Timeline


and

03/19/10:

23:42 Changeset [56298] by mjs@apple.com
  • 2 edits in trunk/WebKitSite

2010-03-19 Maciej Stachowiak <mjs@apple.com>

Reviewed by Mark Rowe.

http://webkit.org/ has an HTML5 validation error
https://bugs.webkit.org/show_bug.cgi?id=36406

  • nav.inc:
22:29 Changeset [56297] by hamaji@chromium.org
  • 3 edits
    2 adds in trunk

2010-03-19 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by Dan Bernstein.

WebCore::RenderButton::styleDidChange ReadAV@NULL (6739b7fe455ecb54a6812c0866c3b47c)
https://bugs.webkit.org/show_bug.cgi?id=34641

  • fast/css/first-letter-block-form-controls-crash-expected.txt: Added.
  • fast/css/first-letter-block-form-controls-crash.html: Added.

2010-03-19 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by Dan Bernstein.

WebCore::RenderButton::styleDidChange ReadAV@NULL (6739b7fe455ecb54a6812c0866c3b47c)
https://bugs.webkit.org/show_bug.cgi?id=34641

Don't dig into buttons and menu lists when finding which element
should be modified by :first-letter pseudo class. Even before
this change, we didn't dig into inline buttons and menu lists as
they are replaced so this issue wasn't found long time.

Test: fast/css/first-letter-block-form-controls-crash.html

  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::updateFirstLetter):
21:29 Changeset [56296] by mrowe@apple.com
  • 1 edit
    1 add in trunk/LayoutTests

Land a test result that should have been added along with the test in r56286.

  • transitions/transition-end-event-destroy-iframe-expected.txt: Copied from LayoutTests/transitions/transition-end-event-destroy-renderer-expected.txt.
20:55 Changeset [56295] by eric@webkit.org
  • 11 edits
    2 copies
    2 adds in trunk

2010-03-19 Adam Barth <abarth@webkit.org>

Reviewed by Daniel Bates.

Change XSSAuditor block syntax
https://bugs.webkit.org/show_bug.cgi?id=34436

Update tests to account for the new blocking syntax. Added test for
disabling the XSS filter.

  • http/tests/security/xssAuditor/malformed-xss-protection-header-expected.txt:
  • http/tests/security/xssAuditor/malformed-xss-protection-header.html:
  • http/tests/security/xssAuditor/no-protection-script-tag-expected.txt: Added.
  • http/tests/security/xssAuditor/no-protection-script-tag.html: Added.
  • http/tests/security/xssAuditor/resources/echo-head-base-href.pl:
  • http/tests/security/xssAuditor/resources/echo-intertag-click-and-notify.pl:
  • http/tests/security/xssAuditor/resources/echo-intertag.pl:
  • http/tests/security/xssAuditor/xss-protection-parsing-01-expected.txt: Added.
  • http/tests/security/xssAuditor/xss-protection-parsing-01.html: Added.

2010-03-19 Adam Barth <abarth@webkit.org>

Reviewed by Daniel Bates.

Change XSSAuditor block syntax
https://bugs.webkit.org/show_bug.cgi?id=34436

Update our blocking syntax to something more reasonable. Also,
implemented a way for a web site to disable the filter.

Tests: http/tests/security/xssAuditor/no-protection-script-tag.html

http/tests/security/xssAuditor/xss-protection-parsing-01.html

  • page/XSSAuditor.cpp: (WebCore::XSSAuditor::xssProtection): (WebCore::XSSAuditor::findInRequest):
  • page/XSSAuditor.h:
  • platform/network/HTTPParsers.cpp: (WebCore::skipToken): (WebCore::parseXSSProtectionHeader):
  • platform/network/HTTPParsers.h: (WebCore::):
20:35 Changeset [56294] by eric@webkit.org
  • 4 edits
    4 adds in trunk/WebKit/chromium

2010-03-19 Kenneth Russell <kbr@google.com>

Reviewed by Darin Fisher.

Add GraphicsContext3D abstraction to WebKit API
https://bugs.webkit.org/show_bug.cgi?id=36262

Added WebGraphicsContext3D to the WebKit API and refactored Chromium's
GraphicsContext3D implementation to use it. All of the OpenGL calls have
been moved out of WebKit/chromium/src/GraphicsContext3D.cpp and into the
WebGraphicsContext3D implementation. GraphicsContext3D is still
responsible for the transfer of rendered output from the
WebGraphicsContext3D to the HTMLCanvasElement.

The GraphicsContext3DInternal class, which is a data member of
GraphicsContext3D for the Chromium port, remains. It is possible to
eliminate this class and thereby one level of delegation, but this is
being deferred.

The needed entry point for a Chrome implementation of
WebGraphicsContext3D has been added to WebKitClient, but it is not being
called yet by GraphicsContext3D. It will be once this patch lands and
Chromium is rolled forward to support this entry point.

This is a large patch, but the transformation is almost entirely
mechanical and there is no change in functionality. Nearly all of
GraphicsContext3D and GraphicsContext3DInternal has been moved to
WebGraphicsContext3DDefaultImpl. The only area where the splitting of
logic is less than mechanical is GraphicsContext3D::beginPaint() and its
callees.

Ran all WebGL layout tests and demos from Khronos site in Chromium.

  • WebKit.gyp:
  • public/WebGraphicsContext3D.h: Added. (WebKit::WebGraphicsContext3D::Attributes::Attributes): (WebKit::WebGraphicsContext3D::~WebGraphicsContext3D):
  • public/WebKitClient.h: (WebKit::WebKitClient::createGraphicsContext3D):
  • src/GraphicsContext3D.cpp: (WebCore::GraphicsContext3DInternal::GraphicsContext3DInternal): (WebCore::GraphicsContext3DInternal::~GraphicsContext3DInternal): (WebCore::GraphicsContext3DInternal::initialize): (WebCore::GraphicsContext3DInternal::platformGraphicsContext3D): (WebCore::GraphicsContext3DInternal::platformTexture): (WebCore::GraphicsContext3DInternal::beginPaint): (WebCore::GraphicsContext3DInternal::endPaint): (WebCore::GraphicsContext3DInternal::reshape): (WebCore::GraphicsContext3DInternal::bindAttribLocation): (WebCore::GraphicsContext3DInternal::bindTexture): (WebCore::GraphicsContext3DInternal::bufferData): (WebCore::GraphicsContext3DInternal::bufferSubData): (WebCore::GraphicsContext3DInternal::getActiveAttrib): (WebCore::GraphicsContext3DInternal::getActiveUniform): (WebCore::GraphicsContext3DInternal::getAttribLocation): (WebCore::GraphicsContext3DInternal::getContextAttributes): (WebCore::GraphicsContext3DInternal::getProgramInfoLog): (WebCore::GraphicsContext3DInternal::getShaderInfoLog): (WebCore::GraphicsContext3DInternal::getShaderSource): (WebCore::GraphicsContext3DInternal::getString): (WebCore::GraphicsContext3DInternal::getUniformLocation): (WebCore::GraphicsContext3DInternal::readPixels): (WebCore::GraphicsContext3DInternal::shaderSource): (WebCore::GraphicsContext3DInternal::texImage2D): (WebCore::GraphicsContext3DInternal::texSubImage2D): (WebCore::GraphicsContext3DInternal::uniform1fv): (WebCore::GraphicsContext3DInternal::uniform1iv): (WebCore::GraphicsContext3DInternal::uniform2fv): (WebCore::GraphicsContext3DInternal::uniform2iv): (WebCore::GraphicsContext3DInternal::uniform3fv): (WebCore::GraphicsContext3DInternal::uniform3iv): (WebCore::GraphicsContext3DInternal::uniform4fv): (WebCore::GraphicsContext3DInternal::uniform4iv): (WebCore::GraphicsContext3DInternal::uniformMatrix2fv): (WebCore::GraphicsContext3DInternal::uniformMatrix3fv): (WebCore::GraphicsContext3DInternal::uniformMatrix4fv): (WebCore::GraphicsContext3D::GraphicsContext3D): (WebCore::GraphicsContext3D::~GraphicsContext3D): (WebCore::GraphicsContext3D::create): (WebCore::GraphicsContext3D::platformGraphicsContext3D): (WebCore::GraphicsContext3D::platformTexture): (WebCore::GraphicsContext3D::texImage2D): (WebCore::GraphicsContext3D::texSubImage2D):
  • src/WebGraphicsContext3D.cpp: Added. (WebKit::WebGraphicsContext3D::createDefault):
  • src/WebGraphicsContext3DDefaultImpl.cpp: Added. (WebKit::WebGraphicsContext3DDefaultImpl::GLConnection::create): (WebKit::WebGraphicsContext3DDefaultImpl::GLConnection::~GLConnection): (WebKit::WebGraphicsContext3DDefaultImpl::VertexAttribPointerState::VertexAttribPointerState): (WebKit::WebGraphicsContext3DDefaultImpl::WebGraphicsContext3DDefaultImpl): (WebKit::WebGraphicsContext3DDefaultImpl::~WebGraphicsContext3DDefaultImpl): (WebKit::WebGraphicsContext3DDefaultImpl::initialize): (WebKit::WebGraphicsContext3DDefaultImpl::makeContextCurrent): (WebKit::WebGraphicsContext3DDefaultImpl::width): (WebKit::WebGraphicsContext3DDefaultImpl::height): (WebKit::WebGraphicsContext3DDefaultImpl::sizeInBytes): (WebKit::createTextureObject): (WebKit::WebGraphicsContext3DDefaultImpl::reshape): (WebKit::WebGraphicsContext3DDefaultImpl::flipVertically): (WebKit::WebGraphicsContext3DDefaultImpl::readBackFramebuffer): (WebKit::WebGraphicsContext3DDefaultImpl::activeTexture): (WebKit::WebGraphicsContext3DDefaultImpl::bindBuffer): (WebKit::WebGraphicsContext3DDefaultImpl::bindFramebuffer): (WebKit::WebGraphicsContext3DDefaultImpl::disableVertexAttribArray): (WebKit::WebGraphicsContext3DDefaultImpl::drawElements): (WebKit::WebGraphicsContext3DDefaultImpl::enableVertexAttribArray): (WebKit::WebGraphicsContext3DDefaultImpl::generateMipmap): (WebKit::WebGraphicsContext3DDefaultImpl::getActiveAttrib): (WebKit::WebGraphicsContext3DDefaultImpl::getActiveUniform): (WebKit::WebGraphicsContext3DDefaultImpl::getContextAttributes): (WebKit::WebGraphicsContext3DDefaultImpl::getError): (WebKit::WebGraphicsContext3DDefaultImpl::getProgramInfoLog): (WebKit::WebGraphicsContext3DDefaultImpl::getShaderInfoLog): (WebKit::WebGraphicsContext3DDefaultImpl::getShaderSource): (WebKit::WebGraphicsContext3DDefaultImpl::getString): (WebKit::WebGraphicsContext3DDefaultImpl::getVertexAttribOffset): (WebKit::WebGraphicsContext3DDefaultImpl::releaseShaderCompiler): (WebKit::WebGraphicsContext3DDefaultImpl::shaderSource): (WebKit::WebGraphicsContext3DDefaultImpl::vertexAttribPointer): (WebKit::WebGraphicsContext3DDefaultImpl::createBuffer): (WebKit::WebGraphicsContext3DDefaultImpl::createFramebuffer): (WebKit::WebGraphicsContext3DDefaultImpl::createProgram): (WebKit::WebGraphicsContext3DDefaultImpl::createRenderbuffer): (WebKit::WebGraphicsContext3DDefaultImpl::createTexture): (WebKit::WebGraphicsContext3DDefaultImpl::deleteBuffer): (WebKit::WebGraphicsContext3DDefaultImpl::deleteFramebuffer): (WebKit::WebGraphicsContext3DDefaultImpl::deleteProgram): (WebKit::WebGraphicsContext3DDefaultImpl::deleteRenderbuffer): (WebKit::WebGraphicsContext3DDefaultImpl::deleteShader): (WebKit::WebGraphicsContext3DDefaultImpl::deleteTexture): (WebKit::WebGraphicsContext3DDefaultImpl::synthesizeGLError):
  • src/WebGraphicsContext3DDefaultImpl.h: Added. (WebKit::WebGraphicsContext3DDefaultImpl::): (WebKit::WebGraphicsContext3DDefaultImpl::GLConnection::chooseFBConfig): (WebKit::WebGraphicsContext3DDefaultImpl::GLConnection::createNewContext): (WebKit::WebGraphicsContext3DDefaultImpl::GLConnection::createPbuffer): (WebKit::WebGraphicsContext3DDefaultImpl::GLConnection::destroyPbuffer): (WebKit::WebGraphicsContext3DDefaultImpl::GLConnection::makeCurrent): (WebKit::WebGraphicsContext3DDefaultImpl::GLConnection::destroyContext): (WebKit::WebGraphicsContext3DDefaultImpl::GLConnection::getCurrentContext): (WebKit::WebGraphicsContext3DDefaultImpl::GLConnection::GLConnection):
20:03 Changeset [56293] by eric@webkit.org
  • 13 edits in trunk/WebCore

2010-03-19 Eric Uhrhane <ericu@chromium.org>

Reviewed by Dmitry Titov.

Refactor DatabaseTracker.cpp for thread safety
https://bugs.webkit.org/show_bug.cgi?id=34991

This enables calling into DatabaseTracker from multiple context threads,
as will happen once Workers can access the Database. It required a fair
amount of reshuffling of locks. I ended up splitting the public
interface [calls that take locks and call private functions] from the
implementations [calls that assert that locks are already held] in order
to avoid lock conflicts. I also had to make sure we weren't sharing
Strings or SecurityOrigins across threads.

No new tests.

Allow access to database handles from multiple threads IFF SQLite is new enough and the user requests it.

  • platform/sql/SQLiteDatabase.cpp: (WebCore::SQLiteDatabase::SQLiteDatabase): (WebCore::SQLiteDatabase::disableThreadingChecks):
  • platform/sql/SQLiteDatabase.h: (WebCore::SQLiteDatabase::sqlite3Handle): (WebCore::SQLiteDatabase::disableThreadingChecks):

Remove an asynchronous call from Database::close back to the execution thread, so that cleanup can be more deterministic.

  • storage/Database.cpp: (WebCore::Database::markAsDeletedAndClose): (WebCore::Database::close):
  • storage/Database.h: (WebCore::Database::):
  • storage/DatabaseDetails.h: (WebCore::DatabaseDetails::DatabaseDetails): (WebCore::DatabaseDetails::thread):
  • storage/DatabaseTask.cpp: (WebCore::DatabaseCloseTask::doPerformTask):
  • storage/DatabaseThread.cpp: (WebCore::DatabaseThread::databaseThread):

Any Strings that get stored in DatabaseTracker, and any Strings returned from DatabaseTracker, are now threadsafeCopies.
Public functions now take all needed locks, then generally call only private functions [there are a few exceptions: deletion functions and origins()].
Private functions no longer take locks.
m_quotaMapGuard becomes m_databaseGuard, and now protects m_database, m_quotaMap, m_proposedDatabases, m_databaseDirectoryPath, m_originsBeingDeleted, m_beingCreated, and m_beingDeleted.
m_proposedDatabases replaces m_proposedDatabase, to account for reentrancy.

  • storage/DatabaseTracker.h:
  • storage/DatabaseTracker.cpp: (WebCore::DatabaseTracker::originQuotaManagerNoLock): (WebCore::DatabaseTracker::originQuotaManager): (WebCore::DatabaseTracker::DatabaseTracker): (WebCore::DatabaseTracker::setDatabaseDirectoryPath): (WebCore::DatabaseTracker::databaseDirectoryPath): (WebCore::DatabaseTracker::trackerDatabasePath): (WebCore::DatabaseTracker::openTrackerDatabase): (WebCore::DatabaseTracker::canEstablishDatabase): (WebCore::DatabaseTracker::hasEntryForOriginNoLock): (WebCore::DatabaseTracker::hasEntryForOrigin): (WebCore::DatabaseTracker::hasEntryForDatabase): (WebCore::DatabaseTracker::originPath): (WebCore::DatabaseTracker::fullPathForDatabaseNoLock): (WebCore::DatabaseTracker::fullPathForDatabase): (WebCore::DatabaseTracker::populateOrigins): (WebCore::DatabaseTracker::origins): (WebCore::DatabaseTracker::databaseNamesForOriginNoLock): (WebCore::DatabaseTracker::databaseNamesForOrigin): (WebCore::DatabaseTracker::detailsForNameAndOrigin): (WebCore::DatabaseTracker::setDatabaseDetails): (WebCore::DatabaseTracker::usageForDatabase): (WebCore::DatabaseTracker::addOpenDatabase): (WebCore::DatabaseTracker::removeOpenDatabase): (WebCore::DatabaseTracker::usageForOriginNoLock): (WebCore::DatabaseTracker::usageForOrigin): (WebCore::DatabaseTracker::quotaForOriginNoLock): (WebCore::DatabaseTracker::quotaForOrigin): (WebCore::DatabaseTracker::setQuota): (WebCore::DatabaseTracker::addDatabase): (WebCore::DatabaseTracker::deleteAllDatabases): (WebCore::DatabaseTracker::deleteOrigin): (WebCore::DatabaseTracker::deleteDatabase): (WebCore::DatabaseTracker::deleteDatabaseFile): (WebCore::DatabaseTracker::setClient): (WebCore::DatabaseTracker::scheduleNotifyDatabaseChanged): (WebCore::DatabaseTracker::notifyDatabasesChanged):

These functions keep track of in-progress deletions and creations, so that we can make sure nobody every deletes a database file while a live database is using it.
(WebCore::DatabaseTracker::canCreateDatabase):
(WebCore::DatabaseTracker::recordCreatingDatabase):
(WebCore::DatabaseTracker::doneCreatingDatabase):
(WebCore::DatabaseTracker::creatingDatabase):
(WebCore::DatabaseTracker::canDeleteDatabase):
(WebCore::DatabaseTracker::recordDeletingDatabase):
(WebCore::DatabaseTracker::doneDeletingDatabase):
(WebCore::DatabaseTracker::deletingDatabase):
(WebCore::DatabaseTracker::canDeleteOrigin):
(WebCore::DatabaseTracker::deletingOrigin):
(WebCore::DatabaseTracker::recordDeletingOrigin):
(WebCore::DatabaseTracker::doneDeletingOrigin):

Any SecurityOrigins stored in OriginQuotaManager are now threadsafeCopies of inputs. There's a new tryLock() function in addition to the existing lock().

  • storage/OriginQuotaManager.cpp: (WebCore::OriginQuotaManager::tryLock): (WebCore::OriginQuotaManager::trackOrigin):
  • storage/OriginQuotaManager.h:
  • page/SecurityOrigin.cpp: (WebCore::SecurityOrigin::databaseIdentifier): Removed DEFINE_STATIC_LOCAL wrapper on a local variable; it appears to have been a small optimization, but it's not thread-safe.
19:35 Changeset [56292] by eric@webkit.org
  • 6 edits
    2 adds in trunk/WebCore

2010-03-19 Luiz Agostini <luiz.agostini@openbossa.org>

Reviewed by Antti Koivisto.

Multiselect popups - rendering
https://bugs.webkit.org/show_bug.cgi?id=36006

The objective of this patch is to create a compile time flag that can be used
to force all <select> elements to be rendered as menu lists.

Theme stylesheet mechanism has been used to change the appearance of the elements.

Class QtStyleOptionWebComboBox has been introduced to provide to QStyle objects
the needed information to render <select multiple> comboboxes.

  • WebCore.pri:
  • WebCore.pro:
  • css/themeQtNoListboxes.css: Added. (select[size][multiple]):
  • dom/SelectElement.h: (WebCore::SelectElementData::usesMenuList):
  • platform/qt/QtStyleOptionWebComboBox.h: Added. (WebCore::QtStyleOptionWebComboBox::QtStyleOptionWebComboBox): (WebCore::QtStyleOptionWebComboBox::multiple): (WebCore::QtStyleOptionWebComboBox::checkMultiple):
  • platform/qt/RenderThemeQt.cpp: (WebCore::RenderThemeQt::extraDefaultStyleSheet): (WebCore::RenderThemeQt::paintMenuList): (WebCore::RenderThemeQt::paintMenuListButton):
  • platform/qt/RenderThemeQt.h:
18:58 Changeset [56291] by eric@webkit.org
  • 25 edits in trunk

2010-03-19 Zhenyao Mo <zmo@google.com>

Reviewed by Oliver Hunt.

Update WebGLArray.slice() to new spec
https://bugs.webkit.org/show_bug.cgi?id-35612

  • fast/canvas/webgl/array-unit-tests-expected.txt: Changed tests to reflect the new slice parameters; add new test cases for default inputs for slice.
  • fast/canvas/webgl/array-unit-tests.html: Ditto.
  • fast/canvas/webgl/webgl-array-invalid-ranges-expected.txt: Changed tests to reflect the new slice parameters.
  • fast/canvas/webgl/webgl-array-invalid-ranges.html: Ditto.

2010-03-19 Zhenyao Mo <zmo@google.com>

Reviewed by Oliver Hunt.

Update WebGLArray.slice() to new spec
https://bugs.webkit.org/show_bug.cgi?id-35612

  • bindings/js/JSWebGLArrayCustom.cpp: Adding support for default inputs. (WebCore::JSWebGLArray::slice):
  • bindings/v8/custom/V8WebGLArrayCustom.cpp: Ditto. (WebCore::V8WebGLArray::sliceCallback):
  • html/canvas/WebGLArray.cpp: Map start/end to offset/length. (WebCore::WebGLArray::calculateOffsetAndLength):
  • html/canvas/WebGLArray.h: Modified slice parameters.
  • html/canvas/WebGLArray.idl: Define custom binding for slice().
  • html/canvas/WebGLByteArray.cpp: Modified slice parameters. (WebCore::WebGLByteArray::slice):
  • html/canvas/WebGLByteArray.h: Ditto.
  • html/canvas/WebGLFloatArray.cpp: Ditto. (WebCore::WebGLFloatArray::slice):
  • html/canvas/WebGLFloatArray.h: Ditto.
  • html/canvas/WebGLIntArray.cpp: Ditto. (WebCore::WebGLIntArray::slice):
  • html/canvas/WebGLIntArray.h: Ditto.
  • html/canvas/WebGLShortArray.cpp: Ditto. (WebCore::WebGLShortArray::slice):
  • html/canvas/WebGLShortArray.h: Ditto.
  • html/canvas/WebGLUnsignedByteArray.cpp: Ditto. (WebCore::WebGLUnsignedByteArray::slice):
  • html/canvas/WebGLUnsignedByteArray.h: Ditto.
  • html/canvas/WebGLUnsignedIntArray.cpp: Ditto. (WebCore::WebGLUnsignedIntArray::slice):
  • html/canvas/WebGLUnsignedIntArray.h: Ditto.
  • html/canvas/WebGLUnsignedShortArray.cpp: Ditto. (WebCore::WebGLUnsignedShortArray::slice):
  • html/canvas/WebGLUnsignedShortArray.h: Ditto.
18:39 Changeset [56290] by eric@webkit.org
  • 1 edit
    4 adds in trunk/LayoutTests

2010-03-19 Adam Barth <abarth@webkit.org>

Reviewed by Darin Adler.

Freeze sandbox attributes on creation
https://bugs.webkit.org/show_bug.cgi?id=34184

Test that allow-forms is frozen on document creation.

  • fast/frames/resources/sandboxed-iframe-form-dynamic-allowed.html: Added.
  • fast/frames/resources/sandboxed-iframe-form-dynamic-disallowed.html: Added.
  • fast/frames/sandboxed-iframe-forms-dynamic-expected.txt: Added.
  • fast/frames/sandboxed-iframe-forms-dynamic.html: Added.
18:20 Changeset [56289] by eric@webkit.org
  • 2 edits
    2 adds in trunk/WebCore

2010-03-19 Yong Li <yong.li@torchmobile.com> and Andy Estes <aestes@apple.com>

Reviewed by Darin Adler.

Ensure the segments inside SharedBuffer are merged before constructing
a CFData with it.

https://bugs.webkit.org/show_bug.cgi?id=50843

  • manual-tests/WebKitSite.webarchive: Added.
  • manual-tests/webarchive-test.html: Added.
  • platform/cf/SharedBufferCF.cpp: (WebCore::SharedBuffer::createCFData):
17:45 Changeset [56288] by oliver@apple.com
  • 4 edits
    2 adds in trunk

2010-03-19 Justin Schuh <jschuh@chromium.org>

Reviewed by Oliver Hunt.

Security: ReadAV@NULL for negative feMorphology filter radius
https://bugs.webkit.org/show_bug.cgi?id=34566

Prevents building the filter if either radius is negative.
Catches negative radii resulting from conversions.

Test: svg/filters/feMorphology-invalid-radius.svg

  • svg/SVGFEMorphologyElement.cpp: (WebCore::SVGFEMorphologyElement::build):
  • svg/graphics/filters/SVGFEMorphology.cpp: (WebCore::FEMorphology::apply):
17:41 Changeset [56287] by dpranke@chromium.org
  • 5 edits in trunk/WebKitTools

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Disable the downstream override expectations temporarily to allow
us to test that we've upstreamed everything correctly. Also, stop
looking at the downstream baselines at all (now you will only be
able to update baselines upstream). In theory this should work, but
if we need to we can always add the downstream dirs back in.

https://bugs.webkit.org/show_bug.cgi?id=36396

  • Scripts/webkitpy/layout_tests/port/chromium.py:
  • Scripts/webkitpy/layout_tests/port/chromium_linux.py:
  • Scripts/webkitpy/layout_tests/port/chromium_mac.py:
  • Scripts/webkitpy/layout_tests/port/chromium_win.py:
17:28 Changeset [56286] by cmarrin@apple.com
  • 3 edits
    2 adds in trunk

Fixed a crash with AnimationController getting deleted out from under itself.


It's possible for the Frame that owns an AnimationController to get deleted
in the EndTransitionEvent (or other animation events) to get deleted in the
event handler. Normally this case is protected against by preventing the Frame
from getting deleted until the end of the runloop. But native uses of the
WebView can subvert this protection. So I added a protector to the
animation event dispatcher to protect it in those cases.

17:08 Changeset [56285] by eric@webkit.org
  • 4 edits in trunk/WebCore

2010-03-19 Luiz Agostini <luiz.agostini@openbossa.org>

Reviewed by Antti Koivisto.

Multiselect Popup - PopupMenuClient extension
https://bugs.webkit.org/show_bug.cgi?id=36178

PopupMenuClient class is the interface used by combobox popup implementations.
It needs to be extended to handle <select multiple> needs.

A new interface named ListPopupMenuClient that inherits from PopupMenuClient was created.
The use of this new class instead of adding methods to PopupMenuClient avoids changes
in other non related implementations of PopupMenuClient.

RenderMenuList has changed to inherit ListPopupMenuClient instead of PopupMenuClient
and to have the new methods implemented.

  • platform/PopupMenuClient.h:
  • rendering/RenderMenuList.cpp: (WebCore::RenderMenuList::listBoxSelectItem): (WebCore::RenderMenuList::multiple):
  • rendering/RenderMenuList.h:
16:42 Changeset [56284] by simon.fraser@apple.com
  • 4 edits
    2 adds in trunk

2010-03-19 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dimitri Glazkov.

https://bugs.webkit.org/show_bug.cgi?id=36387
REGRESSION: Can "uncheck" radio buttons by clicking on them

My change in r56174 caused radio buttons and checkboxes to go down the same code path by virtue
of allowsIndeterminate(), but this broke save/restore behavior on radio buttons.

Fix by saving and restoring both the intermediate and checked state for radio buttons
and checkboxes, as well as the "current radio button" state.

Test: fast/forms/radio-checkbox-restore-indeterminate.html

  • html/HTMLInputElement.cpp: (WebCore::EventHandlingState::EventHandlingState): New struct to hold state between preDispatchEventHandler() and postDispatchEventHandler(). (WebCore::HTMLInputElement::preDispatchEventHandler): Store intermedate and checked state and the current radio button. (WebCore::HTMLInputElement::postDispatchEventHandler): Restore state as appropriate.
  • html/HTMLInputElement.h: Add a comment to clarify how 'intermediate' interacts with 'checked', according to the spec.
16:38 Changeset [56283] by dpranke@chromium.org
  • 2 edits in trunk/WebKitTools

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

fix typo in chromium test expectations overrides routine
https://bugs.webkit.org/show_bug.cgi?id=36397

  • Scripts/webkitpy/layout_tests/port/chromium.py:
16:09 Changeset [56282] by darin@chromium.org
  • 2 edits in trunk/LayoutTests

2010-03-19 Darin Fisher <darin@chromium.org>

Temporarily suppress fast/loader/frame-src-change-added-to-history.html
for Tiger to make the bot green.

https://bugs.webkit.org/show_bug.cgi?id=36382

  • platform/mac-tiger/Skipped:
16:06 Changeset [56281] by dpranke@chromium.org
  • 1 add in trunk/LayoutTests/platform/chromium/test_expectations.txt

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Unreviewed change.

Add in missing test_expectations.txt file that didn't actually
make it into the commit.

https://bugs.webkit.org/show_bug.cgi?id=36326

  • LayoutTests/platform/chromium/test_expectations.txt
16:05 Changeset [56280] by eric@webkit.org
  • 3 edits in trunk/WebCore

2010-03-19 Alexander Pavlov <apavlov@chromium.org>

Reviewed by Pavel Feldman.

Web Inspector: caret moves past prompt in javascript console
https://bugs.webkit.org/show_bug.cgi?id=26602

Having noticed any Element children besides a leading <br> (which are empty
text nodes), WebCore editing facility thinks it should delete the <br>,
thus clearing away the text prompt contents.

  • inspector/front-end/TextPrompt.js: (WebInspector.TextPrompt.prototype.clearAutoComplete): (WebInspector.TextPrompt.prototype._completionsReady):
  • inspector/front-end/utilities.js: (Element.prototype.pruneEmptyTextNodes):
16:02 Changeset [56279] by dpranke@chromium.org
  • 3 edits in trunk

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Flip the Chromium ports to look first for the test expectations
in LayoutTests/platform/chromium and only afterwards look in the
Chromium repo downstream for overrides.

https://bugs.webkit.org/show_bug.cgi?id=36326

  • WebKitTools/Scripts/webkitpy/layout_tests/port/chromium.py:
  • LayoutTests/platform/chromium/test_expectations.txt
15:52 Changeset [56278] by dpranke@chromium.org
  • 1 edit
    2371 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.
Land remaining chromium-linux LayoutTest baselines
(file list omitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

15:43 Changeset [56277] by dpranke@chromium.org
  • 1 edit
    1491 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land chromium-linux/fast/{block,css,forms,text} LayoutTest baselines
(file list omitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-linux/fast/block : Added.
  • platform/chromium-linux/fast/css : Added.
  • platform/chromium-linux/fast/forms : Added.
  • platform/chromium-linux/fast/text : Added.
15:35 Changeset [56276] by dpranke@chromium.org
  • 1 edit
    1453 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land chromium-linux/editing LayoutTest baselines
(file list omitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-linux/editing : Added.
15:33 Changeset [56275] by jhawkins@chromium.org
  • 2 edits in trunk/WebKitTools

2010-03-19 James Hawkins <jhawkins@chromium.org>

Unreviewed.

Add myself to the committers list.

  • Scripts/webkitpy/committers.py:
15:29 Changeset [56274] by dpranke@chromium.org
  • 1 edit
    1893 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land chromium-linux/tables LayoutTest baselines
(file list omitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-linux/tables : Added.
15:18 Changeset [56273] by dpranke@chromium.org
  • 1 edit
    1251 adds in trunk/LayoutTests

010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land chromium-linux/svg LayoutTest baselines
(file list omitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-linux/svg : Added.
15:09 Changeset [56272] by dpranke@chromium.org
  • 1 edit
    1244 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land chromium-linux/css* LayoutTest baselines
(file list omitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-linux/css* : Added.
15:00 Changeset [56271] by dpranke@chromium.org
  • 1 edit
    2661 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land remaining chromium-win/* LayoutTest baselines
(file list omitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-win/* : Added.
14:50 Changeset [56270] by dpranke@chromium.org
  • 1 edit
    1811 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land chromium-win/css* LayoutTest baselines
(file list ommitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-win/css* : Added.
14:38 Changeset [56269] by ossy@webkit.org
  • 2 edits in trunk/LayoutTests

[Qt] Back-forward list dumping is incorrect
https://bugs.webkit.org/show_bug.cgi?id=36392

  • platform/qt/Skipped: fast/loader/frame-src-change-added-to-history.html skipped until fix.
14:34 Changeset [56268] by dpranke@chromium.org
  • 1 edit
    2436 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land chromium-win/fast/[l-z]* LayoutTest baselines
(file list ommitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-win/fast/[l-z]* : Added.
14:24 Changeset [56267] by dpranke@chromium.org
  • 1 edit
    2380 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land chromium-win/fast/{a,c,d,e,f,g,h,i,j}* LayoutTest baselines
(file list ommitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-win/fast/{a,c,d,e,f,g,h,i,j}* : Added.
14:14 Changeset [56266] by xan@webkit.org
  • 2 edits in trunk/LayoutTests

2010-03-19 Xan Lopez <xlopez@igalia.com>

Rubber-stamped by Gustavo Noronha.

r56188 removed lastPosition from the Geolocation code, but the the
geolocation window properties wasn't updated to reflect this. It's
only failing on the GTK+ port, since the other ports have it on
their Skipped list.

  • fast/dom/Window/window-properties-geolocation-expected.txt:
14:11 Changeset [56265] by ossy@webkit.org
  • 2 edits in trunk/LayoutTests

[Qt] REGRESSION (r56209): fast/media/print-restores-previous-mediatype.htm crashes
https://bugs.webkit.org/show_bug.cgi?id=36386

  • platform/qt/Skipped: fast/media/print-restores-previous-mediatype.html skipped until fix.
14:04 Changeset [56264] by dpranke@chromium.org
  • 1 edit
    1216 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land the first part of chromium-win/fast LayoutTest baselines
(file list ommitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-win/fast: Added.
14:01 Changeset [56263] by abarth@webkit.org
  • 3 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Unreviewed.

Fix SheriffBot exception lock when we can't retrieve the first build
from buildbot. (I'll ask Eric to review this change after the fact,
but he's at lunch and I want to get the bot unlocked.)

  • Scripts/webkitpy/buildbot.py:
  • Scripts/webkitpy/buildbot_unittest.py:
13:48 QtWebKitJournal edited by hausmann@webkit.org
(diff)
13:21 Changeset [56262] by ap@apple.com
  • 6 edits in trunk/WebKitTools

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=36380
websocket/tests/frame-lengths.html times out on Tiger bot

https://bugs.webkit.org/show_bug.cgi?id=35041
websocket/tests/frame-lengths.html / websocket/tests/simple-stress.html fail on Windows bot

Double the timeout (from 15 seconds to 30 seconds). We can increase it more, if necessary -
sampling the DRT process on Mac OS X takes much longer anyway, so it's better to avoid
timing out than to detect it early.

  • DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: (LayoutTestController::setWaitToDump):
  • DumpRenderTree/mac/LayoutTestControllerMac.mm:
  • DumpRenderTree/qt/LayoutTestControllerQt.cpp: (LayoutTestController::waitUntilDone):
  • DumpRenderTree/win/LayoutTestControllerWin.cpp:
  • Scripts/run-webkit-tests:
13:20 Changeset [56261] by xan@webkit.org
  • 2 edits in trunk/LayoutTests

2010-03-19 Xan Lopez <xlopez@igalia.com>

Unreviewed, land new expected result for test.

  • platform/gtk/svg/css/getComputedStyle-basic-expected.txt:
13:15 Changeset [56260] by laszlo.1.gombos@nokia.com
  • 4 edits in trunk

2010-03-19 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Unreviewed.

Buildfix for Qt v4.5.

  • Api/qgraphicswebview.cpp: (QGraphicsWebViewOverlay::q):

2010-03-19 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Unreviewed.

Buildfix for Qt v4.5.

  • QtLauncher/main.cpp: (LauncherWindow::LauncherWindow): (LauncherWindow::applyZoom):
13:10 Changeset [56259] by dpranke@chromium.org
  • 1 edit
    2855 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land the chromium-win/tables LayoutTest baselines (file list
ommitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-win/tables: Added.
13:09 Changeset [56258] by dglazkov@chromium.org
  • 2 edits in trunk/WebKit/chromium

Unreviewed, build fix.

Rename enqueueStorageEvent to enqueueEvent to match changes in
http://trac.webkit.org/changeset/56249.

  • src/StorageAreaProxy.cpp:

(WebCore::StorageAreaProxy::storageEvent): Renamed.

13:05 Changeset [56257] by abarth@webkit.org
  • 4 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Operational tweaks to SheriffBot
https://bugs.webkit.org/show_bug.cgi?id=36385

These changes aren't pretty, but they're helpful to make SheriffBot
work operationally. I plan to iterate in these areas, but I wanted to
get this patch landed so I could be running the bot against TOT.

  • Scripts/webkitpy/commands/sheriffbot.py:
  • Scripts/webkitpy/irc/ircbot.py:
  • Scripts/webkitpy/thirdparty/autoinstalled/init.py:
13:01 Changeset [56256] by aroben@apple.com
  • 3 edits in trunk/WebCore

Windows clean build fix after r56192

  • WebCore.vcproj/WebCore.vcproj: Copy headers from

platform/graphics/cg to $WebKitOutputDir as part of the post-build
event so that WebKit can use them. (The only one we currently require
is FontPlatformData.h.)

  • platform/graphics/win/FontPlatformDataCGWin.cpp: Touched to force a

build.

12:56 Changeset [56255] by dpranke@chromium.org
  • 2149 adds in trunk/LayoutTests/platform/chromium-win

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land the chromium-win/svg LayoutTest baselines (file list
ommitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-win/svg: Added.
12:54 Changeset [56254] by beidson@apple.com
  • 1 edit
    1 move in trunk/LayoutTests

Rubberstamped by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=35594

  • fast/loader/api-test-new-window-data-load-base-url.html: Removed.
  • fast/loader/api-test-new-window-data-load-base-url.html-disabled: Copied from fast/loader/api-test-new-window-data-load-base-url.html.
12:41 Changeset [56253] by dpranke@chromium.org
  • 1 edit
    103 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land the chromium-win-vista LayoutTest baselines (file list
ommitted due to length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-win-vista: Added.
12:40 Changeset [56252] by dpranke@chromium.org
  • 1 edit
    141 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land the chromium-win-xp LayoutTest baselines (file list ommitted due to
length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-win-xp: Added.
12:38 Changeset [56251] by dpranke@chromium.org
  • 1 edit
    831 adds in trunk/LayoutTests

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Land the chromium-mac LayoutTest baselines (file list ommitted due to
length).

https://bugs.webkit.org/show_bug.cgi?id=28397

  • platform/chromium-mac: Added.
11:37 Changeset [56250] by joepeck@webkit.org
  • 2 edits in trunk/WebCore

2010-03-19 Joseph Pecoraro <joepeck@webkit.org>

Reviewed by David Kilzer.

Assertion failure in media/video-controls-with-mutation-event-handler.html
https://bugs.webkit.org/show_bug.cgi?id=36376

Fix crashing test. Underlying the media element's volume slider is a
range input, which sanitizes values before expected on the media control.

  • rendering/RenderMedia.cpp: (WebCore::RenderMedia::createVolumeSlider):
11:25 Changeset [56249] by beidson@apple.com
  • 6 edits in trunk/WebCore

3 of the new HTML5 loading events need to be asynchronous.

Reviewed by Darin Adler.

Laying the groundwork for:
https://bugs.webkit.org/show_bug.cgi?id=36201
https://bugs.webkit.org/show_bug.cgi?id=36202
https://bugs.webkit.org/show_bug.cgi?id=36334
https://bugs.webkit.org/show_bug.cgi?id=36335

Document already had an asynchronous event delivery mechanism for storage events, so
we can repurpose that for all async events.

No new tests. (No change in behavior)

  • dom/Document.cpp:

(WebCore::Document::Document):
(WebCore::Document::implicitClose): Use Document::schedule* for the related events.
(WebCore::Document::enqueueEvent): Renamed from enqueueStorageEvent
(WebCore::Document::pendingEventTimerFired): Renamed from "storageEventTimerFired"
(WebCore::Document::statePopped): Use Document::schedulePopstateEvent
(WebCore::Document::enqueuePageshowEvent): All Pageshow events are piped through here.

This will be made asynchronous in a separate patch.

(WebCore::Document::enqueueHashchangeEvent): All Hashchange events are piped through here.

This will be made asynchronous in a separate patch.

(WebCore::Document::enqueuePopstateEvent): All Popstate events are piped through here.

This will be made asynchronous in a separate patch.

  • dom/Document.h:

(WebCore::):

  • history/CachedFrame.cpp:

(WebCore::CachedFrameBase::restore): Use Document::enqueuePageshowEvent

  • loader/FrameLoader.cpp:

(WebCore::FrameLoader::loadInSameDocument): Use Document::enqueueHashchangeEvent

  • storage/StorageEventDispatcher.cpp:

(WebCore::StorageEventDispatcher::dispatch): Use Document::enqueueEvent

11:23 Changeset [56248] by eric@webkit.org
  • 4 edits in trunk/WebKitTools

2010-03-19 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Help sheriff-bot avoid warning about flaky tests (and add more unit testing)
https://bugs.webkit.org/show_bug.cgi?id=36354

  • Scripts/webkitpy/buildbot.py:
    • Make Build creation easier to Mock and test
  • Scripts/webkitpy/buildbot_unittest.py:
    • Test finding green to red transitions and suspect revisions
  • Scripts/webkitpy/commands/queries.py:
    • Make what-broke note when builders have only failed once.
11:19 Changeset [56247] by ap@apple.com
  • 2 edits in trunk/WebKit/mac

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=36337
Log an error when an OOP plug-in sends an unknown object id

Making these LOG_ERROR and not ASSERTs, because I don't want early returns to look
temporary or redundant.

  • Plugins/Hosted/NetscapePluginInstanceProxy.mm: (WebKit::NetscapePluginInstanceProxy::evaluate): (WebKit::NetscapePluginInstanceProxy::invoke): (WebKit::NetscapePluginInstanceProxy::invokeDefault): (WebKit::NetscapePluginInstanceProxy::construct): (WebKit::NetscapePluginInstanceProxy::getProperty): (WebKit::NetscapePluginInstanceProxy::setProperty): (WebKit::NetscapePluginInstanceProxy::removeProperty): (WebKit::NetscapePluginInstanceProxy::hasProperty): (WebKit::NetscapePluginInstanceProxy::hasMethod): (WebKit::NetscapePluginInstanceProxy::enumerate):
11:16 Changeset [56246] by kdecker@apple.com
  • 4 edits in trunk/WebCore

Reviewed by Darin Adler and Brady Eidson.

First step toward:
https://bugs.webkit.org/show_bug.cgi?id=36328

This is the first step torward eliminating the WebKit WebNullPluginView class. The responsibility for this will soon
be in platform-independent code in WebCore. The plan is to change RenderEmbeddedObject and give it the capability of
drawing the missing plug-in text.

  • loader/FrameLoader.cpp: (WebCore::FrameLoader::requestObject): Renamed the renderer parameter to requestEmbeddedObject and made it a RenderEmbeddedObject. (WebCore::FrameLoader::loadPlugin): Likewise.
  • loader/FrameLoader.h: Updated loadPlugin() and requestObject() method signatures accordingly.
11:01 Changeset [56245] by dpranke@chromium.org
  • 2 edits in trunk/WebKitTools

2010-03-19 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Fix the rebaselining tool, which was broken by r36324 when I
added the concept of overridding expectations.

https://bugs.webkit.org/show_bug.cgi?id=36374

  • Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
10:30 Changeset [56244] by darin@chromium.org
  • 2 edits in trunk/WebKit/chromium

2010-03-19 Darin Fisher <darin@chromium.org>

Reviewed by Dimitri Glazkov.

[chromium] querying the current history item during a back/forward
navigation should not clobber the scroll offset, etc.
https://bugs.webkit.org/show_bug.cgi?id=36347

We should only allow the current history item to be modified if we are
no longer loading or if the load is a new navigation (i.e., not a
back/forward/reload variant).

  • src/WebFrameImpl.cpp: (WebKit::WebFrameImpl::currentHistoryItem):
09:41 Changeset [56243] by eric@webkit.org
  • 2 edits in trunk/WebCore

2010-03-19 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Simon Hausmann.

[Qt] Disable TILED_BACKING_STORE if Qt version is earlier than Qt4.6
https://bugs.webkit.org/show_bug.cgi?id=36348

Backing store implementation for QtWebKit requires at least Qt
version 4.6.

No new tests as there is no new functionality.

  • WebCore.pri:
09:16 Changeset [56242] by joepeck@webkit.org
  • 16 edits
    2 adds in trunk

2010-03-19 Joseph Pecoraro <joepeck@webkit.org>

Reviewed by David Kilzer.

<input type=range> does not validate correctly without a renderer and the tests are incorrect
https://bugs.webkit.org/show_bug.cgi?id=36259

Setting value attribute on an <input type=range> to an out-of-range value fires oninput
https://bugs.webkit.org/show_bug.cgi?id=16990

Part 2 of 2: When setting the range element's value, overflows and underflows
are automatically sanitized to valid values. Moved the general case
sanitization code out of the Renderer into HTMLInputElement::sanitizeValue.

  • html/HTMLInputElement.cpp: (WebCore::HTMLInputElement::value): when getting a default value on reset() ensure the defaultValue is provided (WebCore::HTMLInputElement::sanitizeValue): clamp the value within the max/min/step range constraints
  • html/StepRange.cpp: (WebCore::StepRange::StepRange): allow const element in the constructor (WebCore::StepRange::clampValue): clamp from a String value
  • html/StepRange.h: (WebCore::StepRange::defaultValue): easy calculation of the default value for max/min/step range constraints
  • rendering/RenderSlider.cpp: (WebCore::RenderSlider::updateFromElement): no longer handle the general case sanitization in the renderer

2010-03-19 Joseph Pecoraro <joepeck@webkit.org>

Reviewed by David Kilzer.

<input type=range> does not validate correctly without a renderer and the tests are incorrect
https://bugs.webkit.org/show_bug.cgi?id=36259

Setting value attribute on an <input type=range> to an out-of-range value fires oninput
https://bugs.webkit.org/show_bug.cgi?id=16990

When setting the range element's value, overflows and underflows are
automatically sanitized to valid values. Fixed the behavior and removed
tests that were testing for improper behavior.

  • fast/forms/ValidityState-rangeOverflow-expected.txt:
  • fast/forms/ValidityState-rangeUnderflow-expected.txt:
  • fast/forms/input-stepup-stepdown-expected.txt:
  • fast/forms/script-tests/ValidityState-rangeOverflow.js:
  • fast/forms/script-tests/ValidityState-rangeUnderflow.js:
  • fast/forms/script-tests/input-stepup-stepdown.js:
  • fast/forms/script-tests/validationMessage.js:
  • fast/forms/validationMessage-expected.txt:

Added test verifying that bug 16990 was fixed as well.

  • fast/forms/range-input-dynamic-oninput-expected.txt: Added.
  • fast/forms/range-input-dynamic-oninput.html Added.

Improved test because handling of this case changed.

  • fast/forms/range-reset-expected.txt:
  • fast/forms/range-reset.html:
09:16 Changeset [56241] by joepeck@webkit.org
  • 7 edits
    2 adds in trunk/WebCore

2010-03-19 Joseph Pecoraro <joepeck@webkit.org>

Reviewed by David Kilzer.

<input type=range> does not validate correctly without a renderer and the tests are incorrect
https://bugs.webkit.org/show_bug.cgi?id=36259

Part 1 of 2: Refactoring the SliderRange struct out of RenderSlider
into a more appropriate place. Changed the named to StepRange. Changed
from a struct to a class.

Added new files to the build.

  • GNUmakefile.am:
  • WebCore.gypi:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:

Renamed and moved class SliderRange to StepRange.

  • html/StepRange.cpp: Added. (WebCore::StepRange::StepRange): (WebCore::StepRange::clampValue): (WebCore::StepRange::valueFromElement): (WebCore::sliderPosition):
  • html/StepRange.h: Added. (WebCore::StepRange::proportionFromValue): (WebCore::StepRange::valueFromProportion):
  • rendering/RenderSlider.cpp: (WebCore::RenderSlider::updateFromElement): updated to use StepRange (WebCore::RenderSlider::setValueForPosition): updated to use StepRange
08:34 QtWebKitJournal edited by hausmann@webkit.org
(diff)
08:33 Changeset [56240] by darin@chromium.org
  • 2 edits in trunk/LayoutTests

2010-03-19 Darin Fisher <darin@chromium.org>

Fix expected results for fast/loader/frame-src-change-added-to-history.html

  • fast/loader/frame-src-change-added-to-history-expected.txt:
08:27 Changeset [56239] by hausmann@webkit.org
  • 2 edits in trunk/WebCore

https://bugs.webkit.org/show_bug.cgi?id=36366
Repaint loop when painting using scaled CTMs with different translations

Patch by Antti Koivisto <koivisto@iki.fi> on 2010-03-19
Reviewed by Simon Hausmann.

The code in RenderBoxModelScaleObserver::shouldPaintBackgroundAtLowQuality
tests if the scaling has changed from the previous value by comparing the
transformations. The test fails if the scale is the same but the translation
changes. This can lead to infinite repaint loop if the document is painted
in pieces using different translations (for example for tiling).

  • rendering/RenderBoxModelObject.cpp:

(WebCore::RenderBoxModelScaleObserver::shouldPaintBackgroundAtLowQuality):

08:06 QtWebKitBackportingFixes edited by laszlo.1.gombos@nokia.com
(diff)
08:03 Changeset [56238] by hamaji@chromium.org
  • 1 edit
    5 adds in trunk/LayoutTests

2010-03-19 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by Darin Adler.

Need a layout test for finding yensign from documents with Japanese encodings
https://bugs.webkit.org/show_bug.cgi?id=36288

  • editing/selection/find-yensign-and-backslash-expected.txt: Added.
  • editing/selection/find-yensign-and-backslash.html: Added.
  • platform/gtk/editing/selection/find-yensign-and-backslash-expected.txt: Added.
  • platform/qt/editing/selection/find-yensign-and-backslash-expected.txt: Added.
  • platform/win/editing/selection/find-yensign-and-backslash-expected.txt: Added.
07:10 QtWebKitJournal edited by hausmann@webkit.org
(diff)
06:35 Changeset [56237] by pfeldman@chromium.org
  • 5 edits in trunk/WebCore

2010-03-19 Pavel Feldman <pfeldman@chromium.org>

Reviewed by Timothy Hatcher.

Web Inspector: editing shorthands does not always work.

https://bugs.webkit.org/show_bug.cgi?id=36362

  • inspector/InspectorDOMAgent.cpp: (WebCore::InspectorDOMAgent::applyStyleText): (WebCore::InspectorDOMAgent::populateObjectWithStyleProperties): (WebCore::InspectorDOMAgent::shorthandValue):
  • inspector/InspectorDOMAgent.h:
  • inspector/front-end/StylesSidebarPane.js: (WebInspector.StylePropertyTreeElement.prototype):
  • inspector/front-end/inspector.css: (.section .properties li.disabled .enabled-button):
  • inspector/front-end/inspector.js: (WebInspector.startEditing.editingCommitted):
06:10 Changeset [56236] by vestbo@webkit.org
  • 2 edits in trunk/WebCore

[Qt] Don't replace clip when drawing complex text

Reviewed by Simon Hausmann.

  • platform/graphics/qt/FontQt.cpp:
03:05 Changeset [56235] by abarth@webkit.org
  • 4 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Misc bug fixes to make the SheriffBot actually work
https://bugs.webkit.org/show_bug.cgi?id=36355

With these changes, I can actually run the sheriff-bot from start to
finish.

  • Scripts/webkitpy/irc/ircproxy.py:
  • Scripts/webkitpy/patch/patcher.py:
  • Scripts/webkitpy/statusserver.py:
02:51 Changeset [56234] by abarth@webkit.org
  • 3 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Unreviewed.

Actually import the sheriff-bot command so we can run it. Also, move
the bot to #webkit-test so it doesn't cause a ruckus while we test it.

  • Scripts/webkitpy/patch/patcher.py:
  • Scripts/webkitpy/irc/ircbot.py:
02:44 Changeset [56233] by abarth@webkit.org
  • 5 edits
    1 copy in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Second cut at SheriffBot
https://bugs.webkit.org/show_bug.cgi?id=36353

This patch should contain a complete SheriffBot that's capable of
saying reasonable things on IRC. I had to refactor the use of
CommitInfo to make the SheriffBot testable, but I did the minimum
necessary. We should grow webkitcheckout over time to contain the
knowledge of ChangeLogs from scm.

  • Scripts/webkitpy/commands/sheriffbot.py:
  • Scripts/webkitpy/commands/sheriffbot_unittest.py:
  • Scripts/webkitpy/mock_bugzillatool.py:
  • Scripts/webkitpy/patch/patcher.py:
  • Scripts/webkitpy/webkitcheckout.py: Added.
01:44 Changeset [56232] by abarth@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Unreviewed.

Fix Hyatt's IRC nickname.

  • Scripts/webkitpy/committers.py:
01:42 Changeset [56231] by eric@webkit.org
  • 3 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Unreviewed.

Add IRC nicknames for the non-reviewer committers.

  • Scripts/webkitpy/committers.py:
01:10 Changeset [56230] by abarth@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Unreviewed.

Add IRC nicknames for the non-reviewer committers.

  • Scripts/webkitpy/committers.py:
00:59 Changeset [56229] by abarth@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Unreviewed.

More reviewer IRC nicknames.

  • Scripts/webkitpy/committers.py:
00:53 Changeset [56228] by abarth@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Unreviewed.

Add a bunch of IRC nicknames for reviewers.

  • Scripts/webkitpy/committers.py:
00:47 Changeset [56227] by zoltan@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-19 Zoltan Horvath <zoltan@webkit.org>

Reviewed by Oliver Hunt.

Added USE_SYSTEM_MALLOC flag to build-webkit
https://bugs.webkit.org/show_bug.cgi?id=21272

Add system-alloc flag to build-webkit. It makes easy to switch
between system allocator and TCmalloc.

  • Scripts/build-webkit:
00:44 Changeset [56226] by abarth@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Unreviewed "build" fix.

Of course, I caused a regression in the file that isn't tested. :(

  • Scripts/webkitpy/statusserver.py:
00:37 Changeset [56225] by eric@webkit.org
  • 3 edits in trunk/WebKitTools

2010-03-19 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Simplify BuildBot core builder code for easier re-use
https://bugs.webkit.org/show_bug.cgi?id=36350

I simply couldn't see anything through all this Yak-hair.

  • Scripts/webkitpy/buildbot.py:
  • Scripts/webkitpy/commands/queries.py:
00:33 Changeset [56224] by abarth@webkit.org
  • 3 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

committers.py should know IRC nicknames
https://bugs.webkit.org/show_bug.cgi?id=36349

I'll add the actual nicknames in another patch.

  • Scripts/webkitpy/committers.py:
  • Scripts/webkitpy/committers_unittest.py:
00:31 Changeset [56223] by darin@chromium.org
  • 6 edits
    2 adds in trunk

2010-03-18 Darin Fisher <darin@chromium.org>

Reviewed by Adam Barth.

If a frame is already in the document, then setting its "src" attribute
should add the resultant navigation to history.

https://bugs.webkit.org/show_bug.cgi?id=9166

This behavior is specified in section 4.8.2 of the HTML5 spec.

Test: fast/loader/frame-src-change-added-to-history.html

  • html/HTMLFrameElementBase.cpp: (WebCore::HTMLFrameElementBase::openURL): (WebCore::HTMLFrameElementBase::setLocation):
  • html/HTMLFrameElementBase.h:
  • loader/FrameLoader.cpp: (WebCore::FrameLoader::requestFrame):
  • loader/FrameLoader.h:
00:20 Changeset [56222] by hausmann@webkit.org
  • 2 edits in trunk/WebKitTools

https://bugs.webkit.org/show_bug.cgi?id=36318

Patch by Anders Bakken <Anders Bakken <agbakken@gmail.com>> on 2010-03-18
Reviewed by Simon Hausmann.

QtLauncher runs as a GuiClient by default in Qt Embedded which will
make it try to connect to an existing GuiServer. This patch makes it
run like a stand-alone app.

  • QtLauncher/main.cpp:

(LauncherApplication::LauncherApplication):

00:18 Changeset [56221] by ossy@webkit.org
  • 2 edits in trunk/LayoutTests

Unreviewed.

[Qt] Platform specific expected file updated after r56195

  • platform/qt/fast/dom/Window/window-properties-expected.txt: updated.
00:15 Changeset [56220] by abarth@webkit.org
  • 3 edits in trunk/WebKitTools

2010-03-19 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Move find_green_to_red_transition out of "what-broke" onto Builder for easier re-use
https://bugs.webkit.org/show_bug.cgi?id=36345

  • Scripts/webkitpy/buildbot.py:
  • Scripts/webkitpy/commands/queries.py:
00:04 Changeset [56219] by abarth@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-19 Adam Barth <abarth@webkit.org>

Reviewed by Adam Barth.

Actually pass the IRC password to the IRC object
https://bugs.webkit.org/show_bug.cgi?id=36346

I wanted to do this before, but both patches were in flight. This
patch finally closes the loop and makes the IRCProxy system complete.

  • Scripts/webkitpy/patch/patcher.py:

03/18/10:

23:58 Changeset [56218] by abarth@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-18 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Add a StatusServer front end to the SVNRevision table on QueueStatusServer
https://bugs.webkit.org/show_bug.cgi?id=36344

No test because Browser was too hard to mock. :( I couldn't figure
out how to make Mock be a dictionary as well as an object.

  • Scripts/webkitpy/statusserver.py:
23:46 Changeset [56217] by eric@webkit.org
  • 4 edits in trunk/WebKit/gtk

2010-03-18 Philip Chimento <philip.chimento@gmail.com>

Reviewed by Oliver Hunt.

Setting the GObject WebKitWebView property 'window-features' to NULL
causes a crash.
https://bugs.webkit.org/show_bug.cgi?id=36144

  • tests/testwebview.c: Add unit test for this bug.
  • webkit/webkitwebview.cpp: Don't allow the 'window-features' property to be set to NULL.
  • webkit/webkitwebwindowfeatures.cpp: (webkit_web_window_features_equal): Don't examine the members of either web_window_features argument if either is NULL, just return that they are not equal. Additionally, if they are the same object, return that they are equal.
23:46 Changeset [56216] by eric@webkit.org
  • 3 edits
    2 adds in trunk/WebKitTools

2010-03-18 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Split out CommitInfo class and add unit tests
https://bugs.webkit.org/show_bug.cgi?id=36343

Move more logic out of "what-broke" into a shared CommitInfo
class so that it can be used by other commands and unit tested.

  • Scripts/webkitpy/commands/queries.py:
  • Scripts/webkitpy/commitinfo.py: Added.
  • Scripts/webkitpy/commitinfo_unittest.py: Added.
  • Scripts/webkitpy/unittests.py:
23:29 Changeset [56215] by eric@webkit.org
  • 2 edits in trunk/JavaScriptCore

2010-03-18 Tom Callaway <tcallawa@redhat.com>

Reviewed by Darin Adler.

Bug 35429: Fix compile on SPARC64
https://bugs.webkit.org/show_bug.cgi?id=35429

  • wtf/Platform.h: Set WTF_USE_JSVALUE64 for SPARC64
23:24 Changeset [56214] by oliver@apple.com
  • 3 edits
    2 adds in trunk

WebCore: Fix a minor crash with mismatched array sizes in SVG animation
elements.

Patch by Chris Evans <cevans@chromium.org> on 2010-03-18
Reviewed by Nikolas Zimmermann.

https://bugs.webkit.org/show_bug.cgi?id=35606

Test: svg/animations/keypoints-mismatch.svg

  • svg/SVGAnimationElement.cpp:

(WebCore::SVGAnimationElement::startedActiveInterval):
Globally apply validations relevant to all animation types.

LayoutTests: Add test for minor SVG animation crashers.
See https://bugs.webkit.org/show_bug.cgi?id=35606

Patch by Chris Evans <cevans@chromium.org> on 2010-03-18
Reviewed by Nikolas Zimmermann.

  • svg/animations/keypoints-mismatch.svg: Added.
  • svg/animations/keypoints-mismatch-expected.txt: Added.
23:03 Changeset [56213] by eric@webkit.org
  • 2 edits
    2 adds in trunk/LayoutTests

2010-03-18 Ojan Vafai <ojan@chromium.org>

Reviewed by Eric Seidel.

add new results for failing win/linux test after http://trac.webkit.org/changeset/56175
https://bugs.webkit.org/show_bug.cgi?id=36329

Missed this test in r56175 since it does a drag on win/linux, but not a mac.
The new results correctly do the smart-drag behavior of doing a smart-delete
on drag regardless of whether the selection granularity is word granularity.

  • platform/win/editing/selection/drag-text-delay-expected.checksum: Added.
  • platform/win/editing/selection/drag-text-delay-expected.png: Added.
  • platform/win/editing/selection/drag-text-delay-expected.txt:
22:46 Changeset [56212] by eric@webkit.org
  • 10 edits
    6 adds in trunk

2010-03-18 Vangelis Kokkevis <vangelis@chromium.org>

Reviewed by Adam Barth.

Implement software composited graphics layers in Chromium using the Skia library.
https://bugs.webkit.org/show_bug.cgi?id=35557
This is an initial step in the implementation. Layer compositing is functioning
but not optimized in any way. Subsesquent check-ins will be necessary to fine tune
it.

Test: No new exposed functionality so no new tests.

  • WebCore.gypi: Added new source files to the chromium build
  • platform/graphics/GraphicsLayer.h: Added necessary typedef's and forward declarations for Chromium.
  • platform/graphics/skia/GraphicsLayerSkia.cpp: Added.
  • platform/graphics/skia/GraphicsLayerSkia.h: Added. Declaration and implementation of the platform-specific GraphicsLayer class.
  • platform/graphics/skia/LayerRendererSkia.cpp: Added.
  • platform/graphics/skia/LayerRendererSkia.h: Added. Declaration and implementation of the Skia-based software compositor.
  • platform/graphics/skia/LayerSkia.cpp: Added.
  • platform/graphics/skia/LayerSkia.h: Added. Declaration and implementation of a compositable layer that uses a Skia canvas for backing store.

2010-03-18 Vangelis Kokkevis <vangelis@chromium.org>

Reviewed by Adam Barth.

Adding support for the ACCELERATED_COMPOSITING render path to Chromium.
https://bugs.webkit.org/show_bug.cgi?id=35557
Currently compositing of layers is performed via s/w compositor which relies on Skia. This is an initial check-in
and it's only been tested on Windows. Compiling the code requires seting "use_accelerated_compositing=1"
to GYP_DEFINES. The update of layer contents and compositing is fairly inefficient but this will be fixed in
subsequent check-ins.

  • src/ChromeClientImpl.cpp: (WebKit::ChromeClientImpl::attachRootGraphicsLayer): (WebKit::ChromeClientImpl::scheduleCompositingLayerSync):
  • src/ChromeClientImpl.h: (WebKit::ChromeClientImpl::setNeedsOneShotDrawingSynchronization): Added methods required by the RenderLayerCompositor
  • src/WebFrameImpl.cpp: (WebKit::WebFrameImpl::paintWithContext): (WebKit::WebFrameImpl::paint):
  • src/WebFrameImpl.h: Split WebFrameImpl::paint() into two methods to make it possible to call the paint routine with an existing GraphicsContext, which is necessary for painting the root layer into its own backing surface.
  • src/WebViewImpl.cpp: (WebKit::WebViewImpl::WebViewImpl): (WebKit::WebViewImpl::paint): Modified method to handle the accelerated compositing path. Now, when doing accelerated compositing, paint() results in repainting the contents of the root layer and then doing a composite operation. (WebKit::WebViewImpl::setRootGraphicsLayer): (WebKit::WebViewImpl::setAcceleratedCompositing): (WebKit::WebViewImpl::updateRootLayerContents): (WebKit::WebViewImpl::setRootLayerNeedsDisplay):
  • src/WebViewImpl.h: (WebKit::WebViewImpl::isAcceleratedCompositing):
22:32 Changeset [56211] by laszlo.1.gombos@nokia.com
  • 5 edits in trunk

2010-03-18 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Fix LayoutTests/http/tests/appcache/max-size.html
https://bugs.webkit.org/show_bug.cgi?id=36207

  • platform/qt/Skipped:

2010-03-18 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

[Qt] Fix LayoutTests/http/tests/appcache/max-size.html
https://bugs.webkit.org/show_bug.cgi?id=36207

Implement setAppCacheMaximumSize() for Qt.

  • DumpRenderTree/qt/LayoutTestControllerQt.cpp: (LayoutTestController::setAppCacheMaximumSize):
  • DumpRenderTree/qt/LayoutTestControllerQt.h:
22:13 Changeset [56210] by abarth@webkit.org
  • 6 edits
    4 copies
    1 add in trunk/WebKitTools

2010-03-18 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

QueueStatusServer should be able to log broken bots
https://bugs.webkit.org/show_bug.cgi?id=36341

We need to add a new table to the QueueStatusServer to store persistent
information for the SheriffBot. The new table will keep track of which
bots each SVN revision broke.

  • QueueStatusServer/handlers/init.py:
  • QueueStatusServer/handlers/svnrevision.py: Added.
  • QueueStatusServer/handlers/updatebase.py: Added.
  • QueueStatusServer/handlers/updatestatus.py:
  • QueueStatusServer/handlers/updatesvnrevision.py: Added.
  • QueueStatusServer/index.yaml:
  • QueueStatusServer/main.py:
  • QueueStatusServer/model/init.py:
  • QueueStatusServer/model/svnrevision.py: Added.
  • QueueStatusServer/templates/updatesvnrevision.html: Added.
22:11 Changeset [56209] by eric@webkit.org
  • 15 edits
    2 adds in trunk

2010-03-18 Kenneth Rohde Christiansen <kenneth@webkit.org>

Reviewed by Darin Adler.

Add a new test for testing whether the media type is restores
properly after printing a page (changing media type to 'print').

Add expected result and skip the test for platform that do not
implement setMediaType().

  • fast/media/print-restores-previous-mediatype-expected.txt: Added.
  • fast/media/print-restores-previous-mediatype.html: Added.
  • platform/gtk/Skipped:
  • platform/mac/Skipped:
  • platform/win/Skipped:

2010-03-18 Kenneth Rohde Christiansen <kenneth@webkit.org>

Reviewed by Darin Adler.

Make setPrinting() with printing=false, restore the previous
media type in use.

Test: fast/media/print-restores-previous-mediatype.html

  • page/Frame.cpp: (WebCore::Frame::setPrinting):
  • page/FrameView.cpp: (WebCore::FrameView::adjustMediaTypeForPrinting):
  • page/FrameView.h:

2010-03-18 Kenneth Rohde Christiansen <kenneth@webkit.org>

Reviewed by Darin Adler.

Make it possible for the Qt DRT to set the media type from
the LayoutTestController.

  • Api/qwebframe.cpp: (qt_drt_setMediaType):

2010-03-18 Kenneth Rohde Christiansen <kenneth@webkit.org>

Reviewed by Darin Adler.

Add a new method to the Qt LayoutTestController for
changing media type and make the DRT support dry-run printing.

  • DumpRenderTree/qt/DumpRenderTreeQt.cpp: (WebCore::DumpRenderTree::DumpRenderTree): (WebCore::DumpRenderTree::dryRunPrint):
  • DumpRenderTree/qt/DumpRenderTreeQt.h:
  • DumpRenderTree/qt/LayoutTestControllerQt.cpp: (LayoutTestController::setMediaType):
  • DumpRenderTree/qt/LayoutTestControllerQt.h:
21:51 Changeset [56208] by eric@webkit.org
  • 5 edits in trunk/WebKit/qt

2010-03-18 Joe Ligman <joseph.ligman@nokia.com>

Reviewed by Simon Hausmann.

[Qt] New API scrollRecursively has several problems.
https://bugs.webkit.org/show_bug.cgi?id=35873

Remove scrollRecursively from the Qt 4.7 API
Update the internal API to accept a hit test position
for nested scrolling

  • Api/qwebframe.cpp: (webframe_scrollOverflow): (qtwebkit_webframe_scrollRecursively):
  • Api/qwebframe.h:
  • Api/qwebframe_p.h:
  • tests/qwebframe/tst_qwebframe.cpp:
20:58 Changeset [56207] by mrowe@apple.com
  • 5 edits in trunk/LayoutTests

Update test results after r56182. "background-color: rgba(0, 0, 0, 0)" is now reported as "background-color: transparent".

Rubber-stamped by Samuel Weinig.

  • fast/css/getComputedStyle/computed-style-expected.txt:
  • platform/mac/fast/css/getComputedStyle/computed-style-expected.txt:
  • platform/qt/fast/css/getComputedStyle/computed-style-expected.txt:
  • platform/win/fast/css/getComputedStyle/computed-style-expected.txt:
18:33 Changeset [56206] by ddkilzer@apple.com
  • 2 edits in trunk/WebCore

<http://webkit.org/b/36338> Remove unused RenderReplaced::adjustOverflowForBoxShadowAndReflect() declaration

Rubber-stamped by Darin Adler.

This declaration should have been removed in r47440.

  • rendering/RenderReplaced.h:

(WebCore::RenderReplaced::adjustOverflowForBoxShadowAndReflect): Removed.

18:05 Changeset [56205] by jianli@chromium.org
  • 4 edits
    2 adds in trunk

Dragging a PDF image triggers assertion in DragController::startDrag()
https://bugs.webkit.org/show_bug.cgi?id=36247

Reviewed by Dmitry Titov.

WebCore:

Test: fast/images/drag-pdf-as-image.html

  • platform/graphics/cg/PDFDocumentImage.cpp:

(WebCore::PDFDocumentImage::filenameExtension):

  • platform/graphics/cg/PDFDocumentImage.h:

LayoutTests:

  • fast/images/drag-pdf-as-image-expected.txt: Added.
  • fast/images/drag-pdf-as-image.html: Added.
17:21 Changeset [56204] by simon.fraser@apple.com
  • 2 edits in trunk/WebCore

2010-03-18 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein.

Avoid taking a reference to a String owned by a temporary AtomicString.

  • css/CSSSelector.cpp: (WebCore::CSSSelector::RareData::parseNth):
17:17 Changeset [56203] by hyatt@apple.com
  • 2 edits in trunk/LayoutTests/fast

Land new expected results.

16:23 Changeset [56202] by mitz@apple.com
  • 1 edit
    1 add in trunk/LayoutTests

Add missing file from r56186.

  • fast/dynamic/float-remove-above-line-2-expected.txt: Added.
16:03 Changeset [56201] by jianli@chromium.org
  • 6 edits in trunk/WebKit/chromium

Chromium interface change to support Blob.slice.
https://bugs.webkit.org/show_bug.cgi?id=35318

Reviewed by Darin Fisher.

  • features.gypi:
  • public/WebHTTPBody.h:
  • public/WebKitClient.h:

(WebKit::WebKitClient::getFileModificationTime):

  • src/ChromiumBridge.cpp:

(WebCore::ChromiumBridge::getFileModificationTime):

  • src/WebHTTPBody.cpp:

(WebKit::WebHTTPBody::elementAt):
(WebKit::WebHTTPBody::appendFile):
(WebKit::WebHTTPBody::appendFileRange):

16:00 Changeset [56200] by cmarrin@apple.com
  • 4 edits in trunk/WebCore

Fixed parenting issues with HW layers on Windows
https://bugs.webkit.org/show_bug.cgi?id=32449

15:45 Changeset [56199] by dpranke@chromium.org
  • 4 edits in trunk/WebKitTools

2010-03-18 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Add upstream LayoutTests/platform/chromium* directories to the
baseline search path for new-run-webkit-tests in preparation for
upstreaming all of the Chromium baselines. Note that this does
not actually create the directories themselves, but that's okay.

https://bugs.webkit.org/show_bug.cgi?id=36324

  • Scripts/webkitpy/layout_tests/port/chromium_linux.py:
  • Scripts/webkitpy/layout_tests/port/chromium_mac.py:
  • Scripts/webkitpy/layout_tests/port/chromium_win.py:
15:42 Changeset [56198] by ap@apple.com
  • 3 edits
    2 adds in trunk

Reviewed by Anders Carlsson.

https://bugs.webkit.org/show_bug.cgi?id=36327
Test that a plug-in can override Node methods of its element

  • plugins/override-node-method-expected.txt: Added.
  • plugins/override-node-method.html: Added.
15:42 Changeset [56197] by dglazkov@chromium.org
  • 2 edits in trunk/WebCore

Unreviewed, build fix.

[Chromium] Change the order of mac/ and cg/ include dirs to fix compile break
due to http://trac.webkit.org/changeset/56192,

  • WebCore.gyp/WebCore.gyp: Changed the order.
14:51 Changeset [56196] by simon.fraser@apple.com
  • 2 edits in trunk/WebCore

2010-03-18 Simon Fraser <simon.fraser@apple.com>

No review.

Fix brace style after r56170.

  • css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::SelectorChecker::checkOneSelector):
14:39 Changeset [56195] by hyatt@apple.com
  • 5 edits
    2 adds in trunk

https://bugs.webkit.org/show_bug.cgi?id=36325

Reviewed by Oliver Hunt.

Add support for stopImmediatePropagation from DOM Level 3 Events.

Added fast/events/stop-immediate-propagation.html.

  • dom/Event.cpp:

(WebCore::Event::Event):

  • dom/Event.h:

(WebCore::Event::timeStamp):
(WebCore::Event::stopPropagation):
(WebCore::Event::stopImmediatePropagation):
(WebCore::Event::propagationStopped):
(WebCore::Event::immediatePropagationStopped):

  • dom/Event.idl:
  • dom/EventTarget.cpp:

(WebCore::EventTarget::fireEventListeners):

14:26 Changeset [56194] by dglazkov@chromium.org
  • 2 edits in trunk/WebCore

Unreviewed, build fix.

Return frame accessor back to Geolocation, it's used in GeolocationServiceChromium.
The accessor was removed in http://trac.webkit.org/changeset/56188.

  • page/Geolocation.h:

(WebCore::Geolocation::frame): Added back.

14:21 Changeset [56193] by dpranke@chromium.org
  • 4 edits
    1 add in trunk/WebKitTools

2010-03-17 Dirk Pranke <dpranke@chromium.org>

Reviewed by Dimitri Glazkov.

Add the concept of an "overrides" file for expectations so that we
can store test_expectations both upstream and downstream for a port
that runs both in webkit.org and in a separate repository (like
Chromium). Also add some unit tests for the expectations module.

https://bugs.webkit.org/show_bug.cgi?id=36249

  • Scripts/webkitpy/layout_tests/layout_package/test_expectations.py:
  • Scripts/webkitpy/layout_tests/layout_package/test_expectations_test.py: Added.
  • Scripts/webkitpy/layout_tests/port/base.py:
  • Scripts/webkitpy/layout_tests/run_webkit_tests.py:
14:20 Changeset [56192] by bfulgham@webkit.org
  • 5 edits
    1 move
    1 add in trunk/WebCore

Help reduce build problems due to font declarations.
https://bugs.webkit.org/show_bug.cgi?id=36190

Reviewed by Adam Roben.

Functions unchanged, no new tests.

  • WebCore.vcproj/WebCore.vcproj: Add new file, and show movement of FontPlatformData.h to cg/ directory.
  • platform/graphics/cairo/FontPlatformData.h: Merge in WinCairo-specific declarations. Correct constructor declaration to use type float for font size.

(WebCore::FontPlatformData::FontPlatformData):
(WebCore::FontPlatformData::hfont):
(WebCore::FontPlatformData::useGDI):
(WebCore::FontPlatformData::fontFace):
(WebCore::FontPlatformData::setSize):
(WebCore::FontPlatformData::hash):
(WebCore::FontPlatformData::isHashTableDeletedValue):

  • platform/graphics/cairo/FontPlatformDataCairo.cpp:

(WebCore::FontPlatformData::FontPlatformData): Correct

constructor to accept type float for font size.

  • platform/graphics/cg/FontPlatformData.h: Copied from WebCore/platform/graphics/win/FontPlatformData.h.

(WebCore::FontPlatformData::FontPlatformData):
(WebCore::FontPlatformData::hfont):
(WebCore::FontPlatformData::cgFont):
(WebCore::FontPlatformData::operator==):

  • platform/graphics/win/FontPlatformData.h: Removed.
  • platform/graphics/win/FontPlatformDataCairoWin.cpp:

(WebCore::FontPlatformData::operator==): Move implementation

here (previously in header) to match other cairo-based ports.

  • platform/graphics/win/RefCountedHFONT.h: Added. This was extracted from win/FontPlatformData.h, and is now shared by the cg/FontPlatformData.h and cairo/FontPlatformData.h

(WebCore::RefCountedHFONT::create):
(WebCore::RefCountedHFONT::createDeleted):
(WebCore::RefCountedHFONT::~RefCountedHFONT):
(WebCore::RefCountedHFONT::hfont):
(WebCore::RefCountedHFONT::hash):
(WebCore::RefCountedHFONT::RefCountedHFONT):

14:14 Changeset [56191] by simon.fraser@apple.com
  • 3 edits
    2 adds in trunk

2010-03-18 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dave Hyatt.

https://bugs.webkit.org/show_bug.cgi?id=36274
CSS3 :nth-child selector 'odd' keyword case sensitivity test fails

The tests for "odd", "even" etc. for nth-child selectors should be case insensitive.

Test: fast/css/nth-child-odd-case-insensitive.html

  • css/CSSSelector.cpp: (WebCore::CSSSelector::RareData::parseNth):
14:10 Changeset [56190] by hyatt@apple.com
  • 2 edits
    2 adds in trunk

https://bugs.webkit.org/show_bug.cgi?id=36265

Reviewed by Oliver Hunt.

Add support for event.defaultPrevented from DOM level 3.

Added fast/events/defaultprevented.html

  • dom/Event.idl:
13:51 Changeset [56189] by oliver@apple.com
  • 7 edits in trunk/JavaScriptCore

2010-03-18 Oliver Hunt <oliver@apple.com>

Reviewed by Sam Weinig.

Add API to directly expose JSON parsing
https://bugs.webkit.org/show_bug.cgi?id=34887

Add API to expose JSON parsing directly, and add tests to testapi

  • API/JSValueRef.cpp: (JSValueMakeFromJSONString): (JSValueCreateJSONString):
  • API/tests/testapi.c: (main):
  • JavaScriptCore.exp:
  • runtime/JSONObject.cpp: (JSC::JSONStringify):
  • runtime/JSONObject.h:
13:42 Changeset [56188] by darin@apple.com
  • 5 edits in trunk/WebCore

Remove Geolocation.lastPosition, no longer in the spec.
https://bugs.webkit.org/show_bug.cgi?id=36255
rdar://problem/7746357

Reviewed by Kenneth Rohde Christiansen.

  • WebCore.base.exp: Updated since Geolocation's destructor is now non-virtual.
  • page/Geolocation.cpp:

(WebCore::Geolocation::lastPosition): Add an assertion; it's only legal to
call this if access to the location is allowed.

  • page/Geolocation.h: Removed unneeded includes. Made destructor non-virtual,

although it will still be virtual if any of the base classes have a virtual
destructor. Made lastPosition, isAllowed, and isDenied functions private.
Removed unused suspend, resume, setShouldClearCache, shouldClearCache,
and frame functions.

  • page/Geolocation.idl: Removed lastPosition read-only attribute. No longer in

the Geolocation specification.

13:37 Changeset [56187] by hyatt@apple.com
  • 3 edits
    2 adds in trunk

https://bugs.webkit.org/show_bug.cgi?id=36275

Reviewed by Simon Fraser.

Make sure :not does not allow pseudo-elements inside it. (Hooray for another pointless
restriction on :not.)

Added fast/css/invalid-not-with-pseudo-element.html

  • css/CSSGrammar.y:
  • css/CSSSelector.h:

(WebCore::CSSSelector::matchesPseudoElement):

13:33 Changeset [56186] by mitz@apple.com
  • 6 edits
    1 copy in trunk

WebCore: <rdar://problem/7761400> Rework the fix for
https://bugs.webkit.org/show_bug.cgi?id=18722

Reviewed by Darin Adler.

Test: fast/dynamic/float-remove-above-line-2.html

  • rendering/RenderBlock.cpp:

(WebCore::RenderBlock::removeFloatingObject): Treat 0- and less-than-0-height floats
as having a height of 1 so that they intersect with the line they originate on.
(WebCore::RenderBlock::clearFloats): Use numeric_limits.

  • rendering/RenderBlockLineLayout.cpp:

(WebCore::RenderBlock::layoutInlineChildren): Removed the intersection checks here,
so that a float is always included in the float list of the line it originates on, even
if it does not intersect that line. This ensures that every float is accounted for, which
is necessary during incremental layout when adding floats from clean lines.

LayoutTests: Additional test for <rdar://problem/7761400>

Reviewed by Darin Adler.

  • fast/dynamic/float-remove-above-line-2.html: Copied from LayoutTests/fast/dynamic/float-remove-above-line.html.
  • fast/dynamic/float-remove-above-line-expected.txt: Updated.
  • fast/dynamic/float-remove-above-line.html: Fixed inconsequential typo in end tag.
13:29 Changeset [56185] by hyatt@apple.com
  • 1 add in trunk/LayoutTests/fast/css/invalid-import-rule-insertion-expected.txt

Add expected results.

13:19 Changeset [56184] by hausmann@webkit.org
  • 6 edits in trunk

2010-03-18 Antti Koivisto <koivisto@iki.fi>

Reviewed by Kenneth Rohde Christiansen.

https://bugs.webkit.org/show_bug.cgi?id=36102
[Qt] Scaling control API for tiled backing store

Commit the new scale synchronously after unfreeze to avoid ugliness.

  • platform/graphics/TiledBackingStore.cpp: (WebCore::TiledBackingStore::TiledBackingStore): (WebCore::TiledBackingStore::setContentsScale): (WebCore::TiledBackingStore::commitScaleChange): (WebCore::TiledBackingStore::setContentsFrozen):
  • platform/graphics/TiledBackingStore.h:

2010-03-18 Antti Koivisto <koivisto@iki.fi>

Reviewed by Kenneth Rohde Christiansen.

https://bugs.webkit.org/show_bug.cgi?id=36102
[Qt] Scaling control API for tiled backing store

The scale is set by passing the QGraphicsWebView scale to the backing store. The
only new API is the tiledBackingStoreFrozen property which allows disabling
all updates (for example during zoom animation).

  • Api/qgraphicswebview.cpp: (QGraphicsWebViewPrivate::_q_scaleChanged): (QGraphicsWebViewPrivate::updateTiledBackingStoreScale): (QGraphicsWebView::QGraphicsWebView): (QGraphicsWebView::isTiledBackingStoreFrozen): (QGraphicsWebView::setTiledBackingStoreFrozen):
  • Api/qgraphicswebview.h:
  • Api/qwebframe.cpp:
  • Api/qwebframe_p.h:
13:18 QtWebKitRelease20 edited by hausmann@webkit.org
(diff)
13:07 Changeset [56183] by eric@webkit.org
  • 3 edits in trunk/WebKitTools

2010-03-18 Jesus Sanchez-Palencia <jesus.palencia@openbossa.org>

Reviewed by Kenneth Rohde Christiansen.

Add an overlay QGraphicsTextItem to QtLauncher so we can display FPS info
on the launcher and not on the terminal anymore.

[Qt] QtLauncher's FPS info should be displayed on an overlay text item
https://bugs.webkit.org/show_bug.cgi?id=36244

  • QtLauncher/webview.cpp: (WebViewGraphicsBased::WebViewGraphicsBased): (WebViewGraphicsBased::setFrameRateMeasurementEnabled): (WebViewGraphicsBased::updateFrameRate):
  • QtLauncher/webview.h:
12:50 Changeset [56182] by hyatt@apple.com
  • 8 edits
    1 add in trunk

https://bugs.webkit.org/show_bug.cgi?id=36281

Reviewed by Simon Fraser.

Make sure an exception is raised if an @import rule is inserted in the wrong place.

Also make sure (so that this particular test case passes) that rgba(0, 0, 0, 0) is dumped
as transparent by getComputedStyle, since it most commonly occurs in background-color and
that default makes more sense than dumping rgba values.

Added fast/css/invalid-import-insertion.html

  • css/CSSPrimitiveValue.cpp:

(WebCore::CSSPrimitiveValue::createColor):

  • css/CSSStyleSheet.cpp:

(WebCore::CSSStyleSheet::insertRule):

12:49 Changeset [56181] by eric@webkit.org
  • 8 edits
    2 adds in trunk/WebKitTools

2010-03-18 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

First cut at SheriffBot
https://bugs.webkit.org/show_bug.cgi?id=36253

This patch contains a first attempt at writing a sheriff bot.
Currently, we're missing the logic that actually finds the SVN revision
numbers to complain about, but once we have that, we'll have the rest
of the infrustructure to ping IRC and to file bugs.

There's a lot to fill in for the SheriffBot, but this patch give us the
framework in which to do it.

This patch required a bit of refactoring of AbstractQueue because
SheriffBot is the first bot that doesn't process patches (it processes
SVN revisions). Accordingly, I've factored out AbstractPatchQueue to
hold the parts of AbstractQueue that are specific to dealing with
patches. Some of the choices here might not be obvious yet, but we can
tweak them as our needs become clearer.

  • Scripts/webkitpy/commands/queues.py:
  • Scripts/webkitpy/commands/queues_unittest.py:
  • Scripts/webkitpy/commands/sheriffbot.py: Added.
  • Scripts/webkitpy/commands/sheriffbot_unittest.py: Added.
  • Scripts/webkitpy/mock_bugzillatool.py:

Added a MockIRC object to the mock tool.

  • Scripts/webkitpy/multicommandtool.py:

Added a finalize method so the tool can disconnect from IRC
cleanly instead of just droping the socket.

  • Scripts/webkitpy/multicommandtool_unittest.py:
  • Scripts/webkitpy/patch/patcher.py:

Added support for talking to IRC.

  • Scripts/webkitpy/unittests.py:

We should add a commands/unittests.py file at some point to make
the commands module more self-contained.

12:32 Changeset [56180] by eric@webkit.org
  • 5 edits in trunk/WebCore

2010-03-18 Luiz Agostini <luiz.agostini@openbossa.org>

Reviewed by Kenneth Rohde Christiansen.

Multiselect Popup - Listbox click simulation
https://bugs.webkit.org/show_bug.cgi?id=36177

Listbox popups will need to notify the corresponding select elements that a
selection change happened. The current HTMLSelectElement interface does not
allow multiple selections.

The new method listBoxSelectItem will be used for that. I have refactored part
of the mouse handling code in bug 36124 and I am now reusing it here for
<select multiple> popups. All the other cases will handled as they were before to be
sure that no side effects will show up.

  • dom/SelectElement.cpp: (WebCore::SelectElement::updateListBoxSelection): (WebCore::SelectElement::listBoxOnChange):
  • dom/SelectElement.h:
  • html/HTMLSelectElement.cpp: (WebCore::HTMLSelectElement::listBoxPopupClick):
  • html/HTMLSelectElement.h:
12:14 Changeset [56179] by eric@webkit.org
  • 3 edits in trunk/WebKitTools

2010-03-18 Antti Koivisto <koivisto@iki.fi>

Reviewed by Kenneth Rohde Christiansen.

https://bugs.webkit.org/show_bug.cgi?id=36102
[Qt] Scaling control API for tiled backing store

Add animated smooth zooming to Qt launcher when in tiled mode.

  • QtLauncher/main.cpp: (LauncherWindow::LauncherWindow): (LauncherWindow::zoomAnimationFinished): (LauncherWindow::applyZoom): (LauncherWindow::zoomIn): (LauncherWindow::zoomOut):
  • QtLauncher/webview.h: (WebViewGraphicsBased::graphicsWebView):
11:56 Changeset [56178] by eric@webkit.org
  • 4 edits in trunk/WebKitTools

2010-03-18 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Support using IRC accounts with a password
https://bugs.webkit.org/show_bug.cgi?id=36287

Add a global option to specify an IRC password so we can use the
sheriffbot account (which needs a password).

  • Scripts/webkitpy/irc/ircbot.py:
  • Scripts/webkitpy/irc/ircproxy.py:
  • Scripts/webkitpy/patch/patcher.py:
11:37 Changeset [56177] by ap@apple.com
  • 1 edit
    2 copies in trunk/LayoutTests

Not reviewed.

https://bugs.webkit.org/show_bug.cgi?id=36210
plugins/resize-from-plugin.html fails on some platforms

Landing new results for Qt and Gtk, taken from buildbot.

Gtk results are a pass, although the difference in clip rect may indicate some issue to look into.
Qt results are a failure.

  • platform/gtk/plugins/resize-from-plugin-expected.txt: Copied from LayoutTests/plugins/resize-from-plugin-expected.txt.
  • platform/qt/plugins/resize-from-plugin-expected.txt: Copied from LayoutTests/plugins/resize-from-plugin-expected.txt.
11:33 Changeset [56176] by eric@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-18 Eric Seidel <eric@webkit.org>

Just fixing missing parenthesis typo, no review.

  • Scripts/webkitpy/commands/queries.py: '%' has higher precedence than 'or', use parentheses.
11:23 Changeset [56175] by ojan@chromium.org
  • 37 edits in trunk

2010-03-12 Ojan Vafai <ojan@chromium.org>

Reviewed by David Levin.

smartdelete should only occur after double-click
https://bugs.webkit.org/show_bug.cgi?id=35314

  • editing/deleting/non-smart-delete.html:
  • editing/pasteboard/drag-drop-modifies-page.html:
  • editing/pasteboard/page-zoom-expected.txt: Space is removed because we used to incorrectly smartpaste here. New behavior matches NSTextView.
  • editing/style/apply-through-end-of-document-expected.txt: The change here is that selectall + smartpaste doesn't put a BR, but selectall + normal-paste does. Not sure if that's a bug or not, but it's not a side effect of this change.
  • editing/style/style-boundary-005.html:
  • platform/mac/editing/deleting/non-smart-delete-expected.checksum:
  • platform/mac/editing/deleting/non-smart-delete-expected.png:
  • platform/mac/editing/deleting/non-smart-delete-expected.txt:
  • platform/mac/editing/execCommand/paste-1-expected.txt:
  • platform/mac/editing/execCommand/paste-2-expected.txt:
  • platform/mac/editing/pasteboard/drag-drop-modifies-page-expected.checksum:
  • platform/mac/editing/pasteboard/drag-drop-modifies-page-expected.png:
  • platform/mac/editing/pasteboard/drag-drop-modifies-page-expected.txt:
  • platform/mac/editing/selection/expanding-selections-expected.txt:
  • platform/mac/editing/selection/expanding-selections2-expected.txt:
  • platform/mac/editing/selection/triple-click-in-pre-expected.txt: The above three results change because granularity is no longer part of the seleciton, so changes in granularity don't fire selection change notifications.
  • platform/mac/editing/style/style-boundary-005-expected.checksum:
  • platform/mac/editing/style/style-boundary-005-expected.png:
  • platform/mac/editing/style/style-boundary-005-expected.txt:

2010-03-12 Ojan Vafai <ojan@chromium.org>

Reviewed by David Levin.

smartdelete should only occur after double-click
https://bugs.webkit.org/show_bug.cgi?id=35314

  1. Consolidate all notions of selection-granularity into SelectionController.
  2. Now only mouse-based selections store a selection-granularity. This matches NSTextView.

New tests were added in http://trac.webkit.org/changeset/55913.

  • WebCore.base.exp:
  • editing/MoveSelectionCommand.cpp: (WebCore::MoveSelectionCommand::MoveSelectionCommand): (WebCore::MoveSelectionCommand::doApply):
  • editing/MoveSelectionCommand.h: (WebCore::MoveSelectionCommand::create):
  • editing/SelectionController.cpp: (WebCore::SelectionController::SelectionController): (WebCore::SelectionController::setSelection): (WebCore::SelectionController::modify): (WebCore::SelectionController::clear):
  • editing/SelectionController.h: (WebCore::SelectionController::setSelection): (WebCore::SelectionController::granularity):
  • editing/VisibleSelection.cpp: (WebCore::VisibleSelection::VisibleSelection): (WebCore::VisibleSelection::expandUsingGranularity): (WebCore::VisibleSelection::setStartAndEndFromBaseAndExtentRespectingGranularity): (WebCore::VisibleSelection::validate):
  • editing/VisibleSelection.h: (WebCore::operator==):
  • loader/FrameLoader.cpp: (WebCore::FrameLoader::clear):
  • page/DOMSelection.cpp: (WebCore::DOMSelection::extend):
  • page/DragController.cpp: (WebCore::DragController::concludeEditDrag):
  • page/EventHandler.cpp: (WebCore::EventHandler::selectClosestWordFromMouseEvent): (WebCore::EventHandler::selectClosestWordOrLinkFromMouseEvent): (WebCore::EventHandler::handleMousePressEventTripleClick): (WebCore::EventHandler::handleMousePressEventSingleClick): (WebCore::EventHandler::updateSelectionForMouseDrag):
  • page/Frame.cpp: (WebCore::Frame::Frame): (WebCore::Frame::selectionGranularity):
  • page/Frame.h:
  • rendering/RenderTextControl.cpp: (WebCore::RenderTextControl::setSelectionRange):

2010-03-12 Ojan Vafai <ojan@chromium.org>

Reviewed by David Levin.

smartdelete should only occur after double-click
https://bugs.webkit.org/show_bug.cgi?id=35314

  • src/WebFrameImpl.cpp: (WebKit::WebFrameImpl::selectWordAroundPosition):
11:22 Changeset [56174] by simon.fraser@apple.com
  • 4 edits
    2 adds in trunk

2010-03-18 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dave Hyatt.

https://bugs.webkit.org/show_bug.cgi?id=36273
CSS3 :indeterminate and input type=radio test fails

The :indeterminate pseudo-class should apply to both radio buttons and checkboxes.

  • html/HTMLInputElement.cpp: (WebCore::HTMLInputElement::setIndeterminate): Use allowsIndeterminate() to determine if the input supports the indeterminate state. (WebCore::HTMLInputElement::preDispatchEventHandler): Use allowsIndeterminate() to determine whether to handle the indeterminate state. (WebCore::HTMLInputElement::postDispatchEventHandler): Ditto.
  • html/HTMLInputElement.h: (WebCore::HTMLInputElement::allowsIndeterminate): New utility method, returns true for radio buttons and checkboxes.
11:11 Changeset [56173] by dglazkov@chromium.org
  • 4 edits in trunk/WebCore

Unreviewed, build fix.

[V8] Turn npObjectTypeInfo into a function, make everyone happy.

  • bindings/v8/NPV8Object.cpp:

(WebCore::npObjectTypeInfo): Added new function.
(npCreateV8ScriptObject):

  • bindings/v8/NPV8Object.h: Turned npObjectTypeInfo into function.
  • bindings/v8/V8NPObject.cpp: Replaced references to function
11:06 Changeset [56172] by ap@apple.com
  • 6 edits
    1 copy in trunk

Reviewed by Adam Roben and Anders Carlsson.

https://bugs.webkit.org/show_bug.cgi?id=36210
plugins/resize-from-plugin.html fails on some platforms

WebKitTools:

Turns out that most platforms don't use "cross-platform" main.cpp. Copied code added for
the test to their versions of the file.

  • DumpRenderTree/unix/TestNetscapePlugin/TestNetscapePlugin.cpp: (webkit_test_plugin_set_window):
  • DumpRenderTree/win/TestNetscapePlugin/main.cpp: (NPP_SetWindow):

LayoutTests:

  • platform/win/plugins/resize-from-plugin-expected.txt: Copied from LayoutTests/plugins/resize-from-plugin-expected.txt. Windows results are different, probably because windowed plug-ins just get the whole window as clip rect.
  • platform/gtk/Skipped:
  • platform/qt/Skipped: Unskipped the tests to see how they work with TestNetscapePlugin fixes from this patch.
11:03 Changeset [56171] by hyatt@apple.com
  • 2 edits
    3 adds in trunk

https://bugs.webkit.org/show_bug.cgi?id=36283

Reviewed by Simon Fraser.

The wrong constants were being used for the code that attempted to compact the background-repeat shorthand to
a single value.

Added fast/backgrounds/background-repeat-computed-style.html

  • css/CSSComputedStyleDeclaration.cpp:

(WebCore::fillRepeatToCSSValue):

10:57 Changeset [56170] by simon.fraser@apple.com
  • 3 edits
    1 move
    1 add
    1 delete in trunk

2010-03-18 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dave Hyatt.

https://bugs.webkit.org/show_bug.cgi?id=36271
CSS3 :enabled on an input type=hidden element fails

The CSS3 selectors spec now makes it clear that hidden inputs should respect
the :enabled and :disabled pseudo-classes.

Test: fast/forms/hidden-input-enabled.html

  • css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::SelectorChecker::checkOneSelector): Remove checks for hidden inputs.
10:13 Changeset [56169] by enrica@apple.com
  • 3 edits
    4 adds in trunk

WebCore: Intro text at Star Wars demo is clipped.
<rdar://problem/7560979>
https://bugs.webkit.org/show_bug.cgi?id=33909

Reviewed by Simon Fraser.

Test: compositing/repaint/layer-repaint.html

We are flipping the coordinates for drawing, therefore
we must flip them also when we invalidate the rectangles.

  • platform/graphics/win/GraphicsLayerCACF.cpp:

(WebCore::WebLayer::setNeedsDisplay):

LayoutTests: Intro text at Star Wars demo is clipped.
<rdar://problem/7560979>
https://bugs.webkit.org/show_bug.cgi?id=33909

Reviewed by Simon Fraser.

We are flipping the coordinates for drawing, therefore
we must flip them also when we invalidate the rectangles.

  • compositing/repaint/layer-repaint.html: Added.
  • platform/mac/compositing/repaint/layer-repaint-expected.checksum: Added.
  • platform/mac/compositing/repaint/layer-repaint-expected.png: Added.
  • platform/mac/compositing/repaint/layer-repaint-expected.txt: Added.
09:59 Changeset [56168] by japhet@chromium.org
  • 4 edits in trunk

2010-03-18 Nate Chapin <japhet@chromium.org>

Unreviewed, Chromium build fix.

Missed a #include in V8Collection.h.

  • bindings/v8/V8Collection.h:
09:50 Changeset [56167] by hyatt@apple.com
  • 2 edits
    2 adds in trunk

Fix for https://bugs.webkit.org/show_bug.cgi?id=36284.

Reviewed by Simon Fraser.

Make sure the CSS parser properly rejects negative border widths.

Added fast/borders/negative-border-width.html

  • css/CSSParser.cpp:

(WebCore::CSSParser::parseValue):

09:38 Changeset [56166] by japhet@chromium.org
  • 44 edits
    1 add
    1 delete in trunk

2010-03-18 Nate Chapin <japhet@chromium.org>

Reviewed by Dimitri Glazkov.

[V8] Remove V8Index.h and all references to V8ClassIndex.


https://bugs.webkit.org/show_bug.cgi?id=33477

  • WebCore.gypi:
  • bindings/scripts/CodeGeneratorV8.pm:
  • bindings/v8/NPV8Object.cpp:
  • bindings/v8/NPV8Object.h:
  • bindings/v8/V8Collection.h:
  • bindings/v8/V8DOMWindowShell.cpp:
  • bindings/v8/V8DOMWindowShell.h:
  • bindings/v8/V8DOMWrapper.cpp:
  • bindings/v8/V8DOMWrapper.h:
  • bindings/v8/V8GCController.cpp:
  • bindings/v8/V8Helpers.cpp:
  • bindings/v8/V8Index.h: Removed.
  • bindings/v8/V8IsolatedContext.h:
  • bindings/v8/V8NPObject.cpp:
  • bindings/v8/V8Proxy.cpp:
  • bindings/v8/V8Proxy.h:
  • bindings/v8/V8SVGPODTypeWrapper.h:
  • bindings/v8/WorkerContextExecutionProxy.cpp:
  • bindings/v8/WorkerContextExecutionProxy.h:
  • bindings/v8/WrapperTypeInfo.h: Move WrapperTypeInfo struct into its own file from V8Index.h.
  • bindings/v8/custom/V8CanvasRenderingContext2DCustom.cpp:
  • bindings/v8/custom/V8DOMWindowCustom.cpp:
  • bindings/v8/custom/V8EventSourceConstructor.cpp:
  • bindings/v8/custom/V8HTMLAudioElementConstructor.cpp:
  • bindings/v8/custom/V8HTMLAudioElementConstructor.h:
  • bindings/v8/custom/V8HTMLImageElementConstructor.cpp:
  • bindings/v8/custom/V8HTMLImageElementConstructor.h:
  • bindings/v8/custom/V8HTMLOptionElementConstructor.cpp:
  • bindings/v8/custom/V8HTMLOptionElementConstructor.h:
  • bindings/v8/custom/V8HistoryCustom.cpp:
  • bindings/v8/custom/V8LocationCustom.cpp:
  • bindings/v8/custom/V8WebGLArrayCustom.h:
  • bindings/v8/custom/V8WebGLByteArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLFloatArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLIntArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLRenderingContextCustom.cpp:
  • bindings/v8/custom/V8WebGLShortArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLUnsignedByteArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLUnsignedIntArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLUnsignedShortArrayCustom.cpp:
  • bindings/v8/custom/V8WebKitPointConstructor.cpp:
09:20 Changeset [56165] by yurys@chromium.org
  • 3 edits in trunk/WebCore

2010-03-18 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Preserve console history between debugging sessions.

https://bugs.webkit.org/show_bug.cgi?id=36223

  • inspector/front-end/ConsoleView.js: (WebInspector.ConsoleView.prototype._settingsLoaded): (WebInspector.ConsoleView.prototype._enterKeyPressed.printResult): (WebInspector.ConsoleView.prototype._enterKeyPressed):
  • inspector/front-end/Settings.js: (WebInspector.Settings.prototype._load):
09:12 Changeset [56164] by mitz@apple.com
  • 1 edit in trunk/WebCore/ChangeLog

Added reviewer name for r56155

08:58 Changeset [56163] by eric@webkit.org
  • 4 edits in trunk

2010-03-18 Kim Grönholm <kim.gronholm@nomovok.com>

Reviewed by Simon Hausmann.

[Qt] GraphicsLayer: Pausing and resuming of animations don't work as expected
https://bugs.webkit.org/show_bug.cgi?id=36219

The play-state layout test now works, and can be removed from the skip list.

  • platform/qt/Skipped: animations/play-state.html is removed from the skipped list

2010-03-18 Kim Grönholm <kim.gronholm@nomovok.com>

Reviewed by Simon Hausmann.

[Qt] GraphicsLayer: Pausing and resuming of animations don't work as expected
https://bugs.webkit.org/show_bug.cgi?id=36219

addAnimation and pauseAnimation were handling animation state changes in
the wrong way.
Apparently WebCore calls addAnimation for resuming, so we have to make
sure to resume an existing animation if that happens. Also, timeOffset
is now used for synchronization of the Qt animation clock with the
WebCore clock, as opposed to using it as a delay timer.
Both those fixes were necessary to get the layout test to work.

LayoutTests/animations/play-state.html now passed on Qt.

  • platform/graphics/qt/GraphicsLayerQt.cpp: (WebCore::AnimationQtBase::AnimationQtBase): (WebCore::TransformAnimationQt::updateState): (WebCore::GraphicsLayerQt::addAnimation): (WebCore::GraphicsLayerQt::pauseAnimation):
08:41 Changeset [56162] by eric@webkit.org
  • 3 edits
    2 adds in trunk

2010-03-18 Adam Langley <agl@chromium.org>

Reviewed by David Levin.

[chromium] Fix inverted logic in per-strike renderer patch.

I screwed up in r55089 and got one of the conditions backwards. Thanks
to Roman Tsisyk for pointing it out.

https://bugs.webkit.org/show_bug.cgi?id=35495

  • platform/chromium/fast/text/chromium-linux-fontconfig-renderstyle.html: Added.

2010-03-18 Adam Langley <agl@chromium.org>

Reviewed by David Levin.

[chromium] Fix inverted logic in per-strike renderer patch.

I screwed up in r55089 and got one of the conditions backwards. Thanks
to Roman Tsisyk for pointing it out.

https://bugs.webkit.org/show_bug.cgi?id=35495

Test: platform/chromium/fast/text/chromium-linux-fontconfig-renderstyle.html

  • platform/graphics/chromium/FontPlatformDataLinux.cpp: (WebCore::FontPlatformData::querySystemForRenderStyle):
08:23 Changeset [56161] by pfeldman@chromium.org
  • 7 edits in trunk/WebCore

2010-03-18 Pavel Feldman <pfeldman@chromium.org>

Reviewed by Timothy Hatcher.

Web Inspector: Show inherited styles for a selected DOM element

https://bugs.webkit.org/show_bug.cgi?id=28039

  • English.lproj/localizedStrings.js:
  • inspector/InspectorDOMAgent.cpp: (WebCore::InspectorDOMAgent::getStyles): (WebCore::InspectorDOMAgent::populateObjectWithStyleProperties):
  • inspector/front-end/DOMAgent.js: (WebInspector.CSSStyleDeclaration):
  • inspector/front-end/StylesSidebarPane.js: (WebInspector.StylesSidebarPane.prototype._update): (WebInspector.StylesSidebarPane.prototype._containsInherited): (WebInspector.StylePropertiesSection): (WebInspector.StylePropertiesSection.prototype.isPropertyInherited): (WebInspector.StylePropertiesSection.prototype.onpopulate):
07:59 Changeset [56160] by eric@webkit.org
  • 3 edits in trunk/WebKitTools

2010-03-18 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Kenneth Rohde Christiansen.

Add support to run-launcher to open the EFL example browser.
http://webkit.org/b/36181

  • Scripts/webkitdirs.pm:
  • Scripts/run-launcher:
07:49 Changeset [56159] by tonikitoo@webkit.org
  • 2 edits in trunk/WebCore

Spatial Navigation: Add isNull and document convenience methods to FocusCandidate
https://bugs.webkit.org/show_bug.cgi?id=36167

Reviewed by Gustavo Noronha.
Patch by Antonio Gomes <tonikitoo@webkit.org>

It turns out that Spatial Navigation related code (in FocusController.cpp for
instance) can be simplified and look better if FocusCandidate class offer some
convinience method. This patch introduces a couple of them (isNull and a Document
getter).

A followup refactoring patch will be making use of these helper methods.

  • page/SpatialNavigation.h:

(WebCore::FocusCandidate::FocusCandidate): Added a "Node*" parameter
to FocusCandidate's constructor.
(WebCore::FocusCandidate::isNull): In the Spatial Navigation logic, a NULL
FocusCandidate object is one that does not hold a valid reference to a "Node*".
(WebCore::FocusCandidate::document): This method returns a pointer for the FocusCandidate
node's document, if |node| is a valid pointer.

07:42 Changeset [56158] by eric@webkit.org
  • 2 edits in trunk/WebCore

2010-03-18 Noam Rosenthal <noam.rosenthal@nokia.com>

Reviewed by Antti Koivisto.

[Qt] [Performance] GraphicsLayerQt updates the scene too often
https://bugs.webkit.org/show_bug.cgi?id=36158

This fix makes sure that flushChanges is only called when necessary,
by calling the notifySync function asynchronously, which makes sure flushChanges() is called
after the WebCore compositor has made all its changes.

This has shown a visual improvement on several test-cases.

  • Makefile:
  • platform/graphics/qt/GraphicsLayerQt.cpp: (WebCore::GraphicsLayerQtImpl::): (WebCore::GraphicsLayerQtImpl::notifySyncRequired): (WebCore::GraphicsLayerQtImpl::notifyChange): (WebCore::GraphicsLayerQtImpl::flushChanges): (WebCore::GraphicsLayerQt::setMaskLayer): (WebCore::GraphicsLayerQt::setPosition): (WebCore::GraphicsLayerQt::setAnchorPoint): (WebCore::GraphicsLayerQt::setSize): (WebCore::GraphicsLayerQt::setTransform): (WebCore::GraphicsLayerQt::setChildrenTransform): (WebCore::GraphicsLayerQt::setPreserves3D): (WebCore::GraphicsLayerQt::setMasksToBounds): (WebCore::GraphicsLayerQt::setDrawsContent): (WebCore::GraphicsLayerQt::setBackgroundColor): (WebCore::GraphicsLayerQt::clearBackgroundColor): (WebCore::GraphicsLayerQt::setContentsOpaque): (WebCore::GraphicsLayerQt::setBackfaceVisibility): (WebCore::GraphicsLayerQt::setOpacity): (WebCore::GraphicsLayerQt::setContentsRect):
06:31 Changeset [56157] by eric@webkit.org
  • 4 edits in trunk

2010-03-18 Sergio Villar Senin <svillar@igalia.com>

Reviewed by Xan Lopez.

[GTK] Failing tests http/tests/misc/image-blocked-src-change.html
& http/tests/misc/image-blocked-src-no-change.html
https://bugs.webkit.org/show_bug.cgi?id=36227

Removed tests from the list of Skipped:
http/tests/misc/image-blocked-src-no-change.html
http/tests/misc/image-blocked-src-change.html
http/tests/security/local-CSS-from-remote.html
http/tests/security/local-video-poster-from-remote.html
http/tests/security/frame-loading-via-document-write.html
http/tests/security/local-iFrame-from-remote.html
http/tests/security/local-video-src-from-remote.html
http/tests/security/local-image-from-remote.html
http/tests/security/local-JavaScript-from-remote.html

  • platform/gtk/Skipped:

2010-03-18 Sergio Villar Senin <svillar@igalia.com>

Reviewed by Xan Lopez.

[GTK] Failing tests http/tests/misc/image-blocked-src-change.html
& http/tests/misc/image-blocked-src-no-change.html
https://bugs.webkit.org/show_bug.cgi?id=36227

  • DumpRenderTree/gtk/DumpRenderTree.cpp: (webViewConsoleMessage): print only the file name instead of the whole URI when printing messages with local URI's
05:47 Changeset [56156] by eric@webkit.org
  • 4 edits in trunk/WebKitTools

2010-03-18 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Teach what-broke how to look up reviewer and author Committer objects by name
https://bugs.webkit.org/show_bug.cgi?id=36264

  • Scripts/webkitpy/commands/queries.py:
    • Add committer_by_name lookups for both reviewer and author
    • Improve printing in the cases where lookups fail.
  • Scripts/webkitpy/committers.py:
    • Add committer_by_name
  • Scripts/webkitpy/committers_unittest.py:
    • Test committer_by_name
05:30 Changeset [56155] by eric@webkit.org
  • 3 edits in trunk/WebCore

2010-03-18 Stephen White <senorblanco@chromium.org>

Parsing of rgb() colors in HTML canvas is still fairly slow.
findNamedColor() was showing up as hot, so this patch attempts to
avoid calling it by putting the rgb() check ahead of the named
colors. It also removes a call to Color::Color() from
CanvasStyle, which was doing a redundant check for the hex
color and named colors, which CSSParser already does. Finally, it
changes the ad-hoc parser to check for negative values, and clamps them
at zero. This avoids the lex/yacc path for rgb() colors with
negative color components.

https://bugs.webkit.org/show_bug.cgi?id=36199

Covered by fast/canvas/set-colors.html.

  • css/CSSParser.cpp: (WebCore::parseInt): (WebCore::CSSParser::parseColor):
  • html/canvas/CanvasStyle.cpp: (WebCore::CanvasStyle::applyFillColor):
05:13 Changeset [56154] by eric@webkit.org
  • 2 edits in trunk/LayoutTests

2010-03-18 Chang Shu <chang.shu@nokia.com>

Reviewed by Kenneth Rohde Christiansen.

https://bugs.webkit.org/show_bug.cgi?id=36220

[Qt] Unskip the tests that pass after the fix for bug 36139.
http/tests/appcache/foreign-iframe-main.html
http/tests/appcache/manifest-with-empty-file.html
http/tests/appcache/resource-redirect.html

  • platform/qt/Skipped:
04:56 Changeset [56153] by eric@webkit.org
  • 1 edit
    2 adds in trunk/WebKit

2010-03-18 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Kenneth Rohde Christiansen.

Add EFL port files to efl/ewk.
http://webkit.org/b/35931

  • efl/ewk/ewk_settings.cpp: Added.
  • efl/ewk/ewk_settings.h: Added.
01:55 Changeset [56152] by ddkilzer@apple.com
  • 2 edits in trunk/WebCore

<http://webkit.org/b/36246> Node::removeAllEventListenersSlowCase is declared but not used

Reviewed by Sam Weinig.

  • dom/Node.h:

(WebCore::Node::removeAllEventListenersSlowCase): Removed
declaration. Implementation was removed in r48701.

03/17/10:

22:37 Changeset [56151] by abarth@webkit.org
  • 5 edits in trunk/WebKitTools

2010-03-17 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

create-rollout should actually fill out the description
https://bugs.webkit.org/show_bug.cgi?id=36261

  • Scripts/webkitpy/commands/download.py:

The % operator was applied to the wrong string.

  • Scripts/webkitpy/commands/download_unittest.py:
  • Scripts/webkitpy/commands/upload_unittest.py:
  • Scripts/webkitpy/mock_bugzillatool.py:

Add support for seeing what we actually do with create_bug.

22:02 Changeset [56150] by weinig@apple.com
  • 2 edits in trunk/JavaScriptCore

Update WebKit availability macros for release after 4.0.

Reviewed by Darin Adler and Mark Rowe.

  • API/WebKitAvailability.h:
21:56 Changeset [56149] by eric@webkit.org
  • 1 edit
    2 adds in trunk/WebCore

2010-03-17 Stephan Aßmus <superstippi@gmx.de>

Reviewed by Oliver Hunt.

[Haiku] Implement ImageBuffer support
https://bugs.webkit.org/show_bug.cgi?id=35288

Covered by existing tests.

The StillImage class implements WebCore::Image by wrapping a native
BBitmap. It will be needed to implement WebCore::ImageBuffer. This
solution is just like it's done in the Qt port.

  • platform/graphics/haiku/StillImageHaiku.cpp: Added. (WebCore::StillImage::StillImage): (WebCore::StillImage::destroyDecodedData): (WebCore::StillImage::decodedSize): (WebCore::StillImage::size): (WebCore::StillImage::nativeImageForCurrentFrame): (WebCore::StillImage::draw):
  • platform/graphics/haiku/StillImageHaiku.h: Added. (WebCore::StillImage::create):
21:06 Changeset [56148] by zecke@webkit.org
  • 3 edits in trunk/WebCore

[Gtk+] Avoid making a copy of the HTTPHeaderMap.

ResourceRequest::httpHeaderFields() is returning a const
reference but we are storing it as a simple value. This
means that we run either the copy constructor or the assignment
operator and make a copy of the RefPtrs.

This extra copy can be avoided easily by storing
const reference.

  • platform/network/soup/ResourceRequestSoup.cpp:

(WebCore::ResourceRequest::toSoupMessage):

  • platform/network/soup/ResourceResponseSoup.cpp:

(WebCore::ResourceResponseSoup::toSoupMessage):

20:58 Changeset [56147] by eric@webkit.org
  • 1 edit
    56 adds in trunk/WebKit

2010-03-17 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Kenneth Rohde Christiansen.

Add default theme files to efl/DefaultTheme.
http://webkit.org/b/36240

  • efl/DefaultTheme/widget/radio/img_radio_on.png: Added.
  • efl/DefaultTheme/radio/img_radio_off_hover.png: Added.
  • efl/DefaultTheme/radio/img_radio_on_hover.png: Added.
  • efl/DefaultTheme/radio/img_radio_off.png: Added.
  • efl/DefaultTheme/radio/img_radio_off_focus.png: Added.
  • efl/DefaultTheme/radio/img_radio_on_focus.png: Added.
  • efl/DefaultTheme/combo/combo_normal.png: Added.
  • efl/DefaultTheme/combo/combo_focus_button.png: Added.
  • efl/DefaultTheme/combo/combo_hover_button.png: Added.
  • efl/DefaultTheme/combo/combo_hover.png: Added.
  • efl/DefaultTheme/combo/combo_focus.png: Added.
  • efl/DefaultTheme/combo/combo_press_button.png: Added.
  • efl/DefaultTheme/combo/combo_normal_button.png: Added.
  • efl/DefaultTheme/combo/combo_press.png: Added.
  • efl/DefaultTheme/combo/icon.png: Added.
  • efl/DefaultTheme/file/file_normal.png: Added.
  • efl/DefaultTheme/file/file_press.png: Added.
  • efl/DefaultTheme/file/file_hover.png: Added.
  • efl/DefaultTheme/file/file_focus.png: Added.
  • efl/DefaultTheme/search/cancel/cancel_normal_button.png: Added.
  • efl/DefaultTheme/search/cancel/cancel_normal_button2.png: Added.
  • efl/DefaultTheme/search/decoration/decoration_normal_button.png: Added.
  • efl/DefaultTheme/search/field/field_focused.png: Added.
  • efl/DefaultTheme/search/field/field_normal.png: Added.
  • efl/DefaultTheme/search/field/field_hovered.png: Added.
  • efl/DefaultTheme/entry/img_normal.png: Added.
  • efl/DefaultTheme/entry/img_hovered.png: Added.
  • efl/DefaultTheme/entry/img_focused.png: Added.
  • efl/DefaultTheme/check/img_check_off_hover.png: Added.
  • efl/DefaultTheme/check/img_check_on.png: Added.
  • efl/DefaultTheme/check/img_check_off_focus.png: Added.
  • efl/DefaultTheme/check/img_check_on_focus.png: Added.
  • efl/DefaultTheme/check/img_check_off.png: Added.
  • efl/DefaultTheme/check/img_check_on_hover.png: Added.
  • efl/DefaultTheme/scrollbar/scrollbar_knob_h.png: Added.
  • efl/DefaultTheme/scrollbar/scrollbar_knob_v.png: Added.
  • efl/DefaultTheme/scrollbar/scrollbar_hilight.png: Added.
  • efl/DefaultTheme/scrollbar/scrollbar_v.png: Added.
  • efl/DefaultTheme/scrollbar/scrollbar_h.png: Added.
  • efl/DefaultTheme/button/img_button_normal.png: Added.
  • efl/DefaultTheme/button/img_button_focus.png: Added.
  • efl/DefaultTheme/button/img_button_hover.png: Added.
  • efl/DefaultTheme/button/img_button_press.png: Added.
20:41 Changeset [56146] by eric@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-17 Jesus Sanchez-Palencia <jesus.palencia@openbossa.org>

Reviewed by Kenneth Rohde Christiansen.

Accelerated Compositing is now default on QtWebKit so the option
in QtLauncher must be true as default.

[Qt] QtLauncher's Accelerated Compositing option must be true as default
https://bugs.webkit.org/show_bug.cgi?id=36234

  • QtLauncher/main.cpp: (LauncherWindow::applyPrefs): (LauncherWindow::toggleAcceleratedCompositing): (LauncherApplication::handleUserOptions):
20:24 Changeset [56145] by eric@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-17 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Commit queue should ignore (probably red) builders when landing rollouts
https://bugs.webkit.org/show_bug.cgi?id=36169

When landing a rollout, the builders are probably red, so we need to
ignore them in the subprocess too. Also, we might as well update the
working copy because we haven't validated anything about the current
revision prior to trying to land.

This change is testable, but it requires changing the mock executive to
log its arguments. That will generate a lot of expectation changes, so
I'd like to do that in a separate patch.

  • Scripts/webkitpy/commands/queues.py:
20:07 Changeset [56144] by eric@webkit.org
  • 2 edits in trunk/WebKit/chromium

2010-03-17 Garret Kelly <gdk@chromium.org>

Reviewed by Darin Fisher.

PlatformTouchEventBuilder should not be resizing the touch point
array before appending touch points to it.
https://bugs.webkit.org/show_bug.cgi?id=36231

  • src/WebInputEventConversion.cpp: (WebKit::PlatformTouchEventBuilder::PlatformTouchEventBuilder):
19:51 Changeset [56143] by eric@webkit.org
  • 2 edits in trunk/WebCore

2010-03-17 Alexander Pavlov <apavlov@chromium.org>

Reviewed by Pavel Feldman.

Web Inspector: When switching resources in the Resources panel, the Content tab is always selected
https://bugs.webkit.org/show_bug.cgi?id=36222

  • inspector/front-end/ResourcesPanel.js: (WebInspector.ResourcesPanel.prototype.showResource):
19:25 Changeset [56142] by oliver@apple.com
  • 5 edits in trunk

2010-03-17 Oliver Hunt <oliver@apple.com>

Reviewed by Gavin Barraclough.

undefined, NaN, and Infinity should be ReadOnly
https://bugs.webkit.org/show_bug.cgi?id=36263

Simply add the ReadOnly flag to these properties.

  • runtime/JSGlobalObject.cpp: (JSC::JSGlobalObject::reset):

2010-03-17 Oliver Hunt <oliver@apple.com>

Reviewed by Gavin Barraclough.

undefined, NaN, and Infinity should be ReadOnly
https://bugs.webkit.org/show_bug.cgi?id=36263

Add tests for the attributes of undefined, NaN and Infinity on the
global object

  • fast/js/getOwnPropertyDescriptor-expected.txt:
  • fast/js/resources/getOwnPropertyDescriptor.js:
19:16 Changeset [56141] by eric@webkit.org
  • 3 edits in trunk/WebKit/qt

2010-03-17 Antti Koivisto <koivisto@iki.fi>

Reviewed by Tor Arne Vestbø.

Fix Qt build with tiled backing store disabled.

  • Api/qwebframe.cpp:
  • Api/qwebframe_p.h:
18:59 Changeset [56140] by eric@webkit.org
  • 6 edits in trunk

2010-03-17 Chang Shu <chang.shu@nokia.com>

Reviewed by Laszlo Gombos.

https://bugs.webkit.org/show_bug.cgi?id=36139
[Qt] Unskip the test that now passes.

platform/qt/Skipped:

2010-03-17 Chang Shu <chang.shu@nokia.com>

Reviewed by Laszlo Gombos.

https://bugs.webkit.org/show_bug.cgi?id=36139
[Qt] Clean up cache while setting cache quota. This behavior matches other platforms,
such as mac and gtk.

  • Api/qwebsettings.cpp:

(QWebSettings::setOfflineWebApplicationCacheQuota):

2010-03-17 Chang Shu <chang.shu@nokia.com>

Reviewed by Laszlo Gombos.

https://bugs.webkit.org/show_bug.cgi?id=36139
[Qt] Clean up cache each time DumpRenderTree starts. This behavior
matches other platforms, such as mac and gtk.

  • DumpRenderTree/qt/DumpRenderTreeQt.cpp: (WebCore::DumpRenderTree::DumpRenderTree):
18:32 Changeset [56139] by beidson@apple.com
  • 3 edits in trunk/WebCore

databaseIdentifiers are not filtered for slashes
<rdar://problem/7708789> and https://bugs.webkit.org/show_bug.cgi?id=36243

Reviewed by Darin Adler.

In addition to filtering for '/' and '\', to remove the directory vulnerability
on all platforms, it seems worth it to also escape other characters that are
obviously dangerous or illegal to have in a filename (mostly inspired by the
Windows illegal-character list).

No new tests - It's unclear how a test could possibly work into our testing
infrastructure.

  • page/SecurityOrigin.cpp:

(WebCore::): Added a 128-bool table "needsEscaping" that has a true/false answer

for lower-ASCII.

(WebCore::SecurityOrigin::SecurityOrigin):
(WebCore::SecurityOrigin::createFromDatabaseIdentifier): "Unescape" the host

component using the KURL utility.

(WebCore::shouldEscapeUChar): If the char is lower-ASCII, look it up in the

needsEscaping table. Otherwise, let it pass.

(WebCore::encodedHost): Modeled after KURL's host escaping method, but targeted

for the escaping considerations for the host component. We don't use the KURL
version because this one is "different" enough because it operates on UTF16 and
has its own "is bad character?" decider.

(WebCore::SecurityOrigin::databaseIdentifier): Create the escaped host if needed,

and use it instead of the unescaped host.

  • page/SecurityOrigin.h:
18:17 Changeset [56138] by abarth@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-17 Adam Barth <abarth@webkit.org>

Unreviewed.

Turns out this needs to be a string.

  • Scripts/webkitpy/bugzilla.py:
18:09 Changeset [56137] by eric@webkit.org
  • 12 edits in trunk/WebKitTools

2010-03-17 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Add more infrastructure for sheriff-bot, including making what-broke more awesome
https://bugs.webkit.org/show_bug.cgi?id=36254

  • Scripts/webkitpy/bugzilla.py:
    • Made the various URL methods return None when passed None.
  • Scripts/webkitpy/bugzilla_unittest.py:
    • Test that the url methods work as expected.
  • Scripts/webkitpy/buildbot.py:
    • Add a static Build.build_url so that its possible to generate a build url without a Build object.
    • Give users a URL in _fetch_xmlrpc_build_dictionary error message.
  • Scripts/webkitpy/changelogs.py:
    • Add a new ChangeLogEntry class to encapsulate entry-parsing logic.
    • Add is_path_to_changelog to greatly simplify SCM.modified_changelogs code.
    • Make ChangeLog.parse_latest_entry_from_file a public method.
  • Scripts/webkitpy/changelogs_unittest.py:
    • Add tests for new ChangeLog entry parsing.
  • Scripts/webkitpy/commands/queries.py:
    • Make "what-broke" not print "ok" builders, only failing ones.
    • Print much more information on failing builders, including links and authorship/reviewer information.
  • Scripts/webkitpy/commands/queues_unittest.py:
    • Use a fake_checkout path since fixing the cwd (as part of fixing scm_unittests.py) was breaking tests.
  • Scripts/webkitpy/mock_bugzillatool.py:
    • Move MockSCM away from using os.getcwd() as that was fragile (and wrong).
  • Scripts/webkitpy/patch/patcher.py:
    • Remove code which was broken now that this file has moved.
    • Code was also redundant now that SCM.find_checkout_root() exists.
  • Scripts/webkitpy/scm.py:
    • Greatly simplify modified_changelogs now that I understand list comprehensions.
    • Expect ChangeLogEntry objects instead of raw strings.
    • Add changed_files_for_revision, committer_email_for_revision and contents_at_revision
    • Add commit_with_message argument to all sites since someone half-added it before. :(
    • Get rid of copy/paste code using _status_regexp()
  • Scripts/webkitpy/scm_unittest.py:
    • Fix these tests!
    • Add new tests for new scm code.
    • Fix spelling of "awsome" to "awesome".
17:37 Changeset [56136] by eric@webkit.org
  • 1 edit
    4 adds in trunk/LayoutTests

2010-03-17 Drew Wilson <atwilson@chromium.org>

Reviewed by David Levin.

Add tests for "load shared worker in iframe" case.
https://bugs.webkit.org/show_bug.cgi?id=36203

Makes sure that creating a worker simultaneously in a document and in
a child iframe works (you end up with a single instance).

  • fast/workers/resources/shared-worker-count-connections.js: Added.
  • fast/workers/resources/shared-worker-iframe.html: Added.
  • fast/workers/shared-worker-in-iframe-expected.txt: Added.
  • fast/workers/shared-worker-in-iframe.html: Added.
17:10 Changeset [56135] by darin@apple.com
  • 2 edits in trunk/JavaScriptCore

Speed up Math.round a little by removing unneeded special case
https://bugs.webkit.org/show_bug.cgi?id=36107

Reviewed by Oliver Hunt.

Test: fast/js/math.html

  • runtime/MathObject.cpp:

(JSC::mathProtoFuncRound): This function had a special case for numbers
between -0.5 and -0.0 to return -0.0. But the algorithm in the function
already yields -0.0 for those cases, so the extra checking and branching
is unneeded.

16:40 Changeset [56134] by eric@webkit.org
  • 2 edits
    3 adds in trunk/LayoutTests

2010-03-17 Shu Chang <Chang.Shu@nokia.com>

Reviewed by Eric Seidel.

[Qt] Rebaseline the test and remove it from Skipped.
https://bugs.webkit.org/show_bug.cgi?id=36059

  • platform/qt/Skipped:
  • platform/qt/fast/canvas/canvas-text-baseline-expected.checksum: Added.
  • platform/qt/fast/canvas/canvas-text-baseline-expected.png: Added.
  • platform/qt/fast/canvas/canvas-text-baseline-expected.txt: Added.
16:39 Changeset [56133] by abarth@webkit.org
  • 5 edits in trunk/WebKitTools

2010-03-17 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Change post-rollout to create-rollout and have it make a new bug
instead of posting the rollout to the old bug.
https://bugs.webkit.org/show_bug.cgi?id=36250

The new bug blocks the old bug instead of adding more complexity to the
old bug. One tricky question is whether to create the bug if we're
unable to create a rollout patch. In this patch, we do create the bug,
but we might revist this question in the future.

  • Scripts/webkitpy/bugzilla.py:
  • Scripts/webkitpy/commands/download.py:
  • Scripts/webkitpy/commands/download_unittest.py:
  • Scripts/webkitpy/steps/createbug.py:
16:18 Changeset [56132] by dbates@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-17 Daniel Bates <dbates@rim.com>

Rubber-stamped by David Levin.

Add myself to the list of reviewers.

  • Scripts/webkitpy/committers.py:
16:01 WebKit Team edited by dbates@webkit.org
Moved my entry from Committer to Reviewer. (diff)
15:38 Changeset [56131] by dimich@chromium.org
  • 5 edits
    3 adds in trunk

[v8] Avoid reentry into v8 after TerminateExecution() on a worker thread.
https://bugs.webkit.org/show_bug.cgi?id=36239

Reviewed by Dimitri Glazkov.

WebCore:

Test: fast/workers/termination-with-port-messages.html

  • bindings/v8/V8Proxy.h:

(WebCore::throwError): Avoid throwing if v8 execution is terminated anyways.

  • bindings/v8/WorkerScriptController.h:

(WebCore::WorkerScriptController::proxy): Return V8 proxy as 0 if execution is terminating. Callers to check for 0 before invoking v8.

  • bindings/v8/custom/V8WorkerContextCustom.cpp:

(WebCore::SetTimeoutOrInterval): Bail out if v8 proxy is 0.
(WebCore::toV8): Ditto

LayoutTests:

  • fast/workers/resources/termination-with-port-messages.js: Added.

(echo):
(onmessage):

  • fast/workers/termination-with-port-messages-expected.txt: Added.
  • fast/workers/termination-with-port-messages.html: Added.
15:23 Changeset [56130] by eric@webkit.org
  • 1 edit
    2 adds in trunk/WebKit

2010-03-17 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Kenneth Rohde Christiansen.

Add EFL example browser to efl/EWebLauncher.
http://webkit.org/b/36176

  • efl/EWebLauncher/main.c: Added.
14:58 Changeset [56129] by dimich@chromium.org
  • 2 edits in trunk/WebKit/chromium

[chromium] Roll Chromium builder's DEPS to current revs.
https://bugs.webkit.org/show_bug.cgi?id=36241

Reviewed by Dimitri Glazkov.

  • DEPS:
14:47 Changeset [56128] by eric@webkit.org
  • 11 edits
    1 delete in trunk/WebCore

2010-03-17 José Millán Soto <jmillan@igalia.com>

Reviewed by Adam Barth.

Change the way of prefetching DNS to allow prefetching function to use full URL
https://bugs.webkit.org/show_bug.cgi?id=35589

Creates a new function, ResourceHandle::prepareForURL, whose default implementation
calls prefetchDNS with the host of the given URL as the parameter.
Soup implementation of ResourceHandle::prepareForURL executes soup_session_prepare_for_uri.
Soup implementation of prefetchDNS is removed.
Changed various prefetchDNS calls to ResourceHandle::prepareForURL calls.

  • GNUmakefile.am:
  • html/HTMLAnchorElement.cpp: (WebCore::HTMLAnchorElement::parseMappedAttribute):
  • html/HTMLLinkElement.cpp: (WebCore::HTMLLinkElement::process):
  • page/Chrome.cpp: (WebCore::Chrome::mouseDidMoveOverElement):
  • platform/network/DNS.h:
  • platform/network/ResourceHandle.cpp: (WebCore::ResourceHandle::prepareForURL):
  • platform/network/ResourceHandle.h:
  • platform/network/chromium/DNSChromium.cpp: (WebCore::ResourceHandle::prepareForURL):
  • platform/network/soup/DNSSoup.cpp: Removed.
  • platform/network/soup/ResourceHandleSoup.cpp: (WebCore::ResourceHandle::prepareForURL):
  • wml/WMLAElement.cpp: (WebCore::WMLAElement::parseMappedAttribute):
14:11 Changeset [56127] by kbr@google.com
  • 15 edits
    2 deletes in trunk

2010-03-17 Kenneth Russell <kbr@google.com>

Reviewed by Darin Fisher.

Revert WebGL context attributes changes 33416 / r56074 and 36200 / r56093
https://bugs.webkit.org/show_bug.cgi?id=36233

The patch for bug 33416, which added multisampling support to the
WebGL back buffer, uncovered some OpenGL driver bugs on the build
bots which need further investigation to determine the appropriate
workaround. Reverting this change, the minor build fix in 36189,
and the skipping of the affected tests in bug 36200.

Built and tested WebKit and Chromium and ran all WebGL layout
tests in both.

  • src/GraphicsContext3D.cpp: (WebCore::GraphicsContext3DInternal::GraphicsContext3DInternal): (WebCore::GraphicsContext3DInternal::~GraphicsContext3DInternal): (WebCore::GraphicsContext3DInternal::reshape): (WebCore::GraphicsContext3DInternal::beginPaint): (WebCore::GraphicsContext3DInternal::bindFramebuffer): (WebCore::GraphicsContext3D::readPixels):

2010-03-17 Kenneth Russell <kbr@google.com>

Reviewed by Darin Fisher.

Revert WebGL context attributes changes 33416 / r56074 and 36200 / r56093
https://bugs.webkit.org/show_bug.cgi?id=36233

The patch for bug 33416, which added multisampling support to the
WebGL back buffer, uncovered some OpenGL driver bugs on the build
bots which need further investigation to determine the appropriate
workaround. Reverting this change, the minor build fix in 36189,
and the skipping of the affected tests in bug 36200.

Built and tested WebKit and Chromium and ran all WebGL layout
tests in both.

  • bindings/v8/custom/V8HTMLCanvasElementCustom.cpp: (WebCore::V8HTMLCanvasElement::getContextCallback):
  • platform/graphics/GraphicsContext3D.h:
  • platform/graphics/mac/Canvas3DLayer.h:
  • platform/graphics/mac/Canvas3DLayer.mm: (-[Canvas3DLayer drawInCGLContext:pixelFormat:forLayerTime:displayTime:]):
  • platform/graphics/mac/GraphicsContext3DMac.cpp: (WebCore::GraphicsContext3D::GraphicsContext3D): (WebCore::GraphicsContext3D::~GraphicsContext3D): (WebCore::GraphicsContext3D::reshape): (WebCore::GraphicsContext3D::bindFramebuffer): (WebCore::GraphicsContext3D::readPixels):
  • platform/graphics/mac/GraphicsLayerCA.mm: (WebCore::GraphicsLayerCA::setContentsToGraphicsContext3D):

2010-03-17 Kenneth Russell <kbr@google.com>

Reviewed by Darin Fisher.

Revert WebGL context attributes changes 33416 / r56074 and 36200 / r56093
https://bugs.webkit.org/show_bug.cgi?id=36233

The patch for bug 33416, which added multisampling support to the
WebGL back buffer, uncovered some OpenGL driver bugs on the build
bots which need further investigation to determine the appropriate
workaround. Reverting this change, the minor build fix in 36189,
and the skipping of the affected tests in bug 36200.

Built and tested WebKit and Chromium and ran all WebGL layout
tests in both.

  • fast/canvas/webgl/context-attributes-alpha-depth-stencil-antialias-expected.txt: Removed.
  • fast/canvas/webgl/context-attributes-alpha-depth-stencil-antialias.html: Removed.
  • fast/canvas/webgl/context-attributes-expected.txt:
  • fast/canvas/webgl/context-attributes.html:
  • fast/canvas/webgl/gl-get-calls-expected.txt:
  • fast/canvas/webgl/gl-get-calls.html:
  • platform/mac/Skipped:
14:02 Changeset [56126] by yael.aharon@nokia.com
  • 8 edits in trunk/WebCore

Optimize painting for HTMLProgressElement
https://bugs.webkit.org/show_bug.cgi?id=36113
Cache the progress bar's position, so it is not required to recalculate
it for each repaint.
Call repaintRectangle() when the position changes.

Reviewed by Darin Adler.

No new tests, as no new functionality introduced.

  • html/HTMLProgressElement.cpp:
  • platform/qt/RenderThemeQt.cpp:
  • platform/qt/RenderThemeQt.h:
  • rendering/RenderProgress.cpp:
  • rendering/RenderProgress.h:
  • rendering/RenderTheme.cpp:
  • rendering/RenderTheme.h:
13:52 Changeset [56125] by abarth@webkit.org
  • 3 edits
    9 adds in trunk/WebKitTools

2010-03-17 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Add a way for the bots to send messages to IRC
https://bugs.webkit.org/show_bug.cgi?id=36235

We'll use these classes to notify #webkit about bad SVN revisions.
This patch just has some skeleton code for us to play with.

  • Scripts/webkitpy/irc/init.py: Added.
  • Scripts/webkitpy/irc/ircbot.py: Added.

A bot that knows how to talk to IRC.

  • Scripts/webkitpy/irc/ircproxy.py: Added.

We need to run the bot on its own thread because the irclib needs
its own mainloop. This class provides an abstraction of the
threading.

  • Scripts/webkitpy/irc/messagepump.py: Added.
  • Scripts/webkitpy/irc/messagepump_unittest.py: Added.
  • Scripts/webkitpy/irc/threadedmessagequeue.py: Added.

A thread-safe message queue for sending messages from the main
thread to the IRC thread.

  • Scripts/webkitpy/irc/threadedmessagequeue_unittest.py: Added.
  • Scripts/webkitpy/irc/unittests.py: Added.
  • Scripts/webkitpy/thirdparty/autoinstalled/init.py:

Autoinstall irclib

  • Scripts/webkitpy/unittests.py:
11:47 Changeset [56124] by kov@webkit.org
  • 2 edits in trunk/JavaScriptCore

2010-03-17 Mike Homey <glandium@debian.org>

Reviewed by Gustavo Noronha.

Build fix for SPARC. Fix missing macro value.

  • wtf/Platform.h:
11:44 Changeset [56123] by victorw@chromium.org
  • 2 edits in trunk/WebKitTools

2010-03-17 Victor Wang <victorw@chromium.org>

Reviewed by Dimitri Glazkov.

Fix image_diff syntax in webkitpy/port/base.py.
The syntax is wrong if diff_filename is specified.

https://bugs.webkit.org/show_bug.cgi?id=36230

  • Scripts/webkitpy/layout_tests/port/base.py:
11:22 Changeset [56122] by ap@apple.com
  • 2 edits in trunk/LayoutTests
  • platform/gtk/Skipped: Also, resize-from-plugin.html got skipped twice tonight, removing one of the lines.
11:15 Changeset [56121] by bweinstein@apple.com
  • 1 edit in trunk/WebKit/win/Interfaces/WebKit.idl

Touch WebKit.idl to fix the build

11:11 Changeset [56120] by tonikitoo@webkit.org
  • 2 edits in trunk/WebCore

Mac Buildfix: Using ASSERT_UNUSED instead of ASSERT.

Reviewed by Simon fraser.
Patch by Antonio Gomes <tonikitoo@webkit.org>

  • dom/SelectElement.cpp:

(WebCore::SelectElement::listBoxDefaultEventHandler):

11:02 Changeset [56119] by ap@apple.com
  • 2 edits in trunk/LayoutTests
  • platform/gtk/Skipped: Updated to reference a newly filed bug.
11:00 Changeset [56118] by enrica@apple.com
  • 8 edits in trunk

Missing support for showing compositing layers borders and repaint count on Windows.
<rdar://problem/7760736>
<https://bugs.webkit.org/show_bug.cgi?id=36197>

Reviewed by Darin Adler.

WebCore:

  • platform/graphics/win/GraphicsLayerCACF.cpp:

(WebCore::WebLayer::drawInContext): Modified to use the correct font on Windows.

WebKit/win:

  • Interfaces/IWebPreferencesPrivate.idl:
  • WebPreferenceKeysPrivate.h:
  • WebPreferences.cpp:

(WebPreferences::initializeDefaultSettings):
(WebPreferences::showDebugBorders):
(WebPreferences::setShowDebugBorders):
(WebPreferences::showRepaintCounter):
(WebPreferences::setShowRepaintCounter):

  • WebPreferences.h:
  • WebView.cpp:

(WebView::notifyPreferencesChanged):

10:37 Changeset [56117] by tonikitoo@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-16 Diego Gonzalez <diego.gonzalez@openbossa.org>

Reviewed by Kenneth Christiansen.

Provide to QtLauncher a way to change the ViewportUpdateMode
when it's in graphics based mode.

[Qt] Make QtLaucher able to select the ViewportUpdateMode
https://bugs.webkit.org/show_bug.cgi?id=36175

  • QtLauncher/main.cpp: (LauncherWindow::selectViewportUpdateMode): (LauncherWindow::createChrome):
10:25 Changeset [56116] by tonikitoo@webkit.org
  • 3 edits in trunk/WebCore

Multiselect Popup - SelectElement refactoring. https://bugs.webkit.org/show_bug.cgi?id=36124

Reviewed by Antti Koivisto.

As it is today it is not possible to the menulists popups to tell SelectElement
that more then one option is selected. If you select one the other ones
selected state will turn off.

For <select multiple> I needed to extend the API used by menulist popups. This
is the first step. Just refactoring SelectElement without any behavior change.
Separating code that will be reused in future patches.

  • dom/SelectElement.cpp:

(WebCore::SelectElement::updateSelectedState):
(WebCore::SelectElement::listBoxDefaultEventHandler):

  • dom/SelectElement.h:
10:09 Changeset [56115] by japhet@chromium.org
  • 1 edit
    1 copy
    1 delete in trunk/WebCore

2010-03-17 Nate Chapin <japhet@chromium.org>

Unreviewed, Chromium build fix.

Delete the correct file this time.

  • bindings/v8/custom/V8XMLHttpRequestCustom.cpp: Copied from WebCore/bindings/v8/custom/V8XMLHttpRequestCustom.cpp.
  • bindings/v8/custom/V8XMLHttpRequestUploadCustom.cpp: Removed.
09:51 Changeset [56114] by japhet@chromium.org
  • 14 edits
    1 delete in trunk/WebCore

2010-03-17 Nate Chapin <japhet@chromium.org>

Reviewed by Dimitri Glazkov.

[V8] Remove custom bindings callbacks that aren't in use.
(The relevant .idl doesn't mark them as V8Custom or Custom).


https://bugs.webkit.org/show_bug.cgi?id=33066

  • Android.v8bindings.mk:
  • WebCore.gypi:
  • bindings/scripts/CodeGeneratorV8.pm:
  • bindings/v8/custom/V8DOMWindowCustom.cpp:
  • bindings/v8/custom/V8WebGLArrayCustom.h:
  • bindings/v8/custom/V8WebGLByteArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLFloatArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLIntArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLShortArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLUnsignedByteArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLUnsignedIntArrayCustom.cpp:
  • bindings/v8/custom/V8WebGLUnsignedShortArrayCustom.cpp:
  • bindings/v8/custom/V8XMLHttpRequestCustom.cpp:
  • bindings/v8/custom/V8XMLHttpRequestUploadCustom.cpp: Removed.
09:11 Changeset [56113] by zoltan@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-17 Zoltan Horvath <zoltan@webkit.org>

Reviewed by Darin Adler.

Ambiguous error message when building for unspecified platform
https://bugs.webkit.org/show_bug.cgi?id=30203

Add an extra line information to the error message.

  • Scripts/webkitdirs.pm:
08:24 Changeset [56112] by steveblock@google.com
  • 3 edits in trunk/WebCore

2010-03-17 Steve Block <steveblock@google.com>

Reviewed by Dimitri Glazkov.

Adds ENABLE(WORKERS) guards to V8 bindings
https://bugs.webkit.org/show_bug.cgi?id=36221

Build fix only, no new tests.

  • bindings/v8/V8DOMWrapper.cpp: (WebCore::V8DOMWrapper::instantiateV8Object):
  • bindings/v8/V8Proxy.cpp: (WebCore::toV8Context):
07:57 ImprovingLifeOnWindows edited by dbates@webkit.org
Rolled out changeset 56044 as to broke the Windows Debug (test) bot. So, … (diff)
07:40 Changeset [56111] by pfeldman@chromium.org
  • 2 edits in trunk/LayoutTests

2010-03-17 Pavel Feldman <pfeldman@chromium.org>

Not reviewed: update mimetype expectations in inspector tests.

  • inspector/timeline-network-resource-expected.txt:
07:30 Changeset [56110] by pfeldman@chromium.org
  • 3 edits
    1 delete in trunk

2010-03-17 Pavel Feldman <pfeldman@chromium.org>

Not reviewed: GTK tests fix - removed custom expectation that was
added with no good reason.

  • inspector/timeline-test.js:
  • platform/gtk/inspector/timeline-network-resource-expected.txt: Removed.
07:20 Changeset [56109] by yurys@chromium.org
  • 2 edits in trunk/WebCore

2010-03-17 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Remove unused reference to InspectorController from InspectorFrontend.

https://bugs.webkit.org/show_bug.cgi?id=36218

  • inspector/InspectorFrontend.h:
07:08 Changeset [56108] by pfeldman@chromium.org
  • 14 edits in trunk

2010-03-17 Ilya Tikhonovsky <loislo@chromium.org>

Reviewed by Pavel Feldman.

WebInspector: Timeline improvements - next iteration.

The top level records should be collapsed.
Virtually linked events should be nested in Timeline like Send Request and
corresponding Received Responce, Timer Install and Timer Fire etc.
It should be possible to see Main Resource request.

https://bugs.webkit.org/show_bug.cgi?id=36122

  • inspector/InspectorController.cpp: (WebCore::InspectorController::resetScriptObjects): (WebCore::InspectorController::didReceiveResponse):
  • inspector/InspectorTimelineAgent.cpp: (WebCore::InspectorTimelineAgent::willReceiveResourceResponse): (WebCore::InspectorTimelineAgent::didReceiveResourceResponse):
  • inspector/InspectorTimelineAgent.h:
  • inspector/front-end/TimelinePanel.js: (WebInspector.TimelinePanel): (WebInspector.TimelinePanel.prototype._createStatusbarButtons): (WebInspector.TimelinePanel.prototype.addRecordToTimeline): (WebInspector.TimelinePanel.prototype._findParentRecord): (WebInspector.TimelinePanel.prototype._innerAddRecordToTimeline): (WebInspector.TimelinePanel.prototype.resize): (WebInspector.TimelinePanel.prototype._createTopLevelRecord): (WebInspector.TimelinePanel.prototype.reset): (WebInspector.TimelinePanel.prototype._clean): (WebInspector.TimelinePanel.prototype._refresh): (WebInspector.TimelinePanel.prototype._refreshRecords): (WebInspector.TimelinePanel.prototype._addToRecordsWindow): (WebInspector.TimelineCalculator.prototype.computeBarGraphPercentages): (WebInspector.TimelineCalculator.prototype.updateBoundaries): (WebInspector.TimelineRecordGraphRow): (WebInspector.TimelineRecordGraphRow.prototype.update): (WebInspector.TimelinePanel.FormattedRecord): (WebInspector.TimelinePanel.FormattedRecord.prototype.get children): (WebInspector.TimelinePanel.FormattedRecord.prototype._generatePopupContent):
  • inspector/front-end/inspector.css:
  • loader/ResourceLoader.cpp: (WebCore::ResourceLoader::didReceiveResponse):
06:51 Changeset [56107] by apavlov@chromium.org
  • 11 edits in trunk/WebCore

2010-03-17 Alexander Pavlov <apavlov@chromium.org>

Reviewed by Pavel Feldman.

Web Inspector: Reimplement style-related audits using native API
https://bugs.webkit.org/show_bug.cgi?id=36172

  • inspector/InspectorBackend.cpp: (WebCore::InspectorBackend::getAllStyles):
  • inspector/InspectorBackend.h:
  • inspector/InspectorBackend.idl:
  • inspector/InspectorDOMAgent.cpp: (WebCore::InspectorDOMAgent::getAllStyles): (WebCore::InspectorDOMAgent::getMatchedCSSRules): (WebCore::InspectorDOMAgent::buildObjectForStyleSheet): (WebCore::InspectorDOMAgent::buildObjectForRule):
  • inspector/InspectorDOMAgent.h:
  • inspector/InspectorFrontend.cpp: (WebCore::InspectorFrontend::didGetAllStyles):
  • inspector/InspectorFrontend.h:
  • inspector/front-end/AuditRules.js: (WebInspector.AuditRules.evaluateInTargetWindow): (WebInspector.AuditRules.UnusedCssRule.prototype.doRun.evalCallback.selectorsCallback): (WebInspector.AuditRules.UnusedCssRule.prototype.doRun.evalCallback.routine): (WebInspector.AuditRules.UnusedCssRule.prototype.doRun.evalCallback): (WebInspector.AuditRules.UnusedCssRule.prototype.doRun.routine): (WebInspector.AuditRules.UnusedCssRule.prototype.doRun): (WebInspector.AuditRules.ImageDimensionsRule.prototype.doRun): (WebInspector.AuditRules.CssInHeadRule.prototype.doRun): (WebInspector.AuditRules.StylesScriptsOrderRule.prototype.doRun):
  • inspector/front-end/DOMAgent.js:
  • inspector/front-end/InjectedScript.js: (injectedScriptConstructor):
05:47 Changeset [56106] by ossy@webkit.org
  • 4 edits in trunk

WebKit/qt: [Qt] Enable accelerated compositing by default
https://bugs.webkit.org/show_bug.cgi?id=35866

Reviewed by Kenneth Rohde Christiansen.

  • Api/qwebsettings.cpp:

(QWebSettings::QWebSettings):

LayoutTests: [Qt] GraphicsLayer: fix test regressions
https://bugs.webkit.org/show_bug.cgi?id=36191

Reviewed by Kenneth Rohde Christiansen.

Enabling accelerated compositing (AC) by default make 3 tests fail.
AC is more important now than these tests, so they are skipped until
fix and AC is enabled.

  • platform/qt/Skipped:
    • animations/fill-mode-transform.html skipped
    • animations/play-state.html skipped
    • animations/simultaneous-start-left.html skipped
05:10 Changeset [56105] by steveblock@google.com
  • 6 edits in trunk/WebCore

2010-03-17 Steve Block <steveblock@google.com>

Reviewed by Jeremy Orlow.

Replace instances of bzero in WebCore/bridge common code with memset
https://bugs.webkit.org/show_bug.cgi?id=36214

Syntax change only, no new tests.

  • bridge/jni/JNIUtility.cpp: (JSC::Bindings::getJNIField):
  • bridge/jni/jsc/JNIBridgeJSC.cpp: (JavaField::dispatchValueFromInstance):
  • bridge/jni/jsc/JNIUtilityPrivate.cpp: (JSC::Bindings::convertValueToJValue):
  • bridge/jni/v8/JNIUtilityPrivate.cpp: (JSC::Bindings::convertNPVariantToJValue):
  • bridge/jni/v8/JavaNPObjectV8.cpp: (JSC::Bindings::AllocJavaNPObject):
03:59 Changeset [56104] by yurys@chromium.org
  • 5 edits in trunk/WebCore

2010-03-17 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Mark inspector resources as changed so that they are reloaded on the next frontend connection.


Split resetScriptObjects into code specific for page navigation and insepctor frontend disconnection.

https://bugs.webkit.org/show_bug.cgi?id=36183

  • WebCore.xcodeproj/project.pbxproj: Added InspectorWorkerResource.h to the project file.
  • inspector/InspectorController.cpp: (WebCore::InspectorController::disconnectFrontend): (WebCore::InspectorController::unbindAllResources): (WebCore::InspectorController::didCommitLoad):
  • inspector/InspectorController.h:
  • inspector/InspectorResource.cpp: (WebCore::InspectorResource::releaseScriptObject):
02:57 Changeset [56103] by pfeldman@chromium.org
  • 2 edits in trunk/WebCore

2010-03-17 Pavel Feldman <pfeldman@chromium.org>

Reviewed by Timothy Hatcher.

Web Inspector: Capture main resource load events in timeline panel.

https://bugs.webkit.org/show_bug.cgi?id=36188

  • inspector/front-end/TimelinePanel.js: (WebInspector.TimelinePanel.prototype._createStatusbarButtons):
01:05 Changeset [56102] by mjs@apple.com
  • 3 edits in trunk/LayoutTests

Disable new plugin tests from r56079 on Gtk and Qt, since they are failing.

  • platform/gtk/Skipped:
  • platform/qt/Skipped:
00:51 Changeset [56101] by xan@webkit.org
  • 2 edits in trunk/LayoutTests

2010-03-17 Xan Lopez <xlopez@igalia.com>

Skip a couple of failing tests in the bots.

  • platform/gtk/Skipped:
00:39 Changeset [56100] by zoltan@webkit.org
  • 2 edits in trunk/WebCore

2010-03-17 Zoltan Horvath <zoltan@webkit.org>

Reviewed by Eric Seidel.

Allow custom memory allocation control for AtomicString class
https://bugs.webkit.org/show_bug.cgi?id=35907

Inherits the following class from FastAllocBase because it is
instantiated by 'new':

class name - instantiated at: WebCore/'location'
AtomicString - WebCore/html/HTMLTableElement.cpp:525

  • platform/text/AtomicString.h:

03/16/10:

23:50 Changeset [56099] by abarth@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-16 Adam Barth <abarth@webkit.org>

No review, rolling out r56044.
http://trac.webkit.org/changeset/56044
https://bugs.webkit.org/show_bug.cgi?id=36048

This patch broke Windows Debug (Tests)

  • Scripts/webkitdirs.pm:
20:35 QtWebKitBackportingFixes edited by laszlo.1.gombos@nokia.com
(diff)
18:53 Changeset [56098] by jhawkins@chromium.org
  • 2 edits in trunk/WebKit/chromium

2010-03-16 James Hawkins <jhawkins@chromium.org>

Reviewed by David Levin.

[Chromium] Size the WebVector of forms after determining how many
forms are valid and will be returned.

https://bugs.webkit.org/show_bug.cgi?id=36204

No new tests, as this is only triggered by Chromium's use of the
WebFormElement API.

  • src/WebFrameImpl.cpp: (WebKit::WebFrameImpl::forms):
18:05 Changeset [56097] by jam@chromium.org
  • 4 edits in trunk/LayoutTests/platform

update skipped test locations since I moved my test before commit

17:31 Changeset [56096] by jam@chromium.org
  • 13 edits
    3 adds in trunk

2010-03-16 John Abd-El-Malek <jam@chromium.org>

Reviewed by Darin Fisher.

Give keyboard focus to PluginDocuments by default
https://bugs.webkit.org/show_bug.cgi?id=36147

Test: http/tests/plugins/plugin-document-has-focus.html

  • loader/PluginDocument.cpp: (WebCore::PluginDocument::pluginNode):
  • loader/PluginDocument.h:
  • page/EventHandler.cpp: (WebCore::eventTargetNodeForDocument):

2010-03-16 John Abd-El-Malek <jam@chromium.org>

Reviewed by Darin Fisher.

Give keyboard focus to PluginDocuments by default
https://bugs.webkit.org/show_bug.cgi?id=36147

  • http/tests/plugins/plugin-document-has-focus-expected.txt: Added.
  • http/tests/plugins/plugin-document-has-focus.html: Added.
  • http/tests/plugins/resources/plugin-document-has-focus.pl: Added.
  • platform/gtk/Skipped:
  • platform/mac/Skipped:
  • platform/qt/Skipped:
  • platform/win/Skipped:

2010-03-16 John Abd-El-Malek <jam@chromium.org>

Reviewed by Darin Fisher.

Give keyboard focus to PluginDocuments by default
https://bugs.webkit.org/show_bug.cgi?id=36147

  • DumpRenderTree/TestNetscapePlugIn.subproj/PluginObject.cpp: (pluginAllocate):
  • DumpRenderTree/TestNetscapePlugIn.subproj/PluginObject.h:
  • DumpRenderTree/TestNetscapePlugIn.subproj/main.cpp: (NPP_New): (NPP_SetWindow): (handleEventCarbon): (handleEventCocoa):
17:23 Changeset [56095] by mitz@apple.com
  • 3 edits in trunk/WebKitLibraries

Fix incorrect glyph advances when using the Core Graphics (non-GDI) glyph look.

Reviewed by Sam Weinig.

  • win/lib/WebKitSystemInterface.lib:
  • win/lib/WebKitSystemInterface_debug.lib:
17:10 Changeset [56094] by cfleizach@apple.com
  • 5 edits
    2 adds in trunk

REGRESSION (r53857): Crash when tabbing to <map>
https://bugs.webkit.org/show_bug.cgi?id=36017

Reviewed by Darin Adler.

WebCore:

Test: fast/events/imagemap-norender-crash.html

  • html/HTMLMapElement.cpp:

(WebCore::HTMLMapElement::imageElement):

LayoutTests:

Also fix the path to js-test-post.js in tab-imagemap.html.

  • fast/events/imagemap-norender-crash-expected.txt: Added.
  • fast/events/imagemap-norender-crash.html: Added.
  • fast/events/tab-imagemap-expected.txt:
  • fast/events/tab-imagemap.html:
16:39 Changeset [56093] by kbr@google.com
  • 2 edits in trunk/LayoutTests

2010-03-16 Kenneth Russell <kbr@google.com>

Reviewed by Eric Seidel.

Temporarily disable WebGL layout tests failing due to multisampling
https://bugs.webkit.org/show_bug.cgi?id=36200

  • platform/mac/Skipped:
16:14 Changeset [56092] by barraclough@apple.com
  • 2 edits in trunk/JavaScriptCore

Bug 36083 - REGRESSION (r55772-r55834): Crash in JavaScriptCore RegExp code on PowerPC

Reviewed by Oliver Hunt, Darin Adler.

The problem is a bug in our port of PCRE - that a read may take place from the first character in an
empty string. For the time being, revert to using a valid pointer in the data segment rather than
an invalid non-null pointer into the zero-page for the empty string's data pointer. A better fix for
this will be to remove PCRE.

  • runtime/UStringImpl.cpp:

(JSC::UStringImpl::empty):

16:05 Changeset [56091] by weinig@apple.com
  • 2 edits in trunk/WebKitTools

Fix run-webkit-httpd on Windows.

Reviewed by Mark Rowe.

  • Scripts/webkitperl/httpd.pm:
15:33 Changeset [56090] by ap@apple.com
  • 2 edits in trunk/WebKitTools

Tiger build fix.

  • DumpRenderTree/TestNetscapePlugIn.subproj/PluginObject.cpp: (pluginGetProperty): Added more type casts to shut down warnings.
15:30 Changeset [56089] by ojan@chromium.org
  • 7 edits in trunk/LayoutTests

2010-03-15 Ojan Vafai <ojan@chromium.org>

Reviewed by Alexey Proskuryakov.

fix webkit win test failures for smart delete
https://bugs.webkit.org/show_bug.cgi?id=36141

Works around https://bugs.webkit.org/show_bug.cgi?id=36140.

  • editing/deleting/smart-delete-004.html:
  • editing/pasteboard/smart-paste-002.html:
  • editing/pasteboard/smart-paste-005.html:
  • editing/pasteboard/smart-paste-007.html:
  • platform/mac/editing/pasteboard/smart-paste-002-expected.txt:
  • platform/mac/editing/pasteboard/smart-paste-005-expected.txt:
15:26 Changeset [56088] by bfulgham@webkit.org
  • 3 edits in trunk/WebCore

Build fix. No Review.

Correct missing includes in cURL network backend introduced
by recent changes in WebKit.

  • platform/network/curl/ResourceHandleCurl.cpp: Add missing includes for SharedBuffer and PassRefPtr.

(WebCore::ResourceHandle::loadResourceSynchronously): Correct

constructor signature for ResourceHandle to match @r55688.

  • platform/network/curl/ResourceHandleManager.cpp: Add missing include for RetainPtr.
15:18 ImprovingLifeOnWindows edited by dbates@webkit.org
Support for detecting the presence of the Windows Platform SDK was … (diff)
15:10 QtWebKitGraphics edited by noam.rosenthal@nokia.com
(diff)
14:26 Changeset [56087] by kov@webkit.org
  • 1 copy in releases/WebKitGTK/webkit-1.1.90

Tagging 1.1.90.

14:26 Changeset [56086] by ap@apple.com
  • 4 edits in trunk/WebKitTools

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=36184
YouTube video resizing doesn't work with OOP plug-ins

Added a resizeTo() method, which calls resizePlugin() in JS with the same arguments,
and a lastSetWindowArguments property, which returns a string describing the last NPWindow
passed to NPN_SetWindow.

  • DumpRenderTree/TestNetscapePlugIn.subproj/PluginObject.cpp: (pluginGetProperty): (testResizeTo): (pluginInvoke):
  • DumpRenderTree/TestNetscapePlugIn.subproj/PluginObject.h:
  • DumpRenderTree/TestNetscapePlugIn.subproj/main.cpp: (NPP_SetWindow):
14:07 Changeset [56085] by darin@apple.com
  • 12 edits in trunk/JavaScriptCore

Rolled out r56081 since it broke the Windows build.

13:58 Changeset [56084] by kbr@google.com
  • 2 edits in trunk/WebKit/chromium

2010-03-16 Kenneth Russell <kbr@google.com>

Reviewed by Darin Fisher.

Build breakage from 33416
https://bugs.webkit.org/show_bug.cgi?id=36189

No new tests. Built release Chromium.

  • src/GraphicsContext3D.cpp: (WebCore::GraphicsContext3DInternal::reshape):
13:49 Changeset [56083] by xan@webkit.org
  • 2 edits in trunk

2010-03-16 Xan Lopez <xlopez@igalia.com>

Rubber-stamped by Gustavo Noronha.

Update library version for 1.1.90 release.

  • configure.ac:
13:46 Changeset [56082] by xan@webkit.org
  • 4 edits in trunk

2010-03-16 Xan Lopez <xlopez@igalia.com>

Reviewed by Gustavo Noronha.

Bump version for 1.1.90 release.

  • configure.ac:

WebKit/gtk:

2010-03-16 Xan Lopez <xlopez@igalia.com>

Reviewed by Gustavo Noronha.

Update for 1.1.90 release.

  • NEWS:
13:46 Changeset [56081] by darin@apple.com
  • 12 edits in trunk/JavaScriptCore

Added arrayLength function and used it instead of hand-written code.

Reviewed by Sam Weinig.

Removed PtrAndFlags.h. The file was removed long ago.

  • interpreter/Interpreter.cpp:

(JSC::Interpreter::privateExecute): Use arrayLength.

  • runtime/Collector.h:

(JSC::CollectorBitmap::count): Remove WTF:: prefix from use
of bitCount. In WTF we inject things into the global namespace
and this is no exception.

  • runtime/DatePrototype.cpp:

(JSC::formatLocaleDate): Use arrayLength.

  • runtime/JSFunction.h:

(JSC::JSFunction::nativeFunction): Remove WTF:: prefix as above.
(JSC::JSFunction::scopeChain): Ditto.
(JSC::JSFunction::setScopeChain): Ditto.
(JSC::JSFunction::setNativeFunction): Ditto.

  • runtime/JSGlobalObject.cpp:

(JSC::JSGlobalObject::reset): Use arrayLength.

  • runtime/JSImmediate.h:

(JSC::reinterpretDoubleToIntptr): Remove WTF:: prefix as above.
(JSC::reinterpretIntptrToDouble): Ditto.

  • runtime/JSONObject.cpp:

(JSC::Stringifier::appendQuotedString): Use arrayLength.
(JSC::Stringifier::toJSON): Ditto.
(JSC::Stringifier::appendStringifiedValue): Ditto.

  • runtime/SmallStrings.cpp:

(JSC::SmallStrings::SmallStrings): Ditto.

  • runtime/UString.cpp:

(JSC::UString::from): Use arrayLength.

  • wtf/StdLibExtras.h:

(WTF::bitwise_cast): Marked this inline!
(WTF::arrayLength): Added function template (actually written by Sam).
Added using declarations for public things in WTF namespace defined
in this file.

13:44 Changeset [56080] by kov@webkit.org
  • 2 edits in trunk/WebKit/gtk

2010-03-16 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Fix one too many empty lines in documentation of
window-obejct-cleared signal, which caused the documentation to be
rendered funny. Thanks to Martin Robinson for noticing.

  • webkit/webkitwebview.cpp:
13:40 Changeset [56079] by ap@apple.com
  • 5 edits
    3 adds in trunk

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=36184
YouTube video resizing doesn't work with OOP plug-ins

Test: plugins/resize-from-plugin.html

We were calling _WKPHResizePluginInstance synchronously or asynchronously, depending on
whether the size has changed. But sync and async messages are not necessarily delivered in
order - plug-in host listens only to the former while waiting for a response to a message it
sent (a call to invoke() in this case).

  • Plugins/Hosted/NetscapePluginInstanceProxy.h:
  • Plugins/Hosted/NetscapePluginInstanceProxy.mm: (WebKit::NetscapePluginInstanceProxy::resize):
  • Plugins/Hosted/WebHostedNetscapePluginView.mm: (-[WebHostedNetscapePluginView updateAndSetWindow]):
13:31 Changeset [56078] by kov@webkit.org
  • 2 edits in trunk/WebKit/gtk

2010-03-16 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Unreviewed. Add missing symbol to GeoLocation documentation
section.

  • docs/webkitgtk-sections.txt:
12:35 Changeset [56077] by kov@webkit.org
  • 1 edit
    1 add in trunk/WebKit/gtk/po

2010-03-16 Matej Urbančič <mateju@svn.gnome.org>

Reviewed by Gustavo Noronha.

Slovenian translation.

  • sl.po: Added.
12:14 Changeset [56076] by kov@webkit.org
  • 1 edit
    1 add in trunk/WebKit/gtk/po

2010-03-16 António Lima <amrlima@gmail.com>

Reviewed by Gustavo Noronha.

Translation for pt (Portuguese)
https://bugs.webkit.org/show_bug.cgi?id=36148

  • pt.po: Added.
11:56 Changeset [56075] by eric@webkit.org
  • 1 edit
    2 adds in trunk/WebKit

2010-03-16 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Kenneth Rohde Christiansen.

Add EFL port files to efl/WebCoreSupport.
http://webkit.org/b/35916

  • efl/WebCoreSupport/ContextMenuClientEfl.cpp: Added.
  • efl/WebCoreSupport/ContextMenuClientEfl.h: Added.
11:53 Changeset [56074] by kbr@google.com
  • 14 edits
    2 adds in trunk

2010-03-16 Zhenyao Mo <zmo@google.com>

Reviewed by Darin Fisher.

Hook up WebGLContextAttributes to OpenGL context creation code
https://bugs.webkit.org/show_bug.cgi?id=33416

Test: fast/canvas/webgl/context-attributes-alpha-depth-stencil-antialias.html

  • src/GraphicsContext3D.cpp: Hook up WebGLContextAttributes to OpenGL context creation code for Windows.

2010-03-16 Zhenyao Mo <zmo@google.com>

Reviewed by Darin Fisher.

Hook up WebGLContextAttributes to OpenGL context creation code
https://bugs.webkit.org/show_bug.cgi?id=33416

Test: fast/canvas/webgl/context-attributes-alpha-depth-stencil-antialias.html

  • bindings/v8/custom/V8HTMLCanvasElementCustom.cpp: Fix an index bug.
  • platform/graphics/GraphicsContext3D.h: Add members/functions for multisampling/stencil buffer purpose.
  • platform/graphics/mac/Canvas3DLayer.h: Add GraphicsContext3D as a member of Canvas3DLayer.
  • platform/graphics/mac/Canvas3DLayer.mm: Add multisampling support.
  • platform/graphics/mac/GraphicsContext3DMac.cpp: Hook up WebGLContextAttributes to OpenGL context creation code for Mac.
  • platform/graphics/mac/GraphicsLayerCA.mm: Adjust to modified Canvas3DLayer init call.

2010-03-16 Zhenyao Mo <zmo@google.com>

Reviewed by Darin Fisher.

Hook up WebGLContextAttributes to OpenGL context creation code
https://bugs.webkit.org/show_bug.cgi?id=33416

  • fast/canvas/webgl/context-attributes-alpha-depth-stencil-antialias-expected.txt: Added.
  • fast/canvas/webgl/context-attributes-alpha-depth-stencil-antialias.html: Added.
  • fast/canvas/webgl/context-attributes-expected.txt: WebGL context attributes behavior changed with this fix.
  • fast/canvas/webgl/context-attributes.html: Ditto.
  • fast/canvas/webgl/gl-get-calls-expected.txt: Stencil buffer is enabled, thus Stencil Bits is no longer 0.
  • fast/canvas/webgl/gl-get-calls.html: Ditto.
11:18 Changeset [56073] by eric@webkit.org
  • 7 edits in trunk

2010-03-16 Joanmarie Diggs <joanmarie.diggs@gmail.com>

Reviewed by Xan Lopez.

Added layout tests to get the number of rows and columns.

https://bugs.webkit.org/show_bug.cgi?id=35504
[Gtk] Evaluate and fix AtkTable for layout tables

  • platform/gtk/accessibility/table-hierarchy.html:
  • platform/gtk/accessibility/table-hierarchy-expected.txt:

2010-03-16 Joanmarie Diggs <joanmarie.diggs@gmail.com>

Reviewed by Xan Lopez.

https://bugs.webkit.org/show_bug.cgi?id=35504
[Gtk] Evaluate and fix AtkTable for layout tables

Causes "layout tables" to be exposed as AccessibilityTables for Gtk.

  • accessibility/AccessibilityTable.cpp: (AccessibilityTable::isTableExposableThroughAccessibility):

2010-03-16 Joanmarie Diggs <joanmarie.diggs@gmail.com>

Reviewed by Xan Lopez.

https://bugs.webkit.org/show_bug.cgi?id=35504
[Gtk] Evaluate and fix AtkTable for layout tables

Implements rowCount and columnCount for Gtk in DRT.

  • DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp: (AccessibilityUIElement::rowCount): (AccessibilityUIElement::columnCount):
10:41 Changeset [56072] by mrobinson@webkit.org
  • 5 edits in trunk/WebKit/gtk

2010-03-16 Martin Robinson <mrobinson@webkit.org>

Reviewed by Xan Lopez.

can't input korean into lower all input box except adress input box in webkit gtk launcher
https://bugs.webkit.org/show_bug.cgi?id=32290

Make the GTK+ EditorClient properly handle different types of input module
behavior such as commit and preedit signals that happen outside of key event
filtering and multiple times in a row. Filter keyup events as well as keydown
events and call gtk_im_context_focus_{in/out} when the WebView focus changes.

Added tests for this behavior to the GTK+ unit tests.

  • WebCoreSupport/EditorClientGtk.cpp: (WebKit::imContextCommitted): Handle this signal properly when it happens outside of key event filtering. (WebKit::imContextPreeditChanged): Immediately update the preedit state and do not reject empty preedits, so that cancellation works properly. (WebKit::EditorClient::updatePendingComposition): Add this method, which handles the situation where a commit signal happens when there is still a pending commit. (WebKit::EditorClient::shouldBeginEditing): Clear pending composition state before editing starts. (WebKit::EditorClient::shouldEndEditing): Clear pending composition state before editing ends. (WebKit::EditorClient::handleKeyboardEvent): No longer special case preedits which happen during key event filtering. When confirming a pending composition use insertText instead of confirmComposition. (WebKit::EditorClient::handleInputMethodKeydown):
  • WebCoreSupport/EditorClientGtk.h: Make pendingComposition a member, so that multiple WebViews do not share state. (WebKit::EditorClient::webView): Added. (WebKit::EditorClient::treatContextCommitAsKeyEvent): Added. (WebKit::EditorClient::clearPendingComposition): Added.
  • tests/testkeyevents.c: (test_keypress_events_load_status_cb): (map_event_cb): (setup_keyevent_test): (test_keypress_events): (element_text_equal_to): (test_ime_load_status_cb): (test_ime): (main):
  • webkit/webkitwebview.cpp: (webkit_web_view_key_release_event): (webkit_web_view_focus_in_event):
10:33 Changeset [56071] by kov@webkit.org
  • 2 edits in trunk/WebCore

Unreviewed. Build fix for distcheck.

10:26 Changeset [56070] by darin@chromium.org
  • 9 edits in trunk

2010-03-15 Darin Fisher <darin@chromium.org>

Reviewed by Adam Barth.

https://bugs.webkit.org/show_bug.cgi?id=36152
[chromium] Add support for history.pushState and history.replaceState

This is covered by the existing tests in fast/loader/stateobjects

09:58 Changeset [56069] by aroben@apple.com
  • 2 edits in trunk/WebCore

Fix linker warnings when building WebCore on Windows

Rubber-stamped by Steve Falkenburg.

  • WebCore.vcproj/WebCore.vcproj: Exclude JSDOMFormData.cpp from the

build, since it's already included in DerivedSources.cpp. Also let VS
sort this file as it saw fit.

09:53 QtWebKitGraphics edited by noam.rosenthal@nokia.com
(diff)
09:49 Changeset [56068] by antti@apple.com
  • 6 edits in trunk

WebCore: https://bugs.webkit.org/show_bug.cgi?id=36121
[Qt] Make WebKit scrollbars work with tiling

Reviewed by Simon Hausmann.

Translate coordinates correctly.

  • page/Frame.cpp:

(WebCore::Frame::tiledBackingStorePaintEnd):

WebKit/qt: https://bugs.webkit.org/show_bug.cgi?id=36121
[Qt] Make WebKit scrollbars work with tiling

Reviewed by Simon Hausmann.

  • Use the scrollbar overlay (implemented for accelerated compositing) when in tiled mode.
  • Make overlay compile unconditionally, enable on demand. This removes bunch of unneeded ifdefs.
  • Update the scroll position to the backing store as needed.
  • Renamed some methods.
  • Api/qgraphicswebview.cpp:

(QGraphicsWebViewPrivate::):
(QGraphicsWebViewPrivate::createOrDeleteOverlay):
(QGraphicsWebViewPrivate::setRootGraphicsLayer):
(QGraphicsWebViewPrivate::updateCompositingScrollPosition):
(QGraphicsWebViewPrivate::scroll):
(QGraphicsWebViewPrivate::update):
(QGraphicsWebViewPrivate::graphicsItemVisibleRect):
(QGraphicsWebView::paint):
(QGraphicsWebView::setPage):
(QGraphicsWebView::updateGeometry):
(QGraphicsWebView::setGeometry):

  • Api/qwebframe.cpp:

(QWebFramePrivate::renderFromTiledBackingStore):

  • Api/qwebframe_p.h:
09:21 Changeset [56067] by pfeldman@chromium.org
  • 2 edits in trunk/LayoutTests

2010-03-16 Pavel Feldman <pfeldman@chromium.org>

Not reviewed: inspector http test fix that migrates it to the new inspector api.

  • http/tests/inspector/inspect-iframe-from-different-domain.html:
09:19 Changeset [56066] by zoltan@webkit.org
  • 2 edits in trunk/JavaScriptCore

2010-03-16 Zoltan Horvath <zoltan@webkit.org>

Reviewed by Darin Adler.

Remove extra <new> include and add guards to operator new/delete definitions
https://bugs.webkit.org/show_bug.cgi?id=35967

Remove extra <new> header include from FastAlloc.cpp since it is included in
FastAlloc.h. Add ENABLE(GLOBAL_FASTMALLOC_NEW) macro guard to operator
new/delete/new []/delete [] definitions.

  • wtf/FastMalloc.cpp:
09:18 Changeset [56065] by jwieczorek@webkit.org
  • 2 edits in trunk/WebCore

2010-03-16 Jakub Wieczorek <jwieczorek@webkit.org>

Reviewed by Eric Seidel.

[Qt] MediaPlayerPrivate: Initialize m_mediaPlayerControl to 0 to
avoid crashes when no media services can be found.

https://bugs.webkit.org/show_bug.cgi?id=36142

  • platform/graphics/qt/MediaPlayerPrivateQt.cpp: (WebCore::MediaPlayerPrivate::MediaPlayerPrivate):
09:09 Changeset [56064] by abarth@webkit.org
  • 2 edits in trunk/WebCore

2010-03-16 Adam Barth <abarth@webkit.org>

Unreviewed.

noscript tag should render when @sandbox disables JavaScript
https://bugs.webkit.org/show_bug.cgi?id=36092

Add missing "!" in plugin code. Should fix plugins/embed-inside-object
on Gtk and Qt.

  • plugins/PluginView.cpp: (WebCore::PluginView::load):
09:00 Changeset [56063] by japhet@chromium.org
  • 2 edits
    1 delete in trunk/WebCore

2010-03-16 Nate Chapin <japhet@chromium.org>

Reviewed by Dimitri Glazkov.

[V8] Delete DerivedSourcesAllInOne.cpp.

https://bugs.webkit.org/show_bug.cgi?id=33048

  • Android.v8bindings.mk:
  • bindings/v8/DerivedSourcesAllInOne.cpp: Removed.
08:39 Changeset [56062] by hausmann@webkit.org
  • 2 edits in trunk/JavaScriptCore

Roll out r55889, as it breaks the Qt port on Leopard

2010-03-11 Kent Hansen <kent.hansen@nokia.com>

Reviewed by Simon Hausmann.

Mac OS X: Use deployment target to determine whether memory tagging should be enabled
https://bugs.webkit.org/show_bug.cgi?id=34888

Use TARGETING_TIGER rather than BUILDING_ON_TIGER to determine the tags.

07:35 Changeset [56061] by kov@webkit.org
  • 2 edits in trunk/WebCore

2010-03-16 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Revert r56006 as it caused functional issues. We will work on a
testcase, and rework the fix, and land when we have this a bit
more testable.

  • platform/gtk/RenderThemeGtk.cpp: (WebCore::RenderThemeGtk::paintMediaPlayButton):
07:19 Changeset [56060] by robert@webkit.org
  • 2 edits in trunk/WebKit/qt

2010-03-15 Robert Hogan <robert@webkit.org>

Reviewed by Simon Hausmann.

[Qt] Fix crash in QWebView::setPage()

tst_qwebpage was crashing on setPage(0)

https://bugs.webkit.org/show_bug.cgi?id=36137

  • Api/qwebview.cpp: (QWebView::setPage):
06:58 Changeset [56059] by pfeldman@chromium.org
  • 5 edits in trunk/WebCore

2010-03-16 Pavel Feldman <pfeldman@chromium.org>

Reviewed by Timothy Hatcher.

Web Inspector: Errors when inspecting styles of non-renderable elements in XHTML.

https://bugs.webkit.org/show_bug.cgi?id=35025

  • inspector/InspectorDOMAgent.cpp: (WebCore::InspectorDOMAgent::getStyles):
  • inspector/front-end/DOMAgent.js: (WebInspector.DOMNode.prototype._addAttribute): (WebInspector.DOMWindow.prototype.Object):
  • inspector/front-end/MetricsSidebarPane.js:
  • inspector/front-end/StylesSidebarPane.js: (WebInspector.StylesSidebarPane.prototype.update.callback): (WebInspector.StylesSidebarPane.prototype.update): (WebInspector.StylesSidebarPane.prototype._update):
06:37 Changeset [56058] by yurys@chromium.org
  • 2 edits in trunk/WebKit/chromium

2010-03-16 Yury Semikhatsky <yurys@chromium.org>

Unreviewed.

Fix Chromium Mac build: remove unused code.

  • src/WebDevToolsAgentImpl.cpp: (WebKit::WebDevToolsAgentImpl::createInspectorFrontendProxy):
06:32 Changeset [56057] by tonikitoo@webkit.org
  • 6 edits in trunk

Spatial Navigation: Add a scrollIntoView call when focusing an element.
https://bugs.webkit.org/show_bug.cgi?id=36020

Reviewed by Simon Fraser.
Patch by Antonio Gomes <tonikitoo@webkit.org>

WebCore:

When focusing an element in Spatial Navigation logic, it is desired to make
this element visible in the current viewport. For that to happen, there
could be a call to Element's scrollIntoView method at focusing time. However
for visual aspects, it is preferable to scroll to an inflated rect of |element|
in order to consider the focus outline width.
As Element's scrollIntoView method does not provide this flexibility, patch adds
a custom scrollIntoView method to SpatialNavigation.h .

  • page/FocusController.cpp:

(WebCore::FocusController::advanceFocusDirectionally):

  • page/SpatialNavigation.cpp:

(WebCore::scrollIntoView):

  • page/SpatialNavigation.h:

LayoutTests:

  • platform/gtk/Skipped: Unskip snav-iframe-with-offscreen-element.html
05:10 Changeset [56056] by kov@webkit.org
  • 2 edits in trunk/LayoutTests

2010-03-16 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

[GStreamer] media/video-played-collapse.html
https://bugs.webkit.org/show_bug.cgi?id=36165

Skip apparently unreliable test. It used to fail locally and pass
on the bots, but the situation is inversed now. We need to
investigate this properly.

  • platform/gtk/Skipped:
04:57 Changeset [56055] by kov@webkit.org
  • 3 edits in trunk/WebCore

2010-03-10 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Eric Seidel.

[GTK] Position queries are lagging
https://bugs.webkit.org/show_bug.cgi?id=34372

Normalize timeout priorities through the GStreamer media player as
a whole. This should help avoid that any of them is starved.

  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp: (WebCore::MediaPlayerPrivate::volumeChanged): (WebCore::MediaPlayerPrivate::processBufferingStats): (WebCore::MediaPlayerPrivate::muteChanged):
  • platform/graphics/gstreamer/VideoSinkGStreamer.cpp: (webkit_video_sink_render):
  • platform/graphics/gstreamer/WebKitWebSourceGStreamer.cpp: (webKitWebSrcNeedDataCb): (webKitWebSrcEnoughDataCb): (webKitWebSrcSeekDataCb):
04:45 Changeset [56054] by kov@webkit.org
  • 2 edits in trunk/WebCore

2010-03-10 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Eric Seidel.

[GTK] Position queries are lagging
https://bugs.webkit.org/show_bug.cgi?id=34372

Use default priority also for parsing data: URIs.

  • platform/network/soup/ResourceHandleSoup.cpp: (WebCore::startData):
04:44 Changeset [56053] by kov@webkit.org
  • 2 edits in trunk/WebCore

2010-03-10 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Eric Seidel.

[GTK] Position queries are lagging
https://bugs.webkit.org/show_bug.cgi?id=34372

Normalize SharedTimers priorities to always be
G_PRIORITY_DEFAULT. As it is currently, timers that should happen
immediately (0 miliseconds delay) end up delayed by a lot of time
due to glib starving lower priority sources.

  • platform/gtk/SharedTimerGtk.cpp: (WebCore::setSharedTimerFireTime):
04:44 QtWebKitBugs edited by jocelyn.turcotte@nokia.com
Make the new bug link easier to spot on the page (diff)
04:36 QtWebKitBugs edited by jocelyn.turcotte@nokia.com
(diff)
04:35 QtWebKitBugs edited by jocelyn.turcotte@nokia.com
(diff)
04:11 Changeset [56052] by xan@webkit.org
  • 4 edits in trunk

2010-03-16 Xan Lopez <xlopez@igalia.com>

Reviewed by Gustavo Noronha.

Add support for Fast Mobile Scrolling in the build system.

  • configure.ac:

WebCore:

2010-03-16 Xan Lopez <xlopez@igalia.com>

Reviewed by Gustavo Noronha.

Add support for Fast Mobile Scrolling in the build system.

  • GNUmakefile.am:
03:59 Changeset [56051] by yurys@chromium.org
  • 62 edits
    2 copies
    1 add in trunk

2010-03-11 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Introduce InspectorFrontendClient that provides InspectorFrontend with an interface to the embedder. InspectorClient now serves as a delegate for InspectorController and does not contain methods for managing inspector frontend window. That allows to create remote InspectorFrontendHost.

Introduce InspectorFrontendClient that would provide InspectorFrontend with an interface to the embedder
https://bugs.webkit.org/show_bug.cgi?id=35036

  • GNUmakefile.am:
  • WebCore.Inspector.exp:
  • WebCore.base.exp:
  • WebCore.gypi:
  • WebCore.order:
  • WebCore.pro:
  • WebCore.xcodeproj/project.pbxproj:
  • inspector/ConsoleMessage.cpp: (WebCore::ConsoleMessage::addToFrontend):
  • inspector/ConsoleMessage.h:
  • inspector/InspectorClient.h:
  • inspector/InspectorController.cpp: (WebCore::InspectorController::InspectorController): (WebCore::InspectorController::~InspectorController): (WebCore::InspectorController::inspectedPageDestroyed): (WebCore::InspectorController::windowVisible): (WebCore::InspectorController::frontendReady): (WebCore::InspectorController::willCloseFrontend): (WebCore::InspectorController::addConsoleMessage): (WebCore::InspectorController::setInspectorFrontendClient): (WebCore::InspectorController::inspectedWindowScriptObjectCleared): (WebCore::InspectorController::setFrontend): (WebCore::InspectorController::show): (WebCore::InspectorController::close): (WebCore::InspectorController::releaseDOMAgent): (WebCore::InspectorController::populateScriptObjects): (WebCore::InspectorController::didCommitLoad): (WebCore::InspectorController::getProfile): (WebCore::InspectorController::enableDebugger):
  • inspector/InspectorController.h: (WebCore::InspectorController::hasInspectorFrontendClient):
  • inspector/InspectorFrontend.cpp: (WebCore::InspectorFrontend::InspectorFrontend): (WebCore::InspectorFrontend::~InspectorFrontend): (WebCore::InspectorFrontend::addConsoleMessage): (WebCore::InspectorFrontend::bringToFront): (WebCore::InspectorFrontend::inspectedURLChanged):
  • inspector/InspectorFrontend.h:
  • inspector/InspectorFrontendClient.h: Added. (WebCore::InspectorFrontendClient::~InspectorFrontendClient):
  • inspector/InspectorFrontendClientLocal.cpp: Added. (WebCore::FrontendMenuProvider::create): (WebCore::FrontendMenuProvider::disconnect): (WebCore::FrontendMenuProvider::FrontendMenuProvider): (WebCore::FrontendMenuProvider::~FrontendMenuProvider): (WebCore::FrontendMenuProvider::populateContextMenu): (WebCore::FrontendMenuProvider::contextMenuItemSelected): (WebCore::FrontendMenuProvider::contextMenuCleared): (WebCore::InspectorFrontendClientLocal::InspectorFrontendClientLocal): (WebCore::InspectorFrontendClientLocal::~InspectorFrontendClientLocal): (WebCore::InspectorFrontendClientLocal::windowObjectCleared): (WebCore::InspectorFrontendClientLocal::frontendLoaded): (WebCore::InspectorFrontendClientLocal::canAttachWindow): (WebCore::InspectorFrontendClientLocal::changeAttachedWindowHeight): (WebCore::InspectorFrontendClientLocal::moveWindowBy): (WebCore::InspectorFrontendClientLocal::showContextMenu): (WebCore::InspectorFrontendClientLocal::setAttachedWindow): (WebCore::InspectorFrontendClientLocal::restoreAttachedWindowHeight): (WebCore::InspectorFrontendClientLocal::constrainedAttachedWindowHeight):
  • inspector/InspectorFrontendClientLocal.h: Added.
  • inspector/InspectorFrontendHost.cpp: (WebCore::InspectorFrontendHost::InspectorFrontendHost): (WebCore::InspectorFrontendHost::~InspectorFrontendHost): (WebCore::InspectorFrontendHost::loaded): (WebCore::InspectorFrontendHost::attach): (WebCore::InspectorFrontendHost::detach): (WebCore::InspectorFrontendHost::closeWindow): (WebCore::InspectorFrontendHost::bringToFront): (WebCore::InspectorFrontendHost::inspectedURLChanged): (WebCore::InspectorFrontendHost::canAttachWindow): (WebCore::InspectorFrontendHost::setAttachedWindowHeight): (WebCore::InspectorFrontendHost::moveWindowBy): (WebCore::InspectorFrontendHost::showContextMenu):
  • inspector/InspectorFrontendHost.h: (WebCore::InspectorFrontendHost::create):
  • inspector/InspectorFrontendHost.idl:
  • inspector/front-end/InspectorFrontendHostStub.js: (.WebInspector.InspectorFrontendHostStub.prototype.bringToFront): (.WebInspector.InspectorFrontendHostStub.prototype.inspectedURLChanged):
  • inspector/front-end/inspector.js: (WebInspector.loaded): (WebInspector.close): (WebInspector.bringToFront): (WebInspector.inspectedURLChanged):
  • loader/EmptyClients.h: (WebCore::EmptyInspectorClient::openInspectorFrontend):
  • loader/FrameLoader.cpp: (WebCore::FrameLoader::dispatchDidClearWindowObjectInWorld):
  • page/Page.cpp: (WebCore::Page::Page): (WebCore::Page::~Page):
  • page/Page.h:
  • platform/ContextMenu.cpp: (WebCore::ContextMenu::populate): (WebCore::ContextMenu::addInspectElementItem):

2010-03-16 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Introduce InspectorFrontendClient that provides InspectorFrontend with an interface to the embedder. InspectorClient now serves as a delegate for InspectorController and does not contain methods for managing inspector frontend window. That allows to create remote InspectorFrontendHost.

Introduce InspectorFrontendClient that would provide InspectorFrontend with an interface to the embedder
https://bugs.webkit.org/show_bug.cgi?id=35036

  • WebCoreSupport/WebInspectorClient.h:
  • WebCoreSupport/WebInspectorClient.mm: (WebInspectorClient::WebInspectorClient): (WebInspectorClient::inspectorDestroyed): (WebInspectorClient::openInspectorFrontend): (WebInspectorClient::highlight): (WebInspectorClient::hideHighlight): (WebInspectorFrontendClient::WebInspectorFrontendClient): (WebInspectorFrontendClient::frontendLoaded): (WebInspectorFrontendClient::localizedStringsURL): (WebInspectorFrontendClient::hiddenPanels): (WebInspectorFrontendClient::bringToFront): (WebInspectorFrontendClient::closeWindow): (WebInspectorFrontendClient::attachWindow): (WebInspectorFrontendClient::detachWindow): (WebInspectorFrontendClient::setAttachedWindowHeight): (WebInspectorFrontendClient::inspectedURLChanged): (WebInspectorFrontendClient::updateWindowTitle): (-[WebInspectorWindowController dealloc]): (-[WebInspectorWindowController windowShouldClose:]): (-[WebInspectorWindowController close]): (-[WebInspectorWindowController showWindow:]): (-[WebInspectorWindowController attach]): (-[WebInspectorWindowController detach]): (-[WebInspectorWindowController attached]): (-[WebInspectorWindowController setFrontendClient:]): (-[WebInspectorWindowController destroyInspectorView]): (-[WebNodeHighlighter initWithInspectedWebView:]): (-[WebNodeHighlighter dealloc]): (-[WebNodeHighlighter highlightNode:]): (-[WebNodeHighlighter hideHighlight]): (-[WebNodeHighlighter didAttachWebNodeHighlight:]): (-[WebNodeHighlighter willDetachWebNodeHighlight:]):
  • WebInspector/WebInspector.mm: (-attach:): (-detach:):

2010-03-16 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Introduce InspectorFrontendClient that provides InspectorFrontend with an interface to the embedder. InspectorClient now serves as a delegate for InspectorController and does not contain methods for managing inspector frontend window. That allows to create remote InspectorFrontendHost.

Introduce InspectorFrontendClient that would provide InspectorFrontend with an interface to the embedder
https://bugs.webkit.org/show_bug.cgi?id=35036

  • WebCoreSupport/WebInspectorClient.cpp: (WebInspectorClient::WebInspectorClient): (WebInspectorClient::~WebInspectorClient): (WebInspectorClient::openInspectorFrontend): (WebInspectorClient::highlight): (WebInspectorClient::hideHighlight): (WebInspectorClient::updateHighlight): (WebInspectorFrontendClient::WebInspectorFrontendClient): (WebInspectorFrontendClient::~WebInspectorFrontendClient): (WebInspectorFrontendClient::frontendLoaded): (WebInspectorFrontendClient::localizedStringsURL): (WebInspectorFrontendClient::hiddenPanels): (WebInspectorFrontendClient::bringToFront): (WebInspectorFrontendClient::closeWindow): (WebInspectorFrontendClient::attachWindow): (WebInspectorFrontendClient::detachWindow): (WebInspectorFrontendClient::setAttachedWindowHeight): (WebInspectorFrontendClient::inspectedURLChanged): (WebInspectorFrontendClient::closeWindowWithoutNotifications): (WebInspectorFrontendClient::showWindowWithoutNotifications): (WebInspectorFrontendClient::destroyInspectorView): (WebInspectorFrontendClient::updateWindowTitle): (WebInspectorFrontendClient::onGetMinMaxInfo): (WebInspectorFrontendClient::onSize): (WebInspectorFrontendClient::onClose): (WebInspectorFrontendClient::onSetFocus): (WebInspectorFrontendClient::onWebViewWindowPosChanging): (WebInspectorWndProc): (WebInspectorFrontendClient::windowReceivedMessage):
  • WebCoreSupport/WebInspectorClient.h: (WebInspectorClient::frontendClosing):
  • WebInspector.cpp: (WebInspector::attach): (WebInspector::detach):

2010-03-16 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Introduce InspectorFrontendClient that provides InspectorFrontend with an interface to the embedder. InspectorClient now serves as a delegate for InspectorController and does not contain methods for managing inspector frontend window. That allows to create remote InspectorFrontendHost.

Introduce InspectorFrontendClient that would provide InspectorFrontend with an interface to the embedder
https://bugs.webkit.org/show_bug.cgi?id=35036

  • Api/qwebinspector.cpp: (QWebInspector::hideEvent): (QWebInspector::closeEvent):
  • Api/qwebinspector.h:
  • Api/qwebpage.h:
  • WebCoreSupport/InspectorClientQt.cpp: (WebCore::InspectorClientQt::openInspectorFrontend): (WebCore::InspectorFrontendClientQt::InspectorFrontendClientQt): (WebCore::InspectorFrontendClientQt::frontendLoaded): (WebCore::InspectorFrontendClientQt::localizedStringsURL): (WebCore::InspectorFrontendClientQt::hiddenPanels): (WebCore::InspectorFrontendClientQt::bringToFront): (WebCore::InspectorFrontendClientQt::closeWindow): (WebCore::InspectorFrontendClientQt::attachWindow): (WebCore::InspectorFrontendClientQt::detachWindow): (WebCore::InspectorFrontendClientQt::setAttachedWindowHeight): (WebCore::InspectorFrontendClientQt::inspectedURLChanged): (WebCore::InspectorFrontendClientQt::updateWindowTitle):
  • WebCoreSupport/InspectorClientQt.h:

2010-03-16 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Introduce InspectorFrontendClient that provides InspectorFrontend with an interface to the embedder. InspectorClient now serves as a delegate for InspectorController and does not contain methods for managing inspector frontend window. That allows to create remote InspectorFrontendHost.

Introduce InspectorFrontendClient that would provide InspectorFrontend with an interface to the embedder
https://bugs.webkit.org/show_bug.cgi?id=35036

  • WebCoreSupport/InspectorClientGtk.cpp: (WebKit::notifyWebViewDestroyed): (WebKit::InspectorClient::InspectorClient): (WebKit::InspectorClient::inspectorDestroyed): (WebKit::InspectorClient::openInspectorFrontend): (WebKit::InspectorClient::highlight): (WebKit::InspectorClient::hideHighlight): (WebKit::InspectorClient::populateSetting): (WebKit::InspectorClient::storeSetting): (WebKit::InspectorFrontendClient::InspectorFrontendClient): (WebKit::InspectorFrontendClient::~InspectorFrontendClient): (WebKit::InspectorFrontendClient::destroyInspectorWindow): (WebKit::InspectorFrontendClient::localizedStringsURL): (WebKit::InspectorFrontendClient::hiddenPanels): (WebKit::InspectorFrontendClient::bringToFront): (WebKit::InspectorFrontendClient::closeWindow): (WebKit::InspectorFrontendClient::attachWindow): (WebKit::InspectorFrontendClient::detachWindow): (WebKit::InspectorFrontendClient::setAttachedWindowHeight): (WebKit::InspectorFrontendClient::inspectedURLChanged):
  • WebCoreSupport/InspectorClientGtk.h:

2010-03-16 Yury Semikhatsky <yurys@chromium.org>

Reviewed by Pavel Feldman.

Introduce InspectorFrontendClient that provides InspectorFrontend with an interface to the embedder. InspectorClient now serves as a delegate for InspectorController and does not contain methods for managing inspector frontend window. That allows to create remote InspectorFrontendHost.

Introduce InspectorFrontendClient that would provide InspectorFrontend with an interface to the embedder
https://bugs.webkit.org/show_bug.cgi?id=35036

  • src/InspectorClientImpl.cpp: (WebKit::InspectorClientImpl::openInspectorFrontend):
  • src/InspectorClientImpl.h:
  • src/WebDevToolsAgentImpl.cpp: (WebKit::WebDevToolsAgentImpl::attach): (WebKit::WebDevToolsAgentImpl::didCommitProvisionalLoad): (WebKit::WebDevToolsAgentImpl::createInspectorFrontendProxy): (WebKit::WebDevToolsAgentImpl::setInspectorFrontendProxyToInspectorController):
  • src/WebDevToolsAgentImpl.h:
  • src/WebDevToolsFrontendImpl.cpp: (WebKit::WebDevToolsFrontendImpl::WebDevToolsFrontendImpl): (WebKit::WebDevToolsFrontendImpl::jsBringToFront): (WebKit::WebDevToolsFrontendImpl::jsInspectedURLChanged):
  • src/WebDevToolsFrontendImpl.h:
  • src/js/InjectDispatch.js: (close):
03:17 Changeset [56050] by pfeldman@chromium.org
  • 2 edits in trunk/WebKit/chromium

2010-03-16 Pavel Feldman <pfeldman@chromium.org>

Not reviewed: chromium tests fix.

  • src/js/InspectorControllerImpl.js: (devtools.InspectorBackendImpl):
02:32 Changeset [56049] by pfeldman@chromium.org
  • 19 edits in trunk

2010-03-16 Pavel Feldman <pfeldman@chromium.org>

Reviewed by Timothy Hatcher.

Web Inspector: migrate to native styles inspector in
order to inspect styles from foreighn domains.

https://bugs.webkit.org/show_bug.cgi?id=36117

  • inspector/InspectorBackend.cpp: (WebCore::InspectorBackend::getStyles): (WebCore::InspectorBackend::getInlineStyle): (WebCore::InspectorBackend::getComputedStyle): (WebCore::InspectorBackend::applyStyleText): (WebCore::InspectorBackend::setStyleText): (WebCore::InspectorBackend::setStyleProperty): (WebCore::InspectorBackend::toggleStyleEnabled): (WebCore::InspectorBackend::setRuleSelector): (WebCore::InspectorBackend::addRule):
  • inspector/InspectorBackend.h:
  • inspector/InspectorBackend.idl:
  • inspector/InspectorDOMAgent.cpp: (WebCore::InspectorDOMAgent::InspectorDOMAgent): (WebCore::InspectorDOMAgent::getStyles): (WebCore::InspectorDOMAgent::getInlineStyle): (WebCore::InspectorDOMAgent::getComputedStyle): (WebCore::InspectorDOMAgent::getMatchedCSSRules): (WebCore::InspectorDOMAgent::getAttributeStyles): (WebCore::InspectorDOMAgent::applyStyleText): (WebCore::InspectorDOMAgent::setStyleText): (WebCore::InspectorDOMAgent::setStyleProperty): (WebCore::InspectorDOMAgent::toggleStyleEnabled): (WebCore::InspectorDOMAgent::setRuleSelector): (WebCore::InspectorDOMAgent::addRule): (WebCore::InspectorDOMAgent::bindStyle): (WebCore::InspectorDOMAgent::bindRule): (WebCore::InspectorDOMAgent::buildObjectForStyle): (WebCore::InspectorDOMAgent::populateObjectWithStyleProperties): (WebCore::InspectorDOMAgent::buildObjectForRule): (WebCore::InspectorDOMAgent::uniqueStyleProperties): (WebCore::InspectorDOMAgent::longhandProperties): (WebCore::InspectorDOMAgent::shorthandPriority): (WebCore::InspectorDOMAgent::ruleAffectsNode): (WebCore::InspectorDOMAgent::toArray):
  • inspector/InspectorDOMAgent.h:
  • inspector/InspectorFrontend.cpp: (WebCore::InspectorFrontend::didGetEventListenersForNode): (WebCore::InspectorFrontend::didGetStyles): (WebCore::InspectorFrontend::didGetComputedStyle): (WebCore::InspectorFrontend::didGetInlineStyle): (WebCore::InspectorFrontend::didApplyStyleText): (WebCore::InspectorFrontend::didSetStyleText): (WebCore::InspectorFrontend::didSetStyleProperty): (WebCore::InspectorFrontend::didToggleStyleEnabled): (WebCore::InspectorFrontend::didSetRuleSelector): (WebCore::InspectorFrontend::didAddRule):
  • inspector/InspectorFrontend.h:
  • inspector/front-end/DOMAgent.js: (WebInspector.CSSStyleDeclaration):
  • inspector/front-end/ElementsTreeOutline.js:
  • inspector/front-end/InjectedScript.js:
  • inspector/front-end/MetricsSidebarPane.js: (WebInspector.MetricsSidebarPane.prototype.update): (WebInspector.MetricsSidebarPane.prototype.editingCommitted):
  • inspector/front-end/StylesSidebarPane.js: (WebInspector.StylesSidebarPane.prototype.update): (WebInspector.StylePropertiesSection.prototype.editingSelectorCommitted.callback): (WebInspector.StylePropertiesSection.prototype.editingSelectorCommitted): (WebInspector.BlankStylePropertiesSection.prototype.editingSelectorCommitted.callback): (WebInspector.BlankStylePropertiesSection.prototype.editingSelectorCommitted): (WebInspector.StylePropertyTreeElement.prototype): (WebInspector.StylePropertyTreeElement.prototype.):
01:53 Changeset [56048] by hausmann@webkit.org
  • 2 edits in trunk

Add WebKitTools/TestResultServer/index.yaml to gitattributes to ignore for crlf conversion.

  • .gitattributes:
00:39 Changeset [56047] by hamaji@chromium.org
  • 4 edits in trunk/WebCore

2010-03-16 Shinichiro Hamaji <hamaji@chromium.org>

Reviewed by Darin Adler.

RenderText::m_text should be a String, not RefPtr<StringImpl>
https://bugs.webkit.org/show_bug.cgi?id=36010

Refactoring only, so no new tests.

  • platform/text/PlatformString.h: (WebCore::String::makeLower): (WebCore::String::makeUpper): (WebCore::String::makeSecure): (WebCore::String::makeCapitalized): (WebCore::String::containsOnlyASCII):
  • rendering/RenderText.cpp: (WebCore::RenderText::RenderText): (WebCore::RenderText::widthFromCache): (WebCore::RenderText::trimmedPrefWidths): (WebCore::RenderText::containsOnlyWhitespace): (WebCore::RenderText::setTextInternal): (WebCore::RenderText::setText): (WebCore::RenderText::previousOffset): (WebCore::RenderText::previousOffsetForBackwardDeletion): (WebCore::RenderText::nextOffset):
  • rendering/RenderText.h: (WebCore::RenderText::text): (WebCore::RenderText::characters): (WebCore::RenderText::textLength):
00:27 Changeset [56046] by eric@webkit.org
  • 8 edits
    2 adds in trunk

2010-03-16 Adam Barth <abarth@webkit.org>

Reviewed by Darin Adler.

noscript tag should render when @sandbox disables JavaScript
https://bugs.webkit.org/show_bug.cgi?id=36092

Test that the noscript element renders when @sandbox disables
JavaScript.

  • fast/frames/sandboxed-iframe-noscript-expected.txt: Added.
  • fast/frames/sandboxed-iframe-noscript.html: Added.

2010-03-16 Adam Barth <abarth@webkit.org>

Reviewed by Darin Adler.

noscript tag should render when @sandbox disables JavaScript
https://bugs.webkit.org/show_bug.cgi?id=36092

Instead of talking to Settings directly to figure out if JavaScript is
enabled in a frame, we need to talk to the ScriptController. The
ScriptController is better at answering that question because it knows
about @sandbox.

Test: fast/frames/sandboxed-iframe-noscript.html

  • dom/Document.cpp: (WebCore::Document::Document):
  • html/HTMLCanvasElement.cpp: (WebCore::HTMLCanvasElement::createRenderer):
  • html/HTMLElement.cpp: (WebCore::HTMLElement::rendererIsNeeded):
  • html/HTMLParser.cpp: (WebCore::HTMLParser::noscriptCreateErrorCheck): (WebCore::HTMLParser::isInline):
  • plugins/PluginView.cpp: (WebCore::PluginView::load):
00:11 Changeset [56045] by eric@webkit.org
  • 4 edits in trunk/WebKitTools

2010-03-15 Eric Seidel <eric@webkit.org>

Reviewed by Adam Barth.

Add "what-broke" command for debugging when the tree broke
https://bugs.webkit.org/show_bug.cgi?id=36157

This is another step towards automated sheriffing of the webkit tree.
With this logic our scripts are able to determine what revision broke the
tree. Buildbot should do this for us, but unfortunately buildbot doesn't
expose this kind of aggregate information.

  • Scripts/webkitpy/buildbot.py:
    • Add new Builder and Build classes (which will eventually replace the custom dictionaries previously used).
    • Split out more network logic into _fetch methods which will eventually be their own class for mocking.
    • Use XMLRPC to communicate with the buildbot master instead of scraping build pages.
  • Scripts/webkitpy/buildbot_unittest.py:
    • Test the newly added buildbot classes.
  • Scripts/webkitpy/commands/queries.py:
    • Add an experimental what-broke command.
00:08 Changeset [56044] by eric@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-15 Daniel Bates <dbates@rim.com>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=36048

Detect if the Windows Platform SDK is missing when building with
Visual C++ Express Edition and inform the user to download it.

  • Scripts/webkitdirs.pm:

03/15/10:

23:42 Changeset [56043] by eric@webkit.org
  • 6 edits in trunk

2010-03-15 John Gregg <johnnyg@google.com>

Reviewed by David Levin.

Notification object should expose absolute URL of icon
https://bugs.webkit.org/show_bug.cgi?id=35800

Covered by existing tests.

  • notifications/Notification.cpp: (WebCore::Notification::Notification):
  • notifications/Notification.h: (WebCore::Notification::iconURL):

2010-03-15 John Gregg <johnnyg@google.com>

Reviewed by David Levin.

Notification object should expose absolute URL of icon
https://bugs.webkit.org/show_bug.cgi?id=35800

  • public/WebNotification.h:
  • src/WebNotification.cpp: (WebKit::WebNotification::iconURL):
23:25 Changeset [56042] by eric@webkit.org
  • 1 edit
    1 add in trunk/WebCore

2010-03-15 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Holger Freyther.

Add EFL port files to platform/efl.
http://webkit.org/b/35887

  • platform/efl/ClipboardEfl.cpp: Added.
  • platform/efl/PasteboardEfl.cpp: Added.
22:56 Changeset [56041] by tkent@chromium.org
  • 4 edits
    6 adds in trunk

2010-03-15 Kent Tamura <tkent@chromium.org>

Reviewed by Eric Seidel.

Don't submit disabled menu options.
https://bugs.webkit.org/show_bug.cgi?id=35056

  • fast/forms/menulist-disabled-selected-option-expected.txt: Added.
  • fast/forms/menulist-disabled-selected-option.html: Added.
  • fast/forms/menulist-submit-without-selection-expected.txt: Added.
  • fast/forms/menulist-submit-without-selection.html: Added.
  • fast/forms/script-tests/menulist-disabled-selected-option.js: Added.
  • fast/forms/script-tests/menulist-submit-without-selection.js: Added.

2010-03-15 Kent Tamura <tkent@chromium.org>

Reviewed by Eric Seidel.

Don't submit disabled menu options.
https://bugs.webkit.org/show_bug.cgi?id=35056

Test: fast/forms/menulist-disabled-selected-option.html

  • dom/OptionElement.h:
  • dom/SelectElement.cpp: (WebCore::SelectElement::appendFormData): If a selected option is disabled, skip it. Remove code for non-selected menulist because of compatibility with other browsers.
22:55 Changeset [56040] by abarth@webkit.org
  • 7 edits in trunk/WebKitTools

2010-03-15 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Let commit-queue land rollout patches even when the tree is red
https://bugs.webkit.org/show_bug.cgi?id=36155

Now the commit-queue will land patches whose name begins with "ROLLOUT "
even if the tree is red. The patches still go through the usual build
and test process, but they can be landed while the tree is on fire.

  • Scripts/webkitpy/bugzilla.py:
  • Scripts/webkitpy/commands/queues.py:
  • Scripts/webkitpy/commands/queues_unittest.py:
  • Scripts/webkitpy/commands/queuestest.py:
  • Scripts/webkitpy/mock_bugzillatool.py:
22:50 Changeset [56039] by eric@webkit.org
  • 1 edit
    1 add in trunk/WebCore

2010-03-15 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Holger Freyther.

Add EFL port files to platform/efl.
http://webkit.org/b/35884

  • platform/efl/FileSystemEfl.cpp: Added.
22:33 Changeset [56038] by eric@webkit.org
  • 1 edit
    1 add in trunk/WebCore

2010-03-15 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Holger Freyther.

Add EFL port files to platform/efl.
http://webkit.org/b/35882

  • platform/efl/ContextMenuItemEfl.cpp: Added.
22:32 Changeset [56037] by mjs@apple.com
  • 2 edits in trunk/WebKitSite

Tweak canvas perf demo.

  • demos/canvas-perf/canvas.html:
22:15 Changeset [56036] by eric@webkit.org
  • 4 edits in trunk

2010-03-15 Joanmarie Diggs <joanmarie.diggs@gmail.com>

Reviewed by Holger Freyther.

https://bugs.webkit.org/show_bug.cgi?id=35502
[Gtk] Objects of ATK_ROLE_TABLE should not implement AtkText

Simple change to stop tables from implementing AtkText.

  • accessibility/gtk/AccessibilityObjectAtk.cpp: (getInterfaceMaskFromObject):

2010-03-15 Joanmarie Diggs <joanmarie.diggs@gmail.com>

Reviewed by Holger Freyther.

https://bugs.webkit.org/show_bug.cgi?id=35502
[Gtk] Objects of ATK_ROLE_TABLE should not implement AtkText

New test to be sure we do not accidentally implement AtkText for tables

  • tests/testatk.c (testWebkitAtkGetTextInTable): (main):
21:58 Changeset [56035] by eric@webkit.org
  • 6 edits in trunk

2010-03-15 Adam Bergkvist <adam.bergkvist@ericsson.com>

Reviewed by Adam Barth.

According to the updated specification, a data field should always
result in a newline character being appended to the data buffer
regardless if the data buffer contains any data or not. However, upon
event dispatch, the last newline will be removed. This differs from an
older version of the specification where a newline character was
appended before the data value only if the buffer already contained
data. As a result, EventSource now supports receiving events with empty
data or newline characters only. Updated test accordingly.
https://bugs.webkit.org/show_bug.cgi?id=33210

  • http/tests/eventsource/eventsource-parse-event-stream-expected.txt:
  • http/tests/eventsource/eventsource-parse-event-stream.html:
  • http/tests/eventsource/resources/event-stream.php:

2010-03-15 Adam Bergkvist <adam.bergkvist@ericsson.com>

Reviewed by Adam Barth.

According to the updated specification, a data field should always
result in a newline character being appended to the data buffer
regardless if the data buffer contains any data or not. However, upon
event dispatch, the last newline will be removed. This differs from an
older version of the specification where a newline character was
appended before the data value only if the buffer already contained
data. As a result, EventSource now supports receiving events with empty
data or newline characters only.
https://bugs.webkit.org/show_bug.cgi?id=33210

  • page/EventSource.cpp: (WebCore::EventSource::parseEventStreamLine):
21:43 Changeset [56034] by abarth@webkit.org
  • 4 edits
    1 add in trunk/WebKitTools

2010-03-15 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

Add webkit-patch post-rollout to upload rollouts to bugs.webkit.org for easy committing
https://bugs.webkit.org/show_bug.cgi?id=36154

This new command is a mashup of prepare-rollout and post. This command
will be used by an experimental bot to post rollouts of patches that
break things to bugs.webkit.org where they can be landed with the
greatest of ease.

  • Scripts/webkitpy/commands/download.py:
  • Scripts/webkitpy/commands/download_unittest.py:
  • Scripts/webkitpy/steps/init.py:
  • Scripts/webkitpy/steps/postdiffforrevert.py: Added.
21:41 Changeset [56033] by eric@webkit.org
  • 3 edits
    2 adds in trunk

2010-03-15 Valters Svabe <vsvabe@gmail.com>

Reviewed by Darin Adler.

:after selector displays in wrong place with nested div
https://bugs.webkit.org/show_bug.cgi?id=32288

  • fast/css-generated-content/block-after-expected.txt: Added.
  • fast/css-generated-content/block-after.html: Added.

2010-03-15 Valters Svabe <vsvabe@gmail.com>

Reviewed by Darin Adler.

:after selector displays in wrong place with nested div
https://bugs.webkit.org/show_bug.cgi?id=32288

Make sure we don't append things after :after-generated content, if
it's inside an anonymous block.

Test: fast/css-generated-content/block-after.html

  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::addChild):
21:15 Changeset [56032] by abarth@webkit.org
  • 2 edits in trunk/WebKitTools

2010-03-15 Adam Barth <abarth@webkit.org>

Reviewed by Eric Seidel.

webkit-patch rollout should error out on conflicts
https://bugs.webkit.org/show_bug.cgi?id=36151

Instead of blindingly plowing ahead, we now throw an exception if there
are conflicts after applying a reverse diff.

  • Scripts/webkitpy/scm.py:
21:07 Changeset [56031] by eric@webkit.org
  • 1 edit
    1 add in trunk/JavaScriptCore

2010-03-15 Kwang Yul Seo <skyul@company100.net>

Reviewed by Eric Seidel.

[BREWMP] Add a function to create a BREW instance without local variable declarations.
https://bugs.webkit.org/show_bug.cgi?id=34705

Add a template function to create a BREW instance in one line.

  • wtf/brew/ShellBrew.h: Added. (WTF::createInstance):
20:27 Changeset [56030] by darin@chromium.org
  • 2 edits in trunk/WebKit/chromium

2010-03-15 Darin Fisher <darin@chromium.org>

Fix build bustage. We also need to pull down "third_party/tcmalloc"

  • DEPS:
19:36 Changeset [56029] by ggaren@apple.com
  • 2 edits in trunk/JavaScriptCore

Not reviewed.

Removed a now-incorrect comment I forgot to remove in my last check-in.

  • wtf/FastMalloc.cpp:

(WTF::TCMalloc_PageHeap::scavenge):

19:21 Changeset [56028] by ggaren@apple.com
  • 2 edits in trunk/JavaScriptCore

Fixed a portion of:
<rdar://problem/7165917> | https://bugs.webkit.org/show_bug.cgi?id=28676
Safari 4 does not release memory back to the operating system fast enough (28676)

Reviewed by Sam Weinig.

Every few seconds, release a percentage of the minimum unused page count
during that time period.

SunSpider reports no change, command-line or in-browser, Mac or Windows.

  • wtf/FastMalloc.cpp:

(WTF::TCMalloc_PageHeap::init):
(WTF::TCMalloc_PageHeap::signalScavenger):
(WTF::TCMalloc_PageHeap::initializeScavenger): Renamed shouldContinueScavenging
to shouldScavenge, since scavenging is no longer something that we interrupt.

(WTF::TCMalloc_PageHeap::scavenge): The new scavenging algorithm. Fixes
a bug where the old code would release only one item from each size class
per scavenge, potentially leaving large numbers of large-sized objects
unreleased for a long time.

(WTF::TCMalloc_PageHeap::shouldScavenge):
(WTF::TCMalloc_PageHeap::New):
(WTF::TCMalloc_PageHeap::AllocLarge):
(WTF::TCMalloc_PageHeap::Delete):
(WTF::TCMalloc_PageHeap::GrowHeap):
(WTF::TCMalloc_PageHeap::scavengerThread):
(WTF::TCMalloc_PageHeap::periodicScavenge): Updated to track the minimum
value of free_committed_pages_ during a given scavenge period.

17:06 Changeset [56027] by darin@chromium.org
  • 2 edits in trunk/WebKit/chromium

2010-03-15 Darin Fisher <darin@chromium.org>

Fix build bustage. We also need to pull down "gfx"

  • DEPS:
16:54 Changeset [56026] by cfleizach@apple.com
  • 4 edits in trunk

WebKitTools: Unreviewed layout test fix.

VO not able to perform a VO-spacebar on facebook links
https://bugs.webkit.org/show_bug.cgi?id=36132

GTK needs to implement press for this test to work.

  • DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:

(AccessibilityUIElement::press):

LayoutTests: Unreviewed.

VO not able to perform a VO-spacebar on facebook links
https://bugs.webkit.org/show_bug.cgi?id=36132

GTK needs to implement press in DRT before this test can work.

  • platform/gtk/Skipped:
16:28 Changeset [56025] by cfleizach@apple.com
  • 4 edits in trunk

Unreviewed layout test fix.

WebKitTools:

VO not able to perform a VO-spacebar on facebook links
https://bugs.webkit.org/show_bug.cgi?id=36132

Windows needs to implement press in DRT.

  • DumpRenderTree/win/AccessibilityUIElementWin.cpp:

(AccessibilityUIElement::press):

LayoutTests:

VO not able to perform a VO-spacebar on facebook links
https://bugs.webkit.org/show_bug.cgi?id=36132

This test must be skipped until https://bugs.webkit.org/show_bug.cgi?id=36145 is done.

  • platform/win/Skipped:
16:28 Changeset [56024] by darin@chromium.org
  • 5 edits in trunk/WebKit/chromium

2010-03-15 Darin Fisher <darin@chromium.org>

Reviewed by David Levin.

https://bugs.webkit.org/show_bug.cgi?id=36131
Update include paths for chromium Rect, Point, and Size types.

Update DEPS to pull latest chromium code that reflects the
new location of these headers.

  • DEPS:
  • public/WebPoint.h:
  • public/WebRect.h:
  • public/WebSize.h:
16:20 Changeset [56023] by eric@webkit.org
  • 1 edit
    1 add in trunk/WebCore

2010-03-15 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Eric Seidel.

Add EFL port files to platform/efl.
http://webkit.org/b/35885

  • platform/efl/KURLEfl.cpp: Added.
16:06 Changeset [56022] by cfleizach@apple.com
  • 3 edits in trunk/WebKitTools

Unreviewed. Fix break of layout tests on win and gtk.

VO not able to perform a VO-spacebar on facebook links
https://bugs.webkit.org/show_bug.cgi?id=36132

Attempting to implement press action for windows and gtk.

  • DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:

(AccessibilityUIElement::press):

  • DumpRenderTree/win/AccessibilityUIElementWin.cpp:

(AccessibilityUIElement::press):

15:59 Changeset [56021] by barraclough@apple.com
  • 3 edits in trunk/JavaScriptCore

https://bugs.webkit.org/show_bug.cgi?id=35843
Re-land reverted fix to JSString::getIndex()

Reviewed by Sam Weinig.

Calling getIndex() on a JSString in rope form may result in a JSException being thrown
if there is insuficient memory so value(exec) returns UString() with length zero,
which will be passed to jsSingleCharacterSubstring.
Add a slow case function to trap the error & return a safe null value, until the
exception is handled.

  • runtime/JSString.cpp:

(JSC::JSString::getIndexSlowCase):
(JSC::JSString::getStringPropertyDescriptor):

  • runtime/JSString.h:

(JSC::jsSingleCharacterSubstring):
(JSC::JSString::getIndex):
(JSC::jsSingleCharacterString):
(JSC::JSString::getStringPropertySlot):

15:38 Changeset [56020] by cfleizach@apple.com
  • 6 edits in trunk/WebKitTools

VO not able to perform a VO-spacebar on facebook links
https://bugs.webkit.org/show_bug.cgi?id=36132

Reviewed by Beth Dakin.

  • DumpRenderTree/AccessibilityUIElement.cpp:

(pressCallback):
(AccessibilityUIElement::getJSClass):

  • DumpRenderTree/AccessibilityUIElement.h:
  • DumpRenderTree/gtk/AccessibilityUIElementGtk.cpp:

(AccessibilityUIElement::press):

  • DumpRenderTree/mac/AccessibilityUIElementMac.mm:

(AccessibilityUIElement::press):

  • DumpRenderTree/win/AccessibilityUIElementWin.cpp:

(AccessibilityUIElement::press):

15:07 Changeset [56019] by cfleizach@apple.com
  • 3 edits
    2 adds in trunk

VO not able to perform a VO-spacebar on facebook links
https://bugs.webkit.org/show_bug.cgi?id=36132

Reviewed by Beth Dakin.

WebCore:

When a label element is used as a click event handler, and it doesn't have any
corresponding control, it should handle the action itself.

Test: accessibility/label-element-press.html

  • html/HTMLLabelElement.cpp:

(WebCore::HTMLLabelElement::accessKeyAction):

LayoutTests:

  • accessibility/label-element-press-expected.txt: Added.
  • accessibility/label-element-press.html: Added.
14:56 Changeset [56018] by cwzwarich@webkit.org
  • 2 edits in trunk/WebCore

2010-03-15 Cameron Zwarich <zwarich@apple.com>

Reviewed by Geoff Garen.

Bug 36135 - GCController::garbageCollectNow() crashes when called from garbage collection
<https://bugs.webkit.org/show_bug.cgi?id=36135>
<rdar://problem/7752552>

  • bindings/js/GCController.cpp:

(WebCore::GCController::garbageCollectNow): Fail silently if garbage collection is
active rather than hitting the CRASH() reentrancy guard.

14:47 Changeset [56017] by simon.fraser@apple.com
  • 2 edits in trunk/WebCore

2010-03-15 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein.

Drop out of compositing mode when no elements need to composite
https://bugs.webkit.org/show_bug.cgi?id=36134

When no more elements on the page need to be composited, drop out of compositing
mode altogether, including removing the layer-backed view that hosts the layers.

  • rendering/RenderLayerCompositor.cpp: (WebCore::RenderLayerCompositor::updateCompositingLayers): If the list of child layers to attach at the root is empty, we have no composited layers, and can drop out of compositing mode by calling willMoveOffscreen().

(WebCore::RenderLayerCompositor::computeCompositingRequirements): In several places,
use willBeComposited rather than calling needsToBeComposited(), because it avoids
recomputing state that we know already (but assert that we're getting it right).
When we're assessing the root layer in post-order, see if we have any child layers that
are compositing, and also whether the root itself needs to composite. If neither are
true, then we can drop out of compositing mode.

(WebCore::RenderLayerCompositor::needsToBeComposited):
(WebCore::RenderLayerCompositor::requiresCompositingLayer): Move the
"inCompositingMode() && layer->isRootLayer()" test out of requiresCompositingLayer()
and into needsToBeComposited() so that we can call requiresCompositingLayer() separately
on the root layer above.

14:33 Changeset [56016] by eric@webkit.org
  • 1 edit
    1 add in trunk/WebCore

2010-03-15 Leandro Pereira <leandro@profusion.mobi>

Reviewed by Kenneth Rohde Christiansen.

Add EFL port files to platform/efl.
http://webkit.org/b/35888

  • platform/efl/SoundEfl.cpp: Added.
14:21 Changeset [56015] by kenneth@webkit.org
  • 2 edits in trunk/JavaScriptCore

Add a long long version of abs() for MSVC.

Reviewed by Adam Roben.

  • wtf/MathExtras.h:

(abs):

14:02 Changeset [56014] by mrowe@apple.com
  • 2 edits in trunk/WebKitTools

Add a new build slave to replace the existing SnowLeopard Leaks build slave which
appears to be suffering a slow and painful death at the hands of its graphics hardware.

  • BuildSlaveSupport/build.webkit.org-config/config.json:
14:01 Changeset [56013] by simon.fraser@apple.com
  • 19 edits
    2 adds in trunk/LayoutTests

2010-03-15 Simon Fraser <simon.fraser@apple.com>

Rubber-stamped by Dan Bernstein.

Update some pixel results for composting tests; some were incorrect, and others had minor
pixel differences.

  • platform/mac/compositing/color-matching/image-color-matching-expected.checksum:
  • platform/mac/compositing/color-matching/image-color-matching-expected.png:
  • platform/mac/compositing/direct-image-compositing-expected.checksum:
  • platform/mac/compositing/direct-image-compositing-expected.png:
  • platform/mac/compositing/overflow/overflow-scroll-expected.checksum:
  • platform/mac/compositing/overflow/overflow-scroll-expected.png:
  • platform/mac/compositing/reflections/animation-inside-reflection-expected.checksum:
  • platform/mac/compositing/reflections/animation-inside-reflection-expected.png:
  • platform/mac/compositing/reflections/load-video-in-reflection-expected.checksum:
  • platform/mac/compositing/reflections/load-video-in-reflection-expected.png:
  • platform/mac/compositing/reflections/nested-reflection-animated-expected.checksum:
  • platform/mac/compositing/reflections/nested-reflection-animated-expected.png:
  • platform/mac/compositing/reflections/nested-reflection-transition-expected.checksum:
  • platform/mac/compositing/reflections/nested-reflection-transition-expected.png:
  • platform/mac/compositing/reflections/reflection-on-composited-expected.checksum:
  • platform/mac/compositing/reflections/reflection-on-composited-expected.png:
  • platform/mac/compositing/repaint/become-overlay-composited-layer-expected.checksum: Added.
  • platform/mac/compositing/repaint/become-overlay-composited-layer-expected.png: Added.
  • platform/mac/compositing/repaint/layer-repaint-rects-expected.checksum:
  • platform/mac/compositing/repaint/layer-repaint-rects-expected.png:
13:31 Changeset [56012] by enrica@apple.com
  • 18 edits
    4 adds in trunk

WebCore: For continuous scrolling devices on the Mac (trackpads and Mighty/Magic
Mice), set the number of wheel ticks equal to the atomic increment of
the device: a pixel. This ensures that any scrolling action will
generate at least one wheel tick, regardless of the speed/length of
the gesture.

Reviewed by John Sullivan.

https://bugs.webkit.org/show_bug.cgi?id=29601
<rdar://problem/7453254>

Tests: fast/events/continuous-platform-wheelevent-in-scrolling-div.html

fast/events/platform-wheelevent-in-scrolling-div.html

  • platform/mac/WebCoreSystemInterface.h: Modify method signature.
  • platform/mac/WebCoreSystemInterface.mm: Ditto.
  • platform/mac/WheelEventMac.mm:

(WebCore::PlatformWheelEvent::PlatformWheelEvent): Call into
WebKitSystemInterface to determine both wheelTicksX/Y and deltaX/Y.
WKSI will query the event system for number of wheel ticks and
pixels/lines scrolled based on the device type.

WebKit/mac: Updated call to WKGetWheelEventDeltas() to match new method signature.

Reviewed by John Sullivan.

https://bugs.webkit.org/show_bug.cgi?id=29601
<rdar://problem/7453254>

  • WebView/WebDynamicScrollBarsView.mm:

(-[WebDynamicScrollBarsView scrollWheel:]):

WebKitLibraries: Added two new output arguments to WKGetWheelEventDeltas() to return
the number of scroll wheel ticks in the x and y directions.

Reviewed by John Sullivan.

https://bugs.webkit.org/show_bug.cgi?id=29601.
<rdar://problem/7453254>

  • WebKitSystemInterface.h:
  • libWebKitSystemInterfaceLeopard.a:
  • libWebKitSystemInterfaceSnowLeopard.a:
  • libWebKitSystemInterfaceTiger.a:

LayoutTests: These tests verify that scrollwheel events (both continuous and
non-continuous) generate the expected pixel deltas in scrollable
regions, and also generate DOM mousewheel events with expected
wheelDelta values.

Reviewed by John Sullivan.

https://bugs.webkit.org/show_bug.cgi?id=29601
<rdar://problem/7453254>

  • fast/events/continuous-platform-wheelevent-in-scrolling-div-expected.txt: Added.
  • fast/events/continuous-platform-wheelevent-in-scrolling-div.html: Added.
  • fast/events/platform-wheelevent-in-scrolling-div-expected.txt: Added.
  • fast/events/platform-wheelevent-in-scrolling-div.html: Added.
  • fast/events/wheelevent-in-scrolling-div-expected.txt: Rebased

expected results due to change in test output formatting.

  • fast/events/wheelevent-in-scrolling-div.html: Modified to use common

js logging and assertion functions.

  • platform/gtk/Skipped: DRT support needed by tests not yet implemented

in GTK port. See https://bugs.webkit.org/show_bug.cgi?id=36003.

  • platform/mac-tiger/Skipped: DRT support needed by tests not possible

to implement in Tiger.

  • platform/qt/Skipped: DRT support for tests not yet implemented in Qt

port. See https://bugs.webkit.org/show_bug.cgi?id=36004.

  • platform/win/Skipped: DRT support for tests not yet implemented in

Windows port. See https://bugs.webkit.org/show_bug.cgi?id=36002.

13:30 Changeset [56011] by eric@webkit.org
  • 3 edits in trunk/WebKit/chromium

2010-03-15 Marcus Bulach <bulach@chromium.org>

Reviewed by Jeremy Orlow.

Send the frame URL on GeolocationServiceBridgeImpl::startUpdating.
https://bugs.webkit.org/show_bug.cgi?id=36012

  • public/GeolocationServiceBridgeChromium.h:
  • src/GeolocationServiceBridgeChromium.cpp: (WebKit::GeolocationServiceBridgeImpl::startUpdating):
13:18 Changeset [56010] by philn@webkit.org
  • 3 edits in trunk/WebCore

2010-03-15 Philippe Normand <pnormand@igalia.com>

Reviewed by Gustavo Noronha Silva.

[GStreamer] duration query optimizations
https://bugs.webkit.org/show_bug.cgi?id=36116

Don't reattempt duration queries that previously failed and cache
media duration only if it's known.

  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp: (WebCore::MediaPlayerPrivate::MediaPlayerPrivate): (WebCore::MediaPlayerPrivate::duration): (WebCore::MediaPlayerPrivate::updateStates): (WebCore::MediaPlayerPrivate::durationChanged):
  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:
13:11 Changeset [56009] by philn@webkit.org
  • 3 edits in trunk/WebCore

2010-03-15 Philippe Normand <pnormand@igalia.com>

Reviewed by Gustavo Noronha Silva.

[GStreamer] updateStates called for all elements posting a state-change
https://bugs.webkit.org/show_bug.cgi?id=36115

  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp: (WebCore::mediaPlayerPrivateMessageCallback): Update player state only for state-change messages coming from playbin2 directly.
  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h: (WebCore::MediaPlayerPrivate::pipeline): public accessor for the playbin element.
13:10 Changeset [56008] by sullivan@apple.com
  • 5 edits in trunk/WebKit/mac

-[WebFrame setAlwaysHideHorizontal/VerticalScroller:] prevents keyboard scrolling
<https://bugs.webkit.org/show_bug.cgi?id=36125>

Reviewed by Adam Roben.

  • WebView/WebDynamicScrollBarsView.h:

Added instance variables horizontalScrollingAllowedButScrollerHidden and
verticalScrollingAllowedButScrollerHidden. Renamed instance variables
hideHorizontal/VerticalScroller to alwaysHideHorizontal/VerticalScroller for clarity.
Declared methods -horizontalScrollingAllowed and -verticalScrollingAllowed.
Added comments.

  • WebView/WebDynamicScrollBarsView.mm:

(-[WebDynamicScrollBarsView setAlwaysHideHorizontalScroller:]):
Updated for instance variable renaming.
(-[WebDynamicScrollBarsView setAlwaysHideVerticalScroller:]):
Ditto.
(-[WebDynamicScrollBarsView horizontalScrollingAllowed]):
New method, returns YES if the scroller is showing or the only reason that the scroller
is not showing is that setAlwaysHideHorizontalScrolling has been called.
(-[WebDynamicScrollBarsView verticalScrollingAllowed]):
New method, returns YES if the scroller is showing or the only reason that the scroller
is not showing is that setAlwaysHideVerticalScrolling has been called.
(-[WebDynamicScrollBarsView updateScrollers]):
Updated for instance variable renamings. Now updates horizontalScrollingAllowedButScrollerHidden
and verticalScrollingAllowedButScrollerHidden. Now takes the always-hidden state into account
in the early-return code path, to avoid taking it into account twice in the regular code path.

  • WebView/WebFrameView.mm:

(-[WebFrameView _scrollToBeginningOfDocument]):
Use _isScrollable instead of _hasScrollBars.
(-[WebFrameView _scrollToEndOfDocument]):
Ditto.
(-[WebFrameView scrollToBeginningOfDocument:]):
Use _largestScrollableChild instead of _largestChildWithScrollBars.
(-[WebFrameView scrollToEndOfDocument:]):
Ditto.
(-[WebFrameView _pageVertically:]):
Use _isScrollable and _largestScrollableChild instead of _hasScrollBars
and _largestChildWithScrollBars.
(-[WebFrameView _pageHorizontally:]):
Ditto.
(-[WebFrameView _scrollLineVertically:]):
Ditto.
(-[WebFrameView _scrollLineHorizontally:]):
Ditto.
(-[WebFrameView keyDown:]):
Use _largestScrollableChild instead of _largestChildWithScrollBars.
(-[WebFrameView _isScrollable]):
New method, calls -[WebDynamicScrollBarsView horizontalScrollingAllowed] and
-[WebDynamicScrollBarsView verticalScrollingAllowed]
(-[WebFrameView _largestScrollableChild]):
New method, like _largestChildWithScrollBars but uses _isScrollable.
(-[WebFrameView _hasScrollBars]):
Added a comment that this is no longer used by Safari, and can thus probably be
deleted once we no longer want to support it for nightly build compatibility with
old versions of Safari.
(-[WebFrameView _largestChildWithScrollBars]):
Ditto.

  • WebView/WebFrameViewPrivate.h:

Declared -_isScrollable and -_largestScrollableChild. Added comments to
_hasScrollBars and _largestChildWithScrollBars saying that they are no longer
used by Safari, and can thus probably be deleted once we no longer want to
support them for nightly build compatibility with old versions of Safari.

13:08 Changeset [56007] by pkasting@chromium.org
  • 10 edits in trunk/WebCore

Clean up more bits of the open-source image decoders. In addition to
simplifying things somewhat, this makes all the decoders the same in
terms of how they construct "image reader" subclasses and init scale
data.

Reviewed by Adam Barth.

https://bugs.webkit.org/show_bug.cgi?id=36040

No functional change, so no tests.

  • platform/image-decoders/ImageDecoder.h: Reorder a few declarations to try and group/order members slightly better.

(WebCore::ImageDecoder::ImageDecoder):

  • platform/image-decoders/gif/GIFImageDecoder.cpp: Move reader construction into decode(). Prep scale data in setSize(). Remove useless comment.

(WebCore::GIFImageDecoder::setData):
(WebCore::GIFImageDecoder::setSize):
(WebCore::GIFImageDecoder::frameBufferAtIndex):
(WebCore::GIFImageDecoder::decode):

  • platform/image-decoders/gif/GIFImageDecoder.h: Prep scale data in setSize().
  • platform/image-decoders/gif/GIFImageReader.cpp: Prep scale data in setSize().

(GIFImageReader::read):

  • platform/image-decoders/ico/ICOImageDecoder.cpp: Shorten code.

(WebCore::ICOImageDecoder::setSize):

  • platform/image-decoders/jpeg/JPEGImageDecoder.cpp: Fix style. Move reader construction into decode().

(WebCore::JPEGImageReader::close):
(WebCore::JPEGImageReader::skipBytes):
(WebCore::JPEGImageDecoder::decode):

  • platform/image-decoders/jpeg/JPEGImageDecoder.h: Move reader construction into decode().

(WebCore::JPEGImageDecoder::filenameExtension):

  • platform/image-decoders/png/PNGImageDecoder.cpp: Move reader construction into decode(). Track completion on decoder, not reader. Prep scale data in setSize(). Remove useless comment.

(WebCore::PNGImageReader::PNGImageReader):
(WebCore::PNGImageReader::close):
(WebCore::PNGImageReader::decode):
(WebCore::PNGImageDecoder::setSize):
(WebCore::PNGImageDecoder::frameBufferAtIndex):
(WebCore::PNGImageDecoder::headerAvailable):
(WebCore::PNGImageDecoder::pngComplete):
(WebCore::PNGImageDecoder::decode):

  • platform/image-decoders/png/PNGImageDecoder.h: Move reader construction into decode(). Track completion on decoder, not reader. Prep scale data in setSize().

(WebCore::PNGImageDecoder::filenameExtension):
(WebCore::PNGImageDecoder::isComplete):

13:00 Changeset [56006] by philn@webkit.org
  • 2 edits in trunk/WebCore

2010-03-15 Philippe Normand <pnormand@igalia.com>

Reviewed by Gustavo Noronha Silva.

[GTK] video playback position query flood when mouse over the video element
https://bugs.webkit.org/show_bug.cgi?id=35333

Don't trigger a position query only to know which play/pause
button to display.

  • platform/gtk/RenderThemeGtk.cpp: (WebCore::RenderThemeGtk::paintMediaPlayButton):
12:57 Changeset [56005] by dpranke@chromium.org
  • 2 edits in trunk/WebKitTools

2010-03-15 Dirk Pranke <dpranke@chromium.org>

Reviewed by Alexey Proskuryakov.

Fix a minor case where we'd deference a null pointer if we tried
to run new-run-webkit-tests on an unsupported platform (e.g.
Cygwin's python version).

https://bugs.webkit.org/show_bug.cgi?id=36076

  • Scripts/webkitpy/layout_tests/port/factory.py:
10:27 Changeset [56004] by jianli@chromium.org
  • 5 edits
    1 add in trunk/WebCore

[v8] Add V8 bindings for DOMFormData.
https://bugs.webkit.org/show_bug.cgi?id=36026

Reviewed by Nate Chapin.

Test: http/tests/local/send-form-data.html

  • Android.v8bindings.mk:
  • WebCore.gypi:
  • bindings/v8/V8Index.h:
  • bindings/v8/custom/V8DOMFormDataCustom.cpp: Added.
  • bindings/v8/custom/V8XMLHttpRequestCustom.cpp:

(WebCore::V8XMLHttpRequest::sendCallback):

10:27 Changeset [56003] by jianli@chromium.org
  • 29 edits
    6 adds in trunk

WebCore: Add DOMFormData.idl to expose FormData interface.
https://bugs.webkit.org/show_bug.cgi?id=36024

Reviewed by Sam Weinig.

The implementation is based on XMLHttpRequest 2 spec:
http://dev.w3.org/2006/webapi/XMLHttpRequest-2/#formdata

Test: http/tests/local/send-form-data.html

  • Android.derived.jscbindings.mk:
  • Android.derived.v8bindings.mk:
  • Android.jscbindings.mk:
  • DerivedSources.cpp:
  • DerivedSources.make:
  • GNUmakefile.am:
  • WebCore.pri:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • bindings/js/JSDOMFormDataCustom.cpp: Added.
  • bindings/js/JSXMLHttpRequestCustom.cpp:

(WebCore::JSXMLHttpRequest::send):

  • bindings/scripts/CodeGeneratorJS.pm:
  • html/DOMFormData.idl: Added.
  • page/DOMWindow.idl:

LayoutTests: Add DOMFormData.idl to expose FormData interface
https://bugs.webkit.org/show_bug.cgi?id=36024

Reviewed by Sam Weinig.

Add a layout test to test sending FormData via XMLHttpRequest. Also
changed all test results that are affected by exposing FormData.

  • fast/dom/Window/window-properties-expected.txt:
  • fast/dom/Window/window-property-descriptors-expected.txt:
  • fast/dom/constructed-objects-prototypes-expected.txt:
  • fast/dom/prototype-inheritance-2-expected.txt:
  • fast/dom/prototype-inheritance-expected.txt:
  • fast/js/global-constructors-expected.txt:
  • http/tests/local/resources/send-form-data.js: Added.
  • http/tests/local/send-form-data-expected.txt: Added.
  • http/tests/local/send-form-data.html: Added.
  • http/tests/xmlhttprequest/resources/multipart-post-echo.php: Added.
  • platform/gtk/Skipped:
  • platform/gtk/fast/dom/Window/window-properties-expected.txt:
  • platform/qt/Skipped:
  • platform/qt/fast/dom/Window/window-properties-expected.txt:
  • platform/qt/fast/dom/Window/window-property-descriptors-expected.txt:
  • platform/qt/fast/dom/prototype-inheritance-expected.txt:
  • platform/qt/fast/js/global-constructors-expected.txt:
  • platform/win/Skipped:
10:14 Changeset [56002] by eric@webkit.org
  • 14 edits in trunk

2010-03-15 Patrik Persson <patrik.j.persson@ericsson.com>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=32369

Revised test of storage/databases in sandboxed iframes.

Verify that SECURITY_ERR is raised on access to
window.localStorage or window.openDatabase(). Also verify that
window.sessionStorage is now allowed in sandboxed iframes.

WebKit would previously return null references in these cases. The
new behavior is in accordance with HTML5:

  • fast/frames/resources/sandboxed-iframe-storage-disallowed.html: expect exceptions, not null
  • fast/frames/sandboxed-iframe-storage-expected.txt: expect exceptions, not null

2010-03-15 Patrik Persson <patrik.j.persson@ericsson.com>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=32369

Revise iframe sandbox behavior to match the updated HTML5 spec.

  • Enables window.sessionStorage in sandboxed iframes.
  • Raises SECURITY_ERR exceptions when window.localStorage or window.openDatabase() is blocked by iframe sandboxing.

Note: window.sessionStorage does not raise exceptions.

WebKit would previously return null references in these cases. The
new behavior is in accordance with HTML5:

http://dev.w3.org/html5/webstorage/ (sections 4.2 and 4.3)
http://dev.w3.org/html5/webdatabase/ (section 4.1)
http://www.mail-archive.com/whatwg@lists.whatwg.org/msg19786.html

  • inspector/InspectorController.cpp: (WebCore::InspectorController::selectDOMStorage): exception handling
  • inspector/InspectorDOMStorageResource.cpp: (WebCore::InspectorDOMStorageResource::handleEvent): exception handling
  • page/DOMWindow.cpp: (WebCore::DOMWindow::sessionStorage): re-enabled in iframe sandbox (WebCore::DOMWindow::localStorage): raise exception rather than return null (WebCore::DOMWindow::openDatabase): raise exception rather than return null
  • page/DOMWindow.h: added exceptions to interface
  • page/DOMWindow.idl: added exceptions to interface
  • page/SecurityOrigin.h: (WebCore::SecurityOrigin::canAccessLocalStorage): renamed function to reflect its purpose
  • storage/StorageEventDispatcher.cpp: (WebCore::StorageEventDispatcher::dispatch): exception handling

2010-03-15 Patrik Persson <patrik.j.persson@ericsson.com>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=32369

Revise iframe sandbox behavior to match the updated HTML5 spec.

  • Enables window.sessionStorage in sandboxed iframes.
  • Raises SECURITY_ERR exceptions when window.localStorage or window.openDatabase() is blocked by iframe sandboxing.

Note: window.sessionStorage does not raise exceptions.

WebKit would previously return null references in these cases. The
new behavior is in accordance with HTML5:

http://dev.w3.org/html5/webstorage/ (sections 4.2 and 4.3)
http://dev.w3.org/html5/webdatabase/ (section 4.1)
http://www.mail-archive.com/whatwg@lists.whatwg.org/msg19786.html

  • src/StorageAreaProxy.cpp: (WebCore::StorageAreaProxy::storageEvent): exception handling
  • src/StorageEventDispatcherImpl.cpp: (WebCore::StorageEventDispatcherImpl::dispatchStorageEvent): exception handling
09:41 Changeset [56001] by darin@apple.com
  • 3 edits in trunk/WebKitTools

Tell Subversion about more directories that expect to have .pyc files.

  • Scripts/webkitpy/layout_tests: Added property svn:ignore.
  • Scripts/webkitpy/layout_tests/port: Added property svn:ignore.
09:33 Changeset [56000] by eric@webkit.org
  • 2 edits in trunk/JavaScriptCore

2010-03-15 Gabor Loki <loki@webkit.org>

Reviewed by Gavin Barraclough.

Combine ctiTrampolines on ARM and Thumb-2
https://bugs.webkit.org/show_bug.cgi?id=36014

  • jit/JITStubs.cpp: (JSC::JITThunks::JITThunks):
09:22 Changeset [55999] by apavlov@chromium.org
  • 9 edits
    1 add in trunk/WebCore

2010-03-15 Alexander Pavlov <apavlov@chromium.org>

Reviewed by Timothy Hatcher.

Web Inspector: Implement a progress indicator in the Audits panel
when resources are reloading
https://bugs.webkit.org/show_bug.cgi?id=35971

  • English.lproj/localizedStrings.js:
  • GNUmakefile.am:
  • WebCore.gypi:
  • inspector/front-end/AuditLauncherView.js: (WebInspector.AuditLauncherView.prototype.updateResourceTrackingState): (WebInspector.AuditLauncherView.prototype.get totalResources): (WebInspector.AuditLauncherView.prototype.set totalResources): (WebInspector.AuditLauncherView.prototype.get loadedResources): (WebInspector.AuditLauncherView.prototype.set loadedResources): (WebInspector.AuditLauncherView.prototype._resetResourceCount): (WebInspector.AuditLauncherView.prototype.resourceStarted): (WebInspector.AuditLauncherView.prototype.resourceFinished): (WebInspector.AuditLauncherView.prototype.reset): (WebInspector.AuditLauncherView.prototype._setAuditRunning): (WebInspector.AuditLauncherView.prototype._launchButtonClicked): (WebInspector.AuditLauncherView.prototype._createCategoryElement): (WebInspector.AuditLauncherView.prototype._createLauncherUI): (WebInspector.AuditLauncherView.prototype._updateResourceProgress): (WebInspector.AuditLauncherView.prototype._updateButton):
  • inspector/front-end/AuditsPanel.js: (WebInspector.AuditsPanel.prototype.resourceStarted): (WebInspector.AuditsPanel.prototype.resourceFinished): (WebInspector.AuditsPanel.prototype.reset):
  • inspector/front-end/Images/spinner.gif: Added.
  • inspector/front-end/WebKit.qrc:
  • inspector/front-end/audits.css: (.audit-launcher-view .resource-progress > img):
  • inspector/front-end/inspector.js: (WebInspector.updateResource):
08:55 QtWebKitBackportingFixes edited by laszlo.1.gombos@nokia.com
(diff)
08:54 Changeset [55998] by eric@webkit.org
  • 4 edits in trunk

2010-03-15 Tony Chang <tony@chromium.org>

Reviewed by Darin Adler.

option-pageup/down should move cursor on chromium mac
https://bugs.webkit.org/show_bug.cgi?id=36108

Remove waitUntilDone/notifyDone from the layout test, it's not needed
and causes timeouts on failure.

  • editing/input/option-page-up-down.html:

2010-03-15 Tony Chang <tony@chromium.org>

Reviewed by Darin Adler.

option-pageup/down should move cursor on chromium mac
https://bugs.webkit.org/show_bug.cgi?id=36108

  • src/EditorClientImpl.cpp: (WebKit::):
08:40 Changeset [55997] by sullivan@apple.com
  • 5 edits in trunk/WebKit/mac

Method name and parameter name mistakes from recent SPI addition
<https://bugs.webkit.org/show_bug.cgi?id=36119>

Reviewed by Dan Bernstein.

  • WebView/WebDynamicScrollBarsView.h:

Renamed instance variable and method name from "setAllowXXX" to "setAllowsXXX".

  • WebView/WebDynamicScrollBarsView.mm:

(-[WebDynamicScrollBarsView setAllowsScrollersToOverlapContent:]):
Updated for renamed instance variable and method.
(-[WebDynamicScrollBarsView setAlwaysHideHorizontalScroller:]):
Changed parameter name from shouldBeVisible to shouldBeHidden.
(-[WebDynamicScrollBarsView setAlwaysHideVerticalScroller:]):
Ditto.
(-[WebDynamicScrollBarsView contentViewFrame]):
Updated for renamed instance variable.
(-[WebDynamicScrollBarsView tile]):
Ditto.
(-[WebDynamicScrollBarsView reflectScrolledClipView:]):
Ditto.

  • WebView/WebFrame.mm:

(-[WebFrame setAllowsScrollersToOverlapContent:]):
Renamed method from setAllowXXX, and updated for same change in WebDynamicScrollBarsView.

  • WebView/WebFramePrivate.h:

Renamed method name from "setAllowXXX" to "setAllowsXXX".

08:34 Changeset [55996] by philn@webkit.org
  • 3 edits in trunk/WebCore

2010-03-15 Philippe Normand <pnormand@igalia.com>

Reviewed by Gustavo Noronha Silva.

[GStreamer] replace g_idle_add / g_timeout_add calls with Timers in the gstreamer player
https://bugs.webkit.org/show_bug.cgi?id=35735

Replaced g_idle_add calls with Timers immediately started once
only. The g_timeout_add was replaced by a periodic Timer.

  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp: (WebCore::MediaPlayerPrivate::MediaPlayerPrivate): (WebCore::MediaPlayerPrivate::~MediaPlayerPrivate): (WebCore::MediaPlayerPrivate::volumeChangedTimerFired): (WebCore::MediaPlayerPrivate::volumeChanged): (WebCore::MediaPlayerPrivate::processBufferingStats): (WebCore::MediaPlayerPrivate::fillTimerFired): (WebCore::MediaPlayerPrivate::maxTimeLoaded): (WebCore::MediaPlayerPrivate::updateStates): (WebCore::MediaPlayerPrivate::muteChangedTimerFired): (WebCore::MediaPlayerPrivate::muteChanged):
  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:
08:20 Changeset [55995] by kov@webkit.org
  • 5 edits in trunk

2010-03-12 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Eric Carlson.

media/video-preload.html fails
https://bugs.webkit.org/show_bug.cgi?id=35793

Only effectively load, and start buffering when playing, or when
the preload attribute is set.

Test: media/video-preload.html

  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp: (WebCore::MediaPlayerPrivate::MediaPlayerPrivate): (WebCore::MediaPlayerPrivate::load): (WebCore::MediaPlayerPrivate::commitLoad): (WebCore::MediaPlayerPrivate::prepareToPlay): (WebCore::MediaPlayerPrivate::setPreload):
  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:
07:46 Changeset [55994] by laszlo.1.gombos@nokia.com
  • 2 edits in trunk/WebKit/qt

2010-03-15 Laszlo Gombos <laszlo.1.gombos@nokia.com>

Unreviewed.

[Qt] Build fix if TILED_BACKING_STORE is disabled.

  • Api/qwebframe.cpp: (QWebFramePrivate::renderContentsLayerAbsoluteCoords):
07:34 Changeset [55993] by eric@webkit.org
  • 5 edits
    2 adds in trunk

2010-03-15 Shu Chang <Chang.Shu@nokia.com>

Reviewed by Holger Freyther.

[Qt] Change and add expected results after the fix and also remove the test from Skipped.
https://bugs.webkit.org/show_bug.cgi?id=36060

  • platform/qt/Skipped:
  • platform/qt/fast/canvas/fillrect-gradient-zero-stops-expected.checksum: Added.
  • platform/qt/fast/canvas/fillrect-gradient-zero-stops-expected.png: Added.
  • platform/qt/fast/canvas/fillrect-gradient-zero-stops-expected.txt:

2010-03-15 Shu Chang <Chang.Shu@nokia.com>

Reviewed by Holger Freyther.

[Qt] The behavior of QGradient with no stops is defined differently from HTML5 spec,
where the latter requires the gradient to be transparent black. Explicitly added a
transparent black color stop to match the HTML5 behavior.
https://bugs.webkit.org/show_bug.cgi?id=36060

  • platform/graphics/qt/GradientQt.cpp: (WebCore::Gradient::platformGradient):
07:13 Changeset [55992] by tonikitoo@webkit.org
  • 2 edits in trunk/WebCore

Some minor code clean ups in SpatialNavigation.cpp

Rubber-stamped by Holger Freyther.
Patch by Antonio Gomes <tonikitoo@webkit.org>

  • page/SpatialNavigation.cpp:

(WebCore::distanceInDirection):
(WebCore::deflateIfOverlapped):

06:02 Changeset [55991] by pfeldman@chromium.org
  • 2 edits in trunk/WebCore

2010-03-13 Ilya Tikhonovsky <loislo@chromium.org>

Reviewed by Timothy Hatcher.

WebInspector: Position of GraphBar elements is calculating more carefully
and they fully visible at the bottom of Timeline panel.


http://bugs.webkit.org/show_bug.cgi?id=35966

  • inspector/front-end/TimelinePanel.js: (WebInspector.TimelineRecordGraphRow.prototype.update):
02:10 Changeset [55990] by eric@webkit.org
  • 6 edits in trunk

2010-03-15 MORITA Hajime <morrita@google.com>

Reviewed by Alexey Proskuryakov.

Selection.modify extends too far with 'lineboundary'.
https://bugs.webkit.org/show_bug.cgi?id=33413

  • editing/selection/extend-selection-expected.txt:
  • platform/win/editing/selection/extend-selection-expected.txt: Updated result to correct expectation that described the wrong behaviour for the selection expansion with 'lineboundary' granularity.

2010-03-15 MORITA Hajime <morrita@google.com>

Reviewed by Alexey Proskuryakov.

Selection.modify extends too far with 'lineboundary'.
https://bugs.webkit.org/show_bug.cgi?id=33413

Selection.modify() with 'lineboundary' granularity implies just
"Go to the end of the line", unlike selection expansion with
other type of granularities. This change handled LineGranularity
as special case, to look-up end of line with UPSTREAM affinity.
Doing this prevents look-up algorithm to go next line.

Test: editing/selection/extend-selection-expected.txt

  • dom/Position.cpp: (WebCore::Position::getInlineBoxAndOffset): Handled an edge case for node look-up with UPSTREAM.
  • editing/SelectionController.cpp: (WebCore::SelectionController::modifyExtendingForward): Added UPSTREAM tweak for the case for LineGranularity.
01:34 Changeset [55989] by philn@webkit.org
  • 2 edits in trunk/WebKit/gtk

2010-03-09 Philippe Normand <pnormand@igalia.com>

Reviewed by Holger Freyther.

[GTK] GTK_WIDGET_IS_SENSITIVE is deprecated in GTK+ 2.20
https://bugs.webkit.org/show_bug.cgi?id=35909

  • webkit/webkitwebview.cpp: GTK_WIDGET_IS_SENSITIVE has been deprecated in gtk 2.20. Use gtk_widget_is_sensitive when available.
01:33 Changeset [55988] by yuzo@google.com
  • 6 edits
    5 adds in trunk

2010-03-14 Yuzo Fujishima <yuzo@google.com>

Reviewed by Shinichiro Hamaji.

Always call PrintContext.end() from PrintContext::pageNumberForElement().

Without this change, WebKit remains in printing mode after determining the page number if the specified element is found.

https://bugs.webkit.org/show_bug.cgi?id=36049

  • platform/gtk/Skipped:
  • platform/mac/printing/return-from-printing-mode-expected.checksum: Added.
  • platform/mac/printing/return-from-printing-mode-expected.png: Added.
  • platform/mac/printing/return-from-printing-mode-expected.txt: Added.
  • platform/qt/Skipped:
  • platform/win/Skipped:
  • printing/return-from-printing-mode.html: Added.

2010-03-14 Yuzo Fujishima <yuzo@google.com>

Reviewed by Shinichiro Hamaji.

Always call PrintContext.end() from PrintContext::pageNumberForElement().

Without this change, WebKit remains in printing mode after determining the page number if the specified element is found.

https://bugs.webkit.org/show_bug.cgi?id=36049

Test: printing/return-from-printing-mode.html

  • page/PrintContext.cpp: (WebCore::PrintContext::pageNumberForElement):
01:19 Changeset [55987] by philn@webkit.org
  • 2 edits in trunk/WebCore

2010-03-11 Philippe Normand <pnormand@igalia.com>

Reviewed by Gustavo Noronha Silva.

[GStreamer] progressTimer never stopped after EOS
https://bugs.webkit.org/show_bug.cgi?id=36007

Mark the pipeline as paused only if playback hasn't ended
yet. When didEnd() is triggered, the pipeline is paused (but
paused() still returns false) and a timeupdate event is fired,
those 2 conditions allow the HTMLMediaElement to stop the
progressTimer (in updatePlayerState, via mediaPlayerTimeChanged).

  • platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp: (WebCore::MediaPlayerPrivate::updateStates):
00:28 Changeset [55986] by eric@webkit.org
  • 1 edit
    5 adds in trunk/WebKitTools

2010-03-15 Adam Barth <abarth@webkit.org>

Reviewed by Darin Adler.

Operational scripts from running the EWS
https://bugs.webkit.org/show_bug.cgi?id=36097

These are the scripts I use to manage the EWS on EC2. If someone other
than me wants to run the EWS, these scripts might be helpful.

  • EWSTools/boot.sh: Added.
  • EWSTools/create-webkit-git: Added.
  • EWSTools/screen-config: Added.
  • EWSTools/start-queue.sh: Added.
00:04 Changeset [55985] by darin@apple.com
  • 2 edits in trunk/WebKit/mac

2010-03-14 Darin Adler <darin@apple.com>

Reviewed by Mark Rowe.

Remove unneeded dependency on non-string identifier for an NSTableColumn
https://bugs.webkit.org/show_bug.cgi?id=36106

  • WebView/WebTextCompletionController.mm: (-[WebTextCompletionController _buildUI]): Use init instead of initWithIdentifier: because the table has only one column and that column does not need an identifier.
Note: See TracTimeline for information about the timeline view.