 2009 22:07:29 +0000 (22:07 +0000)]
    Reviewed by Darin Adler.

        Rename KURL single argument constructor to avoid confusion

        * platform/KURL.h: The constructor that used to be single argument should now be invoked
        as KURL(ParsedURLString, myString).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47907 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore: Fix for https://bugs.webkit.org/show_bug.cgi?id=28635 [CSS3
bdakin@apple.com [Mon, 31 Aug 2009 22:05:24 +0000 (22:05 +0000)]
WebCore: Fix for https://bugs.webkit.org/show_bug.cgi?id=28635 [CSS3
Backgrounds and Borders] Add support for 2-keyword values for

Reviewed by Darin Adler.

This patch allows background-repeat to take two values by making
background-repeat just like background-position internally. There
is a little extra legwork for background-repeat because the spec
indicates that its computed value should be equivalent to how it
was specified. I keep track of the specified thing by setting the
implicit flag whenever background-repeat is defined with only one
value (since internally, we now store this as 2 values.)

Here we can't access the implicit flag, so for backwards-
compatibility's sake, we always return one value when that makes
* css/CSSComputedStyleDeclaration.cpp:

Here we check for the implicit flag and return one value when it is
* css/CSSMutableStyleDeclaration.cpp:

Add support for CSSPropertyBackgroundRepeatX and
* css/CSSParser.cpp:
* css/CSSParser.h:

Get rid of mappings to RepeatXFill and RepeatYFill since we don't
need those parts of the EFillRepeat enum anymore.
* css/CSSPrimitiveValueMappings.h:
(WebCore::CSSPrimitiveValue::operator EFillRepeat):

Add CSSPropertyBackgroundRepeatX and CSSPropertyBackgroundRepeatY
* css/CSSPropertyLonghand.cpp:

Add background-repeat-x and -y.
* css/CSSPropertyNames.in:

Break repeat into x and y.
* css/CSSStyleSelector.cpp:
* css/CSSStyleSelector.h:

There is no more RepeatXFill. Instead, look for
fillLayer->repeatX() == RepeatFill
* rendering/RenderBoxModelObject.cpp:

Break m_repeat into m_repeatX and m_repeatY
* rendering/style/FillLayer.cpp:
* rendering/style/FillLayer.h:

Again, break backgroundRepeat into backgroundRepeatX and
* rendering/style/RenderStyle.h:

We don't need RepeatXFill or RepeatYFill. Now that we store two
values, they are represented by Repeat-NoRepeat and NoRepeat-
Repeat, respectively.
* rendering/style/RenderStyleConstants.h:

LayoutTests: Tests for https://bugs.webkit.org/show_bug.cgi?id=28635 [CSS3
Backgrounds and Borders] Add support for 2-keyword values for

Reviewed by Darin Adler.

This test is updated to test for two values.
* fast/backgrounds/repeat/parsing-background-repeat-expected.txt:
* fast/backgrounds/repeat/resources/parsing-background-repeat.js:

These tests are updated to reflect new order.
* fast/css/background-position-serialize-expected.txt:
* fast/css/remove-shorthand-expected.txt:

Reflect new order in results, and make the test dump as text.
* fast/inspector/style.html:
* platform/mac/fast/inspector/style-expected.checksum: Removed.
* platform/mac/fast/inspector/style-expected.png: Removed.
* platform/mac/fast/inspector/style-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47906 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-08-31 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Mon, 31 Aug 2009 21:00:02 +0000 (21:00 +0000)]
2009-08-31  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dan Bernstein.

        Compositing layer size needs to take into account software reflections on children.

        When computing the bounds of a composited layer, take software-rendered
        reflections into account.

        Test: compositing/reflections/reflection-in-composited.html

        * rendering/RenderLayerCompositor.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47905 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoWebCore:
dglazkov@chromium.org [Mon, 31 Aug 2009 20:37:51 +0000 (20:37 +0000)]

2009-08-31  Mark Mentovai  <mark@chromium.org>

        Reviewed by Dave Hyatt.


        Perform a layout prior to checking whether the scrollbar modes are
        off, on, or automatic.  The modes may change during layout.

        * platform/ScrollView.cpp:


2009-08-31  Mark Mentovai  <mark@chromium.org>

        Reviewed by Dave Hyatt.


        Perform a layout prior to checking whether the scrollbar modes are
        off, on, or automatic.  The modes may change during layout.

        * WebView/WebDynamicScrollBarsView.mm:
        (-[WebDynamicScrollBarsView updateScrollers]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47904 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-08-31 Patrick Mueller <Patrick_Mueller@us.ibm.com>
eric@webkit.org [Mon, 31 Aug 2009 20:15:17 +0000 (20:15 +0000)]
2009-08-31  Patrick Mueller  <Patrick_Mueller@us.ibm.com>

        Reviewed by Timothy Hatcher.

        Inspector Request Headers Should Show Data Sent With Request

        Manual test added - see below.

        * English.lproj/localizedStrings.js:
        * inspector/InspectorResource.cpp:
        * inspector/InspectorResource.h:
        * inspector/front-end/Resource.js:
        * inspector/front-end/ResourceView.js:
        * inspector/front-end/inspector.css:
        * inspector/front-end/inspector.js:
        * manual-tests/inspector/display-form-data.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47903 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoFixed typo.
barraclough@apple.com [Mon, 31 Aug 2009 19:31:58 +0000 (19:31 +0000)]
Fixed typo.

Patch by Robert Agoston <Agoston.Robert@stud.u-szeged.hu> on 2009-08-31
Reviewed by Gavin Barraclough.

* parser/Parser.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47902 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-08-31 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Mon, 31 Aug 2009 19:01:11 +0000 (19:01 +0000)]
2009-08-31  Simon Fraser  <simon.fraser@apple.com>

        Followup from r47877; add test for the Obj-C wrapper class for WebkitCSSTransformValue

        * platform/mac/fast/dom/wrapper-classes-objc-expected.txt:
        * platform/mac/fast/dom/wrapper-classes-objc.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47901 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-08-28 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Mon, 31 Aug 2009 18:37:47 +0000 (18:37 +0000)]
2009-08-28  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dan Bernstein.

        Fix positioning error when a compositing, absolutely-positioned element has an ancestor with opacity.

        For absolutely-positioned layers, convertToLayerCoords() jumps to the positioned
        ancestor, since the layer x and y are relative to that ancestor. However, this could
        skip over the ancestorLayer, thus giving the wrong result.

        Fix by duplicating the enclosingPositionedAncestor() logic in convertToLayerCoords(),
        and checking for ancestorLayer along the way. If found, compute offset of both to
        the enclosingPositionedAncestor() and subtract.

        This also fixes a positioning bug with abs. positioned elements in reflections,
        so there is a new reflection test with a pixel result.

        Tests: compositing/geometry/abs-position-inside-opacity.html

        * rendering/RenderLayer.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47900 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoFall back to a Release version of Safari if a Debug one doesn't exist
aroben@apple.com [Mon, 31 Aug 2009 18:12:19 +0000 (18:12 +0000)]
Fall back to a Release version of Safari if a Debug one doesn't exist


Reviewed by Sam Weinig.

* Scripts/webkitdirs.pm:
(safariPath): If the user is working with a Debug build, but there's
no Debug version of Safari present, fall back to using a Release
version of Safari.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47899 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoMake safariPath() work for Debug builds of Safari on Windows
aroben@apple.com [Mon, 31 Aug 2009 17:54:41 +0000 (17:54 +0000)]
Make safariPath() work for Debug builds of Safari on Windows


Reviewed by Sam Weinig.

* Scripts/webkitdirs.pm:
(safariPath): If the user is working with a Debug build, add the
_debug suffix to Safari.exe.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47898 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years ago2009-08-31 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Mon, 31 Aug 2009 17:20:59 +0000 (17:20 +0000)]
2009-08-31  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Timothy Hatcher.

        WebInspector: pass call frames into the frontend as a part of
        pausedScript event.


        * inspector/InspectorController.cpp:
        * inspector/InspectorFrontend.cpp:
        * inspector/InspectorFrontend.h:
        * inspector/front-end/ScriptsPanel.js:
        * inspector/front-end/inspector.js:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47897 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoGive WebKit clients access to the window features specified in window.open
aroben@apple.com [Mon, 31 Aug 2009 15:06:38 +0000 (15:06 +0000)]
Give WebKit clients access to the window features specified in window.open

Reviewed by Steve Falkenburg.

* COMVariantSetter.h: Added COMVariantSetters for bool and float.

* Interfaces/IWebUIDelegatePrivate.idl: Added IWebUIDelegatePrivate2,
which contains a new version of createWebViewWithRequest that also
passes the window features as an IPropertyBag. This is analagous to
the -createWebViewWithRequest:windowFeatures: method added to
WebUIDelegate on Mac in r27452. Also added key strings for use with
the window features property bag.

* WebCoreSupport/WebChromeClient.cpp:
(createWindowFeaturesPropertyBag): Added this helper function to turn
a WebCore::WindowFeatures into an IPropertyBag for passing to WebKit
(WebChromeClient::createWindow): If the UI delegate implements
IWebUIDelegatePrivate2, call the new createWebViewWithRequest function
that takes a window features property bag. If the delegate doesn't
implement the new function, fall back to the old functions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47896 268f45cc-cd09-0410-ab3c-d52691b4dbfc

11 years agoA little clean-up in WebChromeClient::createWindow
aroben@apple.com [Mon, 31 Aug 2009 15:05:11 +0000 (15:05 +0000)]
A little clean-up in WebChromeClient::createWindow

Reviewed by Steve Falkenburg.

* WebCoreSupport/WebChromeClient.cpp:
(WebChromeClient::createWindow): Refactored to share more code between
the dialog and non-dialog cases.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@47895 268f45cc-cd09-0410-ab3c-d52691b4dbfc