WebKit-https.git
10 years ago2010-07-13 Andreas Kling <andreas.kling@nokia.com>
andreas.kling@nokia.com [Tue, 13 Jul 2010 11:53:05 +0000 (11:53 +0000)]
2010-07-13  Andreas Kling  <andreas.kling@nokia.com>

        Unreviewed. Adding myself as committer.

        * Scripts/webkitpy/common/config/committers.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63194 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Steve Block <steveblock@google.com>
steveblock@google.com [Tue, 13 Jul 2010 11:35:13 +0000 (11:35 +0000)]
2010-07-12  Steve Block  <steveblock@google.com>

        Reviewed by Jeremy Orlow.

        DeviceOrientationEvent should use optional properties
        https://bugs.webkit.org/show_bug.cgi?id=41607

        This change adds a new DeviceOrientation class which takes care of which of the
        optional properties are present. DeviceOrientationEvent owns an instance of
        DeviceOrientation, rather than owning the properties directly.
        DeviceOrientationEvent now requires custom bindings.

        Test: fast/dom/DeviceOrientation/optional-event-properties.html

        * Android.mk:
        * Android.jscbindings.mk:
        * Android.v8bindings.mk:
        * CMakeLists.txt:
        * GNUmakefile.am:
        * WebCore.gypi:
        * WebCore.pro:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * bindings/js/JSBindingsAllInOne.cpp:
        * bindings/js/JSDeviceOrientationEventCustom.cpp: Added.
        (WebCore::JSDeviceOrientationEvent::alpha):
        (WebCore::JSDeviceOrientationEvent::beta):
        (WebCore::JSDeviceOrientationEvent::gamma):
        (WebCore::JSDeviceOrientationEvent::initDeviceOrientationEvent):
        * bindings/v8/custom/V8DeviceOrientationEventCustom.cpp: Added.
        (WebCore::V8DeviceOrientationEvent::alphaAccessorGetter):
        (WebCore::V8DeviceOrientationEvent::betaAccessorGetter):
        (WebCore::V8DeviceOrientationEvent::gammaAccessorGetter):
        (WebCore::V8DeviceOrientationEvent::initDeviceOrientationEventCallback):
        * dom/DeviceOrientation.cpp: Added.
        * dom/DeviceOrientation.h: Added.
        (WebCore::DeviceOrientation::create):
        (WebCore::DeviceOrientation::canProvideAlpha):
        (WebCore::DeviceOrientation::alpha):
        (WebCore::DeviceOrientation::canProvideBeta):
        (WebCore::DeviceOrientation::beta):
        (WebCore::DeviceOrientation::canProvideGamma):
        (WebCore::DeviceOrientation::gamma):
        (WebCore::DeviceOrientation::DeviceOrientation):
        * dom/DeviceOrientationEvent.cpp:
        (WebCore::DeviceOrientationEvent::DeviceOrientationEvent):
        (WebCore::DeviceOrientationEvent::initDeviceOrientationEvent):
        * dom/DeviceOrientationEvent.h:
        (WebCore::DeviceOrientationEvent::create):
        (WebCore::DeviceOrientationEvent::orientation):
        * dom/DeviceOrientationEvent.idl:
2010-07-12  Steve Block  <steveblock@google.com>

        Reviewed by Jeremy Orlow.

        DeviceOrientationEvent should use optional properties
        https://bugs.webkit.org/show_bug.cgi?id=41607

        * fast/dom/DeviceOrientation/optional-event-properties-expected.txt: Added.
        * fast/dom/DeviceOrientation/optional-event-properties.html: Added.
        * fast/dom/DeviceOrientation/script-tests/optional-event-properties.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63193 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-13 Andreas Kling <andreas.kling@nokia.com>
andreas.kling@nokia.com [Tue, 13 Jul 2010 11:04:53 +0000 (11:04 +0000)]
2010-07-13  Andreas Kling  <andreas.kling@nokia.com>

        Reviewed by Antti Koivisto.

        CSS3 background: Number of layers should be determined by background-image element count
        https://bugs.webkit.org/show_bug.cgi?id=41201

        Change FillLayer culling logic to discard all layers
        after the first one without an image set.

        Manual test: css3-background-layer-count.html

        * manual-tests/css3-background-layer-count.html: Added.
        * rendering/style/FillLayer.cpp:
        (WebCore::FillLayer::cullEmptyLayers):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63192 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoPrevent assertion/duplicate loads for non-deferred subtitute-data loads
vestbo@webkit.org [Tue, 13 Jul 2010 10:47:04 +0000 (10:47 +0000)]
Prevent assertion/duplicate loads for non-deferred subtitute-data loads

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=30879

MainResourceLoader uses the member m_initialRequest to store requests for future
deferred loads. When doing the actual load in handleDataLoadNow(), we therefore
have to clear this request so that subsequent entries into the loader will not
start yet another load.

This can happen as a result of a PageGroupLoadDeferrer going out of scope when
returning from Chrome::runJavaScriptAlert(), which calls setDeferredLoading(false),
but only in the case of using both substitute-data and non-deferred main resource
load together. That's why two new DRT functions were added:

 * queueLoadHTMLString()
 * setDeferMainResourceLoad()

The change adds DRT hooks for Mac, Win and Qt for these two functions. For Mac
and Win the hook uses new SPI in WebDataSource. For Qt a new static member was
added to the FrameLoaderClientQt and accessed though DumpRenderTreeSupportQt.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63191 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-13 Yoshiki Hayashi <yhayashi@google.com>
commit-queue@webkit.org [Tue, 13 Jul 2010 10:25:28 +0000 (10:25 +0000)]
2010-07-13  Yoshiki Hayashi  <yhayashi@google.com>

        Reviewed by Kent Tamura.

        Make sure correct Nodes are passed to childrenChanged so that
        :last-child gets properly applied when fragment is inserted.
        https://bugs.webkit.org/show_bug.cgi?id=37944

        * fast/css/last-child-innerhtml-expected.txt: Added.
        * fast/css/last-child-innerhtml.html: Added.
2010-07-13  Yoshiki Hayashi  <yhayashi@google.com>

        Reviewed by Kent Tamura.

        Make sure correct Nodes are passed to childrenChanged so that
        :last-child gets properly applied when fragment is inserted.
        https://bugs.webkit.org/show_bug.cgi?id=37944

        Test: fast/css/last-child-innerhtml.html

        * dom/ContainerNode.cpp:
        (WebCore::ContainerNode::replaceChild):
        (WebCore::ContainerNode::appendChild):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63190 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-13 Ryosuke Niwa <rniwa@webkit.org>
rniwa@webkit.org [Tue, 13 Jul 2010 10:11:54 +0000 (10:11 +0000)]
2010-07-13  Ryosuke Niwa  <rniwa@webkit.org>

        Reviewed by Kent Tamura.

        InsertListCommand's modifyRange and doApply should be merged
        https://bugs.webkit.org/show_bug.cgi?id=41403

        Isolated the code in doApply to insert and remove lists for single paragraph into doApplyForSingleParagraph.
        Merged the code in modifyRange into doApply and cleaned up.

        No test is added since this is a clean up.

        * editing/InsertListCommand.cpp:
        (WebCore::InsertListCommand::InsertListCommand): m_forceCreateList is no longer initialized
        (WebCore::InsertListCommand::doApply): Isolated the code to insert/remove lists in doApplyForSingleParagraph
        (WebCore::InsertListCommand::doApplyForSingleParagraph): Insert/remove lists for single paragraph
        * editing/InsertListCommand.h: Added doApplyForSingleParagraph and removed m_forceCreateList

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63189 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-13 Philippe Normand <pnormand@igalia.com>
philn@webkit.org [Tue, 13 Jul 2010 10:05:42 +0000 (10:05 +0000)]
2010-07-13  Philippe Normand  <pnormand@igalia.com>

        Rubber-stamped by Xan Lopez.

        Updated the copyright headers of the files modified in r62476.

        * WebView/WebView.mm:
        * WebView/WebViewData.h:
        * WebView/WebViewInternal.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63188 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-13 Maciej Stachowiak <mjs@apple.com>
mjs@apple.com [Tue, 13 Jul 2010 10:02:52 +0000 (10:02 +0000)]
2010-07-13  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Nikolas Zimmermann.

        Activate test fonts for layout tests in WebKitTestRunner (on Mac)
        https://bugs.webkit.org/show_bug.cgi?id=42153

        * WebKitTestRunner/InjectedBundle/ActivateFonts.h: Added.
        * WebKitTestRunner/InjectedBundle/InjectedBundle.cpp:
        (WTR::InjectedBundle::initialize): Activate fonts on startup.
        * WebKitTestRunner/InjectedBundle/mac: Added.
        * WebKitTestRunner/InjectedBundle/mac/ActivateFonts.mm: Added.
        (WTR::activateFonts): Activate our fonts.
        * WebKitTestRunner/InjectedBundle/win: Added.
        * WebKitTestRunner/InjectedBundle/win/ActivateFonts.cpp: Added.
        (WTR::activateFonts): Dummy version.
        * WebKitTestRunner/WebKitTestRunner.xcodeproj/project.pbxproj: Copy fonts to bundle;
        add more files.
        * WebKitTestRunner/fonts: Added.
        * WebKitTestRunner/fonts/AHEM____.TTF: Added.
        * WebKitTestRunner/fonts/ColorBits-A.png: Added.
        * WebKitTestRunner/fonts/ColorBits.ttf: Added.
        * WebKitTestRunner/fonts/WebKit Layout Tests 2.ttf: Added.
        * WebKitTestRunner/fonts/WebKit Layout Tests.ttf: Added.
        * WebKitTestRunner/fonts/WebKitWeightWatcher100.ttf: Added.
        * WebKitTestRunner/fonts/WebKitWeightWatcher200.ttf: Added.
        * WebKitTestRunner/fonts/WebKitWeightWatcher300.ttf: Added.
        * WebKitTestRunner/fonts/WebKitWeightWatcher400.ttf: Added.
        * WebKitTestRunner/fonts/WebKitWeightWatcher500.ttf: Added.
        * WebKitTestRunner/fonts/WebKitWeightWatcher600.ttf: Added.
        * WebKitTestRunner/fonts/WebKitWeightWatcher700.ttf: Added.
        * WebKitTestRunner/fonts/WebKitWeightWatcher800.ttf: Added.
        * WebKitTestRunner/fonts/WebKitWeightWatcher900.ttf: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63187 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoAdd tests that require layoutTestController.layerTreeAsText to WebKit2 skipped list
mjs@apple.com [Tue, 13 Jul 2010 09:49:33 +0000 (09:49 +0000)]
Add tests that require layoutTestController.layerTreeAsText to WebKit2 skipped list
https://bugs.webkit.org/show_bug.cgi?id=42147

Reviewed by Steve Falkenburg.

* platform/mac-wk2/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63186 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago Unreviewed.
abecsi@webkit.org [Tue, 13 Jul 2010 09:39:06 +0000 (09:39 +0000)]
Unreviewed.

    Remove extra line added in r63183 "resources/whitespace01.dat: PASS"
    from html5lib/runner-expected.txt (which was probably added accidentally),
    because no bot produces this output in the tree.

* html5lib/runner-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63185 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoRemove check-header-includes, as it didn't turn out to be very useful.
dumi@chromium.org [Tue, 13 Jul 2010 08:26:40 +0000 (08:26 +0000)]
Remove check-header-includes, as it didn't turn out to be very useful.
https://bugs.webkit.org/show_bug.cgi?id=41970

Reviewed by Eric Seidel.

* Scripts/check-header-includes: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63184 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Eric Seidel <eric@webkit.org>
eric@webkit.org [Tue, 13 Jul 2010 08:23:02 +0000 (08:23 +0000)]
2010-07-12  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Fix typo in "close the cell" which caused assertion
        https://bugs.webkit.org/show_bug.cgi?id=42138

        * html5lib/resources/tables01.dat:
2010-07-12  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Fix typo in "close the cell" which caused assertion
        https://bugs.webkit.org/show_bug.cgi?id=42138

        * html/HTMLTreeBuilder.cpp:
        (WebCore::HTMLTreeBuilder::closeTheCell):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63183 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-13 Kent Tamura <tkent@chromium.org>
tkent@chromium.org [Tue, 13 Jul 2010 08:11:09 +0000 (08:11 +0000)]
2010-07-13  Kent Tamura  <tkent@chromium.org>

        Unreviewed, build fix.

        Fix a build error by r63179 for Chromium.

        * tests/PopupMenuTest.cpp:
        (WebKit::TestPopupMenuClient::itemLabel): Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63182 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoWindows build fix.
sfalken@apple.com [Tue, 13 Jul 2010 07:01:26 +0000 (07:01 +0000)]
Windows build fix.
Don't build MiniBrowser except for Debug_Internal.

* WebKit2.sln:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63181 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Eric Seidel <eric@webkit.org>
eric@webkit.org [Tue, 13 Jul 2010 06:53:16 +0000 (06:53 +0000)]
2010-07-12  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Fix line number handling in HTMLTreeBuilder to fix a zillion layout tests
        https://bugs.webkit.org/show_bug.cgi?id=42143

        Covered by a zillion layout tests.

        * html/HTMLTreeBuilder.cpp:
        (WebCore::HTMLTreeBuilder::processEndTag):
        (WebCore::HTMLTreeBuilder::processScriptStartTag):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63180 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Jay Civelli <jcivelli@chromium.org>
jcivelli@chromium.org [Tue, 13 Jul 2010 06:41:09 +0000 (06:41 +0000)]
2010-07-12  Jay Civelli  <jcivelli@chromium.org>

        Reviewed by Kent Tamura.

        Adding right aligned text called labels to PopupMenus.
        https://bugs.webkit.org/show_bug.cgi?id=41964

        * platform/PopupMenuClient.h:
        * platform/chromium/PopupMenuChromium.cpp:
        (WebCore::PopupListBox::paintRow):
        * rendering/RenderMenuList.cpp:
        (WebCore::RenderMenuList::itemLabel):
        * rendering/RenderMenuList.h:
        * rendering/RenderTextControlSingleLine.cpp:
        (WebCore::RenderTextControlSingleLine::itemLabel):
        * rendering/RenderTextControlSingleLine.h:
2010-07-12  Jay Civelli  <jcivelli@chromium.org>

        Reviewed by Kent Tamura.

        Making the autofill popup use labels to show the label
        associated with an item.
        https://bugs.webkit.org/show_bug.cgi?id=41964

        * src/AutoFillPopupMenuClient.cpp:
        (WebKit::AutoFillPopupMenuClient::getSuggestion):
        (WebKit::AutoFillPopupMenuClient::getLabel):
        (WebKit::AutoFillPopupMenuClient::itemLabel):
        (WebKit::AutoFillPopupMenuClient::setSuggestions):
        (WebKit::AutoFillPopupMenuClient::convertListIndexToInternalIndex):
        * src/AutoFillPopupMenuClient.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63179 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Steve Falkenburg <sfalken@apple.com>
sfalken@apple.com [Tue, 13 Jul 2010 06:32:34 +0000 (06:32 +0000)]
2010-07-12  Steve Falkenburg  <sfalken@apple.com>

        Reviewed by Maciej Stachowiak.

        Make WebKit2 be built by build-webkit (so it will be built by build.webkit.org bots)
        https://bugs.webkit.org/show_bug.cgi?id=40922

        Always build WebKit2 on Windows. Necessary since our WebKit build on
        Windows is packaged into a DLL with WebKit2.

        * Scripts/build-webkit: Always build WebKit2 on Windows.
2010-07-12  Steve Falkenburg  <sfalken@apple.com>

        Reviewed by Maciej Stachowiak.

        Make WebKit2 be built by build-webkit (so it will be built by build.webkit.org bots)
        https://bugs.webkit.org/show_bug.cgi?id=40922

        Add additional dependent projects. Necessary since our WebKit build on
        Windows is packaged into a DLL with WebKit2.

        * WebKit2.sln:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63178 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoOne http test timing out on Windows can cause all future tests to time out
bweinstein@apple.com [Tue, 13 Jul 2010 06:24:43 +0000 (06:24 +0000)]
One http test timing out on Windows can cause all future tests to time out
https://bugs.webkit.org/show_bug.cgi?id=42146

Reviewed by Steve Falkenburg.

There have been multiple times on the Windows bots where one http test timing out causes all future
http tests to time out. This is because the httpd process becomes unresponsive. When an http test
times out, we should restart httpd on Cygwin to prevent this.

* Scripts/old-run-webkit-tests: If an http test timed out on Cygwin, close httpd and restart it.
* Scripts/webkitperl/httpd.pm: Add additional logic to closeHTTPD to teach it about taskkill if
    using kill fails.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63177 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Martin Robinson <mrobinson@igalia.com>
mrobinson@webkit.org [Tue, 13 Jul 2010 06:21:10 +0000 (06:21 +0000)]
2010-07-12  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Dirk Schulze.

        [Cairo] Incorrect Test for Text Fill
        https://bugs.webkit.org/show_bug.cgi?id=42123

        Unskip tests which cover text stroking and add pixel dump results for
        them via the GTK+ port.

        * platform/gtk/Skipped: Unskip passing tests.
        * platform/gtk/fast/text/stroking-decorations-expected.checksum: Added.
        * platform/gtk/fast/text/stroking-decorations-expected.png: Added.
        * platform/gtk/fast/text/stroking-decorations-expected.txt: Added.
        * platform/gtk/fast/text/stroking-expected.checksum: Added.
        * platform/gtk/fast/text/stroking-expected.png: Added.
        * platform/gtk/fast/text/stroking-expected.txt: Added.
2010-07-12  Mihnea Ovidenea  <mihnea@adobe.com>

        Reviewed by Dirk Schulze.

        [Cairo] Incorrect Test for Text Fill
        https://bugs.webkit.org/show_bug.cgi?id=42123

        Correct test used for text shadow.  It currently checks for
        equality with cTextFill.  However, this test fails if the
        text drawing mode is set to 'cTextFill | cTextStroke'.  Fix
        is to modify the test to be like other Cairo uses of the
        text drawing mode, and check for the cTextFill bit being
        set, not for equality with the bit.

        * platform/graphics/cairo/FontCairo.cpp:
        (WebCore::Font::drawGlyphs):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63176 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Martin Robinson <mrobinson@igalia.com>
mrobinson@webkit.org [Tue, 13 Jul 2010 06:05:05 +0000 (06:05 +0000)]
2010-07-12  Martin Robinson  <mrobinson@igalia.com>

        Unreviewed.

        Fix a small error in the Skipped list.

        * platform/gtk/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63175 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoWindows build fix.
sfalken@apple.com [Tue, 13 Jul 2010 06:01:31 +0000 (06:01 +0000)]
Windows build fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63174 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Sheriff Bot <webkit.review.bot@gmail.com>
tkent@chromium.org [Tue, 13 Jul 2010 05:54:26 +0000 (05:54 +0000)]
2010-07-12  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r63171.
        http://trac.webkit.org/changeset/63171
        https://bugs.webkit.org/show_bug.cgi?id=42144

        Broke Chromium canary (Requested by tkent on #webkit).

        * WebKit.gyp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63173 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Tue, 13 Jul 2010 05:41:18 +0000 (05:41 +0000)]
2010-07-12  Pavel Feldman  <pfeldman@chromium.org>

        Not reviewed: rebaseline chromium tests.

        * platform/chromium-win-xp/fast/gradients/radial-centered-expected.checksum: Copied from LayoutTests/platform/chromium-win/fast/gradients/radial-centered-expected.checksum.
        * platform/chromium-win-xp/fast/gradients/radial-centered-expected.png: Copied from LayoutTests/platform/chromium-win/fast/gradients/radial-centered-expected.png.
        * platform/chromium-win-xp/fast/gradients/radial-centered-expected.txt: Copied from LayoutTests/platform/chromium-win/fast/gradients/radial-centered-expected.txt.
        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63172 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Kent Tamura <tkent@chromium.org>
tkent@chromium.org [Tue, 13 Jul 2010 05:37:31 +0000 (05:37 +0000)]
2010-07-12  Kent Tamura  <tkent@chromium.org>

        Reviewed by David Levin.

        [Chromium] Upstreaming inspector_resources target
        https://bugs.webkit.org/show_bug.cgi?id=41669

        'inspector_resource' is currently in src/webkit/webkit.gyp of
        Chromium tree. We move it to WebKit tree because:
        - DumpRenderTree in WebKit tree needs it, and
        - It's hard to refer to WebKit.gypi and WebCore.gypi from
          webkit_support in both cases of Chromium-tree build and
          WebKit-tree build.

        * WebKit.gyp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63171 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Viet-Trung Luu <viettrungluu@chromium.org>
commit-queue@webkit.org [Tue, 13 Jul 2010 05:30:28 +0000 (05:30 +0000)]
2010-07-12  Viet-Trung Luu  <viettrungluu@chromium.org>

        Reviewed by Darin Fisher.

        [Chromium] Pass key down events on to plugins.
        https://bugs.webkit.org/show_bug.cgi?id=42113

        * src/WebPluginContainerImpl.cpp:
        (WebKit::WebPluginContainerImpl::handleKeyboardEvent):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63170 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Maciej Stachowiak <mjs@apple.com>
mjs@apple.com [Tue, 13 Jul 2010 05:24:09 +0000 (05:24 +0000)]
2010-07-12  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by David Levin.

        Add failing animation and canvas tests to WebKit2 Skipped list
        https://bugs.webkit.org/show_bug.cgi?id=42141

        * platform/mac-wk2/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63169 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Eric Seidel <eric@webkit.org>
eric@webkit.org [Tue, 13 Jul 2010 05:10:10 +0000 (05:10 +0000)]
2010-07-12  Eric Seidel  <eric@webkit.org>

        Reviewed by Sam Weinig.

        Fix a typo in the adoption agency causing test failures
        https://bugs.webkit.org/show_bug.cgi?id=42133

        * html5lib/runner-expected-html5.txt:
2010-07-12  Eric Seidel  <eric@webkit.org>

        Reviewed by Sam Weinig.

        Fix a typo in the adoption agency causing test failures
        https://bugs.webkit.org/show_bug.cgi?id=42133

        The new behavior actually differs from old webkit.
        <p><b><p>TEST
        was not bold in the old parser, but is bold now.
        This matches Minefield and the HTML5 spec.

        Covered by two tests in html5lib/runner.html.

        * html/HTMLTreeBuilder.cpp:
        (WebCore::HTMLTreeBuilder::callTheAdoptionAgency):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63168 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoFix WebKit2 to use the same compiler-selection logic as the other projects.
mrowe@apple.com [Tue, 13 Jul 2010 04:27:12 +0000 (04:27 +0000)]
Fix WebKit2 to use the same compiler-selection logic as the other projects.

Rubber-stamped by Maciej Stachowiak.

* WebKit2.xcodeproj/project.pbxproj: Don't hard-code WebKit2 to build with GCC 4.2.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63167 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Tue, 13 Jul 2010 04:04:22 +0000 (04:04 +0000)]
2010-07-12  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Joseph Pecoraro.

        Web Inspector: provide starts and ends for network phases instead of duration.

        https://bugs.webkit.org/show_bug.cgi?id=42091

        * inspector/InspectorResource.cpp:
        (WebCore::InspectorResource::updateResponse):
        (WebCore::InspectorResource::updateScriptObject):
        (WebCore::InspectorResource::buildObjectForTiming):
        * platform/network/ResourceLoadTiming.h:
        (WebCore::ResourceLoadTiming::deepCopy):
        (WebCore::ResourceLoadTiming::operator==):
        (WebCore::ResourceLoadTiming::ResourceLoadTiming):
        * platform/network/ResourceResponseBase.cpp:
        (WebCore::ResourceResponseBase::ResourceResponseBase):
        (WebCore::ResourceResponseBase::wasCached):
        (WebCore::ResourceResponseBase::setWasCached):
        * platform/network/ResourceResponseBase.h:
2010-07-12  Pavel Feldman  <pfeldman@chromium.org>

        Reviewed by Joseph Pecoraro.

        Web Inspector: provide starts and ends for network phases instead of duration.

        https://bugs.webkit.org/show_bug.cgi?id=42091

        * public/WebURLLoadTiming.h:
        * public/WebURLResponse.h:
        * src/WebURLLoadTiming.cpp:
        (WebKit::WebURLLoadTiming::proxyStart):
        (WebKit::WebURLLoadTiming::setProxyStart):
        (WebKit::WebURLLoadTiming::proxyEnd):
        (WebKit::WebURLLoadTiming::setProxyEnd):
        (WebKit::WebURLLoadTiming::dnsStart):
        (WebKit::WebURLLoadTiming::setDNSStart):
        (WebKit::WebURLLoadTiming::dnsEnd):
        (WebKit::WebURLLoadTiming::setDNSEnd):
        (WebKit::WebURLLoadTiming::connectStart):
        (WebKit::WebURLLoadTiming::setConnectStart):
        (WebKit::WebURLLoadTiming::connectEnd):
        (WebKit::WebURLLoadTiming::setConnectEnd):
        (WebKit::WebURLLoadTiming::sendStart):
        (WebKit::WebURLLoadTiming::setSendStart):
        (WebKit::WebURLLoadTiming::sendEnd):
        (WebKit::WebURLLoadTiming::setSendEnd):
        (WebKit::WebURLLoadTiming::receiveHeadersStart):
        (WebKit::WebURLLoadTiming::setReceiveHeadersStart):
        (WebKit::WebURLLoadTiming::receiveHeadersEnd):
        (WebKit::WebURLLoadTiming::setReceiveHeadersEnd):
        (WebKit::WebURLLoadTiming::sslStart):
        (WebKit::WebURLLoadTiming::setSSLStart):
        (WebKit::WebURLLoadTiming::sslEnd):
        (WebKit::WebURLLoadTiming::setSSLEnd):
        * src/WebURLResponse.cpp:
        (WebKit::WebURLResponse::wasCached):
        (WebKit::WebURLResponse::setWasCached):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63166 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Tue, 13 Jul 2010 03:41:11 +0000 (03:41 +0000)]
2010-07-12  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        HTML5 Parser: document.write after onload blows away document
        https://bugs.webkit.org/show_bug.cgi?id=40745

        Update test to expect new behavior.

        * http/tests/misc/write-while-waiting.html:
2010-07-12  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        HTML5 Parser: document.write after onload blows away document
        https://bugs.webkit.org/show_bug.cgi?id=40745

        Rather than blowing away the document when we get a document.write call
        after the document is closed, we new ignore the write.  This
        technically violates the spec (which requires us to blow away the
        document), but blowing away the document breaks too many web sites.

        Rather than this patch, we could go back to our old behavior (which was
        to append the bytes just before EOF), but implementing this approach
        (suggested by Henri) will let us gather data about whether his approach
        is workable.

        See also: http://www.w3.org/Bugs/Public/show_bug.cgi?id=9767

        * dom/Document.cpp:
        (WebCore::Document::write):
        * html/HTMLDocumentParser.cpp:
        (WebCore::HTMLDocumentParser::insert):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63165 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Maciej Stachowiak <mjs@apple.com>
mjs@apple.com [Tue, 13 Jul 2010 03:39:08 +0000 (03:39 +0000)]
2010-07-12  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Dan Bernstein.

        Remove use of auto_ptr in WebKitTestRunner
        https://bugs.webkit.org/show_bug.cgi?id=42134

        Replaced auto_ptr with PassOwnPtr / OwnPtr.

        * WebKitTestRunner/InjectedBundle/InjectedBundlePage.cpp:
        (WTR::WKStringToUTF8):
        (WTR::InjectedBundlePage::dump):
        (WTR::InjectedBundlePage::addMessageToConsole):
        * WebKitTestRunner/TestInvocation.cpp:
        (WTR::WKStringToUTF8):
        (WTR::TestInvocation::didRecieveMessageFromInjectedBundle):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63164 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoNeed support for a WebKit2-specific Skipped list (and initially add accessibility...
mjs@apple.com [Tue, 13 Jul 2010 03:13:04 +0000 (03:13 +0000)]
Need support for a WebKit2-specific Skipped list (and initially add accessibility tests to it)
https://bugs.webkit.org/show_bug.cgi?id=42132

Reviewed by Sam Weinig.

WebKitTools:

* Scripts/old-run-webkit-tests: In --webkit-test-runner mode, add mac-wk2 to list of
platform directories.

LayoutTests:

* platform/mac-wk2: Added.
* platform/mac-wk2/Skipped: Added. Initially add the accessibility
directory (with two blocking bugs).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63163 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Tony Gentilcore <tonyg@chromium.org>
tonyg@chromium.org [Tue, 13 Jul 2010 02:57:46 +0000 (02:57 +0000)]
2010-07-12  Tony Gentilcore  <tonyg@chromium.org>

        Reviewed by Oliver Hunt.

        Strip BOMs from source before passing to V8
        https://bugs.webkit.org/show_bug.cgi?id=42102

        This extra copy may carry a performance penalty. We should investigate
        whether this allows any simplification in v8/scanner.cc:SkipJavaScriptWhiteSpace().

        No new tests because no new functionality.

        * bindings/v8/ScriptSourceCode.h:
        (WebCore::ScriptSourceCode::ScriptSourceCode):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63162 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Tue, 13 Jul 2010 02:45:28 +0000 (02:45 +0000)]
2010-07-12  Gustavo Noronha Silva  <gns@gnome.org>

        Unreviewed. make distcheck fix.

        * GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63161 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Sheriff Bot <webkit.review.bot@gmail.com>
commit-queue@webkit.org [Tue, 13 Jul 2010 02:45:26 +0000 (02:45 +0000)]
2010-07-12  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r63156.
        http://trac.webkit.org/changeset/63156
        https://bugs.webkit.org/show_bug.cgi?id=42129

        broke chromium compile (Requested by jamesr on #webkit).

        * WebKit.gyp:
        * public/WebAccessibilityObject.h:
        * public/WebAttribute.h: Removed.
        * public/WebDocumentType.h: Removed.
        * public/WebElement.h:
        * public/WebNamedNodeMap.h: Removed.
        * public/WebNode.h:
        * public/WebRenderStyle.h: Removed.
        * src/WebAccessibilityObject.cpp:
        * src/WebAttribute.cpp: Removed.
        * src/WebDocumentType.cpp: Removed.
        * src/WebElement.cpp:
        * src/WebNamedNodeMap.cpp: Removed.
        * src/WebNode.cpp:
        * src/WebRenderStyle.cpp: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63160 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Yuta Kitamura <yutak@chromium.org>
commit-queue@webkit.org [Tue, 13 Jul 2010 02:42:38 +0000 (02:42 +0000)]
2010-07-12  Yuta Kitamura  <yutak@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Fix crash caused by unintentional deletion of worker bridge and channel.

        WebSocket: Crash caused by calling close() within onmessage handler
        https://bugs.webkit.org/show_bug.cgi?id=41507

        * websocket/tests/workers/close-in-onmessage-crash-expected.txt: Added.
        * websocket/tests/workers/close-in-onmessage-crash.html: Added.
        * websocket/tests/workers/resources/close-in-onmessage-crash.js: Added.
        (runTests.try.ws.onopen):
        (runTests.try.ws.onmessage): Calling close() inside the onmessage handler
        causes the crash.
        (runTests.try.ws.onclose):
        (runTests):
2010-07-12  Yuta Kitamura  <yutak@chromium.org>

        Reviewed by Alexey Proskuryakov.

        Fix crash caused by unintentional deletion of worker bridge and channel.

        WebSocket: Crash caused by calling close() within onmessage handler
        https://bugs.webkit.org/show_bug.cgi?id=41507

        Test: websocket/tests/workers/close-in-onmessage-crash.html

        * websockets/WebSocket.cpp:
        (WebCore::WebSocket::close): bufferedAmount() may call WebSocket::didClose(),
        which causes m_channel to get freed.
        * websockets/WorkerThreadableWebSocketChannel.cpp:
        (WebCore::WorkerThreadableWebSocketChannel::Bridge::send): Add reference to
        the bridge because waitForMethodCompletion() may dereference the bridge.
        (WebCore::WorkerThreadableWebSocketChannel::Bridge::bufferedAmount): Ditto.
        (WebCore::WorkerThreadableWebSocketChannel::Bridge::waitForMethodCompletion):
        The root cause is a call to WorkerRunLoop::runInMode in this function.
        It may call WebSocket::didClose() even inside WebSocket::close(), which frees
        everything including the worker bridge and the channel.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63159 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Eric Seidel <eric@webkit.org>
eric@webkit.org [Tue, 13 Jul 2010 02:05:38 +0000 (02:05 +0000)]
2010-07-12  Eric Seidel  <eric@webkit.org>

        Unreviewed.  Just skipping a test on Gtk due to missing DRT features.

        REGRESSION (HTML5 parser?): Impossible to get past the CAPTCHA on postcode.royalmail.com
        https://bugs.webkit.org/show_bug.cgi?id=41797

        * platform/gtk/Skipped:
         - Gtk is missing DRT api.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63158 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Maciej Stachowiak <mjs@apple.com>
mjs@apple.com [Tue, 13 Jul 2010 01:58:45 +0000 (01:58 +0000)]
2010-07-12  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Anders Carlsson.

        WKView should override setFrameSize: instead of setFrame:
        https://bugs.webkit.org/show_bug.cgi?id=42127

        * UIProcess/API/mac/WKView.mm:
        (-[WKView setFrameSize:]): Override this method instead of setFrame:,
        since that is the right way to do it.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63157 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Dominic Mazzoni <dmazzoni@google.com>
commit-queue@webkit.org [Tue, 13 Jul 2010 01:45:47 +0000 (01:45 +0000)]
2010-07-12  Dominic Mazzoni  <dmazzoni@google.com>

        Reviewed by Darin Fisher.

        Add methods to WebAccessibilityObject to access the Node and
        Document, and then add methods to access some document, node,
        and element properties needed for Windows accessibility.

        https://bugs.webkit.org/show_bug.cgi?id=41569

        * WebKit.gyp:
        * public/WebAccessibilityObject.h:
        * public/WebAttribute.h: Added.
        (WebKit::WebAttribute::~WebAttribute):
        (WebKit::WebAttribute::WebAttribute):
        (WebKit::WebAttribute::operator=):
        * public/WebDocumentType.h: Added.
        (WebKit::WebDocumentType::WebDocumentType):
        (WebKit::WebDocumentType::operator=):
        (WebKit::WebDocumentType::assign):
        * public/WebElement.h:
        * public/WebNamedNodeMap.h: Added.
        (WebKit::WebNamedNodeMap::~WebNamedNodeMap):
        (WebKit::WebNamedNodeMap::WebNamedNodeMap):
        (WebKit::WebNamedNodeMap::operator=):
        * public/WebNode.h:
        * public/WebRenderStyle.h: Added.
        (WebKit::WebRenderStyle::~WebRenderStyle):
        (WebKit::WebRenderStyle::WebRenderStyle):
        (WebKit::WebRenderStyle::operator=):
        * src/WebAccessibilityObject.cpp:
        (WebKit::WebAccessibilityObject::node):
        (WebKit::WebAccessibilityObject::document):
        * src/WebAttribute.cpp: Added.
        (WebKit::WebAttribute::reset):
        (WebKit::WebAttribute::assign):
        (WebKit::WebAttribute::WebAttribute):
        (WebKit::WebAttribute::localName):
        (WebKit::WebAttribute::value):
        * src/WebDocumentType.cpp: Added.
        (WebKit::WebDocumentType::name):
        (WebKit::WebDocumentType::WebDocumentType):
        (WebKit::WebDocumentType::operator=):
        (WebKit::WebDocumentType::operator PassRefPtr<DocumentType>):
        * src/WebElement.cpp:
        (WebKit::WebElement::attributes):
        * src/WebNamedNodeMap.cpp: Added.
        (WebKit::WebNamedNodeMap::reset):
        (WebKit::WebNamedNodeMap::assign):
        (WebKit::WebNamedNodeMap::WebNamedNodeMap):
        (WebKit::WebNamedNodeMap::length):
        (WebKit::WebNamedNodeMap::attributeItem):
        * src/WebNode.cpp:
        (WebKit::WebNode::element):
        (WebKit::WebNode::hasComputedStyle):
        (WebKit::WebNode::computedStyle):
        * src/WebRenderStyle.cpp: Added.
        (WebKit::WebRenderStyle::reset):
        (WebKit::WebRenderStyle::assign):
        (WebKit::WebRenderStyle::WebRenderStyle):
        (WebKit::WebRenderStyle::display):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63156 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Maciej Stachowiak <mjs@apple.com>
mjs@apple.com [Tue, 13 Jul 2010 01:36:57 +0000 (01:36 +0000)]
2010-07-12  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Anders Carlsson.

        Make WebKitTestRunner resize the view specially for the W3C SVG tests.
        https://bugs.webkit.org/show_bug.cgi?id=42126

        * WebKitTestRunner/PlatformWebView.h:
        * WebKitTestRunner/TestInvocation.cpp:
        (WTR::TestInvocation::TestInvocation): Store the path as a C
        string for later use.
        (WTR::TestInvocation::~TestInvocation):
        (WTR::sizeWebViewForCurrentTest): Resize the WebView to the proper
        size, depending on whether this is a W3C SVG test.
        (WTR::TestInvocation::invoke): Call the size function.
        * WebKitTestRunner/TestInvocation.h:
        * WebKitTestRunner/mac/PlatformWebViewMac.mm:
        (WTR::PlatformWebView::resizeTo): Implement.
        * WebKitTestRunner/win/PlatformWebViewWin.cpp:
        (WTR::PlatformWebView::resizeTo): Placeholder.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63155 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Tue, 13 Jul 2010 01:28:13 +0000 (01:28 +0000)]
2010-07-12  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        REGRESSION (HTML5 parser?): Impossible to get past the CAPTCHA on postcode.royalmail.com
        https://bugs.webkit.org/show_bug.cgi?id=41797

        This test is a bit subtle, but it captures a reduction of what I think
        is going on with the site.

        Essentially, the real parser doesn't see the <img> tag, but the preload
        scanner does.  Prior to this patch, we were preload scanning this
        region of the document even though we weren't supposed to.

        * http/tests/loading/preload-slow-loading-expected.txt: Added.
        * http/tests/loading/preload-slow-loading.php: Added.
2010-07-12  Adam Barth  <abarth@webkit.org>

        Reviewed by Eric Seidel.

        REGRESSION (HTML5 parser?): Impossible to get past the CAPTCHA on postcode.royalmail.com
        https://bugs.webkit.org/show_bug.cgi?id=41797

        Once we resume parsing after script execution, we want to clear the
        preload scanner so we don't scan any bytes it might have accumulated.

        Test: http/tests/loading/preload-slow-loading.php

        * html/HTMLDocumentParser.cpp:
        (WebCore::HTMLDocumentParser::resumeParsingAfterScriptExecution):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63154 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoReviewed by Jian Li.
ajwong@chromium.org [Tue, 13 Jul 2010 01:15:06 +0000 (01:15 +0000)]
Reviewed by Jian Li.

Adding directional property enums back into
CSSComputedStyleDeclaration::getPropertyCSSValue().

https://bugs.webkit.org/show_bug.cgi?id=42122

The four directional -webkit- properties --  CSSPropertyWebkitMarginEnd,
CSSPropertyWebkitMarginStart, CSSPropertyWebkitPaddingEnd, and
CSSPropertyWebkitPaddingStart -- are resolved into other css
properties before the swtich statement via a call to
CSSProperty::resolveDirectionAwareProperty().  Thus, they are never
seen by the switch statement.  However, if you leave out a potential
enum value from the switch, gcc will generate a warning if -Wall is
specified.  This warning breaks the Chromium build.  To avoid this,
we add in the enum values and ASSERT_NOT_REACHED() if they are hit.

* css/CSSComputedStyleDeclaration.cpp:
(WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63153 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoWebKit/win: <rdar://problem/8113038> WebKit1 and WebKit2 should build as a single DLL
sfalken@apple.com [Tue, 13 Jul 2010 00:54:16 +0000 (00:54 +0000)]
WebKit/win: <rdar://problem/8113038> WebKit1 and WebKit2 should build as a single DLL
https://bugs.webkit.org/show_bug.cgi?id=40921

Reviewed by Alice Liu.

Pre-WebKit2 WebKit now builds into a static library named WebKitLib.lib.
WebKit.dll now links in WebCore.lib, WebKitLib.lib and WebKit2 code.

This is a first step. We'll likely want to migrate the remainder of
the non-deprecated WebKit code (strings, DLLMain, resources) into WebKit2.

* WebKit.vcproj/WebKit.def: Removed.
* WebKit.vcproj/WebKit.make: Don't fail if WebKit doesn't produce a DLL.
* WebKit.vcproj/WebKit.vcproj: Build WebKit as a static lib.
Renamed project name to WebKitLib to avoid confusion and naming conflicts.
Generate intermediate pdb file for debuggability of static lib.
Removed DLL-related options.
* WebKit.vcproj/deleteButton.png: Removed.
* WebKit.vcproj/deleteButtonPressed.png: Removed.
* WebKit.vcproj/fsVideoAudioVolumeHigh.png: Removed.
* WebKit.vcproj/fsVideoAudioVolumeLow.png: Removed.
* WebKit.vcproj/fsVideoExitFullscreen.png: Removed.
* WebKit.vcproj/fsVideoPause.png: Removed.
* WebKit.vcproj/fsVideoPlay.png: Removed.
* WebKit.vcproj/missingImage.png: Removed.
* WebKit.vcproj/nullplugin.png: Removed.
* WebKit.vcproj/panEastCursor.png: Removed.
* WebKit.vcproj/panIcon.png: Removed.
* WebKit.vcproj/panNorthCursor.png: Removed.
* WebKit.vcproj/panNorthEastCursor.png: Removed.
* WebKit.vcproj/panNorthWestCursor.png: Removed.
* WebKit.vcproj/panSouthCursor.png: Removed.
* WebKit.vcproj/panSouthEastCursor.png: Removed.
* WebKit.vcproj/panSouthWestCursor.png: Removed.
* WebKit.vcproj/panWestCursor.png: Removed.
* WebKit.vcproj/searchCancel.png: Removed.
* WebKit.vcproj/searchCancelPressed.png: Removed.
* WebKit.vcproj/searchMagnifier.png: Removed.
* WebKit.vcproj/searchMagnifierResults.png: Removed.
* WebKit.vcproj/textAreaResizeCorner.png: Removed.
* WebKit.vcproj/verticalTextCursor.png: Removed.
* WebKit.vcproj/zoomInCursor.png: Removed.
* WebKit.vcproj/zoomOutCursor.png: Removed.

WebKit2: <rdar://problem/8113038> WebKit1 and WebKit2 should build as a single DLL
https://bugs.webkit.org/show_bug.cgi?id=40921

Reviewed by Alice Liu.

Pre-WebKit2 WebKit now builds into a static library named WebKitLib.lib.
WebKit.dll now links in WebCore.lib, WebKitLib.lib and WebKit2 code.

This is a first step. We'll likely want to migrate the remainder of
the non-deprecated WebKit code (strings, DLLMain, resources) into WebKit2.

* UIProcess/API/C/WKBase.h: Updated to new BUILDING_ name.
* WebProcess/InjectedBundle/API/c/WKBundleBase.h: Updated to new BUILDING_ name.
* WebProcess/WebCoreSupport/win/WebCoreLocalizedStrings.cpp: Removed.
* WebProcess/win/DllMain.cpp: Removed. Overlaps with implementation in WebKitLib.lib.
* WebProcess/win/WebLocalizableStrings.cpp: Removed. Overlaps with implementation in WebKitLib.lib.
* WebProcess/win/WebLocalizableStrings.h: Removed. Overlaps with implementation in WebKitLib.lib.
* win/WebKit2.def: Added. Copied from WebKit project.
* win/WebKit2.rc: Added resources previously in WebKit.
* win/WebKit2.vcproj: Changed project name to WebKit so we will link output to WebKit.dll.
Removed implementations overlapping with WebKitLib (WebCoreLocalizedStrings, DllMain, WebLocalizableStrings, WebProcessMain).
* win/WebKit2Common.vsprops: Use a framework name of WebKit instead of WebKit2 to reflect project name.
Renamed BUILDING_WEBKIT2 to BUILDING_WEBKIT since there is now just a single WebKit.dll.
* win/WebKit2WebProcess.vcproj: Link against WebKit instead of WebKit2 due to renaming.
* win/deleteButton.png: Copied from ../WebKit/win/WebKit.vcproj/deleteButton.png.
* win/deleteButtonPressed.png: Copied from ../WebKit/win/WebKit.vcproj/deleteButtonPressed.png.
* win/fsVideoAudioVolumeHigh.png: Copied from ../WebKit/win/WebKit.vcproj/fsVideoAudioVolumeHigh.png.
* win/fsVideoAudioVolumeLow.png: Copied from ../WebKit/win/WebKit.vcproj/fsVideoAudioVolumeLow.png.
* win/fsVideoExitFullscreen.png: Copied from ../WebKit/win/WebKit.vcproj/fsVideoExitFullscreen.png.
* win/fsVideoPause.png: Copied from ../WebKit/win/WebKit.vcproj/fsVideoPause.png.
* win/fsVideoPlay.png: Copied from ../WebKit/win/WebKit.vcproj/fsVideoPlay.png.
* win/missingImage.png: Copied from ../WebKit/win/WebKit.vcproj/missingImage.png.
* win/nullplugin.png: Copied from ../WebKit/win/WebKit.vcproj/nullplugin.png.
* win/panEastCursor.png: Copied from ../WebKit/win/WebKit.vcproj/panEastCursor.png.
* win/panIcon.png: Copied from ../WebKit/win/WebKit.vcproj/panIcon.png.
* win/panNorthCursor.png: Copied from ../WebKit/win/WebKit.vcproj/panNorthCursor.png.
* win/panNorthEastCursor.png: Copied from ../WebKit/win/WebKit.vcproj/panNorthEastCursor.png.
* win/panNorthWestCursor.png: Copied from ../WebKit/win/WebKit.vcproj/panNorthWestCursor.png.
* win/panSouthCursor.png: Copied from ../WebKit/win/WebKit.vcproj/panSouthCursor.png.
* win/panSouthEastCursor.png: Copied from ../WebKit/win/WebKit.vcproj/panSouthEastCursor.png.
* win/panSouthWestCursor.png: Copied from ../WebKit/win/WebKit.vcproj/panSouthWestCursor.png.
* win/panWestCursor.png: Copied from ../WebKit/win/WebKit.vcproj/panWestCursor.png.
* win/searchCancel.png: Copied from ../WebKit/win/WebKit.vcproj/searchCancel.png.
* win/searchCancelPressed.png: Copied from ../WebKit/win/WebKit.vcproj/searchCancelPressed.png.
* win/searchMagnifier.png: Copied from ../WebKit/win/WebKit.vcproj/searchMagnifier.png.
* win/searchMagnifierResults.png: Copied from ../WebKit/win/WebKit.vcproj/searchMagnifierResults.png.
* win/textAreaResizeCorner.png: Copied from ../WebKit/win/WebKit.vcproj/textAreaResizeCorner.png.
* win/verticalTextCursor.png: Copied from ../WebKit/win/WebKit.vcproj/verticalTextCursor.png.
* win/zoomInCursor.png: Copied from ../WebKit/win/WebKit.vcproj/zoomInCursor.png.
* win/zoomOutCursor.png: Copied from ../WebKit/win/WebKit.vcproj/zoomOutCursor.png.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63151 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Tue, 13 Jul 2010 00:40:47 +0000 (00:40 +0000)]
2010-07-12  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dan Bernstein.

        Don't go into compositing mode for 0x0 plugins
        https://bugs.webkit.org/show_bug.cgi?id=34009

        Don't use compositing for small (0-height or width, or 1x1) plugins, or iframes whose
        height or width is zero.

        Previously we made all compositing decisions inside styleChanged(). However,
        now that plugin and iframe compositing behavior depends on renderer size, we have
        to wait until layout before deciding whether to composite these. This behavior
        change is controlled by the m_compositingDependsOnGeometry flag. When set,
        updateCompositingLayers() always does a hierarchy update.

        Tests: compositing/iframes/iframe-size-from-zero.html
               compositing/iframes/iframe-size-to-zero.html
               compositing/plugins/1x1-composited-plugin.html
               compositing/plugins/large-to-small-composited-plugin.html
               compositing/plugins/small-to-large-composited-plugin.html

        * page/FrameView.cpp:
        (WebCore::FrameView::updateCompositingLayers): No longer bail if usesCompositing() is false; we
        have to always enter updateCompositingLayers().
        (WebCore::FrameView::repaintFixedElementsAfterScrolling): Ditto
        (WebCore::FrameView::enterCompositingMode): Remove bogus return of a void.

        * rendering/RenderLayerCompositor.h: Add m_compositingDependsOnGeometry.
        * rendering/RenderLayerCompositor.cpp:
        (WebCore::RenderLayerCompositor::RenderLayerCompositor): Initialize m_compositingDependsOnGeometry to false.
        (WebCore::RenderLayerCompositor::updateCompositingLayers): If m_compositingDependsOnGeometry is true,
        we always need to run through the layer hierarchy looking for things which need to be composited, even if
        we're not (yet) in compositing mode.

        (WebCore::RenderLayerCompositor::computeCompositingRequirements): Because we can enter compositing mode
        on the fly inside updateCompositingLayers() now, the state of willBeComposited needs to be updated
        when processing the root layer, for the case where the compositing mode changes.

        (WebCore::RenderLayerCompositor::requiresCompositingForPlugin): Set the m_compositingDependsOnGeometry
        flag if we see a potentially-composited plugin. Once we have layout information, only composite the plugin
        if height * width > 1.

        (WebCore::RenderLayerCompositor::requiresCompositingForIFrame): Set the m_compositingDependsOnGeometry
        flag if we see a potentially-composited iframe. Once we have layout information, only composite the plugin
        if height or width is greater than zero.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63149 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Maciej Stachowiak <mjs@apple.com>
mjs@apple.com [Tue, 13 Jul 2010 00:37:51 +0000 (00:37 +0000)]
2010-07-12  Maciej Stachowiak  <mjs@apple.com>

        Reviewed by Anders Carlsson.

        Get rid of auto_ptr use in WebKit2
        https://bugs.webkit.org/show_bug.cgi?id=42119

        Replace all use of auto_ptr with OwnPtr/PassOwnPtr.

        * Platform/CoreIPC/Connection.cpp:
        (CoreIPC::Connection::sendMessage):
        (CoreIPC::Connection::waitForMessage):
        (CoreIPC::Connection::sendSyncMessage):
        (CoreIPC::Connection::processIncomingMessage):
        (CoreIPC::Connection::sendOutgoingMessages):
        (CoreIPC::Connection::dispatchMessages):
        * Platform/CoreIPC/Connection.h:
        (CoreIPC::Connection::Message::Message):
        (CoreIPC::Connection::send):
        (CoreIPC::Connection::sendSync):
        (CoreIPC::Connection::waitFor):
        * Platform/CoreIPC/mac/ConnectionMac.cpp:
        (CoreIPC::Connection::sendOutgoingMessage):
        (CoreIPC::createArgumentDecoder):
        (CoreIPC::Connection::receiveSourceEventHandler):
        * Platform/CoreIPC/qt/ConnectionQt.cpp:
        (CoreIPC::Connection::readyReadHandler):
        (CoreIPC::Connection::sendOutgoingMessage):
        * Platform/CoreIPC/win/ConnectionWin.cpp:
        (CoreIPC::Connection::readEventHandler):
        (CoreIPC::Connection::sendOutgoingMessage):
        * Platform/RunLoop.cpp:
        (RunLoop::performWork):
        (RunLoop::scheduleWork):
        * Platform/RunLoop.h:
        * Platform/WorkItem.h:
        (WorkItem::create):
        * Platform/WorkQueue.h:
        * Platform/mac/WorkQueueMac.cpp:
        (WorkQueue::executeWorkItem):
        (WorkQueue::scheduleWork):
        (WorkQueue::EventSource::EventSource):
        (WorkQueue::registerMachPortEventHandler):
        * Platform/qt/WorkQueueQt.cpp:
        (WorkQueue::connectSignal):
        (WorkQueue::scheduleWork):
        * Platform/win/WorkQueueWin.cpp:
        (WorkQueue::registerHandle):
        (WorkQueue::scheduleWork):
        (WorkQueue::performWork):
        * UIProcess/ChunkedUpdateDrawingAreaProxy.cpp:
        (WebKit::ChunkedUpdateDrawingAreaProxy::paint):
        * UIProcess/WebProcessProxy.cpp:
        (WebKit::WebProcessProxy::sendMessage):
        (WebKit::WebProcessProxy::didFinishLaunching):
        * UIProcess/WebProcessProxy.h:
        (WebKit::WebProcessProxy::send):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63148 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoUnreviewed, build fix.
ajwong@chromium.org [Tue, 13 Jul 2010 00:37:35 +0000 (00:37 +0000)]
Unreviewed, build fix.

[chromium] rebaseline new test expectations.

* platform/chromium-mac/media/video-source-expected.txt:
* platform/chromium-mac/media/video-src-expected.txt:
* platform/chromium-win/media/video-source-expected.txt:
* platform/chromium-win/media/video-src-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63147 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoAX converts AtomicStrings to .string() more than needed
cfleizach@apple.com [Mon, 12 Jul 2010 23:53:17 +0000 (23:53 +0000)]
AX converts AtomicStrings to .string() more than needed
https://bugs.webkit.org/show_bug.cgi?id=42056

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63139 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Tony Chang <tony@chromium.org>
tony@chromium.org [Mon, 12 Jul 2010 23:38:39 +0000 (23:38 +0000)]
2010-07-12  Tony Chang  <tony@chromium.org>

        Reviewed by David Hyatt.

        crash in FrameView::detachCustomScrollbars
        https://bugs.webkit.org/show_bug.cgi?id=41196

        * scrollbars/hidden-iframe-scrollbar-crash-expected.txt: Added.
        * scrollbars/hidden-iframe-scrollbar-crash.html: Added.
2010-07-12  Tony Chang  <tony@chromium.org>

        Reviewed by David Hyatt.

        crash in FrameView::detachCustomScrollbars
        https://bugs.webkit.org/show_bug.cgi?id=41196

        Test: scrollbars/hidden-iframe-scrollbar-crash.html

        * page/FrameView.cpp:
        (WebCore::FrameView::detachCustomScrollbars):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63138 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Eric Seidel <eric@webkit.org>
eric@webkit.org [Mon, 12 Jul 2010 23:16:54 +0000 (23:16 +0000)]
2010-07-12  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        HTML tags should break out of foreign content
        https://bugs.webkit.org/show_bug.cgi?id=42106

        * html5lib/runner-expected-html5.txt:
2010-07-12  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        HTML tags should break out of foreign content
        https://bugs.webkit.org/show_bug.cgi?id=42106

        Implement another paragraph of the spec to pass another
        bunch of foreign content tests.

        This fixes a bunch of tests in html5lib/runner.html.
        After this change we only have 4 remaining foreign content failures.

        * html/HTMLTreeBuilder.cpp:
        (WebCore::HTMLTreeBuilder::processStartTag):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63135 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Zhenyao Mo <zmo@google.com>
zmo@google.com [Mon, 12 Jul 2010 23:13:18 +0000 (23:13 +0000)]
2010-07-12  Zhenyao Mo  <zmo@google.com>

        Reviewed by Darin Fisher.

        Bring bufferData and clear to GLES2 compliant
        https://bugs.webkit.org/show_bug.cgi?id=41574

        * fast/canvas/webgl/invalidPassedParams-expected.txt: Add wrong param test cases for bufferData and clear.
        * fast/canvas/webgl/invalidPassedParams.html: Ditto.
        * fast/canvas/webgl/resources/desktop-gl-constants.js: Add ACCUM_BUFFER_BIT enum.
2010-07-12  Zhenyao Mo  <zmo@google.com>

        Reviewed by Darin Fisher.

        Bring bufferData and clear to GLES2 compliant
        https://bugs.webkit.org/show_bug.cgi?id=41574

        * html/canvas/WebGLRenderingContext.cpp:
        (WebCore::WebGLRenderingContext::bufferData): Call validateBufferDataUsage().
        (WebCore::WebGLRenderingContext::clear): Check mask.
        (WebCore::WebGLRenderingContext::validateBufferDataUsage): Check usage.
        * html/canvas/WebGLRenderingContext.h: Declare validateBufferDataUsage.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63134 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Mario Sanchez Prada <msanchez@igalia.com>
commit-queue@webkit.org [Mon, 12 Jul 2010 22:51:43 +0000 (22:51 +0000)]
2010-07-12  Mario Sanchez Prada  <msanchez@igalia.com>

        Reviewed by Xan Lopez.

        Ensure DRT loads GAIL (Gtk+ module), for a11y tests
        https://bugs.webkit.org/show_bug.cgi?id=38648

        Add the GTK_MODULES envvar (set to "gail") to the clean
        environment when running DRT for the Gtk+ port

        * Scripts/old-run-webkit-tests:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63133 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago Build fix. Copy WKArray.h to the include directory.
adachan@apple.com [Mon, 12 Jul 2010 22:37:35 +0000 (22:37 +0000)]
    Build fix.  Copy WKArray.h to the include directory.

        * win/WebKit2Generated.make:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63132 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Eric Seidel <eric@webkit.org>
eric@webkit.org [Mon, 12 Jul 2010 22:24:23 +0000 (22:24 +0000)]
2010-07-12  Eric Seidel  <eric@webkit.org>

        Unreviewed.  Attempt to fix Chromium Windows build.

        Always generate SVGNames and MathMLNames for all ports (to support HTML5)
        https://bugs.webkit.org/show_bug.cgi?id=42050

        Another way to fix this might be to mark all the .in files with
        svn:eol=native.  But fixing the perl to be more robust against
        stray whitespace seems to be a better long-term fix.

        No functional change, thus no tests.

        * bindings/scripts/InFilesParser.pm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63131 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Martin Robinson <mrobinson@igalia.com>
mrobinson@webkit.org [Mon, 12 Jul 2010 22:22:28 +0000 (22:22 +0000)]
2010-07-12  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Dirk Schulze.

        [GTK] Support pixel tests
        https://bugs.webkit.org/show_bug.cgi?id=31518

        Finish support for pixel tests on GTK+.

        * DumpRenderTree/PixelDumpSupport.cpp: Remove unecessary RetainPtr include (only
        works on CoreFoundation systems) and add missing cstdio include.
        * DumpRenderTree/cairo/PixelDumpSupportCairo.cpp: Switch to using the MD5 support found
        in JSC library. MD5.cpp and MD5.h are only available for the Windows build.
        (computeMD5HashStringForBitmapContext): The number of bytes per row should be the row stride
        of the image, not the row stride multiplied by the width. Use JSC MD5 implementation to calculate
        the MD5 hash here. According to glibc manpages, using snprintf to build a string in unsupported
        by the ANSI standard and this fails on Linux, so unroll the loop here.
        * DumpRenderTree/cairo/PixelDumpSupportCairo.h: Correct some constructor misbehavior.
        (BitmapContext::BitmapContext): There is no reason to check the value of the
        m_context member in the constructor and it certainly should not be freed. Instead
        just initialize it with the incoming value.
        * DumpRenderTree/gtk/DumpRenderTree.cpp: Fix the order of includes.
        (dump): Actually call dumpWebViewAsPixelsAndCompareWithExpected when it is time to
        dump pixels.
        * DumpRenderTree/gtk/PixelDumpSupportGtk.cpp: Added.
        (createBitmapContextFromWebView):
        * GNUmakefile.am: Add new files to the build.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63130 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoRe-do an Xcode project change that Xcode decided not to save.
mrowe@apple.com [Mon, 12 Jul 2010 22:12:45 +0000 (22:12 +0000)]
Re-do an Xcode project change that Xcode decided not to save.

* WebKit2.xcodeproj/project.pbxproj: Remove WebKit2.exp from the project.
For some reason it was being copied in to the framework wrapper.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63129 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Martin Robinson <mrobinson@igalia.com>
mrobinson@webkit.org [Mon, 12 Jul 2010 22:07:49 +0000 (22:07 +0000)]
2010-07-12  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Gustavo Noronha Silva.

        [GTK] r63100 broke the testwebview test
        https://bugs.webkit.org/show_bug.cgi?id=42114

        * tests/testwebview.c: Disable a failing assertion in test_webkit_web_view_adjustments
        until we can get to the bottom of the failure.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63128 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Andreas Kling <andreas.kling@nokia.com>
commit-queue@webkit.org [Mon, 12 Jul 2010 22:01:43 +0000 (22:01 +0000)]
2010-07-12  Andreas Kling  <andreas.kling@nokia.com>

        Reviewed by Kenneth Rohde Christiansen.

        Remove dependency on PlatformString.h in Color.h
        https://bugs.webkit.org/show_bug.cgi?id=42109

        * platform/graphics/Color.cpp:
        * platform/graphics/Color.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63126 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoRemove the exports file from WebKit2.
mrowe@apple.com [Mon, 12 Jul 2010 21:58:27 +0000 (21:58 +0000)]
Remove the exports file from WebKit2.

Reviewed by Sam Weinig.

* Configurations/Base.xcconfig: Have symbols default to hidden visibility.
* Configurations/WebKit2.xcconfig: Remove the export file.
* UIProcess/API/mac/WKView.h: Export the WKView class.
* WebProcess/WebKitMain.cpp: Export the WebKitMain function.
* mac/WebKit2.exp: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63125 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoUpdate WebKitSystemInterface.
mrowe@apple.com [Mon, 12 Jul 2010 21:58:21 +0000 (21:58 +0000)]
Update WebKitSystemInterface.

Rubber-stamped by Sam Weinig.

* libWebKitSystemInterfaceLeopard.a:
* libWebKitSystemInterfaceSnowLeopard.a:
* libWebKitSystemInterfaceTiger.a:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63124 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Eric Seidel <eric@webkit.org>
eric@webkit.org [Mon, 12 Jul 2010 21:45:59 +0000 (21:45 +0000)]
2010-07-12  Eric Seidel  <eric@webkit.org>

        Unreviewed, build fix.

        Update HTMLTreeBuilder now that MathMLNames is always generated
        https://bugs.webkit.org/show_bug.cgi?id=42059

        Fix the Windows project file to build MathMLNames.*.
        Also added MathMLElementFactory.* which is a NOOP now, but will
        prevent folks from breaking the build when they turn MathML on.

        * WebCore.vcproj/WebCore.vcproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63123 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago https://bugs.webkit.org/show_bug.cgi?id=13075
ap@apple.com [Mon, 12 Jul 2010 21:38:42 +0000 (21:38 +0000)]
    https://bugs.webkit.org/show_bug.cgi?id=13075
        XMLHttpRequest with failed authentication should set status to 401

        https://bugs.webkit.org/show_bug.cgi?id=6871
        <rdar://problem/3363403> 401 error page is never shown

        Enable xmlhttprequest/failed-auth.html on Windows. It now passes on my machine.

        * platform/win/Skipped: Also removed two userscript tests that don't exist (run-webkit-tests
        loudly complained about those).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63122 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoWindows build fix.
sfalken@apple.com [Mon, 12 Jul 2010 21:33:24 +0000 (21:33 +0000)]
Windows build fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63121 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-09 Darin Adler <darin@apple.com>
darin@apple.com [Mon, 12 Jul 2010 21:32:34 +0000 (21:32 +0000)]
2010-07-09  Darin Adler  <darin@apple.com>

        Reviewed by Geoffrey Garen.

        String to number coercion is not spec compliant
        https://bugs.webkit.org/show_bug.cgi?id=31349

        ToNumber should ignore NBSP (\u00a0)
        https://bugs.webkit.org/show_bug.cgi?id=25490

        * runtime/JSGlobalObjectFunctions.cpp:
        (JSC::parseIntOverflow): Added a version that works on UChar.
        * runtime/JSGlobalObjectFunctions.h: Ditto.

        * runtime/UString.cpp:
        (JSC::isInfinity): Added helper functions.
        (JSC::UString::toDouble): Use isStrWhiteSpace instead of
        isSASCIISpace to define what we should skip. Got rid of the
        code that used CString and UTF8String, instead processing the
        UChar of the string directly, except for when we call strtod.
        For strtod, use our own home-grown conversion function that
        does not try to do any UTF-16 processing. Tidied up the logic
        a bit as well.
2010-07-09  Darin Adler  <darin@apple.com>

        Reviewed by Geoffrey Garen.

        String to number coercion is not spec compliant
        https://bugs.webkit.org/show_bug.cgi?id=31349

        * fast/js/ToNumber-expected.txt: Updated to expect more tests to pass.
        * fast/js/parseFloat-expected.txt: Ditto.
        * fast/js/sputnik/Conformance/09_Type_Conversion/9.3_ToNumber/9.3.1_ToNumber_from_String/S9.3.1_A2-expected.txt: Ditto.
        * fast/js/sputnik/Conformance/09_Type_Conversion/9.3_ToNumber/9.3.1_ToNumber_from_String/S9.3.1_A3_T1-expected.txt: Ditto.
        * fast/js/sputnik/Conformance/09_Type_Conversion/9.3_ToNumber/9.3.1_ToNumber_from_String/S9.3.1_A3_T2-expected.txt: Ditto.
        * fast/js/sputnik/Conformance/15_Native_Objects/15.1_The_Global_Object/15.1.2/15.1.2.3_parseFloat/S15.1.2.3_A2_T10-expected.txt: Ditto.
        * fast/js/sputnik/Conformance/15_Native_Objects/15.1_The_Global_Object/15.1.2/15.1.2.3_parseFloat/S15.1.2.3_A2_T3-expected.txt: Ditto.
        * fast/js/sputnik/Conformance/15_Native_Objects/15.1_The_Global_Object/15.1.2/15.1.2.3_parseFloat/S15.1.2.3_A2_T8-expected.txt: Ditto.
        * fast/js/sputnik/Conformance/15_Native_Objects/15.1_The_Global_Object/15.1.2/15.1.2.3_parseFloat/S15.1.2.3_A2_T9-expected.txt: Ditto.
        * fast/js/sputnik/Conformance/15_Native_Objects/15.1_The_Global_Object/15.1.2/15.1.2.3_parseFloat/S15.1.2.3_A6-expected.txt: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63120 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Eric Carlson <eric.carlson@apple.com>
eric.carlson@apple.com [Mon, 12 Jul 2010 21:26:27 +0000 (21:26 +0000)]
2010-07-12  Eric Carlson  <eric.carlson@apple.com>

        Reviewed by Darin Adler.

        Update media element's handling of empty 'src' attribute
        https://bugs.webkit.org/show_bug.cgi?id=42001

        * html/HTMLMediaElement.cpp:
        (WebCore::HTMLMediaElement::src): Return getNonEmptyURLAttribute().
        (WebCore::HTMLMediaElement::selectMediaResource): Call noneSupported() for empty 'src'.
        (WebCore::HTMLMediaElement::selectNextSourceChild): Use getNonEmptyURLAttribute() to convert
        'src' to URL instead of document()->completeURL(). Don't consider a <source> with an empty 'src'.

        * html/HTMLMediaElement.idl: Add 'NonEmpty' option to 'src' attribute.

        * html/HTMLSourceElement.cpp:
        (WebCore::HTMLSourceElement::src): Return getNonEmptyURLAttribute().
        (WebCore::HTMLSourceElement::isURLAttribute): New, 'src' is a URL attribute.
        * html/HTMLSourceElement.h:

        * html/HTMLSourceElement.idl: Add 'NonEmpty' option to 'src' attribute.

        * html/HTMLVideoElement.cpp:
        (WebCore::HTMLVideoElement::parseMappedAttribute): Use getNonEmptyURLAttribute() to convert
        'poster' to URL instead of document()->completeURL().

        * html/HTMLVideoElement.idl:  Add 'NonEmpty' option to 'poster' attribute.

2010-07-12  Eric Carlson  <eric.carlson@apple.com>

        Reviewed by NOBODY (OOPS!).

        Update media element's handling of empty 'src' attribute
        https://bugs.webkit.org/show_bug.cgi?id=42001

        * fast/dom/URL-attribute-reflection-expected.txt: Update results now that audio, video, and
        source return 'non-empty URL'.

        * media/video-poster-expected.txt:
        * media/video-poster.html: Test empty poster url.

        * media/video-source-error-no-candidate-expected.txt:
        * media/video-source-error-no-candidate.html: Empty 'src' should fires an error event.

        * media/video-source-expected.txt:
        * media/video-source.html: Test <source> element's 'src' attribute and property as well
        as currentSrc.

        * media/video-src-empty-expected.txt: Removed.
        * media/video-src-empty.html: Removed, the test is no longer valid.

        * media/video-src-none-expected.txt:
        * media/video-src-none.html: Updated to test empty 'src'. Add tests to ensure that an error
        is fired when load() is called and the 'src' is empty, but not when 'src' is missing.

        * media/video-src.html:
        * media/video-src-expected.txt: Test 'src' attribute and property as well as currentSrc.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63119 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Martin Robinson <mrobinson@igalia.com>
mrobinson@webkit.org [Mon, 12 Jul 2010 21:24:32 +0000 (21:24 +0000)]
2010-07-12  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Xan Lopez.

        [GTK] make dist is broken because of missing headers and other miscellaneous reasons
        https://bugs.webkit.org/show_bug.cgi?id=42107

        * GNUmakefile.am: Add missing header to the sources list.
2010-07-12  Martin Robinson  <mrobinson@igalia.com>

        Reviewed by Xan Lopez.

        [GTK] make dist is broken because of missing headers and other miscellaneous reasons
        https://bugs.webkit.org/show_bug.cgi?id=42107

        * GNUmakefile.am: Remove InspectorFrontend2.idl from the list of IDL files, so
        that it is not built into libWebCoreJS. Add missing header to the source list.
        Make sure to distribute the new file: WebCore/inspector/CodeGeneratorInspector.pm.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63118 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Mon, 12 Jul 2010 21:20:45 +0000 (21:20 +0000)]
2010-07-12  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        Bump to 1.3.3

        * configure.ac:

WebKit/gtk:

2010-07-12  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        Update for 1.3.3

        * NEWS:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63117 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Eric Seidel <eric@webkit.org>
eric@webkit.org [Mon, 12 Jul 2010 21:13:09 +0000 (21:13 +0000)]
2010-07-12  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        HTMLTreeBuilder needs to support mixing SVG and MathML content
        https://bugs.webkit.org/show_bug.cgi?id=42096

        * html5lib/runner-expected-html5.txt:
2010-07-12  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        HTMLTreeBuilder needs to support mixing SVG and MathML content
        https://bugs.webkit.org/show_bug.cgi?id=42096

        This is just a direct transcription of another paragraph of the
        HTML5 spec.

        This improved a couple results in html5lib/runner.html, but more
        work to do yet to pass all the foreign content tests.

        * html/HTMLTreeBuilder.cpp:
        (WebCore::HTMLTreeBuilder::processStartTag):
        (WebCore::HTMLTreeBuilder::processEndTag):
        (WebCore::HTMLTreeBuilder::processUsingSecondaryInsertionModeAndAdjustInsertionMode):
        * html/HTMLTreeBuilder.h:
        * mathml/mathtags.in:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63116 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Eric Seidel <eric@webkit.org>
eric@webkit.org [Mon, 12 Jul 2010 21:09:17 +0000 (21:09 +0000)]
2010-07-12  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Update HTMLTreeBuilder now that MathMLNames is always generated
        https://bugs.webkit.org/show_bug.cgi?id=42059

        We pass a bunch more tests now that our MathML code path is enabled.

        * html5lib/runner-expected-html5.txt:
2010-07-12  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        Update HTMLTreeBuilder now that MathMLNames is always generated
        https://bugs.webkit.org/show_bug.cgi?id=42059

        Fix the HTMLTreeBuilder MathML code path to compile and remove
        the MathML and SVG guards now that SVGNames and MathMLNames are
        always compiled into ever port after:
        https://bugs.webkit.org/show_bug.cgi?id=42050

        This fixed a whole bunch of libhtml5 tests now that we have the
        mathml code paths enabled.

        * html/HTMLTreeBuilder.cpp:
        (WebCore::HTMLTreeBuilder::processStartTagForInBody):
        (WebCore::HTMLTreeBuilder::processStartTag):
        (WebCore::HTMLTreeBuilder::resetInsertionModeAppropriately):
        (WebCore::HTMLTreeBuilder::processEndTag):
        * page/Frame.cpp:
        (WebCore::Frame::Frame):
         - Always init SVGNames and MathML names.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63115 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Mon, 12 Jul 2010 21:06:37 +0000 (21:06 +0000)]
2010-07-12  Adam Barth  <abarth@webkit.org>

        Unreviewed.

        Remove stray "raise e" that got included in a previous patch.  This
        caused the EWS bots to turn red instead of purple when a patch failed
        to apply.

        * Scripts/webkitpy/tool/commands/stepsequence.py:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63114 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoWindows build fix
sfalken@apple.com [Mon, 12 Jul 2010 21:06:25 +0000 (21:06 +0000)]
Windows build fix

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63113 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Simon Fraser <simon.fraser@apple.com>
simon.fraser@apple.com [Mon, 12 Jul 2010 21:02:22 +0000 (21:02 +0000)]
2010-07-12  Simon Fraser  <simon.fraser@apple.com>

        Reviewed by Dan Bernstein.

        Need to do a layout when RenderLayers come and go because of compositing
        https://bugs.webkit.org/show_bug.cgi?id=42108

        If we create or destroy RenderLayers for reasons other than style changes
        (e.g. because of composited iframes or plugins), then we need to ensure
        that we do a layout.

        Test: compositing/iframes/layout-on-compositing-change.html

        * rendering/RenderObject.cpp:
        (WebCore::RenderObject::adjustStyleDifference):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63112 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-11 Eric Seidel <eric@webkit.org>
eric@webkit.org [Mon, 12 Jul 2010 20:53:58 +0000 (20:53 +0000)]
2010-07-11  Eric Seidel  <eric@webkit.org>

        Reviewed by Darin Adler.

        Always generate SVGNames and MathMLNames for all ports (to support HTML5)
        https://bugs.webkit.org/show_bug.cgi?id=42050

        Historically, FOONames has only been generate when ENABLE(FOO) is defined.
        However, for HTML5 parser support, we need to have access to the SVG
        and MathML tag names regardless of whether we the engine is configured
        to render SVG or MathML content.

        This change enables generation of SVGNames and MathMLNames on all ports and
        makes it so that ports can include FOOElementFactory.* regardless of whether
        ENABLE(FOO) is defined (and have it do the right thing).

        No functional change (yet) so no tests.

        * DerivedSources.make:
        * GNUmakefile.am:
        * WebCore.gyp/WebCore.gyp:
        * WebCore.pri:
        * dom/make_names.pl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63111 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Sheriff Bot <webkit.review.bot@gmail.com>
commit-queue@webkit.org [Mon, 12 Jul 2010 20:53:45 +0000 (20:53 +0000)]
2010-07-12  Sheriff Bot  <webkit.review.bot@gmail.com>

        Unreviewed, rolling out r63101.
        http://trac.webkit.org/changeset/63101
        https://bugs.webkit.org/show_bug.cgi?id=42103

        Broke one API test (Requested by xan_ on #webkit).

        * Scripts/old-run-webkit-tests:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63110 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-10 Eric Seidel <eric@webkit.org>
eric@webkit.org [Mon, 12 Jul 2010 20:50:45 +0000 (20:50 +0000)]
2010-07-10  Eric Seidel  <eric@webkit.org>

        Reviewed by Adam Barth.

        make_names.pl should always generate all names in Names.* files
        https://bugs.webkit.org/show_bug.cgi?id=42023

        Only the *ElementFactory files need to have conditional contents
        based on enabled features.  WebCore should always have all known
        names for SVG, MathML, XML, XLink, HTML, etc. generated in the
        various *Names files, even if features are disabled.

        make_names.pl is kinda a big hack at this point.  I tried to clean
        up a little as I went.  The way I made *Names include all names was to
        read the .in files twice, once using the preprocessor and once without.

        * dom/make_names.pl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63109 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago Reviewed by Adam Roben.
ap@apple.com [Mon, 12 Jul 2010 20:39:47 +0000 (20:39 +0000)]
    Reviewed by Adam Roben.

        https://bugs.webkit.org/show_bug.cgi?id=13075
        XMLHttpRequest with failed authentication should set status to 401

        https://bugs.webkit.org/show_bug.cgi?id=6871
        <rdar://problem/3363403> 401 error page is never shown

        Fix crashes in Windows DumpRenderTree.

        * DumpRenderTree/win/ResourceLoadDelegate.cpp: (ResourceLoadDelegate::didReceiveAuthenticationChallenge):
        If we return an error, WebKit will call continueWithoutCredentialForAuthenticationChallenge()
        again on a destroyed sender.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63108 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoWindows failure-to-launch fix
aroben@apple.com [Mon, 12 Jul 2010 20:33:54 +0000 (20:33 +0000)]
Windows failure-to-launch fix

* DumpRenderTree/win/DumpRenderTree.vcproj: Copy CoreVideo.dll and
CoreVideo.pdb into WebKitOutputDir in the post-build event, too, like
we already do for CoreFoundation, CFNetwork, etc.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63107 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoUnreviewed, build fix
ajwong@chromium.org [Mon, 12 Jul 2010 20:18:53 +0000 (20:18 +0000)]
Unreviewed, build fix

[chromium] update test_expectations, and rebaseline a changed svg test.

* platform/chromium-win/fast/gradients/radial-centered-expected.checksum:
* platform/chromium-win/fast/gradients/radial-centered-expected.png:
* platform/chromium-win/fast/gradients/radial-centered-expected.txt:
* platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63106 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago https://bugs.webkit.org/show_bug.cgi?id=13075
ap@apple.com [Mon, 12 Jul 2010 20:05:52 +0000 (20:05 +0000)]
    https://bugs.webkit.org/show_bug.cgi?id=13075
        XMLHttpRequest with failed authentication should set status to 401

        https://bugs.webkit.org/show_bug.cgi?id=6871
        <rdar://problem/3363403> 401 error page is never shown

        The fix doesn't help Tiger, update results of one test and skip another.

        * platform/mac-tiger/Skipped: Skip 401-alternative-content.php.
        * platform/mac-tiger/http/tests/xmlhttprequest/failed-auth-expected.txt:
        Copied from http/tests/xmlhttprequest/failed-auth-expected.txt, updated for failure result
        in async case.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63105 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoMove WebKit2.vcproj's settings into .vsprops files
aroben@apple.com [Mon, 12 Jul 2010 19:22:17 +0000 (19:22 +0000)]
Move WebKit2.vcproj's settings into .vsprops files

This makes it easier to make changes that affect all configurations.

Fixes <http://webkit.org/b/42097> WebKit2 should use .vsprops files

Reviewed by Steve Falkenburg.

* win/WebKit2.vcproj: Moved settings from here to the files below.
* win/WebKit2Apple.vsprops: Added. Links against Apple-specific
libraries.
* win/WebKit2CFLite.vsprops: Added. Links against CFLite.
* win/WebKit2Common.vsprops: Added. Contains settings shared by all
configurations.
* win/WebKit2DirectX.vsprops: Added. Contains settings to help with
linking against DirectX.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63104 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoStop generating stripped symbols for Release builds
aroben@apple.com [Mon, 12 Jul 2010 19:21:56 +0000 (19:21 +0000)]
Stop generating stripped symbols for Release builds

It turns out we can strip the symbols after-the-fact using PDBCopy.

Fixes <http://webkit.org/b/42085>.

Reviewed by Steve Falkenburg.

JavaScriptCore:

* JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCoreCommon.vsprops:
Removed the pre-link event, which just created the public\sym
directory.

WebKit2:

* win/WebKit2.vcproj: Removed the no-longer-needed Release override of
the StripPrivateSymbols attribute. (This attribute is no longer set in
release.vsprops, so doesn't need to be overridden.)

WebKitLibraries:

* win/tools/vsprops/release.vsprops: Removed the StripPrivateSymbols
attribute, which caused link.exe to generate a stripped PDB file for
each project.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63103 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Andreas Kling <andreas.kling@nokia.com>
commit-queue@webkit.org [Mon, 12 Jul 2010 18:34:15 +0000 (18:34 +0000)]
2010-07-12  Andreas Kling  <andreas.kling@nokia.com>

        Reviewed by Oliver Hunt.

        Canvas: Move fillRect() save/restore into GraphicsContext implementations
        https://bugs.webkit.org/show_bug.cgi?id=42088

        Saving the platform painter state is an expensive operation,
        so don't do it in fillRect() for platforms that don't need it. (CG, Qt)

        * html/canvas/CanvasRenderingContext2D.cpp:
        (WebCore::CanvasRenderingContext2D::fillRect):
        * platform/graphics/cairo/GraphicsContextCairo.cpp:
        (WebCore::GraphicsContext::fillRect):
        * platform/graphics/skia/GraphicsContextSkia.cpp:
        (WebCore::GraphicsContext::fillRect):
        * platform/graphics/wince/GraphicsContextWince.cpp:
        (WebCore::GraphicsContext::fillRect):
        * platform/graphics/wx/GraphicsContextWx.cpp:
        (WebCore::GraphicsContext::fillRect):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63102 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Mario Sanchez Prada <msanchez@igalia.com>
commit-queue@webkit.org [Mon, 12 Jul 2010 18:18:32 +0000 (18:18 +0000)]
2010-07-12  Mario Sanchez Prada  <msanchez@igalia.com>

        Reviewed by Xan Lopez.

        Ensure DRT loads GAIL (Gtk+ module), for a11y tests
        https://bugs.webkit.org/show_bug.cgi?id=38648

        Add the GTK_MODULES envvar (set to "gail") to the clean
        environment when running DRT for the Gtk+ port

        * Scripts/old-run-webkit-tests:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63101 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Nate Chapin <japhet@chromium.org>
japhet@chromium.org [Mon, 12 Jul 2010 18:14:17 +0000 (18:14 +0000)]
2010-07-12  Nate Chapin  <japhet@chromium.org>

        Reviewed by Darin Fisher.

        Ensure that a cache policy that forces validation is cleared once
        the load event is fired, rather than only doing so at the next
        navigation. This leads to a lot of unnecessary load on AJAX-y
        websites.

        https://bugs.webkit.org/show_bug.cgi?id=41813

        Test: http/tests/xmlhttprequest/cache-headers-after-reload.html

        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::handledOnloadEvents): Reset m_loadType to FrameLoadTypeStandard.
        (WebCore::FrameLoader::addExtraFieldsToRequest): Only respect the original request's cache policy if the
            DocumentLoader is still loading, and handle the other cache policy settings that were scattered around the loader.
        (WebCore::FrameLoader::loadResourceSynchronously): Merge cachePolicy setting into FrameLoader::addExtraFieldsToRequest.
        * loader/SubresourceLoader.cpp:
        (WebCore::SubresourceLoader::create): Merge cachePolicy setting into FrameLoader::addExtraFieldsToRequest.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63100 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Andreas Kling <andreas.kling@nokia.com>
commit-queue@webkit.org [Mon, 12 Jul 2010 18:00:52 +0000 (18:00 +0000)]
2010-07-12  Andreas Kling  <andreas.kling@nokia.com>

        Reviewed by Oliver Hunt.

        CSS color parsing optimizations
        https://bugs.webkit.org/show_bug.cgi?id=42073

        - Avoid instantiating a CSSParser in parseColor() unless necessary.
        - Fixed hex color fast-path to support strings starting with '#'.
        - Avoid allocating a new string for the value part of a '#' color.

        * css/CSSParser.cpp:
        (WebCore::CSSParser::parseColor):
        * platform/graphics/Color.cpp:
        (WebCore::Color::parseHexColor):
        (WebCore::Color::Color):
        * platform/graphics/Color.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63099 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Mon, 12 Jul 2010 17:42:54 +0000 (17:42 +0000)]
2010-07-12  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        Suffix .mo files with the GTK+ API version so that they can be
        parallel installable.

        * configure.ac:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63098 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoWebKit2 build fix for Windows.
bweinstein@apple.com [Mon, 12 Jul 2010 17:31:37 +0000 (17:31 +0000)]
WebKit2 build fix for Windows.

* WebProcess/Plugins/Netscape/NetscapePlugin.cpp: Remove some stub function definitions.
* win/WebKit2.vcproj: Add PluginController.h to the vcproj.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63097 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoUnreviewed.
ossy@webkit.org [Mon, 12 Jul 2010 17:28:07 +0000 (17:28 +0000)]
Unreviewed.

[Qt] Some inspector test fail intermittently
https://bugs.webkit.org/show_bug.cgi?id=42090

Add tests to the Skipped list until fix to avoid false positive alarms.

* platform/qt/Skipped:
 - inspector/timeline-layout.html skipped.
 - inspector/timeline-parse-html.html skipped.
 - inspector/timeline-recalculate-styles.html skipped.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63096 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago Reviewed by Darin Adler.
ap@apple.com [Mon, 12 Jul 2010 16:57:21 +0000 (16:57 +0000)]
    Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=13075
        XMLHttpRequest with failed authentication should set status to 401

        https://bugs.webkit.org/show_bug.cgi?id=6871
        <rdar://problem/3363403> 401 error page is never shown

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63095 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoAdd a PluginController class, use it for invalidation and getting the user agent
andersca@apple.com [Mon, 12 Jul 2010 16:52:46 +0000 (16:52 +0000)]
Add a PluginController class, use it for invalidation and getting the user agent
https://bugs.webkit.org/show_bug.cgi?id=42084

Reviewed by Adam Roben.

WebCore:

* WebCore.exp.in:
Export Widget::convertToContainingWindow.

WebKit2:

* WebKit2.xcodeproj/project.pbxproj:
Add PluginController.h

* WebProcess/Plugins/DummyPlugin.cpp:
(WebKit::DummyPlugin::initialize):
* WebProcess/Plugins/DummyPlugin.h:
Pass the PluginController to initialize.

* WebProcess/Plugins/Netscape/NetscapeBrowserFuncs.cpp:
Implement NPN_UserAgent, NPN_MemAlloc, NPN_MemFree, NPN_InvalidateRect and NPN_InvalidateRegion.

* WebProcess/Plugins/Netscape/NetscapePlugin.cpp:
(WebKit::NetscapePlugin::NetscapePlugin):
Initialize m_pluginController to null.

(WebKit::NetscapePlugin::invalidate):
Ask the plug-in controller to invalidate.

(WebKit::NetscapePlugin::userAgent):´
Ask the plug-in controller for the user agent.

(WebKit::NetscapePlugin::initialize):
Set the m_pluginController member variable.

* WebProcess/Plugins/Plugin.h:
Make initialize take a PluginController.

* WebProcess/Plugins/PluginController.h: Added.

* WebProcess/Plugins/PluginView.cpp:
(WebKit::PluginView::initializePlugin):
Pass the PluginController to initialize.

(WebKit::PluginView::invalidateRect):
Tell the host window to invalidate the given rect.

(WebKit::PluginView::invalidate):
Call invalidateRect.

(WebKit::PluginView::userAgent):
Ask the frame loader client for the user agent.

* WebProcess/Plugins/PluginView.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63094 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoAdd WARN_UNUSED_RETURN to the smart pointer "leak" member functions
andersca@apple.com [Mon, 12 Jul 2010 16:41:13 +0000 (16:41 +0000)]
Add WARN_UNUSED_RETURN to the smart pointer "leak" member functions
https://bugs.webkit.org/show_bug.cgi?id=42086

Reviewed by Dan Bernstein.

* wtf/OwnPtr.h:
* wtf/PassOwnPtr.h:
* wtf/PassRefPtr.h:
(WTF::PassRefPtr::releaseRef):
(WTF::NonNullPassRefPtr::leakRef):
(WTF::NonNullPassRefPtr::releaseRef):
* wtf/RetainPtr.h:
(WTF::RetainPtr::releaseRef):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63093 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Mon, 12 Jul 2010 16:07:17 +0000 (16:07 +0000)]
2010-07-12  Pavel Feldman  <pfeldman@chromium.org>

        Not reviewed: chromium test expectations updated.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63092 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Mon, 12 Jul 2010 15:48:39 +0000 (15:48 +0000)]
2010-07-12  Pavel Feldman  <pfeldman@chromium.org>

        Not reviewed: chromium test expectations updated.

        * platform/chromium/test_expectations.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63091 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Andreas Kling <andreas.kling@nokia.com>
commit-queue@webkit.org [Mon, 12 Jul 2010 15:40:22 +0000 (15:40 +0000)]
2010-07-12  Andreas Kling  <andreas.kling@nokia.com>

        Reviewed by Tor Arne Vestbø.

        [Qt] Add location bar (Ctrl+L) shortcut in QtTestBrowser
        https://bugs.webkit.org/show_bug.cgi?id=42082

        * QtTestBrowser/main.cpp:
        (LauncherWindow::createChrome):
        * QtTestBrowser/mainwindow.cpp:
        (MainWindow::openLocation):
        * QtTestBrowser/mainwindow.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63090 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Andreas Kling <andreas.kling@nokia.com>
commit-queue@webkit.org [Mon, 12 Jul 2010 15:18:01 +0000 (15:18 +0000)]
2010-07-12  Andreas Kling  <andreas.kling@nokia.com>

        Reviewed by Simon Hausmann.

        [Qt] Dragging within webkit with a drag created via Javascript ends up misinterpreting the data
        https://bugs.webkit.org/show_bug.cgi?id=41457

        Rebaseline test now that Qt doesn't corrupt non-breaking spaces on clipboard roundtrip.

        * platform/qt/editing/pasteboard/onpaste-text-html-expected.txt:
2010-07-12  Andreas Kling  <andreas.kling@nokia.com>

        Reviewed by Simon Hausmann.

        [Qt] Dragging within webkit with a drag created via Javascript ends up misinterpreting the data
        https://bugs.webkit.org/show_bug.cgi?id=41457

        Treat non-special-cased mime-types as Unicode strings in ClipboardQt's getData()
        Fixes corruption when retrieving data that was set with anything other than text/plain

        Also use QMimeData::setHtml() when applicable to be consistent with PasteboardQt.

        * platform/qt/ClipboardQt.cpp:
        (WebCore::isHtmlMimeType):
        (WebCore::ClipboardQt::getData):
        (WebCore::ClipboardQt::setData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63089 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Steve Block <steveblock@google.com>
steveblock@google.com [Mon, 12 Jul 2010 15:06:28 +0000 (15:06 +0000)]
2010-07-12  Steve Block  <steveblock@google.com>

        Reviewed by Jeremy Orlow.

        add ANDROID to STORE_FONT_CUSTOM_PLATFORM_DATA
        https://bugs.webkit.org/show_bug.cgi?id=32273

        Tested by existing tests, just adding ANDROID to the list of platforms that use this feature.

        * loader/CachedFont.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63088 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoMake the Python autoinstaller not use a dead SourceForge server
aroben@apple.com [Mon, 12 Jul 2010 14:58:21 +0000 (14:58 +0000)]
Make the Python autoinstaller not use a dead SourceForge server

Fixes <http://webkit.org/b/42080> webkit-patch is broken due to
offline SourceForge server

Reviewed by Anders Carlsson.

* Scripts/webkitpy/thirdparty/__init__.py: Use
surfnet.dl.sourceforge.net instead of hivelocity.dl.sourceforge.net,
which seems to be down.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63087 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Pavel Feldman <pfeldman@chromium.org>
pfeldman@chromium.org [Mon, 12 Jul 2010 14:54:25 +0000 (14:54 +0000)]
2010-07-12  Pavel Feldman  <pfeldman@chromium.org>

        Not reviewed: rebaseline chromium layout tests.

        * platform/chromium-linux/fast/dom/HTMLLinkElement/prefetch-expected.txt: Added.
        * platform/chromium-linux/fast/gradients/radial-centered-expected.checksum:
        * platform/chromium-linux/fast/gradients/radial-centered-expected.png:
        * platform/chromium-linux/fast/gradients/radial-centered-expected.txt: Added.
        * platform/chromium-linux/fast/inline/continuation-outlines-with-layers-2-expected.checksum: Added.
        * platform/chromium-linux/fast/inline/continuation-outlines-with-layers-2-expected.png: Added.
        * platform/chromium-linux/fast/inline/continuation-outlines-with-layers-2-expected.txt: Added.
        * platform/chromium-mac/fast/dom/HTMLLinkElement/prefetch-expected.txt: Added.
        * platform/chromium-mac/fast/gradients/radial-centered-expected.checksum: Removed.
        * platform/chromium-mac/fast/gradients/radial-centered-expected.png: Removed.
        * platform/chromium-mac/fast/gradients/radial-centered-expected.txt: Removed.
        * platform/chromium-mac/fast/inline/continuation-outlines-with-layers-2-expected.checksum: Added.
        * platform/chromium-mac/fast/inline/continuation-outlines-with-layers-2-expected.png: Added.
        * platform/chromium-win-vista/fast/gradients/radial-centered-expected.checksum: Added.
        * platform/chromium-win-vista/fast/gradients/radial-centered-expected.png: Added.
        * platform/chromium-win-vista/fast/gradients/radial-centered-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63086 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years agoMake run-webkit-tests print how many tests timed out when exiting early
aroben@apple.com [Mon, 12 Jul 2010 14:49:52 +0000 (14:49 +0000)]
Make run-webkit-tests print how many tests timed out when exiting early

The number of timed-out tests is taken into account when deciding
whether to exit early. Leaving it out of the output just makes the
script look buggy (because it might say something like "Exiting early
after 0 crashes.").

Fixes <http://webkit.org/b/42077> run-webkit-tests prints confusing
messages when exiting early due to crashes and time-outs

Reviewed by Anders Carlsson.

* Scripts/old-run-webkit-tests:
(stopRunningTestsEarlyIfNeeded): When stopping, print the number of
timed-out tests, too.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63085 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Lucas De Marchi <lucas.demarchi@profusion.mobi>
leandro@webkit.org [Mon, 12 Jul 2010 14:27:02 +0000 (14:27 +0000)]
2010-07-12  Lucas De Marchi  <lucas.demarchi@profusion.mobi>

        Unreviewed.

        [EFL] Move ScriptConcotrollerEfl.cpp from CMakeLists.txt to
        CMakeListsEfl.txt.

        * CMakeLists.txt:
        * CMakeListsEfl.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63084 268f45cc-cd09-0410-ab3c-d52691b4dbfc

10 years ago2010-07-12 Lucas De Marchi <lucas.demarchi@profusion.mobi>
leandro@webkit.org [Mon, 12 Jul 2010 14:23:22 +0000 (14:23 +0000)]
2010-07-12  Lucas De Marchi  <lucas.demarchi@profusion.mobi>

        Unreviewed build fix after r60050.

        * CMakeLists.txt: Add WebCore/bindings to the include path.
2010-07-12  Lucas De Marchi  <lucas.demarchi@profusion.mobi>

        Unreviewed build fix after r60050.

        * CMakeLists.txt: Add WebCore/bindings to the include path.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@63083 268f45cc-cd09-0410-ab3c-d52691b4dbfc