⚠ Archived content — this site is no longer maintained.   Current WebKit documentation is at docs.webkit.org.

Timeline



Nov 6, 2007:

11:23 PM Changeset in webkit [27503] by mjs
  • 2 edits in trunk/JavaScriptCore

Reviewed by Eric.


  • only collect when the heap is full, unless we have lots of extra cost garbage


1.1% SunSpider speedup.


This shouldn't hit memory use much since the extra space in those
blocks hangs around either way.

  • kjs/collector.cpp: (KJS::Collector::heapAllocate): (KJS::Collector::collect): Fix logic error that reversed the sense of collect's return value.
11:00 PM Changeset in webkit [27502] by aroben
  • 1 edit
    1 move in trunk/LayoutTests

Move a Mac-specific test to platform/mac

  • editing/pasteboard/5583362.html: Removed.
  • platform/mac/editing/pasteboard/5583362.html: Added.
10:03 PM Changeset in webkit [27501] by oliver
  • 3 edits in trunk/JavaScriptCore

Avoid unnecessarily boxing the result from post inc/decrement for 0.3% gain in sunspider

Reviewed by Maciej

We now convert the common 'for (...; ...; <var>++) ...' to the semantically identical
'for (...; ...; ++<var>) ...'.

9:44 PM JavaScript performance improvement ideas edited by eric@webkit.org
(diff)
9:41 PM Changeset in webkit [27500] by eric@webkit.org
  • 2 edits in trunk/WebKitTools

2007-11-06 Eric Seidel <​eric@webkit.org>

  • Scripts/build-testkjs: build fix... too many $$
9:31 PM Changeset in webkit [27499] by bdakin
  • 4 edits
    4 adds in trunk

WebCore:

Reviewed by Sam.

<rdar://problem/5575812> REGRESSION:When using absolute positioning
with overflow:auto div, WebKit seems to add an additional 15px

  • rendering/RenderBox.cpp: (WebCore::RenderBox::containingBlockWidthForPositioned): We need to subtract off the vertical scrollbar width too.

LayoutTests:

Reviewed by Sam.

New test for <rdar://problem/5575812> REGRESSION:When using
absolute positioning with overflow:auto div, WebKit seems to add an
additional 15px

  • fast/overflow/overflow-auto-position-absolute.html: Added.
  • platform/mac/fast/overflow/overflow-auto-position-absolute-expected.checksum: Added.
  • platform/mac/fast/overflow/overflow-auto-position-absolute-expected.png: Added.
  • platform/mac/fast/overflow/overflow-auto-position-absolute-expected.txt: Added.


Test with changed results. The measurements in the rtl example now
more closely mirror the ltr example.

  • platform/mac/fast/overflow/unreachable-overflow-rtl-bug-expected.txt:
9:29 PM Changeset in webkit [27498] by aroben
  • 5 edits in trunk/WebKit/win

Change WebLocalizableStrings to take UTF-8 C strings

This matches the way things work on the Mac, and will allow source
files containing localizable strings to be shared between Mac and
Windows. The old functions have not been removed for compatibility
reasons, but are now just wrappers around the new UTF-8 functions.

Reviewed by Ada.

  • WebKit.vcproj/WebKit.def: Added new functions.
  • WebKit.vcproj/WebKit_debug.def: Ditto.
  • WebLocalizableStrings.cpp: (copyLocalizedStringFromBundle): Changed to take a WebCore::String representing the key. (localizedString): Refactored from WebLocalizedString. Takes a WebCore::String representing the key. (localizedLPCTSTR): Ditto for WebLocalizedLPCTSTR. (WebLocalizedStringUTF8): Added. Takes a UTF-8 C string as the key. (WebLocalizedLPCTSTRUTF8): Ditto. (WebLocalizedString): Changed to call localizedString. (WebLocalizedLPCTSTR): Changed to call localizedLPCTSTR.
  • WebLocalizableStrings.h: Changed macros to use the new UTF-8 functions.
9:24 PM Changeset in webkit [27497] by eric@webkit.org
  • 2 edits in trunk/WebKitTools

2007-11-06 Eric Seidel <​eric@webkit.org>

Reviewed by Mark Rowe.

  • Scripts/build-testkjs: return xcodebuild's exit status, instead of grep's
9:00 PM JavaScript performance improvement ideas edited by eric@webkit.org
(diff)
7:01 PM Changeset in webkit [27496] by mrowe@apple.com
  • 2 edits
    24 moves
    1 add in trunk/WebKit

Move Mac files from WebKit into WebKit/mac.

5:13 PM Changeset in webkit [27495] by justing
  • 3 edits
    5 adds in trunk

WebCore:

Reviewed by Dan Bernstein.


<rdar://problem/5583387> ASSERTION FAILED: !refChild->hasTagName(bodyTag) when pasting newline in plain text into rich text Mail
<rdar://problem/5583362> REGRESSION (5523.10.3-TOT): Newlines stripped when pasting plain text in Mail

  • editing/markup.cpp: (WebCore::createFragmentFromText): Put paragraphs of text into clones of the block that encloses the input context, unless that block is the body, which shouldn't be cloned. In that case, use regular divs, as we did before r27369.

LayoutTests:

Reviewed by Dan Bernstein.

<rdar://problem/5583387> ASSERTION FAILED: !refChild->hasTagName(bodyTag) when pasting newline in plain text into rich text Mail
<rdar://problem/5583362> REGRESSION (5523.10.3-TOT): Newlines stripped when pasting plain text in Mail

  • editing/pasteboard/5583362.html: Added.
  • editing/resources/plaintext-pasteboard-data.dat: Added.
  • platform/mac/editing/pasteboard/5583362-expected.checksum: Added.
  • platform/mac/editing/pasteboard/5583362-expected.png: Added.
  • platform/mac/editing/pasteboard/5583362-expected.txt: Added.
5:03 PM Changeset in webkit [27494] by eseidel
  • 2 edits in trunk/JavaScriptCore

2007-11-06 Eric Seidel <​eric@webkit.org>

Reviewed by darin.

This fixes a regressed layout test for string + object


SunSpider claims this was an overall 0.3% speedup, although some individual tests were slower.

  • kjs/nodes.cpp: (KJS::add): remove erroneous "fast path" for string + *
4:17 PM JavaScript performance improvement ideas edited by Darin Adler
(diff)
4:12 PM JavaScript performance improvement ideas edited by eric@webkit.org
(diff)
4:11 PM JavaScript performance improvement ideas edited by eric@webkit.org
(diff)
4:06 PM Changeset in webkit [27493] by ggaren@apple.com
  • 3 edits in trunk/JavaScriptCore

Reviewed by Eric Seidel.


Added toJSNumber, a fast path for converting a JSValue to a JS number,
and deployed it in postfix expressions. In the fast case this
eliminates a call to jsNumber.


0.4% speedup on SunSpider.

  • ChangeLog:
  • kjs/nodes.cpp: (KJS::PostIncResolveNode::evaluate): (KJS::PostIncLocalVarNode::evaluate): (KJS::PostDecResolveNode::evaluate): (KJS::PostDecLocalVarNode::evaluate): (KJS::PostIncBracketNode::evaluate): (KJS::PostDecBracketNode::evaluate): (KJS::PostIncDotNode::evaluate): (KJS::PostDecDotNode::evaluate): (KJS::UnaryPlusNode::evaluate):
  • kjs/value.h: (KJS::JSValue::toJSNumber):
4:04 PM Changeset in webkit [27492] by mrowe@apple.com
  • 2 edits in trunk/WebCore

2007-11-06 Christian Dywan <​christian@twotoasts.de>

Reviewed by Darin.

Fix ​http://bugs.webkit.org/show_bug.cgi?id=15828
Bug 15828: WebKit GTK include and lib directory is installed in qt4-named directory

  • WebCore.pro: Use sane default install paths for the gtk port.
4:02 PM Changeset in webkit [27491] by mrowe@apple.com
  • 2 edits in trunk/WebKit/gtk

2007-11-06 Rodney Dawes <​dobey@wayofthemonkey.com>

Fix ​http://bugs.webkit.org/attachment.cgi?id=17043&action=view
Bug 15766: [GTK] WebKit sometimes spews binary data as text/plain into iframes

FrameLoaderClient::objectContentType needs to check with the MIMETypeRegistry
to determine whether the given MIME type is displayable as an image or non-image.

  • WebCoreSupport/FrameLoaderClientGtk.cpp: (FrameLoaderClient::objectContentType): Change logic to match that in the Windows and Mac ports.
3:19 PM Changeset in webkit [27490] by antti
  • 2 edits in trunk/LayoutTests

Correct test results.

  • platform/mac/fast/layers/layer-visibility-expected.txt:
3:00 PM Changeset in webkit [27489] by antti
  • 5 edits in trunk/LayoutTests

Reviewed by Hyatt.


Update test to cover
<rdar://problem/5521068>
Visibility not propagated correctly for children of a layer with z-index


The fix for this was (accidentally) already checked in with r27277,
with retroactive r=hyatt.

  • fast/layers/layer-visibility.html:
  • platform/mac/fast/layers/layer-visibility-expected.checksum:
  • platform/mac/fast/layers/layer-visibility-expected.png:
  • platform/mac/fast/layers/layer-visibility-expected.txt:
2:28 PM Changeset in webkit [27488] by justing
  • 4 edits
    4 adds in trunk

WebCore:

Reviewed by Darin Adler.


<rdar://problem/5576619>
REGRESSION: Caret disappears after deleting the last character in inline hole (15714)

  • editing/TypingCommand.cpp: (WebCore::TypingCommand::deleteSelection): Like the other TypingCommands, including both of the other deletion commands, call typingAddedToOpenCommand(), which takes the command's endingSelection and sets it as selection.

LayoutTests:

Reviewed by Darin Adler.


<rdar://problem/5576619> REGRESSION: Caret disappears after deleting the last character in inline hole (15714)

  • platform/mac/editing/input/5576619-expected.checksum: Added.
  • platform/mac/editing/input/5576619-expected.png: Added.
  • platform/mac/editing/input/5576619-expected.txt: Added.
  • platform/mac/editing/input/5576619.html: Added.
  • platform/mac/editing/input/text-input-controller-expected.txt:
1:37 PM Changeset in webkit [27487] by Darin Adler
  • 2 edits in trunk/JavaScriptCore

Reviewed by Maciej.

There was a mistake in the algorithm used to find an empty slot in the property
map entries vector; when we were putting in a new property value and not overwriting
an existing deleted sentinel, we would enlarge the entries vector, but would not
overwrite the stale data that's in the new part. It was easy to pin this down by
turning on property map consistency checks -- I never would have landed with this
bug if I had run the regression tests once with consistency checks on!

  • kjs/property_map.cpp: (KJS::PropertyMap::put): Changed logic for the case where foundDeletedElement is false to always use the item at the end of the entries vector. Also allowed me to merge with the logic for the "no deleted sentinels at all" case.
1:03 PM Changeset in webkit [27486] by mitz@apple.com
  • 4 edits
    6 adds in trunk

WebCore:

Reviewed by Antti Koivisto and Dave Hyatt.

  • fix <rdar://problem/5582961> Incorrect layout and floating object list corruption when CSS decreases a block's height

Test: fast/block/float/overhanging-after-height-decrease.html

  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::layoutBlock): If after calculating the height it turns out that there are overhanging floats that were not overhanging before, rescan children with overhanging floats and add them. (WebCore::RenderBlock::layoutBlockChildren): Added a parameter that returns the lowest float bottom of any of the children. (WebCore::RenderBlock::addOverhangingFloats): Changed to return the lowest float bottom.
  • rendering/RenderBlock.h:

LayoutTests:

Reviewed by Antti Koivisto and Dave Hyatt.

  • test for <rdar://problem/5582961> Incorrect layout and floating object list corruption when CSS decreases a block's height
  • fast/block/float/overhanging-after-height-decrease.html: Added.
  • platform/mac-leopard/fast/block: Added.
  • platform/mac-leopard/fast/block/float: Added.
  • platform/mac-leopard/fast/block/float/overhanging-after-height-decrease-expected.checksum: Added.
  • platform/mac-leopard/fast/block/float/overhanging-after-height-decrease-expected.png: Added.
  • platform/mac/fast/block/float/overhanging-after-height-decrease-expected.txt: Added.
11:53 AM Changeset in webkit [27485] by adele
  • 8 edits
    4 adds in trunk

WebCore:

Reviewed by Darin.

Switched all uses of HTMLImageLoader to use OwnPtrs.

  • html/HTMLInputElement.h:
  • html/HTMLInputElement.cpp: (WebCore::HTMLInputElement::init): (WebCore::HTMLInputElement::~HTMLInputElement): (WebCore::HTMLInputElement::setInputType): (WebCore::HTMLInputElement::parseMappedAttribute): (WebCore::HTMLInputElement::attach):
  • html/HTMLObjectElement.h:
  • html/HTMLObjectElement.cpp: (WebCore::HTMLObjectElement::HTMLObjectElement): (WebCore::HTMLObjectElement::~HTMLObjectElement): (WebCore::HTMLObjectElement::parseMappedAttribute): (WebCore::HTMLObjectElement::attach):
  • html/HTMLVideoElement.h:
  • html/HTMLVideoElement.cpp: (WebCore::HTMLVideoElement::HTMLVideoElement): (WebCore::HTMLVideoElement::attach): (WebCore::HTMLVideoElement::detach): (WebCore::HTMLVideoElement::parseMappedAttribute):

LayoutTests:

Reviewed by Darin.

Added test for poster attribute.

  • media/content/abe.png: Added.
  • media/content/greenbox.png: Added.
  • media/video-poster-expected.txt: Added.
  • media/video-poster.html: Added.
11:49 AM Changeset in webkit [27484] by oliver
  • 2 edits in trunk/JavaScriptCore

Fix previous patch to use a 3 bit shift, a 16 bit shift causes a regression in sunspider.

RS=Darin

11:34 AM Changeset in webkit [27483] by oliver
  • 2 edits in trunk/JavaScriptCore

Replace boolean comparisons in AddNode with mask comparisons for a 0.2% improvement in sunspider.

Reviewed by Darin

10:56 AM Changeset in webkit [27482] by eseidel
  • 2 edits in trunk/JavaScriptCore

2007-11-06 Eric Seidel <​eric@webkit.org>

Reviewed by darin.


SunSpider claims this is a 1.1% speedup.

  • kjs/nodes.cpp: (KJS::throwOutOfMemoryError): Added, non inline. (KJS::addSlowCase): renamed from add(), non inline. (KJS::add): add fast path for String + String, Number + Number and String + *
10:43 AM Changeset in webkit [27481] by Darin Adler
  • 4 edits in trunk/WebKit/win

Reviewed by Adam and Steve.

  • added hooks needed to implement showModalDialog on Windows
  • Interfaces/IWebUIDelegate.idl: Added canRunModal, createModalDialog, and runModal functions to the end of IWebUIDelegate3.
  • WebChromeClient.h: Added uiDelegate and uiDelegate2 helper functions, so it's easier to write client functions.
  • WebChromeClient.cpp: (WebChromeClient::createWindow): Implemented dialog case. Calls IWebUIDelegate3. (WebChromeClient::canRunModal): Implemented. Calls IWebUIDelegate3. (WebChromeClient::runModal): Ditto. (WebChromeClient::uiDelegate): Added. (WebChromeClient::uiDelegate2): Added. (WebChromeClient::uiDelegate3): Added.
10:11 AM Changeset in webkit [27480] by antti
  • 5 edits
    6 adds in trunk

WebCore:

Reviewed by Darin.


Trigger media load on on src attribute changes as specified in new HTML5 draft.

Tests: media/video-src-change.html

media/video-src-remove.html
media/video-src-set.html

  • html/HTMLMediaElement.cpp: (WebCore::HTMLMediaElement::attributeChanged):
  • html/HTMLMediaElement.h:

LayoutTests:

Reviewed by Darin.


Tests for media src attribute changes.

  • media/video-src-change-expected.txt: Added.
  • media/video-src-change.html: Added.
  • media/video-src-remove-expected.txt: Added.
  • media/video-src-remove.html: Added.
  • media/video-src-set-expected.txt: Added.
  • media/video-src-set.html: Added.
  • media/video-test.js:
10:10 AM Changeset in webkit [27479] by ap
  • 2 edits in trunk/WebKit/win

Rubber-stamped by Adam Roben.

Windows build fix.

  • WebFrame.cpp: (WebFrame::string): plainText() returns a String now.
9:48 AM Changeset in webkit [27478] by mitz@apple.com
  • 3 edits
    5 adds in trunk

WebCore:

Reviewed by Darin Adler.

Test: fast/repaint/make-children-non-inline.html

  • rendering/RenderBlock.cpp: (WebCore::RenderBlock::makeChildrenNonInline): Repaint the block. This is needed because the inline children may be repositioned as they move into new anonymous blocks, but those blocks have no knowledge of where their children used to be, so they cannot invalidate those areas.

LayoutTests:

Reviewed by Darin Adler.

  • fast/repaint/make-children-non-inline.html: Added.
  • platform/mac/fast/repaint/make-children-non-inline-expected.txt: Added.
  • platform/mac-leopard/fast/repaint: Added.
  • platform/mac-leopard/fast/repaint/make-children-non-inline-expected.checksum: Added.
  • platform/mac-leopard/fast/repaint/make-children-non-inline-expected.png: Added.
8:07 AM Changeset in webkit [27477] by ap
  • 10 edits in trunk

Reviewed by Darin.

​http://bugs.webkit.org/show_bug.cgi?id=15847
Some editing cleanup

No change in functionality.

WebCore:

  • editing/TextIterator.cpp: (WebCore::plainText):
  • editing/TextIterator.h: Made WebCore::plainText() return String instead of DeprecatedString.
  • bridge/mac/WebCoreAXObject.mm: (-[WebCoreAXObject textUnderElement]): (-[WebCoreAXObject value]): (-[WebCoreAXObject doAXBoundsForTextMarkerRange:]): (-[WebCoreAXObject doAXNextSentenceEndTextMarkerForTextMarker:]): (-[WebCoreAXObject doAXPreviousSentenceStartTextMarkerForTextMarker:]): Updated for the above change. There is no need to explicitly convert to NSString now.
  • editing/EditCommand.cpp: (WebCore::EditCommand::EditCommand): m_startingSelection and m_endingSelection are actually initialized in constructor body, so the work done in initializer list was wasted.

(WebCore::EditCommand::apply): Moved some stars.
(WebCore::EditCommand::unapply): Ditto.
(WebCore::EditCommand::reapply): Ditto.
(WebCore::EditCommand::setStartingSelection): The loop exit condition was evaluated twice,
removed one of the checks.

  • editing/SelectionController.cpp: (WebCore::SelectionController::toString): plainText() result type now matches what we need here.
  • page/mac/WebCoreFrameBridge.h:
  • page/mac/WebCoreFrameBridge.mm: Removed unused -[WebCoreFrameBridge deleteKeyPressedWithSmartDelete:granularity:] and -[WebCoreFrameBridge forwardDeleteKeyPressedWithSmartDelete:granularity:].

WebKit:

  • WebView/WebHTMLView.mm: (-[WebHTMLView deleteToEndOfLine:]): (-[WebHTMLView deleteToEndOfParagraph:]): WebCore had a duplicate of the same logic already. We are passing a boundary value to a function that expects granularity, this may need to be straightened out in the future.
1:08 AM Changeset in webkit [27476] by eseidel
  • 2 edits in trunk/JavaScriptCore

2007-11-06 Eric Seidel <​eric@webkit.org>

Reviewed by mjs.


Avoid more UString creation.


SunSpider claims this is a 0.4% speedup.

  • kjs/regexp_object.cpp: (KJS::RegExpObjectImp::construct): use UString::find(UChar)

Nov 5, 2007:

11:47 PM Changeset in webkit [27475] by mrowe@apple.com
  • 2 edits in trunk/WebCore

2007-11-05 Mark Rowe <​mrowe@apple.com>

Mac build fix.

  • bindings/js/JSHTMLDocumentCustom.cpp: (WebCore::writeHelper):
11:30 PM Changeset in webkit [27474] by mrowe@apple.com
  • 2 edits in trunk/JavaScriptCore

2007-11-05 Mark Rowe <​mrowe@apple.com>

Mac build fix.

  • kjs/array_object.cpp: (KJS::ArrayProtoFunc::callAsFunction):
11:20 PM Changeset in webkit [27473] by aroben
  • 3 edits in trunk/WebCore

Don't crash if SafariTheme can't be loaded

PlatformScrollBarSafari and RenderThemeSafari were not checking
whether SafariTheme was successfully loaded. All other uses of
SafariTheme already check this.

Reviewed by Steve.

  • platform/win/PlatformScrollBarSafari.cpp: (WebCore::PlatformScrollbar::paintButton): (WebCore::PlatformScrollbar::paintTrack): (WebCore::PlatformScrollbar::paintThumb):
  • rendering/RenderThemeSafari.cpp: (WebCore::RenderThemeSafari::isControlStyled): (WebCore::RenderThemeSafari::paintCheckbox): (WebCore::RenderThemeSafari::paintRadio): (WebCore::RenderThemeSafari::paintButton): (WebCore::RenderThemeSafari::paintTextField): (WebCore::RenderThemeSafari::paintCapsLockIndicator): (WebCore::RenderThemeSafari::paintTextArea): (WebCore::RenderThemeSafari::paintMenuList): (WebCore::RenderThemeSafari::paintSliderThumb): (WebCore::RenderThemeSafari::paintSearchField): (WebCore::RenderThemeSafari::paintSearchFieldCancelButton): (WebCore::RenderThemeSafari::paintSearchFieldResultsDecoration): (WebCore::RenderThemeSafari::paintSearchFieldResultsButton):
11:20 PM Changeset in webkit [27472] by aroben
  • 2 edits in trunk/JavaScriptCore

Windows build fix

  • kjs/list.h:
11:09 PM Changeset in webkit [27471] by mrowe@apple.com
  • 2 edits in trunk/JavaScriptCore

2007-11-05 Mark Rowe <​mrowe@apple.com>

Build fix. Add missing #include.

  • kjs/operations.cpp:
10:48 PM Changeset in webkit [27470] by eseidel
  • 2 edits in trunk/JavaScriptCore

2007-11-05 Eric Seidel <​eric@webkit.org>

Reviewed by mjs.


Remove another call to toString(exec)


SunSpider claims this is a 0.5% speedup.

  • kjs/operations.cpp: (KJS::equal): remove another toString
10:39 PM Changeset in webkit [27469] by kevino
  • 1 edit
    6 adds in trunk/WebCore

Adding the rest of wx port's graphics impl.

Rubber-stamped by Maciej Stachowiak.

10:24 PM Changeset in webkit [27468] by eseidel
  • 2 edits in trunk/JavaScriptCore

2007-11-05 Eric Seidel <​eric@webkit.org>

  • kjs/operations.cpp: (KJS::equal): correct broken change.
10:17 PM Changeset in webkit [27467] by eseidel
  • 2 edits in trunk/JavaScriptCore

2007-11-05 Eric Seidel <​eric@webkit.org>

Reviewed by mjs.

Remove one more call to toString(exec).


SunSpider claims this is a 0.7% speedup.

  • kjs/operations.cpp: (KJS::equal): remove a call to toString()
9:52 PM Changeset in webkit [27466] by kevino
  • 11 edits
    4 adds in trunk/WebCore

wx port bug fixes and new files to get platform/wx building on trunk.

Rubber-stamped by Maciej Stachowiak.

9:43 PM Changeset in webkit [27465] by adele
  • 12 edits in trunk/WebCore

Reviewed by Maciej.

Fix for <rdar://problem/5579999> Add poster attribute for video element

  • html/HTMLAttributeNames.in: Added poster attribute.
  • html/HTMLImageLoader.cpp: (WebCore::HTMLImageLoader::updateFromElement): Calls imageSourceAttributeName instead of having special cases for the different kinds of elements.
  • dom/Element.cpp: (WebCore::Element::imageSourceAttributeName): Added.
  • dom/Element.h:
  • html/HTMLObjectElement.cpp: (WebCore::HTMLObjectElement::imageSourceAttributeName): Added.
  • html/HTMLObjectElement.h:
  • html/HTMLMediaElement.cpp: (WebCore::HTMLMediaElement::movieNetworkStateChanged): Calls updatePosterImage when the network state is empty and when the first frame has been loaded.
  • html/HTMLMediaElement.h: (WebCore::HTMLMediaElement::rendererIsNeeded): Made this inline. HTMLVideoElement now also implements this.
  • html/HTMLVideoElement.idl: Added case for poster attribute.
  • html/HTMLVideoElement.h: Added image loader and flag to keep track of whether or not the poster image should be shown.
  • html/HTMLVideoElement.cpp: (WebCore::HTMLVideoElement::HTMLVideoElement): Initialize m_imageLoader and m_shouldShowPosterImage. (WebCore::HTMLVideoElement::rendererIsNeeded): Calls HTMLElement::rendererIsNeeded since HTMLMediaElements have renderer by default. (WebCore::HTMLVideoElement::createRenderer): Create a RenderImage or RenderVideo depending on whether or not the poster image should be shown. (WebCore::HTMLVideoElement::attach): Set up image loader and RenderImage if necessary. (WebCore::HTMLVideoElement::detach): Delete image loader if its no longer needed. (WebCore::HTMLVideoElement::parseMappedAttribute): Added case to process poster attribute. (WebCore::HTMLVideoElement::poster): Added getter for poster attribute. (WebCore::HTMLVideoElement::setPoster): Added setter for poster attribute. (WebCore::HTMLVideoElement::isURLAttribute): Added. (WebCore::HTMLVideoElement::imageSourceAttributeName): Added. (WebCore::HTMLVideoElement::updatePosterImage): Added. Updates m_shouldShowPosterImage and if its changed, detaches and attaches so the renderer is correct.
9:39 PM JavaScript performance improvement ideas edited by mjs@apple.com
(diff)
9:25 PM Changeset in webkit [27464] by aroben
  • 5 edits in trunk

Fix <rdar://5563572> SVG image support is turned off

WebCore:

Fix <rdar://5563572> SVG image support is turned off

Turned on SVG images for all platforms.

Reviewed by Eric.

  • WebCore.vcproj/WebCore.vcproj: Added SVGImage.{cpp,h}
  • loader/CachedImage.cpp: (WebCore::CachedImage::createImage): Removed platform checks for SVGImage.

LayoutTests:

Remove a now-passing test from the Windows Skipped file

Reviewed by Eric.

  • platform/win/Skipped:
8:55 PM Changeset in webkit [27463] by eseidel
  • 2 edits in trunk/WebCore

Fix spelling errors in ChangeLogs, no review

8:43 PM Changeset in webkit [27462] by antti
  • 2 edits in trunk/WebCore

Reviewed by Darin.

QTMovieView can generate callbacks during paint. This can lead to crashes.


Delay callbacks so they get handled after painting is completed. No test case,
I don't know how to reliably reproduce this.

  • platform/graphics/mac/MoviePrivateQTKit.mm: (WebCore::MoviePrivate::MoviePrivate): (WebCore::MoviePrivate::~MoviePrivate): (WebCore::MoviePrivate::paint): (-[WebCoreMovieObserver initWithCallback:WebCore::]): (-[WebCoreMovieObserver disconnect]): (-[WebCoreMovieObserver loadStateChanged:]): (-[WebCoreMovieObserver rateChanged:]): (-[WebCoreMovieObserver sizeChanged:]): (-[WebCoreMovieObserver timeChanged:]): (-[WebCoreMovieObserver volumeChanged:]): (-[WebCoreMovieObserver didEnd:]): (-[WebCoreMovieObserver setDelayCallbacks:]):
8:12 PM Changeset in webkit [27461] by antti
  • 4 edits in trunk/WebCore

Reviewed by Adam.

Add some missing WebCore* prefixes to ObjC classes

  • platform/mac/FileChooserMac.mm: (WebCore::FileChooser::FileChooser):
  • platform/mac/SharedBufferMac.mm: (WebCore::SharedBuffer::createNSData): (WebCore::SharedBuffer::createCFData):
  • platform/mac/SharedTimerMac.cpp: (WebCore::setSharedTimerFireTime):
6:53 PM Changeset in webkit [27460] by alp
  • 3 edits in trunk/WebKit/gtk

2007-11-05 Christian Dywan <​christian@twotoasts.de>

Reviewed by Maciej.

​http://bugs.webkit.org/show_bug.cgi?id=15409
FrameLoaderClientGtk hardcodes data, including platform to Linux i686

Compute a proper user agent string.

Patch includes fixes by Alp.

  • WebCoreSupport/FrameLoaderClientGtk.cpp: (WebKit::agentPlatform): (WebKit::agentOS): (WebKit::composeUserAgent): (WebKit::FrameLoaderClient::userAgent):
  • WebCoreSupport/FrameLoaderClientGtk.h:
6:10 PM Changeset in webkit [27459] by adachan
  • 6 edits in trunk

<rdar://problem/5579772> Regression: AltGr does not work
Testing whether alt-key is down is not the right test for system key event.
Added a m_isSystemKey flag in PlatformKeyboardEvent to keep track of whether
this is a system key event, check that flag instead in handleEditingKeyboardEvent().

Reviewed by Oliver.


4:35 PM Changeset in webkit [27458] by mrowe@apple.com
  • 2 edits in trunk/JavaScriptCore

2007-11-05 Mark Rowe <​mrowe@apple.com>

Gtk build fix.

  • pcre/pcre.pri:
4:02 PM Changeset in webkit [27457] by adachan
  • 2 edits in trunk/WebKit/win

Build fix.

Reviewed by Tristan.

  • WebChromeClient.h:
3:56 PM Changeset in webkit [27456] by antti
  • 3 edits in trunk/WebKitLibraries

Reviewed by Sam.


Update WKQTMovieViewSetDrawSynchronously.

  • libWebKitSystemInterfaceLeopard.a:
  • libWebKitSystemInterfaceTiger.a:
3:21 PM Changeset in webkit [27455] by sullivan@apple.com
  • 2 edits in trunk/WebKit
  • WebView/WebView.mm: (-[WebView _searchWithSpotlightFromMenu:]):

Teeny style tweak to test svn access on other machine

3:12 PM Changeset in webkit [27454] by sullivan@apple.com
  • 2 edits in trunk/WebKit
  • WebView/WebView.mm: (-[WebView computedStyleForElement:pseudoElement:]):


Teeny style tweak to test svn access

3:05 PM Changeset in webkit [27453] by mrowe@apple.com
  • 2 edits in trunk/JavaScriptCore

2007-11-05 Mark Rowe <​mrowe@apple.com>

Gtk build fix.

  • kjs/list.cpp:
2:43 PM Changeset in webkit [27452] by tristan
  • 23 edits in trunk

qt:

Reviewed by Darin Adler.

  • WebCoreSupport/ChromeClientQt.cpp: (WebCore::ChromeClientQt::createWindow):
  • WebCoreSupport/ChromeClientQt.h: Revised to use new WebCore ChromeClient createWindow API.

WebCore:

Reviewed by Darin Adler.

  • WebCore.xcodeproj/project.pbxproj: Set WindowFeatures.h as a Private header.


  • loader/FrameLoader.cpp: (WebCore::FrameLoader::createWindow): Revised to use a single createWindow function instead of two createWindow functions and one createModalDialog function. The logic for this is now addressed in WebKit in an effort to make this easier to follow.


  • page/Chrome.cpp: (WebCore::Chrome::createWindow):
  • page/Chrome.h:
  • page/ChromeClient.h:
  • page/ContextMenuController.cpp: (WebCore::openNewWindow):
  • platform/graphics/svg/SVGImageEmptyClients.h: (WebCore::SVGEmptyChromeClient::createWindow): Revised to take new additional windowFeatures parameter.

WebKit:

Reviewed by Darin Adler.

  • DefaultDelegates/WebDefaultUIDelegate.m: (-[WebDefaultUIDelegate webView:createWebViewWithRequest:windowFeatures:]): Forward the UI delegate to call webView:createWebViewWithRequest: if this method doesn't exist.


  • WebCoreSupport/WebFrameLoaderClient.mm: (WebFrameLoaderClient::dispatchCreatePage):
  • Plugins/WebBaseNetscapePluginView.mm: (-[WebBaseNetscapePluginView loadPluginRequest:]):
  • WebView/WebView.mm: (-[WebView _openNewWindowWithRequest:]): Revised to use new webView:createWebViewWithRequest:windowFeatures: callback.


  • WebCoreSupport/WebChromeClient.h:
  • WebCoreSupport/WebChromeClient.mm: (WebChromeClient::createWindow): Added a new createWindow that accepts 3 parameters, so we can pass up windowFeatures to the chrome. Removed createModalDialog to use new createWindow function.


  • WebView/WebUIDelegatePrivate.h: Added new webView:createWebViewWithRequest:windowFeatures: method.

win:

Reviewed by Darin Adler.


Part of the WebKit/WebCore API changes for
<rdar://problem/5368188>

  • WebChromeClient.cpp: (WebChromeClient::createWindow): Removed usage of createModalDialog and revised to use new createWindow ChromeClient API.
2:02 PM Changeset in webkit [27451] by ggaren@apple.com
  • 2 edits in trunk/JavaScriptCore

Touched a file to test my new HTTP access.

  • kjs/scope_chain.cpp:
1:56 PM Changeset in webkit [27450] by ggaren
  • 1 edit
    1 add in trunk/JavaScriptGlue

Build fix.

  • ForwardingHeaders/kjs/value.h: Added.
1:35 PM Changeset in webkit [27449] by alp
  • 2 edits in trunk/JavaScriptCore

2007-11-05 Alp Toker <​alp@atoker.com>

Unreviewed build fix for qmake-based ports.

Someone with a better understanding of qmake still needs to sort out
the INCLUDEPATH/DEPENDPATH mess.

1:27 PM Changeset in webkit [27448] by ggaren
  • 15 edits
    1 add in trunk

JavaScriptCore:

Reviewed by Darin Adler.


​http://bugs.webkit.org/show_bug.cgi?id=15835

Switched List implementation from a custom heap allocator to an inline
Vector, for a disappointing .5% SunSpider speedup.


Also renamed List::slice to List::getSlice because "get" is the
conventional prefix for functions returning a value through an out
parameter.

  • kjs/array_object.cpp: (KJS::ArrayProtoFunc::callAsFunction): Removed some redundant function calls and memory accesses.
  • kjs/bool_object.cpp: (BooleanObjectImp::construct): Removed questionable use of iterator.
  • kjs/list.cpp:
  • kjs/list.h: New List class, implemented in terms of Vector. Two interesting differences:
    1. The inline capacity is 8, not 5. Many of the Lists constructed during a SunSpider run are larger than 5; almost none are larger than 8.
  1. The growth factor is 4, not 2. Since we can guarantee that Lists aren't long-lived, we can grow them more aggressively, to avoid excessive copying.
  • kjs/regexp_object.cpp: (RegExpObjectImp::construct): Removed redundant function calls.
  • kjs/string_object.cpp: (KJS::StringObjectImp::construct): Removed questionable use of iterator.
  • wtf/Vector.h: (WTF::::uncheckedAppend): Added a fast, unchecked version of append.

WebCore:

Reviewed by Darin Adler.


​http://bugs.webkit.org/show_bug.cgi?id=15835

Small adaptations to new KJS::List class.

  • bindings/js/kjs_window.cpp: (KJS::WindowFunc::callAsFunction): (KJS::ScheduledAction::ScheduledAction):

WebKit:

Reviewed by Darin Adler.


​http://bugs.webkit.org/show_bug.cgi?id=15835

Small adaptations to new KJS::List class.

  • ForwardingHeaders/kjs/value.h: Added.
1:09 PM Changeset in webkit [27447] by mrowe@apple.com
  • 3 edits in trunk/JavaScriptCore

2007-11-05 Mark Rowe <​mrowe@apple.com>

Reviewed by Alp Toker.

Add DEPENDPATH to JavaScriptCore and pcre to help qmake with dependencies.

10:57 AM Changeset in webkit [27446] by aroben
  • 3 edits in trunk/WebKitTools

Add support on Windows for WEBKIT_TESTFONTS

This environment variable lets you specify where the fonts to be used
by DumpRenderTree reside. The Qt port is already using this, so I'm
just following their lead.

Reviewed by Darin.

  • DumpRenderTree/win/DumpRenderTree.cpp: (exePath): Refactored code out of initialize(). (fontsPath): Returns either $WEBKIT_TESTFONTS or DumpRenderTree.resources. (initialize): Use the new fontsPath function. (main): Use the new exePath function.
  • Scripts/run-webkit-tests: Propagate the WEBKIT_TESTFONTS environment variable to DRT, like Qt does.
10:57 AM Changeset in webkit [27445] by aroben
  • 4 edits in trunk/WebKit/win

Add IWebCache::disabled

This matches the Mac WebCache class.

Reviewed by Sam.

  • Interfaces/IWebCache.idl:
  • WebCache.cpp: (WebCache::disabled):
  • WebCache.h:
10:57 AM Changeset in webkit [27444] by aroben
  • 9 edits
    3 adds in trunk

Add IWebInspector and a way to get one from a WebView

WebCore:

Allow passing a base class pointer to COMPtr::copyRefTo

Reviewed by Sam.

  • platform/win/COMPtr.h:

WebKit/win:

Add IWebInspector and a way to get one from a WebView

This API matches the Mac one added in r27266.

Reviewed by Sam.

  • Interfaces/IWebInspector.idl: Added.
  • Interfaces/IWebViewPrivate.idl: Added a new inspector method to match Mac.
  • WebInspector.cpp: Added. (WebInspector::createInstance): (WebInspector::WebInspector): (WebInspector::~WebInspector): (WebInspector::webViewClosed): (WebInspector::QueryInterface): (WebInspector::AddRef): (WebInspector::Release): (WebInspector::show): (WebInspector::showConsole): (WebInspector::showTimeline): (WebInspector::close): (WebInspector::attach): (WebInspector::detach):
  • WebInspector.h: Added.
  • WebKit.vcproj/Interfaces.vcproj: Added IWebInspector.idl.
  • WebKit.vcproj/WebKit.vcproj: Added WebInspector.{cpp,h}.
  • WebKit.vcproj/WebKitGUID.vcproj: Added IWebInspector_i.c.
  • WebView.cpp: (WebView::close): Notify the WebInspector that we're closing. (WebView::inspector): Added.
  • WebView.h:
10:56 AM Changeset in webkit [27443] by aroben
  • 2 edits in trunk/WebKit/win

Cache strings that come directly from WebLocalizedString

Reviewed by Steve.

  • WebLocalizableStrings.cpp: (copyLocalizedStringFromBundle): This used to be WebLocalizedString, but is now just a static helper function. (findCachedStringInMap): Added static helper. (findCachedString): Refactored and cleaned up code from WebLocalizedLPCTSTR. (cacheString): Ditto. (WebLocalizedString): Rewrote to use the new helper functions. (WebLocalizedLPCTSTR): Changed to use the new helper functions.
10:55 AM Changeset in webkit [27442] by aroben
  • 2 edits in trunk/WebKit/win

Only override the default user agent string if we're actually given a custom one

Reviewed by Sam.

  • WebView.cpp: (WebView::setCustomUserAgent):
10:52 AM Changeset in webkit [27441] by mitz@apple.com
  • 3 edits
    2 adds in trunk

WebCore:

Reviewed by Oliver Hunt.

  • fix ASSERTION FAILED: !HashTranslator::equal(KeyTraits::emptyValue(), key) when a class attribute is all-whitespace

Test: fast/dom/class-all-whitespace.html

  • dom/StyledElement.cpp: (WebCore::StyledElement::parseMappedAttribute): Check if there is any non-whitespace character in the class attribute.

LayoutTests:

Reviewed by Oliver Hunt.

  • test that an all-whitespace class attribute does not cause an assertion failure
  • fast/dom/class-all-whitespace-expected.txt: Added.
  • fast/dom/class-all-whitespace.html: Added.
10:23 AM Changeset in webkit [27440] by beidson@apple.com
  • 3 edits in trunk/WebCore

Reviewed by Sam

Add transaction blocking to the DatabaseAuthorizer in preparation for the new version
of the sql storage API

  • storage/DatabaseAuthorizer.cpp: (WebCore::DatabaseAuthorizer::allowTransaction):
  • storage/DatabaseAuthorizer.h:
8:34 AM Changeset in webkit [27439] by mrowe@apple.com
  • 2 edits in trunk/WebCore

2007-11-05 Mark Rowe <​mrowe@apple.com>

Reviewed by Alp Toker.

Have getMIMETypeForExtension return a null string when no MIME type is known
for the extension rather than returning "text/plain". This prevents plugin data
being dumped into object elements when plugins are disabled.

  • platform/gtk/MIMETypeRegistryGtk.cpp: (WebCore::MIMETypeRegistry::getMIMETypeForExtension):
7:01 AM Changeset in webkit [27438] by mrowe@apple.com
  • 2 edits in trunk/WebKitTools

2007-11-05 Mark Rowe <​mrowe@apple.com>

Rubber-stamped by Alp Toker.

Remove buggy "autocomplete" from GtkLauncher as it causes more problems than it solves.

  • GtkLauncher/main.cpp: (goToURLBarText): (main):
6:39 AM Changeset in webkit [27437] by mrowe@apple.com
  • 3 edits in trunk/WebKit/gtk

2007-11-05 Mark Rowe <​mrowe@apple.com>

Rubber-stamped by Alp Toker.

Remove unused m_firstData member from FrameLoaderClientGtk.

  • WebCoreSupport/FrameLoaderClientGtk.cpp:
  • WebCoreSupport/FrameLoaderClientGtk.h:
6:39 AM Changeset in webkit [27436] by mrowe@apple.com
  • 2 edits in trunk/WebKit/gtk

2007-11-05 Mark Rowe <​mrowe@apple.com>

Reviewed by Alp Toker.

Fix ​http://bugs.webkit.org/show_bug.cgi?id=15842
Bug 15842: [Gtk] about:blank doesn't work

  • WebCoreSupport/FrameLoaderClientGtk.cpp: (WebKit::FrameLoaderClient::finishedLoading): Set the encoding on the frame loader to get work done that is normally done when the first bit of data is received, even in the case of a document with no data (like about:blank).
6:30 AM Changeset in webkit [27435] by ap
  • 2 edits in trunk/LayoutTests

Reviewed by Darin.

​http://bugs.webkit.org/show_bug.cgi?id=15841
fast/dom/Document/early-document-access.html crashes under GuardMalloc

  • fast/dom/Document/early-document-access.html: Don't check for open() timeout; wait for popup window to close as long as it takes.
Note: See TracTimeline for information about the timeline view.