13 years ago2008-01-02 Holger Hans Peter Freyther <zecke@selfish.org>
zecke@webkit.org [Thu, 3 Jan 2008 02:05:01 +0000 (02:05 +0000)]
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:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29104 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago - fix buffer overruns seen on buildbot
darin@apple.com [Thu, 3 Jan 2008 02:00:26 +0000 (02:00 +0000)]
    - 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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29103 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago * platform/win/Skipped: removing a fixed test
alice.liu@apple.com [Thu, 3 Jan 2008 01:29:06 +0000 (01:29 +0000)]
    * platform/win/Skipped: removing a fixed test
        this was fixed with r28836 but was left behind on the skipped list

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29102 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
darin@apple.com [Thu, 3 Jan 2008 01:25:51 +0000 (01:25 +0000)]

        Reviewed by Maciej.

        - fix http://bugs.webkit.org/show_bug.cgi?id=16657
          Acid3 failure since table.caption and table.thead do not work for nodes added by appendChild
        - fix http://bugs.webkit.org/show_bug.cgi?id=16659
          Acid3 expects HTMLTableElement.rows to include a <tr> element that is an immediate child of the <table>

        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.

        - test for http://bugs.webkit.org/show_bug.cgi?id=16657
          Acid3 failure since table.caption and table.thead do not work for nodes added by appendChild
        - test for http://bugs.webkit.org/show_bug.cgi?id=16659
          Acid3 expects HTMLTableElement.rows to include a <tr> element that is an immediate child of the <table>

        * 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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29101 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago - try to fix Qt build
darin@apple.com [Thu, 3 Jan 2008 01:11:10 +0000 (01:11 +0000)]
    - try to fix Qt build

        * wtf/unicode/qt4/UnicodeQt4.h:
        (WTF::Unicode::foldCase): Add some missing const.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29100 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Alice and Tim.
darin@apple.com [Thu, 3 Jan 2008 00:58:41 +0000 (00:58 +0000)]
    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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29099 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Anders.
darin@apple.com [Thu, 3 Jan 2008 00:32:19 +0000 (00:32 +0000)]
    Reviewed by Anders.

        - http://bugs.webkit.org/show_bug.cgi?id=16712
          change StringImpl to take and return PassRefPtr instead of raw pointers

        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
        (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.
        * 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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29098 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Oliver Hunt.
timothy@apple.com [Thu, 3 Jan 2008 00:07:38 +0000 (00:07 +0000)]
    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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29097 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoJavaScriptCore:
alice.liu@apple.com [Wed, 2 Jan 2008 23:53:46 +0000 (23:53 +0000)]

        Reviewed by Sam Weinig.

        need to export ASCIICType.h for use in DRT

        * JavaScriptCore.vcproj/WTF/WTF.vcproj:
        * wtf/ASCIICType.h:


        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:
        using isupper will cause an assertion for inputs outside of ascii range. use isASCIIUpper instead.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29096 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago * Scripts/do-webcore-rename: Some more name change plans.
darin@apple.com [Wed, 2 Jan 2008 23:46:09 +0000 (23:46 +0000)]
    * Scripts/do-webcore-rename: Some more name change plans.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29095 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
alice.liu@apple.com [Wed, 2 Jan 2008 22:58:50 +0000 (22:58 +0000)]

        Reviewed by Maciej.

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

        * platform/ScrollView.h:
        * platform/win/ScrollViewWin.cpp:
         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:
        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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29094 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Maciej.
alice.liu@apple.com [Wed, 2 Jan 2008 22:37:42 +0000 (22:37 +0000)]
    Reviewed by Maciej.

        Added handling for page up and page down in EventSender

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29093 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago * platform/win/Skipped: Added some more failures.
aroben@apple.com [Wed, 2 Jan 2008 22:23:17 +0000 (22:23 +0000)]
    * platform/win/Skipped: Added some more failures.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29092 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Beth Dakin.
weinig@apple.com [Wed, 2 Jan 2008 21:32:39 +0000 (21:32 +0000)]
    Reviewed by Beth Dakin.

        Cleanup error_object.h/cpp.

        * kjs/JSGlobalObject.cpp:
        * kjs/error_object.cpp:
        * kjs/error_object.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29091 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Fixed a typo (pointed out in review, but I somehow missed it at first).
ap@webkit.org [Wed, 2 Jan 2008 21:07:26 +0000 (21:07 +0000)]
    Fixed a typo (pointed out in review, but I somehow missed it at first).

        * editing/EditorCommand.cpp: (WebCore::executeDelete):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29090 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Get the Windows bots closer to green
aroben@apple.com [Wed, 2 Jan 2008 20:24:23 +0000 (20:24 +0000)]
    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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29089 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Darin.
ap@webkit.org [Wed, 2 Jan 2008 19:14:00 +0000 (19:14 +0000)]
    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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29088 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Darin.
ap@webkit.org [Wed, 2 Jan 2008 18:31:13 +0000 (18:31 +0000)]
    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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29087 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Darin.
ap@webkit.org [Wed, 2 Jan 2008 17:34:41 +0000 (17:34 +0000)]
    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:
        * loader/FrameLoader.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29086 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoAutotools build fix.
mrowe@apple.com [Wed, 2 Jan 2008 16:01:11 +0000 (16:01 +0000)]
Autotools build fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29085 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoAdd missing dependencies to some GNUmakefile.am rules.
mrowe@apple.com [Wed, 2 Jan 2008 15:14:24 +0000 (15:14 +0000)]
Add missing dependencies to some GNUmakefile.am rules.

Rubber-stamped by Alp Toker.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29084 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoAutotools fix. Add missing dependency on grammar.y.
mrowe@apple.com [Wed, 2 Jan 2008 14:42:23 +0000 (14:42 +0000)]
Autotools fix.  Add missing dependency on grammar.y.

Rubber-stamped by Alp Toker.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29083 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoAutotools build fix. Make can expand $@ to any of the targets for the rule,
mrowe@apple.com [Wed, 2 Jan 2008 14:42:02 +0000 (14:42 +0000)]
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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29082 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2008-01-02 Luca Bruno <lethalman88@gmail.com>
alp@webkit.org [Wed, 2 Jan 2008 12:56:30 +0000 (12:56 +0000)]
2008-01-02  Luca Bruno  <lethalman88@gmail.com>

        Reviewed by Darin.

        [GTK] run-launcher sets wrong LD_LIBRARY_PATH

        * Scripts/run-launcher:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29081 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2008-01-02 Luca Bruno <lethalman88@gmail.com>
alp@webkit.org [Wed, 2 Jan 2008 10:25:57 +0000 (10:25 +0000)]
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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29080 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2008-01-02 Alp Toker <alp@atoker.com>
alp@webkit.org [Wed, 2 Jan 2008 09:07:31 +0000 (09:07 +0000)]
2008-01-02  Alp Toker  <alp@atoker.com>

        GTK+ autotools build fix. Track changes in r29073.

        * GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29079 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago - fix release build
darin@apple.com [Wed, 2 Jan 2008 07:03:21 +0000 (07:03 +0000)]
    - 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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29078 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Scripting MIME Types application/ecmascript, application/javascript not viewable
ddkilzer@apple.com [Wed, 2 Jan 2008 06:59:10 +0000 (06:59 +0000)]
    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

        * 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().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29077 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoFix wx build
weinig@apple.com [Wed, 2 Jan 2008 06:54:37 +0000 (06:54 +0000)]
Fix wx build

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29076 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoJavaScriptCore:
darin@apple.com [Wed, 2 Jan 2008 06:44:22 +0000 (06:44 +0000)]

        Reviewed by Eric.

        - fix for http://bugs.webkit.org/show_bug.cgi?id=16695
          JSC allows non-identifier codepoints in identifiers (affects Acid3)

        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.

        - test for http://bugs.webkit.org/show_bug.cgi?id=16695
          JSC allows non-identifier codepoints in identifiers (affects Acid3)

        * 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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29075 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoFix a stray executable bit.
darin@apple.com [Wed, 2 Jan 2008 06:38:08 +0000 (06:38 +0000)]
Fix a stray executable bit.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29074 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
weinig@apple.com [Wed, 2 Jan 2008 05:46:41 +0000 (05:46 +0000)]

        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:
        * bindings/js/kjs_binding.cpp:
        * 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:
        * dom/DOMCoreException.idl:
        * dom/Event.h:
        * dom/EventException.h: Copied from WebCore/dom/DOMCoreException.h.
        * dom/EventException.idl: Copied from WebCore/dom/DOMCoreException.idl.
        * dom/EventTargetNode.cpp:
        * dom/ExceptionBase.cpp: Copied from WebCore/dom/DOMCoreException.cpp.
        * dom/ExceptionBase.h: Copied from WebCore/dom/DOMCoreException.h.
        * dom/ExceptionCode.cpp:
        * dom/ExceptionCode.h:
        * dom/Range.cpp:
        * dom/RangeException.h:
        * dom/RangeException.idl:
        * page/DOMWindow.idl:
        * svg/SVGColor.cpp:
        * svg/SVGException.h:
        * svg/SVGException.idl:
        * svg/SVGLocatable.cpp:
        * xml/XMLHttpRequest.cpp:
        * xml/XMLHttpRequest.h:
        * xml/XMLHttpRequestException.h: Copied from WebCore/dom/DOMCoreException.h.
        * xml/XMLHttpRequestException.idl: Copied from WebCore/dom/DOMCoreException.idl.
        * xml/XPathEvaluator.h:
        * xml/XPathException.h: Copied from WebCore/dom/DOMCoreException.h.
        * xml/XPathException.idl: Copied from WebCore/dom/DOMCoreException.idl.
        * xml/XPathParser.cpp:
        * xml/XPathResult.cpp:


        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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29073 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Dan.
ddkilzer@apple.com [Wed, 2 Jan 2008 00:38:18 +0000 (00:38 +0000)]
    Reviewed by Dan.

        - fix http://bugs.webkit.org/show_bug.cgi?id=16700
          Fix -[WebDefaultPolicyDelegate webView:decidePolicyForMIMEType:request:frame:decisionListener:]

        * DefaultDelegates/WebDefaultPolicyDelegate.m: Check return value of
        -[NSFileManager fileExistsAtPath:isDirectory:] before using the value
        of isDirectory.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29072 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Remove JSDomExceptionConstructor.lut.h from clean step
weinig@apple.com [Tue, 1 Jan 2008 20:44:05 +0000 (20:44 +0000)]
    Remove JSDomExceptionConstructor.lut.h from clean step
        as it no longer exists.

        * GNUmakefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29071 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago - Windows build fix
mitz@apple.com [Tue, 1 Jan 2008 20:42:46 +0000 (20:42 +0000)]
    - Windows build fix

        * WebCore.vcproj/WebCore.vcproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29070 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoAttempt to fix the Windows bots until they can get the Windows SDK installed.
pewtermoose@webkit.org [Tue, 1 Jan 2008 19:52:11 +0000 (19:52 +0000)]
Attempt to fix the Windows bots until they can get the Windows SDK installed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29069 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago - rolled scope chain optimization out; it was breaking the world
darin@apple.com [Tue, 1 Jan 2008 19:47:21 +0000 (19:47 +0000)]
    - rolled scope chain optimization out; it was breaking the world

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29068 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Geoff.
darin@apple.com [Tue, 1 Jan 2008 19:35:37 +0000 (19:35 +0000)]
    Reviewed by Geoff.

        - http://bugs.webkit.org/show_bug.cgi?id=16685
          eliminate List::empty() to cut down on PIC branches

        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
        (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
        (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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29067 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoJavaScriptCore:
darin@apple.com [Tue, 1 Jan 2008 19:16:06 +0000 (19:16 +0000)]

        Reviewed by Geoff.

        - fix http://bugs.webkit.org/show_bug.cgi?id=16648
          REGRESSION (r28165): Yuku.com navigation prints "jsRegExpExecute failed with result -2"
          <rdar://problem/5646486> REGRESSION (r28165): Layout test fast/regex/test1 fails intermittently

        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.

        - updated test results for https://bugs.webkit.org/show_bug.cgi?id=16648

        * 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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29066 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Geoff.
darin@apple.com [Tue, 1 Jan 2008 19:13:40 +0000 (19:13 +0000)]
    Reviewed by Geoff.

        - http://bugs.webkit.org/show_bug.cgi?id=16683
          speed up function calls by making ScopeChain::push cheaper

        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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29065 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Darin Adler.
ggaren@apple.com [Tue, 1 Jan 2008 18:57:03 +0000 (18:57 +0000)]
    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().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29064 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Try again to fix the builds
weinig@apple.com [Tue, 1 Jan 2008 18:05:45 +0000 (18:05 +0000)]
    Try again to fix the builds

        * DerivedSources.make:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29063 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Fix non-mac builds.
weinig@apple.com [Tue, 1 Jan 2008 17:45:22 +0000 (17:45 +0000)]
    Fix non-mac builds.

        * WebCore.pro:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCoreSources.bkl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29062 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Alexey.
eric@webkit.org [Tue, 1 Jan 2008 10:40:57 +0000 (10:40 +0000)]
    Reviewed by Alexey.

        Don't replace \ with / in data: urls

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

        * platform/KURL.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29061 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2008-01-01 Alp Toker <alp@atoker.com>
alp@webkit.org [Tue, 1 Jan 2008 09:05:54 +0000 (09:05 +0000)]
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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29060 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Oliver.
darin@apple.com [Tue, 1 Jan 2008 08:43:59 +0000 (08:43 +0000)]
    Reviewed by Oliver.

        - http://bugs.webkit.org/show_bug.cgi?id=16684
          eliminate debugger overhead from function body execution

        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.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29059 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
weinig@apple.com [Tue, 1 Jan 2008 03:04:16 +0000 (03:04 +0000)]

        Reviewed by Darin.

        Patch for http://bugs.webkit.org/show_bug.cgi?id=16637
        Acid3 expects ExeceptionCode constants to be defined on DOMException objects

        - Make DOMException a real JS object.

        Test: fast/dom/DOMException/prototype-object.html

        * DerivedSources.make:
        * WebCore.xcodeproj/project.pbxproj:

        This is no longer needed as the autogenerated classes now includes the
        * bindings/js/JSDOMExceptionConstructor.cpp: Removed.
        * bindings/js/JSDOMExceptionConstructor.h: Removed.

        Create on demand and use the new class for DOMExceptions.
        * bindings/js/kjs_binding.cpp:

        Remove no longer needed custom constructor getter.
        * bindings/js/kjs_window.cpp:

        Don't expose DOMCoreException as the name of class by special casing
        the user visible class name to be DOMException.
        * bindings/scripts/CodeGeneratorJS.pm:

        The DOMException class/file needs to be named DOMCoreException because there is
        name conflict with one of the Objective-C bindings classes.  It should be renamed
        to DOMException when the Objective-C bindings are moved into WebKit.
        * dom/DOMCoreException.cpp: Added.
        * dom/DOMCoreException.h: Added.
        * dom/DOMCoreException.idl: Added.
        * page/DOMWindow.idl:


        Reviewed by Darin.

        Test for http://bugs.webkit.org/show_bug.cgi?id=16637
        Acid3 expects ExeceptionCode constants to be defined on DOMException objects

        * fast/dom/DOMException: Added.
        * fast/dom/DOMException/prototype-object-expected.txt: Added.
        * fast/dom/DOMException/prototype-object.html: Added.
        * fast/dom/DOMException/resources: Added.
        * fast/dom/DOMException/resources/prototype-object.js: Added.
        * fast/dom/Window/window-properties-expected.txt:
        * http/tests/xmlhttprequest/connection-error-sync-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29058 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Re-enable querySelector and querySelectorAll and touch the necessary files...
weinig@apple.com [Tue, 1 Jan 2008 02:01:46 +0000 (02:01 +0000)]
    Re-enable querySelector and querySelectorAll and touch the necessary files to not
        kill the windows build.

        * WebCore.vcproj/build-generated-files.sh:
        * bindings/scripts/CodeGeneratorCOM.pm:
        * dom/Document.idl:
        * dom/Element.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29057 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago * fast/dom/HTMLFormElement/elements-not-in-document-expected.txt: Updated...
darin@apple.com [Tue, 1 Jan 2008 01:45:15 +0000 (01:45 +0000)]
    * fast/dom/HTMLFormElement/elements-not-in-document-expected.txt: Updated results to expect success.
        I accdidentally landed an expected failure instead!

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29056 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago - fix Windows build
darin@apple.com [Tue, 1 Jan 2008 01:37:55 +0000 (01:37 +0000)]
    - fix Windows build

        * dom/Document.idl: Temporarily disable querySelector and querySelectorAll, since they are showing
        up as pure virtual functions. Sam can fix this later.
        * dom/Element.idl: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29055 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
mitz@apple.com [Tue, 1 Jan 2008 01:32:00 +0000 (01:32 +0000)]

        Reviewed by Darin Adler.

        - fix http://bugs.webkit.org/show_bug.cgi?id=14134
          <rdar://problem/5655160> REGRESSION (r25353): Whitespace nodes ignored between inline list items

        Test: fast/dynamic/create-renderer-for-whitespace-only-text.html

        * dom/Node.cpp:
        (WebCore::Node::attach): Added code to check if this node's renderer
        has become the "previous renderer" of any sibling text node, and if so,
        ensure that that node gets a renderer if it now needs one.
        (WebCore::Node::createRendererIfNeeded): Removed the assertion that the
        node is not attached.


        Reviewed by Darin Adler.

        - test and updated results for http://bugs.webkit.org/show_bug.cgi?id=14134
          <rdar://problem/5655160> REGRESSION (r25353): Whitespace nodes ignored between inline list items

        * fast/dynamic/create-renderer-for-whitespace-only-text.html: Added.
        * fast/dynamic/style-access-late-stylesheet-load-expected.txt:
        * platform/mac-leopard/fast/dynamic: Added.
        * platform/mac-leopard/fast/dynamic/create-renderer-for-whitespace-only-text-expected.checksum: Added.
        * platform/mac-leopard/fast/dynamic/create-renderer-for-whitespace-only-text-expected.png: Added.
        * platform/mac/editing/deleting/delete-block-merge-contents-001-expected.txt:
        * platform/mac/editing/deleting/delete-block-merge-contents-019-expected.txt:
        * platform/mac/editing/deleting/delete-block-merge-contents-020-expected.txt:
        * platform/mac/editing/inserting/editable-html-element-expected.txt:
        * platform/mac/editing/inserting/editing-empty-divs-expected.txt:
        * platform/mac/editing/pasteboard/paste-TIFF-expected.txt:
        * platform/mac/editing/selection/4983858-expected.txt:
        * platform/mac/editing/selection/5136696-expected.txt:
        * platform/mac/editing/selection/drag-to-contenteditable-iframe-expected.txt:
        * platform/mac/editing/style/remove-underline-across-paragraph-expected.txt:
        * platform/mac/editing/style/remove-underline-across-paragraph-in-bold-expected.txt:
        * platform/mac/editing/style/remove-underline-after-paragraph-expected.txt:
        * platform/mac/editing/style/remove-underline-after-paragraph-in-bold-expected.txt:
        * platform/mac/editing/style/remove-underline-from-stylesheet-expected.txt:
        * platform/mac/fast/dynamic/create-renderer-for-whitespace-only-text-expected.txt: Added.
        * platform/mac/fast/dynamic/move-node-with-selection-expected.txt:
        * platform/mac/fast/forms/input-align-expected.txt:
        * platform/mac/fast/inline/positionedLifetime-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29054 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
darin@apple.com [Tue, 1 Jan 2008 01:17:22 +0000 (01:17 +0000)]

        Reviewed by Mitz.

        - fix http://bugs.webkit.org/show_bug.cgi?id=16641
          Acid3 reveals HTMLFormElement.elements fails to update when element name changes

        Test: fast/dom/HTMLFormElement/elements-not-in-document.html

        This was a bug specific to forms that are not in the document tree.
        The fix was to change the code to increment the document version number to match
        up with other document change tracking. Maybe at some point we can clean these up
        so we don't have so many competing change notification systems.

        * dom/ContainerNode.cpp:
        (WebCore::ContainerNode::replaceChild): Removed bogus comment.
        (WebCore::ContainerNode::addChild): Added an explicit incDOMTreeVersion
        call here, since this code path bypasses the subtree-modified event code.

        * dom/Element.cpp:
        (WebCore::Element::setAttribute): Remove the inDocument() check -- not all HTML
        collections are for things in the document.
        (WebCore::Element::setAttributeMap): Ditto.

        * dom/EventTargetNode.cpp:
        (WebCore::EventTargetNode::dispatchSubtreeModifiedEvent): Added a call to
        incDOMTreeVersion here; covers most cases of tree structure changes.

        * dom/Node.cpp:
        (WebCore::Node::attach): Remove call to incDOMTreeVersion -- creating a renderer
        has nothing to do with changes to the DOM tree!
        (WebCore::Node::detach): Ditto.

        * html/HTMLFormElement.cpp:
        (WebCore::HTMLFormElement::registerFormElement): Remove call to incDOMTreeVersion.
        This is handled at a lower level and doesn't need to be here.
        (WebCore::HTMLFormElement::removeFormElement): Ditto.


        Reviewed by Mitz.

        - test for http://bugs.webkit.org/show_bug.cgi?id=16641
          Acid3 reveals HTMLFormElement.elements fails to update when element name changes

        * fast/dom/HTMLFormElement: Added.
        * fast/dom/HTMLFormElement/elements-not-in-document-expected.txt: Added.
        * fast/dom/HTMLFormElement/elements-not-in-document.html: Added.
        * fast/dom/HTMLFormElement/resources: Added.
        * fast/dom/HTMLFormElement/resources/TEMPLATE.html: Added.
        * fast/dom/HTMLFormElement/resources/elements-not-in-document.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29053 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Suggested by Antti.
darin@apple.com [Tue, 1 Jan 2008 00:20:19 +0000 (00:20 +0000)]
    Suggested by Antti.

        * Scripts/webkitdirs.pm: Turned off the QuickTime requirement for Windows until
        we get it installed on the build bots.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29052 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
darin@apple.com [Tue, 1 Jan 2008 00:18:16 +0000 (00:18 +0000)]

        Reviewed by Darin.

        Patch for http://bugs.webkit.org/show_bug.cgi?id=14994
        Support for MessageEvent and cross-domain messaging

        Test: http/tests/messaging/cross-domain-message-send.html

        * DerivedSources.make:
        * WebCore.pro:
        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * WebCoreSources.bkl:
        * bindings/js/JSDOMWindowCustom.cpp:
        (WebCore::JSDOMWindow::customGetOwnPropertySlot): Allow cross-domain access to the
        postMessage function.
        * bindings/js/JSEventCustom.cpp:
        * dom/Event.cpp:
        * dom/Event.h:
        * dom/EventNames.h: New event name
        * dom/MessageEvent.cpp: Added.
        * dom/MessageEvent.h: Added.
        * dom/MessageEvent.idl: Added.
        * page/DOMWindow.cpp:
        (WebCore::DOMWindow::postMessage): Added.
        * page/DOMWindow.h:
        * page/DOMWindow.idl:


        Reviewed by Darin.

        Test for http://bugs.webkit.org/show_bug.cgi?id=14994
        Support for MessageEvent and cross-domain messaging

        * fast/dom/Window/window-properties-expected.txt:
        * http/tests/messaging: Added.
        * http/tests/messaging/cross-domain-message-send-expected.txt: Added.
        * http/tests/messaging/cross-domain-message-send.html: Added.
        * http/tests/messaging/resources: Added.
        * http/tests/messaging/resources/cross-domain-message-receive.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29051 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Suggested by Antti.
darin@apple.com [Mon, 31 Dec 2007 23:56:16 +0000 (23:56 +0000)]
    Suggested by Antti.

        - turn ENABLE_VIDEO back off for Windows until we can install QuickTime on the build bots

        * WebCore.vcproj/WebCore.vcproj: Removed ENABLE_VIDEO.
        * WebCore.vcproj/build-generated-files.sh: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29050 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-30 Alp Toker <alp@atoker.com>
alp@webkit.org [Mon, 31 Dec 2007 07:48:54 +0000 (07:48 +0000)]
2007-12-30  Alp Toker  <alp@atoker.com>

        Reviewed by Dan Bernstein.

        Silence warning.

        * dom/NodeList.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29049 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Darin.
pewtermoose@webkit.org [Mon, 31 Dec 2007 07:06:01 +0000 (07:06 +0000)]
    Reviewed by Darin.

        Bug 16578: Windows Web Inspector window needs minimum size

        * WebInspectorClient.cpp:
        * WebInspectorClient.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29048 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Sam.
eric@webkit.org [Mon, 31 Dec 2007 07:03:31 +0000 (07:03 +0000)]
    Reviewed by Sam.

        More small cleanup to array_object.cpp

        * kjs/array_object.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29047 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Sam.
eric@webkit.org [Mon, 31 Dec 2007 07:03:05 +0000 (07:03 +0000)]
    Reviewed by Sam.

        Apply wkstyle to array_object.cpp

        * kjs/array_object.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29046 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Sam.
eric@webkit.org [Mon, 31 Dec 2007 07:02:26 +0000 (07:02 +0000)]
    Reviewed by Sam.

        Remove maxInt/minInt, replacing with std:max/min<int>()

        * kjs/array_object.cpp:
        * kjs/operations.cpp:
        * kjs/operations.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29045 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
weinig@apple.com [Mon, 31 Dec 2007 07:02:05 +0000 (07:02 +0000)]

        Reviewed by Oliver Hunt.

        Fix for http://bugs.webkit.org/show_bug.cgi?id=16387
        Variable names can be enumerated across domains

        Test: http/tests/security/cross-frame-access-enumeration.html

        * bindings/js/kjs_window.cpp:
        (KJS::Window::getPropertyNames): Override method to test same-origin policy.
        * bindings/js/kjs_window.h:


        Reviewed by Oliver Hunt.

        Test for http://bugs.webkit.org/show_bug.cgi?id=16387

        * http/tests/security/cross-frame-access-enumeration-expected.txt: Added.
        * http/tests/security/cross-frame-access-enumeration.html: Added.
        * http/tests/security/resources/cross-frame-iframe-for-enumeration-test.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29044 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Sam.
eric@webkit.org [Mon, 31 Dec 2007 07:01:44 +0000 (07:01 +0000)]
    Reviewed by Sam.

        * Scripts/do-webcore-rename: Add a few more planned renames s/(\w+)Imp/\1/

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29043 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago - added missing results file
darin@apple.com [Mon, 31 Dec 2007 06:37:16 +0000 (06:37 +0000)]
    - added missing results file

        * fast/events/event-instanceof-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29042 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
weinig@apple.com [Sun, 30 Dec 2007 22:24:47 +0000 (22:24 +0000)]

        Reviewed by Oliver Hunt.

        Patch for http://bugs.webkit.org/show_bug.cgi?id=10686
        event instanceof MouseEvent throws exception

        Add JS constructors for all the Event types.

        Test: fast/events/event-instanceof.html

        * WebCore.xcodeproj/project.pbxproj:
        * dom/KeyboardEvent.idl:
        * dom/MouseEvent.idl:
        * dom/MutationEvent.idl:
        * dom/OverflowEvent.idl:
        * dom/ProgressEvent.idl:
        * dom/TextEvent.idl:
        * dom/UIEvent.idl:
        * dom/WheelEvent.idl:
        * page/DOMWindow.idl:


        Reviewed by Oliver Hunt.

        Test for http://bugs.webkit.org/show_bug.cgi?id=10686
        event instanceof MouseEvent throws exception

        * fast/dom/Window/window-properties-expected.txt:
        * fast/events/event-instanceof.html: Added.
        * fast/events/resources/event-instanceof.js: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29041 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-30 Alp Toker <alp@atoker.com>
alp@webkit.org [Sun, 30 Dec 2007 17:44:37 +0000 (17:44 +0000)]
2007-12-30  Alp Toker  <alp@atoker.com>

        Build fix for older autoconf versions.

        * configure.ac:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29040 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
ddkilzer@apple.com [Sun, 30 Dec 2007 17:05:52 +0000 (17:05 +0000)]

        Reviewed by Darin.

        - fix http://bugs.webkit.org/show_bug.cgi?id=15359
          JPEG image not shown when height is specified as percentage inside a table

        The problem occurs when a replaced element (image, canvas, etc.) with
        a percent-height attribute is contained by a table cell with an auto-
        or percent-height attribute.  If there are no other conditions to cause
        the table cell's height to expand, an available height of zero will
        always be returned.  In these cases, the intrinsic height of the
        replaced element should be used if it is greater than the available
        height of the table cell.

        Tests: fast/replaced/table-percent-height.html

        * rendering/RenderBox.cpp:


        Reviewed by Darin.

        - test for http://bugs.webkit.org/show_bug.cgi?id=15359
          JPEG image not shown when height is specified as percentage inside a table

        This test was added specifically for this bug.  It tests the height of
        replaced elements (canvas, embed, img, object, button, input, isindex,
        select, textarea ) within table cells.

        * fast/replaced/table-percent-height-expected.txt: Added.
        * fast/replaced/table-percent-height.html: Added.

        These test results were updated because they are more correct.

        * platform/mac/fast/replaced/width100percent-image-expected.checksum: Updated.
        * platform/mac/fast/replaced/width100percent-image-expected.png: Updated.
        * platform/mac/fast/replaced/width100percent-image-expected.txt: Updated.

        Moved tests and results for these two tests from tables/mozilla_expected_failures/bugs
        to tables/mozilla/bugs since these results have been fixed.

        * platform/mac/tables/mozilla/bugs/bug137388-1-expected.checksum: Copied from LayoutTests/platform/mac/tables/mozilla_expected_failures/bugs/bug137388-1-expected.checksum.
        * platform/mac/tables/mozilla/bugs/bug137388-1-expected.png: Copied from LayoutTests/platform/mac/tables/mozilla_expected_failures/bugs/bug137388-1-expected.png.
        * platform/mac/tables/mozilla/bugs/bug137388-1-expected.txt: Copied from LayoutTests/platform/mac/tables/mozilla_expected_failures/bugs/bug137388-1-expected.txt.
        * platform/mac/tables/mozilla/bugs/bug137388-2-expected.checksum: Copied from LayoutTests/platform/mac/tables/mozilla_expected_failures/bugs/bug137388-2-expected.checksum.
        * platform/mac/tables/mozilla/bugs/bug137388-2-expected.png: Copied from LayoutTests/platform/mac/tables/mozilla_expected_failures/bugs/bug137388-2-expected.png.
        * platform/mac/tables/mozilla/bugs/bug137388-2-expected.txt: Copied from LayoutTests/platform/mac/tables/mozilla_expected_failures/bugs/bug137388-2-expected.txt.
        * platform/mac/tables/mozilla_expected_failures/bugs/bug137388-1-expected.checksum: Removed.
        * platform/mac/tables/mozilla_expected_failures/bugs/bug137388-1-expected.png: Removed.
        * platform/mac/tables/mozilla_expected_failures/bugs/bug137388-1-expected.txt: Removed.
        * platform/mac/tables/mozilla_expected_failures/bugs/bug137388-2-expected.checksum: Removed.
        * platform/mac/tables/mozilla_expected_failures/bugs/bug137388-2-expected.png: Removed.
        * platform/mac/tables/mozilla_expected_failures/bugs/bug137388-2-expected.txt: Removed.
        * tables/mozilla/bugs/bug137388-1.html: Copied from LayoutTests/tables/mozilla_expected_failures/bugs/bug137388-1.html.
        * tables/mozilla/bugs/bug137388-2.html: Copied from LayoutTests/tables/mozilla_expected_failures/bugs/bug137388-2.html.
        * tables/mozilla_expected_failures/bugs/bug137388-1.html: Removed.
        * tables/mozilla_expected_failures/bugs/bug137388-2.html: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29039 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-30 Mark Rowe <mrowe@apple.com>
mrowe@apple.com [Sun, 30 Dec 2007 16:21:36 +0000 (16:21 +0000)]
2007-12-30  Mark Rowe  <mrowe@apple.com>

        Unreviewed.  Update WordPress to 2.3.2, which includes security fixes.

        * blog/wp-admin/admin.php:
        * blog/wp-admin/includes/file.php:
        * blog/wp-admin/install.php:
        * blog/wp-admin/setup-config.php:
        * blog/wp-app.php:
        * blog/wp-includes/formatting.php:
        * blog/wp-includes/functions.php:
        * blog/wp-includes/pluggable.php:
        * blog/wp-includes/post.php:
        * blog/wp-includes/query.php:
        * blog/wp-includes/taxonomy.php:
        * blog/wp-includes/version.php:
        * blog/wp-includes/wp-db.php:
        * blog/wp-mail.php:
        * blog/wp-settings.php:
        * blog/xmlrpc.php:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29038 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-30 Luca Bruno <lethalman88@gmail.com>
alp@webkit.org [Sun, 30 Dec 2007 11:52:03 +0000 (11:52 +0000)]
2007-12-30  Luca Bruno  <lethalman88@gmail.com>

        Reviewed by Alp Toker.

        Crash in CURL for empty POST

        We have to set POST even when the data is empty, otherwise cURL will
        hang while waiting for a response.

        * platform/network/curl/ResourceHandleManager.cpp
        (ResourceHandleManager::setupPOST): allow empty POST

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29037 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Sam.
eric@webkit.org [Sun, 30 Dec 2007 08:38:39 +0000 (08:38 +0000)]
    Reviewed by Sam.

        Update Number.toString to properly throw exceptions.
        Cleanup code in Number.toString implementation.

        * kjs/number_object.cpp:
        * kjs/object.cpp:
        (KJS::Error::create): Remove bogus debug lines.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29036 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-29 Jan Michael Alonzo <jmalonzo@unpluggable.com>
alp@webkit.org [Sun, 30 Dec 2007 06:47:16 +0000 (06:47 +0000)]
2007-12-29  Jan Michael Alonzo  <jmalonzo@unpluggable.com>

        Reviewed by Alp Toker.

        Enable Database, XPath and XSLT features by default

        Move all SVG related stuff inside SVG block. Don't include
        SVGNames, SVGElementFactory, and XLinkNames when SVG is not

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29035 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Darin.
ap@webkit.org [Sun, 30 Dec 2007 06:13:40 +0000 (06:13 +0000)]
    Reviewed by Darin.

        FCKEditor: Images disappear on drag/drop and copy/paste

        Test: editing/pasteboard/drag-image-in-about-blank-frame.html

        * editing/markup.cpp: (WebCore::createFragmentFromMarkup): Don't use "about:blank" as a
        base URL, just like we don't use an empty one.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29034 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-29 Jan Michael Alonzo <jmalonzo@unpluggable.com>
alp@webkit.org [Sun, 30 Dec 2007 03:33:44 +0000 (03:33 +0000)]
2007-12-29  Jan Michael Alonzo  <jmalonzo@unpluggable.com>

        Reviewed by Alp Toker.

        autotools update and fixes

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29033 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Oliver.
darin@apple.com [Sun, 30 Dec 2007 00:33:20 +0000 (00:33 +0000)]
    Reviewed by Oliver.

        - fix http://bugs.webkit.org/show_bug.cgi?id=16663
          leak bot shows createCStringFromNPVariant result leaking

        * DumpRenderTree/TestNetscapePlugIn.subproj/PluginObject.cpp:
        (pluginInvoke): Added a missing free.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29032 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoReviewed by Alexey.
zimmermann@webkit.org [Sat, 29 Dec 2007 16:06:23 +0000 (16:06 +0000)]
Reviewed by Alexey.

Respect horiz-origin-x / horiz-origin-y / vert-origin-x / vert-origin-y properties when drawing SVG Fonts.
(Fixes fonts-elem-05-t.svg in a --svg-fonts build)

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29031 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-29 Alp Toker <alp@atoker.com>
alp@webkit.org [Sat, 29 Dec 2007 14:00:56 +0000 (14:00 +0000)]
2007-12-29  Alp Toker  <alp@atoker.com>

        Fix typo.

        * hosted/sunspider.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29030 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoReviewed by Oliver.
zimmermann@webkit.org [Sat, 29 Dec 2007 13:50:15 +0000 (13:50 +0000)]
Reviewed by Oliver.

Further SVG Font work. Parse all <glyph> attributes, using SVGGlyphElement::buildGlyphIdentifier.
SVGFontElement::collectGlyphs() now uses this method. Per character advance values work well now.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29029 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-29 Alp Toker <alp@atoker.com>
alp@webkit.org [Sat, 29 Dec 2007 13:03:59 +0000 (13:03 +0000)]
2007-12-29  Alp Toker  <alp@atoker.com>

        Reviewed by Maciej.

        Provide the current URL in the SunSpider results page. Useful for
        copying benchmark results out of browsers without a location bar or
        where the location bar can't handle long URLs.

        * hosted/sunspider-results.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29028 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoReviewed by Oliver.
zimmermann@webkit.org [Sat, 29 Dec 2007 12:50:57 +0000 (12:50 +0000)]
Reviewed by Oliver.

Refactor SVGAnimationElement::parseValues into a generic function parseDelimitedString.
SVGAnimationElement needs to parse semicolon-seperated strings, SVGGlyphElement comma-seperated.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29027 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoReviewed by Maciej.
zimmermann@webkit.org [Sat, 29 Dec 2007 11:59:50 +0000 (11:59 +0000)]
Reviewed by Maciej.

Add new helper structure SVGFontData - FontData holds this object as OwnPtr.
Store several attribute values there (horiz-adv-x, horiz-origin-x etc..)

To optimize for the common case ('FontData' used for HTML rendering) it feels
better to hold one OwnPtr in FontData, than several floats.

Parse all <font> attributes in SVGFontFaceElement::createFontData.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29026 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago - try to fix Windows and WX builds (broken by SVG Fonts check-in)
darin@apple.com [Sat, 29 Dec 2007 07:42:17 +0000 (07:42 +0000)]
    - try to fix Windows and WX builds (broken by SVG Fonts check-in)

        * platform/graphics/win/FontWin.cpp:
        (WebCore::Font::drawGlyphs): Pass font size to FontData::ascent, now that it requires it.
        I can't see how it can be right to require the font size for ascent and descent,
        but not for other metrics functions in FontData.

        * platform/graphics/wx/FontWx.cpp:
        (WebCore::Font::drawGlyphs): Pass font size to FontData::ascent and FontData::descent.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29025 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Mitz.
ap@webkit.org [Sat, 29 Dec 2007 07:35:17 +0000 (07:35 +0000)]
    Reviewed by Mitz.

        fast/xpath/namespace-vs-predicate.xhtml fails unexpectedly in no-SVG build

        * fast/xpath/namespace-vs-predicate.xhtml: Use (hopefully) correct code to remove an SVG element.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29024 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago - checked in these files that were supposed to be part of
darin@apple.com [Sat, 29 Dec 2007 07:27:29 +0000 (07:27 +0000)]
    - checked in these files that were supposed to be part of
          the previous check-in, but somehow weren't

        * fast/js/number-tofixed-expected.txt: Updated.
        * fast/js/number-toprecision-expected.txt: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29023 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-28 Eric Seidel <eric@webkit.org>
eric@webkit.org [Sat, 29 Dec 2007 05:51:32 +0000 (05:51 +0000)]
2007-12-28  Eric Seidel  <eric@webkit.org>

        Reviewed by Oliver.

        ASSERT when debugging via Drosera due to missed var lookup optimization.

        No test case possible.

        * kjs/nodes.cpp:
        * kjs/nodes.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29022 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Mark Rowe.
mitz@apple.com [Sat, 29 Dec 2007 04:19:30 +0000 (04:19 +0000)]
    Reviewed by Mark Rowe.

        - fix http://bugs.webkit.org/show_bug.cgi?id=16650
          <rdar://problem/5664872> REGRESSION (r28278-r28314): ATSUI uses LTR writing direction for all text runs

        Covered by existing pixel tests.

        * platform/graphics/mac/FontMac.mm:
        (WebCore::Font::drawComplexText): Corrected to maintain the style
        information in the adjusted text run when passing it to
        ATSUILayoutParameters. Prior to r28298 the style was passed separately.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29021 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Oliver.
eric@webkit.org [Sat, 29 Dec 2007 04:01:13 +0000 (04:01 +0000)]
    Reviewed by Oliver.

        Fix (-0).toFixed() and re-factor a little
        Fix (-0).toExponential() and printing of trailing 0s in toExponential
        Fix toPrecision(nan) handling

        * kjs/number_object.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29020 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoBuild fix, not reviewed.
zimmermann@webkit.org [Sat, 29 Dec 2007 03:27:31 +0000 (03:27 +0000)]
Build fix, not reviewed.
Add wtf/OwnPtr.h include, to fix --svg-fonts build.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29019 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Sam.
eric@webkit.org [Sat, 29 Dec 2007 03:17:11 +0000 (03:17 +0000)]
    Reviewed by Sam.

        More changes to make number code readable

        * kjs/number_object.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29018 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Sam.
eric@webkit.org [Sat, 29 Dec 2007 03:16:43 +0000 (03:16 +0000)]
    Reviewed by Sam.

        More small cleanups to toPrecision

        * kjs/number_object.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29017 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Sam.
eric@webkit.org [Sat, 29 Dec 2007 03:16:13 +0000 (03:16 +0000)]
    Reviewed by Sam.

        More small attempts to make number code readable

        * kjs/number_object.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29016 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-28 Eric Seidel <eric@webkit.org>
eric@webkit.org [Sat, 29 Dec 2007 03:15:39 +0000 (03:15 +0000)]
2007-12-28  Eric Seidel  <eric@webkit.org>

        Reviewed by Sam.

        Break out callAsFunction implementations into static functions

        * kjs/number_object.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29015 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Sam.
eric@webkit.org [Sat, 29 Dec 2007 03:14:52 +0000 (03:14 +0000)]
    Reviewed by Sam.

        Apply wkstyle/astyle and fix placement of *

        * kjs/number_object.cpp:
        * kjs/object.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29014 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-28 Christian Dywan <christian@twotoasts.de>
alp@webkit.org [Sat, 29 Dec 2007 02:48:38 +0000 (02:48 +0000)]
2007-12-28  Christian Dywan  <christian@twotoasts.de>

        Reviewed by Alp Toker.

        [GTK] webkit_web_view_has_selection returns the opposite result

        Invert the return value.

        * WebView/webkitwebview.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29013 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoReviewed by Eric.
zimmermann@webkit.org [Sat, 29 Dec 2007 02:41:53 +0000 (02:41 +0000)]
Reviewed by Eric.

Fixes: http://bugs.webkit.org/show_bug.cgi?id=10649 (WebKit SVG needs SVG Fonts support)
Begin implementation of SVG Fonts module. Basic documents using SVG Fonts already work.
Only local, in-document fonts who declare their glyphs using the <glyph d="..."> path syntax
are supported. (<glyph> containing arbitary SVG content as child elements, not supported yet).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29012 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Alexey Proskuryakov.
mitz@apple.com [Fri, 28 Dec 2007 07:50:44 +0000 (07:50 +0000)]
    Reviewed by Alexey Proskuryakov.

        - fix http://bugs.webkit.org/show_bug.cgi?id=16628
          ASSERTION FAILED: m_resizeLayer (running layout tests)

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::~RenderLayer): Avoid calling
        EventHandler::resizeLayerDestroyed() if the document is being destroyed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29011 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
mitz@apple.com [Fri, 28 Dec 2007 03:22:42 +0000 (03:22 +0000)]

        Reviewed by Oliver Hunt.

        - fix http://bugs.webkit.org/show_bug.cgi?id=16603
          <rdar://problem/5664199> Crash when resizing text field

        Test: fast/layers/resize-layer-deletion-crash.html

        The event handler has only a weak reference to the layer that is
        currently in resize mode, so it is the layer's responsibility to let
        the event handler know if it has been destroyed while in that mode.

        * page/EventHandler.cpp:
        (WebCore::EventHandler::resizeLayerDestroyed): Added. Resets
        m_resizeLayer to 0.
        * page/EventHandler.h:
        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::~RenderLayer): Added a call to
        EventHandler::resizeLayerDestroyed() if the layer is in resize mode.


        Reviewed by Oliver Hunt.

        - test for http://bugs.webkit.org/show_bug.cgi?id=16603
          <rdar://problem/5664199> Crash when resizing text field

        * fast/layers/resize-layer-deletion-crash-expected.txt: Added.
        * fast/layers/resize-layer-deletion-crash.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29010 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Alice Liu.
weinig@apple.com [Fri, 28 Dec 2007 01:25:22 +0000 (01:25 +0000)]
    Reviewed by Alice Liu.

        * platform/win/Skipped: Remove xss-DENIED-from-data-url-sub-frame-to-data-url-sub-frame.html
        since it is now passing.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29009 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
weinig@apple.com [Fri, 28 Dec 2007 00:57:51 +0000 (00:57 +0000)]

        Reviewed by Sam Weinig.


        The same-origin check was missing in the implementation of
        setTimeout, setInterval, addEventListener, and removeEventListener.

        Suppose <http://www.badguy.com/> contains an iframe to
        <http://www.goodguy.com/>. Now www.badguy.com can steal
        www.goodguy.com cookies by running this code:

        setTimeout.call(frames[0], "alert(document.cookie)", 1000);

        This patch changes the behavior so that setTimeout to does
        nothing and returns an undefined value if the caller is not
        permitted to script the window whose setTimeout method is being
        called. The same applies to setInterval, addEventListener, and

        Tests: http/tests/security/cross-frame-access-call.html

        * bindings/js/kjs_window.cpp:


        Reviewed by Sam Weinig.

        Tests for http://bugs.webkit.org/show_bug.cgi?id=16539

        * http/tests/security/cross-frame-access-call-expected.txt: Added.
        * http/tests/security/cross-frame-access-call.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29008 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years agoWebCore:
mitz@apple.com [Fri, 28 Dec 2007 00:11:46 +0000 (00:11 +0000)]

        Reviewed by Dave Hyatt and Sam Weinig.

        - fix <rdar://problem/5605937> Inspector: Disclosure triangle not drawn in node's properties panel until click

        Test: fast/layers/add-layer-with-nested-stacking.html
        Test: fast/layers/remove-layer-with-nested-stacking.html

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::addChild): If the new child is overflow-only
        but has children of its own, dirty the stacking context's z-order lists
        since the grandchildren might need to be in them.
        (WebCore::RenderLayer::removeChild): Similarly for the old child.


        Reviewed by Dave Hyatt and Sam Weinig.

        - pixel tests for <rdar://problem/5605937> Inspector: Disclosure triangle not drawn in node's properties panel until click

        * fast/layers/add-layer-with-nested-stacking.html: Added.
        * fast/layers/remove-layer-with-nested-stacking.html: Added.
        * platform/mac/fast/layers/add-layer-with-nested-stacking-expected.checksum: Added.
        * platform/mac/fast/layers/add-layer-with-nested-stacking-expected.png: Added.
        * platform/mac/fast/layers/add-layer-with-nested-stacking-expected.txt: Added.
        * platform/mac/fast/layers/remove-layer-with-nested-stacking-expected.checksum: Added.
        * platform/mac/fast/layers/remove-layer-with-nested-stacking-expected.png: Added.
        * platform/mac/fast/layers/remove-layer-with-nested-stacking-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29007 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago2007-12-27 Jan Michael Alonzo <jmalonzo@unpluggable.com>
alp@webkit.org [Thu, 27 Dec 2007 23:14:27 +0000 (23:14 +0000)]
2007-12-27  Jan Michael Alonzo  <jmalonzo@unpluggable.com>

        Reviewed by Alp Toker.

        [GTK] Check for deprecated API use (G_DISABLE_DEPRECATED etc.)

        Added the flags mentioned in the bug only when doing 'debug'
        builds. -DGST_DISABLE_DEPRECATED only added when video is enabled.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29006 268f45cc-cd09-0410-ab3c-d52691b4dbfc

13 years ago Reviewed by Sam.
eric@webkit.org [Thu, 27 Dec 2007 22:20:18 +0000 (22:20 +0000)]
    Reviewed by Sam.

        ASSERT in JavaScriptCore while viewing WICD test case

        * kjs/nodes.cpp:
        (KJS::ForInNode::execute): move KJS_CHECK_EXCEPTION to proper place

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@29005 268f45cc-cd09-0410-ab3c-d52691b4dbfc