18:36 Changeset [29203] by weinig@apple.com


Reviewed by Eric Seidel.

Patch for http://bugs.webkit.org/show_bug.cgi?id=16758
ASSERT when using TreeWalker methods for a current node outside of the root (Acid3)

  • Ensure that returned nodes are within the root node, or return 0, in adherence with the spec.

Test: fast/dom/TreeWalker/TreeWalker-currentNode.html

  • dom/TreeWalker.cpp: (WebCore::TreeWalker::parentNode): (WebCore::TreeWalker::firstChild): (WebCore::TreeWalker::lastChild): (WebCore::TreeWalker::previousSibling): (WebCore::TreeWalker::nextSibling): (WebCore::TreeWalker::previousNode): (WebCore::TreeWalker::nextNode):


Reviewed by Eric Seidel.

Test for http://bugs.webkit.org/show_bug.cgi?id=16758
ASSERT when using TreeWalker methods for a current node outside of the root (Acid3)

  • fast/dom/TreeWalker: Added.
  • fast/dom/TreeWalker/TreeWalker-currentNode.html: Added.
  • fast/dom/TreeWalker/resources: Added.
  • fast/dom/TreeWalker/resources/TreeWalker-currentNode.js: Added.
18:24 Changeset [29202] by darin@apple.com
  • Scripts/do-webcore-rename: Some more renaming plans.
18:16 Changeset [29201] by darin@apple.com
  • API/JSRetainPtr.h: One more file that needed the change below.
18:14 Changeset [29200] by darin@apple.com
  • wtf/OwnPtr.h: OwnPtr needs the same fix as RefPtr below.
18:12 Changeset [29199] by aroben@apple.com

Build fix.

Reviewed by Maciej.

  • wtf/RetainPtr.h: Use PtrType instead of T* because of the RemovePointer magic.
14:46 Changeset [29198] by darin@apple.com

Rubber stamped by Maciej Stachowiak.

  • cut down own PIC branches by using a pointer-to-member-data instead of a pointer-to-member-function in WTF smart pointers
  • wtf/OwnArrayPtr.h:
  • wtf/OwnPtr.h:
  • wtf/PassRefPtr.h:
  • wtf/RefPtr.h:
  • wtf/RetainPtr.h: Use a pointer to the m_ptr member instead of the get member. The GCC compiler generates better code for this idiom.
03:44 Changeset [29197] by alp@webkit.org

2008-01-05 Alp Toker <alp@atoker.com>

Rubber-stamped by Mark Rowe.

Fix configure script output with correct descriptions of the default
configuration options.

02:23 Changeset [29196] by alp@webkit.org

2008-01-05 Sylvain Pasche <sylvain.pasche@gmail.com>

Reviewed by Alp Toker.

Allow autogen.sh invocation from a separate build directory.

  • autogen.sh:
00:26 Changeset [29195] by mrowe@apple.com

2008-01-05 Henry Mason <hmason@mac.com>

Reviewed by Maciej Stachowiak.

Bug 16738: Collector block offset could be stored as an cell offset instead of a byte offset

Gives a 0.4% SunSpider boost and prettier code.

  • kjs/collector.cpp: Switched to cell offsets from byte offsets (KJS::Collector::heapAllocate): (KJS::Collector::sweep):
00:02 Changeset [29194] by mrowe@apple.com

Have the two malloc zones print useful diagnostics if their free method are unexpectedly invoked.
Due to <rdar://problem/5671357> this can happen if an application attempts to free a pointer that
was not allocated by any registered malloc zone on the system.

Reviewed by Maciej Stachowiak.


21:45 Changeset [29193] by mitz@apple.com
  • update the list of tests failing on Windows because of font differences and uncomment a test that has been re-enabled on Mac.
  • platform/win/Skipped:
21:20 Changeset [29192] by oliver@apple.com

Fix bounds computation bugs responsible for http://bugs.webkit.org/show_bug.cgi?id=16015
and other image repaint bugs.

Reviewed by Beth Dakin.

We now cache the full local bounds for the <image> element, as otherwise certain
combinations of attribute changes could result in incorrect dirty rects.
Additionally we no longer use any of the integer bounds fields on RenderObject for
determining repaint bounds (this was the principle cause of bug #16015).

I also removed the outline painting code as it was both wrong, and not correctly
repainted. I feel safe doing this as no other browser or viewer supports outline
properties on svg elements.

I was unable to make a testcase for this unfortunately, despite seemingly deterministic
behaviour :(

21:13 Changeset [29191] by bdakin@apple.com


Reviewed by Oliver.

Fix for http://bugs.webkit.org/show_bug.cgi?id=16704 input with
type="hidden" matches :enabled/:disabled (Acid3 bug)

Prevent :enabled and :disabled from applying to input type="hidden"

  • css/CSSStyleSelector.cpp: (WebCore::CSSStyleSelector::checkOneSelector): Rather than allowing :enabled and :disabled to apply to all controls, only allow it to apply to non-"hidden" controls
  • dom/Element.h: (WebCore::Element::isInputTypeHidden):
  • html/HTMLInputElement.h: (WebCore::HTMLInputElement::isInputTypeHidden):


Reviewed by Oliver.

Test for http://bugs.webkit.org/show_bug.cgi?id=16704 input with
type="hidden" matches :enabled/:disabled (Acid3 bug)

  • fast/forms/hidden-input-not-enabled-expected.txt: Added.
  • fast/forms/hidden-input-not-enabled.html: Added.
20:05 Changeset [29190] by weinig@apple.com

Add missing expected result.

  • fast/images/border-expected.txt: Added.
20:04 Changeset [29189] by weinig@apple.com


Reviewed by Oliver Hunt.

  • Match the spec when calling getFloatValue, getStringValue, getCounterValue, getRectValue and getRGBColorValue of CSSPrimitiveValue by throwing exceptions if the type of the CSSPrimitiveValue is not the same as the type requested.
  • Fix the intermittent assertion failure seen in svg/css/glyph-orientation-rounding-test.xhtml

Test: fast/css/CSSPrimitiveValue-exceptions.html

  • css/CSSPrimitiveValue.cpp: (WebCore::CSSPrimitiveValue::getDoubleValue): (WebCore::CSSPrimitiveValue::getStringValue): (WebCore::CSSPrimitiveValue::getCounterValue): (WebCore::CSSPrimitiveValue::getRectValue): (WebCore::CSSPrimitiveValue::getRGBColorValue): (WebCore::CSSPrimitiveValue::getPairValue):
  • css/CSSPrimitiveValue.h: (WebCore::CSSPrimitiveValue::getFloatValue): (WebCore::CSSPrimitiveValue::getIntValue): (WebCore::CSSPrimitiveValue::getCounterValue): (WebCore::CSSPrimitiveValue::getRectValue): (WebCore::CSSPrimitiveValue::getDashboardRegionValue):
  • css/CSSPrimitiveValue.idl:


Reviewed by Oliver Hunt.

Throw exceptions for the CSSPrimitiveValue getters, matching the spec.

  • fast/css/CSSPrimitiveValue-exceptions-expected.txt: Added.
  • fast/css/CSSPrimitiveValue-exceptions.html: Added.
  • fast/css/resources/CSSPrimitiveValue-exceptions.js: Added.
  • fast/dom/prototypes.html:
  • fast/dom/wrapper-classes.html:
  • platform/mac/fast/dom/wrapper-classes-objc.html:
  • svg/css/glyph-orientation-rounding-test-expected.txt:
  • svg/css/glyph-orientation-rounding-test.xhtml: Pass in a unit so this doesn't throw. This patch also fixes the associated intermittent failure seen in this test.
18:32 Changeset [29188] by antti@apple.com

Reviewed by Darin.

Windows part of <rdar://problem/5647034>
Media tests crash if an old version of QuickTime is installed

Check QuickTime version on Windows too.

17:51 Changeset [29187] by weinig@apple.com

Reviewed by Kevin McCullough.

  • Splits Objective-C tests out of wrapper-classes.html and into platform specific test.
  • Adds tests for .proto and .constructor to JS wrapper tests.
  • fast/dom/wrapper-classes.html:
  • platform/mac/fast/dom/wrapper-classes-expected.txt:
  • platform/mac/fast/dom/wrapper-classes-objc-expected.txt: Added.
  • platform/mac/fast/dom/wrapper-classes-objc.html: Copied from fast/dom/wrapper-classes.html.
  • platform/win/Skipped:
17:45 Changeset [29186] by slewis@apple.com

Move svg/batik/text/smallFonts.svg
to skip list since it is still failing on tiger.

  • platform/mac-tiger/Skipped:
17:11 Changeset [29185] by darin@apple.com


Reviewed by Maciej and Alice.

  • fix <rdar://problem/4404302> Borders where there should be none (canadasmountains.com)

Test: fast/images/border.html

  • html/HTMLImageElement.cpp: (WebCore::HTMLImageElement::parseMappedAttribute): Change the code that handles cases where the border value is not a number to use the value "0" for the border width instead of not setting the border width and style at all. This matches other browsers, and makes more sense.


Reviewed by Maciej and Alice.

  • test for <rdar://problem/4404302> Borders where there should be none (canadasmountains.com)
  • fast/images/border.html: Added.
  • fast/images/resources/TEMPLATE.html: Copied from fast/js/resources/TEMPLATE.html.
  • fast/images/resources/border.js: Added.
16:32 Changeset [29184] by alice.liu@apple.com


Reviewed by Alexey and Darin.

Adding "home" and "end" to DRT's keydown since
DOM key events can't handle keyIdentifiers at this point.

  • DumpRenderTree/mac/EventSendingController.mm: (-[EventSendingController keyDown:withModifiers:]):
  • DumpRenderTree/win/EventSender.cpp: (keyDownCallback):


Reviewed by Alexey and Darin.

  • editing/selection/move-begin-end.html: add a codepath that will test the same kind of functionality for home and end keys
  • platform/win/Skipped: removing section of tests that have been resolved
16:28 Changeset [29183] by antti@apple.com

Try to fix 64-bit build

  • platform/graphics/mac/MediaPlayerPrivateQTKit.mm: (WebCore::MediaPlayerPrivate::isAvailable):
16:15 Changeset [29182] by slewis@apple.com

Fix another stringImpl leak.

  • platform/text/StringImpl.cpp: remove an extra allocation (WebCore::StringImpl::StringImpl):
15:34 Changeset [29181] by alice.liu@apple.com

Reviewed by Mitz.

  • platform/graphics/win/FontCacheWin.cpp: fix loop condition that was causing crash
14:56 Changeset [29180] by oliver@apple.com

Update tiger layout test results for svg/batik/text/smallFonts, and move to mac-tiger platform


14:43 Changeset [29179] by timothy@apple.com

Reviewed by Adam Roben.

<rdar://problem/5671059> Always show the Timeline and Console buttons in the Inspector

  • page/inspector/inspector.css: Remove styles related to the toggle button and make the area always visible.
  • page/inspector/inspector.html: Remove the toggle button.
  • page/inspector/inspector.js: Remove code to toggle the status area.
14:35 Changeset [29178] by ap@webkit.org

Reviewed by Adam Roben.

Restore compatibility with shipping Safari beta.


REGRESSION: Crash on Safari Start

  • Interfaces/IWebViewPrivate.idl: Should have added a new method at the end.


Temporarily restore BackwardDelete for Windows nightlies to work correctly.

  • editing/EditorCommand.cpp: (WebCore::CommandEntry::):
14:27 Changeset [29177] by ap@webkit.org

Reviewed by Darin.

<rdar://problem/5611712> xsl:sort does not use a case folding sort, and the 'case-order' attribute is ignored (16077)

Enable the fix on Windows.

14:08 Changeset [29176] by antti@apple.com

Fix build.

  • bindings/js/JSHTMLElementWrapperFactory.cpp:
13:44 Changeset [29175] by antti@apple.com

Reviewed by Darin.

Partial fix for <rdar://problem/5647034>
Media tests crash if an old version of QuickTime is installed

Disable media support if QuickTime is not current enough (>=7.3).

Windows patch coming soon.

  • bindings/js/JSHTMLElementWrapperFactory.cpp: (WebCore::createJSHTMLWrapper):
  • bindings/js/kjs_window.cpp: (KJS::Window::getValueProperty):
  • html/HTMLElementFactory.cpp: (WebCore::audioConstructor): (WebCore::videoConstructor): (WebCore::sourceConstructor):
  • platform/graphics/MediaPlayer.cpp: (WebCore::MediaPlayer::isAvailable):
  • platform/graphics/MediaPlayer.h:
  • platform/graphics/mac/MediaPlayerPrivateQTKit.h:
  • platform/graphics/mac/MediaPlayerPrivateQTKit.mm: (WebCore::MediaPlayerPrivate::isAvailable):
13:02 Changeset [29174] by timothy@apple.com

Reviewed by Darin Adler.

<rdar://problem/5604409> JavaScript privilege escalation when Web Inspector accesses page unsafely (16011)

Check if the property is a getter before asking for the value.
If the property is a getter, we no longer show the value.

  • page/inspector/PropertiesSidebarPane.js:
  • page/inspector/inspector.css:
12:54 Changeset [29173] by alp@webkit.org

2008-01-04 Alp Toker <alp@atoker.com>

Rubber-stamped by Mark Rowe.

Remove AM_MAINTAINER_MODE, as recommended by the automake manual. The
"rebuild rules" will now be enabled by default.

This obviates the need to pass '--enable-maintainer-mode' to

  • configure.ac:
11:31 Changeset [29172] by aroben@apple.com

Fix <rdar://4968855> Shift-PageUp/Shift-PageDown only select one line of text


Fix <rdar://4968855> Shift-PageUp/Shift-PageDown only select one line of text

Test: platform/win/editing/selection/shift-page-down-up.html

Reviewed by Oliver.

  • WebView.cpp: Correct the commands associated with Shift-PageUp/Shift-PageDown.


Added a test for <rdar://4968855> Shift-PageUp/Shift-PageDown only select one line of text

Reviewed by Oliver.

  • platform/win/editing/selection/shift-page-up-down-expected.txt: Added.
  • platform/win/editing/selection/shift-page-up-down.html: Added.
11:25 Changeset [29171] by oliver@apple.com

<rdar://problem/5646437> REGRESSION (r28015): svg/batik/text/smallFonts fails

Reviewed by Darin

Regenerated expected output as filter elements are no longer generated in a default
build. This wasn't done when filters were turned off earlier due to other (now gone)
layout failures i didn't want to put in the expected output.

11:23 Changeset [29170] by mitz@apple.com

Reviewed by Darin Adler.

  • fix synthetic bold and italic on Windows

Covered by an existing test.

  • platform/graphics/win/FontPlatformDataWin.cpp: (WebCore::FontPlatformData::FontPlatformData):
09:18 Changeset [29169] by aroben@apple.com
  • platform/win/Skipped: Added some more failures.
06:52 Changeset [29168] by alp@webkit.org

2008-01-04 Alp Toker <alp@atoker.com>

GTK+ autotools build fix. Terminate empty rules.

06:49 Changeset [29167] by hausmann@webkit.org

fix DRT after the changes to QWebPage.

The great thing is that we actually don't have
any regressions from the QWebPage change :)

Signed-off-by: Simon

06:49 Changeset [29166] by hausmann@webkit.org

Remove most dependencies of Widget/ScrollView onto native QWidgets.

This also brings the code closer in line with the Windows code. Seems
to work nicely on first try :)

Signed-off-by: Simon

06:48 Changeset [29165] by hausmann@webkit.org
  • Remove the todo from QWebPage and move the code to QWebView.

Signed-off-by: Simon

06:48 Changeset [29164] by hausmann@webkit.org
  • Move the various event methods to QWebPagePrivate. This is similar to QTextControl as well.

Signed-off-by: Simon

06:48 Changeset [29163] by hausmann@webkit.org
  • Forward the event from QWebView to QWebPage through QObject::event. This is similar to the way QTextControl is working.

Signed-off-by: Simon

06:47 Changeset [29162] by hausmann@webkit.org

make QWebPage a QObject and get things to compile.

Nothing works currently though.

Signed-off-by: Simon

04:47 QtWebKitGitInstructions edited by bmeyer@trolltech.com
Joe Doe needs quotes otherwise the name is Joe (diff)
04:26 Changeset [29161] by hausmann@webkit.org

Added some preliminary class documentation for QWebView, fixed a missing const() and some missing Q_PROPERTYs

Signed-off-by: Lars

04:25 Changeset [29160] by hausmann@webkit.org

add a viewportSize to QWebPage.

This is required to eventually make it a QObject only.

Signed-off-by: Simon

04:25 Changeset [29159] by hausmann@webkit.org

take a QString as identifier in QWebFrame::addToJSWindowObject.

Signed-off-by: Simon

04:24 Changeset [29158] by hausmann@webkit.org

Call the frame arguments for the javascript callbacks "originatingFrame"

Signed-off-by: Lars

04:24 Changeset [29157] by hausmann@webkit.org

Moved the QWebPage::addToHistory signal into QWebHistoryInterface

Signed-off-by: Lars

04:23 Changeset [29156] by hausmann@webkit.org

moved title(), url(), icon() and initialLayoutComplete() from QWebPage to QWebFrame

Signed-off-by: Simon

04:19 Changeset [29155] by alp@webkit.org

Fix bug URL in ChangeLog entry

04:18 Changeset [29154] by alp@webkit.org

2008-01-04 Alp Toker <alp@atoker.com>

Reviewed by Mark Rowe.

make -j is failing with the autotools based system

Support parallel code generation. Nearly every use of explicit
multiple targets was a potential concurrency bug, though in practice
the bison rules were the first to be noticed because they took longer
to complete and broke the build immediately.

  • GNUmakefile.am:
02:44 Changeset [29153] by alp@webkit.org

2008-01-04 Alp Toker <alp@atoker.com>

Reviewed by Mark Rowe.

Re-use a single static dummy surface rather than creating and
destroying a surface for each CairoPath.

  • platform/graphics/cairo/CairoPath.h: (WebCore::CairoPath::CairoPath):
00:16 Changeset [29152] by mrowe@apple.com

Tiger build fix.

  • platform/Threading.h: The OSAtomic functions take non-volatile pointers on Tiger.
00:05 Changeset [29151] by alp@webkit.org

2008-01-04 Alp Toker <alp@atoker.com>

GTK+ DRT build fix for breakage introduced in r29149.

  • DumpRenderTree/gtk/LayoutTestControllerGtk.cpp: (LayoutTestController::execCommand):


23:55 Changeset [29150] by mrowe@apple.com

Use platform-provided atomic operations in place of inline assembly to increase portability.

Reviewed by Maciej Stachowiak.

  • platform/Threading.h:


23:31 Changeset [29149] by ap@webkit.org

Reviewed by Darin.

<rdar://problem/5463489> A number of layout tests should be using execCommand instead of textInputController

Added layoutTestController.execCommand to access editor commands that are not available
via document.execCommand.

23:15 Changeset [29148] by oliver@apple.com

Fix <rdar://problem/5668517> REGRESSION: Major under painting issues in SVG (carto.net dock example)

Revied by Maciej.

We need to cache the absolute bounds of the <image>,
as there's no reliable way to recompute the old bounding
box one we have started layout.

22:34 Changeset [29147] by alp@webkit.org

2008-01-03 Xan Lopez <xan@gnome.org>

Reviewed by Alp Toker.

[GTK] Signal "navigation-requested" does not react correctly on
return TRUE from callbacks

  • WebView/webkitwebview.cpp: use our own accumulator for signals returning WebKitNavigationResponse. The emission will be stopped when any callback returns anything but WEBKIT_NAVIGATION_RESPONSE_ACCEPT.
22:23 Changeset [29146] by alp@webkit.org

2008-01-03 Alp Toker <alp@atoker.com>

Reviewed by Mark Rowe.

Support building in Scratchbox, which has a version of make that fails
on wildcard syntax. Use a vpath to match IDL files instead.

  • GNUmakefile.am:
22:00 Changeset [29145] by darin@apple.com
  • fix Safari build
  • WebKit.xcodeproj/project.pbxproj: Mark header private (fixes Safari build).
21:41 Changeset [29144] by mrowe@apple.com

Update Tiger library to a G3-friendly version.

20:38 Changeset [29143] by mitz@apple.com

Reviewed by Sam Weinig.

  • skip tests that are failing because font fallback results in a different font being used on Windows than on Mac.
  • platform/win/Skipped:
20:22 Changeset [29142] by jhoneycutt@apple.com

2008-01-03 Jon Honeycutt <jhoneycutt@apple.com>

Reviewed by Darin.

<rdar://problem/5504775> PDF page will not load first time after Adobe
Reader install, unless browser is relaunched

Refresh and re-search the plugin database if the MIME type is not

  • plugins/win/PluginDatabaseWin.cpp: (WebCore::PluginDatabaseWin::isMIMETypeRegistered):
  • plugins/win/PluginDatabaseWin.h:
20:03 Changeset [29141] by adele@apple.com

Update libraries for <rdar://problem/4106190> Include "Where from" metadata in drag-and-dropped images

  • WebKitSystemInterface.h:
  • libWebKitSystemInterfaceLeopard.a:
  • libWebKitSystemInterfaceTiger.a:
18:04 Changeset [29140] by mitz@apple.com

Reviewed by Darin Adler.

  • platform/graphics/win/FontCacheWin.cpp: (WebCore::linkedFontEnumProc): Added. This callback is used to fetch a valid LOGFONT for a given family. (WebCore::getLinkedFonts): Added. Returns a vector of font families linked to the given font family by the Windows registry key HKLM\Software\...\FontLink\SystemLink. The registry values typically differ based on the installed language version of Windows. (WebCore::FontCache::getFontDataForCharacters): Changed to not use MLang font mapping, which is Windows code page based, except for characters in the range U+2000..U+200F. Instead, this function gets the font Uniscribe would use for the character. However, that font might not actually contain the character, in which case GDI font linking would substitute a different font. Therefore, this function walks the linked font list until it finds a font that actually contains the character.
17:28 Changeset [29139] by alice.liu@apple.com


Reviewed by Darin, Mitz.

fixed <rdar://5130762> mousedowns in different locations increase the clickcount incorrectly

the clickcount should not increase if the last click was at a different location. it's a new click.

  • DumpRenderTree/mac/DumpRenderTree.mm: (runTest):
  • DumpRenderTree/mac/EventSendingController.h:
  • DumpRenderTree/mac/EventSendingController.mm: (-[EventSendingController mouseDown]): (-[EventSendingController mouseUp]):


Reviewed by Darin, Mitz.

updating expected results after fixing clickcount for
<rdar://5130762> mousedowns in different locations increase the clickcount incorrectly

  • fast/forms/focus-selection-input-expected.txt:
  • fast/forms/focus-selection-textarea-expected.txt:
  • platform/mac/fast/events/objc-event-api-expected.txt:
  • platform/mac/fast/inline/dirtyLinesForInline-expected.txt:
  • platform/mac/fast/text/textIteratorNilRenderer-expected.txt:
  • platform/mac/svg/custom/pointer-events-image-expected.txt:
  • platform/mac/svg/custom/pointer-events-path-expected.txt:
  • platform/mac/svg/custom/pointer-events-text-expected.txt:
  • platform/mac/svg/custom/text-hit-test-expected.txt:
17:17 Changeset [29138] by darin@apple.com


Reviewed by Mitz.

Test: fast/table/border-changes.html

  • html/HTMLTableElement.h:
  • html/HTMLTableElement.cpp: (WebCore::HTMLTableElement::parseMappedAttribute): Check the border type before and after parsing attributes, rather than doing this only for the rules attribute. (WebCore::HTMLTableElement::cellBorders): Added. (WebCore::HTMLTableElement::getSharedCellDecl): Changed to use cellBorders to factor out the rule about what type of borders to use.


Reviewed by Mitz.

  • fast/table/border-changes-expected.txt: Added.
  • fast/table/border-changes.html: Added.
  • fast/table/resources: Added.
  • fast/table/resources/TEMPLATE.html: Added.
  • fast/table/resources/border-changes.js: Added.
17:00 Changeset [29137] by adachan@apple.com

Replaced the use of editor command "ForwardDelete" with "DeleteForward" to be consistent
with our use of "DeleteBackward".

Rubber-stamped by Jon.

  • WebView.cpp:
16:13 Changeset [29136] by darin@apple.com

Reviewed by Adam.

  • Scripts/update-webkit: Make this work a little better for the people at Apple who have a directory named Internal.
14:45 Changeset [29135] by kdecker@apple.com

Reviewed by Darin.

Fixed: <rdar://problem/4106190> Include "Where from" metadata in drag-and-dropped images

  • Misc/WebNSFileManagerExtras.h:
  • Misc/WebNSFileManagerExtras.m: (-[NSFileManager _webkit_setMetadataURL:referrer:atPath:]): Added new method. Uses WebKitSystemInterface to set "Where from:" metadata information.
  • WebView/WebHTMLView.mm: (-[WebHTMLView namesOfPromisedFilesDroppedAtDestination:]): Added "Where from:" metadata for drag and dropped images.
14:10 Changeset [29134] by weinig@apple.com

Reviewed by Darin.

Pass the prototype of WebCore JS objects up the constructor chain
rather than explicitly setting using setPrototype. This removes many
redundant settings of the prototype on construction. To avoid a CG
hazard, the prototype must be constructed before before calling the
constructor of the JS object.

  • JS objects that inherit from DOMObject, which all bindings objects (except Window) do, now can't implicitly have a jsNull prototype, but must explicitly pass it up the construction chain.
  • bindings/js/JSCSSRuleCustom.cpp: (WebCore::toJS):
  • bindings/js/JSCSSValueCustom.cpp: (WebCore::toJS):
  • bindings/js/JSDocumentCustom.cpp: (WebCore::toJS):
  • bindings/js/JSEventCustom.cpp: (WebCore::toJS):
  • bindings/js/JSEventTargetNode.cpp: (WebCore::JSEventTargetNode::JSEventTargetNode):
  • bindings/js/JSEventTargetNode.h:
  • bindings/js/JSHTMLAllCollection.h: (WebCore::JSHTMLAllCollection::JSHTMLAllCollection):
  • bindings/js/JSHTMLAudioElementConstructor.cpp: (WebCore::JSHTMLAudioElementConstructor::JSHTMLAudioElementConstructor):
  • bindings/js/JSHTMLCollectionCustom.cpp: (WebCore::getNamedItems): (WebCore::toJS):
  • bindings/js/JSHTMLElementWrapperFactory.cpp: (WebCore::createJSHTMLWrapper):
  • bindings/js/JSHTMLFormElementCustom.cpp: (WebCore::JSHTMLFormElement::nameGetter):
  • bindings/js/JSHTMLInputElementBase.cpp: (WebCore::JSHTMLInputElementBase::JSHTMLInputElementBase):
  • bindings/js/JSHTMLInputElementBase.h:
  • bindings/js/JSHTMLOptionElementConstructor.cpp: (WebCore::JSHTMLOptionElementConstructor::JSHTMLOptionElementConstructor):
  • bindings/js/JSLocation.cpp: (WebCore::JSLocation::JSLocation):
  • bindings/js/JSLocation.h:
  • bindings/js/JSNamedNodesCollection.cpp: (WebCore::JSNamedNodesCollection::JSNamedNodesCollection):
  • bindings/js/JSNamedNodesCollection.h:
  • bindings/js/JSNodeCustom.cpp: (WebCore::toJS):
  • bindings/js/JSSVGElementWrapperFactory.cpp: (WebCore::createJSSVGWrapper):
  • bindings/js/JSSVGPathSegCustom.cpp: (WebCore::toJS):
  • bindings/js/JSStyleSheetCustom.cpp: (WebCore::toJS):
  • bindings/js/JSXMLHttpRequest.cpp: (KJS::JSXMLHttpRequestConstructorImp::JSXMLHttpRequestConstructorImp): (KJS::JSXMLHttpRequestConstructorImp::construct): (KJS::JSXMLHttpRequest::JSXMLHttpRequest):
  • bindings/js/JSXMLHttpRequest.h:
  • bindings/js/JSXSLTProcessor.cpp: (KJS::JSXSLTProcessor::JSXSLTProcessor): (KJS::XSLTProcessorConstructorImp::XSLTProcessorConstructorImp): (KJS::XSLTProcessorConstructorImp::implementsConstruct): (KJS::XSLTProcessorConstructorImp::construct):
  • bindings/js/JSXSLTProcessor.h:
  • bindings/js/kjs_binding.h: (KJS::DOMObject::DOMObject): (KJS::cacheDOMObject): (KJS::cacheSVGDOMObject):
  • bindings/js/kjs_css.cpp: (WebCore::JSRGBColor::JSRGBColor): (WebCore::getJSRGBColor):
  • bindings/js/kjs_css.h:
  • bindings/js/kjs_events.cpp: (WebCore::JSClipboard::JSClipboard): (WebCore::toJS):
  • bindings/js/kjs_events.h:
  • bindings/js/kjs_html.cpp: (WebCore::ImageConstructorImp::ImageConstructorImp):
  • bindings/js/kjs_navigator.cpp: (KJS::Navigator::Navigator): (KJS::PluginBase::PluginBase):
  • bindings/js/kjs_navigator.h:
  • bindings/js/kjs_window.cpp: (KJS::Window::Window): (KJS::Window::location): (KJS::Window::getValueProperty):
  • bindings/js/kjs_window.h:
  • bindings/scripts/CodeGeneratorJS.pm:
13:40 Changeset [29133] by ap@webkit.org

Reviewed by Darin.

Hixie's XMLHttpRequest/018.html test fails

This was fixed earlier, just adding a test case.

  • http/tests/xmlhttprequest/web-apps/018-expected.txt: Added.
  • http/tests/xmlhttprequest/web-apps/018.html: Added.
  • http/tests/xmlhttprequest/web-apps/resources: Added.
  • http/tests/xmlhttprequest/web-apps/resources/018-test.xml: Added.
13:21 Changeset [29132] by alice.liu@apple.com

Reviewed by Darin.

This fixes pageup/down in iframes. test for this is fast/frames/iframe-scroll-page-up-down.html

  • WebView/WebHTMLView.mm: (-[WebHTMLView doCommandBySelector:]): Have the editor handle all the commands it supports instead of just text commands. If not handled by the editor, the webview will handle the command.
11:35 Changeset [29131] by adachan@apple.com

Regression: backspace stopped working on text inputs

  • Replaced the use of editor command "BackwardDelete" with "DeleteBackward" since it's no longer supported.

Reviewed by Adam.

  • WebEditorClient.cpp: (WebEditorClient::textWillBeDeletedInTextField):
  • WebView.cpp:
11:00 Changeset [29130] by zecke@webkit.org

2008-01-03 Holger Hans Peter Freyther <zecke@selfish.org>

Reviewed by Alp.

-This is from http://bugs.webkit.org/show_bug.cgi?id=16115

Change the Gtk ContextMenuItem code to generate the GtkMenuItem
or GtkCheckMenuItem on the fly. Currently we will create a
GtkCheckMenuItem if the ContextMenuItem has been checked. What needs
to be done is to change WebCore to tell the platform code if an item
is checkable or not.

  • platform/ContextMenuItem.h: (WebCore::PlatformMenuItemDescription::PlatformMenuItemDescription):
  • platform/gtk/ContextMenuGtk.cpp: (WebCore::ContextMenu::appendItem):
  • platform/gtk/ContextMenuItemGtk.cpp: (WebCore::ContextMenuItem::ContextMenuItem): (WebCore::ContextMenuItem::~ContextMenuItem): (WebCore::ContextMenuItem::createNativeMenuItem): (WebCore::ContextMenuItem::releasePlatformDescription): (WebCore::ContextMenuItem::type): (WebCore::ContextMenuItem::setType): (WebCore::ContextMenuItem::action): (WebCore::ContextMenuItem::setAction): (WebCore::ContextMenuItem::platformSubMenu): (WebCore::ContextMenuItem::setSubMenu): (WebCore::ContextMenuItem::setChecked):
10:37 Changeset [29129] by aroben@apple.com

Use HTTP::Date instead of Date::Parse because it's installed by Cygwin by default

Rubberstamped by Mark.

  • Scripts/update-webkit-auxiliary-libs:
10:28 Changeset [29128] by mitz@apple.com

Rubber-stamped by Adam Roben.

  • update the project hierarchy to match the on-disk organization of the platform directory.
  • WebCore.vcproj/WebCore.vcproj:
10:19 Changeset [29127] by aroben@apple.com

Fix Bug 15663: update-webkit re-downloads WebKitAuxiliaryLibrary unnecessarily


Added a fuzz factor into the Last-Modified comparison for downloading

The zip file is served from a set of mirrors who give Last-Modified
times that are off by 1-3 seconds from each other. This was causing
the build bots to redownload WebKitAuxiliaryLibrary for every build,
which would then cause all of WebCore to rebuild each time.

Reviewed by Mark.

  • Scripts/update-webkit-auxiliary-libs: Check if the new zip file is at least 30 seconds newer than the old one -- otherwise we assume that the difference in time is due to the mirrors being slightly offset from each other. (sub lastModifiedToUnixTime): Added.
09:15 Changeset [29126] by kdecker@apple.com

Reviewed by Darin.

Fixed: <rdar://problem/5660603> QuickDraw plug-ins can cause a 100% reproducible assertion failure in AppKit (breaks Safari UI)

  • Plugins/WebBaseNetscapePluginView.mm: (-[WebBaseNetscapePluginView updateAndSetWindow]): Simplified an early return for non-QuickDraw plug-ins and switched to using the more NSView friendly version of lockFocus, lockFocusIfCanDraw.
07:34 Changeset [29125] by aroben@apple.com

Fix a buffer overrun and a leak introduced in r29098

Reviewed by Mark.

  • platform/text/StringImpl.cpp: (WebCore::StringImpl::StringImpl): Only allocate one buffer, and make it be big enough to hold the string contents plus the null terminator.
06:40 Changeset [29124] by ap@webkit.org

Windows build fix.

  • DumpRenderTree/ForwardingHeaders/wtf/HashTraits.h: Added (needed by COMPtr.h).
05:53 Changeset [29123] by hausmann@webkit.org

Moved QWebPage::open to QWebFrame::load and added setHtml.

Signed-off-by: Lars

05:50 Changeset [29122] by hausmann@webkit.org

Added the first revision of QWebView and started moving functionality from QWebPave over to QWebView and QWebFrame.

Signed-off-by: Lars

05:40 Changeset [29121] by hausmann@webkit.org

Fix compilation with gcc 4.3, include limits.h for INT_MAX

04:02 Changeset [29120] by alp@webkit.org

2008-01-03 Alp Toker <alp@atoker.com>

Suggested by Mark Rowe.

Fix indentation and remove trailing whitespace.

  • platform/network/curl/ResourceHandleManager.cpp:
03:57 QtWebKitContrib edited by hausmann@webkit.org
03:56 QtWebKitContrib edited by hausmann@webkit.org
03:53 QtWebKitContrib edited by hausmann@webkit.org
03:31 QtWebKitContrib edited by hausmann@webkit.org
03:04 Changeset [29119] by alp@webkit.org

2008-01-03 Luca Bruno <lethalman88@gmail.com>

Reviewed by Alp Toker.

Fix HTTP POST-based logins to sites like Facebook, GMail by ensuring
that the two POST methods don't conflict.

  • platform/network/curl/ResourceHandleManager.cpp: (WebCore::ResourceHandleManager::setupPOST):
01:30 QtWebKitTodo edited by zander@kde.org
00:49 Changeset [29118] by darin@apple.com
  • tests/mozilla/expected.html: The fix for bug 16696 also fixed a test case, ecma_3/RegExp/perlstress-002.js, so updated results to expect that test to succeed.


23:50 Changeset [29117] by weinig@apple.com
  • Scripts/do-webcore-rename: Yet more renaming ideas.
23:44 Changeset [29116] by darin@apple.com
  • touched some files to try to get the Windows buildbot building again
23:37 Changeset [29115] by darin@apple.com

Oops, fix another.

23:37 Changeset [29114] by darin@apple.com

Oops, fix one.

23:36 Changeset [29113] by darin@apple.com
  • Scripts/do-webcore-rename: More renaming ideas. (Maciej, please merge yours with mine.)
22:47 Changeset [29112] by mitz@apple.com

Reviewed by Sam Weinig.

  • fix small caps rendering

Covered by an existing test,

  • platform/graphics/win/FontDataWin.cpp: (WebCore::FontData::smallCapsFontData):
22:45 Changeset [29111] by mrowe@apple.com

Merge r27238 to fix a hang during the layout tests.

22:39 Changeset [29110] by darin@apple.com


Reviewed by Geoff.

Test: fast/regex/early-acid3-86.html

The problem was with the cutoff point between backreferences and octal
escape sequences. We need to determine the cutoff point by counting the
total number of capturing brackets, which requires an extra pass through
the expression when compiling it.

  • pcre/pcre_compile.cpp: (CompileData::CompileData): Added numCapturingBrackets. Removed some unused fields. (compileBranch): Use numCapturingBrackets when calling checkEscape. (calculateCompiledPatternLength): Use numCapturingBrackets when calling checkEscape, and also store the bracket count at the end of the compile. (jsRegExpCompile): Call calculateCompiledPatternLength twice -- once to count the number of brackets and then a second time to calculate the length.


Reviewed by Geoff.

  • fast/regex/early-acid3-86-expected.txt: Added.
  • fast/regex/early-acid3-86.html: Added.
  • fast/regex/resources/early-acid3-86.js: Added.
  • fast/regex/test1-expected.txt: Updated for a few cases where we now fail. But these "failures" represent us replacing PCRE behavior with semantics that are correct for JavaScript regular expressions.
22:11 Changeset [29109] by darin@apple.com

Reviewed and landed by Darin.

  • kjs/nodes.cpp: (KJS::DoWhileNode::execute): Added a missing return.
19:56 Changeset [29108] by antti@apple.com

Reviewed by Oliver.

Calculate video position and size within the renderer box in WebCore. This
way the aspect ratio calculation is not needed in each MediaPlayer implementation.

This fixes video aspect ratio on Windows.

Covered by an existing pixel test.

  • platform/graphics/mac/MediaPlayerPrivateQTKit.mm: (WebCore::MediaPlayerPrivate::createQTMovieView):
  • rendering/RenderVideo.cpp: (WebCore::RenderVideo::videoBox): (WebCore::RenderVideo::paintReplaced): (WebCore::RenderVideo::updatePlayer):
  • rendering/RenderVideo.h:
18:33 Changeset [29107] by adachan@apple.com

Fix windows build.

  • WebCore.vcproj/WebCore.vcproj:
18:09 Changeset [29106] by sullivan@apple.com

Reviewed by Dan Bernstein

  • fixed <rdar://problem/5579010> REGRESSION: Safari inserts newlines in tab names when U+2028 is present
  • loader/DocumentLoader.cpp: (WebCore::canonicalizedTitle): in the code that replaces control characters with white space, also replace the unicode line separator and paragraph separator characters
18:07 Changeset [29105] by zecke@webkit.org

2008-01-02 Holger Hans Peter Freyther <zecke@selfish.org>

Rubber stamped by Alp.

Remove GDK_MULTIHEAD_SAFE and GTK_MULTIHEAD_SAFE because they break
the build. At least people doing a debug build on Ubuntu Hardy will see
the breakage. I was asked to leave the flags inside the files to ease future

  • GNUmakefile.am:
  • WebKit.pri:
18:05 Changeset [29104] by zecke@webkit.org

2008-01-02 Holger Hans Peter Freyther <zecke@selfish.org>

Rubber stamped by Darin.

Coding Style fix. Do not use an else because we use a return in the
if branch.

  • WebView/webkitwebview.cpp:
18:00 Changeset [29103] by darin@apple.com
  • fix buffer overruns seen on buildbot
  • platform/text/StringImpl.cpp: (WebCore::countCharacter): Added this. (WebCore::StringImpl::toCoordsArray): Use countCharacter instead of incorrect find expression. (WebCore::StringImpl::toLengthArray): Ditto. (WebCore::StringImpl::replace): Added assertions.
17:29 Changeset [29102] by alice.liu@apple.com
  • platform/win/Skipped: removing a fixed test this was fixed with r28836 but was left behind on the skipped list
17:25 Changeset [29101] by darin@apple.com


Reviewed by Maciej.

Tests: dom/html/level2/html/HTMLCollection07.html


  • GNUmakefile.am: Added HTMLTableRowsCollection.
  • WebCore.pro: Ditto.
  • WebCore.vcproj/WebCore.vcproj: Ditto.
  • WebCore.xcodeproj/project.pbxproj: Ditto.
  • WebCoreSources.bkl: Ditto.
  • dom/XMLTokenizer.cpp: Took out stray include.
  • html/HTMLCollection.cpp: (WebCore::HTMLCollection::itemAfter): Removed all the table rows code, since we now use a separate class for that collection. Also got rid of the distinct types for custom collections that don't need them (use Other for both).
  • html/HTMLCollection.h: Also made firstItem non-virtual because it doesn't need to be virtual.
  • html/HTMLFormCollection.cpp: (WebCore::HTMLFormCollection::HTMLFormCollection): Use Other instead of FormElements for the HTMLCollection type.
  • html/HTMLTableElement.cpp: (WebCore::HTMLTableElement::HTMLTableElement): Eliminated m_head, m_foot, m_firstBody, and m_caption. (WebCore::HTMLTableElement::caption): Added non-inline version. Finds the caption rather than keeping a pointer to it. (WebCore::HTMLTableElement::setCaption): Rewrote. (WebCore::HTMLTableElement::tHead): Ditto. (WebCore::HTMLTableElement::setTHead): Ditto. (WebCore::HTMLTableElement::tFoot): Ditto. (WebCore::HTMLTableElement::setTFoot): Ditto. (WebCore::HTMLTableElement::createTHead): Ditto. (WebCore::HTMLTableElement::deleteTHead): Ditto. (WebCore::HTMLTableElement::createTFoot): Ditto. (WebCore::HTMLTableElement::deleteTFoot): Ditto. (WebCore::HTMLTableElement::createCaption): Ditto. (WebCore::HTMLTableElement::deleteCaption): Ditto. (WebCore::HTMLTableElement::lastBody): Added. (WebCore::HTMLTableElement::insertRow): Rewrote to use a loop based on code in HTMLTableRowsCollection. This is different from the old code mainly in how it handles rows outside any section. (WebCore::HTMLTableElement::deleteRow): Ditto. (WebCore::HTMLTableElement::addChild): Removed code to set the various members. Keeping pointers to these was a possible source of serious bugs too, including crashes with stale pointers, although I didn't write any test cases to prove those bugs existed. (WebCore::HTMLTableElement::parseMappedAttribute): Changed the rules code to visit all cells, not just the cells of the first body. I believe this fixed rendering on some table tests. I think the code visits too many cells and also the use of recursion is overkill, but I didn't try to fix that. (WebCore::HTMLTableElement::rows): Changed to use the new HTMLTableRowsCollection.
  • html/HTMLTableElement.h: Changed functions to return PassRefPtr, which can be important if strange things like DOM mutation events take things ot of the tree before they are safely referenced by JavaScript wrappers. Also changed functions to take PassRefPtr and added exceptions. Removed unneeded firstTBody and setTBody functions and childrenChanged function override, as well as unused Rules and Frame enums. Removed m_head, m_foot, m_firstBody, and m_caption, and added lastBody function. Removed unneeded friend declaration for HTMLTableCellElement.
  • html/HTMLTableElement.idl: Allow the setteres for caption, tHead, and tFoot to raise exceptions.
  • html/HTMLTableRowsCollection.cpp: Added. Implements the HTML 5 rule for which rows are in the collection in which order.
  • html/HTMLTableRowsCollection.h: Added.
  • loader/FTPDirectoryDocument.cpp: (WebCore::FTPDirectoryTokenizer::appendEntry): Use the standard insertRow function instead of coming up with our own way of inserting a row. Simplifies things -- we can remove the code to create a tbody element.


Reviewed by Maciej.

  • fast/dom/HTMLTableElement/early-acid3-65-excerpt-expected.txt: Added.
  • fast/dom/HTMLTableElement/early-acid3-65-excerpt.html: Added.
  • fast/dom/HTMLTableElement/early-acid3-66-excerpt-expected.txt: Added.
  • fast/dom/HTMLTableElement/early-acid3-66-excerpt.html: Added.
  • fast/dom/HTMLTableElement/resources/early-acid3-65-excerpt.js: Added.
  • fast/dom/HTMLTableElement/resources/early-acid3-66-excerpt.js: Added.
  • dom/html/level2/html/HTMLCollection07-expected.txt: Updated to reflect success.
  • dom/html/level2/html/HTMLCollection08-expected.txt: Updated to reflect success.
  • dom/xhtml/level2/html/HTMLCollection07-expected.txt: Updated to reflect success.
  • dom/xhtml/level2/html/HTMLCollection08-expected.txt: Updated to reflect success.
  • fast/dom/HTMLTableElement/resources/rows.js: Changed to expect HTML 5 behavior for rows outside table sections.
  • fast/dom/HTMLTableElement/rows-expected.txt: Updated.
  • platform/mac/tables/mozilla/bugs/bug30418-expected.checksum: Updated to reflect better results.
  • platform/mac/tables/mozilla/bugs/bug30418-expected.png: Ditto.
  • platform/mac/tables/mozilla/bugs/bug30418-expected.txt: Ditto.
17:11 Changeset [29100] by darin@apple.com
  • try to fix Qt build
  • wtf/unicode/qt4/UnicodeQt4.h: (WTF::Unicode::foldCase): Add some missing const.
16:58 Changeset [29099] by darin@apple.com

Reviewed by Alice and Tim.

  • try to fix GTK and Qt builds
  • platform/win/ScrollViewWin.cpp: (WebCore::ScrollView::scroll): Improve logic slightly for the case of vertical scrolling when there's no vertical scroll bar.
  • platform/gtk/ScrollViewGtk.cpp: (WebCore::ScrollView::scroll): Copy the code from Windows. Maybe this should be factored differently.
  • platform/qt/ScrollViewQt.cpp: (WebCore::ScrollView::scroll): Ditto.
16:32 Changeset [29098] by darin@apple.com

Reviewed by Anders.

Also eliminated use of const StringImpl. Since StringImpl is immutable there
is no distinction between a const and non-const one at the moment.

  • WebCore.base.exp: Updated.
  • css/CSSHelper.cpp: (WebCore::parseURL): Make String directly, not by making a StringImpl.
  • dom/Attr.cpp: (WebCore::Attr::createTextChild): Convert AtomicString to String with domString, not via StringImpl. (WebCore::Attr::setValue): Remove unneed call to impl() when passing a String to a function that takes a String.
  • dom/CDATASection.cpp: Removed unused constructor. (WebCore::CDATASection::cloneNode): Added a now-needed .get(). (WebCore::CDATASection::createNew): Changed function to take a PassRefPtr.
  • dom/CDATASection.h:
  • dom/CharacterData.cpp: (WebCore::CharacterData::CharacterData): Removed unneeded initialization and ref() now that the string is a RefPtr. Also updated to not call "new StringImpl". (WebCore::CharacterData::~CharacterData): Removed unneeded deref() since it's a RefPtr. (WebCore::CharacterData::setData): More of that. (WebCore::CharacterData::substringData): Ditto. (WebCore::CharacterData::appendData): Ditto. (WebCore::CharacterData::insertData): Ditto. (WebCore::CharacterData::deleteData): Ditto. (WebCore::CharacterData::replaceData): Ditto. (WebCore::CharacterData::nodeValue): Ditto. (WebCore::CharacterData::dispatchModifiedEvent): Ditto. (WebCore::CharacterData::dump): Ditto.
  • dom/CharacterData.h: Changed to use a RefPtr. I could have used a String instead, but since String adds extra branches to handle 0, I figured it was more conservative to just use RefPtr. Later it would be good to figure out which is preferred style and be more consistent. Maybe we'll phase out StringImpl, or maybe we'll go the other way and use it more since it can be more efficient.
  • dom/DOMImplementation.cpp: (WebCore::addString): Changed set to use String rather than StringImpl. (WebCore::isSVG10Feature): Ditto. (WebCore::isSVG11Feature): Ditto. (WebCore::DOMImplementation::createDocument): Replaced custom code to find a colon with a call to String::find.
  • dom/Range.cpp: (WebCore::Range::insertNode): Updated since the result of splitText is now a PassRefPtr.
  • dom/Text.cpp: (WebCore::Text::splitText): Updated since str is now a RefPtr. Also made the result of this function be a PassRefPtr. (WebCore::Text::createRenderer): Ditto. (WebCore::Text::createNew): Made the parameter and result both be PassRefPtr.
  • dom/Text.h:
  • html/HTMLElement.cpp: (WebCore::HTMLElement::nodeName): Use String::upper.
  • html/HTMLInputElement.cpp: (WebCore::numGraphemeClusters): Remove now-unneeded const. (WebCore::numCharactersInGraphemeClusters): Ditto.
  • html/HTMLTokenizer.cpp: (WebCore::HTMLTokenizer::processToken): Updated for function name change.
  • platform/text/AtomicString.cpp: (WebCore::CStringTranslator::translate): Updated since there is no longer a constructor that takes a string.
  • platform/text/PlatformString.h: Added new constructors that take PassRefPtr and RefPtr. Removed misleading comment.
  • platform/text/String.cpp: (WebCore::String::String): Changed to use StringImpl::create, which handles the empty string automatically. (WebCore::String::append): Ditto. (WebCore::String::charactersWithNullTermination): Similar. (WebCore::String::format): Ditto.
  • platform/text/StringHash.h: Took out unneeded const.
  • platform/text/StringImpl.cpp: (WebCore::deleteUCharVector): Changed to take a const pointer since the buffers are now const UChar buffers. (WebCore::StringImpl::StringImpl): Removed some constructors. Got rid of the separate init functions. The constructors are now private and used only in the create functions and one or two other places. (WebCore::StringImpl::containsOnlyWhitespace): Removed now-meaningless const. (WebCore::StringImpl::substring): Ditto. Also changed return value to be a PassRefPtr. (WebCore::StringImpl::characterStartingAt): Ditto. (WebCore::StringImpl::toLength): Ditto. (WebCore::StringImpl::toCoordsArray): Ditto. (WebCore::StringImpl::toLengthArray): Ditto. (WebCore::StringImpl::isLower): Ditto. (WebCore::StringImpl::lower): Ditto. Changed to use Vector and adopt so we don't have to use new directly here. Makes empty string handling more consistent. (WebCore::StringImpl::upper): Ditto. (WebCore::StringImpl::secure): Ditto. (WebCore::StringImpl::foldCase): Ditto. (WebCore::StringImpl::stripWhiteSpace): Ditto. (WebCore::StringImpl::simplifyWhiteSpace): Ditto. (WebCore::StringImpl::capitalize): Ditto. (WebCore::StringImpl::toInt): Removed now-meaningless const. (WebCore::StringImpl::toInt64): Ditto. (WebCore::StringImpl::toUInt64): Ditto. (WebCore::StringImpl::toDouble): Ditto. (WebCore::StringImpl::toFloat): Ditto. (WebCore::StringImpl::find): Ditto. (WebCore::StringImpl::reverseFind): Ditto. (WebCore::StringImpl::endsWith): Ditto. (WebCore::StringImpl::replace): Ditto. (WebCore::equal): Ditto. (WebCore::equalIgnoringCase): Ditto. (WebCore::StringImpl::ascii): Ditto. (WebCore::StringImpl::defaultWritingDirection): Ditto. (WebCore::StringImpl::createStrippingNullCharacters): Ditto. (WebCore::StringImpl::adopt): Added special case so this uses the shared empty string like other functions. Also optimized the common case where the vector happens to already have the right size so we don't do a fastRealloc at all in those cases. (WebCore::StringImpl::create): Added. These are now the public functions for creating new StringImpl objects. They all implement the shared empty string. (WebCore::StringImpl::createWithTerminatingNullCharacter):
  • platform/text/StringImpl.h:
  • platform/text/cf/StringCF.cpp: (WebCore::String::String): Use StringImpl::create instead of new StringImpl.
  • platform/text/cf/StringImplCF.cpp: (WebCore::StringImpl::createCFString): Removed now-obsolete const.
  • platform/text/mac/StringImplMac.mm: (WebCore::StringImpl::operator NSString *): Ditto.
  • platform/text/mac/StringMac.mm: (WebCore::String::String): Use StringImpl::create instead of new StringImpl.
  • platform/text/qt/StringQt.cpp: (WebCore::String::String): Ditto.
  • platform/text/wx/StringWx.cpp: (WebCore::String::String): Ditto.
  • rendering/RenderBR.cpp: (WebCore::RenderBR::RenderBR): Ditto.
  • rendering/RenderSVGInlineText.cpp: (WebCore::RenderSVGInlineText::RenderSVGInlineText): Use PassRefPtr.
  • rendering/RenderSVGInlineText.h:
  • rendering/RenderText.cpp: (WebCore::charactersAreAllASCII): Removed now-unneeded const.
  • rendering/RenderTextFragment.cpp: (WebCore::RenderTextFragment::originalText): Use RefPtr.
16:07 Changeset [29097] by timothy@apple.com

Reviewed by Oliver Hunt.

<rdar://problem/5618086> WebInspector does not expand the DOM tree after being closed

Closing the Web Inspector causes the DOM tree outline to be torn down, clearing the
internal element lookup tables. The represented DOM node object still holds the identifier
it was assigned, and a later call to findTreeElement will use that original identifier
against a cleared lookup table. In that case we need to fallback on DOM ancestor lookup.

  • page/inspector/treeoutline.js: (TreeOutline.prototype.findTreeElement): If the DOM node already had a treeElementIdentifier, but the TreeOutline no longer has the element in the _knownTreeElements list do an ancestor lookup instead of an early return.
15:53 Changeset [29096] by alice.liu@apple.com


Reviewed by Sam Weinig.

need to export ASCIICType.h for use in DRT

  • JavaScriptCore.vcproj/WTF/WTF.vcproj:
  • wtf/ASCIICType.h: (WTF::isASCIIUpper):


Reviewed by Sam Weinig.

fixing assertion hit with editing/selection/move-begin-end.html

  • DumpRenderTree/ForwardingHeaders/wtf/ASCIICType.h: Added.
  • DumpRenderTree/win/DumpRenderTree.vcproj:
  • DumpRenderTree/win/EventSender.cpp: (keyDownCallback): using isupper will cause an assertion for inputs outside of ascii range. use isASCIIUpper instead.
15:46 Changeset [29095] by darin@apple.com
  • Scripts/do-webcore-rename: Some more name change plans.
14:58 Changeset [29094] by alice.liu@apple.com


Reviewed by Maciej.

Fixed <rdar://5283861> (problems scrolling in gmail message content area)

  • platform/ScrollView.h:
  • platform/win/ScrollViewWin.cpp: (WebCore::ScrollView::scroll): Changed return value to bool to reflect success of scroll attempt


Reviewed by Maciej.

Fixed <rdar://5283861> (problems scrolling in gmail message content area)

  • WebView.cpp: (WebView::keyDown): bubble scrolling from a key event


Reviewed by Maciej.

Added test for <rdar://5283861> (problems scrolling in gmail message content area)

  • fast/frames/iframe-scroll-page-up-down-expected.txt: Added.
  • fast/frames/iframe-scroll-page-up-down.html: Added.
  • fast/frames/resources/iframe-scroll-page-up-down-1.html: Added.
  • fast/frames/resources/iframe-scroll-page-up-down-2.html: Added.
14:37 Changeset [29093] by alice.liu@apple.com

Reviewed by Maciej.

Added handling for page up and page down in EventSender

  • DumpRenderTree/mac/EventSendingController.mm: (-[EventSendingController keyDown:withModifiers:]):
  • DumpRenderTree/win/EventSender.cpp: (keyDownCallback):
14:23 Changeset [29092] by aroben@apple.com
  • platform/win/Skipped: Added some more failures.
13:32 Changeset [29091] by weinig@apple.com

Reviewed by Beth Dakin.

Cleanup error_object.h/cpp.

  • kjs/JSGlobalObject.cpp: (KJS::JSGlobalObject::reset):
  • kjs/error_object.cpp: (KJS::): (KJS::ErrorInstance::ErrorInstance): (KJS::ErrorPrototype::ErrorPrototype): (KJS::ErrorProtoFuncToString::ErrorProtoFuncToString): (KJS::ErrorProtoFuncToString::callAsFunction): (KJS::ErrorObjectImp::ErrorObjectImp): (KJS::ErrorObjectImp::implementsConstruct): (KJS::ErrorObjectImp::construct): (KJS::ErrorObjectImp::callAsFunction): (KJS::NativeErrorPrototype::NativeErrorPrototype): (KJS::NativeErrorImp::NativeErrorImp): (KJS::NativeErrorImp::implementsConstruct): (KJS::NativeErrorImp::construct): (KJS::NativeErrorImp::callAsFunction): (KJS::NativeErrorImp::mark):
  • kjs/error_object.h: (KJS::ErrorInstance::classInfo): (KJS::NativeErrorImp::classInfo):
13:07 Changeset [29090] by ap@webkit.org

Fixed a typo (pointed out in review, but I somehow missed it at first).

  • editing/EditorCommand.cpp: (WebCore::executeDelete):
12:24 Changeset [29089] by aroben@apple.com

Get the Windows bots closer to green

  • platform/win/Skipped: Added some new failures.
  • platform/win/fast/dom/Window/window-properties-expected.txt: Updated these results.
11:14 Changeset [29088] by ap@webkit.org

Reviewed by Darin.

Fix fast/events/arrow-keys-on-body.html for real.

  • DumpRenderTree/win/EventSender.cpp: (keyDownCallback): Pass proper keyData for WM_KEYUP, too.
10:31 Changeset [29087] by ap@webkit.org

Reviewed by Darin.

Resolved several FIXMEs in EditorCommand.
Made recently added commands hidden from JS again.
Removed BackwardDelete implementation, which used to be dead code, but got exposed now.

Tests: editing/execCommand/delete-no-scroll.html


09:34 Changeset [29086] by ap@webkit.org

Reviewed by Darin.

action=mailto + method=get - The generated mailto URI is incorrect and the hvalues are encoded twice

Submitted data only includes first input item

Reworked encoding of mailto URLs to match other browsers.
Moved most of related logic from FrameLoader::submitForm() to HTMLFormElement::submit().

Tests: fast/forms/mailto/advanced-get.html


  • html/HTMLFormElement.cpp: (WebCore::HTMLFormElement::submit):
  • loader/FrameLoader.cpp: (WebCore::FrameLoader::submitForm):
08:01 Changeset [29085] by mrowe@apple.com

Autotools build fix.

07:14 Changeset [29084] by mrowe@apple.com

Add missing dependencies to some GNUmakefile.am rules.

Rubber-stamped by Alp Toker.

06:42 Changeset [29083] by mrowe@apple.com

Autotools fix. Add missing dependency on grammar.y.

Rubber-stamped by Alp Toker.

06:42 Changeset [29082] by mrowe@apple.com

Autotools build fix. Make can expand $@ to any of the targets for the rule,
while we always want to use the name of the .cpp file as the output file.

Reviewed by Alp Toker.

04:56 Changeset [29081] by alp@webkit.org

2008-01-02 Luca Bruno <lethalman88@gmail.com>

Reviewed by Darin.

[GTK] run-launcher sets wrong LD_LIBRARY_PATH

  • Scripts/run-launcher:
03:33 QtWebKitContrib edited by hausmann@webkit.org
02:25 Changeset [29080] by alp@webkit.org

2008-01-02 Luca Bruno <lethalman88@gmail.com>

Reviewed by Alp Toker.

[GTK] ContextMenu and ContextMenuItem lacks an implementation

Add context menu support.

Based on a patch by Holger Freyther.

01:07 Changeset [29079] by alp@webkit.org

2008-01-02 Alp Toker <alp@atoker.com>

GTK+ autotools build fix. Track changes in r29073.

  • GNUmakefile.am:


23:03 Changeset [29078] by darin@apple.com
  • fix release build
  • bindings/js/kjs_binding.cpp: (KJS::setDOMException): Initialize to avoid uninitialized variable warning. Removed default so we get a warning if there's a missing case.
22:59 Changeset [29077] by ddkilzer@apple.com

Scripting MIME Types application/ecmascript, application/javascript not viewable

Reviewed by Darin.

This patch consolidates the list of acceptable MIME types for JavaScript
source into the MIMETypeRegistry class, and replaces checks for these
types with a call to MIMETypeRegistry::isSupportedJavaScriptMIMEType().

No tests added since viewing JavaScript source is not testable.

  • dom/DOMImplementation.cpp: (WebCore::DOMImplementation::isTextMIMEType): Use MIMETypeRegistry::isSupportedJavaScriptMIMEType() instead of a single hard-coded MIME type, "application/x-javascript".
  • html/HTMLScriptElement.cpp: (WebCore::HTMLScriptElement::shouldExecuteAsJavaScript): Moved list of JavaScript MIME types from here to MIMETypeRegistry::initialiseSupportedJavaScriptMIMETypes().
  • platform/MIMETypeRegistry.cpp: (WebCore::initialiseSupportedJavaScriptMIMETypes): Added. List of MIME types came from HTMLScriptElement::shouldExecuteAsJavaScript(). (WebCore::initialiseSupportedNonImageMimeTypes): Remove single hard-coded MIME type, "application/x-javascript", from the list. (WebCore::initialiseMIMETypeRegistry): Initialise supportedJavaScriptMIMETypes, then pre-populate supportedNonImageMIMETypes with values in supportedJavaScriptMIMETypes. (WebCore::MIMETypeRegistry::isSupportedJavaScriptMIMEType): Added.
  • platform/MIMETypeRegistry.h: Added isSupportedJavaScriptMIMEType().
22:54 Changeset [29076] by weinig@apple.com

Fix wx build

22:44 Changeset [29075] by darin@apple.com


Reviewed by Eric.

Test: fast/js/kde/parse.html

  • kjs/lexer.cpp: (KJS::Lexer::lex): Added additional states to distinguish Unicode escapes at the start of identifiers from ones inside identifiers. Rejected characters that don't pass the isIdentStart and isIdentPart tests. (KJS::Lexer::convertUnicode): Removed incorrect FIXME comment.
  • kjs/lexer.h: Added new states to distinguish \u escapes at the start of identifiers from \u escapes inside identifiers.


Reviewed by Eric.

  • fast/js/kde/parse-expected.txt: Updated.
  • fast/js/kde/resources/parse.js: Added tests that cover both the non-ASCII characters themselves and the same characters parsed as \u sequences.
22:38 Changeset [29074] by darin@apple.com

Fix a stray executable bit.

21:46 Changeset [29073] by weinig@apple.com


Reviewed by Darin.

Patch for http://bugs.webkit.org/show_bug.cgi?id=16691
Use real JS objects for the rest of the DOMExceptions (EventException, RangeException, etc)

  • Create JS objects for all the different types of exceptions, each with there own prototype and constructor.
  • Abstract all the logic and storage for exception classes in to an ExceptionBase class.
  • Move specialized ExceptionCodes into the exception classes. (ie. the EventExceptionCode enum is now in EventException).

Tests: fast/dom/DOMException/EventException.html


  • DerivedSources.make:
  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCore.xcodeproj/project.pbxproj:
  • WebCoreSources.bkl:
  • bindings/js/JSSVGMatrixCustom.cpp: (WebCore::JSSVGMatrix::inverse): (WebCore::JSSVGMatrix::rotateFromVector):
  • bindings/js/kjs_binding.cpp: (KJS::setDOMException):
  • bindings/scripts/CodeGeneratorJS.pm: Use the constant values defined in the IDL as the ObjC bindings do.
  • bindings/scripts/CodeGeneratorObjC.pm:
  • dom/DOMCoreException.cpp: Removed.
  • dom/DOMCoreException.h: (WebCore::DOMCoreException::DOMCoreException):
  • dom/DOMCoreException.idl:
  • dom/Event.h:
  • dom/EventException.h: Copied from WebCore/dom/DOMCoreException.h. (WebCore::EventException::EventException): (WebCore::EventException::):
  • dom/EventException.idl: Copied from WebCore/dom/DOMCoreException.idl.
  • dom/EventTargetNode.cpp: (WebCore::EventTargetNode::dispatchEvent):
  • dom/ExceptionBase.cpp: Copied from WebCore/dom/DOMCoreException.cpp. (WebCore::ExceptionBase::ExceptionBase): (WebCore::ExceptionBase::toString):
  • dom/ExceptionBase.h: Copied from WebCore/dom/DOMCoreException.h. (WebCore::ExceptionBase::code):
  • dom/ExceptionCode.cpp: (WebCore::getExceptionCodeDescription):
  • dom/ExceptionCode.h: (WebCore::):
  • dom/Range.cpp: (WebCore::Range::insertNode): (WebCore::Range::checkNodeWOffset): (WebCore::Range::checkNodeBA): (WebCore::Range::selectNode): (WebCore::Range::selectNodeContents): (WebCore::Range::surroundContents):
  • dom/RangeException.h: (WebCore::RangeException::RangeException): (WebCore::RangeException::):
  • dom/RangeException.idl:
  • page/DOMWindow.idl:
  • svg/SVGColor.cpp: (WebCore::SVGColor::setRGBColor):
  • svg/SVGException.h: (WebCore::SVGException::SVGException): (WebCore::SVGException::):
  • svg/SVGException.idl:
  • svg/SVGLocatable.cpp: (WebCore::SVGLocatable::getTransformToElement):
  • xml/XMLHttpRequest.cpp: (WebCore::XMLHttpRequest::dispatchEvent): (WebCore::XMLHttpRequest::open): (WebCore::XMLHttpRequest::send):
  • xml/XMLHttpRequest.h:
  • xml/XMLHttpRequestException.h: Copied from WebCore/dom/DOMCoreException.h. (WebCore::XMLHttpRequestException::XMLHttpRequestException): (WebCore::XMLHttpRequestException::):
  • xml/XMLHttpRequestException.idl: Copied from WebCore/dom/DOMCoreException.idl.
  • xml/XPathEvaluator.h:
  • xml/XPathException.h: Copied from WebCore/dom/DOMCoreException.h. (WebCore::XPathException::XPathException): (WebCore::XPathException::):
  • xml/XPathException.idl: Copied from WebCore/dom/DOMCoreException.idl.
  • xml/XPathParser.cpp: (WebCore::XPath::Parser::parseStatement):
  • xml/XPathResult.cpp: (WebCore::XPathResult::convertTo): (WebCore::XPathResult::numberValue): (WebCore::XPathResult::stringValue): (WebCore::XPathResult::booleanValue): (WebCore::XPathResult::singleNodeValue): (WebCore::XPathResult::snapshotLength): (WebCore::XPathResult::iterateNext): (WebCore::XPathResult::snapshotItem):


Reviewed by Darin.

Tests for http://bugs.webkit.org/show_bug.cgi?id=16691
Use real JS objects for the rest of the DOMExceptions (EventException, RangeException, etc)

  • fast/dom/DOMException/EventException-expected.txt: Added.
  • fast/dom/DOMException/EventException.html: Copied from LayoutTests/fast/dom/DOMException/prototype-object.html.
  • fast/dom/DOMException/RangeException-expected.txt: Added.
  • fast/dom/DOMException/RangeException.html: Copied from LayoutTests/fast/dom/DOMException/prototype-object.html.
  • fast/dom/DOMException/XPathException-expected.txt: Added.
  • fast/dom/DOMException/XPathException.html: Copied from LayoutTests/fast/dom/DOMException/prototype-object.html.
  • fast/dom/DOMException/resources/EventException.js: Copied from LayoutTests/fast/dom/DOMException/resources/prototype-object.js.
  • fast/dom/DOMException/resources/RangeException.js: Copied from LayoutTests/fast/dom/DOMException/resources/prototype-object.js.
  • fast/dom/DOMException/resources/XPathException.js: Copied from LayoutTests/fast/dom/DOMException/resources/prototype-object.js.
  • fast/dom/Window/window-properties-expected.txt:
  • http/tests/xmlhttprequest/XMLHttpRequestException-expected.txt: Added.
  • http/tests/xmlhttprequest/XMLHttpRequestException.html: Added.
  • svg/custom/SVGException-expected.txt: Added.
  • svg/custom/SVGException.html: Copied from LayoutTests/fast/dom/DOMException/prototype-object.html.
  • svg/custom/resources/SVGException.js: Copied from LayoutTests/fast/dom/DOMException/resources/prototype-object.js.
16:38 Changeset [29072] by ddkilzer@apple.com

Reviewed by Dan.

  • DefaultDelegates/WebDefaultPolicyDelegate.m: Check return value of -[NSFileManager fileExistsAtPath:isDirectory:] before using the value of isDirectory.
12:44 Changeset [29071] by weinig@apple.com

Remove JSDomExceptionConstructor.lut.h from clean step
as it no longer exists.

  • GNUmakefile.am:
12:42 Changeset [29070] by mitz@apple.com
  • Windows build fix
  • WebCore.vcproj/WebCore.vcproj:
11:52 Changeset [29069] by pewtermoose@webkit.org

Attempt to fix the Windows bots until they can get the Windows SDK installed.

11:47 Changeset [29068] by darin@apple.com
  • rolled scope chain optimization out; it was breaking the world
11:35 Changeset [29067] by darin@apple.com

Reviewed by Geoff.

Also included one other speed-up -- remove the call to reserveCapacity from
FunctionBodyNode::processDeclarations in all but the most unusual cases.

Together these make SunSpider 1.016x as fast.

  • JavaScriptCore.exp: Updated.
  • kjs/ExecState.cpp: (KJS::globalEmptyList): Added. Called only when creating global ExecState instances. (KJS::ExecState::ExecState): Broke constructor up into three separate functions, for the three separate node types. Also went through each of the three and streamlined as much as possible, removing dead code. This prevents us from having to access the global in the function body version of the constructor.
  • kjs/ExecState.h: Added emptyList(). Replaced the constructor with a set of three that are specific to the different node types that can create new execution state objects.
  • kjs/array_object.cpp: (KJS::ArrayProtoFuncToLocaleString::callAsFunction): Use exec->emptyList() instead of List::empty(). (KJS::ArrayProtoFuncConcat::callAsFunction): Ditto. (KJS::ArrayProtoFuncSlice::callAsFunction): Ditto. (KJS::ArrayProtoFuncSplice::callAsFunction): Ditto. (KJS::ArrayProtoFuncFilter::callAsFunction): Ditto.
  • kjs/function.cpp: (KJS::FunctionImp::callAsFunction): Updated to call new ExecState constructor. (KJS::GlobalFuncImp::callAsFunction): Ditto (for eval).
  • kjs/function_object.cpp: (FunctionObjectImp::construct): Use exec->emptyList() instead of List::empty().
  • kjs/list.cpp: Removed List::empty.
  • kjs/list.h: Ditto.
  • kjs/nodes.cpp: (KJS::ElementNode::evaluate): Use exec->emptyList() instead of List::empty(). (KJS::ArrayNode::evaluate): Ditto. (KJS::ObjectLiteralNode::evaluate): Ditto. (KJS::PropertyListNode::evaluate): Ditto. (KJS::FunctionBodyNode::processDeclarations): Another speed-up. Check the capacity before calling reserveCapacity, because it doesn't get inlined the local storage vector is almost always big enough -- saving the function call overhead is a big deal. (KJS::FuncDeclNode::makeFunction): Use exec->emptyList() instead of List::empty(). (KJS::FuncExprNode::evaluate): Ditto.
  • kjs/object.cpp: (KJS::tryGetAndCallProperty): Ditto.
  • kjs/property_slot.cpp: (KJS::PropertySlot::functionGetter): Ditto.
  • kjs/string_object.cpp: (KJS::StringProtoFuncSplit::callAsFunction): Ditto.
11:16 Changeset [29066] by darin@apple.com


Reviewed by Geoff.

Fixes 34 failing test cases in the fast/regex/test1.html test.

Restored the stack which prevents infinite loops for brackets that match the empty
string; it had been removed as an optimization.

Unfortunately, restoring this stack causes the regular expression test in SunSpider
to be 1.095x as slow and the overall test to be 1.004x as slow. Maybe we can find
a correct optimization to restore the speed!

It's possible the original change was on the right track but just off by one.

  • pcre/pcre_exec.cpp: Add back eptrblock, but name it BracketChainNode. (MatchStack::pushNewFrame): Add back the logic needed here. (startNewGroup): Ditto. (match): Ditto.


Reviewed by Geoff.

  • fast/regex/test1-expected.txt: Update results changed by restoring the logic to avoid failing on infinite repeats of brackets that match the empty string.
11:13 Changeset [29065] by darin@apple.com

Reviewed by Geoff.

This gives a 1.019x speedup on SunSpider.

After doing this, I realized this probably will be obsolete when the optimization
to avoid creating an activation object is done. When we do that one we should check
if rolling this out will speed things up, since this does add overhead at the time
you copy the scope chain.

  • kjs/object.h: Removed the ScopeChain::release function. It was marked inline, and called in exactly one place, so moved it there. No idea why it was in this header file!
  • kjs/scope_chain.cpp: Removed the overload of the ScopeChain::push function that takes another ScopeChain. It was unused. I think we used it over in WebCore at one point, but not any more.
  • kjs/scope_chain.h: Changed ScopeChainNode into a struct rather than a class, got rid of its constructor so we can have one that's uninitialized, and moved the refCount into a derived struct, ScopeChainHeapNode. Made _node mutable so it can be changed in the moveToHeap function. Changed the copy constructor and assignment operator to call moveToHeap, since the top node can't be shared when it's embedded in another ScopeChain object. Updated functions as needed to handle the case where the first object isn't on the heap or to add casts for cases where it's guaranteed to be. Changed the push function to always put the new node into the ScopeChain object; it will get put onto the heap when needed later.
10:57 Changeset [29064] by ggaren@apple.com

Reviewed by Darin Adler.

Fixed slight logic error in reserveCapacity, where we would reallocate
the storage buffer unnecessarily.

  • wtf/Vector.h: (WTF::::reserveCapacity): No need to grow the buffer if newCapacity is equal to capacity().
10:14 BuildingGtk edited by sylvain.pasche@gmail.com
link to ApplicationsGtk (diff)
10:12 ApplicationsGtk edited by sylvain.pasche@gmail.com
added PyWebKitGtk (diff)
10:05 Changeset [29063] by weinig@apple.com

Try again to fix the builds

  • DerivedSources.make:
09:45 Changeset [29062] by weinig@apple.com

Fix non-mac builds.

  • WebCore.pro:
  • WebCore.vcproj/WebCore.vcproj:
  • WebCoreSources.bkl:
02:40 Changeset [29061] by eric@webkit.org

Reviewed by Alexey.

Don't replace \ with / in data: urls

Test: fast/loader/url-data-replace-backslash.html

  • platform/KURL.cpp: (WebCore::KURL::init):
01:05 Changeset [29060] by alp@webkit.org

2008-01-01 Alp Toker <alp@atoker.com>

GTK+ autotools build fix. Track changes in r29051, r29058 and pass the
correct parameter to AM_INIT_AUTOMAKE.

00:43 Changeset [29059] by darin@apple.com

Reviewed by Oliver.

Speeds SunSpider up 1.003x. That's a small amount, but measurable.

  • JavaScriptCore.exp: Updated.
  • kjs/Parser.h: (KJS::Parser::parse): Create the node with a static member function named create() instead of using new explicitly.
  • kjs/grammar.y: Changed calls to new FunctionBodyNode to use FunctionBodyNode::create().
  • kjs/nodes.cpp: (KJS::ProgramNode::create): Added. Calls new. (KJS::EvalNode::create): Ditto. (KJS::FunctionBodyNode::create): Ditto, but creates FunctionBodyNodeWithDebuggerHooks when a debugger is present. (KJS::FunctionBodyNode::execute): Removed debugger hooks. (KJS::FunctionBodyNodeWithDebuggerHooks::FunctionBodyNodeWithDebuggerHooks): Added. (KJS::FunctionBodyNodeWithDebuggerHooks::execute): Calls the debugger, then the code, then the debugger again.
  • kjs/nodes.h: Added create functions, made the constructors private and protected.
Note: See TracTimeline for information about the timeline view.