Timeline
Nov 8, 2009:
- 11:34 PM Changeset in webkit [50640] by
-
- 4 edits1 add in trunk
The XML tokenizer reports a parse error twice if it occurs before the document element is found.
https://bugs.webkit.org/show_bug.cgi?id=31144
XMLTokenizer::doEnd() uses an additional logic to report a parse failure in
documents that end prematurely but are not considered invalid by QXmlStream.
This is to stay compatible with the libxml2 implementation.
However, that code path would be also hit in situations when it should not,
i.e. the error would have already been caught and handled. As a result, the
same error would be reported twice.
No new tests, because the problem is already covered by
fast/parser/xml-declaration-missing-ending-mark.html.
Add the Qt-specific result for that test and unskip it.
- 10:39 PM Changeset in webkit [50639] by
-
- 5 edits in trunk/WebCore
2009-11-08 Keishi Hattori <casey.hattori@gmail.com>
Reviewed by Pavel Feldman.
Web Inspector: Inspector should support copy() in the command line
https://bugs.webkit.org/show_bug.cgi?id=31238
- inspector/InspectorBackend.cpp: (WebCore::InspectorBackend::copyText): Added.
- inspector/InspectorBackend.h: Added copyText
- inspector/InspectorBackend.idl: Added copyText
- inspector/front-end/InjectedScript.js: (InjectedScript._copy): Added. (InjectedScript._ensureCommandLineAPIInstalled):
- 10:00 PM Changeset in webkit [50638] by
-
- 2 edits in trunk/WebCore
2009-11-08 Drew Wilson <atwilson@chromium.org>
Reviewed by Dimitri Glazkov.
V8 WorkerContextExecutionProxy does not handle SharedWorkers
https://bugs.webkit.org/show_bug.cgi?id=31226
Now checks to see what type of context is active and creates the
appropriate wrapper (DEDICATEDWORKERCONTEXT vs SHAREDWORKERCONTEXT).
Added support for converting to SharedWorkers and SharedWorkerContexts.
Test: Existing layout tests cover this case (start passing in Chrome).
- bindings/v8/WorkerContextExecutionProxy.cpp: (WebCore::WorkerContextExecutionProxy::initContextIfNeeded): Now generates the right type of DOMWrapper for SharedWorkerContexts. (WebCore::WorkerContextExecutionProxy::convertToV8Object): Added support for SHAREDWORKERCONTEXT. (WebCore::WorkerContextExecutionProxy::convertEventTargetToV8Object): Added support for SharedWorker and SharedWorkerContext.
- 9:52 PM Changeset in webkit [50637] by
-
- 3 edits4 adds in trunk
2009-11-08 Johnny Ding <johnnyding.webkit@gmail.com>
Reviewed by Dimitri Glazkov.
https://bugs.webkit.org/show_bug.cgi?id=31224
[V8] Return StyleSheet object instead of HTMLStyleElement w/document.styleSheets named property getter.
- fast/dom/StyleSheet/get-stylesheet-byname-expected.txt: Added.
- fast/dom/StyleSheet/get-stylesheet-byname.html: Added.
- fast/dom/StyleSheet/script-tests: Added.
- fast/dom/StyleSheet/script-tests/get-stylesheet-byname.js: Added.
2009-11-08 Johnny Ding <johnnyding.webkit@gmail.com>
Reviewed by Dimitri Glazkov.
https://bugs.webkit.org/show_bug.cgi?id=31224
[V8] Return StyleSheet object instead of HTMLStyleElement w/document.styleSheets named property getter.
Test: fast/dom/StyleSheet/get-stylesheet-byname.html
- bindings/v8/custom/V8StyleSheetListCustom.cpp: (WebCore::NAMED_PROPERTY_GETTER):
- 8:52 PM Changeset in webkit [50636] by
-
- 3 edits in trunk/WebKit/mac
Made the full-screen video HUD appear when playback stops, such as when
the end of the video is reached.
Reviewed by Eric Carlson.
- WebView/WebVideoFullscreenHUDWindowController.h: Cleaned up.
- WebView/WebVideoFullscreenHUDWindowController.mm:
(-[WebVideoFullscreenHUDWindowController scheduleTimeUpdate]): Updated
for the renaming of -updateRate to -updatePlayButton.
(-[WebVideoFullscreenHUDWindowController updatePlayButton]): Renamed
-updateRate to this.
(-[WebVideoFullscreenHUDWindowController updateRate]): This method now
responds to changes to the playback rate by updating the play button
and showing or hiding the HUD as necessary.
(-[WebVideoFullscreenHUDWindowController togglePlaying:]): Now only
toggles playing. UI updates are driven by -updateRate being called.
(-[WebVideoFullscreenHUDWindowController playing]): Cleaned up.
- 8:46 PM Changeset in webkit [50635] by
-
- 2 edits in trunk/WebCore
2009-11-08 Pavel Feldman <pfeldman@chromium.org>
Reviewed by Timothy Hatcher.
Web Inspector: "Search again" on scripts panel switches
to the script this search started with.
- inspector/front-end/Panel.js: (WebInspector.Panel.prototype.jumpToNextSearchResult):
- 6:43 PM Changeset in webkit [50634] by
-
- 3 edits in trunk/JavaScriptCore
- wtf/DateMath.cpp:
(WTF::parseDateFromNullTerminatedCharacters):
(JSC::gregorianDateTimeToMS):
(JSC::msToGregorianDateTime):
(JSC::parseDateFromNullTerminatedCharacters):
- wtf/DateMath.h:
(JSC::GregorianDateTime::GregorianDateTime):
Reviewed by NOBODY (speculative snow leopard and windows build fixes).
- 6:33 PM Changeset in webkit [50633] by
-
- 9 edits in trunk
JavaScriptCore: Hopefully, the last build fix.
Reviewed by NOBODY (chromium build fix).
Create better separation in DateMath about the JSC
and non-JSC portions. Also, only expose the non-JSC
version in the exports.
- JavaScriptCore.exp:
- JavaScriptCore.vcproj/JavaScriptCore/JavaScriptCore.def:
- wtf/DateMath.cpp:
(WTF::parseDateFromNullTerminatedCharacters):
(JSC::getUTCOffset):
(JSC::gregorianDateTimeToMS):
(JSC::msToGregorianDateTime):
(JSC::parseDateFromNullTerminatedCharacters):
- wtf/DateMath.h:
(JSC::gmtoffset):
JavaScriptGlue: Added the use jsc define for files that use
this config file and DateMath.h. This should
be able to go away when DateMath is properly
split into wtf and jsc portions which is this bug:
https://bugs.webkit.org/show_bug.cgi?id=31246
Reviewed by NOBODY (chromium build fix).
- config.h:
WebCore: * platform/network/HTTPParsers.cpp:
(WebCore::parseDate): Changed this to not
use a date parser that needs ExecState passed.
Reviewed by NOBODY (chromium build fix).
- 5:28 PM Changeset in webkit [50632] by
-
- 14 edits in trunk/WebCore
2009-11-08 Kent Tamura <tkent@chromium.org>
Reviewed by Darin Adler.
Simplify Icon interface.
https://bugs.webkit.org/show_bug.cgi?id=31154
- Remove Icon::createIconForFile(). createIconForFiles() covers createIconForFile()'s role.
- Remove FileChooser::chooseIcon()
- Change the parameter types of FileChooser constructor and the factory method, String -> const Vector<String>&, in order to support initialization with multiple files.
- Remove the icon loading code in IconChromiumWin.cpp, which doesn't work because of the sandbox.
No tests because it's just a refactoring.
- platform/FileChooser.cpp: (WebCore::FileChooser::FileChooser): (WebCore::FileChooser::create): (WebCore::FileChooser::chooseFile): (WebCore::FileChooser::chooseFiles):
- platform/FileChooser.h:
- platform/graphics/Icon.h:
- platform/graphics/chromium/IconChromiumLinux.cpp:
- platform/graphics/chromium/IconChromiumMac.cpp:
- platform/graphics/chromium/IconChromiumWin.cpp: (WebCore::Icon::createIconForFiles):
- platform/graphics/gtk/IconGtk.cpp: (WebCore::Icon::createIconForFiles):
- platform/graphics/haiku/IconHaiku.cpp:
- platform/graphics/mac/IconMac.mm: (WebCore::Icon::createIconForFiles):
- platform/graphics/qt/IconQt.cpp: (WebCore::Icon::createIconForFiles):
- platform/graphics/win/IconWin.cpp: (WebCore::Icon::createIconForFiles):
- platform/graphics/wx/IconWx.cpp:
- rendering/RenderFileUploadControl.cpp: (WebCore::RenderFileUploadControl::RenderFileUploadControl):
- 5:18 PM Changeset in webkit [50631] by
-
- 4 edits3 adds in trunk
2009-11-08 Daniel Bates <dbates@webkit.org>
Reviewed by Adam Barth.
Allows same-origin plugin-based content to load.
Test: http/tests/security/xssAuditor/object-src-inject.html
- page/XSSAuditor.cpp: (WebCore::XSSAuditor::canLoadExternalScriptFromSrc): Modified to call XSSAuditor::isSameOriginResource. (WebCore::XSSAuditor::canLoadObject): Ditto. (WebCore::XSSAuditor::canSetBaseElementURL): Ditto. (WebCore::XSSAuditor::isSameOriginResource): Added.
- page/XSSAuditor.h:
2009-11-08 Daniel Bates <dbates@webkit.org>
Reviewed by Adam Barth.
Tests that the XSSAuditor prevents loading plugin-based content that is not
from the same-origin as the enclosing page.
- http/tests/security/xssAuditor/object-src-inject-expected.txt: Added.
- http/tests/security/xssAuditor/object-src-inject.html: Added.
- http/tests/security/xssAuditor/resources/echo-object-src.pl: Added.
- 3:54 PM Changeset in webkit [50630] by
-
- 2 edits in trunk/WebCore
Unreviewed build fix for chromium.
- platform/graphics/skia/GraphicsContextSkia.cpp:
(WebCore::GraphicsContext::fillPath):
(WebCore::GraphicsContext::fillRect):
(WebCore::GraphicsContext::strokePath):
(WebCore::GraphicsContext::strokeRect):
- 3:42 PM Changeset in webkit [50629] by
-
- 2 edits in trunk/WebCore
Unreviewed build fix for chromium.
Build fix for https://bugs.webkit.org/show_bug.cgi?id=31219
Clean up GraphicsContext's current concept of ColorSpace
- platform/graphics/skia/GraphicsContextSkia.cpp:
(WebCore::GraphicsContext::fillPath):
(WebCore::GraphicsContext::fillRect):
(WebCore::GraphicsContext::strokePath):
(WebCore::GraphicsContext::strokeRect):
- 1:02 PM Changeset in webkit [50628] by
-
- 3 edits in trunk/JavaScriptCore
For the change in DateMath.
Reviewed by NOBODY (chromium build fix).
- config.h:
- wtf/DateMath.cpp:
- 11:47 AM Changeset in webkit [50627] by
-
- 3 edits in trunk/WebCore
2009-11-08 Pavel Feldman <pfeldman@chromium.org>
Reviewed by Timothy Hatcher.
Web Inspector: [REGRESSION] committing style edit
clears elements panel selection.
- inspector/front-end/ElementsTreeOutline.js: (WebInspector.ElementsTreeOutline.prototype.update):
- inspector/front-end/StylesSidebarPane.js: (WebInspector.StylePropertyTreeElement.prototype.): (WebInspector.StylePropertyTreeElement.prototype):
- 11:09 AM Changeset in webkit [50626] by
-
- 3 edits in trunk/LayoutTests
2009-11-08 Shu Chang <Chang.Shu@nokia.com>
Reviewed by Holger Freyther.
Update texts in test description after test case change in r50579.
https://bugs.webkit.org/show_bug.cgi?id=31122
- editing/selection/shrink-selection-after-shift-pagedown-expected.txt:
- editing/selection/shrink-selection-after-shift-pagedown.html:
- 11:01 AM Changeset in webkit [50625] by
-
- 27 edits in trunk
2009-11-08 Janne Koskinen <janne.p.koskinen@digia.com>
Reviewed by Holger Freyther.
ResourceRequest to be class instead of struct
https://bugs.webkit.org/show_bug.cgi?id=30670
Started as a compilation fix for Symbian where the compiler makes a distinction between
class and struct in function argument signatures.
Changed all forward declarations of ResourceRequest to have class in the forward
declaration instead of struct and changed the definition of ResourceRequest to be class
and added access qualifiers where missing. Additionally two references of friend
struct ResourceRequestBase changed to class instead.
- history/HistoryItem.h:
- inspector/InspectorController.h:
- inspector/InspectorResource.h:
- loader/DocumentThreadableLoader.h:
- loader/FrameLoaderClient.h:
- loader/MainResourceLoader.h:
- loader/ResourceLoadNotifier.h:
- loader/SubresourceLoader.h:
- loader/SubresourceLoaderClient.h:
- loader/ThreadableLoader.h:
- loader/WorkerThreadableLoader.h:
- loader/appcache/ApplicationCache.h:
- loader/appcache/ApplicationCacheHost.h:
- platform/CrossThreadCopier.h:
- platform/network/ResourceHandle.h:
- platform/network/ResourceHandleClient.h:
- platform/network/ResourceRequestBase.h:
- platform/network/cf/ResourceRequest.h:
- platform/network/cf/ResourceRequestCFNet.h:
- platform/network/chromium/ResourceRequest.h:
- platform/network/curl/ResourceRequest.h:
- platform/network/qt/ResourceRequest.h:
- platform/network/soup/ResourceRequest.h:
- xml/XMLHttpRequest.h:
2009-11-08 Janne Koskinen <janne.p.koskinen@digia.com>
Reviewed by Holger Freyther.
ResourceRequest to be class instead of struct
https://bugs.webkit.org/show_bug.cgi?id=30670
Started as a compilation fix for Symbian where the compiler makes a distinction between
class and struct in function argument signatures.
Changed forward declaration of ResourceRequest to have class in the forward
declaration instead of struct.
- WebDownload.h:
- 10:51 AM Changeset in webkit [50624] by
-
- 2 edits in trunk/WebKitTools
2009-11-08 Shu Chang <Chang.Shu@nokia.com>
Reviewed by Holger Freyther.
[Qt] Added support for key code 8 (backspace) in EventSenderQt.
This helps to pass the test case below. Also replaced hardcoded
code numbers with defined constants.
https://bugs.webkit.org/show_bug.cgi?id=31185
Test: editing/undo/undo-deleteWord.html
- DumpRenderTree/qt/EventSenderQt.cpp: (EventSender::keyDown):
- 9:32 AM Changeset in webkit [50623] by
-
- 6 edits4 adds in trunk
<rdar://problem/7363434> Crash inside RenderObject::localToAbsolute
below FrameView::layout
https://bugs.webkit.org/show_bug.cgi?id=31093
Reviewed by Simon Fraser.
WebCore:
Test: fast/block/positioning/relative-positioned-inline-container.html
In <http://trac.webkit.org/changeset/19148>, setStaticY() was changed
to mark the object for layout, doing so without marking its ancestors.
However, RenderBlock::skipLeadingWhitespace and
RenderBlock::skipTrailingWhitespace() call setStaticY() on a relative-
positioned inline container, causing it to be marked for layout without
ever going back to give it layout, and thus layout could end with a
dirty object still in the tree, leading to all sorts of badness.
The fix is to revert setStaticY() to not marking the object dirty, and
instead do it in the call sites that require it, which are in
RenderBlock and RenderFlexibleBox.
- rendering/RenderBlock.cpp:
(WebCore::RenderBlock::adjustPositionedBlock):
- rendering/RenderFlexibleBox.cpp:
(WebCore::RenderFlexibleBox::layoutHorizontalBox):
(WebCore::RenderFlexibleBox::layoutVerticalBox):
- rendering/RenderLayer.cpp:
- rendering/RenderLayer.h:
(WebCore::RenderLayer::setStaticY):
LayoutTests:
- fast/block/positioning/relative-positioned-inline-container-expected.checksum: Added.
- fast/block/positioning/relative-positioned-inline-container-expected.png: Added.
- fast/block/positioning/relative-positioned-inline-container-expected.txt: Added.
- fast/block/positioning/relative-positioned-inline-container.html: Added.
- 9:26 AM Changeset in webkit [50622] by
-
- 2 edits in trunk/WebKitTools
[Qt] [DRT] Fix wrong logic in LayoutTestController processWork
https://bugs.webkit.org/show_bug.cgi?id=31164
Patch by Antonio Gomes <tonikitoo@webkit.org> on 2009-11-05
Reviewed by Holger Freyther.
Fixed wrong logic to assume WorkQueue is done in QT's DRT.
- DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::processWork):
- 9:25 AM Changeset in webkit [50621] by
-
- 4 edits in trunk/WebKitTools
[Qt][DRT] Replace queueScript by queueNonLoadingScript and queueLoadingScript method
https://bugs.webkit.org/show_bug.cgi?id=31158
Patch by Antonio Gomes <tonikitoo@webkit.org> on 2009-11-07
Reviewed by Holger Freyther.
By invoking a script queue'd by queueScript(), 'true' was beeing returned
always, which from WorkQueue prospective means that a load has been started
and the queue processing should stop and wait for the load to finish.
Spinning it off into a loading and a non-loading variants was the solution
adopted by Mac's DRT to work around this problem. The former keeps returning
'true' while the later executes the script synchronously and returns 'false'
making it possible to the WorkQueue to proceed right away.
- DumpRenderTree/qt/LayoutTestControllerQt.cpp:
(LayoutTestController::processWork):
(LayoutTestController::queueLoadingScript):
(LayoutTestController::queueNonLoadingScript):
- DumpRenderTree/qt/LayoutTestControllerQt.h:
- DumpRenderTree/qt/WorkQueueItem.h:
(LoadingScriptItem::LoadingScriptItem):
(LoadingScriptItem::invoke):
(NonLoadingScriptItem::NonLoadingScriptItem):
(NonLoadingScriptItem::invoke):
Nov 7, 2009:
- 7:59 PM Changeset in webkit [50620] by
-
- 2 edits in trunk/WebKitTools
Invoke prepare-ChangeLog via an absolute path rather than assuming it can be found in PATH.
Rubber-stamped by Cameron Zwarich.
- Scripts/commit-log-editor:
- 7:59 PM Changeset in webkit [50619] by
-
- 4 edits in trunk/WebKitTools
Fix <https://bugs.webkit.org/show_bug.cgi?id=28168>.
commit-log-editor does not support all the email address configurations that prepare-Changelog supports
Reviewed by Darin Adler.
Move logic for determining the name and email address to use in a ChangeLog entry from
prepare-ChangeLog to VCSUtils so that commit-log-editor can use it. It wants to check
whether the author of the patch matches committer, and therefore needs access to the
email address that would be used in a ChangeLog entry.
Based on a patch by Pierre d'Herbemont.
- Scripts/VCSUtils.pm:
- Scripts/commit-log-editor:
- Scripts/prepare-ChangeLog:
- Scripts/webkitdirs.pm:
- 6:31 PM Changeset in webkit [50618] by
-
- 2 edits1 add in trunk/WebCore
2009-11-07 Daniel Bates <dbates@webkit.org>
Reviewed by Darin Adler.
Fixes an issue where the onchange event handler is not fired when the
input field is autocompleted.
We cannot test this using DRT since DRT cannot emulate autocompletion.
So, a manual-test is included.
Tests: manual-tests/autocompletion-fire-onchange.html
- manual-tests/autocompletion-fire-onchange.html: Added.
- rendering/RenderTextControl.cpp: (WebCore::RenderTextControl::setInnerTextValue):
- 2:40 PM Changeset in webkit [50617] by
-
- 1 edit4 adds in trunk/LayoutTests
2009-11-07 Dirk Pranke <dpranke@chromium.org>
Reviewed by Darin Adler.
Add in missing expectations.
- platform/mac/fast/forms/input-list-button-size-expected.checksum: Added.
- platform/mac/fast/forms/input-list-button-size-expected.png: Added.
- platform/mac/fast/forms/select-dirty-parent-pref-widths-expected.png: Added.
- platform/mac/fast/text/text-letter-spacing-expected.png: Added.
- 12:06 PM Changeset in webkit [50616] by
-
- 2 edits in trunk/WebCore
2009-11-07 Benjamin Otte <otte@gnome.org>
Reviewed by Holger Freyther.
Mark redrawn areas on image surfaces as dirty.
This is required to conform to the Cairo API, but is currently only
used by debugging tools like cairo-trace.
- platform/graphics/cairo/ImageBufferCairo.cpp: (WebCore::ImageBuffer::platformTransformColorSpace): (WebCore::putImageData):
- 1:03 AM Changeset in webkit [50615] by
-
- 2 edits in trunk/WebCore
2009-11-07 Joanmarie Diggs <joanmarie.diggs@gmail.com>
Reviewed by Xan Lopez.
https://bugs.webkit.org/show_bug.cgi?id=30878
[Gtk] atk_text_get_text() fails in entries when the end_offset is -1
If the end_offset is -1, use the String length as the end_offset.
- accessibility/gtk/AccessibilityObjectWrapperAtk.cpp: (webkit_accessible_text_get_text):
- 12:00 AM Changeset in webkit [50614] by
-
- 8 edits in trunk/WebCore
Fix for https://bugs.webkit.org/show_bug.cgi?id=31219 Clean up
GraphicsContext's current concept of ColorSpace
Reviewed by Adele Peterson.
ColorSpace is now called ColorType. The variables on the state we
appropriately re-named as well. I removed strokeColorSpace() and
fillColorSpace() from GraphicsContext since they were never called.
- platform/graphics/GraphicsContext.cpp:
(WebCore::GraphicsContext::setStrokeColor):
(WebCore::GraphicsContext::setFillColor):
(WebCore::GraphicsContext::setStrokePattern):
(WebCore::GraphicsContext::setFillPattern):
(WebCore::GraphicsContext::setStrokeGradient):
(WebCore::GraphicsContext::setFillGradient):
- platform/graphics/GraphicsContext.h:
(WebCore::):
- platform/graphics/GraphicsContextPrivate.h:
(WebCore::GraphicsContextState::GraphicsContextState):
- platform/graphics/cairo/GraphicsContextCairo.cpp:
(WebCore::GraphicsContext::fillPath):
(WebCore::GraphicsContext::strokePath):
- platform/graphics/cg/GraphicsContextCG.cpp:
(WebCore::calculateDrawingMode):
(WebCore::GraphicsContext::drawPath):
(WebCore::GraphicsContext::fillPath):
(WebCore::GraphicsContext::strokePath):
(WebCore::GraphicsContext::fillRect):
(WebCore::GraphicsContext::strokeRect):
- platform/graphics/qt/GraphicsContextQt.cpp:
(WebCore::GraphicsContext::fillPath):
(WebCore::GraphicsContext::strokePath):
(WebCore::GraphicsContext::fillRect):
- platform/graphics/wince/GraphicsContextWince.cpp:
(WebCore::GraphicsContext::fillPath):
(WebCore::GraphicsContext::fillRect):
