WebKit.git
12 years ago2009-05-22 Antonio Gomes <antonio.gomes@openbossa.org>
kov@webkit.org [Fri, 22 May 2009 19:40:08 +0000 (19:40 +0000)]
2009-05-22  Antonio Gomes   <antonio.gomes@openbossa.org>

        Reviewed by Gustavo Noronha.

        Make Gtk build not bail out if gtk-doc-tools is not installed.

        Warning message dumped instead.

        * autogen.sh:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44054 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-22 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>
kov@webkit.org [Fri, 22 May 2009 19:07:10 +0000 (19:07 +0000)]
2009-05-22  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Reviewed by Xan Lopez.

        Add big warnings about the glib unicode backend being slow and
        incomplete, since it is a work in progress.

        * autotools/webkit.m4:
        * configure.ac:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44053 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Fri, 22 May 2009 19:06:07 +0000 (19:06 +0000)]
2009-05-21  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Maciej Stachowiak.

        https://bugs.webkit.org/show_bug.cgi?id=24818
        [GTK] http auth dialog pops up twice after a cancelled atempt

        Workaround libsoup bug
        http://bugzilla.gnome.org/show_bug.cgi?id=583462, which made
        libsoup emit the 'authenticate' signal more often than it was
        expected.

        We'll be able to remove this when we depend on libsoup > 2.26.2.

        * webkit/webkitsoupauthdialog.c:
        (session_authenticate):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44052 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-22 Peter Kasting <pkasting@google.com>
pkasting@chromium.org [Fri, 22 May 2009 18:39:04 +0000 (18:39 +0000)]
2009-05-22  Peter Kasting  <pkasting@google.com>

        Reviewed by Darin Fisher.

        https://bugs.webkit.org/show_bug.cgi?id=25709 part one
        Indent contents of namespaces in header files, to comply with the
        style guide.  No functional change.

        * platform/image-decoders/ImageDecoder.h:
        (WebCore::RGBA32Buffer::):
        (WebCore::RGBA32Buffer::RGBA32Buffer):
        (WebCore::RGBA32Buffer::clear):
        (WebCore::RGBA32Buffer::bytes):
        (WebCore::RGBA32Buffer::rect):
        (WebCore::RGBA32Buffer::height):
        (WebCore::RGBA32Buffer::status):
        (WebCore::RGBA32Buffer::duration):
        (WebCore::RGBA32Buffer::disposalMethod):
        (WebCore::RGBA32Buffer::hasAlpha):
        (WebCore::RGBA32Buffer::setRect):
        (WebCore::RGBA32Buffer::ensureHeight):
        (WebCore::RGBA32Buffer::setStatus):
        (WebCore::RGBA32Buffer::setDuration):
        (WebCore::RGBA32Buffer::setDisposalMethod):
        (WebCore::RGBA32Buffer::setHasAlpha):
        (WebCore::RGBA32Buffer::setRGBA):
        (WebCore::ImageDecoder::m_failed):
        (WebCore::ImageDecoder::~ImageDecoder):
        (WebCore::ImageDecoder::setData):
        (WebCore::ImageDecoder::size):
        (WebCore::ImageDecoder::frameCount):
        (WebCore::ImageDecoder::repetitionCount):
        (WebCore::ImageDecoder::supportsAlpha):
        (WebCore::ImageDecoder::failed):
        (WebCore::ImageDecoder::setFailed):
        (WebCore::ImageDecoder::clearFrameBufferCache):
        * platform/image-decoders/bmp/BMPImageDecoder.h:
        (WebCore::BMPImageDecoder::filenameExtension):
        * platform/image-decoders/gif/GIFImageDecoder.h:
        (WebCore::GIFImageDecoder::filenameExtension):
        (WebCore::GIFImageDecoder::frameDurationAtIndex):
        (WebCore::GIFImageDecoder::):
        * platform/image-decoders/ico/ICOImageDecoder.h:
        (WebCore::ICOImageDecoder::filenameExtension):
        * platform/image-decoders/jpeg/JPEGImageDecoder.h:
        (WebCore::JPEGImageDecoder::filenameExtension):
        (WebCore::JPEGImageDecoder::supportsAlpha):
        (WebCore::JPEGImageDecoder::reader):
        (WebCore::JPEGImageDecoder::setSize):
        * platform/image-decoders/png/PNGImageDecoder.h:
        (WebCore::PNGImageDecoder::filenameExtension):
        (WebCore::PNGImageDecoder::reader):
        (WebCore::PNGImageDecoder::decodingFailed):
        * platform/image-decoders/xbm/XBMImageDecoder.h:
        (WebCore::XBMImageDecoder::filenameExtension):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44051 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoWebCore
kov@webkit.org [Fri, 22 May 2009 18:06:51 +0000 (18:06 +0000)]
WebCore

2009-05-22  Dominik Röttsches  <dominik.roettsches@access-company.com>

        Reviewed by Gustavo Noronha.

        https://bugs.webkit.org/show_bug.cgi?id=15914
        [GTK] Implement Unicode functionality using GLib

        Minor changes that allow building a hybrid version of WebKit
        that has WTF Unicode backend based on GLib functions while
        text codecs and TextBreakIterator remain based on ICU.

        Fixed some includes that were assuming a WTF ICU backend and
        introduced temporary usage of WTF_USE_GLIB_ICU_UNICODE_HYBRID
        macro.

        * platform/ThreadGlobalData.cpp:
        (WebCore::ThreadGlobalData::ThreadGlobalData):
        (WebCore::ThreadGlobalData::~ThreadGlobalData):
        * platform/ThreadGlobalData.h:
        * platform/text/TextBoundariesICU.cpp:
        * platform/text/TextCodecICU.h:
        * platform/text/TextEncoding.cpp:
        (WebCore::TextEncoding::encode):
        * platform/text/TextEncodingRegistry.cpp:
        (WebCore::buildBaseTextCodecMaps):
        (WebCore::extendTextCodecMaps):

JavaScriptCore

2009-05-22  Dominik Röttsches  <dominik.roettsches@access-company.com>

        Reviewed by Gustavo Noronha.

        https://bugs.webkit.org/show_bug.cgi?id=15914
        [GTK] Implement Unicode functionality using GLib

        Original patch by Jürg Billeter and Naiem Shaik.
        Implementing WTF Unicode functionality based on GLib.

        * GNUmakefile.am:
        * wtf/unicode/Unicode.h:
        * wtf/unicode/glib: Added.
        * wtf/unicode/glib/UnicodeGLib.cpp: Added.
        (WTF::Unicode::foldCase):
        (WTF::Unicode::toLower):
        (WTF::Unicode::toUpper):
        (WTF::Unicode::direction):
        (WTF::Unicode::umemcasecmp):
        * wtf/unicode/glib/UnicodeGLib.h: Added.
        (WTF::Unicode::):
        (WTF::Unicode::toLower):
        (WTF::Unicode::toUpper):
        (WTF::Unicode::toTitleCase):
        (WTF::Unicode::isArabicChar):
        (WTF::Unicode::isFormatChar):
        (WTF::Unicode::isSeparatorSpace):
        (WTF::Unicode::isPrintableChar):
        (WTF::Unicode::isDigit):
        (WTF::Unicode::isPunct):
        (WTF::Unicode::mirroredChar):
        (WTF::Unicode::category):
        (WTF::Unicode::isLower):
        (WTF::Unicode::digitValue):
        (WTF::Unicode::combiningClass):
        (WTF::Unicode::decompositionType):
        * wtf/unicode/glib/UnicodeMacrosFromICU.h: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44050 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoUnreviewed build fix.
abarth@webkit.org [Fri, 22 May 2009 16:59:58 +0000 (16:59 +0000)]
Unreviewed build fix.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44049 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoWebCore:
abarth@webkit.org [Fri, 22 May 2009 16:45:32 +0000 (16:45 +0000)]
WebCore:

2009-05-22  Adam Barth  <abarth@webkit.org>

        Reviewed by Maciej Stachowiak.

        https://bugs.webkit.org/show_bug.cgi?id=25955

        Remove the policyBaseURL parameter from setCookie.  This parameter is
        redundant with the document parameter.

        * WebCore.base.exp:
        * WebCore.order:
        * dom/Document.cpp:
        (WebCore::Document::setCookie):
        * platform/CookieJar.h:
        * platform/mac/CookieJar.mm:
        (WebCore::setCookies):
        * platform/network/ResourceRequestBase.cpp:
        (WebCore::ResourceRequestBase::firstPartyForCookies):
        (WebCore::ResourceRequestBase::setFirstPartyForCookies):
        * platform/network/chromium/CookieJarChromium.cpp:
        (WebCore::setCookies):
        * platform/network/curl/CookieJarCurl.cpp:
        (WebCore::setCookies):
        * platform/network/soup/CookieJarSoup.cpp:
        (WebCore::setCookies):
        * platform/network/win/CookieJarCFNetWin.cpp:
        (WebCore::setCookies):
        * platform/network/win/CookieJarWin.cpp:
        (WebCore::setCookies):
        * platform/qt/CookieJarQt.cpp:
        (WebCore::setCookies):

WebKit/mac:

2009-05-22  Adam Barth  <abarth@webkit.org>

        Reviewed by Maciej Stachowiak.

        https://bugs.webkit.org/show_bug.cgi?id=25955

        Remove the policyBaseURL parameter from setCookie.  This parameter is
        redudant with the document parameter.

        * Plugins/Hosted/NetscapePluginInstanceProxy.mm:
        (WebKit::NetscapePluginInstanceProxy::setCookies):
        * Plugins/WebNetscapePluginView.mm:
        (-[WebNetscapePluginView setVariable:forURL:value:length:]):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44048 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-22 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Fri, 22 May 2009 16:44:47 +0000 (16:44 +0000)]
2009-05-22  Adam Barth  <abarth@webkit.org>

        Reviewed by Maciej Stachowiak.

        Rename policyBaseURL and mainDocumentURL to firstPartyForCookies.  The
        previous names were confusing and inaccurate.

        * WebCore.order:
        * dom/Document.cpp:
        (WebCore::Document::setCookie):
        * dom/Document.h:
        (WebCore::Document::firstPartyForCookies):
        (WebCore::Document::setFirstPartyForCookies):
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::begin):
        (WebCore::FrameLoader::updateFirstPartyForCookies):
        (WebCore::FrameLoader::setFirstPartyForCookies):
        (WebCore::FrameLoader::open):
        (WebCore::FrameLoader::addExtraFieldsToRequest):
        (WebCore::FrameLoader::loadResourceSynchronously):
        * loader/FrameLoader.h:
        * loader/MainResourceLoader.cpp:
        (WebCore::MainResourceLoader::willSendRequest):
        * platform/CookieJar.h:
        * platform/mac/CookieJar.mm:
        (WebCore::setCookies):
        * platform/network/ResourceRequestBase.cpp:
        (WebCore::ResourceRequestBase::adopt):
        (WebCore::ResourceRequestBase::copyData):
        (WebCore::ResourceRequestBase::firstPartyForCookies):
        (WebCore::ResourceRequestBase::setFirstPartyForCookies):
        (WebCore::equalIgnoringHeaderFields):
        * platform/network/ResourceRequestBase.h:
        * platform/network/cf/ResourceRequestCFNet.cpp:
        (WebCore::ResourceRequest::doUpdatePlatformRequest):
        (WebCore::ResourceRequest::doUpdateResourceRequest):
        * platform/network/chromium/CookieJarChromium.cpp:
        (WebCore::setCookies):
        * platform/network/mac/ResourceRequestMac.mm:
        (WebCore::ResourceRequest::doUpdateResourceRequest):
        (WebCore::ResourceRequest::doUpdatePlatformRequest):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44047 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Fix the call to editingCancelled to be this._editingCancelled, since
timothy@apple.com [Fri, 22 May 2009 16:08:58 +0000 (16:08 +0000)]
    Fix the call to editingCancelled to be this._editingCancelled, since
        editingCancelled doesn't exist and throws an exception.

        Reviewed by Timothy Hatcher.

        * inspector/front-end/ElementsTreeOutline.js:
        (WebInspector.ElementsTreeElement.prototype._attributeEditingCommitted):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44046 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-22 Andre Pedralho <andre.pedralho@openbossa.org>
hausmann@webkit.org [Fri, 22 May 2009 15:33:21 +0000 (15:33 +0000)]
2009-05-22  Andre Pedralho  <andre.pedralho@openbossa.org>

        Reviewed by Simon Hausmann.

        Fixed --makeargs usage in Qt build.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44045 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Eric Seidel.
ap@webkit.org [Fri, 22 May 2009 15:00:39 +0000 (15:00 +0000)]
    Reviewed by Eric Seidel.

        https://bugs.webkit.org/show_bug.cgi?id=25928
        Several errors detected by cppcheck

        Fix all (the only one) problem in cross-platform code found by the tool. Port maintainers
        should take a look at other reports.

        * css/CSSParser.cpp: (WebCore::CSSParser::addProperty): Don't leak the property even if there
        are billions of those.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44044 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-22 Zoltan Horvath <horvath.zoltan.6@stud.u-szeged.hu>
hausmann@webkit.org [Fri, 22 May 2009 14:26:50 +0000 (14:26 +0000)]
2009-05-22  Zoltan Horvath  <horvath.zoltan.6@stud.u-szeged.hu>

        Reviewed by Simon Hausmann.

        Correcting error message on missing WEBKIT_TESTFONTS
        environment varible in Qt-port.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44043 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-22 Kenneth Rohde Christiansen <kenneth.christiansen@openbossa.org>
hausmann@webkit.org [Fri, 22 May 2009 14:26:39 +0000 (14:26 +0000)]
2009-05-22  Kenneth Rohde Christiansen  <kenneth.christiansen@openbossa.org>

        Reviewed by Simon Hausmann.

        Skip tests depending on platform dependent results.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44042 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-22 Shinichiro Hamaji <hamaji@google.com>
eric@webkit.org [Fri, 22 May 2009 13:24:06 +0000 (13:24 +0000)]
2009-05-22  Shinichiro Hamaji  <hamaji@google.com>

        Reviewed by Oliver Hunt.

        Remove optimization path alpha=0 case from GraphicContext(CG|Skia).
        This optimization doesn't make sense for some composite mode (e.g., 'copy' operation).

        https://bugs.webkit.org/show_bug.cgi?id=25956

        * platform/graphics/cg/GraphicsContextCG.cpp:
        (WebCore::GraphicsContext::drawRect):
        (WebCore::GraphicsContext::drawLine):
        (WebCore::GraphicsContext::strokeArc):
        (WebCore::GraphicsContext::drawConvexPolygon):
        (WebCore::GraphicsContext::fillPath):
        (WebCore::GraphicsContext::strokePath):
        (WebCore::GraphicsContext::fillRect):
        (WebCore::GraphicsContext::fillRoundedRect):
        (WebCore::GraphicsContext::strokeRect):
        * platform/graphics/skia/GraphicsContextSkia.cpp:
        (WebCore::GraphicsContext::drawConvexPolygon):
        (WebCore::GraphicsContext::drawEllipse):
        (WebCore::GraphicsContext::fillPath):
        (WebCore::GraphicsContext::fillRect):
        (WebCore::GraphicsContext::strokePath):
        (WebCore::GraphicsContext::strokeRect):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44041 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Antonio Gomes <antonio.gomes@openbossa.org>
hausmann@webkit.org [Fri, 22 May 2009 13:12:04 +0000 (13:12 +0000)]
2009-05-21  Antonio Gomes   <antonio.gomes@openbossa.org>

        Reviewed by Simon Hausmann.

        [Qt] Additional test for resource load (CSS file) for QWebFrame.

        * tests/qwebframe/qwebframe.qrc:
        * tests/qwebframe/style.css: Added.
        * tests/qwebframe/tst_qwebframe.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44040 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-22 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Fri, 22 May 2009 08:37:05 +0000 (08:37 +0000)]
2009-05-22  Xan Lopez  <xlopez@igalia.com>

        Unreviewed build fix.

        Revert the whole thing, since we can't figure out a way to make it
        work in all platforms.

        * rendering/SVGInlineTextBox.cpp:
        (WebCore::pathForDecoration):
        * svg/SVGLength.cpp:
        (WebCore::SVGLength::PercentageOfViewport):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44039 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-22 Simon Hausmann <simon.hausmann@nokia.com>
hausmann@webkit.org [Fri, 22 May 2009 08:36:57 +0000 (08:36 +0000)]
2009-05-22  Simon Hausmann  <simon.hausmann@nokia.com>

        Unreviewed Qt build fix.

        Provide dummy implementation of pure virtual EditorClient method.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44038 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-22 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Fri, 22 May 2009 08:20:33 +0000 (08:20 +0000)]
2009-05-22  Xan Lopez  <xlopez@igalia.com>

        Try to fix Windows build.

        Apparently it can't tell if '1.0f' is float or double, so cast to
        float.

        * rendering/SVGInlineTextBox.cpp:
        (WebCore::pathForDecoration):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44037 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-22 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Fri, 22 May 2009 08:11:03 +0000 (08:11 +0000)]
2009-05-22  Xan Lopez  <xlopez@igalia.com>

        Try to fix breakage from the previous patch.

        * rendering/SVGInlineTextBox.cpp:
        (WebCore::pathForDecoration):
        * svg/SVGLength.cpp:
        (WebCore::SVGLength::PercentageOfViewport):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44036 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-15 Fridrich Strba <fridrich.strba@bluewin.ch>
xan@webkit.org [Fri, 22 May 2009 07:28:03 +0000 (07:28 +0000)]
2009-05-15  Fridrich Strba  <fridrich.strba@bluewin.ch>

        Reviewed by Maciej Stachowiak.

        https://bugs.webkit.org/show_bug.cgi?id=25805
        In debug build, mingw win32 build has problems with powf

        Change C-style calls to powf and sqrtf to C++-style calls to
        overloaded pow and sqrt. This solves also a problem with mingw
        compiler during the debug build.

        * rendering/SVGInlineTextBox.cpp:
        (WebCore::pathForDecoration):
        * svg/SVGLength.cpp:
        (WebCore::SVGLength::PercentageOfViewport):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44035 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Fri, 22 May 2009 06:20:44 +0000 (06:20 +0000)]
2009-05-21  Xan Lopez  <xlopez@igalia.com>

        Unreviewed build fix.

        Add MacroAssemblerCodeRef.h to file list.

        * GNUmakefile.am:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44034 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Stephanie Lewis <slewis@apple.com>
slewis@apple.com [Fri, 22 May 2009 03:57:33 +0000 (03:57 +0000)]
2009-05-21  Stephanie Lewis  <slewis@apple.com>

        Reviewed by Mark Rowe.

        <rdar://problem/6913680> Update image caching options.
        Disable extra ImageIO cache because we already handle caching decoded image data.

        * platform/graphics/cg/ImageSourceCG.cpp:
        (WebCore::imageSourceOptions):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44031 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Gavin Barraclough <barraclough@apple.com>
barraclough@apple.com [Fri, 22 May 2009 02:34:41 +0000 (02:34 +0000)]
2009-05-21  Gavin Barraclough  <barraclough@apple.com>

        Reviewed by Darin Adler.
        Addition of MacroAssemblerCodeRef.h rubber stamped by Geoff Garen.

        Refactor JIT code-handle objects.  The representation of generated code is currently
        a bit of a mess.  We have a class JITCode which wraps the pointer to a block of
        generated code, but this object does not reference the executable pool meaning that
        external events (the pool being derefed) could make the pointer become invalid.
        To overcome this both the JIT and Yarr implement further (and similar) objects to
        wrap the code pointer with a RefPtr to the pool.  To add to the mire, as well as the
        CodeBlock containing a handle onto the code the FunctionBodyNode also contains a
        copy of the code pointer which is used almost (but not entirely) uniquely to access
        the JIT code for a function.

        Rationalization of all this:

            * Add a new type 'MacroAssembler::CodeRef' as a handle for a block of JIT generated code.
            * Change the JIT & Yarr to internally handle code using CodeRefs.
            * Move the CodeRef (formerly anow defunct JITCodeRef) from CodeBlock to its owner node.
            * Remove the (now) redundant code pointer from FunctionBodyNode.

        While tidying this up I've made the PatchBuffer return code in new allocations using a CodeRef,
        and have enforced an interface that the PatchBuffer will always be used, and 'finalizeCode()' or
        'finalizeCodeAddendum()' will always be called exactly once on the PatchBuffer to complete code generation.

        This gives us a potentially useful hook ('PatchBuffer::performFinalization()') at the end of generation,
        which may have a number of uses.  It may be helpful should we wish to switch our generation
        model to allow RW/RX exclusive memory, and it may be useful on non-cache-coherent platforms to
        give us an oportunity to cache flush as necessary.

        No performance impact.

        * assembler/AbstractMacroAssembler.h:
        (JSC::AbstractMacroAssembler::ProcessorReturnAddress::relinkCallerToTrampoline):
        (JSC::AbstractMacroAssembler::CodeRef::CodeRef):
        (JSC::AbstractMacroAssembler::CodeRef::trampolineAt):
        (JSC::AbstractMacroAssembler::PatchBuffer::PatchBuffer):
        (JSC::AbstractMacroAssembler::PatchBuffer::~PatchBuffer):
        (JSC::AbstractMacroAssembler::PatchBuffer::link):
        (JSC::AbstractMacroAssembler::PatchBuffer::linkTailRecursive):
        (JSC::AbstractMacroAssembler::PatchBuffer::patch):
        (JSC::AbstractMacroAssembler::PatchBuffer::complete):
        (JSC::AbstractMacroAssembler::PatchBuffer::finalize):
        (JSC::AbstractMacroAssembler::PatchBuffer::entry):
        * bytecode/CodeBlock.cpp:
        (JSC::CodeBlock::CodeBlock):
        (JSC::CodeBlock::reparseForExceptionInfoIfNecessary):
        (JSC::CodeBlock::setJITCode):
        * bytecode/CodeBlock.h:
        (JSC::CodeBlock::getBytecodeIndex):
        (JSC::CodeBlock::executablePool):
        * interpreter/CallFrameClosure.h:
        * interpreter/Interpreter.cpp:
        (JSC::Interpreter::execute):
        (JSC::Interpreter::prepareForRepeatCall):
        * jit/JIT.cpp:
        (JSC::JIT::privateCompile):
        (JSC::JIT::privateCompileCTIMachineTrampolines):
        (JSC::JIT::linkCall):
        * jit/JIT.h:
        * jit/JITCode.h:
        (JSC::JITCode::JITCode):
        (JSC::JITCode::operator bool):
        (JSC::JITCode::addressForCall):
        (JSC::JITCode::offsetOf):
        (JSC::JITCode::execute):
        (JSC::JITCode::size):
        (JSC::JITCode::executablePool):
        (JSC::JITCode::HostFunction):
        * jit/JITPropertyAccess.cpp:
        (JSC::JIT::privateCompilePutByIdTransition):
        (JSC::JIT::privateCompilePatchGetArrayLength):
        (JSC::JIT::privateCompileGetByIdProto):
        (JSC::JIT::privateCompileGetByIdSelfList):
        (JSC::JIT::privateCompileGetByIdProtoList):
        (JSC::JIT::privateCompileGetByIdChainList):
        (JSC::JIT::privateCompileGetByIdChain):
        * jit/JITStubs.cpp:
        (JSC::JITStubs::cti_vm_dontLazyLinkCall):
        (JSC::JITStubs::cti_vm_lazyLinkCall):
        * parser/Nodes.cpp:
        (JSC::ProgramNode::generateJITCode):
        (JSC::EvalNode::generateJITCode):
        (JSC::FunctionBodyNode::FunctionBodyNode):
        (JSC::FunctionBodyNode::createNativeThunk):
        (JSC::FunctionBodyNode::generateJITCode):
        * parser/Nodes.h:
        (JSC::ScopeNode::generatedJITCode):
        (JSC::ScopeNode::getExecutablePool):
        (JSC::ScopeNode::setJITCode):
        (JSC::ProgramNode::jitCode):
        (JSC::EvalNode::jitCode):
        (JSC::FunctionBodyNode::jitCode):
        * runtime/RegExp.cpp:
        (JSC::RegExp::match):
        * yarr/RegexJIT.cpp:
        (JSC::Yarr::RegexGenerator::compile):
        (JSC::Yarr::jitCompileRegex):
        (JSC::Yarr::executeRegex):
        * yarr/RegexJIT.h:
        (JSC::Yarr::RegexCodeBlock::RegexCodeBlock):
        (JSC::Yarr::RegexCodeBlock::pcreFallback):
        (JSC::Yarr::RegexCodeBlock::setFallback):
        (JSC::Yarr::RegexCodeBlock::operator bool):
        (JSC::Yarr::RegexCodeBlock::set):
        (JSC::Yarr::RegexCodeBlock::execute):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44030 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Albert J. Wong <ajwong@chromium.org>
levin@chromium.org [Fri, 22 May 2009 01:55:25 +0000 (01:55 +0000)]
2009-05-21  Albert J. Wong  <ajwong@chromium.org>

        Reviewed by Eric Seidel.

        Bug 25903: Create a default UI theme for media controls in Chromium.
        https://bugs.webkit.org/show_bug.cgi?id=25903

        Implement a default UI for chromium.  Add a style sheet for Chromium
        media controls with good defaults and implemented basic draw functions
        for play/pause & mute buttons.

        * css/mediaControlsChromium.css: Added.
        * rendering/RenderThemeChromiumLinux.cpp:
        (WebCore::RenderThemeChromiumWin::extraMediaControlsStyleSheet): Export
        our custom media controls style sheet.
        (WebCore::RenderThemeChromiumLinux::paintMediaButtonInternal): Paint
        buttons respecting chromium media controls color scheme.
        (WebCore::RenderThemeChromiumLinux::paintMediaPlayButton): Paint logic
        specific to play/pause button.
        (WebCore::RenderThemeChromiumLinux::paintMediaMuteButton): Paint
        specific to mute button.
        * rendering/RenderThemeChromiumLinux.h: Appropriate header changes.
        * rendering/RenderThemeChromiumWin.cpp:
        (WebCore::RenderThemeChromiumWin::extraMediaControlsStyleSheet): Export
        our custom media controls style sheet.
        (WebCore::RenderThemeChromiumWin::paintMediaButtonInternal): Paint
        buttons respecting chromium media controls color scheme.
        (WebCore::RenderThemeChromiumWin::paintMediaPlayButton): Paint logic
        specific to play/pause button.
        (WebCore::RenderThemeChromiumWin::paintMediaMuteButton): Paint specific
        to mute button.
        (WebCore::RenderThemeChromiumWin::setDefaultFontSize): Appropriate
        header changes.
        * rendering/RenderThemeChromiumWin.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44029 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoChecked in layout tests I accidentally left out of my last check-in.
ggaren@apple.com [Fri, 22 May 2009 01:46:49 +0000 (01:46 +0000)]
Checked in layout tests I accidentally left out of my last check-in.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44027 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Eric Seidel <eric@webkit.org>
eric@webkit.org [Fri, 22 May 2009 01:35:59 +0000 (01:35 +0000)]
2009-05-21  Eric Seidel  <eric@webkit.org>

        Rubber-stamped by Antti Koivisto.

        Make it possible to write js-only http tests by exposing a /js-test-resources alias.
        https://bugs.webkit.org/show_bug.cgi?id=25915

        Second attempt at landing, now using -c instead of -C to make sure
        that mod_alias is loaded on Apache 1.x instances before using Alias.

        * Scripts/run-webkit-httpd:
        * Scripts/run-webkit-tests:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44020 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoReviewed by Kevin Ollivier.
kevino@webkit.org [Fri, 22 May 2009 01:10:32 +0000 (01:10 +0000)]
Reviewed by Kevin Ollivier.

When stripping the query from a file:// URL, do not leave a trailing question mark.

https://bugs.webkit.org/show_bug.cgi?id=25940

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44018 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoReviewed by Kevin Ollivier.
kevino@webkit.org [Fri, 22 May 2009 00:57:40 +0000 (00:57 +0000)]
Reviewed by Kevin Ollivier.

Notify ImageObserver::didDraw to help the cache purge alive resources.

https://bugs.webkit.org/show_bug.cgi?id=25942

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44017 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago<rdar://problem/6910264> REGRESSION: Cached DOM global object property access fails...
oliver@apple.com [Fri, 22 May 2009 00:28:22 +0000 (00:28 +0000)]
<rdar://problem/6910264> REGRESSION: Cached DOM global object property access fails in browser (25921)
<https://bugs.webkit.org/show_bug.cgi?id=25921>

Reviewed by Maciej Stachowiak.

When caching properties on the global object we need to ensure that we're
not attempting to cache through a shell object.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44016 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Evan Martin <evan@chromium.org>
levin@chromium.org [Fri, 22 May 2009 00:15:48 +0000 (00:15 +0000)]
2009-05-21  Evan Martin  <evan@chromium.org>

        Reviewed by Eric Seidel.

        Flesh out font fallback for Chromium's Skia port.
        getLastResortFallbackFont should fall back on well-known fonts.
        https://bugs.webkit.org/show_bug.cgi?id=25860

        * platform/graphics/chromium/FontCacheLinux.cpp:
        (WebCore::FontCache::getLastResortFallbackFont): try known font names.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44015 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Windows build fix.
sfalken@apple.com [Fri, 22 May 2009 00:09:13 +0000 (00:09 +0000)]
    Windows build fix.

        * JavaScriptCore.vcproj/jsc/jscCommon.vsprops:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44014 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by John Sullivan.
mitz@apple.com [Thu, 21 May 2009 23:32:40 +0000 (23:32 +0000)]
    Reviewed by John Sullivan.

        - speculative fix for <rdar://problem/6890716> crash at
          FontCache::purgeInactiveFontData + 199

        * platform/graphics/FontCache.cpp:
        (WebCore::FontCache::purgeInactiveFontData): Deleting a SimpleFontData
        can cause releaseFontData() to be called, which modifies
        gInactiveFontData. Therefore, avoid deleting SimpleFontData instances
        while iterating over gInactiveFontData and delete them afterwards.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44011 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoWebCore:
ggaren@apple.com [Thu, 21 May 2009 23:26:06 +0000 (23:26 +0000)]
WebCore:

2009-05-21  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Sam Weinig.

        Test for <rdar://problem/6910832> | https://bugs.webkit.org/show_bug.cgi?id=25907
        Incorrect URL returned to the DOM while the user drags a file

        * page/DragController.cpp:
        (WebCore::DragController::dragExited):
        (WebCore::DragController::tryDHTMLDrag): Don't base our decision on KURL,
        since that only looks at the text of the document's URL. Do base our
        decision on the securityOrigin(), which knows more about the document's
        actual origin.

LayoutTests:

2009-05-21  Geoffrey Garen  <ggaren@apple.com>

        Reviewed by Sam Weinig.

        Tests for <rdar://problem/6910832> | https://bugs.webkit.org/show_bug.cgi?id=25907
        Incorrect URL returned to the DOM while the user drags a file

        * http/tests/local/drag-over-remote-content-expected.txt: Added.
        * http/tests/local/drag-over-remote-content.html: Added.
        * http/tests/security/drag-over-remote-content-iframe-expected.txt: Added.
        * http/tests/security/drag-over-remote-content-iframe.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44010 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoCorrect change log
mitz@apple.com [Thu, 21 May 2009 23:23:27 +0000 (23:23 +0000)]
Correct change log

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44009 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Rubber-stamped by Mark Rowe.
mitz@apple.com [Thu, 21 May 2009 23:22:24 +0000 (23:22 +0000)]
    Rubber-stamped by Mark Rowe.

        - fix https://bugs.webkit.org/show_bug.cgi?id=25947
          Assertion failure in FrameLoader::transitionToCommitted on launch in
          Safari with r43985

        Rolled out r43985.

        * platform/MIMETypeRegistry.cpp:
        (WebCore::initializeSupportedNonImageMimeTypes):
        (WebCore::MIMETypeRegistry::isSupportedNonImageMIMEType):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@44000 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Sam Weinig <sam@webkit.org>
weinig@apple.com [Thu, 21 May 2009 23:00:42 +0000 (23:00 +0000)]
2009-05-21  Sam Weinig  <sam@webkit.org>

        Reviewed by Brady Eidson and Mark Rowe.

        Fix for <rdar://problem/6901522>
        REGRESSION: Office 2008: Preview with browser is not shown, and Finder window is in the front of Safari

        If the NSURLRequest is annotated with a "ForceHTMLMIMEType" property, force
        the MIME type to be text/html.

        * platform/network/mac/ResourceHandleMac.mm:
        (-[WebCoreResourceHandleAsDelegate connection:didReceiveResponse:]):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43993 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoBlind windows fix after last commit.
kevino@webkit.org [Thu, 21 May 2009 22:56:18 +0000 (22:56 +0000)]
Blind windows fix after last commit.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43992 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>
kov@webkit.org [Thu, 21 May 2009 22:42:10 +0000 (22:42 +0000)]
2009-05-21  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Fixing test to actually use the correct names for the frames.

        * http/tests/mime/accept-all-text-types.html:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43991 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoReviewed by Kevin Ollivier.
kevino@webkit.org [Thu, 21 May 2009 22:29:46 +0000 (22:29 +0000)]
Reviewed by Kevin Ollivier.

Use an environment variable to control which logs are on and off.
Also, make a generic function for translating channel name to its log.

https://bugs.webkit.org/show_bug.cgi?id=25936

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43990 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>
kov@webkit.org [Thu, 21 May 2009 22:15:15 +0000 (22:15 +0000)]
2009-05-21  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Disable 2 more tests that sometimes fail.

        * platform/gtk/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43989 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Thu, 21 May 2009 22:12:39 +0000 (22:12 +0000)]
2009-05-21  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Mark Rowe.

        Bug 25945: Add support for MADV_FREE to TCMalloc
        <https://bugs.webkit.org/show_bug.cgi?id=25945>
        <rdar://problem/6910754>

        Add support for MADV_FREE to TCMalloc_SystemRelease for platforms that
        don't also support MADV_FREE_REUSE. The code is identical to the MADV_DONTNEED
        case except for the advice passed to madvise(), so combining the two cases
        makes the most sense.

        * wtf/Platform.h: Only define HAVE_MADV_FREE when not building on Tiger or
        Leopard, because while it is defined on these platforms it actually does
        nothing.
        * wtf/TCSystemAlloc.cpp:
        (TCMalloc_SystemRelease): use MADV_FREE if it is available; otherwise use
        MADV_DONTNEED.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43988 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agowx build fix after method added to EditorClient.
kevino@webkit.org [Thu, 21 May 2009 22:07:13 +0000 (22:07 +0000)]
wx build fix after method added to EditorClient.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43987 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>
kov@webkit.org [Thu, 21 May 2009 21:54:50 +0000 (21:54 +0000)]
2009-05-21  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Adding file I forgot to add to last commit.

        * http/tests/mime/resources/accept-all-text-types.cgi: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43986 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>
kov@webkit.org [Thu, 21 May 2009 21:47:55 +0000 (21:47 +0000)]
2009-05-21  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Reviewed by Darin Adler.

        Test that WebKit accepts any text/* type.

        * http/tests/mime/accept-all-text-types-expected.txt: Added.
        * http/tests/mime/accept-all-text-types.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43985 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Darin Fisher <darin@chromium.org>
darin@chromium.org [Thu, 21 May 2009 21:36:20 +0000 (21:36 +0000)]
2009-05-21  Darin Fisher  <darin@chromium.org>

        Fix-up coding style.

        * WebCoreSupport/WebEditorClient.mm:
        (WebEditorClient::getAutoCorrectSuggestionForMisspelledWord):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43984 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Darin Fisher <darin@chromium.org>
darin@chromium.org [Thu, 21 May 2009 21:36:10 +0000 (21:36 +0000)]
2009-05-21  Darin Fisher  <darin@chromium.org>

        Fix-up coding style.

        * WebCoreSupport/WebEditorClient.cpp:
        (WebEditorClient::getAutoCorrectSuggestionForMisspelledWord):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43983 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Darin Fisher <darin@chromium.org>
darin@chromium.org [Thu, 21 May 2009 21:35:10 +0000 (21:35 +0000)]
2009-05-21  Darin Fisher  <darin@chromium.org>

        Fix build bustage.

        * WebCoreSupport/EditorClientGtk.cpp:
        (WebKit::EditorClient::getAutoCorrectSuggestionForMisspelledWord): Add stub for method.
        * WebCoreSupport/EditorClientGtk.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43982 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 David Levin <levin@chromium.org>
levin@chromium.org [Thu, 21 May 2009 21:19:26 +0000 (21:19 +0000)]
2009-05-21  David Levin  <levin@chromium.org>

        Reviewed by Darin Fisher.

        Bug 25946: Chromium needs MIMETypeRegistry::getMediaMIMETypeForExtension implementation.
        https://bugs.webkit.org/show_bug.cgi?id=25946

        A very simple implementation to keep it working.  A bug has been filed in the chromium db
        (http://code.google.com/p/chromium/issues/detail?id=12441) to investigate doing something more
        complete.

        * platform/chromium/MimeTypeRegistryChromium.cpp:
        (WebCore::MIMETypeRegistry::getMediaMIMETypeForExtension):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43981 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Siddhartha Chattopadhyaya <sidchat@google.com>
darin@chromium.org [Thu, 21 May 2009 21:02:28 +0000 (21:02 +0000)]
2009-05-20  Siddhartha Chattopadhyaya  <sidchat@google.com>

        Reviewed by Justin Garcia.

        Add automatic spell correction support in WebKit

        https://bugs.webkit.org/show_bug.cgi?id=24953

        * editing/Editor.cpp:
        (WebCore::findFirstMisspellingInRange):
        (WebCore::Editor::advanceToNextMisspelling):
        (WebCore::Editor::markMisspellingsAfterTypingToPosition):
        (WebCore::markAllMisspellingsInRange):
        (WebCore::markMisspellingsOrBadGrammar):
        (WebCore::Editor::markMisspellings):
        (WebCore::Editor::markBadGrammar):
        (WebCore::Editor::markMisspellingsAndBadGrammar):
        * editing/Editor.h:
        * loader/EmptyClients.h:
        (WebCore::EmptyEditorClient::getAutoCorrectSuggestionForMisspelledWord):
        * page/EditorClient.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43980 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Christian Dywan <christian@twotoasts.de>
christian@webkit.org [Thu, 21 May 2009 20:57:53 +0000 (20:57 +0000)]
2009-05-21  Christian Dywan  <christian@twotoasts.de>

        Reviewed by Darin Adler.

        Text files which have sub MIME types are treated as not displayable
        http://bugs.webkit.org/show_bug.cgi?id=24903

        * platform/MIMETypeRegistry.cpp:
        (WebCore::initializeSupportedNonImageMimeTypes):
        (WebCore::MIMETypeRegistry::isSupportedNonImageMIMEType): Regard any
        MIME type beginning with "text/" as supported and remove all "text/"
        types from the list.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43979 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoLayoutTests
kov@webkit.org [Thu, 21 May 2009 20:54:43 +0000 (20:54 +0000)]
LayoutTests

2009-05-21  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Reviewed by Xan Lopez.

        Disable failing HTTP tests now that we are enabling HTTP on the
        bot.

        * platform/gtk/Skipped:

WebKitTools

2009-05-21  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>

        Reviewed by Xan Lopez.

        Default to running HTTP tests for GTK+.

        * Scripts/run-webkit-tests:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43978 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Darin Adler <darin@apple.com>
darin@apple.com [Thu, 21 May 2009 20:30:19 +0000 (20:30 +0000)]
2009-05-21  Darin Adler  <darin@apple.com>

        Rolled the httpd changes out, since many machines (including the bots)
        have an older version of Apache that does not support the Alias command.
        Or maybe it's something else, but the bots are failing.

        * Scripts/run-webkit-httpd: Rolled back.
        * Scripts/run-webkit-tests: Ditto.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43977 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoFix <https://bugs.webkit.org/show_bug.cgi?id=25917> / <rdar://problem/6910066>.
mrowe@apple.com [Thu, 21 May 2009 19:57:26 +0000 (19:57 +0000)]
Fix <https://bugs.webkit.org/show_bug.cgi?id=25917> / <rdar://problem/6910066>.
Bug 25917: REGRESSION (r43559?): Javascript debugger crashes when pausing page

Reviewed by Oliver Hunt.

The debugger currently retrieves the arguments object from an activation rather than pulling
it from a call frame.  This is unreliable to due to the recent optimization to lazily create
the arguments object.  In the long-term it should stop doing that (<rdar://problem/6911886>),
but for now we force eager creation of the arguments object when debugging.

* bytecompiler/BytecodeGenerator.cpp:
(JSC::BytecodeGenerator::BytecodeGenerator):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43976 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Eric Seidel <eric@webkit.org>
ggaren@apple.com [Thu, 21 May 2009 19:25:08 +0000 (19:25 +0000)]
2009-05-21  Eric Seidel  <eric@webkit.org>

        Reviewed by Geoffrey Garen.

        Make it possible to write js-only http tests by exposing a /js-test-resources alias.
        https://bugs.webkit.org/show_bug.cgi?id=25915

        * Scripts/run-webkit-httpd:
        * Scripts/run-webkit-tests:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43975 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by David Kilzer.
ap@webkit.org [Thu, 21 May 2009 18:51:07 +0000 (18:51 +0000)]
    Reviewed by David Kilzer.

        https://bugs.webkit.org/show_bug.cgi?id=25890
        Implement Network logging channel on Mac

        * platform/network/mac/ResourceHandleMac.mm:
        (WebCore::ResourceHandle::~ResourceHandle):
        (WebCore::ResourceHandle::start):
        (WebCore::ResourceHandle::cancel):
        (WebCore::ResourceHandle::setDefersLoading):
        (-[WebCoreResourceHandleAsDelegate connection:willSendRequest:redirectResponse:]):
        (-[WebCoreResourceHandleAsDelegate connectionShouldUseCredentialStorage:]):
        (-[WebCoreResourceHandleAsDelegate connection:didReceiveAuthenticationChallenge:]):
        (-[WebCoreResourceHandleAsDelegate connection:didCancelAuthenticationChallenge:]):
        (-[WebCoreResourceHandleAsDelegate connection:didReceiveResponse:]):
        (-[WebCoreResourceHandleAsDelegate connection:didReceiveData:lengthReceived:]):
        (-[WebCoreResourceHandleAsDelegate connection:willStopBufferingData:]):
        (-[WebCoreResourceHandleAsDelegate connection:didSendBodyData:totalBytesWritten:totalBytesExpectedToWrite:]):
        (-[WebCoreResourceHandleAsDelegate connectionDidFinishLoading:]):
        (-[WebCoreResourceHandleAsDelegate connection:didFailWithError:]):
        (-[WebCoreResourceHandleAsDelegate connection:willCacheResponse:]):
        Added logging for async requests (as these are most interesting).

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43974 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Darin Adler.
ap@webkit.org [Thu, 21 May 2009 18:19:11 +0000 (18:19 +0000)]
    Reviewed by Darin Adler.

        <rdar://problem/6907542> REGRESSION (r43143): Hang in RenderLineBoxList::dirtyLinesFromChangedChild
        when clicking link to load Flash animation (http://www.roambi.com)

        * WebCoreSupport/WebEditorClient.mm: (WebEditorClient::respondToChangedSelection):
        Don't call -[NSApplication updateWindows] here, WebHTMLView can take care of this.

        * WebView/WebHTMLView.mm:
        (-[WebHTMLView inputContext]): Use a precomputed boolean stored in WebHTMLViewPrivate, as
        calling isTextInput() is not always safe.
        (-[WebHTMLView textStorage]): Ditto.
        (-[WebHTMLView _updateSelectionForInputManager]): Update _private->exposeInputContext when
        selection changes, and let AppKit update its cache if necessary.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43973 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Eric Carlson <eric.carlson@apple.com>
eric.carlson@apple.com [Thu, 21 May 2009 18:17:31 +0000 (18:17 +0000)]
2009-05-21  Eric Carlson  <eric.carlson@apple.com>

        Reviewed by David Kilzer.

        <rdar://problem/6902197> WebCore doesn't know about all of the MIME types supported by QTKit

        Add a static table to map from some common file extensions to their MIME types so it
        is possible to detect support even when the platform mapping is incomplete. Use the
        table to lookup types not found by UTI on the OSX, and when looking for a media engine
        with a generic or missing type.

        * platform/MIMETypeRegistry.cpp:
        (WebCore::initializeMediaTypeMaps): New, initialize the mediaMIMETypeForExtensionMap.
        (WebCore::MIMETypeRegistry::getMediaMIMETypeForExtension): New. Look for the specified
        MIME type in the static mapping table if the platform specific getMIMETypeForExtension
        doesn't find a match.
        * platform/MIMETypeRegistry.h: Declare getMediaMIMETypeForExtension.

        * platform/graphics/MediaPlayer.cpp:
        (WebCore::MediaPlayer::load): When the type passed is empty, "application/octet-stream", or
        "text/plain", try to look it up based on the file extension with getMediaMIMETypeForExtension.

        * platform/graphics/mac/MediaPlayerPrivateQTKit.mm:
        (WebCore::addFileTypesToCache): See if getMediaMIMETypeForExtension can map from file extension
        to MIME type if UTI fails.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43972 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Andre Pedralho <andre.pedralho@openbossa.org>
kov@webkit.org [Thu, 21 May 2009 16:36:13 +0000 (16:36 +0000)]
2009-05-21  Andre Pedralho  <andre.pedralho@openbossa.org>

        Reviewed by Gustavo Noronha.

        Added --makeargs flag to the build script.

        * Scripts/build-webkit:
        * Scripts/webkitdirs.pm:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43971 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Darin Adler.
ap@webkit.org [Thu, 21 May 2009 16:29:24 +0000 (16:29 +0000)]
    Reviewed by Darin Adler.

        <rdar://problem/5579292> REGRESSION (2-3): "Default default" encoding for Korean changed
        from Korean (Windows, DOS) to Korean (ISO 2022-KR), which breaks some sites

        * WebView/WebPreferences.mm:
       (+[WebPreferences _setInitialDefaultTextEncodingToSystemEncoding]): Update the existing fix
       for the changed result of CFStringConvertEncodingToIANACharSetName().

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43970 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Darin Adler.
ap@webkit.org [Thu, 21 May 2009 16:03:50 +0000 (16:03 +0000)]
    Reviewed by Darin Adler.

        <rdar://problem/6883758> REGRESSION (r43143): First character typed with input method does
        not go into inline hole (seen with Chinese & Kotoeri on top Chinese website www.baidu.com)

        * WebCoreSupport/WebEditorClient.mm: (WebEditorClient::respondToChangedSelection):
        Call -[NSApplication updateWindows] to make AppKit re-fetch the input context when selection
        changes. Since we use SelectionController to check whether the view is editable, it is
        important not to re-fetch the context too early, e.g. from a focus changed notification.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43969 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Eric Seidel <eric@webkit.org>
eric@webkit.org [Thu, 21 May 2009 14:05:05 +0000 (14:05 +0000)]
2009-05-21  Eric Seidel  <eric@webkit.org>

        Reviewed by Alexey Proskuryakov.

        Clean up DragController
        https://bugs.webkit.org/show_bug.cgi?id=25926

        I renamed m_document to m_documentUnderMouse to better document what it does
        I could call it m_lastDocumentUnderMouse to be most-accurate, but this seemed
        long enough.

        I also saw copy/paste code involving clearing the selection when moving out of
        one document and into another.  I moved that code into a function called
        mouseMovedIntoDocument.

        I also got rid of a couple needless null-checks after confirming via source
        inspection they were not ever possible.
        In Mac WebKit it's not possible to have a Page that doesn't have a mainFrame()
        I left an ASSERT(m_page->mainFrame()) in case some other port ever violates this.
        It's also not possible to return a document from documentAtPoint that is not
        in a frame (as such a document would not be rendered).

        No functional changes, thus no tests.

        * page/DragController.cpp:
        (WebCore::DragController::DragController):
        (WebCore::DragController::dragIsMove):
        (WebCore::DragController::dragExited):
        (WebCore::DragController::performDrag):
        (WebCore::DragController::mouseMovedIntoDocument):
        (WebCore::DragController::dragEnteredOrUpdated):
        (WebCore::DragController::tryDocumentDrag):
        (WebCore::DragController::operationForLoad):
        (WebCore::DragController::concludeEditDrag):
        (WebCore::DragController::canProcessDrag):
        (WebCore::DragController::tryDHTMLDrag):
        (WebCore::DragController::placeDragCaret):
        * page/DragController.h:
        (WebCore::DragController::documentUnderMouse):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43968 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Eric Seidel <eric@webkit.org>
eric@webkit.org [Thu, 21 May 2009 12:50:30 +0000 (12:50 +0000)]
2009-05-21  Eric Seidel  <eric@webkit.org>

        Reviewed by Alexey Proskuryakov.

        Rename DragController::dragOperation() to sourceDragOperation() for clarity
        Remove unused setDragOperation()

        There are two times DragOperation is used.  The source specifies
        what drag operations it supports (in a mask) and the destination
        picks a subset of those as DragOperations it supports if a drop were
        to occur.  I was confused when I first saw this accessor as to which
        it meant.  It turns out that this is only used when WebKit starts a
        drag, and we keep it around so that we can provide this information
        to other Mac OS X applications when they ask.  I've renamed the method
        for clarity so the next person in this code will be less confused.

        No functional changes, thus no tests.

        * page/DragController.h:
        (WebCore::DragController::sourceDragOperation):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43967 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Reviewed by Darin Adler.
ap@webkit.org [Thu, 21 May 2009 12:34:31 +0000 (12:34 +0000)]
    Reviewed by Darin Adler.

        https://bugs.webkit.org/show_bug.cgi?id=19517
        <rdar://problem/6007112> DOM modification of textArea causes Access Violation (NULL pointer?)

        This no longer happens with ToT, just adding a test.

        * fast/parser/head-parsing-19517-expected.txt: Added.
        * fast/parser/head-parsing-19517.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43966 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Rubber-stamped by Mark Rowe.
mitz@apple.com [Thu, 21 May 2009 09:30:05 +0000 (09:30 +0000)]
    Rubber-stamped by Mark Rowe.

        - correct a copyright header accidently reverted in r43964

        * WebKitSystemInterface.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43965 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoWebCore:
mitz@apple.com [Thu, 21 May 2009 09:16:42 +0000 (09:16 +0000)]
WebCore:

        Reviewed by Anders Carlsson.

        - WebCore part of <rdar://problem/6901751> REGRESSION (r35515): Tiger
          crash painting the selection on registration page of car2go.com

        A bug in old versions of Core Graphics causes memory corruption to occur
        when clipping under certain conditions. Make the clipping functions
        fail silently under those conditions.

        Test: fast/block/float/selection-gap-clip-out-tiger-crash.html

        * WebCore.Tiger.exp: Added wkCGContextIsSafeToClip
        * platform/graphics/cg/GraphicsContextCG.cpp:
        (WebCore::safeCGContextClip): Added a wrapper around CGContextClip that,
        on Tiger, bails out if wkCGContextIsSafeToClip() returns false.
        (WebCore::safeCGContextEOClip): Ditto for CGContextEOClip.
        (WebCore::safeCGContextClipToRect): Ditto for CGContextClipToRect.
        (WebCore::GraphicsContext::fillPath): Changed to call the safe variants
        of CGContextClip, CGContextEOClip and CGContextClipToRect.
        (WebCore::GraphicsContext::strokePath): Ditto.
        (WebCore::GraphicsContext::fillRect): Ditto.
        (WebCore::GraphicsContext::clip): Ditto.
        (WebCore::GraphicsContext::clipOut): Ditto.
        (WebCore::GraphicsContext::clipOutEllipseInRect): Ditto.
        (WebCore::GraphicsContext::clipPath): Ditto.
        (WebCore::GraphicsContext::addInnerRoundedRectClip): Ditto.
        (WebCore::GraphicsContext::strokeRect): Ditto.
        * platform/mac/WebCoreSystemInterface.h: Added wkCGContextIsSafeToClip.
        * platform/mac/WebCoreSystemInterface.mm: Added wkCGContextIsSafeToClip.

WebKit/mac:

        Reviewed by Anders Carlsson.

        - WebKit part of <rdar://problem/6901751> REGRESSION (r35515): Tiger
          crash painting the selection on registration page of car2go.com

        * WebCoreSupport/WebSystemInterface.m:
        (InitWebCoreSystemInterface): Added CGContextIsSafeToClip.

WebKitLibraries:

        Reviewed by Anders Carlsson.

        - WebKitSystemInterface part of <rdar://problem/6901751> REGRESSION
          (r35515): Tiger crash painting the selection on registration page of
          car2go.com

        * WebKitSystemInterface.h:
        * libWebKitSystemInterfaceTiger.a:

LayoutTests:

        Reviewed by Anders Carlsson.

        - test for <rdar://problem/6901751> REGRESSION (r35515): Tiger crash
          painting the selection on registration page of car2go.com

        * fast/block/float/selection-gap-clip-out-tiger-crash-expected.txt: Added.
        * fast/block/float/selection-gap-clip-out-tiger-crash.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43964 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Thu, 21 May 2009 09:08:31 +0000 (09:08 +0000)]
2009-05-21  Xan Lopez  <xlopez@igalia.com>

        Unreviewed build fix.

        https://bugs.webkit.org/show_bug.cgi?id=25910
        WebkitGTK r43960 cannot compile

        Do not include local copy of guriescape header, which was
        removed. We now use the copy in glib.

        * platform/gtk/FileSystemGtk.cpp:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43963 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Cameron Zwarich <zwarich@apple.com>
cwzwarich@webkit.org [Thu, 21 May 2009 08:41:29 +0000 (08:41 +0000)]
2009-05-21  Cameron Zwarich  <zwarich@apple.com>

        Reviewed by Maciej Stachowiak.

        Bug 25912: Harden NumberPrototype.cpp by removing use of strcpy()
        <https://bugs.webkit.org/show_bug.cgi?id=25912>

        This causes no change on SunSpider.

        * runtime/NumberPrototype.cpp:
        (JSC::integerPartNoExp): replace strcpy() with memcpy(), ASSERT that the
        temporary buffer has sufficient space to store the result, and move the
        explicit null-termination closer to the memcpy() for easier visual inspection
        of the code.
        (JSC::fractionalPartToString): replace strcpy() with memcpy(), and ASSERT
        that the temporary buffer has sufficient space to store the result. There
        is no explicit null-termination because this is done by the caller. The
        same is already true for exponentialPartToString().
        (JSC::numberProtoFuncToExponential): replace strcpy() with memcpy(), explicitly
        null-terminate the result, and ASSERT that the temporary buffer has sufficient
        space to store the result.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43962 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Eric Seidel <eric@webkit.org>
eric@webkit.org [Thu, 21 May 2009 08:02:27 +0000 (08:02 +0000)]
2009-05-21  Eric Seidel  <eric@webkit.org>

        Reviewed by Alexey Proskuryakov.

        Fix file-input-files-access which was passing for the wrong reasons.

        The log was pushing the drop target off screen and the drops were failing!

        * editing/pasteboard/file-input-files-access-expected.txt:
        * editing/pasteboard/resources/file-input-files-access.js:
        (testOrderedDraggingWithDirectory):
        (runTest):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43961 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Sam Weinig <sam@webkit.org>
weinig@apple.com [Thu, 21 May 2009 05:52:44 +0000 (05:52 +0000)]
2009-05-20  Sam Weinig  <sam@webkit.org>

        Reviewed by Cameron Zwarich.

        Cleanup the JSGlobalData when exiting early with the usage statement in jsc.

        * jsc.cpp:
        (printUsageStatement):
        (parseArguments):
        (jscmain):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43960 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Adam Barth <abarth@webkit.org>
abarth@webkit.org [Thu, 21 May 2009 04:06:34 +0000 (04:06 +0000)]
2009-05-20  Adam Barth  <abarth@webkit.org>

        Reviewed by Dimitri Glazkov.

        https://bugs.webkit.org/show_bug.cgi?id=25706

        It turns out the CurrentContext wasn't currect either because it gave
        us the frame for the write() function itself.  We actually want the
        *calling* context.

        * bindings/v8/custom/V8HTMLDocumentCustom.cpp:
        (WebCore::CALLBACK_FUNC_DECL):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43958 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Thu, 21 May 2009 02:52:47 +0000 (02:52 +0000)]
2009-05-20  Gustavo Noronha Silva  <gns@gnome.org>

        Skipping a new test.

        * platform/gtk/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43957 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Thu, 21 May 2009 02:45:26 +0000 (02:45 +0000)]
2009-05-20  Gustavo Noronha Silva  <gns@gnome.org>

        Skipping test that fails intermitently.

        * platform/gtk/Skipped:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43956 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago[GTK] Implement Image::loadPlatformResource.
zecke@webkit.org [Thu, 21 May 2009 02:43:19 +0000 (02:43 +0000)]
[GTK] Implement Image::loadPlatformResource.

Install the four resources we are using. For the icons
where an icon name is specified by freedeskop.org try to
map the WebCore name to that name and fallback to the
internal icon. For everything else use the internal icon
name. Remove the temporary link stub as this is not
required any more.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43955 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago[GTK] Gtk+ wants to have a themable urlIcon too.
zecke@webkit.org [Thu, 21 May 2009 02:42:42 +0000 (02:42 +0000)]
[GTK] Gtk+ wants to have a themable urlIcon too.

https://bugs.webkit.org/show_bug.cgi?id=19357

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43954 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Gustavo Noronha Silva <gns@gnome.org>
kov@webkit.org [Thu, 21 May 2009 02:41:18 +0000 (02:41 +0000)]
2009-05-20  Gustavo Noronha Silva  <gns@gnome.org>

        Rubber-stamped by Mark Rowe.

        Add GTK+-specific results for canvas-composite-alpha. This test
        has 'failing' cases that pass on GTK+/cairo.

        * platform/gtk/fast/canvas/canvas-composite-alpha-expected.txt: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43953 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Stephanie Lewis <slewis@apple.com>
slewis@apple.com [Thu, 21 May 2009 02:22:32 +0000 (02:22 +0000)]
2009-05-20  Stephanie Lewis  <slewis@apple.com>

        Update the order files.  <rdar://problem/6881750> Generate new order files.

        * JavaScriptCore.order:
       * WebCore.order:
        * WebKit.order:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43951 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoFix some assertion failures in plug-in tests on the Windows debug build bot.
mrowe@apple.com [Thu, 21 May 2009 02:14:14 +0000 (02:14 +0000)]
Fix some assertion failures in plug-in tests on the Windows debug build bot.

Reviewed by Gavin Barraclough.

* plugins/PluginStream.cpp:
(WebCore::PluginStream::startStream): Switch to using protocolIsJavaScript.
* plugins/PluginView.cpp:
(WebCore::scriptStringIfJavaScriptURL): Ditto.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43949 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoWebCore:
beidson@apple.com [Wed, 20 May 2009 23:58:25 +0000 (23:58 +0000)]
WebCore:

2009-05-20  Brady Eidson  <beidson@apple.com>

        Reviewed by Darin Adler.

        More protocolInHTTPFamily() cleanup.

        * xml/XMLHttpRequest.cpp:
        (WebCore::XMLHttpRequest::send):

WebKit/win:

2009-05-20  Brady Eidson  <beidson@apple.com>

        Reviewed by Darin Adler.

        More protocolInHTTPFamily() cleanup.

        * WebHistory.cpp:
        (WebHistory::visitedURL):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43946 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoPrevent search engines from getting creative in their descriptions of webkit.org...
mrowe@apple.com [Wed, 20 May 2009 23:57:19 +0000 (23:57 +0000)]
Prevent search engines from getting creative in their descriptions of webkit.org content.

Rubber-stamped by Eric Seidel.

* blog/wp-content/themes/webkit/header.php:
* header.inc:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43944 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Christian Dywan <christian@twotoasts.de>
christian@webkit.org [Wed, 20 May 2009 23:37:34 +0000 (23:37 +0000)]
2009-05-21  Christian Dywan  <christian@twotoasts.de>

        Reviewed by Jan Alonzo.

        * platform/network/soup/ResourceHandleSoup.cpp:
        (WebCore::parseDataUrl): Always use Glib Base64 decoding.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43943 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Sam Weinig <sam@webkit.org>
weinig@apple.com [Wed, 20 May 2009 23:36:47 +0000 (23:36 +0000)]
2009-05-20  Sam Weinig  <sam@webkit.org>

        Reviewed by Oliver Hunt.

        Fix for <rdar://problem/6905475>
        Office 2008: Entourage: Crash when you switch to "My Day" view

        The host function callback callObjCFallbackObject was not annotated
        with JSC_HOST_CALL which meant that when it was called by the JIT, which
        expects the JSC_HOST_CALL calling convention, the arguments are not in the
        expected places.

        * bridge/objc/objc_runtime.mm:
        (JSC::Bindings::callObjCFallbackObject): Added missing JSC_HOST_CALL
        annotation on callObjCFallbackObject.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43942 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 David Levin <levin@chromium.org>
levin@chromium.org [Wed, 20 May 2009 23:32:17 +0000 (23:32 +0000)]
2009-05-20  David Levin  <levin@chromium.org>

        Not Reviewed, build fix.

        Need to disable this assert while the chormium code is fixed for it.

        * platform/KURLGoogle.cpp:
        (WebCore::KURL::protocolIs):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43941 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoAdded bugs.webkit.org URL
ddkilzer@apple.com [Wed, 20 May 2009 23:13:04 +0000 (23:13 +0000)]
Added bugs.webkit.org URL

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43940 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago Test for <rdar://problem/5666078> Cookie parsing terminates at the first...
ddkilzer@apple.com [Wed, 20 May 2009 23:09:43 +0000 (23:09 +0000)]
    Test for <rdar://problem/5666078> Cookie parsing terminates at the first semicolon, ignoring quotes (16699)

        Reviewed by Brady Eidson.

        * http/tests/cookies/double-quoted-value-with-semi-colon-expected.txt: Added.
        * http/tests/cookies/double-quoted-value-with-semi-colon.html: Added.
        * http/tests/cookies/resources/double-quoted-value-with-semi-colon.js: Added.
        * platform/mac-tiger/Skipped: Added
        double-quoted-value-with-semi-colon.html since CFNetwork on
        Tiger stops parsing cookie values when it sees a semi-colon,
        even inside double quotes.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43939 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 David Levin <levin@chromium.org>
levin@chromium.org [Wed, 20 May 2009 23:01:52 +0000 (23:01 +0000)]
2009-05-20  David Levin  <levin@chromium.org>

        Not Reviewed, build fix.

        Change KURLGoogle.cpp to mirror the KURL.cpp changes done in
        http://trac.webkit.org/changeset/43929.  Fix assert placement.

        * platform/KURLGoogle.cpp:
        (WebCore::KURL::protocolIs):
        (WebCore::protocolIs):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43938 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Brady Eidson <beidson@apple.com>
beidson@apple.com [Wed, 20 May 2009 22:35:10 +0000 (22:35 +0000)]
2009-05-20  Brady Eidson  <beidson@apple.com>

        Reviewed by Darin Adler.

        <rdar://problem/6905336> REGRESSION: "Clear History" does not save empty history to disk

        * History/WebHistory.mm:
        (-[WebHistoryPrivate data]): If there are no entries, return an empty NSData instead of nil.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43937 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Nate Chapin <japhet@google.com>
levin@chromium.org [Wed, 20 May 2009 22:26:50 +0000 (22:26 +0000)]
2009-05-20  Nate Chapin  <japhet@google.com>

        Reviewed by Dimitri Glazkov.

        Upstream V8 bindings for V8DomWindow.
        https://bugs.webkit.org/show_bug.cgi?id=25869

        * bindings/v8/custom/V8DOMWindowCustom.cpp:
        (WebCore::V8Custom::WindowSetTimeoutImpl): Moved from v8_custom.cpp.
        (WebCore::isAscii): Moved from v8_custom.cpp.
        (WebCore::convertBase64): Moved from v8_custom.cpp.
        (WebCore::CALLBACK_FUNC_DECL): Moved from v8_custom.cpp.
        (WebCore::eventNameFromAttributeName): Moved from v8_custom.cpp.
        (WebCore::ACCESSOR_SETTER): Moved from v8_custom.cpp.
        (WebCore::ACCESSOR_GETTER): Moved from v8_custom.cpp.
        (WebCore::NAMED_ACCESS_CHECK): Moved from v8_custom.cpp.
        (WebCore::INDEXED_ACCESS_CHECK): Moved from v8_custom.cpp.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43936 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-21 Christian Dywan <christian@twotoasts.de>
christian@webkit.org [Wed, 20 May 2009 22:18:34 +0000 (22:18 +0000)]
2009-05-21  Christian Dywan  <christian@twotoasts.de>

        Rubberstamped by Gustavo Noronha.

        * platform/network/curl/ResourceHandleManager.cpp:
        (WebCore::parseDataUrl): Remove Gtk/ Glib Base64 decoding.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43935 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Christian Dywan <christian@twotoasts.de>
christian@webkit.org [Wed, 20 May 2009 22:04:00 +0000 (22:04 +0000)]
2009-05-20  Christian Dywan  <christian@twotoasts.de>

        Reviewed by Gustavo Noronha.

        * GNUmakefile.am:
        * platform/gtk/guriescape.c:
        * platform/gtk/guriescape.h: Remove copies of g_uri_escape_string and
        g_uri_unescape_string now that we require Glib 2.16.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43934 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 David Levin <levin@chromium.org>
levin@chromium.org [Wed, 20 May 2009 20:55:24 +0000 (20:55 +0000)]
2009-05-20  David Levin  <levin@chromium.org>

        Not Reviewed, build fix.

        Change KURLGoogle.cpp to mirror changes done in
        http://trac.webkit.org/changeset/43929.

        * platform/KURLGoogle.cpp:
        (WebCore::protocolIsJavaScript):
        (WebCore::protocolIs):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43933 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Brady Eidson <beidson@apple.com>
beidson@apple.com [Wed, 20 May 2009 19:53:22 +0000 (19:53 +0000)]
2009-05-20  Brady Eidson  <beidson@apple.com>

        Rubberstamped by Sam Weinig

        At all call sites that check if a KURL has either the http or https protocol,
        use the KURL::protocolInHTTPFamily() accessor, instead.

        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::iconURL):
        (WebCore::FrameLoader::didOpenURL):
        (WebCore::FrameLoader::open):

        * loader/appcache/ApplicationCache.cpp:
        (WebCore::ApplicationCache::requestIsHTTPOrHTTPSGet):

        * loader/loader.cpp:
        (WebCore::Loader::load):

        * xml/XMLHttpRequest.cpp:
        (WebCore::XMLHttpRequest::send):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43932 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Anders Carlsson <andersca@apple.com>
andersca@apple.com [Wed, 20 May 2009 19:39:09 +0000 (19:39 +0000)]
2009-05-20  Anders Carlsson  <andersca@apple.com>

        Reviewed by Darin Adler and Kevin Decker.

        WebKit side of <rdar://problem/6895072>

        Pass the localization as a launch property.

        * Plugins/Hosted/NetscapePluginHostManager.mm:
        (WebKit::NetscapePluginHostManager::spawnPluginHost):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43931 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoFix ChangeLog typo
beidson@apple.com [Wed, 20 May 2009 19:36:48 +0000 (19:36 +0000)]
Fix ChangeLog typo

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43930 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoWebCore:
beidson@apple.com [Wed, 20 May 2009 19:31:01 +0000 (19:31 +0000)]
WebCore:

2009-05-20  Brady Eidson  <beidson@apple.com>

        Reviewed by Darin Adler.

        Fix <rdar://problem/6904941> and https://bugs.webkit.org/show_bug.cgi?id=25300

        KURL parsing changes back in r30243 and r30815 made javascript urls that use the form
        "javascript://" be detected as invalid hierarchical URLs.  When a KURL is marked as
        invalid, you can no longer ask it a question such as "is your protocol javascript?"
        Therefore FrameLoader wouldn't recognize them as javascript URLs and instead try to
        navigated to them.

        URL parsing rules aside, such URLs are in use in the wild (Microsoft's Virtual Earth
        being a high profile example) and actually represent valid javascript.

        Whenever checking for javascript urls, the new protocolIsJavaScript() should be used
        as it functions on a String which doesn't have to pass KURLs parsing rules.

        Test: fast/loader/javascript-url-hierarchical-execution.html

        * bindings/js/JSAttrCustom.cpp:
        (WebCore::JSAttr::setValue): Use protocolIsJavaScript().
        * bindings/js/JSDOMWindowCustom.cpp:
        (WebCore::JSDOMWindow::setLocation): Ditto.
        (WebCore::createWindow): Ditto.
        (WebCore::JSDOMWindow::open): Ditto.
        * bindings/js/JSElementCustom.cpp:
        (WebCore::allowSettingSrcToJavascriptURL): Ditto.
        * bindings/js/JSHTMLFrameElementCustom.cpp:
        (WebCore::allowSettingJavascriptURL): Ditto.
        * bindings/js/JSHTMLIFrameElementCustom.cpp:
        (WebCore::JSHTMLIFrameElement::setSrc): Ditto.
        * bindings/js/JSLocationCustom.cpp:
        (WebCore::navigateIfAllowed): Ditto.
        (WebCore::JSLocation::reload): Ditto.
        * editing/markup.cpp:
        (WebCore::appendQuotedURLAttributeValue): Ditto.
        * html/HTMLAnchorElement.cpp:
        (WebCore::HTMLAnchorElement::parseMappedAttribute): Ditto.
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::requestFrame): Ditto.
        (WebCore::FrameLoader::submitForm): Ditto.
        (WebCore::FrameLoader::executeIfJavaScriptURL): Ditto.

        * platform/KURL.cpp:
        (WebCore::KURL::init): Use protocolIsJavaScript().
        (WebCore::KURL::protocolIs): ASSERT that the protocol being asked about is not javascript.
           Such checks should go through protocolIsJavaScript() instead.
        (WebCore::encodeRelativeString): Use protocolIsJavaScript().
        (WebCore::protocolIsJavaScript):
        * platform/KURL.h:

LayoutTests:

2009-05-20  Brady Eidson  <beidson@apple.com>

        Reviewed by Darin Adler.

        Fix <rdar://problem/6904941> and https://bugs.webkit.org/show_bug.cgi?id=25300

        * fast/loader/javascript-url-hierarchical-execution-expected.txt: Added.
        * fast/loader/javascript-url-hierarchical-execution.html: Added.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43929 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years agoVersioning.
mrowe@apple.com [Wed, 20 May 2009 19:27:18 +0000 (19:27 +0000)]
Versioning.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43926 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago[plugins] PluginPackage*.cpp for Gtk+/Qt misassigned hasproperty and hasmethod
zecke@webkit.org [Wed, 20 May 2009 17:35:22 +0000 (17:35 +0000)]
[plugins] PluginPackage*.cpp for Gtk+/Qt misassigned hasproperty and hasmethod

https://bugs.webkit.org/show_bug.cgi?id=24510

PluginPackageGtk.cpp and PluginPackageQt.cpp assigned _NPN_HasMethod
to hasproperty and _NPN_HasProperty to hashmethod. This commit is fixing
it. It is also adding test code to PluginObject.cpp and it will invoke
this code from a new test.

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43923 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Sam Weinig <sam@webkit.org>
weinig@apple.com [Wed, 20 May 2009 17:31:31 +0000 (17:31 +0000)]
2009-05-20  Sam Weinig  <sam@webkit.org>

        Reviewed by Timothy Hatcher.

        Make absoluteRectsForRange and absoluteQuadsForRange non-virtual
        and only exist on RenderText.

        * dom/Range.cpp:
        (WebCore::Range::textRects):
        * inspector/InspectorController.cpp:
        (WebCore::InspectorController::drawNodeHighlight): Use absoluteQuads instead
        of absoluteQuadsForRange since no range is being specifiec.
        * rendering/RenderObject.cpp:
        * rendering/RenderObject.h:
        * rendering/RenderText.h:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43922 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago - fix the build after r43918
mitz@apple.com [Wed, 20 May 2009 17:05:20 +0000 (17:05 +0000)]
    - fix the build after r43918

        * WebCoreSupport/WebChromeClient.h:
        (WebChromeClient::setCursor):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43921 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Xan Lopez <xlopez@igalia.com>
xan@webkit.org [Wed, 20 May 2009 16:59:40 +0000 (16:59 +0000)]
2009-05-20  Xan Lopez  <xlopez@igalia.com>

        Reviewed by Gustavo Noronha.

        Implement AtkObject::focus-event and
        AtkObject::state-changed:focused signal emission.

        Adds a new handleFocusedUIElementChangedWithRenderers, only in the
        GTK port, called at the same point than
        handleFocusedUIElementChanged but with two parameters, the old and
        the newly focused RenderObjects. We need this, since the ATK
        signals require us to pass the objects involved in the focus
        change as arguments.

        * accessibility/AXObjectCache.h:
        * accessibility/gtk/AXObjectCacheAtk.cpp:
        (WebCore::AXObjectCache::handleFocusedUIElementChangedWithRenderers):
        * dom/Document.cpp:
        (WebCore::Document::setFocusedNode):

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43920 268f45cc-cd09-0410-ab3c-d52691b4dbfc

12 years ago2009-05-20 Ariya Hidayat <ariya.hidayat@nokia.com>
ariya@webkit.org [Wed, 20 May 2009 16:23:03 +0000 (16:23 +0000)]
2009-05-20  Ariya Hidayat  <ariya.hidayat@nokia.com>

        Reviewed by Simon Hausmann and Holger Freyther.

        [Qt] Add renderHints property to QWebView.

        * Api/qwebview.cpp:
        (QWebViewPrivate::QWebViewPrivate):
        (QWebView::renderHints):
        (QWebView::setRenderHints):
        (QWebView::setRenderHint):
        (QWebView::paintEvent):
        * Api/qwebview.h:
        * tests/qwebview/qwebview.pro: Added.
        * tests/qwebview/tst_qwebview.cpp: Added.
        (tst_QWebView::tst_QWebView):
        (tst_QWebView::~tst_QWebView):
        (tst_QWebView::init):
        (tst_QWebView::cleanup):
        (tst_QWebView::renderHints):
        * tests/tests.pro:

git-svn-id: http://svn.webkit.org/repository/webkit/trunk@43919 268f45cc-cd09-0410-ab3c-d52691b4dbfc