Not reviewed, gdk build fix.

        * platform/gdk/FrameGdk.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23992 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by John.

        Test for Repro crash due to infinite recursion in HTMLParser::handleError @ youos.com
        <rdar://problem/5237811>

        * fast/table/incomplete-table-in-fragment-hang-expected.txt: Added.
        * fast/table/incomplete-table-in-fragment-hang.html: Added.

WebCore:

        Reviewed by John.

        Fix Repro crash due to infinite recursion in HTMLParser::handleError @ youos.com
        <rdar://problem/5237811>

        It is possible to add table parts (thead etc) without table ancestor to a document fragment. If a new table element
        was added to such a part, as in

        div.innerHTML = '<tbody><table>';

        the parser error handling code would try to pop the previous table as normal. However since
        the table does not actually exist nothing would happen and parser would go to infinite recursion.

        Solution here is to pop table parts one by one when handling the error inside a fragment instead of trying to pop
        the table straight away (as it might not exist).

        * html/HTMLParser.cpp:
        (WebCore::HTMLParser::handleError):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23991 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Patch from Qing Zhao to fix combobox rendering

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23989 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Compile

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23988 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Mitz.

        Test for http://bugs.webkit.org/show_bug.cgi?id=14504
        REGRESSION (r23573): Amazon Book Excerpt can not be viewed in recent nightly

        * fast/dom/Window/window-property-shadowing-expected.txt: Added.
        * fast/dom/Window/window-property-shadowing.html: Added.

WebCore:

        Reviewed by Mitz.

        Patch for http://bugs.webkit.org/show_bug.cgi?id=14504
        REGRESSION (r23573): Amazon Book Excerpt can not be viewed in recent nightly

        - Made it possible (again) to shadow window properties with local variable names.

        Test: fast/dom/Window/window-property-shadowing.html

        * bindings/js/JSDOMWindowCustom.cpp:
        (WebCore::JSDOMWindow::customPut):
        * bindings/js/kjs_window.cpp:
        (KJS::Window::put):
        * bindings/scripts/CodeGeneratorJS.pm:
        * page/DOMWindow.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23987 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Adam.

        Fixed <rdar://problem/5225119> support smart pasting on Windows

        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
            Added new files
        * editing/SmartReplace.cpp: Added.
        (WebCore::isCharacterSmartReplaceExempt):
            Added non-CF implementation placeholder
        * editing/SmartReplace.h: Added.
        * editing/SmartReplaceCF.cpp: Added.
        (WebCore::getSmartSet):
        (WebCore::isCharacterSmartReplaceExempt):
            Added new implementation

        * bridge/win/FrameWin.cpp:
        * page/Frame.h:
        * page/mac/FrameMac.mm:
            Removed old declaration and implementation

        * editing/ReplaceSelectionCommand.cpp:
        (WebCore::ReplaceSelectionCommand::doApply):
            Changed to call the new function instead of old

        * page/mac/WebCoreFrameBridge.h:
        * page/mac/WebCoreFrameBridge.mm:
            Removed old declaration and implementation
        (-[WebCoreFrameBridge smartInsertForString:replacingRange:beforeString:afterString:]):
            Changed to call the new function instead of old

        * platform/win/PasteboardWin.cpp:
        (WebCore::Pasteboard::Pasteboard):
            Register new clipboard format
        (WebCore::Pasteboard::writeSelection):
            Set data for smartpaste format on clipboard
        (WebCore::Pasteboard::canSmartReplace):
            Added non-stub implementation

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23986 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Reenable some code for 64-bit now that <rdar://problem/5311640> and <rdar://problem/5311690> are resolved.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23985 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Patch from Qing Zhao: Similar to the PlatformScrollbar fixes, adjust the
painter to make the styler draw all subcontrols, and in this case the focus
rect, in the proper location

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23984 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2007-07-04  Mitz Pettel  <mitz@webkit.org>

        Reviewed by Maciej.

        - fix http://bugs.webkit.org/show_bug.cgi?id=14495
          REGRESSION: Wrong cursor is displayed during drag and drop when text is selected
          <rdar://problem/5312384>

        No test possible because DumpRenderTree cannot query the current cursor.

        * page/EventHandler.cpp:
        (WebCore::EventHandler::selectCursor): Changed into a member function and
        added that if the last mouse down event could not have started a
        selection, then the cursor should not change to an I-beam unconditionally.
        (WebCore::EventHandler::handleMouseMoveEvent):
        * page/EventHandler.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23983 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2007-07-04  Holger Hans Peter Freyther  <zecke@selfish.org>

        Reviewed by Rob Buis.

        A little progress in tracking dependencies.

        * WebCore.pro: Add DEPENDPATH for the Gdk port.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23982 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    * StringsNotToBeLocalized.txt: Updated for recent changes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23978 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin.

        Remove HIWebView in 64-bit.

        Also disable a few calls that are now gone in 64-bit and filed these bugs:

        <rdar://problem/5311653> WebKit needs to adopt HIGetMousePosition where we used GetGlobalMouse before
        <rdar://problem/5311648> WebKit needs to move off of CopyEvent in WebBaseNetscapePluginView
        <rdar://problem/5311640> WebKit needs a new solution for HISearchWindowShow on 64-bit

        * Carbon/CarbonUtils.m:
        * Carbon/CarbonWindowAdapter.m:
        * Carbon/CarbonWindowContentView.m:
        * Carbon/CarbonWindowFrame.m:
        * Carbon/HIViewAdapter.m:
        (SetViewNeedsDisplay):
        * Carbon/HIWebView.m:
        (Draw):
        (SyncFrame):
        * Configurations/WebKit.xcconfig:
        * Plugins/WebBaseNetscapePluginView.mm:
        (+[WebBaseNetscapePluginView getCarbonEvent:]):
        (TSMEventHandler):
        * WebKit.LP64.exp: Added.
        * WebView/WebView.mm:
        (-[WebView _searchWithSpotlightFromMenu:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23977 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Disable the strict aliasing warning until <rdar://problem/5311093> is fixed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23976 268f45cc-cd09-0410-ab3c-d52691b4dbfc

64-bit build fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23975 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fixed update-webkit-localizable-strings to work with new extract-localizable-strings

Got rid of extract-webkit-localizable-strings because it was trying to
do update-webkit-localizable-strings' job.

Rubberstamped by Darin.

* Scripts/extract-webkit-localizable-strings: Removed.
* Scripts/update-webkit-localizable-strings: Renamed from
extract-localizable-strings.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23974 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Merge the Windows and Mac localized strings and exceptions files

WebKit:

        Merge the Windows and Mac localized strings and exceptions files

        Reviewed by Darin and Anders.

        * English.lproj/Localizable.strings: Added Windows strings.
        * StringsNotToBeLocalized.txt: Renamed from WebKit/English.lproj/StringsNotToBeLocalized.txt.

WebKit/win:

        Merge the Windows and Mac localized strings and exceptions files

        Reviewed by Darin and Anders.

        * English.lproj/Localizable.strings: Added Mac strings.
        * English.lproj/StringsNotToBeLocalized.txt: Removed.

WebKitTools:

        Modify scripts to keep Mac/Windows localized strings in sync

        extract-webkit-localizable-strings now handles extracting strings from
        both the Mac and Windows WebKit ports and updating their respective
        Localizable.strings files. extract-localizable-strings is now really a
        piece of plumbing rather than a top-level tool.

        Reviewed by Darin and Anders.

        * Scripts/extract-localizable-strings: Changed to handle multiple
        directories.
        * Scripts/extract-webkit-localizable-strings: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23972 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Removed printf I accidently left in.

        * Misc/WebNSURLExtras.m: (mapHostNames):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23971 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Tim Hatcher.

        - tests for <rdar://problem/5221297> Font family name in the cssText for a DOMCSSStyleDeclaration needs quotes

        * editing/style/font-family-with-space-expected.txt: Added.
        * editing/style/font-family-with-space.html: Added.

        * fast/inspector/style-expected.txt: Updated expected results to expect "Lucida Grande" in quote marks.

WebCore:

        Reviewed by Tim Hatcher.

        - fix <rdar://problem/5221297> Font family name in the cssText for a DOMCSSStyleDeclaration needs quotes

        Test: editing/style/font-family-with-space.html

        * css/FontFamilyValue.cpp:
        (WebCore::isValidCSSIdentifier): Added. Implements the same rule that the CSS lexer does.
        (WebCore::quoteStringIfNeeded): Changed to call isValidCSSIdentifier instead of just
        checking for a leading "#" character.
        * editing/markup.cpp:
        (WebCore::escapeTextForMarkup): Changed to take a String parameter for better efficiency.
        (WebCore::renderedText): Changed to return a String for better efficiency.
        (WebCore::addNamespace): Updated to pass String to escapeTextForMarkup.
        (WebCore::startMarkup): Updated to pass String to escapeTextForMarkup. Added missing call
        to escapeTextForMarkup in the special case for the style property.
        (WebCore::createMarkup): Changed from single quotes to double quotes and also added missing
        call to escapeTextForMarkup in two special cases for the style property.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23969 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Updating comment

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23967 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin.

        Twelfth round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        - Add casts and accompanying FIXMEs to avoid remaining compiler errors.

        * WebCore.xcodeproj/project.pbxproj:
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::opened):
        (WebCore::FrameLoader::sendRemainingDelegateMessages):
        * loader/FrameLoader.h:
        * loader/ResourceLoader.cpp:
        (WebCore::ResourceLoader::didReceiveData):
        * platform/network/mac/ResourceHandleMac.mm:
        (-[WebCoreResourceHandleAsDelegate connection:didReceiveData:lengthReceived:]):
        (-[WebCoreResourceHandleAsDelegate connection:willStopBufferingData:]):
        * platform/network/mac/ResourceResponseMac.mm:
        (WebCore::ResourceResponse::nsURLResponse):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23964 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Darin.

        Test for: <rdar://problem/5292988> domain names shouldn't contain ignorable characters

        * fast/encoding/idn-security-expected.txt: Added.
        * fast/encoding/idn-security.html: Added.
        * win/Skipped: Blocked by <rdar://problem/5301954>.

WebKit:

        Reviewed by Darin.

        Fix for: <rdar://problem/5292988> domain names shouldn't contain ignorable characters

        * Misc/WebNSURLExtras.m:
        (isLookalikeCharacter): Renamed.  Also excludes any non-printable character,
         any character considered as whitespace that isn't already converted to a space by ICU, any ignorable character,
         and any character excluded in Mozilla's blacklist: http://kb.mozillazine.org/Network.IDN.blacklist_chars
        (allCharactersInIDNScriptWhiteList):

WebKitTools:

        Reviewed by Darin.

        Extended DumpRenderTree to test encoding and decoding host names.

        * DumpRenderTree/DumpRenderTree.m:
        (+[LayoutTestController isSelectorExcludedFromWebScript:]): Added cases for encodeHostName and decodeHostName.
        (+[LayoutTestController webScriptNameForSelector:]): ditto.
        (-[LayoutTestController decodeHostName:]): Added.
        (-[LayoutTestController encodeHostName:]): Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23963 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin.

        Eleventh round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        Add functions to test long long and unsigned long long.

        * DumpRenderTree/DumpRenderTree.m:
        (+[LayoutTestController isSelectorExcludedFromWebScript:]):
        (+[LayoutTestController webScriptNameForSelector:]):
        (-[LayoutTestController objCLongLongRoundTrip:]):
        (-[LayoutTestController objCUnsignedLongLongRoundTrip:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23961 268f45cc-cd09-0410-ab3c-d52691b4dbfc

JavaScriptCore:

        Reviewed by Darin.

        Eleventh round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        - Fixes a real bug where where we were setting long long and unsigned long long
          values to a long field.

        * bindings/objc/objc_utility.mm:
        (KJS::Bindings::convertValueToObjcValue):

LayoutTests:

        Reviewed by Darin.

        Eleventh round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        * fast/objc: Added.
        * fast/objc/longlongTest-expected.txt: Added.
        * fast/objc/longlongTest.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23960 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Maciej.

        - fix <rdar://problem/5310848> WebDataSource lifetime problem -- may be cause of the leaks seen on the buildbot

        * WebView/WebDataSource.mm:
        (-[WebDataSourcePrivate dealloc]): Added a call to the new detachDataSource function.
        (-[WebDataSourcePrivate finalize]): Ditto.

        * WebView/WebDocumentLoaderMac.h: Added detachDataSource function to be used when the
        WebDataSource is deallocated. Added retain/releaseDataSource helper functions to be
        used to retain and release the data source object. Replaced the m_hasEverBeenDetached
        boolean with a more primitive and hence easier to understand m_isDataSourceRetained boolean.

        * WebView/WebDocumentLoaderMac.mm:
        (WebDocumentLoaderMac::WebDocumentLoaderMac): Initialize m_isDataSourceRetained to false.
        (WebDocumentLoaderMac::setDataSource): Call retainDataSource instead of calling HardRetain
        on the dataSource parameter. Also updated a comment.
        (WebDocumentLoaderMac::attachToFrame): Call retainDataSource unconditionally rather than
        trying to use m_hasEverBeenDetached to decide if a retain is needed. Also got rid of an
        assertion that m_loadingResources is empty -- not important any more.
        (WebDocumentLoaderMac::detachFromFrame): Call releaseDataSource instead of using
        HardRelease, but only if m_loadingResources is empty. If it's non-empty, then we'll
        do the releaseDataSource later in decreaseLoadCount.
        (WebDocumentLoaderMac::increaseLoadCount): Call retainDataSource unconditionally
        rather than calling HardRetain only if the old set of resources was empty.
        (WebDocumentLoaderMac::decreaseLoadCount): Call releaseDataSource if m_loadingResources
        is empty and we're not attached to a frame. If we are attached to a frame, then
        we'll do the releaseDataSource later in detachFromFrame.
        (WebDocumentLoaderMac::retainDataSource): Added. Calls CFRetain, but only if the data
        source is not already retained (according to the boolean).
        (WebDocumentLoaderMac::releaseDataSource): Added. Calls CFRelease, but only if the data
        source is currently retained (according to the boolean).
        (WebDocumentLoaderMac::detachDataSource): Added. Sets m_dataSource to nil. Since this
        is only called from WebDataSource's dealloc and finalize methods, it won't ever be called
        when the m_isDataSourceRetained boolean is true.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23957 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Adam.

        <rdar://problem/5093862>
        Nothing shown in plugin area when trying to play content using the DivX plugin

        Defer calling SetWindow for the DivX plugin, just like we do with the WMP plugin.

        * plugins/win/PluginViewWin.cpp:
        (WebCore::PluginViewWin::determineQuirks):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23956 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Brady Eidson.

        Tenth round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        - Add explicit casts.

        * kjs/dtoa.cpp:
        (Bigint::):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23955 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    - forgot to check in one file in the fix for <rdar://problem/5307880>
          some classes need finalize methods because of non-trivial work done in dealloc methods

        * WebView/WebView.mm: (-[WebViewPrivate finalize]): Delete identifierMap so it doesn't leak.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23954 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Brady Eidson.

        Ninth round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        - Fix function signature to use unsigned long instead of unsigned long long.

        * loader/ProgressTracker.cpp:
        (WebCore::ProgressTracker::completeProgress):
        * loader/ProgressTracker.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23953 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Oliver.

        Eighth round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        - Convert SVG implementation files to use floats instead of doubles
          to match the spec/IDLs.

        * bindings/scripts/CodeGeneratorJS.pm:
        * ksvg2/svg/SVGAngle.cpp:
        * ksvg2/svg/SVGAnimatedTemplate.h:
        * ksvg2/svg/SVGAnimationElement.cpp:
        (WebCore::SVGAnimationElement::getEndTime):
        (WebCore::SVGAnimationElement::getStartTime):
        (WebCore::SVGAnimationElement::getCurrentTime):
        (WebCore::SVGAnimationElement::getSimpleDuration):
        * ksvg2/svg/SVGAnimationElement.h:
        * ksvg2/svg/SVGFitToViewBox.cpp:
        (WebCore::SVGFitToViewBox::parseViewBox):
        * ksvg2/svg/SVGLength.cpp:
        (WebCore::SVGLength::setValueAsString):
        * ksvg2/svg/SVGLinearGradientElement.cpp:
        (WebCore::SVGLinearGradientElement::buildGradient):
        * ksvg2/svg/SVGMatrix.idl:
        * ksvg2/svg/SVGNumber.idl:
        * ksvg2/svg/SVGParserUtilities.cpp:
        (WebCore::SVGPathParser::parseSVG):
        (WebCore::SVGPathParser::calculateArc):
        (WebCore::SVGPathParser::svgLineToHorizontal):
        (WebCore::SVGPathParser::svgLineToVertical):
        (WebCore::SVGPathParser::svgCurveToCubicSmooth):
        (WebCore::SVGPathParser::svgCurveToQuadratic):
        (WebCore::SVGPathParser::svgCurveToQuadraticSmooth):
        (WebCore::SVGPathParser::svgArcTo):
        * ksvg2/svg/SVGParserUtilities.h:
        * ksvg2/svg/SVGPathElement.cpp:
        (WebCore::SVGPathElement::SVGPathElement):
        (WebCore::SVGPathElement::getTotalLength):
        (WebCore::SVGPathElement::getPointAtLength):
        (WebCore::SVGPathElement::getPathSegAtLength):
        (WebCore::SVGPathElement::createSVGPathSegMovetoAbs):
        (WebCore::SVGPathElement::createSVGPathSegMovetoRel):
        (WebCore::SVGPathElement::createSVGPathSegLinetoAbs):
        (WebCore::SVGPathElement::createSVGPathSegLinetoRel):
        (WebCore::SVGPathElement::createSVGPathSegCurvetoCubicAbs):
        (WebCore::SVGPathElement::createSVGPathSegCurvetoCubicRel):
        (WebCore::SVGPathElement::createSVGPathSegCurvetoQuadraticAbs):
        (WebCore::SVGPathElement::createSVGPathSegCurvetoQuadraticRel):
        (WebCore::SVGPathElement::createSVGPathSegArcAbs):
        (WebCore::SVGPathElement::createSVGPathSegArcRel):
        (WebCore::SVGPathElement::createSVGPathSegLinetoHorizontalAbs):
        (WebCore::SVGPathElement::createSVGPathSegLinetoHorizontalRel):
        (WebCore::SVGPathElement::createSVGPathSegLinetoVerticalAbs):
        (WebCore::SVGPathElement::createSVGPathSegLinetoVerticalRel):
        (WebCore::SVGPathElement::createSVGPathSegCurvetoCubicSmoothAbs):
        (WebCore::SVGPathElement::createSVGPathSegCurvetoCubicSmoothRel):
        (WebCore::SVGPathElement::createSVGPathSegCurvetoQuadraticSmoothAbs):
        (WebCore::SVGPathElement::createSVGPathSegCurvetoQuadraticSmoothRel):
        (WebCore::SVGPathElement::svgMoveTo):
        (WebCore::SVGPathElement::svgLineTo):
        (WebCore::SVGPathElement::svgLineToHorizontal):
        (WebCore::SVGPathElement::svgLineToVertical):
        (WebCore::SVGPathElement::svgCurveToCubic):
        (WebCore::SVGPathElement::svgCurveToCubicSmooth):
        (WebCore::SVGPathElement::svgCurveToQuadratic):
        (WebCore::SVGPathElement::svgCurveToQuadraticSmooth):
        (WebCore::SVGPathElement::svgArcTo):
        (WebCore::SVGPathElement::parseMappedAttribute):
        * ksvg2/svg/SVGPathElement.h:
        * ksvg2/svg/SVGPathSegArc.cpp:
        (WebCore::SVGPathSegArcAbs::SVGPathSegArcAbs):
        (WebCore::SVGPathSegArcAbs::setX):
        (WebCore::SVGPathSegArcAbs::x):
        (WebCore::SVGPathSegArcAbs::setY):
        (WebCore::SVGPathSegArcAbs::y):
        (WebCore::SVGPathSegArcAbs::setR1):
        (WebCore::SVGPathSegArcAbs::r1):
        (WebCore::SVGPathSegArcAbs::setR2):
        (WebCore::SVGPathSegArcAbs::r2):
        (WebCore::SVGPathSegArcAbs::setAngle):
        (WebCore::SVGPathSegArcAbs::angle):
        (WebCore::SVGPathSegArcRel::SVGPathSegArcRel):
        (WebCore::SVGPathSegArcRel::setX):
        (WebCore::SVGPathSegArcRel::x):
        (WebCore::SVGPathSegArcRel::setY):
        (WebCore::SVGPathSegArcRel::y):
        (WebCore::SVGPathSegArcRel::setR1):
        (WebCore::SVGPathSegArcRel::r1):
        (WebCore::SVGPathSegArcRel::setR2):
        (WebCore::SVGPathSegArcRel::r2):
        (WebCore::SVGPathSegArcRel::setAngle):
        (WebCore::SVGPathSegArcRel::angle):
        * ksvg2/svg/SVGPathSegArc.h:
        * ksvg2/svg/SVGPathSegCurvetoCubic.cpp:
        (WebCore::SVGPathSegCurvetoCubicAbs::SVGPathSegCurvetoCubicAbs):
        (WebCore::SVGPathSegCurvetoCubicAbs::setX):
        (WebCore::SVGPathSegCurvetoCubicAbs::x):
        (WebCore::SVGPathSegCurvetoCubicAbs::setY):
        (WebCore::SVGPathSegCurvetoCubicAbs::y):
        (WebCore::SVGPathSegCurvetoCubicAbs::setX1):
        (WebCore::SVGPathSegCurvetoCubicAbs::x1):
        (WebCore::SVGPathSegCurvetoCubicAbs::setY1):
        (WebCore::SVGPathSegCurvetoCubicAbs::y1):
        (WebCore::SVGPathSegCurvetoCubicAbs::setX2):
        (WebCore::SVGPathSegCurvetoCubicAbs::x2):
        (WebCore::SVGPathSegCurvetoCubicAbs::setY2):
        (WebCore::SVGPathSegCurvetoCubicAbs::y2):
        (WebCore::SVGPathSegCurvetoCubicRel::SVGPathSegCurvetoCubicRel):
        (WebCore::SVGPathSegCurvetoCubicRel::setX):
        (WebCore::SVGPathSegCurvetoCubicRel::x):
        (WebCore::SVGPathSegCurvetoCubicRel::setY):
        (WebCore::SVGPathSegCurvetoCubicRel::y):
        (WebCore::SVGPathSegCurvetoCubicRel::setX1):
        (WebCore::SVGPathSegCurvetoCubicRel::x1):
        (WebCore::SVGPathSegCurvetoCubicRel::setY1):
        (WebCore::SVGPathSegCurvetoCubicRel::y1):
        (WebCore::SVGPathSegCurvetoCubicRel::setX2):
        (WebCore::SVGPathSegCurvetoCubicRel::x2):
        (WebCore::SVGPathSegCurvetoCubicRel::setY2):
        (WebCore::SVGPathSegCurvetoCubicRel::y2):
        * ksvg2/svg/SVGPathSegCurvetoCubic.h:
        * ksvg2/svg/SVGPathSegCurvetoCubicSmooth.cpp:
        (WebCore::SVGPathSegCurvetoCubicSmoothAbs::SVGPathSegCurvetoCubicSmoothAbs):
        (WebCore::SVGPathSegCurvetoCubicSmoothAbs::setX):
        (WebCore::SVGPathSegCurvetoCubicSmoothAbs::x):
        (WebCore::SVGPathSegCurvetoCubicSmoothAbs::setY):
        (WebCore::SVGPathSegCurvetoCubicSmoothAbs::y):
        (WebCore::SVGPathSegCurvetoCubicSmoothAbs::setX2):
        (WebCore::SVGPathSegCurvetoCubicSmoothAbs::x2):
        (WebCore::SVGPathSegCurvetoCubicSmoothAbs::setY2):
        (WebCore::SVGPathSegCurvetoCubicSmoothAbs::y2):
        (WebCore::SVGPathSegCurvetoCubicSmoothRel::SVGPathSegCurvetoCubicSmoothRel):
        (WebCore::SVGPathSegCurvetoCubicSmoothRel::setX):
        (WebCore::SVGPathSegCurvetoCubicSmoothRel::x):
        (WebCore::SVGPathSegCurvetoCubicSmoothRel::setY):
        (WebCore::SVGPathSegCurvetoCubicSmoothRel::y):
        (WebCore::SVGPathSegCurvetoCubicSmoothRel::setX2):
        (WebCore::SVGPathSegCurvetoCubicSmoothRel::x2):
        (WebCore::SVGPathSegCurvetoCubicSmoothRel::setY2):
        (WebCore::SVGPathSegCurvetoCubicSmoothRel::y2):
        * ksvg2/svg/SVGPathSegCurvetoCubicSmooth.h:
        * ksvg2/svg/SVGPathSegCurvetoQuadratic.cpp:
        (WebCore::SVGPathSegCurvetoQuadraticAbs::SVGPathSegCurvetoQuadraticAbs):
        (WebCore::SVGPathSegCurvetoQuadraticAbs::setX):
        (WebCore::SVGPathSegCurvetoQuadraticAbs::x):
        (WebCore::SVGPathSegCurvetoQuadraticAbs::setY):
        (WebCore::SVGPathSegCurvetoQuadraticAbs::y):
        (WebCore::SVGPathSegCurvetoQuadraticAbs::setX1):
        (WebCore::SVGPathSegCurvetoQuadraticAbs::x1):
        (WebCore::SVGPathSegCurvetoQuadraticAbs::setY1):
        (WebCore::SVGPathSegCurvetoQuadraticAbs::y1):
        (WebCore::SVGPathSegCurvetoQuadraticRel::SVGPathSegCurvetoQuadraticRel):
        (WebCore::SVGPathSegCurvetoQuadraticRel::setX):
        (WebCore::SVGPathSegCurvetoQuadraticRel::x):
        (WebCore::SVGPathSegCurvetoQuadraticRel::setY):
        (WebCore::SVGPathSegCurvetoQuadraticRel::y):
        (WebCore::SVGPathSegCurvetoQuadraticRel::setX1):
        (WebCore::SVGPathSegCurvetoQuadraticRel::x1):
        (WebCore::SVGPathSegCurvetoQuadraticRel::setY1):
        (WebCore::SVGPathSegCurvetoQuadraticRel::y1):
        * ksvg2/svg/SVGPathSegCurvetoQuadratic.h:
        (WebCore::SVGPathSegCurvetoQuadraticRel::toString):
        * ksvg2/svg/SVGPathSegCurvetoQuadraticSmooth.cpp:
        (WebCore::SVGPathSegCurvetoQuadraticSmoothAbs::SVGPathSegCurvetoQuadraticSmoothAbs):
        (WebCore::SVGPathSegCurvetoQuadraticSmoothAbs::setX):
        (WebCore::SVGPathSegCurvetoQuadraticSmoothAbs::x):
        (WebCore::SVGPathSegCurvetoQuadraticSmoothAbs::setY):
        (WebCore::SVGPathSegCurvetoQuadraticSmoothAbs::y):
        (WebCore::SVGPathSegCurvetoQuadraticSmoothRel::SVGPathSegCurvetoQuadraticSmoothRel):
        (WebCore::SVGPathSegCurvetoQuadraticSmoothRel::setX):
        (WebCore::SVGPathSegCurvetoQuadraticSmoothRel::x):
        (WebCore::SVGPathSegCurvetoQuadraticSmoothRel::setY):
        (WebCore::SVGPathSegCurvetoQuadraticSmoothRel::y):
        * ksvg2/svg/SVGPathSegCurvetoQuadraticSmooth.h:
        * ksvg2/svg/SVGPathSegLineto.cpp:
        (WebCore::SVGPathSegLinetoAbs::SVGPathSegLinetoAbs):
        (WebCore::SVGPathSegLinetoAbs::setX):
        (WebCore::SVGPathSegLinetoAbs::x):
        (WebCore::SVGPathSegLinetoAbs::setY):
        (WebCore::SVGPathSegLinetoAbs::y):
        (WebCore::SVGPathSegLinetoRel::SVGPathSegLinetoRel):
        (WebCore::SVGPathSegLinetoRel::setX):
        (WebCore::SVGPathSegLinetoRel::x):
        (WebCore::SVGPathSegLinetoRel::setY):
        (WebCore::SVGPathSegLinetoRel::y):
        * ksvg2/svg/SVGPathSegLineto.h:
        * ksvg2/svg/SVGPathSegLinetoHorizontal.cpp:
        (WebCore::SVGPathSegLinetoHorizontalAbs::SVGPathSegLinetoHorizontalAbs):
        (WebCore::SVGPathSegLinetoHorizontalAbs::setX):
        (WebCore::SVGPathSegLinetoHorizontalAbs::x):
        (WebCore::SVGPathSegLinetoHorizontalRel::SVGPathSegLinetoHorizontalRel):
        (WebCore::SVGPathSegLinetoHorizontalRel::setX):
        (WebCore::SVGPathSegLinetoHorizontalRel::x):
        * ksvg2/svg/SVGPathSegLinetoHorizontal.h:
        * ksvg2/svg/SVGPathSegLinetoVertical.cpp:
        (WebCore::SVGPathSegLinetoVerticalAbs::SVGPathSegLinetoVerticalAbs):
        (WebCore::SVGPathSegLinetoVerticalAbs::setY):
        (WebCore::SVGPathSegLinetoVerticalAbs::y):
        (WebCore::SVGPathSegLinetoVerticalRel::SVGPathSegLinetoVerticalRel):
        (WebCore::SVGPathSegLinetoVerticalRel::setY):
        (WebCore::SVGPathSegLinetoVerticalRel::y):
        * ksvg2/svg/SVGPathSegLinetoVertical.h:
        * ksvg2/svg/SVGPathSegList.cpp:
        (WebCore::SVGPathSegList::getPathSegAtLength):
        * ksvg2/svg/SVGPathSegList.h:
        * ksvg2/svg/SVGPathSegMoveto.cpp:
        (WebCore::SVGPathSegMovetoAbs::SVGPathSegMovetoAbs):
        (WebCore::SVGPathSegMovetoAbs::setX):
        (WebCore::SVGPathSegMovetoAbs::x):
        (WebCore::SVGPathSegMovetoAbs::setY):
        (WebCore::SVGPathSegMovetoAbs::y):
        (WebCore::SVGPathSegMovetoRel::SVGPathSegMovetoRel):
        (WebCore::SVGPathSegMovetoRel::setX):
        (WebCore::SVGPathSegMovetoRel::x):
        (WebCore::SVGPathSegMovetoRel::setY):
        (WebCore::SVGPathSegMovetoRel::y):
        * ksvg2/svg/SVGPathSegMoveto.h:
        * ksvg2/svg/SVGPolyElement.cpp:
        (WebCore::SVGPolyElement::svgPolyTo):
        * ksvg2/svg/SVGRadialGradientElement.cpp:
        (WebCore::SVGRadialGradientElement::buildGradient):
        * ksvg2/svg/SVGSVGElement.cpp:
        (WebCore::SVGSVGElement::viewport):
        (WebCore::SVGSVGElement::getCurrentTime):
        * ksvg2/svg/SVGStopElement.cpp:
        (WebCore::SVGStopElement::SVGStopElement):
        (WebCore::SVGStopElement::parseMappedAttribute):
        * ksvg2/svg/SVGStopElement.h:
        * ksvg2/svg/SVGTransform.cpp:
        (SVGTransform::translate):
        (SVGTransform::scale):
        * ksvg2/svg/SVGTransformDistance.cpp:
        (WebCore::SVGTransformDistance::addToSVGTransform):
        (WebCore::SVGTransformDistance::distance):
        * ksvg2/svg/SVGTransformable.cpp:
        (WebCore::SVGTransformable::parseTransformAttribute):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23951 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Darin.

        <rdar://problem/5289718>
        http://bugs.webkit.org/show_bug.cgi?id=14437
        CrashTracer: [REGRESSION] 76 crashes in Safari at com.apple.WebCore: WebCore::Frame::settings const + 6 (14437)

        Add test from Alexey Proskuryakov.

        * plugins/plugin-remove-subframe-expected.txt: Added.
        * plugins/plugin-remove-subframe.html: Added.

WebCore:

        Reviewed by Darin.

        <rdar://problem/5289718>
        http://bugs.webkit.org/show_bug.cgi?id=14437
        CrashTracer: [REGRESSION] 76 crashes in Safari at com.apple.WebCore: WebCore::Frame::settings const + 6 (14437)

        Based on a patch from Maxime Britto.

        * page/mac/WebCoreFrameBridge.mm:
        (-[WebCoreFrameBridge stringByEvaluatingJavaScriptFromString:forceUserGesture:]):
        If the script caused the frame to go away, return nil. This can only happen if a plugin in a subframe destroys
        its frame.

        (-[WebCoreFrameBridge aeDescByEvaluatingJavaScriptFromString:]):
        ASSERT that this is only called on the main frame.

WebKit:

        Reviewed by Darin.

        * WebView/WebView.mm:
        (-[WebView stringByEvaluatingJavaScriptFromString:]):
        ASSERT that the value returned isn't nil. It can't be nil when invoked on the main frame.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23950 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Written by Darin, reviewed by me

        - fixed <rdar://problem/5249730> REGRESSION (Leopard, around 9A464): Safari opens the same local
          documents more than once due to canonicalization change

        * loader/mac/LoaderNSURLExtras.m:
        (canonicalURL):
        pass URL through KURL so we get KURL's version of canonicalization as well as
        NSURLProtocol's version of canonicalization

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23949 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Darin.

        Seventh round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        - Change from double -> float caused some slight changes.

        * svg/W3C-SVG-1.1/animate-elem-33-t-expected.txt:
        * svg/W3C-SVG-1.1/animate-elem-37-t-expected.txt:
        * svg/W3C-SVG-1.1/animate-elem-39-t-expected.txt:
        * svg/W3C-SVG-1.1/animate-elem-80-t-expected.txt:
        * svg/W3C-SVG-1.1/animate-elem-81-t-expected.txt:
        * svg/W3C-SVG-1.1/animate-elem-82-t-expected.txt:
        * svg/W3C-SVG-1.1/animate-elem-83-t-expected.txt:
        * svg/W3C-SVG-1.1/coords-units-02-b-expected.txt:
        * svg/W3C-SVG-1.1/painting-render-01-b-expected.txt:
        * svg/W3C-SVG-1.1/script-handle-02-b-expected.txt:
        * svg/W3C-SVG-1.1/script-handle-03-b-expected.txt:
        * svg/W3C-SVG-1.1/script-handle-04-b-expected.txt:
        * svg/W3C-SVG-1.1/shapes-circle-02-t-expected.txt:
        * svg/W3C-SVG-1.1/shapes-ellipse-02-t-expected.txt:
        * svg/W3C-SVG-1.1/text-align-01-b-expected.txt:
        * svg/W3C-SVG-1.1/text-align-05-b-expected.txt:
        * svg/W3C-SVG-1.1/types-basicDOM-01-b-expected.txt:

WebCore:

        Reviewed by Darin.

        Seventh round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        - Adds more narrowPrecision functions to FloatPoint, FloatSize and FloatRect.
        - Changes over more implicit conversions to use new functions.

        * bindings/js/kjs_window.cpp:
        (KJS::WindowFunc::callAsFunction):
        * platform/graphics/FloatPoint.cpp:
        (WebCore::FloatPoint::matrixTransform):
        (WebCore::FloatPoint::narrowPrecision):
        * platform/graphics/FloatPoint.h:
        * platform/graphics/FloatRect.cpp:
        (WebCore::FloatRect::narrowPrecision):
        * platform/graphics/FloatRect.h:
        * platform/graphics/FloatSize.cpp:
        (WebCore::FloatSize::narrowPrecision):
        * platform/graphics/FloatSize.h:
        * platform/graphics/Path.cpp:
        (WebCore::Path::createRoundedRectangle):
        (WebCore::Path::createRectangle):
        (WebCore::Path::createEllipse):
        * platform/graphics/mac/ColorMac.mm:
        (WebCore::nsColor):
        * platform/graphics/mac/IconMac.mm:
        (WebCore::Icon::paint):
        * platform/graphics/svg/cg/CgSupport.cpp:
        (WebCore::applyStrokeStyleToContext):
        * platform/graphics/svg/cg/SVGPaintServerGradientCg.cpp:
        (WebCore::CGShadingRefForRadialGradient):
        (WebCore::SVGPaintServerGradient::updateQuartzGradientStopsCache):
        (WebCore::SVGPaintServerGradient::handleBoundingBoxModeAndGradientTransformation):
        * platform/graphics/svg/cg/SVGResourceMaskerCg.mm:
        (WebCore::applyLuminanceToAlphaFilter):
        (WebCore::applyExpandAlphatoGrayscaleFilter):
        (WebCore::transformImageIntoGrayscaleMask):
        * platform/mac/WebCoreTextRenderer.mm:
        (WebCoreDrawTextAtPoint):
        * rendering/RenderPath.cpp:
        (WebCore::RenderPath::mapAbsolutePointToLocal):
        * rendering/RenderSVGContainer.cpp:
        (WebCore::RenderSVGContainer::calcViewport):
        * rendering/RenderSVGImage.cpp:
        (WebCore::RenderSVGImage::nodeAtPoint):
        * rendering/RenderSVGText.cpp:
        (WebCore::RenderSVGText::relativeBBox):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23948 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Adam.

        <rdar://problem/5261523>
        Crash playing MP11 on cnbc.com

        The WMP Netscape plugin still creates a window even when windowless mode is turned on. Work around this by
        always removing the "windowlessvideo" parameter before passing it to the plugin.
        According to http://msdn2.microsoft.com/en-us/library/bb249373.aspx, "The windowlessVideo property is not
        supported for Netscape Navigator. Setting a value for this property in Navigator may yield unexpected results."

        * plugins/win/PluginViewWin.cpp:
        (WebCore::PluginViewWndProc):
        (WebCore::PluginViewWin::paint):
        (WebCore::PluginViewWin::invalidateRect):
        (WebCore::PluginViewWin::determineQuirks):
        (WebCore::PluginViewWin::setParameters):
        (WebCore::PluginViewWin::PluginViewWin):
        (WebCore::PluginViewWin::init):
        * plugins/win/PluginViewWin.h:
        (WebCore::):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23947 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2007-07-04  Mark Rowe  <mrowe@apple.com>

        Unreviewed 64-bit build fixes.

        * WebCoreSupport/WebInspectorClient.mm: Let the compiler know that WebFrameView is a subclass of NSView.
        * WebView/WebDocumentInternal.h: Remove our preprocessor macro once we're done with it.
        * WebView/WebHTMLView.mm:
        (-[WebHTMLView markAllMatchesForText:caseSensitive:limit:]): Fix argument types.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23946 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Patch from Qing Zhao to fix sunken state improperly set.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23945 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Maciej.

        Test for REGRESSION: misplaced selection ring remains after link is no longer focused.
        <rdar://problem/5205580>

        * fast/repaint/focus-layers-expected.checksum: Added.
        * fast/repaint/focus-layers-expected.png: Added.
        * fast/repaint/focus-layers-expected.txt: Added.
        * fast/repaint/focus-layers.html: Added.

WebCore:

        Reviewed by Maciej.

        REGRESSION: misplaced selection ring remains after link is no longer focused.
        <rdar://problem/5205580>

        Fix placement of selection ring when part of the ring is in a different layer.

        * rendering/RenderFlow.cpp:
        (WebCore::RenderFlow::addFocusRingRects): Get coordinates right when moving to a different layer.
        * rendering/RenderObject.cpp:
        (WebCore::RenderObject::setStyle):
        Don't reset state variables when not necessary.  When repaint() was called at the end of setSyle() it would
        get repaint rects wrong in some cases.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23944 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Maciej.

        Blind fix for CrashTracer: [USER] 25 crashes in Safari at com.apple.WebCore: WebCore::Frame::document const + 6
        <rdar://problem/5263939>

        No one has managed to reproduce this but it looks like null check is needed here. There are plausible ways
        m_frame could be null for a live FrameView object.

        * page/FrameView.cpp:
        (WebCore::FrameView::needsLayout):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23943 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Update the Windows Skipped list.

* win/Skipped:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23941 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Update results after r23938.

* fast/inline-block/14498-positionForCoordinates-expected.txt:
* fast/table/click-near-anonymous-table-expected.txt:
* svg/custom/pointer-events-image-expected.txt:
* svg/custom/pointer-events-path-expected.txt:
* svg/custom/text-hit-test-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23940 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin.

        Fix for <rdar://problem/5098378> single-pixel image optimization needs to be rewritten to not depend on floating point bitmap
        Tested by fast/replaced/image-solid-color-with-alpha.html

        Switch from float to byte based bitmap when pulling the color from 1x1 pixel images so
        that we can use the single pixel image optimisation in WebKit/win.

        * platform/graphics/cg/ImageCG.cpp:
        (WebCore::BitmapImage::checkForSolidColor):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23939 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix Bug 14498: RenderContainer::positionForCoordinates contains an order of operations error

WebCore:

        Fix Bug 14498: RenderContainer::positionForCoordinates contains an order of operations error
        http://bugs.webkit.org/show_bug.cgi?id=14498

        There were two bugs:
        1) xPos/yPos were flipped
        2) + binds tighter than ?:, so the order of operations was incorrect

        Reviewed by Darin.

        Test: fast/inline-block/14498-positionForCoordinates.html

        * rendering/RenderContainer.cpp:
        (WebCore::RenderContainer::positionForCoordinates):

LayoutTests:

        Test for Bug 14498: RenderContainer::positionForCoordinates contains an order of operations error
        http://bugs.webkit.org/show_bug.cgi?id=14498

        Reviewed by Darin.

        * fast/inline-block/14498-positionForCoordinates-expected.txt: Added.
        * fast/inline-block/14498-positionForCoordinates.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23938 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Kevin Decker and Tim Hatcher.

        - fix <rdar://problem/5307880> some classes need finalize methods
          because of non-trivial work done in dealloc methods

        * DefaultDelegates/WebScriptDebugServer.m:
        (-[WebScriptDebugServer dealloc]): Added a comment about how this probably won't
        work under GC.
        * WebInspector/WebNodeHighlight.m:
        (-[WebNodeHighlight dealloc]): Ditto.

        * WebView/WebDataSource.mm:
        (+[WebDataSourcePrivate initialize]): Added. Makes finalize run on main thread.
        (-[WebDataSourcePrivate finalize]): Added. Calls deref on the document loader.
        * WebView/WebHTMLView.mm:
        (+[WebHTMLViewPrivate initialize]): Added. Makes finalize run on main thread.
        (-[WebHTMLViewPrivate finalize]): Added. Calls deref on promisedDragTIFFDataSource.

        * WebKit.xcodeproj/project.pbxproj: Let Xcode be Xcode.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23937 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Kevin Decker and Tim Hatcher.

        - fix <rdar://problem/5307906> some classes need finalize methods because
          of non-trivial work done in dealloc methods

        * bindings/objc/DOMRGBColor.mm:
        (+[DOMRGBColor initialize]): Added. Tells finalizer to run on main thread,
        which is needed because we are sharing a CFDictionary here with no locking.

        * bridge/mac/WebCoreAXObject.mm:
        (+[WebCoreAXObject initialize]): Added. Tells finalizer to run on main thread.
        (-[WebCoreAXObject finalize]): Added. Calls detach.

        * WebCore.xcodeproj/project.pbxproj: Let Xcode do what it wants to do.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23936 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin.

        Sixth round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        Use new narrowPrecisionToFloat() and narrowPrecisionToCGFloat() where
        appropriate.

        * html/CanvasRenderingContext2D.cpp:
        (WebCore::CanvasRenderingContext2D::setShadow):
        (WebCore::CanvasRenderingContext2D::applyShadow):
        * html/CanvasStyle.cpp:
        (WebCore::CanvasStyle::applyStrokeColor):
        (WebCore::CanvasStyle::applyFillColor):
        * ksvg2/svg/SVGAngle.cpp:
        * ksvg2/svg/SVGAnimationElement.cpp:
        (WebCore::SVGAnimationElement::parseKeyNumbers):
        (WebCore::parseKeySplines):
        (WebCore::SVGAnimationElement::handleTimerEvent):
        (WebCore::SVGAnimationElement::updateAnimatedValueForElapsedSeconds):
        * ksvg2/svg/SVGPatternElement.cpp:
        (WebCore::SVGPatternElement::buildPattern):
        * ksvg2/svg/SVGRadialGradientElement.cpp:
        (WebCore::SVGRadialGradientElement::buildGradient):
        * platform/graphics/cg/AffineTransformCG.cpp:
        (WebCore::AffineTransform::AffineTransform):
        (WebCore::AffineTransform::setMatrix):
        (WebCore::AffineTransform::map):
        (WebCore::AffineTransform::setA):
        (WebCore::AffineTransform::setB):
        (WebCore::AffineTransform::setC):
        (WebCore::AffineTransform::setD):
        (WebCore::AffineTransform::setE):
        (WebCore::AffineTransform::setF):
        (WebCore::AffineTransform::scale):
        (WebCore::AffineTransform::rotate):
        (WebCore::AffineTransform::translate):
        (WebCore::AffineTransform::shear):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23935 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    * page/mac/WebCoreFrameBridge.mm:
        (-[WebCoreFrameBridge getAllResourceDatas:andResponses:]):
        added a comment

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23933 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Rubberstamped by Darin.

        Rename FloatConversionUtilities.h to FloatConversion.h.
        Fix Win32 build.

        * WebCore.vcproj/WebCore.vcproj:
        * WebCore.xcodeproj/project.pbxproj:
        * platform/DeprecatedString.cpp:
        * platform/FloatConversion.h: Copied from platform/FloatConversionUtilities.h.
        * platform/FloatConversionUtilities.h: Removed.
        * platform/StringImpl.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23932 268f45cc-cd09-0410-ab3c-d52691b4dbfc

JavaScriptCore:

        Reviewed by Kevin McCullough.

        Fourth round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        Add custom piDouble and piFloat constants to use instead of M_PI.

        * kjs/math_object.cpp:
        (MathObjectImp::getValueProperty):
        * wtf/MathExtras.h:
        (wtf_atan2):

WebCore:

        Reviewed by Darin.

        Fifth round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        Add toFloat() method to string classes and use it where appropriate.

        * WebCore.xcodeproj/project.pbxproj:
        * css/cssparser.cpp:
        (WebCore::CSSParser::lex):
        * ksvg2/svg/SVGAngle.cpp:
        (WebCore::SVGAngle::setValueAsString):
        * platform/AtomicString.h:
        (WebCore::AtomicString::toFloat):
        * platform/DeprecatedString.cpp:
        (WebCore::DeprecatedString::toFloat):
        * platform/DeprecatedString.h:
        * platform/FloatConversionUtilities.h: Added.
        (WebCore::narrowPrecisionToFloat):
        (WebCore::narrowPrecisionToCGFloat):
        * platform/PlatformString.h:
        * platform/String.cpp:
        (WebCore::String::toFloat):
        * platform/StringImpl.cpp:
        (WebCore::StringImpl::toFloat):
        * platform/StringImpl.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23931 268f45cc-cd09-0410-ab3c-d52691b4dbfc

JavaScriptCore:

        Reviewed by Kevin McCullough.

        Third round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        Add custom piDouble and piFloat constants to use instead of M_PI.

        * kjs/math_object.cpp:
        (MathObjectImp::getValueProperty):
        * wtf/MathExtras.h:
        (wtf_atan2):

WebCore:

        Reviewed by Kevin McCullough.

        Third round of fixes for implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        Use new piDouble and piFloat constants instead of M_PI.

        * html/CanvasRenderingContext2D.cpp:
        (WebCore::CanvasRenderingContext2D::rotate):
        * ksvg2/svg/SVGParserUtilities.cpp:
        (WebCore::SVGPathParser::calculateArc):
        * platform/graphics/Path.cpp:
        (WebCore::pathLengthApplierFunction):
        (WebCore::Path::createEllipse):
        * platform/graphics/cg/GraphicsContextCG.cpp:
        (WebCore::GraphicsContext::drawEllipse):
        (WebCore::GraphicsContext::strokeArc):
        * platform/graphics/cg/PDFDocumentImage.cpp:
        (WebCore::PDFDocumentImage::setCurrentPage):
        * platform/graphics/svg/filters/cg/SVGFEHelpersCg.h:
        * rendering/RenderPath.cpp:
        (WebCore::drawMarkerWithData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23930 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2007-07-02  Mitz Pettel  <mitz@webkit.org>

        Reviewed by Darin.

        - http://bugs.webkit.org/show_bug.cgi?id=14482
          Add information about the NeedsReduction and HasReduction keywords to quality/reduction.html

        * quality/reduction.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23927 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Justin.

        Fix for
          <rdar://problem/5290113> WebKit does not correctly handle replacement ranges from the IM in -[WebHTMLView insertText:]
          http://bugs.webkit.org/show_bug.cgi?id=13664

        We replicate the logic of -[WebHTMLView setMarkedText:selectedRange:] to handle the Input Method
        feeding us a replacement string through insertText: so we can handle IMs that use insertText to
        replace text.

        * WebView/WebHTMLView.mm:
        (-[WebHTMLView insertText:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23926 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Clarify/fix {Shadow,BorderImage}ParseContext's memory management

Prefast emitted warnings that drew my attention to
{Shadow,BorderImage}ParseContext::failed().  It turned out that these
methods were actually correct, but rather confusing. "failed" really
meant "abort and clean up" rather than "did you fail?", which was
unclear. However, once I figured that out, the "and clean up" part was
still a bit confusing, because all failed() did was to set a flag that
would later cause the ParseContext's members to be deleted in the
destructor. To clear this up, I've gotten rid of the failed() method
altogether. It always returned false, so I've replaced all calls to
it with the value false.

I also noticed that the lifetime management of the ParseContexts'
members was in all cases confusing, and in some cases wrong. The
m_border{Top,Right,Bottom,Left} members of BorderImageParseContext
were leaked whenever a border-image property was successfully parsed.
I fixed that by holding these members in OwnPtrs. The
CSSPrimitiveValue members of {Shadow,BorderImage}ParseContext, which
inherit from Shared, were being explicitly deleted, which is not a
safe way to manage the lifetime of objects that inherit from Shared.
To fix this, I put those members inside RefPtrs. These two changes
allowed me to remove the destructors entirely.

Reviewed by Darin.

All regression tests pass.

* css/cssparser.cpp:
(WebCore::ShadowParseContext::commitValue): Use .release() to avoid
ref-count churn.
(WebCore::ShadowParseContext::commitLength): Use a RefPtr for the new value to
avoid a leak.
(WebCore::CSSParser::parseShadow): Use 'false' instead of
'context.failed()', and use .release() to avoid ref-count churn.
(WebCore::BorderImageParseContext::commitWidth): Updated to use
OwnPtr.
(WebCore::CSSParser::parseBorderImage): Use 'false' instead of
'context.failed'.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23925 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Alexey.

        Fix for <rdar://problem/5306210> Some events are still passed to WebCore despite being handled by the IM
        http://bugs.webkit.org/show_bug.cgi?id=14457

        We have to assume that the IM will consume all events, so we remove the dependency
        on -[WebHTMLView hasMarkedText].

        * WebView/WebHTMLView.mm:
        (-[WebHTMLView _interceptEditingKeyEvent:shouldSaveCommand:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23924 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by John Sullivan.

        <rdar://problem/5297146>
        Crash in Mail at WebCore::Frame::setUserStyleSheetLocation (reproducible w/ particular message)

        * loader/Cache.cpp:
        (WebCore::Cache::requestResource):
        Only insert the resource in the LRU list if the cache is enabled.

        (WebCore::Cache::insertInLRUList):
        (WebCore::Cache::resourceAccessed):
        Assert that the resource is in the cache.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23923 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    * css/tokenizer.flex: Rolled out the fix for 14453.
        The layout test wasn't passing. We'll try again.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23922 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix a crash in ~PluginViewWin()

This crash was introduced in r23827.

Reviewed by Anders.

Fixes dom/html/level2/html/HTMLAnchorElement01.html

* plugins/win/PluginViewWin.cpp:
(WebCore::PluginViewWin::~PluginViewWin): Allow m_plugin to be null.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23920 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Darin.

        Regression test for:
        http://bugs.webkit.org/show_bug.cgi?id=14453
        REGRESSION: www.nzherald.co.nz almost all the formating is gone

        * css2.1/hex-color-regression-expected.checksum: Added.
        * css2.1/hex-color-regression-expected.png: Added.
        * css2.1/hex-color-regression-expected.txt: Added.
        * css2.1/hex-color-regression.html: Added.

WebCore:

        Reviewed by Darin.

        http://bugs.webkit.org/show_bug.cgi?id=14453
        REGRESSION: www.nzherald.co.nz almost all the formating is gone

        * css/tokenizer.flex: Allow hex colors with 2, 4, or 5 hex digits.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23919 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Fixed a no-SVG test failure by removing a node that was affecting the test results differently
        depending on whether SVG support was compiled in or not.

        * fast/xpath/namespace-vs-predicate.xhtml: Remove the r_00 node at the end of the test.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23918 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix some paths to make http tests work again on Windows

Rubberstamped by Sam.

* Scripts/run-webkit-tests:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23917 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Oliver.

        Third round of implicit 64-32 bit conversion errors fixes.
        <rdar://problem/5292262>

        getFloatValue() used to return a double and now returns a float for a small
        loss of precision.

        * fast/css/getFloatValueForUnit-expected.checksum:
        * fast/css/getFloatValueForUnit-expected.png:
        * fast/css/getFloatValueForUnit-expected.txt:
        * fast/css/getFloatValueForUnit.html:

WebCore:

        Reviewed by Oliver.

        Third round of implicit 64-32 bit conversion errors fixes.
        <rdar://problem/5292262>

        - Use new CSSPrimitiveValue::getDoubleValue() and CSSPrimitiveValue::getIntValue()
          methods instead of CSSPrimitiveValue::getFloatValue() where appropriate.

        * css/CSSPrimitiveValue.cpp:
        (WebCore::CSSPrimitiveValue::computeLengthInt):
        (WebCore::CSSPrimitiveValue::computeLengthIntForLength):
        (WebCore::CSSPrimitiveValue::computeLengthShort):
        (WebCore::CSSPrimitiveValue::computeLengthFloat):
        (WebCore::CSSPrimitiveValue::computeLengthDouble):
        (WebCore::CSSPrimitiveValue::getDoubleValue):
        * css/CSSPrimitiveValue.h:
        (WebCore::CSSPrimitiveValue::getDoubleValue):
        (WebCore::CSSPrimitiveValue::getFloatValue):
        (WebCore::CSSPrimitiveValue::getIntValue):
        * css/Counter.h:
        (WebCore::Counter::listStyleNumber):
        * css/MediaQueryEvaluator.cpp:
        (WebCore::parseAspectRatio):
        * css/cssparser.cpp:
        (WebCore::BorderImageParseContext::commitBorderImage):
        * css/cssstyleselector.cpp:
        (WebCore::convertToLength):
        (WebCore::applyCounterList):
        (WebCore::CSSStyleSelector::applyProperty):
        (WebCore::CSSStyleSelector::mapBackgroundSize):
        (WebCore::CSSStyleSelector::mapBackgroundXPosition):
        (WebCore::CSSStyleSelector::mapBackgroundYPosition):
        * html/HTMLTableElement.cpp:
        (WebCore::HTMLTableElement::parseMappedAttribute):
        * ksvg2/css/SVGCSSStyleSelector.cpp:
        (WebCore::CSSStyleSelector::applySVGProperty):
        * ksvg2/misc/KCanvasRenderingStyle.cpp:
        (WebCore::KSVGPainterFactory::cssPrimitiveToLength):
        (WebCore::KSVGPainterFactory::dashArrayFromRenderingStyle):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23916 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Fixed a no-SVG test failure by moving a test that requires SVG into the svg directory.

        * fast/dom/svg-element-attribute-js-null-expected.txt: Removed.
        * fast/dom/svg-element-attribute-js-null.xhtml: Removed.
        * svg/dom/svg-element-attribute-js-null-expected.txt: Copied from fast/dom/svg-element-attribute-js-null-expected.txt.
        * svg/dom/svg-element-attribute-js-null.xhtml: Copied from fast/dom/svg-element-attribute-js-null.xhtml.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23915 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Added a missing file for the test for <rdar://problem/5303567> REGRESSION: XMLHttpRequest.responseXML

        * fast/loader/resources/plist.app: Added. Just some random XML file I found somewhere; seems to make
        the test succeed instead of failing every time!

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23914 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Roll out r23911 as it introduced two layout test failures.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23913 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2007-07-01  Pamela Greene  <pamg.bugs@gmail.com>

        Reviewed by Darin.

        Fix for <http://bugs.webkit.org/show_bug.cgi?id=14320>
        Bug 14320: <canvas> gradient stops at identical offsets are applied in the wrong order
        <rdar://problem/5290084>

        Use std::stable_sort() instead of qsort() to sort gradient stops.
        Adjust the comparison function accordingly.

        * html/CanvasGradient.cpp:
        (WebCore::compareStops):
        (WebCore::CanvasGradient::findStop):

2007-07-01  Pamela Greene  <pamg.bugs@gmail.com>

        Reviewed by Darin.

        Tests for <http://bugs.webkit.org/show_bug.cgi?id=14320>
        Bug 14320: <canvas> gradient stops at identical offsets are applied in the wrong order
        <rdar://problem/5290084>

        Add tests for various combinations of stops, including multiple stops at
        the same offsets, sorted and unsorted.  Update results accordingly.

        * fast/canvas/fillrect_gradient-expected.checksum:
        * fast/canvas/fillrect_gradient-expected.png:
        * fast/canvas/fillrect_gradient-expected.txt:
        * fast/canvas/fillrect_gradient.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23912 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2007-07-01  MorganL  <morganl.webkit@yahoo.com>

        Reviewed by Darin.

        Fix for http://bugs.webkit.org/show_bug.cgi?id=14448
        HTMLObjectElement::data should return an absolute URL

        * html/HTMLObjectElement.cpp:
        (WebCore::HTMLObjectElement::isImageType):
        (WebCore::HTMLObjectElement::data):

2007-07-01  MorganL  <morganl.webkit@yahoo.com>

        Reviewed by Darin.

        Tests for http://bugs.webkit.org/show_bug.cgi?id=14448
        HTMLObjectElement::data should return an absolute URL

        * fast/dom/object-data-absurl-expected.txt: Added.
        * fast/dom/object-data-absurl.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23911 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2007-07-01  Holger Hans Peter Freyther  <zecke@selfish.org>

        Reviewed by Darin.

        m_usesPageCache was not initialized. As the b/f cache is only
        implemented in the Mac port and WebView.mm is taking care of calling
        Settings::setUsesPageCache I decided to make it false by default.

        * page/Settings.cpp:
        (WebCore::Settings::Settings): m_usesPageCache was not initialized

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23910 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2007-07-01  Holger Hans Peter Freyther  <zecke@selfish.org>

        Reviewed by Darin.

        Inherited- and NonInheritedFlags is a bit field. But for comparsion
        the other part of the union _iflags/_niflags is used for comparions.
        This will compare 5 uninitialized bits of _iflags and 22 of _niflags.
        Make sure the unused bits have a specified value.

        * ksvg2/css/SVGRenderStyle.h:
        (WebCore::SVGRenderStyle::setBitDefaults): Make sure the unused bits
        have a value as well.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23909 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2007-07-01  Mitz Pettel  <mitz@webkit.org>

        Reviewed by Darin.

        - http://bugs.webkit.org/show_bug.cgi?id=14481
        Broken links on coding/lgpl-license.html

        * coding/lgpl-license.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23908 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Compile WebCore with /W4 on Windows

Some more warnings were disabled globally, including assignment within
a conditional, unused variable, and various warnings related to
inability to instantiate opaque structs.

In addition, the "unreachable code" warning was turned off for
CSSGrammar.cpp and XPathGrammar.cpp.

Reviewed by Oliver.

* WebCore.vcproj/WebCore.vcproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23907 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Give CString an equality operator to fix a bug in HTMLFormElement::formData

WebCore:

        Give CString an equality operator to fix a bug in HTMLFormElement::formData

        The bug was spotted by MSVC /W4. The problem was that we were comparing
        a CString to a char* using ==, but CString had no equality operator.
        The result was that the CString was being cast to a const char* and a
        pointer comparison was being done, which would (essentially) always
        return false.

        There are two parts to the fix: get rid of CString's const char*
        casting operator, and add an equality operator. Previous uses of the
        casting operator have been changed to use CString::data().

        Test: http/misc/isindex-formdata.html

        Reviewed by Oliver.

        * dom/XMLTokenizer.cpp:
        (WebCore::parseXMLDocumentFragment):
        * html/HTMLDocument.cpp:
        (WebCore::HTMLDocument::determineParseMode):
        * html/HTMLFormElement.cpp:
        (WebCore::HTMLFormElement::formData):
        * loader/TextResourceDecoder.cpp:
        (WebCore::TextResourceDecoder::checkForCSSCharset):
        * platform/CString.cpp:
        (WebCore::operator==):
        * platform/CString.h:
        * platform/KURL.cpp:
        (WebCore::encodeRelativeString):
        * platform/StringImpl.cpp:
        (WebCore::StringImpl::toDouble):
        * platform/network/cf/FormDataStreamCFNet.cpp:
        (WebCore::setHTTPBody):
        * platform/network/mac/FormDataStreamMac.mm:
        (WebCore::setHTTPBody):
        * platform/win/ClipboardUtilitiesWin.cpp:
        (WebCore::markupToCF_HTML):
        * plugins/win/PluginPackageWin.cpp:
        (WebCore::splitString):
        * plugins/win/PluginStreamWin.cpp:
        (WebCore::PluginStreamWin::startStream):
        (WebCore::PluginStreamWin::destroyStream):
        (WebCore::PluginStreamWin::sendJavaScriptStream):
        (WebCore::PluginStreamWin::didFinishLoading):
        * plugins/win/PluginViewWin.cpp:
        (WebCore::PluginViewWin::start):
        (WebCore::createUTF8String):
        (WebCore::PluginViewWin::userAgent):
        * xml/XSLStyleSheet.cpp:
        (WebCore::XSLStyleSheet::locateStylesheetSubResource):
        * xml/XSLTProcessor.cpp:
        (WebCore::xsltParamArrayFromParameterMap):

LayoutTests:

        New test that makes sure we handle putting isindex into form data correctly.

        Reviewed by Oliver.

        * http/tests/misc/isindex-formdata-expected.txt: Added.
        * http/tests/misc/isindex-formdata.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23906 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Initialize two variables that MSVC /W4 isn't smart enough to realize always get initialized

Reviewed by John.

* editing/Editor.cpp:
(WebCore::Editor::advanceToNextMisspelling):
* html/HTMLParser.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23905 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix some possibly unitialized variables found by MSVC /W4

Reviewed by John.

* rendering/bidi.cpp:
(WebCore::RenderBlock::layoutInlineChildren):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23904 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove bogus cast found by MSVC /W4

Reviewed by John and Oliver.

* ksvg2/css/SVGCSSParser.cpp:
(WebCore::CSSParser::parseSVGValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23903 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove unused local functions uncovered by MSVC /W4

Reviewed by John and Oliver.

* platform/win/SearchPopupMenuWin.cpp:
* rendering/RenderText.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23902 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Don't take the address of temporary values to appease MSVC /W4

Reviewed by Oliver.

No regression test possible.

* platform/win/PopupMenuWin.cpp:
(WebCore::PopupMenu::invalidateItem):
(WebCore::PopupMenu::valueChanged):
(WebCore::PopupWndProc):
* platform/win/ScrollViewWin.cpp:
(WebCore::ScrollView::ScrollViewPrivate::scrollBackingStore):
* platform/win/UniscribeController.cpp:
(WebCore::UniscribeController::shapeAndPlaceItem):
* platform/win/WidgetWin.cpp:
(WebCore::Widget::invalidateRect):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23901 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Make an int -> UINT conversion explicit that MSVC /W4 complains about

Reviewed by John and Oliver.

No regression test possible.

* platform/win/SoundWin.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23900 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Remove unreachable code uncovered by MSVC /W4

Reviewed by Oliver.

No regression test possible.

* html/CanvasRenderingContext2D.cpp:
(WebCore::CanvasRenderingContext2D::createPattern):
* html/HTMLAnchorElement.cpp:
(WebCore::HTMLAnchorElement::isLiveLink):
* html/HTMLElement.cpp:
(WebCore::HTMLElement::contentEditable):
* platform/DeprecatedString.cpp:
(WebCore::allocateHandle):
(WebCore::freeHandle):
* plugins/win/npapi.cpp:
(pluginViewForInstance):
* rendering/RenderImage.cpp:
(WebCore::RenderImage::isWidthSpecified):
(WebCore::RenderImage::isHeightSpecified):
* rendering/RenderTheme.cpp:
(WebCore::RenderTheme::isControlStyled):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23899 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Build fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23898 268f45cc-cd09-0410-ab3c-d52691b4dbfc

WebCore:

        Reviewed by Oliver Hunt.

        - WebCore support for accessing the set of rectangles that encompass the selected text

        * page/Frame.h:
        * page/Frame.cpp:
        (WebCore::Frame::selectionTextRects):
        new function, fills in a passed Vector with the line box rects that represent the selected range

        * rendering/RenderText.cpp:
        (WebCore::RenderText::addLineBoxRects):
        fixed off-by-one bug involving use of the confusing function InlineTextBox::end()

        * WebCore.exp:
        exported symbol for this new function

WebKit:

        Reviewed by Oliver Hunt.

        - WebKit support for accessing the set of rectangles that encompass the selected text

        * WebView/WebDocumentPrivate.h:
        added -selectionTextRects to WebDocumentSelection protocol; tweaked comments

        * WebView/WebHTMLView.mm:
        (-[WebHTMLView _selectionDraggingRect]):
        use selectionRect instead of selectionImageRect since they're the same and maybe we can get
        rid of selectionImageRect someday
        (-[WebHTMLView selectionTextRects]):
        added implementation of new protocol method, which calls through to WebCore

        * WebView/WebPDFView.mm:
        (-[WebPDFView selectionTextRects]):
        added simple implementation of new protocol method, which just returns the single selection rect.
        PDFKit doesn't support obtaining multiple rects to describe a multi-line selection.
        (-[WebPDFView selectionImageForcingWhiteText:]):
        use selectionRect instead of selectionImageRect since they're the same and maybe we can get
        rid of selectionImageRect someday

        * Misc/WebSearchableTextView.m:
        (-[WebSearchableTextView selectionTextRects]):
        added no-op implementation of new protocol method to this obsolete class

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23897 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Build fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23895 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Harrison.

        <rdar://problem/5299513> REGRESSION: In Mail, a crash occurs when attempting to undo a newly created To Do

        Backing out a change made in r23787.

        * editing/deleting/4845371-expected.txt:
        * editing/deleting/4845371.html:
        * editing/deleting/5126166-expected.txt:
        * editing/deleting/5290534.html: Removed.
        * editing/deleting/5290534.html-disabled: Added.

WebCore:

        Reviewed by Harrison.

        <rdar://problem/5299513> REGRESSION: In Mail, a crash occurs when attempting to undo a newly created To Do

        Rolled out a change made in r23787.

        * editing/DeleteSelectionCommand.cpp:
        (WebCore::DeleteSelectionCommand::handleGeneralDelete):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23894 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Darin Adler, Anders Carlsson.

        Test for <rdar://problem/5303567> REGRESSION: XMLHttpRequest.responseXML
        returns NULL if response MIME type is not XML -- breaks Wikipedia widget

        * fast/loader/xmlhttprequest-bad-mimetype-expected.txt: Added.
        * fast/loader/xmlhttprequest-bad-mimetype.html: Added.

WebCore:

        Reviewed by Darin Adler, Anders Carlsson.

        Fixed <rdar://problem/5303567> REGRESSION: XMLHttpRequest.responseXML
        returns NULL if response MIME type is not XML -- breaks Wikipedia widget

        The spec says that responseXML must return null for any HTTP response
        with a non-XML MIME type. Technically, that leaves non-HTTP responses
        free to do whatever they want.

        * xml/xmlhttprequest.cpp:
        (WebCore::XMLHttpRequest::getResponseXML): Only return null for non-XML
        MIME types if the response was HTTP.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23893 268f45cc-cd09-0410-ab3c-d52691b4dbfc

JavaScriptCore:

        Reviewed by Darin.

        Second pass at fixing implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        - Add a toFloat() method to JSValue for float conversion.

        * JavaScriptCore.exp:
        * kjs/value.cpp:
        (KJS::JSValue::toFloat):
        * kjs/value.h:

WebCore:

        Reviewed by Darin.

        Second pass at fixing implicit 64-32 bit conversion errors.
        <rdar://problem/5292262>

        - Use new JSValue::toFloat() method instead of toNumber() where
          appropriate.

        * bindings/js/JSCanvasRenderingContext2DCustom.cpp:
        (WebCore::JSCanvasRenderingContext2D::setFillColor):
        (WebCore::JSCanvasRenderingContext2D::setStrokeColor):
        (WebCore::JSCanvasRenderingContext2D::strokeRect):
        (WebCore::JSCanvasRenderingContext2D::drawImage):
        (WebCore::JSCanvasRenderingContext2D::drawImageFromRect):
        (WebCore::JSCanvasRenderingContext2D::setShadow):
        * bindings/js/JSHTMLSelectElementCustom.cpp:
        (WebCore::JSHTMLSelectElement::remove):
        * bindings/js/JSSVGMatrixCustom.cpp:
        (WebCore::JSSVGMatrix::rotateFromVector):
        * bindings/js/kjs_events.cpp:
        (KJS::JSClipboardPrototypeFunction::callAsFunction):
        * bindings/scripts/CodeGeneratorJS.pm:
        * ksvg2/svg/SVGScriptElement.cpp:
        (WebCore::SVGScriptElement::executeScript):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23892 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Rolling out tests for <rdar://problem/5301322>.

        * fast/forms/input-readonly-select-all-expected.checksum: Removed.
        * fast/forms/input-readonly-select-all-expected.png: Removed.
        * fast/forms/input-readonly-select-all-expected.txt: Removed.
        * fast/forms/input-readonly-select-all.html: Removed..
        * fast/forms/textarea-readonly-select-all-expected.checksum: Removed.
        * fast/forms/textarea-readonly-select-all-expected.png: Removed.
        * fast/forms/textarea-readonly-select-all-expected.txt: Removed.
        * fast/forms/textarea-readonly-select-all.html: Removed.

WebCore:

        Rolling out fix for <rdar://problem/5301322> since it broke layout tests.

        * html/HTMLInputElement.h:
        * html/HTMLTextAreaElement.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23891 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Oliver Hunt.

        Layout tests for <rdar://problem/4962298> REGRESSION: Synchronous XHR for
        missing local file throws exception -- breaks Wikipedia widget

        * fast/loader/xmlhttprequest-missing-file-exception-expected.txt: Added.
        * fast/loader/xmlhttprequest-missing-file-exception.html: Added.

WebCore:

        Reviewed by Oliver Hunt.

        Fixed <rdar://problem/4962298> REGRESSION: Synchronous XHR for missing
        local file throws exception -- breaks Wikipedia widget

        * WebCore.xcodeproj/project.pbxproj:
        * xml/xmlhttprequest.cpp:
        (WebCore::XMLHttpRequest::send): Don't throw exceptions for local files.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23889 268f45cc-cd09-0410-ab3c-d52691b4dbfc

2007-06-29  Ada Chan  <adachan@apple.com>

        Reviewed by Adam.

        Fixed a previous fix to a signed/unsigned mismatch.

        * plugins/win/PluginPackageWin.cpp:
        (WebCore::splitString):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23883 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Ada and Kevin McCullough.

        Test for <rdar://problem/5301322> REGRESSION: Select All selects the whole page when used in readonly textareas

        * fast/forms/input-readonly-select-all-expected.checksum: Added.
        * fast/forms/input-readonly-select-all-expected.png: Added.
        * fast/forms/input-readonly-select-all-expected.txt: Added.
        * fast/forms/input-readonly-select-all.html: Added.
        * fast/forms/textarea-readonly-select-all-expected.checksum: Added.
        * fast/forms/textarea-readonly-select-all-expected.png: Added.
        * fast/forms/textarea-readonly-select-all-expected.txt: Added.
        * fast/forms/textarea-readonly-select-all.html: Added.

WebCore:

        Reviewed by Ada and Kevin McCullough.

        Fix for <rdar://problem/5301322> REGRESSION: Select All selects the whole page when used in readonly textareas

        Implemented canSelectAll and selectAll for text fields and textareas, since they're special cases even when they're not editable.

        Tests:
        fast/forms/input-readonly-select-all.html
        fast/forms/textarea-readonly-select-all.html

        * html/HTMLInputElement.h:
        (WebCore::HTMLInputElement::canSelectAll):
        (WebCore::HTMLInputElement::selectAll):
        * html/HTMLTextAreaElement.h:
        (WebCore::HTMLTextAreaElement::canSelectAll):
        (WebCore::HTMLTextAreaElement::selectAll):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23881 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Build fix

This is what happens when you test your patch on two OSs and then
commit from the wrong one.

* rendering/RenderTextControl.cpp:
(WebCore::RenderTextControl::valueChanged):
(WebCore::RenderTextControl::itemText):
(WebCore::RenderTextControl::itemIsSeparator):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23880 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Start compiling WebCore with /W3 under MSVC

One warning was globally disabled: implicit conversion of int to bool.

Reviewed by Darin.

* WebCore.vcproj/WebCore.vcproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23879 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Turn DeprecatedStringData's copy constructor into a static method to appease MSVC /W3

DeprecatedStringData now has no public copy constructor. The single
call to the old one has been replaced with a call to the new
DeprecatedStringData::createAndAdopt static method.

Reviewed by Darin.

All regression tests pass.

* platform/DeprecatedString.cpp:
(WebCore::DeprecatedStringData::createAndAdopt):
(WebCore::DeprecatedStringData::adopt):
(WebCore::DeprecatedString::detachIfInternal):
* platform/DeprecatedString.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23878 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix more signed/unsigned mismatches uncovered by MSVC /W3

Reviewed by John.

No regression tests possible.

* platform/network/cf/ResourceResponseCFNet.cpp:
(WebCore::filenameHasSaneExtension):
* platform/win/FontDataWin.cpp:
(WebCore::FontData::containsCharacters):
* platform/win/PlatformScrollBarSafari.cpp:
(WebCore::):
* platform/win/UniscribeController.cpp:
(WebCore::UniscribeController::advance):
(WebCore::UniscribeController::shapeAndPlaceItem):
* platform/win/WebCoreTextRenderer.cpp:
(WebCore::doDrawTextAtPoint):
* plugins/win/PluginDatabaseWin.cpp:
(WebCore::parseVersionString):
* plugins/win/PluginPackageWin.cpp:
(WebCore::splitString):
* plugins/win/PluginViewWin.cpp:
(WebCore::freeStringArray):
(WebCore::PluginViewWin::invalidateTimerFired):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23877 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Change PopupMenuClient::listSize to return an int to appease warnings from MSVC /W3

Reviewed by John.

No regression test possible.

* html/HTMLSelectElement.cpp:
(WebCore::HTMLSelectElement::menuListDefaultEventHandler):
* platform/PopupMenuClient.h:
* platform/win/PopupMenuWin.cpp:
(WebCore::PopupMenu::calculatePositionAndSize):
(WebCore::PopupMenu::focusFirst):
(WebCore::PopupMenu::focusLast):
(WebCore::PopupMenu::down):
(WebCore::PopupMenu::up):
(WebCore::PopupMenu::paint):
* rendering/RenderMenuList.cpp:
* rendering/RenderMenuList.h:
* rendering/RenderTextControl.cpp:
* rendering/RenderTextControl.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23876 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Fix two typos spotted by MSVC /W3

Both typos were a semicolon following the condition of an if. This
would cause the intended body of the if to always execute.

Unfortunately neither of these fixes are testable.
BackForwardList::goForward is only called from WebBackForwardList, and
the createDragImageFromImage bug would only manifest itself when
CreateCompatibleDC failed.

Reviewed by Mark.

* history/BackForwardList.cpp:
(WebCore::BackForwardList::goForward):
* platform/win/DragImageWin.cpp:
(WebCore::createDragImageFromImage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23875 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Patch from Adam Treat to improve scroll wheel support for Qt.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23874 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Maciej.

        Added WKQTMovieDataRate and WKQTMovieMaxTimeLoaded

        * WebKitSystemInterface.h:
        * libWebKitSystemInterface.a:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23869 268f45cc-cd09-0410-ab3c-d52691b4dbfc

Get WebCore compiling with /W2 on Windows

Reviewed by Steve.

No regression tests affected.

* WebCore.vcproj/WebCore.vcproj: Turn the warning level up to 2, and
disable two warnings: implicit int <=> float conversion, and class vs.
struct inconsistency.
* dom/CharacterData.cpp:
(WebCore::CharacterData::deleteData): Cast to int before negating.
* loader/Cache.cpp:
(WebCore::Cache::remove): Ditto.
* platform/graphics/Path.cpp:
(WebCore::pathLengthApplierFunction): Consistently use floats.
* platform/graphics/cg/GraphicsContextCG.cpp:
(WebCore::GraphicsContext::drawEllipse): Ditto.
* platform/win/GraphicsContextWin.cpp:
(WebCore::GraphicsContext::drawLineForMisspellingOrBadGrammar): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23868 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Maciej Stachowiak.

        Fixed <rdar://problem/5299005> DOM timers fire without a local autorelease
        pool => pathologically high high water mark

        I couldn't measure any performance problems with iBench HTML or PLT.

        * WebCore.xcodeproj/project.pbxproj: Changed .cpp file to be ObjC++.
        * platform/mac/SharedTimerMac.cpp:
        (WebCore::timerFired): Added autorelease pool.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23867 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Reviewed by Beth.

        - test for http://bugs.webkit.org/show_bug.cgi?id=14118
          ASSERTION FAILED: !needsLayout() seen again

        * fast/dynamic/recursive-layout-expected.txt: Added.
        * fast/dynamic/recursive-layout.html: Added.

WebCore:

        Reviewed by Beth.

        - fix http://bugs.webkit.org/show_bug.cgi?id=14118
          ASSERTION FAILED: !needsLayout() seen again

        Test: fast/dynamic/recursive-layout.html

        * page/FrameView.cpp:
        (WebCore::FrameView::layout): Removed the midLayout guards around
        the calls to invalidateSelection() and updateWidgetPositions(). Those
        calls may dirty the layout, but in that case they also update it, so
        they should be allowed to do it. Added an ASSERT to make sure that when
        they return, the root is not marked for layout.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23866 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Adam,

Allow tests to use pre-built roots instead of building the tests.

        * Scripts/run-javascriptcore-tests: added --root option
        * Scripts/run-webkit-tests: added --root option
        * Scripts/webkitdirs.pm: added support for using a root

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23865 268f45cc-cd09-0410-ab3c-d52691b4dbfc

    Reviewed by Darin.

        <rdar://problem/5287569> WEBVIEW: Creating a webview in IB 3.0 (v2.0) NIB later crashes Interface Builder 2.5.4 on Tiger

        Prevent encoding any of the WebView subviews. The subviews are ignored by -[WebView initWithCoder:] and will be recreated.
        The Tiger 2.0 code crashed when the WebView released the subviews in initWithCoder:, so now there are no subviews to release.
        This never happened before because the Tiger 2.0 code and IB wouldn't encode a WebView that has a WebHTMLView.

        * WebView/WebView.mm:
        (-[WebView encodeWithCoder:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23864 268f45cc-cd09-0410-ab3c-d52691b4dbfc

LayoutTests:

        Rubberstamped by Adam Roben.

        Roll out r23861.  It caused the buildbot to fail on intel.

        * http/tests/navigation/goBack-blank-tab-page-expected.txt: Removed.
        * http/tests/navigation/goBack-blank-tab-page.html: Removed.
        * http/tests/navigation/resources/before-go-back.html: Removed.
        * http/tests/navigation/resources/will-go-back.html: Removed.

WebCore:

        Rubberstamped by Adam Roben.

        Roll out r23861.  It caused the buildbot to fail on intel.

        * page/Page.cpp:
        (WebCore::Page::goToItem):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@23863 268f45cc-cd09-0410-ab3c-d52691b4dbfc