Timeline


and

04/02/09:

23:49 Changeset [42198] by simon.fraser@apple.com
  • 3 edits
    4 adds in trunk

2009-04-02 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein

https://bugs.webkit.org/show_bug.cgi?id=25021

Fix repainting an element with a reflection and transform after layout.
Telling the reflection to repaint directly is wrong, because, with
LayoutState enabled, it only takes the reflection's transform into account,
not that of the element. We can fix this, and keep the benefits of LayoutState,
by computing the reflected repaint rect, and repainting that using the original
renderer.

Test: fast/repaint/reflection-repaint-test.html

  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::layoutBlock):
22:39 Changeset [42197] by simon.fraser@apple.com
  • 3 edits
    4 adds in trunk

2009-04-02 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein

https://bugs.webkit.org/show_bug.cgi?id=23307

When repainting during layout (when LayoutState is enabled), we need
to apply the transform to the dirty rect, before applying x(), y() and
the offset storted in the LayoutState.

Also bypass LayoutState in mapLocalToContainer() if we have a repaintContainer,
since LayoutState is always root-relative.

Test: fast/repaint/transform-layout-repaint.html

  • rendering/RenderBox.cpp: (WebCore::RenderBox::mapLocalToContainer): (WebCore::RenderBox::computeRectForRepaint):
20:10 Changeset [42196] by mrowe@apple.com
  • 2 edits in trunk/WebKit/mac

<rdar://problem/6684745> Crash in -[WebView removeSizeObservers] when loading NIB file

Reviewed by Dan Bernstein and Timothy Hatcher.

The implementation of -[NSView initWithCoder:] can result in -viewWillMoveToSuperview:/-viewDidMoveToSuperview: being sent to
our view before we've had a chance to initialize _private, so we need to ensure it is non-nil before dereferencing it in those
methods.

  • WebView/WebView.mm:

(-[WebView removeSizeObservers]): Nil-check _private before dereferencing it.
(-[WebView addSizeObservers]): Ditto.

18:07 Changeset [42195] by mrowe@apple.com
  • 2 edits in trunk/WebCore

Fix the Mac build after r42191.

  • WebCore.base.exp:
18:01 Changeset [42194] by mrowe@apple.com
  • 2 edits in trunk/WebCore

Fix Windows build after r42190.

  • page/EventHandler.cpp:

(WebCore::EventHandler::keyEvent): Update to use the new name of the constant.

17:42 Changeset [42193] by darin@apple.com
  • 1 edit in trunk/WebCore/ChangeLog

Fixed ChangeLog.

2009-04-02 Darin Adler <darin@apple.com>

Reviewed by Brady Eidson.

<rdar://problem/6625840> Previous Page isn't loaded after hitting back button at http://www.breakingnews.ie/

Test: http/tests/navigation/onload-navigation-iframe-2.html

  • html/HTMLFrameElement.cpp: (WebCore::HTMLFrameElement::HTMLFrameElement): Removed unneeded createdByParser flag.
  • html/HTMLFrameElement.h: Ditto.
  • html/HTMLFrameElementBase.cpp: (WebCore::HTMLFrameElementBase::HTMLFrameElementBase): Ditto.
  • html/HTMLFrameElementBase.h: Ditto.
  • html/HTMLFrameOwnerElement.cpp: (WebCore::HTMLFrameOwnerElement::HTMLFrameOwnerElement): Ditto.
  • html/HTMLFrameOwnerElement.h: Ditto.
  • html/HTMLIFrameElement.cpp: (WebCore::HTMLIFrameElement::HTMLIFrameElement): Ditto.
  • html/HTMLIFrameElement.h: Ditto.
  • html/HTMLPlugInElement.cpp: (WebCore::HTMLPlugInElement::HTMLPlugInElement): Ditto.
  • loader/FrameLoader.cpp: (WebCore::FrameLoader::submitForm): Removed unneeded check of the createdByParser flag. The original change that introduced this, http://trac.webkit.org/changeset/25410, does not make it clear why this check was needed. And it seems it is needed.
17:41 Changeset [42192] by darin@apple.com
  • 13 edits
    2 copies in trunk

WebCore:

2009-04-02 Adele Peterson <adele@apple.com>

Reviewed by Darin Adler.

Add a way to get a list of focusable nodes.

  • dom/Document.cpp: (WebCore::Document::setFocusedNode): (WebCore::Document::getFocusableNodes):
  • dom/Document.h:

LayoutTests:

2009-04-02 Darin Adler <darin@apple.com>

Reviewed by Brady Eidson.

<rdar://problem/6625840> Previous Page isn't loaded after hitting back button at http://www.breakingnews.ie/

  • http/tests/navigation/onload-navigation-iframe-2-expected.txt: Copied from LayoutTests/http/tests/navigation/onload-navigation-iframe-expected.txt. Updated for slight difference in the new test.
  • http/tests/navigation/onload-navigation-iframe-2.html: Copied from LayoutTests/http/tests/navigation/onload-navigation-iframe.html. Changed test to use an iframe element already in the HTML.
17:39 Changeset [42191] by adele@apple.com
  • 9 edits
    1 add in trunk

WebCore:

2009-04-02 Adele Peterson <adele@apple.com>

Reviewed by Darin Adler.

Add a way to get a list of focusable nodes.

  • dom/Document.cpp: (WebCore::Document::setFocusedNode): (WebCore::Document::getFocusableNodes):
  • dom/Document.h:

WebKit:

2009-04-02 Adele Peterson <adele@apple.com>

Reviewed by Darin Adler.

  • WebKit.xcodeproj/project.pbxproj: Add WebDOMOperationsInternal.h and make WebDOMOperationsPrivate.h private.

WebKit/mac:

2009-04-02 Adele Peterson <adele@apple.com>

Reviewed by Darin Adler.

Add a way to get a list of focusable nodes.

  • DOM/WebDOMOperations.mm: (-[DOMDocument _focusableNodes]):
  • DOM/WebDOMOperationsInternal.h: Added.
  • DOM/WebDOMOperationsPrivate.h: Make this a private header. Move old methods to WebDOMOperationsInternal.h
  • WebView/WebHTMLView.mm: Use methods from WebDOMOperationsInternal.h
17:19 Changeset [42190] by darin@apple.com
  • 2 edits in trunk/WebCore

2009-04-02 Darin Adler <darin@apple.com>

Reviewed by Sam Weinig.

<rdar://problem/6654255> crash due to null FrameView in EventHandler

Lots of code was added to EventHandler that used view() and page() without
null checking, but both of those can become null due to the frame lifetime.

  • page/EventHandler.cpp: (WebCore::EventHandler::handleMousePressEventSingleClick): (WebCore::EventHandler::eventMayStartDrag): (WebCore::EventHandler::handleAutoscroll): (WebCore::EventHandler::autoscrollTimerFired): (WebCore::EventHandler::setPanScrollCursor): (WebCore::EventHandler::allowDHTMLDrag): (WebCore::EventHandler::hitTestResultAtPoint): (WebCore::EventHandler::stopAutoscrollTimer): (WebCore::EventHandler::selectCursor): (WebCore::EventHandler::handleMousePressEvent): (WebCore::EventHandler::handleMouseMoveEvent): (WebCore::EventHandler::dispatchDragEvent): (WebCore::EventHandler::prepareMouseEvent): (WebCore::EventHandler::dispatchMouseEvent): (WebCore::EventHandler::handleWheelEvent): (WebCore::EventHandler::hoverTimerFired): (WebCore::EventHandler::keyEvent): (WebCore::EventHandler::dragHysteresisExceeded): (WebCore::EventHandler::shouldDragAutoNode): (WebCore::EventHandler::handleDrag): (WebCore::EventHandler::handleTextInputEvent): (WebCore::EventHandler::addPendingFrameUnloadEventCount): (WebCore::EventHandler::removePendingFrameUnloadEventCount): (WebCore::EventHandler::clearPendingFrameUnloadEventCount): (WebCore::EventHandler::addPendingFrameBeforeUnloadEventCount): (WebCore::EventHandler::removePendingFrameBeforeUnloadEventCount): (WebCore::EventHandler::clearPendingFrameBeforeUnloadEventCount): Added null checks for view() and page(). Made minor code cleanups.
14:58 Changeset [42189] by hyatt@apple.com
  • 2 edits in trunk/WebCore

2009-04-02 David Hyatt <hyatt@apple.com>

Reviewed by Oliver Hunt.

Fix regression in <object> layout test. Make sure that even though a 404 image loads, that <object>
still considers it an error for the purposes of firing onerror.

  • html/HTMLImageLoader.cpp: (WebCore::HTMLImageLoader::dispatchLoadEvent):
14:58 Changeset [42188] by pam@chromium.org
  • 4 edits
    3 adds in trunk/LayoutTests

2009-04-02 Pamela Greene <pam@chromium.org>

Reviewed by Darin Adler.

Split no-port test case into separate file, since its result may
depend on the behavior of a DNS proxy on some systems.
https://bugs.webkit.org/show_bug.cgi?id=21530

  • platform/mac-tiger/security/block-test-expected.txt: updated
  • platform/mac-tiger/security/block-test-no-port-expected.txt: Added.
  • platform/mac/security/block-test-expected.txt: updated
  • platform/mac/security/block-test-no-port-expected.txt: Added.
  • security/block-test-no-port.html: Added.
  • security/block-test.html: removed no-port test case
14:48 Changeset [42187] by dglazkov@chromium.org
  • 7 edits in trunk/WebCore

2009-04-02 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Timothy Hatcher.

https://bugs.webkit.org/show_bug.cgi?id=25010
Refactor InspectorController to replace all JS function invocations with
ScriptFunctionCall.

  • bindings/js/ScriptFunctionCall.cpp: (WebCore::ScriptFunctionCall::appendArgument): Added int, UString,

and JSValuePtr methods.

(WebCore::ScriptFunctionCall::call): Added simplified, exception-eating

method.

  • bindings/js/ScriptFunctionCall.h: Decls for above.
  • bindings/js/ScriptObjectQuarantine.cpp: Added helper for Node and DOMWindow. (WebCore::getQuarantinedScriptObject):
  • bindings/js/ScriptObjectQuarantine.h: Decls for above.
  • inspector/InspectorController.cpp: (WebCore::callSimpleFunction): Made into a static. (WebCore::InspectorController::focusNode): Refactored to use ScriptFunctionCall. (WebCore::InspectorController::toggleRecordButton): Ditto. (WebCore::InspectorController::startGroup): Ditto. (WebCore::InspectorController::setAttachedWindow): Ditto. (WebCore::InspectorController::inspectedWindowScriptObjectCleared): Ditto. (WebCore::InspectorController::addScriptProfile): Ditto. (WebCore::InspectorController::didParseSource): Ditto. (WebCore::InspectorController::failedToParseSource): Ditto. (WebCore::InspectorController::didPause): Ditto.
  • inspector/InspectorController.h: Removed callFunction and callSimpleFunction

decls.

14:30 Changeset [42186] by dimich@chromium.org
  • 4 edits in trunk/WebCore

2009-04-02 Jian Li <jianli@chromium.org>

Reviewed by Darin Fisher.

Upstream changes to V8WorkerContextEventListener for V8 bindings.
https://bugs.webkit.org/show_bug.cgi?id=25007

  • bindings/v8/V8ObjectEventListener.h:
  • bindings/v8/V8WorkerContextEventListener.h:
  • bindings/v8/V8WorkerContextEventListener.cpp:
14:29 Changeset [42185] by dimich@chromium.org
  • 1 edit
    2 adds in trunk/WebCore

2009-04-02 Jian Li <jianli@chromium.org>

Reviewed by Darin Fisher.

Upstream V8WorkerContextObjectEventListener for V8 bindings.
https://bugs.webkit.org/show_bug.cgi?id=25005

  • bindings/v8/V8WorkerContextObjectEventListener.cpp: Added.
  • bindings/v8/V8WorkerContextObjectEventListener.h: Added.
14:29 Changeset [42184] by dimich@chromium.org
  • 1 edit
    2 adds in trunk/WebCore

2009-04-02 Jian Li <jianli@chromium.org>

Reviewed by Darin Fisher.

Upstream V8EventListenerList for V8 bindings.
https://bugs.webkit.org/show_bug.cgi?id=25004

  • bindings/v8/V8EventListenerList.cpp: Added.
  • bindings/v8/V8EventListenerList.h: Added.
13:59 BadContent created by wsiegrist@apple.com
13:45 Changeset [42183] by aroben@apple.com
  • 2 edits in trunk/WebKitTools

Build fix for Windows machines that don't have WebKitTools/Scipts in their PATH

  • Scripts/pdevenv: Add WebKitTools/Scripts to PATH when launching Visual Studio/VC++ Express.
13:31 Changeset [42182] by aroben@apple.com
  • 2 edits in trunk/WebKitTools

Use pdevenv when building .vcproj files via our scripts

Reviewed by Mark Rowe.

  • Scripts/webkitdirs.pm: (sub buildVisualStudioProject): Use pdevenv instead of invoking Visual Studio/VC++ Express directly.
13:22 Changeset [42181] by hyatt@apple.com
  • 1 edit in trunk/WebCore/html/HTMLImageLoader.cpp

Fix release build bustage.

13:00 Changeset [42180] by darin@chromium.org
  • 3 edits in trunk/WebCore

2009-04-02 Michael Nordman <michaeln@google.com>

Reviewed by Darin Fisher.

Chromium's ResourceRequest and ResourceResponse structs need some new data members.
https://bugs.webkit.org/show_bug.cgi?id=24897

These fields are needed to facilitate Chromium's implementation of the HTML5
ApplicationCache feature. We need to know what frame (or context) is doing the
requesting, and from what cache the resulting resource was retrieved.

No change in functionality, so no tests.

  • platform/network/chromium/ResourceRequest.h: (WebCore::ResourceRequest::ResourceRequest): (WebCore::ResourceRequest::appCacheContextID): (WebCore::ResourceRequest::setAppCacheContextID):
  • platform/network/chromium/ResourceResponse.h: (WebCore::ResourceResponse::ResourceResponse): (WebCore::ResourceResponse::getAppCacheID): (WebCore::ResourceResponse::setAppCacheID):
12:57 Changeset [42179] by hyatt@apple.com
  • 7 edits
    2 adds in trunk

WebCore:

2009-04-02 David Hyatt <hyatt@apple.com>

Reviewed by Darin Adler

Fix for https://bugs.webkit.org/show_bug.cgi?id=24643.


Make sure images just ignore HTTP error codes and keep loading anyway. Only <object> checks http error codes and falls back. Everyone
else just ignores it and displays the image anyway.

Added http/tests/misc/image-error.html

  • html/HTMLImageLoader.cpp: (WebCore::HTMLImageLoader::notifyFinished):
  • loader/CachedImage.cpp: (WebCore::CachedImage::CachedImage):
  • loader/CachedImage.h: (WebCore::CachedImage::httpStatusCodeError): (WebCore::CachedImage::httpStatusCodeErrorOccurred):
  • loader/CachedResource.h: (WebCore::CachedResource::httpStatusCodeError):
  • loader/loader.cpp: (WebCore::Loader::Host::didReceiveData):

LayoutTests:

2009-04-02 David Hyatt <hyatt@apple.com>

Reviewed by Darin Adler

Add test for https://bugs.webkit.org/show_bug.cgi?id=24643

  • http/tests/misc/image-error-expected.txt: Added.
  • http/tests/misc/image-error.html: Added.
11:42 Changeset [42178] by dimich@chromium.org
  • 2 edits in trunk/WebCore

2009-04-02 Jian Li <jianli@chromium.org>

Reviewed by Dimitri Glazkov.

Add XHR constructor in WorkerContext.idl for V8 bindings.
https://bugs.webkit.org/show_bug.cgi?id=24962

  • workers/WorkerContext.idl:
11:22 Changeset [42177] by pam@chromium.org
  • 3 edits
    3 adds in trunk/LayoutTests

2009-04-02 Pamela Greene <pam@chromium.org>

Reviewed by Darin Adler.

Split long-running regexp out of the main regexp-overflow test file
so VMs (such as V8) that don't limit execution time can still run
the rest of the test cases.
https://bugs.webkit.org/show_bug.cgi?id=18327

  • fast/js/regexp-overflow-expected.txt: updated
  • fast/js/regexp-overflow-too-big-expected.txt: Added.
  • fast/js/regexp-overflow-too-big.html: Added.
  • fast/js/resources/regexp-overflow-too-big.js: Added.
  • fast/js/resources/regexp-overflow.js: removed one test case
11:18 Changeset [42176] by pam@chromium.org
  • 2 edits in trunk/LayoutTests

2009-04-02 Pamela Greene <pam@chromium.org>

Reviewed by Darin Adler.

Minor cleanup: remove unneeded "toLower" on typeof.

  • fast/dom/Window/window-lookup-precedence.html:
10:30 Changeset [42175] by aroben@apple.com
  • 2 edits
    1 add in trunk/WebCore

2009-04-01 Ben Murdoch <benm@google.com>

Fix Bug 19743: Release build fails on 32-bit Windows

<https://bugs.webkit.org/show_bug.cgi?id=19743>

Combines all the HTML element cpp files into one to help reduce the
size of WebCore.lib so we can build in release mode on 32bit Windows.

Reviewed by Adam Roben.

  • WebCore.vcproj/WebCore.vcproj: Excluded all the separate HTML*Element.cpp files from the build. Added HTMLElementsAllInOne.cpp to the build.
  • html/HTMLElementsAllInOne.cpp: Added.
09:47 Changeset [42174] by simon.fraser@apple.com
  • 7 edits in trunk/WebCore

2009-04-02 Simon Fraser <simon.fraser@apple.com>

Reviewed by Darin Adler

https://bugs.webkit.org/show_bug.cgi?id=24999

Optimize hit testing with transforms.

  • rendering/RenderBox.cpp: (WebCore::RenderBox::mapLocalToContainer): (WebCore::RenderBox::mapAbsoluteToLocalPoint): getTransformFromContainer() change to return the matrix by reference.
  • rendering/RenderLayer.cpp: (WebCore::RenderLayer::update3DTransformedDescendantStatus): The method failed to set m_3DTransformedDescendantStatusDirty to false, so did the work every time.

(WebCore::RenderLayer::createLocalTransformState):
Only call the expensive getTransformFromContainer() if there is a transform,
otherwise we just have a translation.

  • rendering/RenderObject.cpp: (WebCore::RenderObject::getTransformFromContainer):
  • rendering/RenderObject.h: Change transformFromContainer() to getTransformFromContainer(), and return the matrix by reference to reduce copying. New enum values for arguments for readability.
  • rendering/TransformState.cpp: (WebCore::TransformState::move): (WebCore::TransformState::applyTransform): (WebCore::HitTestingTransformState::translate): New method that adds a translation to the accumulated matrix, optionally without flattening. Cheaper than applyTransform().

(WebCore::HitTestingTransformState::applyTransform):
(WebCore::HitTestingTransformState::flatten):
(WebCore::HitTestingTransformState::flattenWithTransform):
Add flattenWithTransform(), which is a helper that allows us to
avoid doing the inverse() twice.

  • rendering/TransformState.h: (WebCore::TransformState::): (WebCore::TransformState::move): (WebCore::HitTestingTransformState::): New method on HitTestingTransformState that adds a translation to the accumulated matrix, optionally without flattening. Cheaper than applyTransform(). New enum values for arguments for readability.
09:35 Changeset [42173] by darin@apple.com
  • 2 edits in trunk/JavaScriptCore

2009-04-02 Darin Adler <darin@apple.com>

Reviewed by Kevin Decker.

<rdar://problem/6744471> crash in GC due to uninitialized callFunction pointer

  • runtime/JSGlobalObject.h: (JSC::JSGlobalObject::JSGlobalObjectData::JSGlobalObjectData): Initialize callFunction as we do the other data members that are used in the mark function.
09:25 Changeset [42172] by simon.fraser@apple.com
  • 3 edits in trunk/WebCore

2009-04-02 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dan Bernstein

https://bugs.webkit.org/show_bug.cgi?id=24648

Optimize common code paths in TransformationMatrix.

  • platform/graphics/transforms/TransformationMatrix.cpp: (WebCore::TransformationMatrix::translate): (WebCore::TransformationMatrix::translate3d): Optimize to avoid matrix copy.

(WebCore::TransformationMatrix::isInvertible):
Test for identity and translation matrices before computing
the determinant.

(WebCore::TransformationMatrix::inverse):
Optimize for identity matrix and translations.

  • platform/graphics/transforms/TransformationMatrix.h: (WebCore::TransformationMatrix::isIdentityOrTranslation): Utility method.
09:23 QtWebKitTodo edited by hausmann@webkit.org
(diff)
09:01 Changeset [42171] by kov@webkit.org
  • 2 edits in trunk/WebCore

2009-04-02 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Alexey Proskuryakov.

https://bugs.webkit.org/show_bug.cgi?id=24930
[Gtk] ISO files content is displayed inside the webview instead of being downloaded

Also sniff content of types declared as text/plain. This is
justified by the fact that it is common to have Apache HTTP
servers configured to send text/plain as Content-Type by default.

  • platform/network/soup/ResourceHandleSoup.cpp: (WebCore::gotHeadersCallback):
09:00 Changeset [42170] by kov@webkit.org
  • 2 edits in trunk/WebCore

2009-04-02 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Alexey Proskuryakov.

Protect the handle when notifying the client that the response was
received also in gotChunkCallback, or we crash in didReceiveData
when the load is cancelled in didReceiveResponse.

  • platform/network/soup/ResourceHandleSoup.cpp: (WebCore::gotChunkCallback):
08:50 Changeset [42169] by kov@webkit.org
  • 2 edits in trunk/WebKit/gtk

2009-04-02 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Alexey Proskuryakov.

https://bugs.webkit.org/show_bug.cgi?id=24844
Unit test WebKitDownload

Extra checks to avoid crashing and a critical message when
property accessor functions are called before the download is
started.

  • webkit/webkitdownload.cpp: (_WebKitDownloadPrivate::webkit_download_get_progress): (_WebKitDownloadPrivate::webkit_download_get_elapsed_time):
08:42 Changeset [42168] by kov@webkit.org
  • 2 edits in trunk/WebCore

2009-04-02 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Alexey Proskuryakov.

https://bugs.webkit.org/show_bug.cgi?id=25001
Crash when content type parameters have no value

Handle the case where a Content-Type header parameter has no
value (i.e. no = character), to work-around a bug in libsoup.

  • platform/network/soup/ResourceHandleSoup.cpp: (WebCore::fillResponseFromMessage):
03:29 Changeset [42167] by hausmann@webkit.org
  • 4 edits in trunk

JavaScriptCore:

2009-04-02 Yael Aharon <yael.aharon@nokia.com>

Reviewed by Simon Hausmann

https://bugs.webkit.org/show_bug.cgi?id=24490

Implement WTF::ThreadSpecific in the Qt build using
QThreadStorage.

WebCore:

2009-04-02 Yael Aharon <yael.aharon@nokia.com>

Reviewed by Simon Hausmann.

https://bugs.webkit.org/show_bug.cgi?id=24490

Enable web workers in Qt.

00:54 Changeset [42166] by hausmann@webkit.org
  • 2 edits in trunk/WebKit/qt

Fix the Qt build

00:54 Changeset [42165] by hausmann@webkit.org
  • 2 edits in trunk/WebKit/qt

2009-04-02 Takumi Asaki <takumi.asaki@nokia.com>

Reviewed by Simon Hausmann.

Fix pre-edit handling of text fields with input methods.

The input method sends an empty preeditString() if all characters of
the preedit should be deleted. So inputMethodEvent() has to use
preeditString() if it's empty.

00:12 Changeset [42164] by ap@webkit.org
  • 8 edits
    17 moves
    6 adds in trunk

Reviewed by Darin Adler. Based on work of Julien Chaffraix.

https://bugs.webkit.org/show_bug.cgi?id=22475
REGRESSION: Async XMLHttpRequest never finishes on nonexistent files anymore

https://bugs.webkit.org/show_bug.cgi?id=24886
XHR requests opened when offline never return

WebCore:

  • xml/XMLHttpRequest.cpp: (WebCore::XMLHttpRequest::callReadyStateChangeListener): Only dispatch load event if this wasn't an error. (WebCore::XMLHttpRequest::abort): Fixed formatting. (WebCore::XMLHttpRequest::genericError): Change state to DONE and dispatch readystatechange. The comment saying that this doesn't match Firefox was added in r33559 without explanation or tests, and I don't think that it's accurate. Also, Firefox and Safari 3 both change state to HEADERS_RECEIVED before DONE on error - this doesn't match the spec, and I doubt that any code depends on this, so I went with the spec here.

LayoutTests:

  • fast/loader/xmlhttprequest-missing-file-exception-expected.txt: Removed.
  • fast/loader/xmlhttprequest-missing-file-exception.html: Removed.
  • http/tests/xmlhttprequest/state-after-network-error-expected.txt: Added.
  • http/tests/xmlhttprequest/state-after-network-error.html: Added. Added tests for these issues.
  • http/tests/xmlhttprequest/frame-load-cancelled-abort-expected.txt: As with other errors, the request now reaches DONE state now. This makes the results closer to Firefox again (they changed multiple times recently).
  • http/tests/xmlhttprequest/resources/frame-load-to-cancel.html: Firefox dispatches cancellation events at a later point, when window.parent is already undefined. Changed the test to log all events, making the difference with Firefox less confusing.
  • http/tests/xmlhttprequest/simple-cross-origin-progress-events.html: Fixed a mistake in the test, which could call a non-existent function in some code paths.
  • http/tests/xmlhttprequest/xmlhttprequest-unsafe-redirect-expected.txt: State now reaches DONE on cross-origin check failure, as it should per the spec. This matches Firefox in async case (and in sync case, it simply doesn't dispatch any onreadystatechange events, not sure why).
  • fast/dom/null-document-xmlhttprequest-open-expected.txt: Removed.
  • fast/dom/null-document-xmlhttprequest-open.html: Removed.
  • fast/dom/resources/1251.html: Removed.
  • fast/dom/resources/xmlhttprequest-get-data.xml: Removed.
  • fast/dom/xmlhttprequest-gc-expected.txt: Removed.
  • fast/dom/xmlhttprequest-gc.html: Removed.
  • fast/dom/xmlhttprequest-get-expected.txt: Removed.
  • fast/dom/xmlhttprequest-get.xhtml: Removed.
  • fast/dom/xmlhttprequest-html-response-encoding-expected.txt: Removed.
  • fast/dom/xmlhttprequest-html-response-encoding.html: Removed.
  • fast/dom/xmlhttprequest-invalid-values-expected.txt: Removed.
  • fast/dom/xmlhttprequest-invalid-values.html: Removed.
  • fast/loader/resources/plist.app: Removed.
  • fast/loader/xmlhttprequest-bad-mimetype-expected.txt: Removed.
  • fast/loader/xmlhttprequest-bad-mimetype.html: Removed.
  • fast/xmlhttprequest: Added.
  • fast/xmlhttprequest/null-document-xmlhttprequest-open-expected.txt: Copied from LayoutTests/fast/dom/null-document-xmlhttprequest-open-expected.txt.
  • fast/xmlhttprequest/null-document-xmlhttprequest-open.html: Copied from LayoutTests/fast/dom/null-document-xmlhttprequest-open.html.
  • fast/xmlhttprequest/resources: Added.
  • fast/xmlhttprequest/resources/1251.html: Copied from LayoutTests/fast/dom/resources/1251.html.
  • fast/xmlhttprequest/resources/plist.app: Copied from LayoutTests/fast/loader/resources/plist.app.
  • fast/xmlhttprequest/resources/xmlhttprequest-get-data.xml: Copied from LayoutTests/fast/dom/resources/xmlhttprequest-get-data.xml.
  • fast/xmlhttprequest/xmlhttprequest-bad-mimetype-expected.txt: Copied from LayoutTests/fast/loader/xmlhttprequest-bad-mimetype-expected.txt.
  • fast/xmlhttprequest/xmlhttprequest-bad-mimetype.html: Copied from LayoutTests/fast/loader/xmlhttprequest-bad-mimetype.html.
  • fast/xmlhttprequest/xmlhttprequest-gc-expected.txt: Copied from LayoutTests/fast/dom/xmlhttprequest-gc-expected.txt.
  • fast/xmlhttprequest/xmlhttprequest-gc.html: Copied from LayoutTests/fast/dom/xmlhttprequest-gc.html.
  • fast/xmlhttprequest/xmlhttprequest-get-expected.txt: Copied from LayoutTests/fast/dom/xmlhttprequest-get-expected.txt.
  • fast/xmlhttprequest/xmlhttprequest-get.xhtml: Copied from LayoutTests/fast/dom/xmlhttprequest-get.xhtml.
  • fast/xmlhttprequest/xmlhttprequest-html-response-encoding-expected.txt: Copied from LayoutTests/fast/dom/xmlhttprequest-html-response-encoding-expected.txt.
  • fast/xmlhttprequest/xmlhttprequest-html-response-encoding.html: Copied from LayoutTests/fast/dom/xmlhttprequest-html-response-encoding.html.
  • fast/xmlhttprequest/xmlhttprequest-invalid-values-expected.txt: Copied from LayoutTests/fast/dom/xmlhttprequest-invalid-values-expected.txt.
  • fast/xmlhttprequest/xmlhttprequest-invalid-values.html: Copied from LayoutTests/fast/dom/xmlhttprequest-invalid-values.html.
  • fast/xmlhttprequest/xmlhttprequest-missing-file-exception-expected.txt: Copied from LayoutTests/fast/loader/xmlhttprequest-missing-file-exception-expected.txt.
  • fast/xmlhttprequest/xmlhttprequest-missing-file-exception.html: Copied from LayoutTests/fast/loader/xmlhttprequest-missing-file-exception.html.
  • fast/xmlhttprequest/xmlhttprequest-nonexistent-file-expected.txt: Added.
  • fast/xmlhttprequest/xmlhttprequest-nonexistent-file.html: Added.
  • platform/gtk/Skipped: Moved non-HTTP XMLHttpRequest tests to fast/xmlhttprequest.

04/01/09:

23:49 QtWebKitTodo edited by hausmann@webkit.org
(diff)
21:51 Changeset [42163] by sfalken@apple.com
  • 2 edits in trunk/WebCore

https://bugs.webkit.org/show_bug.cgi?id=24997
Race conditions in icon database threading code


m_syncThreadRunning set was mistakenly eliminated in r27717.
This reintroduces it.

Reviewed by Oliver Hunt.

  • loader/icon/IconDatabase.cpp: (WebCore::IconDatabase::open): Set m_syncThreadRunning.
19:58 Changeset [42162] by mrowe@apple.com
  • 2 edits in trunk/WebKitTools

Windows support for built-product-archive.

  • BuildSlaveSupport/built-product-archive:
18:20 Changeset [42161] by antti@apple.com
  • 6 edits in trunk/WebCore

2009-04-01 Antti Koivisto <antti@apple.com>

Reviewed by Darin Adler.

<rdar://problem/6748609> Use ActiveDOMObject to suspend marquees


Make marquees ActiveDOMObjects, get rid of the special suspension code.

  • html/HTMLMarqueeElement.cpp: (WebCore::HTMLMarqueeElement::HTMLMarqueeElement): (WebCore::HTMLMarqueeElement::canSuspend): (WebCore::HTMLMarqueeElement::suspend): (WebCore::HTMLMarqueeElement::resume):
  • html/HTMLMarqueeElement.h:
  • page/Frame.cpp: (WebCore::Frame::clearTimers):
  • rendering/RenderLayer.cpp:
  • rendering/RenderLayer.h: (WebCore::RenderLayer::marquee):
18:10 Changeset [42160] by dino@apple.com
  • 2 edits in trunk/WebCore

2009-04-01 Dean Jackson <dino@apple.com>

Reviewed by Darin Adler

Make constant values static in previous commit, as suggested
by Darin.

  • rendering/RenderVideo.cpp:
17:57 Changeset [42159] by dino@apple.com
  • 2 edits in trunk/WebCore

2009-04-01 Dean Jackson <dino@apple.com>

Reviewed by Simon Fraser

https://bugs.webkit.org/show_bug.cgi?id=24993

Standalone media should have a smaller instrinsic
height.

  • rendering/RenderVideo.cpp: (WebCore::RenderVideo::RenderVideo):
17:30 Changeset [42158] by darin@apple.com
  • 23 edits
    1 copy
    1 move
    6 adds in trunk

WebCore:

2009-04-01 Darin Adler <darin@apple.com>

Reviewed by Geoff Garen.

Bug 22378: Crash submitting a form when parsing an XHTML document
https://bugs.webkit.org/show_bug.cgi?id=22378
rdar://problem/6388377

Tests: fast/loader/submit-form-while-parsing-1.xhtml

fast/loader/submit-form-while-parsing-2.html

  • WebCore.base.exp: Updated.
  • history/HistoryItem.cpp: (WebCore::HistoryItem::HistoryItem): Renamed m_subItems to m_children. Used uncheckedAppend because we reserveInitialCapacity. (WebCore::HistoryItem::addChildItem): Renamed m_subItems to m_children. Added an assertion that this is only used to add items that don't have duplicate frame names. (WebCore::HistoryItem::setChildItem): Added. Replaces an existing item if any, preserving the isTargetItem flag. (WebCore::HistoryItem::childItemWithTarget): Renamed from childItemWithName for consistency with the other functions here that all call the frame name the "target". Also updated for rename of m_subItems to m_children. (WebCore::HistoryItem::findTargetItem): Renamed from recurseToFindTargetItem. Removed unneeded size check. (WebCore::HistoryItem::targetItem): Changed to always return the top item if no item has the isTargetItem flag set. The old version would instead return 0 in some cases, but return the top item if it had no children. (WebCore::HistoryItem::children): Renamed m_subItems to m_children. (WebCore::HistoryItem::hasChildren): Ditto. (WebCore::HistoryItem::showTreeWithIndent): Ditto.
  • history/HistoryItem.h: Name changes.
  • html/HTMLFormElement.cpp: (WebCore::HTMLFormElement::submit): Create and pass a FormState instead of attaching "recorded form values" and "form about to be submitted" to the frame loader. Parameter work fine for this; there's no need to store state on the FrameLoader.
  • loader/FormState.cpp: (WebCore::FormState::FormState): Adopt a vector instead of copying a hash map. (WebCore::FormState::create): Ditto.
  • loader/FormState.h: Update to use a vector that we adopt instead of hash map that we copy for auto-fill text field values.
  • loader/FrameLoader.cpp: (WebCore::ScheduledRedirection::ScheduledRedirection): Added a new case for form submissions and turned the "wasDuringLoad" state into a separate boolean instead of using a special redirection type to track it. (WebCore::FrameLoader::createWindow): Updated for name and argument change of loadFrameRequest. (WebCore::FrameLoader::urlSelected): Ditto. (WebCore::FrameLoader::submitForm): Removed the "deferred form submission" machinery, replacing it with the "scheduled redirection" mechanism, shared with other kinds of redirection and navigation. Moved frame targeting here so we can schedule the redirection on the right frame. Moved the multiple form submission protection here. Moved the code to implement the rule that m_navigationDuringLoad was used for here too. (WebCore::FrameLoader::didOpenURL): Use the new wasDuringLoad flag instead of the locationChangeDuringLoad type to detect location change during a load. (WebCore::FrameLoader::executeScript): Removed call to now-obsolete function, submitFormAgain. (WebCore::FrameLoader::scheduleLocationChange): Moved the code to stop loading out of this function into scheduleRedirection, so it can be shared with the new scheduleFormSubmission function. (WebCore::FrameLoader::scheduleFormSubmission): Added. Almost the same as scheduleLocationChange, but with the arguments for a form submission. (WebCore::FrameLoader::scheduleRefresh): Updated for the change to the duringLoad flag. (WebCore::FrameLoader::isLocationChange): Added case for formSubmission and removed case for locationChangeDuringLoad. (WebCore::FrameLoader::redirectionTimerFired): Ditto. Also removed unneeded completeURL call and just use KURL constructor to match the other cases. (WebCore::FrameLoader::provisionalLoadStarted): Removed the code to set up the m_navigationDuringLoad, which is no longer needed. The new version of this is in the submitForm function and sets the lockHistory boolean. (WebCore::FrameLoader::scheduleRedirection): Moved the code to stop a load in here that used to be in scheduleLocationChange. (WebCore::FrameLoader::startRedirectionTimer): Added case for formSubmission and removed case for locationChangeDuringLoad. (WebCore::FrameLoader::stopRedirectionTimer): Ditto. (WebCore::FrameLoader::completed): Removed call to now-obsolete function, submitFormAgain. (WebCore::FrameLoader::loadFrameRequest): Renamed from loadFrameRequestWithFormAndValues. Replaced form element and form values argument with a single FormState argument. Changed frame targeting code to use the source frame in the case of a form submission to better match the actual target frame. (WebCore::FrameLoader::loadURL): Don't search for existing frames in the form submission case since we already did that in the submitForm function. (WebCore::FrameLoader::clientRedirected): Changed to work with the m_isExecutingJavaScriptFormAction data member directly instead of taking it as a function parameter. (WebCore::FrameLoader::loadPostRequest): Don't search for existing frames in the form submission case since we already did that in the submitForm function. (WebCore::FrameLoader::addBackForwardItemClippedAtTarget): Moved comment in here that was misplaced elsewhere in the file. (WebCore::FrameLoader::findFrameForNavigation): Changed to use the early return idiom. (WebCore::FrameLoader::recursiveGoToItem): Updated for HistoryItem changes. (WebCore::FrameLoader::childFramesMatchItem): Ditto. (WebCore::FrameLoader::updateHistoryForStandardLoad): Removed the m_navigationDuringLoad logic; that's now handled by setting lockHistory to true in the submitForm function. (WebCore::FrameLoader::updateHistoryForRedirectWithLockedBackForwardList): Use the new setChildItem function so we don't get multiple items for the same frame name in the history item tree in the back/forward list.
  • loader/FrameLoader.h: Renamed loadFrameRequestWithFormAndValues to loadFrameRequest and made it take a form state object instead of the form element and form values. Removed the unused functions loadEmptyDocumentSynchronously, provisionalDocumentLoader, notifyIconChnaged, and updateBaseURLForEmptyDocument. Changed the submitForm function to take a form state argument. Eliminated the clearRecordedFormValues, setFormAboutToBeSubmitted, and recordFormValue functions, which are replaced by the form state arguments to submitForm and loadFrameRequest. Removed the isJavaScriptFormAction argument from the clientRedirected function; instead it looks at a data member directly. Eliminated the submitFormAgain and overload of the submitForm function; these are now subsumed into the remaining submitForm function and the scheduleFormSubmission function. Removed unused and obsolete data members m_navigationDuringLoad, m_deferredFormSubmission, m_formAboutToBeSubmitted and m_formValuesAboutToBeSubmitted.
  • page/ContextMenuController.cpp: (WebCore::ContextMenuController::contextMenuItemSelected): Updated for name and argument change of loadFrameRequest.
  • page/Frame.cpp: (WebCore::Frame::~Frame): Removed call to the now-unneeded clearRecordedFormValues function.

WebKit/mac:

2009-04-01 Darin Adler <darin@apple.com>

Reviewed by Geoff Garen.

Bug 22378: Crash submitting a form when parsing an XHTML document
https://bugs.webkit.org/show_bug.cgi?id=22378
rdar://problem/6388377

  • History/WebHistoryItem.mm: (-[WebHistoryItem targetItem]): Call targetItem directly instead of callling isTargetItem, hasChildren, and recurseToFindTargetItem.
  • WebCoreSupport/WebFrameLoaderClient.mm: (WebFrameLoaderClient::dispatchWillSubmitForm): Updated for the new textFieldValues function in FormState.
  • WebView/WebPDFView.mm: (-[WebPDFView PDFViewWillClickOnLink:withURL:]): Updated for name and argument change of loadFrameRequest.

WebKit/win:

2009-04-01 Darin Adler <darin@apple.com>

Reviewed by Geoff Garen.

Bug 22378: Crash submitting a form when parsing an XHTML document
https://bugs.webkit.org/show_bug.cgi?id=22378
rdar://problem/6388377

  • WebFrame.cpp: (WebFrame::dispatchWillSubmitForm): Updated for the new textFieldValues function in FormState.

LayoutTests:

2009-04-01 Darin Adler <darin@apple.com>

Reviewed by Geoff Garen.

Bug 22378: Crash submitting a form when parsing an XHTML document
https://bugs.webkit.org/show_bug.cgi?id=22378
rdar://problem/6388377

  • fast/loader/onunload-form-submit-crash-2-expected.txt: Updated for change to the test.
  • fast/loader/onunload-form-submit-crash-2.html: Updated test to expect the form submission during the onunload handler to be ineffective.
  • fast/loader/onunload-form-submit-crash-expected.txt: Updated for change to the test.
  • fast/loader/onunload-form-submit-crash.html: Updated test to expect the form submission during the onunload handler to be ineffective.
  • fast/loader/resources/onunload-form-submit-crash2.html: Removed.
  • fast/loader/resources/onunload-form-submit-failure.html: Copied from LayoutTests/fast/loader/resources/onunload-form-submit-crash2.html.
  • fast/loader/resources/onunload-form-submit-success.html: Copied from LayoutTests/fast/loader/resources/onunload-form-submit-crash2.html.
  • fast/loader/resources/submit-form-while-parsing-subframe.html: Added.
  • fast/loader/resources/submit-form-while-parsing-success.html: Added.
  • fast/loader/submit-form-while-parsing-1-expected.txt: Added.
  • fast/loader/submit-form-while-parsing-1.xhtml: Added.
  • fast/loader/submit-form-while-parsing-2-expected.txt: Added.
  • fast/loader/submit-form-while-parsing-2.html: Added.
  • platform/mac/http/tests/navigation/success200-frames-loadsame-expected.txt: Updated to expect proper results, without the duplicate history items for the same frame name that were present before.
16:49 Changeset [42157] by dglazkov@chromium.org
  • 2 edits in trunk/WebCore

2009-04-01 Dimitri Glazkov <dglazkov@chromium.org>

Unreviewed, fixing previous commit.

https://bugs.webkit.org/show_bug.cgi?id=24994
Revert V8DOMMap.cpp change which accidentally snuck into the previous
commit.

  • bindings/v8/V8DOMMap.cpp: Revert change in previous commit.
16:43 Changeset [42156] by dglazkov@chromium.org
  • 5 edits in trunk/WebCore

2009-04-01 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Darin Fisher.

https://bugs.webkit.org/show_bug.cgi?id=24994
Fix miscellaneous merge/upstreaming divergencies in Chromium port.

  • bindings/v8/V8Collection.h: Added an include.
  • platform/KURLGoogle.cpp: (WebCore::KURL::baseAsString): Added to match KURL.cpp.
  • platform/chromium/ClipboardChromium.cpp: Added an include.
15:47 Changeset [42155] by dglazkov@chromium.org
  • 11 edits in trunk/WebCore

2009-04-01 Dimitri Glazkov <dglazkov@chromium.org>

Reviewed by Timothy Hatcher.

https://bugs.webkit.org/show_bug.cgi?id=24989
Refactor InspectorResource to use ScriptObject/FunctionCall.

  • bindings/js/JSInspectorControllerCustom.cpp: (WebCore::JSInspectorController::addResourceSourceToFrame): Tweaked to use

a more appropriate type conversion.

(WebCore::JSInspectorController::getResourceDocumentNode): Tweaked to

accommodate for InspectorResource refactoring.

  • bindings/js/ScriptFunctionCall.cpp: (WebCore::ScriptFunctionCall::appendArgument): Added long long type to match

existing jsNumber call signature.

  • bindings/js/ScriptFunctionCall.h: Declaration for above.
  • bindings/js/ScriptObject.cpp: (WebCore::createEmptyObject): Added. (WebCore::ScriptObject::ScriptObject): (WebCore::ScriptObject::set): Added a bunch of property setters.
  • bindings/js/ScriptObject.h: Declarations for above.
  • inspector/InspectorController.cpp: Moved all InspectorResource-managing code

to InspectorResource.cpp

(WebCore::InspectorController::populateScriptObjects): Changed to use

refactored InspectorResource.

(WebCore::InspectorController::resetScriptObjects): Ditto.
(WebCore::InspectorController::pruneResources): Ditto.
(WebCore::InspectorController::didCommitLoad): Ditto.
(WebCore::InspectorController::addResource): Ditto.
(WebCore::InspectorController::removeResource): Ditto.
(WebCore::InspectorController::didLoadResourceFromMemoryCache): Ditto.
(WebCore::InspectorController::identifierForInitialRequest): Ditto.
(WebCore::InspectorController::willSendRequest): Ditto.
(WebCore::InspectorController::didReceiveResponse): Ditto.
(WebCore::InspectorController::didReceiveContentLength): Ditto.
(WebCore::InspectorController::didFinishLoading): Ditto.
(WebCore::InspectorController::didFailLoading): Ditto,
(WebCore::InspectorController::resourceRetrievedByXMLHttpRequest): Ditto.
(WebCore::InspectorController::scriptImported): Ditto.

  • inspector/InspectorController.h: Removed InspectorResource-managing code decls

and tweaked some signatures to use ScriptString and long long for identifier.

  • inspector/InspectorController.idl: Changed to use long long for identifier.
  • inspector/InspectorResource.cpp: Refactored to use ScriptObject/FunctionCall. (WebCore::InspectorResource::InspectorResource): Ditto. (WebCore::InspectorResource::~InspectorResource): Ditto. (WebCore::InspectorResource::createCached): Added. (WebCore::InspectorResource::updateRequest): Added. (WebCore::InspectorResource::updateResponse): Added. (WebCore::createHeadersObject): Added. (WebCore::InspectorResource::createScriptObject): Added. (WebCore::InspectorResource::updateScriptObject): Added. (WebCore::InspectorResource::releaseScriptObject): Added. (WebCore::InspectorResource::type): Tweaked to use ScriptString. (WebCore::InspectorResource::setXMLHttpResponseText): Added. (WebCore::InspectorResource::sourceString): Tweaked to use ScriptString. (WebCore::InspectorResource::startTiming): Added. (WebCore::InspectorResource::markResponseReceivedTime): Added. (WebCore::InspectorResource::endTiming): Added. (WebCore::InspectorResource::markFailed): Added. (WebCore::InspectorResource::addLength): Added.
  • inspector/InspectorResource.h: Added decls for newly refactored-in methods. (WebCore::InspectorResource::create): Added. (WebCore::InspectorResource::isSameLoader): Added. (WebCore::InspectorResource::markMainResource): Added. (WebCore::InspectorResource::identifier): Added. (WebCore::InspectorResource::requestURL): Added. (WebCore::InspectorResource::frame): Added. (WebCore::InspectorResource::mimeType): Added. (WebCore::InspectorResource::Changes::Changes): Added new class to track

resource changes.

(WebCore::InspectorResource::Changes::hasChange): Added.
(WebCore::InspectorResource::Changes::set): Added.
(WebCore::InspectorResource::Changes::clear): Added.
(WebCore::InspectorResource::Changes::setAll): Added.
(WebCore::InspectorResource::Changes::clearAll): Added.

14:17 Changeset [42154] by darin@chromium.org
  • 3 edits
    4 adds in trunk

2009-04-01 Tony Chang <tony@chromium.org>

Reviewed by Darin Fisher.

Ensure the right click doesn't deselect text under it. This was
happening in the Chromium and GTK+ ports because they don't select
the text under the cursor on right click.


This was regressed in r41715, https://bugs.webkit.org/show_bug.cgi?id=19737


https://bugs.webkit.org/show_bug.cgi?id=24946

Test: fast/events/context-no-deselect.html

  • page/EventHandler.cpp: (WebCore::EventHandler::handleMouseReleaseEvent):
12:48 Changeset [42153] by christian@webkit.org
  • 3 edits
    1 add in trunk

2009-04-01 Christian Dywan <christian@twotoasts.de>

Reviewed by Holger Freyther.

Unit test WebKitDownload
http://bugs.webkit.org/show_bug.cgi?id=24844

  • GNUmakefile.am: Add a unit test for downloading.

2009-04-01 Christian Dywan <christian@twotoasts.de>

Reviewed by Holger Freyther.

  • tests/testdownload.c: Added. (test_webkit_download_create): (main): Add an initial unit test for downloading.
12:32 Changeset [42152] by bolsinga@apple.com
  • 9 edits in trunk

2009-04-01 Greg Bolsinga <bolsinga@apple.com>

Reviewed by Mark Rowe.

https://bugs.webkit.org/show_bug.cgi?id=24990
Put SECTORDER_FLAGS into xcconfig files.

12:26 Changeset [42151] by simon.fraser@apple.com
  • 2 edits in trunk/WebCore

2009-04-01 Simon Fraser <simon.fraser@apple.com>

Reviewed by Dave Hyatt

https://bugs.webkit.org/show_bug.cgi?id=24991

Need to use toRenderBoxModelObject(), not toRenderBox(), to test for layer()
when getting the transform.

  • rendering/RenderObject.cpp: (WebCore::RenderObject::transformFromContainer):
09:55 Changeset [42150] by kov@webkit.org
  • 3 edits in trunk/WebKit/gtk

2009-04-01 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Unreviewed typo fixes, and naming normalization in the
documentation.

  • docs/webkitenvironment.xml:
  • webkit/webkitwebview.cpp:
09:37 Changeset [42149] by kov@webkit.org
  • 2 edits in trunk

2009-04-01 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Unreviewed build fix. Fix make distcheck, after the gtk-doc
integration.

  • GNUmakefile.am:
08:58 QtWebKitTodo edited by hausmann@webkit.org
(diff)
04:41 QtWebKitTodo edited by hausmann@webkit.org
(diff)

03/31/09:

23:43 Changeset [42148] by darin@apple.com
  • 3 edits in trunk/WebCore

2009-03-31 Darin Adler <darin@apple.com>

Rolled out that last change until I can investigate the regression test failures it seems
to have caused.

  • dom/PositionIterator.cpp: (WebCore::PositionIterator::decrement):
  • dom/PositionIterator.h: (WebCore::PositionIterator::PositionIterator):
21:41 Changeset [42147] by darin@apple.com
  • 3 edits in trunk/WebCore

2009-03-31 Darin Adler <darin@apple.com>

Reviewed by Eric Seidel.

Bug 24621: PositionIterator doesn't iterate "after last child" positions when going backwards
https://bugs.webkit.org/show_bug.cgi?id=24621

  • dom/PositionIterator.cpp: (WebCore::PositionIterator::decrement): Make sure that when the parent has no children, we don't ever use Position::uncheckedPreviousOffset. This is consistent with the forward iterator, but also should never arise because of the fix below.
  • dom/PositionIterator.h: (WebCore::PositionIterator::PositionIterator): Fixed so m_offset will always be 0 when the passed-in node has no children. Like the change above, this is consistent with the rest of the class, although in the long run I think it's a bit strange to treat a <p> element with no children differently than a <p> element with children.
17:13 Changeset [42146] by aroben@apple.com
  • 2 edits in trunk/WebKitTools

Make resolve-ChangeLogs -f work when the working tree has spaces in its path

Reviewed by Mark Rowe and David Kilzer.

  • Scripts/resolve-ChangeLogs: (sub fixMergedChangeLogs): Quote the path to resolve-ChangeLogs in case it contains spaces.
15:57 Changeset [42145] by andersca@apple.com
  • 5 edits in trunk/WebKit/mac

2009-03-31 Anders Carlsson <andersca@apple.com>

Reviewed by Adam Roben.

WebKit side of <rdar://problem/6500266>.


  • Plugins/Hosted/NetscapePluginInstanceProxy.h:
  • Plugins/Hosted/NetscapePluginInstanceProxy.mm: (WebKit::NetscapePluginInstanceProxy::wheelEvent): Send the event.


  • Plugins/Hosted/WebHostedNetscapePluginView.mm: (-[WebHostedNetscapePluginView scrollWheel:]): Call NetscapePluginInstanceProxy::wheelEvent. If the plug-in processed the event, don't call super.


  • Plugins/Hosted/WebKitPluginHost.defs: Add definition.
15:13 Changeset [42144] by eric.carlson@apple.com
  • 2 edits in trunk/WebCore

2009-03-31 Eric Carlson <eric.carlson@apple.com>

Reviewed by Simon Fraser.

https://bugs.webkit.org/show_bug.cgi?id=24965
Bug 24965: HTMLMediaElement: network state changes can be missed


  • html/HTMLMediaElement.cpp: (WebCore::HTMLMediaElement::setNetworkState): Always update m_networkState when the state changes, even when no event needs to be fired.
14:55 Changeset [42143] by eric.carlson@apple.com
  • 5 edits in trunk

2009-03-31 Eric Carlson <eric.carlson@apple.com>

Reviewed by Darin Adler.


https://bugs.webkit.org/show_bug.cgi?id=24964
Bug 24964: HTMLMediaElement: 'waiting' event may fire twice

  • html/HTMLMediaElement.cpp: (WebCore::HTMLMediaElement::setReadyState): Only fire 'waiting' event once when ready state drops below HAVE_FUTURE_DATA.

2009-03-31 Eric Carlson <eric.carlson@apple.com>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24964
Bug 24964: HTMLMediaElement: 'waiting' event may fire twice

  • http/tests/media/video-play-stall.html: Fail if 'waiting' fires more than once.
  • media/video-test.js: Fix a typo. (failTest):
14:39 Changeset [42142] by ap@webkit.org
  • 2 edits in trunk/WebKitTools

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24777
WebKit tools are broken on Tiger

  • Scripts/webkitdirs.pm: Don't use arch command with options on Tiger.
14:26 Changeset [42141] by eric.carlson@apple.com
  • 9 edits
    2 adds in trunk

2009-03-31 Eric Carlson <eric.carlson@apple.com>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24846
Bug 24846: HTMLMediaElement should implement 'autobuffer' attribute

Add 'autobuffer' attrubute to media element. 'autobuffer' is a hint that
the author recommends downloading the entire resource optimistically, so the
attribute is made available to the media engine to use if it is able.

Test: media/video-dom-autobuffer.html

  • html/HTMLAttributeNames.in: Add autobuffer.
  • html/HTMLMediaElement.cpp: (WebCore::HTMLMediaElement::parseMappedAttribute): New. (WebCore::HTMLMediaElement::autobuffer): Ditto. (WebCore::HTMLMediaElement::setAutobuffer): Ditto.
  • html/HTMLMediaElement.h: Declare autobuffer and setAutobuffer.
  • html/HTMLMediaElement.idl: Add autobuffer.
  • platform/graphics/MediaPlayer.cpp: (WebCore::MediaPlayer::MediaPlayer): Initialize m_autobuffer. (WebCore::MediaPlayer::autobuffer): New. (WebCore::MediaPlayer::setAutobuffer): Ditto.
  • platform/graphics/MediaPlayer.h: Declare m_autobuffer, autobuffer and setAutobuffer
  • platform/graphics/MediaPlayerPrivate.h: (WebCore::MediaPlayerPrivateInterface::setAutobuffer): Declare setAutobuffer.

2009-03-31 Eric Carlson <eric.carlson@apple.com>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24846
Bug 24846: HTMLMediaElement should implement 'autobuffer' attribute

  • media/video-dom-autobuffer-expected.txt: Added.
  • media/video-dom-autobuffer.html: Added.
13:24 Changeset [42140] by darin@apple.com
  • 2 edits in trunk/WebKit/mac

2009-03-31 Darin Adler <darin@apple.com>

Reviewed by Adele Peterson.

<rdar://problem/6740581> REGRESSION (r41793): Page Down and Page Up don’t work in Leopard Mail

  • WebView/WebHTMLView.mm: (-[WebResponderChainSink tryToPerform:with:]): Added. Without this we would think we had handled an event when we actually hadn't. Specifically, when -[WebFrameView scrollPageDown:] calls tryToPerform on the next responder.
13:23 Changeset [42139] by dimich@chromium.org
  • 2 edits in trunk/WebCore

2009-03-31 Craig Schlenter <craig.schlenter@gmail.com>

Reviewed by Dimitri Glazkov.

https://bugs.webkit.org/show_bug.cgi?id=24550
Check return value of vasprintf. This should keep gcc 4.3.3
happy rather than have it complain about attribute warn_unused_result.

  • dom/XMLTokenizerLibxml2.cpp: (WebCore::XMLTokenizer::error):
13:11 Changeset [42138] by dino@apple.com
  • 2 edits in trunk/WebCore

2009-03-31 Dean Jackson <dino@apple.com>

Style guide violation!

  • loader/MediaDocument.cpp: (WebCore::MediaDocument::defaultEventHandler):
13:05 Changeset [42137] by dino@apple.com
  • 2 edits in trunk/WebCore

2009-03-31 Dean Jackson <dino@apple.com>

Fix Tiger build.

  • loader/MediaDocument.cpp: (WebCore::MediaDocument::defaultEventHandler):
12:56 Changeset [42136] by dimich@chromium.org
  • 2 adds in trunk/WebCore/bindings/v8

Fixing previous commit - forgot to 'git add' new files.

12:55 Changeset [42135] by dimich@chromium.org
  • 1 edit in trunk/WebCore/ChangeLog

2009-03-31 Jian Li <jianli@chromium.org>

Reviewed by Darin Fisher.

Upstream V8DOMMap for v8 bindings.
https://bugs.webkit.org/show_bug.cgi?id=24951

  • bindings/v8/V8DOMMap.cpp: Added.
  • bindings/v8/V8DOMMap.h: Added.
12:50 Changeset [42134] by dino@apple.com
  • 2 edits in trunk/WebCore

2009-03-31 Dean Jackson <dino@apple.com>

Reviewed by Simon Fraser.

https://bugs.webkit.org/show_bug.cgi?id=24955

Spacebar didn't play/pause in standalone MediaDocument

  • loader/MediaDocument.cpp: (WebCore::MediaDocument::defaultEventHandler):
12:10 Changeset [42133] by dimich@chromium.org
  • 2 edits in trunk/WebCore

2009-03-31 Alpha Lam <hclam@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=24901
Added the following methods to MediaPlayerPrivate for Chromium port.
void sizeChanged();
void rateChanged();
void durationChanged();

The above changes are to reflect changes in MediaPlayer.cpp.
More details can be found in the original changeset.
http://trac.webkit.org/changeset/41907

  • platform/graphics/chromium/MediaPlayerPrivateChromium.h:
09:54 Changeset [42132] by darin@apple.com
  • 1 edit
    1 add in trunk/WebKitTools

2009-03-31 Darin Adler <darin@apple.com>

  • Scripts/add-include: Added.
09:46 Changeset [42131] by darin@chromium.org
  • 4 edits in trunk/WebCore

2009-03-31 Rafael Weinstein <rafaelw@chromium.org>

Reviewed by Eric Seidel.

https://bugs.webkit.org/show_bug.cgi?id=24947

Special-case drawing text-shadow on win32, to let GDI draw in cases
when both fill & shadow color are opaque and the shadow has no blur.

  • platform/graphics/chromium/FontChromiumWin.cpp: (WebCore::TransparencyAwareFontPainter::TransparencyAwareGlyphPainter::drawGlyphs): (WebCore::Font::drawComplexText):
  • platform/graphics/skia/SkiaFontWin.cpp: (WebCore::windowsCanHandleDrawTextShadow): (WebCore::windowsCanHandleTextDrawing):
  • platform/graphics/skia/SkiaFontWin.h:
09:21 Changeset [42130] by darin@chromium.org
  • 4 edits in trunk/WebCore

2009-03-31 Jeremy Moskovich <jeremy@chromium.org>

Reviewed by Darin Fisher.

r42055 removed some functions in FrameLoader used by Chrome,
this CL backs out some of those changes to unbreak the Chrome build.

https://bugs.webkit.org/show_bug.cgi?id=24949

  • WebCore.base.exp:
  • loader/FrameLoader.cpp:
  • loader/FrameLoader.h:
09:13 Changeset [42129] by darin@chromium.org
  • 3 edits in trunk/WebCore

2009-03-31 Jeremy Moskovich <jeremy@chromium.org>

Reviewed by Darin Adler.

Build fixes for Chromium.

https://bugs.webkit.org/show_bug.cgi?id=24945

  • dom/Document.h:
  • platform/chromium/ClipboardChromium.cpp: (WebCore::imageToMarkup):
05:42 QtWebKitTodo edited by hausmann@webkit.org
(diff)
03:26 Changeset [42128] by hausmann@webkit.org
  • 2 edits in trunk/WebCore

2009-03-31 Simon Hausmann <simon.hausmann@nokia.com>

Fix the Qt build.

01:11 QtWebKitTodo edited by hausmann@webkit.org
(diff)
00:55 QtWebKitTodo edited by hausmann@webkit.org
(diff)
00:02 Changeset [42127] by mrowe@apple.com
  • 2 edits in trunk/WebKitTools

Fix use of incorrect constant.

  • BuildSlaveSupport/built-product-archive:

03/30/09:

23:29 Changeset [42126] by mrowe@apple.com
  • 1 edit
    1 add in trunk/WebKitTools

Add a script in support of the new build.webkit.org configuration.

Rubber-stamped by Stephanie Lewis.

  • BuildSlaveSupport/built-product-archive: Added.
20:21 Changeset [42125] by kov@webkit.org
  • 3 edits in trunk/WebKit/gtk

2009-03-31 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Holger Freyther.

Updated gtk-doc configuration files for the 1.1.1, 1.1.2, 1.1.3,
and 1.1.4 versions of WebKitGTK+.

  • docs/webkitgtk-docs.sgml:
  • docs/webkitgtk-sections.txt:
20:16 Changeset [42124] by kov@webkit.org
  • 7 edits
    1 add in trunk/WebKit/gtk

2009-03-31 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Holger Freyther.

https://bugs.webkit.org/show_bug.cgi?id=17045
[GTK] Documentation

Added header documentation for WebKitWebNavigationAction and
WebKitWebPolicyDecision.

Added missing documentation pieces for WebKitSoupAuthDialog,
WebKitDownload and WebKitWebPolicyDecision.

Wrote a section about using the WEBKIT_DEBUG environment variable.

  • docs/GNUmakefile.am:
  • docs/webkitenvironment.xml: Added.
  • docs/webkitgtk-docs.sgml:
  • webkit/webkitdownload.cpp:
  • webkit/webkitsoupauthdialog.c:
  • webkit/webkitwebnavigationaction.cpp:
  • webkit/webkitwebpolicydecision.cpp:
19:41 Changeset [42123] by kov@webkit.org
  • 5 edits
    2 adds in trunk

2009-03-30 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Holger Freyther.

Integrate gtk-doc into the Gtk+ buildsystem.

Based on original work by Holger Freyther.

Integrate gtk-doc into our buildsystem. We are forced to
use recursive make for gtk-doc itself as the to be installed
gtk-doc.make (done by gtkdocize) is not working in a non
recursive environment.

The documentation can only be generated when srcdir == builddir.
In all other cases the generated files will not be found. This
limitation can be found in pango and gtk+ as well. There is no
easy way around. The resulting html documentation will be saved
in the srcdir as well.

  • docs/GNUmakefile.am: Added.
  • docs/version.xml.in: Added.
  • docs/webkitgtk-docs.sgml:
19:37 Changeset [42122] by kevino@webkit.org
  • 7 edits in trunk/WebKit/wx

Reviewed by Kevin Ollivier.

Add events to allow apps to override handling for JS prompts and
window object cleared events.

https://bugs.webkit.org/show_bug.cgi?id=24948

19:29 Changeset [42121] by sfalken@apple.com
  • 2 edits in trunk/WebCore

Don't create CFDataRef with a fixed size.
Fixes synchronous XMLHTTPRequests on Windows.


Reviewed by Ada Chan.

  • platform/network/cf/ResourceHandleCFNet.cpp: (WebCore::WebCoreSynchronousLoader::didReceiveData):
17:19 Changeset [42120] by darin@apple.com
  • 4 edits
    2 adds in trunk

WebCore:

2009-03-30 Darin Adler <darin@apple.com>

Reviewed by Sam Weinig.

Bug 24594: PolicyDelegate NavigationAction is WebNavigationTypeOther instead of WebNavigationTypeReload
https://bugs.webkit.org/show_bug.cgi?id=24594
rdar://problem/6682110

Test: fast/loader/reload-policy-delegate.html

  • loader/FrameLoader.cpp: (WebCore::FrameLoader::loadWithDocumentLoader): Set the triggering action here where we know it's the policy loader and can use m_policyLoadType rather than letting it get set inside checkNavigationPolicy.

LayoutTests:

2009-03-30 Darin Adler <darin@apple.com>

Reviewed by Sam Weinig.

Bug 24594: PolicyDelegate NavigationAction is WebNavigationTypeOther instead of WebNavigationTypeReload
https://bugs.webkit.org/show_bug.cgi?id=24594
rdar://problem/6682110

  • fast/frames/resources/iframe-focus.html: Removed bogus but harmless "layoutTestController.loadCount" code.
  • fast/loader/reload-policy-delegate-expected.txt: Added.
  • fast/loader/reload-policy-delegate.html: Added.
17:17 Changeset [42119] by darin@apple.com
  • 3 edits in trunk/WebKitTools

2009-03-30 Darin Adler <darin@apple.com>

Reviewed by Sam Weinig.

Make policy-delegate logging work even for local files by processing the URLs as we
do for other delegates.

  • DumpRenderTree/mac/PolicyDelegate.mm: (-[PolicyDelegate webView:decidePolicyForNavigationAction:request:frame:decisionListener:]): Use _drt_descriptionSuitableForTestResult on the URL.
  • DumpRenderTree/win/PolicyDelegate.cpp: (PolicyDelegate::decidePolicyForNavigationAction): Use urlSuitableForTestResult on the URL.
15:49 Changeset [42118] by timothy@apple.com
  • 2 edits in trunk/WebCore

<rdar://problem/5838871> CrashTracer: 1483 crashes Quicklooking
in Finder (painting without up to date layout)

Reviewed by Dan Bernstein.

  • page/FrameView.cpp: (WebCore::FrameView::paintContents): Add an early return when painting is attempted when layout is needed.
13:14 Changeset [42117] by kov@webkit.org
  • 1 copy in releases/WebKitGTK/webkit-1.1.4

Tagging WebKitGTK+ 1.1.4.

12:59 Changeset [42116] by bolsinga@apple.com
  • 2 edits in trunk/WebCore

2009-03-30 Greg Bolsinga <bolsinga@apple.com>

Reviewed by Simon Fraser.

https://bugs.webkit.org/show_bug.cgi?id=24938


Sort alphabetically.

  • css/CSSComputedStyleDeclaration.cpp: (WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue):
11:33 Changeset [42115] by bolsinga@apple.com
  • 2 edits in trunk/WebCore

2009-03-30 Greg Bolsinga <bolsinga@apple.com>

Reviewed by Simon Fraser.

https://bugs.webkit.org/show_bug.cgi?id=24938


Build fixes for building --3d-rendering and --no-svg

  • rendering/RenderLayerBacking.cpp:
11:27 Changeset [42114] by sfalken@apple.com
  • 2 edits in trunk/WebKitLibraries

2009-03-30 Steve Falkenburg <sfalken@apple.com>

Bump version to 530.

  • win/tools/scripts/VERSION:
11:21 Changeset [42113] by kov@webkit.org
  • 2 edits in trunk/WebKit/gtk

2009-03-30 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Unreviewed. Documenting what changed since the 1.1.3 release, in
the GTK+ port.

  • NEWS:
11:04 Changeset [42112] by bolsinga@apple.com
  • 10 edits in trunk

2009-03-30 Greg Bolsinga <bolsinga@apple.com>

Reviewed by Simon Fraser.


https://bugs.webkit.org/show_bug.cgi?id=24938

Build fixes when building --no-svg


  • css/CSSComputedStyleDeclaration.cpp: (WebCore::CSSComputedStyleDeclaration::getPropertyCSSValue): Add missing CSS properties
  • css/CSSCursorImageValue.cpp: (WebCore::CSSCursorImageValue::updateIfSVGCursorIsUsed): Handle unused param
  • css/CSSParser.cpp: (WebCore::CSSParser::parseValue): Add missing CSS properties
  • css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::applyProperty): Add missing CSS properties
  • dom/Node.cpp: (WebCore::updateSVGElementInstancesAfterEventListenerChange): Handle unused param
  • platform/graphics/Font.cpp: (WebCore::Font::floatWidth): Handle unused param
  • platform/graphics/SimpleFontData.cpp: (WebCore::SimpleFontData::SimpleFontData): Handle unused param

2009-03-30 Greg Bolsinga <bolsinga@apple.com>

Reviewed by Simon Fraser.


https://bugs.webkit.org/show_bug.cgi?id=24938


Build fixes when building --no-svg


DOMHTMLFrameElementPrivate.h and DOMHTMLIFrameElementPrivate.h are only available with ENABLE_SVG.

  • MigrateHeaders.make:
10:46 Changeset [42111] by xan@webkit.org
  • 2 edits in trunk

2009-03-30 Xan Lopez <xlopez@igalia.com>

Unreviewed: bump version to 1.1.4 for release.

  • configure.ac:
10:41 Changeset [42110] by kov@webkit.org
  • 2 edits in trunk/WebKit/gtk

2009-03-30 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Unreviewed. Add information on the bug-fix-only 1.1.3 release.

  • NEWS:
10:37 Changeset [42109] by aroben@apple.com
  • 3 edits in trunk/WebKitTools

Windows build fix

  • DumpRenderTree/win/LayoutTestControllerWin.cpp: (LayoutTestController::waitForPolicyDelegate): Added this empty stub.
  • DumpRenderTree/win/WorkQueueItemWin.cpp: (LoadItem::invoke): (ScriptItem::invoke): Changed to use data members instead of removed member functions.
10:26 Changeset [42108] by darin@apple.com
  • 2 edits in trunk/WebCore

2009-03-30 Darin Adler <darin@apple.com>

Reviewed by Adam Roben.

Bug 24916: REGRESSION: NavigationAction policy dispatch broken
https://bugs.webkit.org/show_bug.cgi?id=24916

  • loader/FrameLoader.cpp: (WebCore::FrameLoader::loadURL): Rearrange the code so that the "event" PassRefPtr is not used twice. This also optimizes the case where a frame name is specified, but it's the name of the frame being loaded. Also called release in all the final uses of FormState.
10:22 Changeset [42107] by darin@apple.com
  • 24 edits in trunk/WebCore

2009-03-30 Darin Adler <darin@apple.com>

Reviewed by Adam Roben.

Bug 24937: NamedNodeMap and NamedAttrMap should not be separate classes
https://bugs.webkit.org/show_bug.cgi?id=24937

Also should fix the Windows COM bindings build.

  • bindings/scripts/CodeGeneratorJS.pm: Removed special case for NamedNodeMap.
  • bindings/scripts/CodeGeneratorObjC.pm: Ditto.
  • dom/NamedAttrMap.h: Renamed class from NamedAttrMap to NamedNodeMap and removed NamedNodeMap base class. Made the NamedNodeMap functions all public and non-virtual. Removed virtualLength.
  • dom/NamedNodeMap.h: Replaced file with just an include of NamedAttrMap.h. As a follow up we will rename the NamedAttrMap files and remove this file.
  • bindings/js/JSNamedNodesCollection.cpp: Renamed from NamedAttrMap to NamedNodeMap.
  • dom/Attr.h: Ditto.
  • dom/Attribute.h: Ditto.
  • dom/Document.cpp: (WebCore::Document::importNode): Ditto.
  • dom/Element.cpp: (WebCore::Element::attributes): Ditto. (WebCore::Element::setAttributeMap): Ditto. (WebCore::Element::createAttributeMap): Ditto. (WebCore::Element::insertedIntoDocument): Ditto. (WebCore::Element::removedFromDocument): Ditto. (WebCore::Element::openTagStartToString): Ditto. (WebCore::Element::removeAttributeNode): Ditto. (WebCore::Element::getAttributeNode): Ditto. (WebCore::Element::getAttributeNodeNS): Ditto. (WebCore::Element::hasAttribute): Ditto. (WebCore::Element::hasAttributeNS): Ditto. (WebCore::Element::normalizeAttributes): Ditto.
  • dom/Element.h: Ditto.
  • dom/NamedAttrMap.cpp: (WebCore::NamedNodeMap::detachAttributesFromElement): Ditto. (WebCore::NamedNodeMap::~NamedNodeMap): Ditto. (WebCore::NamedNodeMap::isMappedAttributeMap): Ditto. (WebCore::NamedNodeMap::getNamedItem): Ditto. (WebCore::NamedNodeMap::getNamedItemNS): Ditto. (WebCore::NamedNodeMap::removeNamedItem): Ditto. (WebCore::NamedNodeMap::removeNamedItemNS): Ditto. (WebCore::NamedNodeMap::setNamedItem): Ditto. (WebCore::NamedNodeMap::item): Ditto. (WebCore::NamedNodeMap::getAttributeItem): Ditto. (WebCore::NamedNodeMap::clearAttributes): Ditto. (WebCore::NamedNodeMap::detachFromElement): Ditto. (WebCore::NamedNodeMap::setAttributes): Ditto. (WebCore::NamedNodeMap::addAttribute): Ditto. (WebCore::NamedNodeMap::removeAttribute): Ditto. (WebCore::NamedNodeMap::mapsEquivalent): Ditto.
  • dom/NamedMappedAttrMap.cpp: (WebCore::NamedMappedAttrMap::clearAttributes): Ditto.
  • dom/NamedMappedAttrMap.h: (WebCore::NamedMappedAttrMap::NamedMappedAttrMap): Ditto.
  • dom/Node.cpp: (WebCore::Node::dumpStatistics): Ditto. (WebCore::Node::isEqualNode): Ditto. (WebCore::Node::isDefaultNamespace): Ditto. (WebCore::Node::lookupNamespaceURI): Ditto. (WebCore::Node::lookupNamespacePrefix): Ditto. (WebCore::Node::compareDocumentPosition): Ditto.
  • dom/Node.h: Ditto.
  • dom/XMLTokenizerLibxml2.cpp: (WebCore::XMLTokenizer::XMLTokenizer): Ditto.
  • editing/ApplyStyleCommand.cpp: (WebCore::isSpanWithoutAttributesOrUnstyleStyleSpan): Ditto. (WebCore::isEmptyFontTag): Ditto. (WebCore::areIdenticalElements): Ditto.
  • editing/markup.cpp: (WebCore::appendStartMarkup): Ditto. (WebCore::completeURLs): Ditto.
  • html/HTMLParser.cpp: (WebCore::HTMLParser::handleError): Ditto.
  • rendering/RenderPartObject.cpp: (WebCore::RenderPartObject::updateWidget): Ditto.
  • svg/SVGAnimatedProperty.h: (WebCore::synchronizeProperty): Ditto.
  • xml/XPathFunctions.cpp: (WebCore::XPath::FunLang::evaluate): Ditto.
  • xml/XPathStep.cpp: (WebCore::XPath::Step::nodesInAxis): Ditto.
09:35 Changeset [42106] by kov@webkit.org
  • 2 edits in trunk/WebCore

2009-03-30 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>

Reviewed by Holger Freyther.

https://bugs.webkit.org/show_bug.cgi?id=24750
[GTK] requests download instead of displaying page

Use soup facilities to append parameters to the content type, to
make that more robust;

  • platform/network/soup/ResourceHandleSoup.cpp: (WebCore::fillResponseFromMessage):
09:24 Changeset [42105] by ddkilzer@apple.com
  • 4 edits in trunk/WebCore

<rdar://problem/6735683> Move SECTORDER_FLAGS from Xcode project to WebCore.xcconfig

Reviewed by Timothy Hatcher, Mark Rowe and Darin Adler.

  • Configurations/Base.xcconfig: Moved Production definition of SECTORDER_FLAGS to here from Xcode project file.
  • Configurations/DebugRelease.xcconfig: Override SECTORDER_FLAGS in Base.xcconfig to the empty string since it is not used for Debug and Release configurations.
  • WebCore.xcodeproj/project.pbxproj: Removed SECTORDER_FLAGS.
08:58 Changeset [42104] by aroben@apple.com
  • 4 edits in trunk/WebKit/win

Build fixes after recent FrameLoader changes

  • WebCoreSupport/WebContextMenuClient.cpp: (WebContextMenuClient::searchWithGoogle): Updated to match urlSelected's current signature. We pass true for the userGesture parameter, since the user must have selected this item from the context menu.
  • WebCoreSupport/WebFrameLoaderClient.cpp: Removed loadURLIntoChildFrame. (WebFrameLoaderClient::createFrame): Changed to call FrameLoader::loadURLIntoChildFrame. This probably should have been done way back around r31281 when that function was added.
  • WebCoreSupport/WebFrameLoaderClient.h: Removed loadURLIntoChildFrame.
08:58 Changeset [42103] by aroben@apple.com
  • 2 edits in trunk/WebCore

Windows build fix after recent Node.idl changes

  • bindings/scripts/CodeGeneratorCOM.pm: Touched this to force sources of classes that derive from Node to rebuild.
08:48 Changeset [42102] by xan@webkit.org
  • 4 edits in trunk/WebKit/gtk

2009-03-30 Xan Lopez <xlopez@igalia.com>

Reviewed by Holger Freyther.

https://bugs.webkit.org/show_bug.cgi?id=24878
[GTK] Use GtkBinding system to handle key events in WebView

Use GtkBinding system for key events in WebKitWebView. Allows the
user to reconfigure the bindings and fixes a bunch of bugs along
the way (we were not checking modifier keys in most actions).

  • webkit/webkitwebview.cpp: (webkit_web_view_move_cursor):
  • webkit/webkitwebview.h:
  • webkitmarshal.list:
08:22 Changeset [42101] by aroben@apple.com
  • 2 edits in trunk/WebCore

Windows build fix

  • svg/graphics/SVGImage.cpp: Added a missing #include.
07:49 Changeset [42100] by dglazkov@chromium.org
  • 7 edits in trunk/WebCore

2009-03-30 Mads Ager <ager@chromium.org>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24879
Add frame accessor to Screen, Console, and BarInfo objects. Add type
accessor to BarInfo objects.

In the V8 JavaScript bindings, we need access to the frame and
type for already created Screen, Console, and BarInfo objects in
order to keep their wrappers alive across GCs.

  • page/BarInfo.cpp: (WebCore::BarInfo::frame): Added. (WebCore::BarInfo::type): Added.
  • page/BarInfo.h:
  • page/Console.cpp: (WebCore::Console::frame): Added.
  • page/Console.h:
  • page/Screen.cpp: (WebCore::Screen::frame): Added.
  • page/Screen.h:
05:21 QtWebKitJournal edited by hausmann@webkit.org
(diff)
02:25 Changeset [42099] by hausmann@webkit.org
  • 3 edits in trunk/WebKit/qt

Document that setHtml/setContent loads only the html/data immediately, not external objects.
2009-03-30 Simon Hausmann <simon.hausmann@nokia.com>

Rubber-stamped by Tor Arne Vestbø.

Document that setHtml/setContent loads only the html/data immediately, not external objects.

02:13 Changeset [42098] by jmalonzo@webkit.org
  • 2 edits in trunk/WebKitTools

2009-03-30 Jan Michael Alonzo <jmalonzo@webkit.org>

Gtk DRT build fix per changeset
http://trac.webkit.org/changeset/42082. Not reviewed.

  • DumpRenderTree/gtk/WorkQueueItemGtk.cpp: (LoadItem::invoke): (ScriptItem::invoke):

03/29/09:

19:49 Changeset [42097] by darin@apple.com
  • 2 edits in trunk/WebCore

2009-03-29 Darin Adler <darin@apple.com>

Requested by Antti Koivisto.

  • loader/DocumentLoader.cpp: (WebCore::DocumentLoader::subresource): Tweaked code to be a little clearer and added a better comment.
19:21 Changeset [42096] by darin@apple.com
  • 3 edits
    2 adds in trunk

WebCore:

2009-03-29 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

Bug 23445: Copying certain hidden text causes a crash
https://bugs.webkit.org/show_bug.cgi?id=23445
rdar://problem/6512520

Test: editing/pasteboard/copy-display-none.html

  • editing/markup.cpp: (WebCore::createMarkup): Added a check for the case where adjusting the start node moves the start of the selection past the end of the range entirely. If we try to iterate we'll never hit the end of the range and will probably crash iterating the rest of the document.

LayoutTests:

2009-03-29 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

Bug 23445: Copying certain hidden text causes a crash
https://bugs.webkit.org/show_bug.cgi?id=23445
rdar://problem/6512520

  • editing/pasteboard/copy-display-none-expected.txt: Added.
  • editing/pasteboard/copy-display-none.html: Added.
18:59 Changeset [42095] by darin@apple.com
  • 3 edits in trunk/WebCore

2009-03-29 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

Bug 24672: ASSERTION FAILURE: !m_purgeableData in WebCore::CachedResource::data() saving a WebArchive
https://bugs.webkit.org/show_bug.cgi?id=24672
rdar://problem/6574263

I couldn't create a test case for this. In fact, the case in the bug doesn't exist any more,
but there is an Apple-internal website I was able to use to reproduce and fix.

  • loader/CachedResource.h: Made makePurgeable public.
  • loader/DocumentLoader.cpp: (WebCore::DocumentLoader::subresource): Call makePurgeable to make the resource non-purgeable so we can get its data if it hasn't yet been purged.
17:06 Changeset [42094] by darin@apple.com
  • 2 edits in trunk/WebCore

2009-03-29 Darin Adler <darin@apple.com>

Fix build again.

  • loader/EmptyClients.h: Fix typo in name of FormState.h.
16:53 Changeset [42093] by darin@apple.com
  • 2 edits in trunk/WebCore

2009-03-29 Darin Adler <darin@apple.com>

Try to fix the release build.

  • history/CachedPage.cpp: Make include unconditional, not debug-only.
16:51 Changeset [42092] by bolsinga@apple.com
  • 2 edits in trunk/WebCore

Undo the previous build fix.

16:44 Changeset [42091] by bolsinga@apple.com
  • 2 edits in trunk/WebCore

Fix build break.

16:29 Changeset [42090] by darin@apple.com
  • 1 edit in trunk/WebCore/ChangeLog

Fix ChangeLog.

16:29 Changeset [42089] by darin@apple.com
  • 3 edits in trunk/WebCore

2009-03-29 Dmitry Titov <dimich@chromium.org>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24926
Fix Chromium build break.

  • bindings/v8/custom/V8DOMWindowCustom.cpp: (WebCore::CALLBACK_FUNC_DECL(DOMWindowShowModalDialog)): renamed 'features' into 'windowFeatures'.
16:26 Changeset [42088] by dimich@chromium.org
  • 2 edits in trunk/WebCore

2009-03-29 Dmitry Titov <dimich@chromium.org>

Reviewed by Darin Adler.

https://bugs.webkit.org/show_bug.cgi?id=24926
Fix Chromium build break.

  • bindings/v8/custom/V8DOMWindowCustom.cpp: (WebCore::CALLBACK_FUNC_DECL(DOMWindowShowModalDialog)): renamed 'features' into 'windowFeatures'.
11:09 Changeset [42087] by darin@apple.com
  • 2 edits in trunk/WebCore

2009-03-29 Darin Adler <darin@apple.com>

Another attempt to fix the Qt build.

  • WebCore.pro: Added CheckedRadioButtons.cpp.
11:00 Changeset [42086] by darin@apple.com
  • 3 edits in trunk/WebKit/qt

2009-03-29 Darin Adler <darin@apple.com>

Try to fix Qt build.

  • WebCoreSupport/EditorClientQt.cpp: Added include of HTMLElement.h.
  • WebCoreSupport/FrameLoaderClientQt.cpp: Added include of FormState.h. (WebCore::FrameLoaderClientQt::prepareForDataSourceReplacement): Removed unneeded call to detachChildren, which is called by FrameLoader right after this. (WebCore::FrameLoaderClientQt::createFrame): Use loadURLIntoChildFrame as other platforms do in the corresponding functions.
10:31 Changeset [42085] by darin@apple.com
  • 2 edits in trunk/WebKitTools

2009-03-29 Darin Adler <darin@apple.com>

  • DumpRenderTree/qt/jsobjects.cpp: Rolled out changes to this file. Maybe this will fix Qt building of DumpRenderTree.
10:15 Changeset [42084] by darin@apple.com
  • 34 edits
    2 copies in trunk

WebCore:

2009-03-29 Darin Adler <darin@apple.com>

Reviewed by Cameron Zwarich.

Bug 24921: remove include of HTMLFormElement.h from Document.h
https://bugs.webkit.org/show_bug.cgi?id=24921

  • GNUmakefile.am: Added CheckedRadioButtons source files.
  • WebCore.scons: Ditto.
  • WebCore.vcproj/WebCore.vcproj: Ditto.
  • WebCore.xcodeproj/project.pbxproj: Ditto.
  • WebCoreSources.bkl: Ditto.
  • bindings/objc/PublicDOMInterfaces.h: Changed type of parentElement to Element.
  • dom/CheckedRadioButtons.cpp: Copied from WebCore/html/HTMLFormElement.cpp. Removed all the code except for CheckedRadioButtons, and moved that out to be a namespace-level class instead of a member of HTMLFormElement.
  • dom/CheckedRadioButtons.h: Copied from WebCore/html/HTMLFormElement.h. Removed all the code except for CheckedRadioButtons, and moved that out to be a namespace-level class instead of a member of HTMLFormElement.
  • dom/Document.h: Removed unneeded includes, including HTMLFormElement.h. Added include of CheckedRadioButtons.h. Changed uses of CheckedRadioButtons class to use a namespace-level class instead of a member of HTMLFormElement.
  • dom/Node.idl: Changed the type of parentElement from Node to Element for two reasons. 1) Node was incorrect, since parentElement returns an Element. 2) The bindings won't compile any more unless they know parentElement's return value is an Element due to include changes.
  • html/HTMLFormElement.cpp: Moved definitions of the functions in the CheckedRadioButtons class to the new CheckedRadioButtons.cpp.
  • html/HTMLFormElement.h: Moved definition of the CheckedRadioButtons class to the new CheckedRadioButtons.h.
  • html/HTMLInputElement.cpp: (WebCore::checkedRadioButtons): Changed use of CheckedRadioButtons class to use a namespace-level class instead of a member of HTMLFormElement.
  • page/Chrome.cpp: (WebCore::Chrome::createWindow): Tweaked formatting.
  • css/CSSGrammar.y: Added newly-needed include.
  • dom/Range.cpp: Ditto.
  • editing/BreakBlockquoteCommand.cpp: Ditto.
  • editing/CompositeEditCommand.cpp: Ditto.
  • editing/InsertLineBreakCommand.cpp: Ditto.
  • editing/ModifySelectionListLevel.cpp: Ditto.
  • editing/RemoveFormatCommand.cpp: Ditto.
  • editing/TextIterator.cpp: Ditto.
  • editing/VisiblePosition.cpp: Ditto.
  • loader/DocLoader.cpp: Ditto.
  • page/AccessibilityRenderObject.cpp: Ditto.
  • page/Page.cpp: Ditto.
  • rendering/RenderBlock.cpp: Ditto.
  • rendering/RenderLayer.cpp: Ditto.
  • rendering/RenderObject.h: Ditto.
  • storage/LocalStorageArea.cpp: Ditto.
  • storage/SessionStorageArea.cpp: Ditto.

WebKit/mac:

2009-03-29 Darin Adler <darin@apple.com>

Reviewed by Cameron Zwarich.

  • Plugins/WebNullPluginView.mm: Added now-needed includes.
  • WebView/WebHTMLRepresentation.mm: Ditto.
  • WebView/WebHTMLView.mm: Ditto.
10:14 Changeset [42083] by darin@apple.com
  • 2 edits in trunk/LayoutTests

2009-03-29 Darin Adler <darin@apple.com>

Reviewed by Cameron Zwarich.

  • fast/forms/targeted-frame-submission.html: Fixes from review comments.
10:11 Changeset [42082] by darin@apple.com
  • 22 edits in trunk

WebKitTools:

2009-03-29 Darin Adler <darin@apple.com>

Reviewed by Cameron Zwarich.

Bug 24922: change WorkQueue-based navigation tests to not depend on synchronous form submission
https://bugs.webkit.org/show_bug.cgi?id=24922

Don't rely on delegate methods to guess whether a queued step starts a load or not, because
script-triggered loads can't be detected that way. Instead have the steps themselves indicate
whether or not a load was triggered.

  • DumpRenderTree/LayoutTestController.cpp: (queueLoadingScriptCallback): Added. Version for scripts that perform loads. (queueNonLoadingScriptCallback): Added. Version for scripts that do not perform loads. (LayoutTestController::staticFunctions): Replaced queueScript with queueLoadingScript and queueNonLoadingScript. (LayoutTestController::queueBackNavigation): Moved here since it's platform-independent now. (LayoutTestController::queueForwardNavigation): Ditto. (LayoutTestController::queueLoadingScript): Replacement for queueScript. Here since it's platform-independent. (LayoutTestController::queueNonLoadingScript): Ditto. (LayoutTestController::queueReload): Moved here since it's platform-independent now.
  • DumpRenderTree/LayoutTestController.h: Ditto.
  • DumpRenderTree/WorkQueue.cpp: (WorkQueue::processWork): Added. Shared by the different platform's work queue implementations.
  • DumpRenderTree/WorkQueue.h: Ditto.
  • DumpRenderTree/WorkQueueItem.h: Changed the invoke function to return true if the item started a load. Removed unused getter functions. Made invoke functions private. Added LoadingScriptItem and NonLoadingScriptItem, making ScriptItem an abstract base.
  • DumpRenderTree/gtk/DumpRenderTree.cpp: (processWork): Use the new WorkQueue::processWork function to implement the new rule.
  • DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: Removed queue functions that are now platform-independent and in LayoutTestController.cpp.
  • DumpRenderTree/gtk/WorkQueueItemGtk.cpp: (LoadItem::invoke): Return true. (ReloadItem::invoke): Ditto. (ScriptItem::invoke): Ditto. (BackForwardItem::invoke): Ditto.
  • DumpRenderTree/mac/FrameLoadDelegate.mm: (-[FrameLoadDelegate processWork:]): Use the new WorkQueue::processWork function to implement the new rule.
  • DumpRenderTree/mac/LayoutTestControllerMac.mm: Removed queue functions that are now platform-independent and in LayoutTestController.cpp.
  • DumpRenderTree/mac/WorkQueueItemMac.mm: (LoadItem::invoke): Return true. (ReloadItem::invoke): Ditto. (ScriptItem::invoke): Ditto. (BackForwardItem::invoke): Ditto.
  • DumpRenderTree/qt/jsobjects.cpp: (LayoutTestController::processWork): Use the new WorkQueue::processWork function to implement the new rule. Removed queue functions that are now platform-independent and in LayoutTestController.cpp.
  • DumpRenderTree/win/FrameLoadDelegate.cpp: (FrameLoadDelegate::processWork): Use the new WorkQueue::processWork function to implement the new rule.
  • DumpRenderTree/win/LayoutTestControllerWin.cpp: Removed queue functions that are now platform-independent and in LayoutTestController.cpp.
  • DumpRenderTree/win/WorkQueueItemWin.cpp: (LoadItem::invoke): Return false when we fail, true when we succeed. (ReloadItem::invoke): Ditto. (ScriptItem::invoke): Ditto. (BackForwardItem::invoke): Ditto.

LayoutTests:

2009-03-29 Darin Adler <darin@apple.com>

Reviewed by Cameron Zwarich.

Bug 24922: change WorkQueue-based navigation tests to not depend on synchronous form submission
https://bugs.webkit.org/show_bug.cgi?id=24922

  • fast/forms/targeted-frame-submission.html: Changed test to wait until the subframe loads before moving on to the next test, using an onload attribute.
  • http/tests/navigation/fallback-anchor-reload.html: Removed use of queueScript here for a script that can instead just be run immedidately.
  • http/tests/navigation/onload-navigation-iframe-timeout.html: Changed test to wait until the subframe loads before moving on to the next test, using an onload attribute.
  • http/tests/navigation/onload-navigation-iframe.html: Ditto.
  • http/tests/navigation/resources/testcode.js: Use queueNonLoadingScript instead of queueScript for scripts that don't involve submitting a form. Use queueLoadingScript for scripts that do involve submitting a form.
08:37 Changeset [42081] by darin@apple.com
  • 4 edits
    2 adds in trunk

WebCore:

2009-03-29 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

<rdar://problem/6015407> attr parsing should allow only identifiers

Test: fast/css/attr-parsing.html

  • css/CSSParser.cpp: (WebCore::CSSParser::parseContent): Allow only CSS_IDENT, and filter out identifiers that start with "-".
  • css/CSSPrimitiveValue.cpp: (WebCore::CSSPrimitiveValue::cssText): Added a case for CSS_ATTR so the test case works. This has the pleasant side effect of fixing a bug too.

LayoutTests:

2009-03-29 Darin Adler <darin@apple.com>

Reviewed by Dan Bernstein.

<rdar://problem/6015407> attr parsing should allow only identifiers

  • fast/css/attr-parsing-expected.txt: Added.
  • fast/css/attr-parsing.html: Added.
08:17 Changeset [42080] by darin@apple.com
  • 1 edit
    4 adds in trunk/LayoutTests

2009-03-29 Darin Adler <darin@apple.com>

Bug 12104: Native Slider: When the thumb's height is specified as a percentage, it is not centered properly
https://bugs.webkit.org/show_bug.cgi?id=12104

Landed the test case from this bug. The bug was fixed by the fix for bug 24740 below.

  • fast/forms/range-thumb-height-percentage.html: Added.
  • platform/mac/fast/forms/range-thumb-height-percentage-expected.checksum: Added.
  • platform/mac/fast/forms/range-thumb-height-percentage-expected.png: Added.
  • platform/mac/fast/forms/range-thumb-height-percentage-expected.txt: Added.
02:45 Changeset [42079] by ap@webkit.org
  • 2 edits in trunk/WebCore

<rdar://problem/6492712> Cross-origin redirects are not handled correctly.

Forgot to save the file after applying changes for review comments.

02:16 Changeset [42078] by ap@webkit.org
  • 5 edits
    4 adds in trunk

Reviewed by Darin Adler.

<rdar://problem/6492712> Cross-origin redirects are not handled correctly.

Test: http/tests/xmlhttprequest/redirect-cross-origin-tripmine.html

  • platform/network/cf/ResourceHandleCFNet.cpp: (WebCore::WebCoreSynchronousLoader::WebCoreSynchronousLoader): (WebCore::ResourceHandle::loadResourceSynchronously): (WebCore::WebCoreSynchronousLoader::willSendRequest): (WebCore::WebCoreSynchronousLoader::didReceiveResponse): (WebCore::WebCoreSynchronousLoader::didReceiveData): (WebCore::WebCoreSynchronousLoader::didFinishLoading): (WebCore::WebCoreSynchronousLoader::didFail): (WebCore::WebCoreSynchronousLoader::didReceiveChallenge): (WebCore::WebCoreSynchronousLoader::load): Match Mac behavior more closely - we shouldn't rely on underlying library handling of synchronous requests.
  • loader/DocumentThreadableLoader.cpp: (WebCore::DocumentThreadableLoader::willSendRequest): -[NSURLConnection cancel] doesn't fully cancel the connection if called from willSendRequest delegate method for a redirect.
  • platform/network/mac/ResourceHandleMac.mm: (-[WebCoreSynchronousLoader connection:willSendRequest:redirectResponse:]): Match async behavior more closely.
Note: See TracTimeline for information about the timeline view.