15 years ago Reviewed by George Staikos.
weinig [Fri, 18 May 2007 06:36:07 +0000 (06:36 +0000)]
    Reviewed by George Staikos.

        Fix Qt build after adding JSHTMLAnchorElementCustom.cpp.

        * WebCore.pro:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21553 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago- patch from Adam Treat to implement frameLoadCompleted()
staikos [Fri, 18 May 2007 06:32:08 +0000 (06:32 +0000)]
- patch from Adam Treat to implement frameLoadCompleted()
- build fix from George Staikos

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21552 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFix preprocessor logic for CG/QT
staikos [Fri, 18 May 2007 06:08:11 +0000 (06:08 +0000)]
Fix preprocessor logic for CG/QT

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21551 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoImplement forceLayoutForNonHTML
staikos [Fri, 18 May 2007 05:32:12 +0000 (05:32 +0000)]
Implement forceLayoutForNonHTML

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21550 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoImplement canShowMIMEType
staikos [Fri, 18 May 2007 05:28:31 +0000 (05:28 +0000)]
Implement canShowMIMEType

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21549 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Add a line gutter to view source mode. The gutter uses counters, and this...
hyatt [Fri, 18 May 2007 05:07:23 +0000 (05:07 +0000)]
    Add a line gutter to view source mode.  The gutter uses counters, and this exposed the fact
        that pseudoStyleForElement was not walking all UA rules properly.

        Reviewed by aroben

        * css/cssstyleselector.cpp:
        * css/view-source.css:
        * html/HTMLViewSourceDocument.cpp:
        * html/HTMLViewSourceDocument.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21548 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
justing [Fri, 18 May 2007 04:17:59 +0000 (04:17 +0000)]

        Reviewed by ggaren

        <rdar://problem/5209984> REGRESSION: Crash on click after clearing Facebook status

        * editing/selection/5209984-expected.txt: Added.
        * editing/selection/5209984.html: Added.


        Reviewed by ggaren

        <rdar://problem/5209984> REGRESSION: Crash on click after clearing Facebook status

        comparePositions() expects non-null positions.  It's
        given a null position by contains() if the current
        selection is a range selection but contains only invisible
        content *and* visible positions cannot be created from its
        end points.  This can happen when there is a range selection
        inside shadow content (like that inside text fields) that
        is made invisible.

        * editing/SelectionController.cpp:
        (WebCore::SelectionController::contains): Check for
        invisible end points.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21547 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
weinig [Fri, 18 May 2007 02:50:44 +0000 (02:50 +0000)]

        Reviewed by Adam.

        Tests for <rdar://problem/5201423> REGRESSION: Links in
        sidebar don't work on Honda website

        * fast/dom/anchor-toString-expected.txt: Added.
        * fast/dom/anchor-toString.html: Added.
        * fast/forms/focus2.html:


        Reviewed by Adam.

        Fix for <rdar://problem/5201423> REGRESSION: Links in
        sidebar don't work on Honda website

        Anchor elements need to have the same behavior for implicit and
        explicit toString() calls.

        Test: fast/dom/anchor-toString.html

        * WebCore.xcodeproj/project.pbxproj:
        * bindings/js/JSHTMLAnchorElementCustom.cpp: Added.
        * html/HTMLAnchorElement.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21546 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Build fix.
ggaren [Thu, 17 May 2007 23:38:51 +0000 (23:38 +0000)]
    Build fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21545 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Hyatt.
adele [Thu, 17 May 2007 23:15:37 +0000 (23:15 +0000)]
    Reviewed by Hyatt.

        Update RenderMenuList PopupClient methods to consider style of inner block.

        * rendering/RenderMenuList.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21544 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Not reviewed, emergency test fix.
mjs [Thu, 17 May 2007 22:57:49 +0000 (22:57 +0000)]
    Not reviewed, emergency test fix.

        - avoid crash on layout tests.

        Patched up my last change so that  widgets don't crash.

        * loader/FrameLoader.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21543 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
antti [Thu, 17 May 2007 22:25:01 +0000 (22:25 +0000)]

        Reviewed by Darin.

        Test for http://bugs.webkit.org/show_bug.cgi?id=13727
        REGRESSION: Reproducible hang setting the selection's extent offset to -1

        * LayoutTests/editing/selection/selection-invalid-offset-expected.txt: Added.
        * LayoutTests/editing/selection/selection-invalid-offset.html: Added.


        Reviewed by Darin.

        Fix for http://bugs.webkit.org/show_bug.cgi?id=13727
        REGRESSION: Reproducible hang setting the selection's extent offset to -1

        Rise INDEX_SIZE_ERR if used selection index is negative. This matches HTML5 for
        getRangeAt() and collapse() which are part of the spec (FF throws NS_ERROR_FAILURE).

        Not throwing exceptions yet for too high offsets (except getRangeAt()) as it is
        unclear if WebCore's definition of offsets is the same as other browsers.

        * bindings/js/kjs_window.cpp:
        * WebCore/editing/SelectionController.cpp:
        * WebCore/editing/SelectionController.h:
        * WebCore/editing/VisiblePosition.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21542 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
adele [Thu, 17 May 2007 21:42:21 +0000 (21:42 +0000)]

        Reviewed by Mitz.

        Test for <rdar://problem/5096035> REGRESSION: rtl text should be left-aligned in <select>s

        * fast/text/international/bidi-menulist-expected.checksum: Added.
        * fast/text/international/bidi-menulist-expected.png: Added.
        * fast/text/international/bidi-menulist-expected.txt: Added.
        * fast/text/international/bidi-menulist.html: Added.


        Reviewed by Mitz.

        Fix for <rdar://problem/5096035> REGRESSION: rtl text should be left-aligned in <select>s

        Test: fast/text/international/bidi-menulist.html

        * rendering/RenderMenuList.cpp:
        (WebCore::RenderMenuList::adjustInnerStyle): Set the inner block style to direction: ltr, and text-align: left.  This means that
         we will ignore those settings in the button to match the way the text is displayed in the popup menu.
        (WebCore::RenderMenuList::createInnerBlock): Call adjustInnerStyle.
        (WebCore::RenderMenuList::setStyle): ditto.
        * rendering/RenderMenuList.h: Added adjustInnerStyle.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21541 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoI accidently mucked around with these tests after generating results.
justing [Thu, 17 May 2007 21:16:48 +0000 (21:16 +0000)]
I accidently mucked around with these tests after generating results.
Correcting them.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21540 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
justing [Thu, 17 May 2007 21:06:12 +0000 (21:06 +0000)]

        Reviewed by harrison

        * editing/execCommand/4924441-expected.checksum: Added.
        * editing/execCommand/4924441-expected.png: Added.
        * editing/execCommand/4924441-expected.txt: Added.
        * editing/execCommand/4924441.html: Added.
        * editing/execCommand/5207369-expected.checksum: Added.
        * editing/execCommand/5207369-expected.png: Added.
        * editing/execCommand/5207369-expected.txt: Added.
        * editing/execCommand/5207369.html: Added.
        * editing/execCommand/5210032-expected.checksum: Added.
        * editing/execCommand/5210032-expected.png: Added.
        * editing/execCommand/5210032-expected.txt: Added.
        * editing/execCommand/5210032.html: Added.


        Reviewed by harrison

        <rdar://problem/4924441> List becomes nested when changing its type
        <rdar://problem/5210032> Content before a list is placed at the end of the list when its made into a list item
        <rdar://problem/5207369> Crash when removing ordering from a nested list

        * editing/InsertListCommand.cpp:
        The outermostEnclosingList of the position before the start isn't
        the previousList if it contains the start.
        Ditto for the position after the end and the nextList.
        If we're putting the newly created list item into the next list,
        insert it at the start of that list, not the end (don't append).
        Avoid nesting newly created lists inside list items.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21539 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago<rdar://problem/4775672> AXTextMarkerRangeForUnorderedTextMarkers returns incorrect...
pyeh [Thu, 17 May 2007 18:35:18 +0000 (18:35 +0000)]
<rdar://problem/4775672> AXTextMarkerRangeForUnorderedTextMarkers returns incorrect order

        * bridge/mac/WebCoreAXObject.mm:
        (-[WebCoreAXObject doAXTextMarkerRangeForUnorderedTextMarkers:]):
        A selection's m_start is not guranteed to equal m_base since m_start can actually be m_base's downstream().
        A more accurate way to determine the order from selection is to ask directly for isBaseFirst().
        In addition, if the two visible positions given are the same position but different affinity, we need to
        handle it here rather than depend on the selection. This is because selection does not retain the
        affinity of both selection start/end visible positions as part of the selection validation.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21538 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
beidson [Thu, 17 May 2007 17:52:44 +0000 (17:52 +0000)]

        Reviewed by Brady

        - pixel test for http://bugs.webkit.org/show_bug.cgi?id=13759
          REGRESSION (r20182-r20184): Incorrect rendering of multipart images

        Created http/tests/multipart and moved the existing multipart test there.

        * http/tests/misc/multipart-expected.txt: Removed.
        * http/tests/misc/multipart.html: Removed.
        * http/tests/misc/resources/1x1-blue.png: Removed.
        * http/tests/misc/resources/multipart.php: Removed.
        * http/tests/multipart/invalid-image-data-expected.checksum: Added.
        * http/tests/multipart/invalid-image-data-expected.png: Added.
        * http/tests/multipart/invalid-image-data-expected.txt: Added.
        * http/tests/multipart/invalid-image-data.html: Added.
        * http/tests/multipart/resources/2x2-green.png: Added. Shared by both
        multipart tests.
        * http/tests/multipart/resources/abe.png: Added.
        * http/tests/multipart/resources/multipart.php: Added. Shared by both
        multipart tests.
        * http/tests/multipart/stop-crash.html: Renamed from misc/multipart.html.


        Reviewed by Brady

        - fix http://bugs.webkit.org/show_bug.cgi?id=13759
          REGRESSION (r20182-r20184): Incorrect rendering of multipart images

        Test: http/tests/multipart/invalid-image-data.html

        * loader/loader.cpp:
        (WebCore::Loader::didReceiveData): Make a copy of the current part's data
        instead of using the resource data, because the latter is going to be
        overwritten by the new part being received.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21536 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-05-17 Oliver Hunt <oliver@apple.com>
oliver [Thu, 17 May 2007 11:32:11 +0000 (11:32 +0000)]
2007-05-17  Oliver Hunt  <oliver@apple.com>


        Reviewed by Justin.

        Fix for <rdar://problem/4244861> Safari fails to create image file
        after releasing dragged image that has changed on source page

        Tell xcode to copy HTMLImageElement.h and HTMLImageLoader.h headers
        across to WebKit as private headers.

        * WebCore.xcodeproj/project.pbxproj:


        Reviewed by Justin.

        Fix for <rdar://problem/4244861> Safari fails to create image file
        after releasing dragged image that has changed on source page

        This patch fixes this bug by manually creating an NSFileWrapper from
        the TIFF promise data for a drag if it is available.  This bypasses
        the problem of the required resource no longer being held due to page
        loads or other constraints.

        We need to leave the old path in place to allow for the case where
        the promised data is not available.

        * WebCoreSupport/WebDragClient.mm:
          Always use the top WebHTMLView as the pasteboard owner, this is
          safe as we only use the owner for resolving promised types.
        * WebView/WebHTMLView.mm:
        (-[WebHTMLView _writeSelectionToPasteboard:]):
        (-[WebHTMLView writeSelectionToPasteboard:types:]):
          Always use the top WebHTMLView as the pasteboard owner.
        (-[WebHTMLView namesOfPromisedFilesDroppedAtDestination:]):
          Add path to create NSFileWrapper from promise data.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21535 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
justing [Thu, 17 May 2007 10:46:00 +0000 (10:46 +0000)]

        Reviewed by harrison

        Demonstrates bug:
        * editing/execCommand/5190926-expected.checksum: Added.
        * editing/execCommand/5190926-expected.png: Added.
        * editing/execCommand/5190926-expected.txt: Added.
        * editing/execCommand/5190926.html: Added.

        Avoided some nesting:
        * editing/execCommand/5142012-2-expected.txt:
        * editing/execCommand/switch-list-type-expected.txt:


        Reviewed by harrison

        Gmail Editor: Crash when removing list style from underlined, multi-item list

        The user underlined some text, then created a list.
        The list was placed inside the inline element carrying
        the underline style, so, that element surrounded the
        list instead of being at the lowest level (just above
        text nodes). This kind of markup is more difficult to
        edit. Filed 5192089 and added a workaround so that we
        don't produce this kind of markup.

        * editing/InsertListCommand.cpp:
        (WebCore::InsertListCommand::doApply): Avoid inline
        elements when inserting the new list.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21534 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-05-17 Oliver Hunt <oliver@apple.com>
oliver [Thu, 17 May 2007 08:54:18 +0000 (08:54 +0000)]
2007-05-17  Oliver Hunt  <oliver@apple.com>

        Reviewed by Adam.

        Fix for <rdar://problem/4343832> Trying to drag a large 6.2MB jpeg image
         out of Safari is unexpectedly slow (4 copies of image plus RTF document on pasteboard)

        This patch causes the construction of the RTF and TIFF data to be delayed until requested.

        We delay TIFF construction from a DOMElement as this may require generating TIFF data
        from the CachedImage, which is slow.  To allow the TIFF data to be created later the
        it's necessary to add a reference to the source DOMElement to the view.

        * Misc/WebNSPasteboardExtras.h:
        * Misc/WebNSPasteboardExtras.mm:
        (-[NSPasteboard _web_writePromisedRTFDFromArchive:containsImage:]):
          Implements the delayed write of RTF data
        (-[NSPasteboard _web_writeImage:element:URL:title:archive:types:source:]):
          Set up the pasteboard to allow the data writing to be delayed
        (-[NSPasteboard _web_declareAndWriteDragImageForElement:URL:title:archive:source:]):
          Now we need to pass the WebHTMLView on to _web_writeImage:element:URL:title:archive:types:source:
        * WebCoreSupport/WebDragClient.mm:
          Clear the dragging pasteboard once the drag has ended to ensure we don't hold references
          to anything longer than we need to.

        * WebView/WebHTMLView.mm:
        (-[WebHTMLViewPrivate dealloc]):
        (-[WebHTMLViewPrivate clear]):
        (-[WebHTMLView pasteboardChangedOwner:]):
          Make sure we clear out the DOMElement reference once it is no longer needed
        (-[WebHTMLView pasteboard:provideDataForType:]):
          Provide delayed data
        (-[WebHTMLView _writeSelectionToPasteboard:]):
          Make sure we set pasteboard ownership correctly
        (-[WebHTMLView promisedDragTIFFDataSource]):
        (-[WebHTMLView setPromisedDragTIFFDataSource:]):
        (-[WebHTMLView writeSelectionToPasteboard:types:]):
        * WebView/WebHTMLViewInternal.h:
        * WebView/WebHTMLViewPrivate.h:

        * WebView/WebView.mm:
        (-[WebView _writeImageForElement:withPasteboardTypes:toPasteboard:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21533 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Allow redundant attributes with the same name in view source mode.
hyatt [Thu, 17 May 2007 08:52:21 +0000 (08:52 +0000)]
    Allow redundant attributes with the same name in view source mode.

        Reviewed by aroben

        * dom/NamedAttrMap.h:
        * html/HTMLTokenizer.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21532 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agogit-svn-id: https://svn.webkit.org/repository/webkit/trunk@21531 268f45cc-cd09-0410...
andersca [Thu, 17 May 2007 08:48:45 +0000 (08:48 +0000)]
git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21531 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoBack out last patch based on discussion with Zack
staikos [Thu, 17 May 2007 08:46:12 +0000 (08:46 +0000)]
Back out last patch based on discussion with Zack

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21530 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoWebCore:
andersca [Thu, 17 May 2007 08:38:44 +0000 (08:38 +0000)]

        Reviewed by Darin.

        <rdar://problem/5207156> Hamachi test tool causes assertion in FormCompletionController in Safari

        This happens because the form delegate gets passed the wrong source frame during the second submit().
        If a frame with the given name is found, FrameLoader::load is called with the same parameters on the found frame.
        This causes the found frame to be used as the source frame, which is wrong.

        The fix is to create the form state object before checking what frame to load the new request in.

        * WebCore.exp:
        * loader/FrameLoader.cpp:
        * loader/FrameLoader.h:
        * manual-tests/submit-form-with-target-twice.html: Added.


        Reviewed by Darin.

        <rdar://problem/5207156> Hamachi test tool causes assertion in FormCompletionController in Safari

        Update for WebCore changes.

        * WebView/WebFrame.mm:
        (-[WebFrame _loadURL:referrer:intoChild:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21529 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Oliver.
mjs [Thu, 17 May 2007 06:52:11 +0000 (06:52 +0000)]
    Reviewed by Oliver.

        <rdar://problem/5205914> REGRESSION: PLT stops after loading first page due to r21479

        * loader/FrameLoader.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21528 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoUpdate the lone view source mode layout test.
hyatt [Thu, 17 May 2007 06:28:57 +0000 (06:28 +0000)]
Update the lone view source mode layout test.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21527 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Improve the syntax-highlighted view source mode. It now preserves all whites...
hyatt [Thu, 17 May 2007 06:27:28 +0000 (06:27 +0000)]
    Improve the syntax-highlighted view source mode.  It now preserves all whitespace properly
        and highlights doctypes and comments (in addition to all the stuff it could highlight before).

        Reviewed by aroben

        * html/HTMLTokenizer.cpp:
        * html/HTMLTokenizer.h:
        * html/HTMLViewSourceDocument.cpp:
        * html/HTMLViewSourceDocument.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21526 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Build fix for some dev configurations
beidson [Thu, 17 May 2007 05:37:13 +0000 (05:37 +0000)]
    Build fix for some dev configurations

        * DumpRenderTree/DumpRenderTree.m:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21525 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Adam.
mjs [Thu, 17 May 2007 04:47:17 +0000 (04:47 +0000)]
    Reviewed by Adam.

        <rdar://problem/5191024> Page never loads completely (emonome.vodpod.com)
        <rdar://problem/5205162> REGRESSION (r21367): In the Address URL field, typing a site name doesn't redirect to site after pressing return
        <rdar://problem/5208951> REGRESSION: Progress bar never completes on link click that downloads (13694)
        <rdar://problem/5201598> REGRESSION: After uploading file to iDisk at mac.com, "loading" status remains and file does not appear in the list

        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::mainReceivedCompleteError): Make sure to always check whether the overall
        load completed, even if we think the current frame load is already complete.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21524 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
mjs [Thu, 17 May 2007 04:21:52 +0000 (04:21 +0000)]

        Reviewed by Justin.

        - added a bunch of layout tests using the new load delegate testing mechanism

        * http/tests/loading/bad-scheme-subframe-expected.txt: Added.
        * http/tests/loading/bad-scheme-subframe.html: Added.
        * http/tests/loading/bad-server-subframe-expected.txt: Added.
        * http/tests/loading/bad-server-subframe.html: Added.
        * http/tests/loading/basic-expected.txt: Added.
        * http/tests/loading/basic.html: Added.
        * http/tests/loading/empty-subframe-expected.txt: Added.
        * http/tests/loading/empty-subframe.html: Added.
        * http/tests/loading/simple-subframe-expected.txt: Added.
        * http/tests/loading/simple-subframe.html: Added.


        Reviewed by Justin.

        - created a new mechanism to log FrameLoadDelegate callbacks in directories named "loading"

        The reason for doing things in this slightly odd way is to make sure we don't miss the load delegates
        that happen before the load commits.

        Basically I moved WaitUntilDoneDelegate into a separate file (FrameLoadDelegate.m) and added
        optional logging for every FrameLoadDelegate callback, including SPI callbacks.

        * DumpRenderTree/DumpRenderTree.h:
        * DumpRenderTree/DumpRenderTree.m:
        (+[LayoutTestController isSelectorExcludedFromWebScript:]):
        (-[LayoutTestController clearBackForwardList]):
        (-[LayoutTestController setUseDashboardCompatibilityMode:]):
        (-[LayoutTestController dumpFrameLoadCallbacks]):
        (-[LayoutTestController setWindowIsKey:]):
        (-[LayoutTestController setMainFrameIsFirstResponder:]):
        (-[LayoutTestController _doLoad:target:]):
        (-[LayoutTestController _doBackOrForwardNavigation:]):
        (-[LayoutTestController queueReload]):
        (-[LayoutTestController queueScript:]):
        (-[LayoutTestController queueLoad:target:]):
        (-[LayoutTestController setAcceptsEditing:]):
        (-[LayoutTestController setTabKeyCyclesThroughElements:]):
        (+[DumpRenderTreeEvent mouseLocation]):
        * DumpRenderTree/DumpRenderTree.xcodeproj/project.pbxproj:
        * DumpRenderTree/DumpRenderTreeDraggingInfo.m:
        (-[DumpRenderTreeDraggingInfo draggingDestinationWindow]):
        * DumpRenderTree/EventSendingController.m:
        (-[EventSendingController mouseDown]):
        (-[EventSendingController mouseUp]):
        (-[EventSendingController mouseMoveToX:Y:]):
        (-[EventSendingController contextClick]):
        * DumpRenderTree/FrameLoadDelegate.h: Added.
        * DumpRenderTree/FrameLoadDelegate.m: Added.
        (-[WebFrame _drt_descriptionSuitableForTestResult]):
        (-[FrameLoadDelegate processWork:]):
        (-[FrameLoadDelegate webView:locationChangeDone:forDataSource:]):
        (-[FrameLoadDelegate webView:didStartProvisionalLoadForFrame:]):
        (-[FrameLoadDelegate webView:didCommitLoadForFrame:]):
        (-[FrameLoadDelegate webView:didFailProvisionalLoadWithError:forFrame:]):
        (-[FrameLoadDelegate webView:didFinishLoadForFrame:]):
        (-[FrameLoadDelegate webView:didFailLoadWithError:forFrame:]):
        (-[FrameLoadDelegate webView:windowScriptObjectAvailable:]):
        (-[FrameLoadDelegate webView:didClearWindowObject:forFrame:]):
        (-[FrameLoadDelegate webView:didReceiveTitle:forFrame:]):
        (-[FrameLoadDelegate webView:didReceiveServerRedirectForProvisionalLoadForFrame:]):
        (-[FrameLoadDelegate webView:didReceiveIcon:forFrame:]):
        (-[FrameLoadDelegate webView:didChangeLocationWithinPageForFrame:]):
        (-[FrameLoadDelegate webView:willPerformClientRedirectToURL:delay:fireDate:forFrame:]):
        (-[FrameLoadDelegate webView:didCancelClientRedirectForFrame:]):
        (-[FrameLoadDelegate webView:willCloseFrame:]):
        (-[FrameLoadDelegate webView:didFirstLayoutInFrame:]):
        (-[FrameLoadDelegate webView:didFinishDocumentLoadForFrame:]):
        (-[FrameLoadDelegate webView:didHandleOnloadEventsForFrame:]):
        * DumpRenderTree/ResourceLoadDelegate.m:
        (-[NSURL _drt_descriptionSuitableForTestResult]):
        * DumpRenderTree/UIDelegate.m:
        (-[UIDelegate webViewFocus:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21523 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
bdakin [Wed, 16 May 2007 21:47:31 +0000 (21:47 +0000)]

        Reviewed by Hyatt.

        Changes for <rdar://problem/5204107> REGRESSION: CSS web forum site
        layout not displaying properly in latest WebKit

        New test:
        * fast/table/tableInsideCaption-expected.checksum: Added.
        * fast/table/tableInsideCaption-expected.png: Added.
        * fast/table/tableInsideCaption-expected.txt: Added.
        * fast/table/tableInsideCaption.html: Added.

        Good changes.
        * tables/mozilla/bugs/bug48028-1-expected.checksum:
        * tables/mozilla/bugs/bug48028-1-expected.png:
        * tables/mozilla/bugs/bug48028-1-expected.txt:
        * tables/mozilla_expected_failures/core/captions1-
        * tables/mozilla_expected_failures/core/captions1-expected.png:
        * tables/mozilla_expected_failures/core/captions1-expected.txt:

        Not great changes. These changes expose other bugs. Filing reports
        for these.
        * tables/mozilla/bugs/bug48028-2-expected.txt:
        * tables/mozilla/bugs/bug55789-expected.checksum:
        * tables/mozilla/bugs/bug55789-expected.png:
        * tables/mozilla/bugs/bug55789-expected.txt:


        Reviewed by Hyatt.

        Fix for <rdar://problem/5204107> REGRESSION: CSS web forum site
        layout not displaying properly in latest WebKit

        Hyatt's http://trac.webkit.org/projects/webkit/changeset/21428
        exposed this bug. We were not handling error cases with <caption>
        correctly. This patch makes them do the same thing as colgroup.

        * html/HTMLParser.cpp:
        * html/HTMLParserErrorCodes.cpp: No need any more for caption-
        specific errors.
        * html/HTMLParserErrorCodes.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21522 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
kmccullo [Wed, 16 May 2007 20:57:47 +0000 (20:57 +0000)]

        Reviewed by Darin and Geoff.

        - rdar://problem/4981886
        - Added tests for allowing JS to close a window that was opened by the DOM.

        * fast/dom/open-and-close-by-DOM-expected.txt: Added.
        * fast/dom/open-and-close-by-DOM.html: Added.
        * fast/dom/resources/close.html: Added.
        * fast/dom/resources/middle.html: Added.


        Reviewed by Darin and Geoff.

        - rdar://problem/4981886
        - Now windows opened by the DOM can be closed by JS.

        * bindings/js/kjs_window.cpp:
        * loader/FrameLoader.cpp:
        * loader/FrameLoader.h:


        Reviewed by Darin and Geoff.

        - rdar://problem/4981886
        - Now windows opened by the DOM can be closed by JS.

        * DumpRenderTree/DumpRenderTree.m:
        (+[LayoutTestController isSelectorExcludedFromWebScript:]):
        (-[LayoutTestController windowCount]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21521 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
darin [Wed, 16 May 2007 20:15:51 +0000 (20:15 +0000)]

        Reviewed by Hyatt.

        - test for <rdar://problem/4928671> Crash due to runaway recursion when fieldset has display: table-row (12066)

        * fast/css/fieldset-display-row-expected.checksum: Added.
        * fast/css/fieldset-display-row-expected.png: Added.
        * fast/css/fieldset-display-row-expected.txt: Added.
        * fast/css/fieldset-display-row.html: Added.


        Reviewed by Hyatt.

        - fix <rdar://problem/4928671> Crash due to runaway recursion when fieldset has display: table-row (12066)

        Test: fast/css/fieldset-display-row.html

        * rendering/RenderContainer.cpp: (WebCore::RenderContainer::addChild):
        Change all decision making about special handling for new children to use
        the class of the render object instead of the style, except for the cases
        where the style is the only way to tell (TABLE_COLUMN_GROUP and TABLE_CAPTION).
        Thus a RenderFieldset with table-row style won't get handled as if it was
        a table row.

        * rendering/RenderTable.cpp: (WebCore::RenderTable::addChild): Same thing.
        Except that here the styles that we ahave to look at are TABLE_CAPTION,

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21520 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoImplement Canvas element in the Qt port.
zack [Wed, 16 May 2007 13:53:52 +0000 (13:53 +0000)]
Implement Canvas element in the Qt port.

Implementation of Context2D was completely missing in the
Qt port. This commit adds bigger parts of it.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21517 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFix shared timers.
zack [Wed, 16 May 2007 13:50:18 +0000 (13:50 +0000)]
Fix shared timers.

Make the timers correctly use miliseconds. Based on a patch
from George.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21516 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Zack
lars [Wed, 16 May 2007 11:12:29 +0000 (11:12 +0000)]
    Reviewed by Zack

        pass the mouse events to the event handler, not the frameview.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21515 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Update layout tests for error handling changes.
lars [Wed, 16 May 2007 11:05:51 +0000 (11:05 +0000)]
 Update layout tests for error handling changes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21514 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoUpdate layout tests for error handling changes.
hyatt [Wed, 16 May 2007 09:25:45 +0000 (09:25 +0000)]
Update layout tests for error handling changes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21513 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFix Qt build.
hyatt [Wed, 16 May 2007 08:47:44 +0000 (08:47 +0000)]
Fix Qt build.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21512 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Add rudimentary error reporting to the HTML parser. This is disabled
hyatt [Wed, 16 May 2007 08:11:05 +0000 (08:11 +0000)]
    Add rudimentary error reporting to the HTML parser.  This is disabled
        for now until it gets beefed up a bit more (and until we work out how
        best to pass the messages up without hurting perf).

        Clean up and remove dead error handling code and incorrect error handling
        code as well.

        Reviewed by mjs

        * WebCore.xcodeproj/project.pbxproj:
        * html/HTMLParser.cpp:
        * html/HTMLParser.h:
        * html/HTMLParserErrorCodes.cpp: Added.
        * html/HTMLParserErrorCodes.h: Added.
        * html/HTMLTokenizer.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21511 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Make FrameLoader write "<html><body>" instead of just "<body>".
hyatt [Wed, 16 May 2007 06:22:17 +0000 (06:22 +0000)]
    Make FrameLoader write "<html><body>" instead of just "<body>".

        Reviewed by aroben

        * loader/FrameLoader.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21510 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agorepair changelog
staikos [Wed, 16 May 2007 05:04:25 +0000 (05:04 +0000)]
repair changelog

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21509 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoMake the loader pass cookies through
staikos [Wed, 16 May 2007 04:47:43 +0000 (04:47 +0000)]
Make the loader pass cookies through

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21508 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-05-15 Adam Treat <adam@staikos.net>
staikos [Wed, 16 May 2007 04:25:38 +0000 (04:25 +0000)]
2007-05-15  Adam Treat  <adam@staikos.net>

        Reviewed by George Staikos.

        - Implement drawArc with proper units.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21507 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-05-15 Adam Treat <adam@staikos.net>
staikos [Wed, 16 May 2007 04:19:14 +0000 (04:19 +0000)]
2007-05-15  Adam Treat  <adam@staikos.net>

        Reviewed by George Staikos.

        - Add missing paint for nested subframes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21506 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
weinig [Wed, 16 May 2007 03:35:24 +0000 (03:35 +0000)]

        Reviewed by Oliver.

        - Yet more tests for http://bugs.webkit.org/show_bug.cgi?id=13722
          Finish testing and fixing the setting of DOM attributes to JS null

          Tests for SVG attributes that take a DOMString.

        * fast/dom/svg-element-attribute-js-null-expected.txt: Added.
        * fast/dom/svg-element-attribute-js-null.xhtml: Added.


        Reviewed by Oliver.

        - Patch for http://bugs.webkit.org/show_bug.cgi?id=13722
          Finish testing and fixing the setting of DOM attributes to JS null

          Convert JS null to null string for SVG attributes.

        Test: fast/dom/svg-element-attribute-js-null.xhtml

        * ksvg2/svg/SVGAngle.idl:
        * ksvg2/svg/SVGElement.idl:
        * ksvg2/svg/SVGLength.idl:
        * ksvg2/svg/SVGScriptElement.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21505 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
weinig [Wed, 16 May 2007 03:24:01 +0000 (03:24 +0000)]

        Reviewed by Oliver.

        - More Tests for http://bugs.webkit.org/show_bug.cgi?id=13722
          Finish testing and fixing the setting of DOM attributes to JS null

          Tests for CanvasRenderingContext2D attributes that take a DOMString.

        * fast/dom/canvasContext2d-element-attribute-js-null-expected.txt: Added.
        * fast/dom/canvasContext2d-element-attribute-js-null.html: Added.


        Reviewed by Oliver.

        - Patch for http://bugs.webkit.org/show_bug.cgi?id=13722
          Finish testing and fixing the setting of DOM attributes to JS null

          Convert JS null to null string for CanvasRenderingContext2D attributes.

        Test: fast/dom/canvasContext2d-element-attribute-js-null.html

        * html/CanvasRenderingContext2D.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21504 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-05-15 Oliver Hunt <oliver@apple.com>
oliver [Wed, 16 May 2007 02:57:54 +0000 (02:57 +0000)]
2007-05-15  Oliver Hunt  <oliver@apple.com>


        Reviewed by Sam and Geoff.

        Removing dead code left behind from drag and drop refactoring.

        * page/mac/WebCoreFrameBridge.h:


        Reviewed by Sam and Geoff.

        Removing dead code left behind from drag and drop refactoring.

        * WebCoreSupport/WebFrameBridge.mm:
        * WebView/WebHTMLView.mm:
        * WebView/WebHTMLViewPrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21503 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
weinig [Wed, 16 May 2007 02:22:45 +0000 (02:22 +0000)]

        Reviewed by Oliver.

        - More Tests for http://bugs.webkit.org/show_bug.cgi?id=13722
          Finish testing and fixing the setting of DOM attributes to JS null

          Tests for CSSOM attributes that take a DOMString.

        * fast/dom/css-element-attribute-js-null-expected.txt: Added.
        * fast/dom/css-element-attribute-js-null.html: Added.


        Reviewed by Oliver.

        - Second Patch for http://bugs.webkit.org/show_bug.cgi?id=13722
          Finish testing and fixing the setting of DOM attributes to JS null

          Convert JS null to null string for remaining CSSOM attributes.

        Test: fast/dom/css-element-attribute-js-null.html

        * bindings/js/kjs_css.cpp:
        * css/CSSPageRule.idl:
        * css/CSSRule.idl:
        * css/CSSStyleDeclaration.idl:
        * css/CSSStyleRule.idl:
        * css/CSSValue.idl:
        * css/MediaList.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21502 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by David Kilzer.
aroben [Tue, 15 May 2007 23:35:59 +0000 (23:35 +0000)]
    Reviewed by David Kilzer.

        prepare-ChangeLog should work with git

        * Scripts/prepare-ChangeLog: Added support for Git.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21501 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-05-15 Bruce Q Hammond <bruceq@apple.com>
bdash [Tue, 15 May 2007 23:19:41 +0000 (23:19 +0000)]
2007-05-15  Bruce Q Hammond  <bruceq@apple.com>

        Reviewed by Darin.

        Correction of previous patch for http://bugs.webkit.org/show_bug.cgi?id=13578
        This corrects the sign of the Y-Axis origin adjustment.

        * Plugins/WebBaseNetscapePluginView.mm:
        (-[WebBaseNetscapePluginView saveAndSetNewPortStateForUpdate:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21500 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Updating expected results for r21074.
justing [Tue, 15 May 2007 22:35:53 +0000 (22:35 +0000)]
    Updating expected results for r21074.

        * css1/cascade/cascade_order-expected.checksum:
        * css1/cascade/cascade_order-expected.png:
        * css2.1/t0402-c71-fwd-parsing-02-f-expected.checksum:
        * css2.1/t0402-c71-fwd-parsing-02-f-expected.png:
        * css2.1/t0505-c16-descendant-01-e-expected.checksum:
        * css2.1/t0505-c16-descendant-01-e-expected.png:
        * css2.1/t050803-c14-classes-00-e-expected.checksum:
        * css2.1/t050803-c14-classes-00-e-expected.png:
        * css2.1/t0509-c15-ids-01-e-expected.checksum:
        * css2.1/t0509-c15-ids-01-e-expected.png:
        * css2.1/t0805-c5518-brdr-t-01-e-expected.checksum:
        * css2.1/t0805-c5518-brdr-t-01-e-expected.png:
        * css2.1/t0805-c5519-brdr-r-02-e-expected.checksum:
        * css2.1/t0805-c5519-brdr-r-02-e-expected.png:
        * css2.1/t0805-c5520-brdr-b-01-e-expected.checksum:
        * css2.1/t0805-c5520-brdr-b-01-e-expected.png:
        * css2.1/t0805-c5521-brdr-l-02-e-expected.checksum:
        * css2.1/t0805-c5521-brdr-l-02-e-expected.png:
        * css2.1/t1205-c563-list-type-00-b-expected.checksum:
        * css2.1/t1205-c563-list-type-00-b-expected.png:
        * css2.1/t1205-c564-list-img-00-b-g-expected.checksum:
        * css2.1/t1205-c564-list-img-00-b-g-expected.png:
        * fast/backgrounds/repeat/noRepeatCorrectClip-expected.checksum:
        * fast/backgrounds/repeat/noRepeatCorrectClip-expected.png:
        * fast/block/float/014-expected.checksum:
        * fast/block/float/014-expected.png:
        * fast/css/001-expected.checksum:
        * fast/css/001-expected.png:
        * fast/css/css2-system-fonts-expected.checksum:
        * fast/css/css2-system-fonts-expected.png:
        * fast/doctypes/001-expected.checksum:
        * fast/doctypes/001-expected.png:
        * fast/doctypes/002-expected.checksum:
        * fast/doctypes/002-expected.png:
        * fast/doctypes/003-expected.checksum:
        * fast/doctypes/003-expected.png:
        * fast/doctypes/004-expected.checksum:
        * fast/doctypes/004-expected.png:
        * fast/dom/Window/window-onFocus-expected.checksum:
        * fast/dom/Window/window-onFocus-expected.png:
        * fast/events/keydown-1-expected.checksum:
        * fast/events/keydown-1-expected.png:
        * fast/lists/002-expected.checksum:
        * fast/lists/002-expected.png:
        * fast/lists/003-expected.checksum:
        * fast/lists/003-expected.png:
        * fast/lists/005-expected.checksum:
        * fast/lists/005-expected.png:
        * fast/lists/007-expected.checksum:
        * fast/lists/007-expected.png:
        * fast/lists/008-expected.checksum:
        * fast/lists/008-expected.png:
        * fast/lists/big-list-marker-expected.checksum:
        * fast/lists/big-list-marker-expected.png:
        * fast/lists/dynamic-marker-crash-expected.checksum:
        * fast/lists/dynamic-marker-crash-expected.png:
        * fast/lists/item-not-in-list-line-wrapping-expected.checksum:
        * fast/lists/item-not-in-list-line-wrapping-expected.png:
        * fast/lists/marker-image-error-expected.checksum:
        * fast/lists/marker-image-error-expected.png:
        * fast/lists/ol-display-types-expected.checksum:
        * fast/lists/ol-display-types-expected.png:
        * fast/lists/scrolled-marker-paint-expected.checksum:
        * fast/lists/scrolled-marker-paint-expected.png:
        * fast/overflow/overflow-rtl-expected.checksum:
        * fast/overflow/overflow-rtl-expected.png:
        * fast/repaint/list-marker-2-expected.checksum:
        * fast/repaint/list-marker-2-expected.png:
        * fast/repaint/list-marker-expected.checksum:
        * fast/repaint/list-marker-expected.png:
        * fast/selectors/001-expected.checksum:
        * fast/selectors/001-expected.png:
        * fast/selectors/013-expected.checksum:
        * fast/selectors/013-expected.png:
        * fast/selectors/015-expected.checksum:
        * fast/selectors/015-expected.png:
        * fast/selectors/166-expected.checksum:
        * fast/selectors/166-expected.png:
        * tables/mozilla/bugs/bug23235-expected.checksum:
        * tables/mozilla/bugs/bug23235-expected.png:
        * tables/mozilla/bugs/bug30692-expected.checksum:
        * tables/mozilla/bugs/bug30692-expected.png:
        * tables/mozilla/bugs/bug3191-expected.checksum:
        * tables/mozilla/bugs/bug3191-expected.png:
        * tables/mozilla/marvin/backgr_index-expected.checksum:
        * tables/mozilla/marvin/backgr_index-expected.png:
        * tables/mozilla/marvin/backgr_layers-opacity-expected.checksum:
        * tables/mozilla/marvin/backgr_layers-opacity-expected.png:
        * tables/mozilla_expected_failures/bugs/bug1010-expected.checksum:
        * tables/mozilla_expected_failures/bugs/bug1010-expected.png:
        * tables/mozilla_expected_failures/marvin/backgr_fixed-bg-expected.checksum:
        * tables/mozilla_expected_failures/marvin/backgr_fixed-bg-expected.png:
        * tables/mozilla_expected_failures/marvin/backgr_layers-show-expected.checksum:
        * tables/mozilla_expected_failures/marvin/backgr_layers-show-expected.png:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21499 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Hyatt.
darin [Tue, 15 May 2007 22:19:19 +0000 (22:19 +0000)]
    Reviewed by Hyatt.

        * css/html4.css: Remove wbr style rule -- not needed any more.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21498 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
beidson [Tue, 15 May 2007 22:03:05 +0000 (22:03 +0000)]

        Reviewed by Brady

        - test for http://bugs.webkit.org/show_bug.cgi?id=13360
          REGRESSION: Crash closing live web cam viewer page

        * http/tests/misc/multipart-expected.txt: Added.
        * http/tests/misc/multipart.html: Added.
        * http/tests/misc/resources/1x1-blue.png: Added.
        * http/tests/misc/resources/multipart.php: Added.


        Reviewed by Brady

        - fix http://bugs.webkit.org/show_bug.cgi?id=13360
          REGRESSION: Crash closing live web cam viewer page

        Test: http/tests/misc/multipart.html

        * loader/ResourceLoader.cpp:
        (WebCore::ResourceLoader::didFail): Avoid calling didFailToLoad() if
        didFinishLoad() has already been called (when the first part of a
        multipart resource arrived).
        (WebCore::ResourceLoader::didCancel): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21497 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Adele Peterson.
ggaren [Tue, 15 May 2007 19:04:50 +0000 (19:04 +0000)]
    Reviewed by Adele Peterson.

        It helps if you swap the right variable.

        * wtf/HashSet.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21494 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
weinig [Tue, 15 May 2007 17:53:39 +0000 (17:53 +0000)]

        Reviewed by Darin.

        - Test for http://bugs.webkit.org/show_bug.cgi?id=13722
          Finish testing and fixing the setting of DOM attributes to JS null

          Tests for Core DOM attributes.

        * fast/dom/coreDOM-element-attribute-js-null-expected.txt: Added.
        * fast/dom/coreDOM-element-attribute-js-null.xhtml: Added.


        Reviewed by Darin.

        - Patch for http://bugs.webkit.org/show_bug.cgi?id=13722
          Finish testing and fixing the setting of DOM attributes to JS null

          Convert JS null to null string for remaining Core DOM attributes.

        Test: fast/dom/coreDOM-element-attribute-js-null.xhtml

        * dom/Attr.cpp:
        (WebCore::Attr::setValue): Don't throw an expection for null values as this
        is not in the spec and Firefox does not do it either.
        * dom/Attr.idl:
        * dom/CharacterData.idl:
        * dom/ProcessingInstruction.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21489 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej Stachowiak.
ggaren [Tue, 15 May 2007 16:47:18 +0000 (16:47 +0000)]
    Reviewed by Maciej Stachowiak.

        Patch: fixed <rdar://problem/5198890> .5% performance regression caused
        by r21307

        The only code r21307 added that runs during the PLT is a frame load delegate
        -respondsToSelector: call inside windowObjectCleared(), so it seems like
        our message dispatch overhead for the frame load delegate is significant.

        This patch is a straight port of Maciej's fix for the same problem in the
        resource load delegate. The solution is simple enough: don't use Objective-C.
        Instead, use a special structure that caches which methods the delegate
        implements, along with pointers to those methods.

        I verified each frame load delegate callback in the debugger, except for
        -webView:didFailLoadWithError:forFrame:, which is not implemented by Safari
        or DumpRenderTree.

        * WebKit/DefaultDelegates/WebDefaultFrameLoadDelegate.h: Removed.
        * WebKit/DefaultDelegates/WebDefaultFrameLoadDelegate.m: Removed.

        * WebCoreSupport/WebFrameBridge.mm:
        (-[WebFrameBridge windowObjectCleared]): I also removed a misleading
        comment here. The JS debugger caches the windowScriptObject, so you do
        need to re-create the debugger every time you invalidate the old
        WebScriptObject wrapper for the window object and create a new one, or
        the debugger will stop working. We could fix this in a number of ways,
        but <rdar://problem/4608404> is not the key issue.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21487 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoUpdate test result now that we have the eventSender object.
lars [Tue, 15 May 2007 14:38:58 +0000 (14:38 +0000)]
Update test result now that we have the eventSender object.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21486 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Extend the QObject JavaScript bindings to work for slots with
lars [Tue, 15 May 2007 13:40:15 +0000 (13:40 +0000)]
    Extend the QObject JavaScript bindings to work for slots with

        Add the EventSender object to DRT. Currently it implements
        mouseDown, mouseUp and mouseMoveTo.

        Make run-webkit-tests --strict work somewhat better for form elements.

        one more test passing for Qt, the first one that uses
        the eventSender in DRT.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21485 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Updating expected results for r21074.
justing [Tue, 15 May 2007 12:45:31 +0000 (12:45 +0000)]
    Updating expected results for r21074.

        * editing/deleting/4866671-expected.checksum:
        * editing/deleting/4866671-expected.png:
        * editing/deleting/delete-first-list-item-expected.checksum:
        * editing/deleting/delete-first-list-item-expected.png:
        * editing/deleting/delete-listitem-002-expected.checksum:
        * editing/deleting/delete-listitem-002-expected.png:
        * editing/deleting/list-item-1-expected.checksum:
        * editing/deleting/list-item-1-expected.png:
        * editing/deleting/merge-into-empty-block-2-expected.checksum:
        * editing/deleting/merge-into-empty-block-2-expected.png:
        * editing/execCommand/4641880-1-expected.checksum:
        * editing/execCommand/4641880-1-expected.png:
        * editing/execCommand/4747450-expected.checksum:
        * editing/execCommand/4747450-expected.png:
        * editing/execCommand/4916402-expected.checksum:
        * editing/execCommand/4916402-expected.png:
        * editing/execCommand/4916583-expected.checksum:
        * editing/execCommand/4916583-expected.png:
        * editing/execCommand/5142012-2-expected.checksum:
        * editing/execCommand/5142012-2-expected.png:
        * editing/execCommand/5144139-1-expected.checksum:
        * editing/execCommand/5144139-1-expected.png:
        * editing/execCommand/create-list-with-hr-expected.checksum:
        * editing/execCommand/create-list-with-hr-expected.png:
        * editing/execCommand/indent-list-item-expected.checksum:
        * editing/execCommand/indent-list-item-expected.png:
        * editing/execCommand/indent-selection-expected.checksum:
        * editing/execCommand/indent-selection-expected.png:
        * editing/execCommand/insert-list-empty-div-expected.checksum:
        * editing/execCommand/insert-list-empty-div-expected.png:
        * editing/execCommand/remove-list-item-1-expected.checksum:
        * editing/execCommand/remove-list-item-1-expected.png:
        * editing/inserting/4959067-expected.checksum:
        * editing/inserting/4959067-expected.png:
        * editing/pasteboard/4861080-expected.checksum:
        * editing/pasteboard/4861080-expected.png:
        * editing/pasteboard/drag-selected-image-to-contenteditable-expected.checksum:
        * editing/pasteboard/drag-selected-image-to-contenteditable-expected.png:
        * editing/pasteboard/drop-link-expected.checksum:
        * editing/pasteboard/drop-link-expected.png:
        * editing/pasteboard/innerText-inline-table-expected.checksum:
        * editing/pasteboard/innerText-inline-table-expected.png:
        * editing/pasteboard/input-field-1-expected.checksum:
        * editing/pasteboard/input-field-1-expected.png:
        * editing/pasteboard/merge-end-list-expected.checksum:
        * editing/pasteboard/merge-end-list-expected.png:
        * editing/pasteboard/merge-start-list-expected.checksum:
        * editing/pasteboard/merge-start-list-expected.png:
        * editing/pasteboard/select-element-1-expected.checksum:
        * editing/pasteboard/select-element-1-expected.png:
        * editing/selection/drag-to-contenteditable-iframe-expected.checksum:
        * editing/selection/drag-to-contenteditable-iframe-expected.png:
        * editing/selection/extend-by-word-002-expected.checksum:
        * editing/selection/extend-by-word-002-expected.png:
        * editing/selection/select-all-iframe-expected.checksum:
        * editing/selection/select-all-iframe-expected.png:
        * editing/selection/selectNode-expected.checksum:
        * editing/selection/selectNode-expected.png:
        * editing/selection/selectNodeContents-expected.checksum:
        * editing/selection/selectNodeContents-expected.png:
        * editing/unsupported-content/list-delete-001-expected.checksum:
        * editing/unsupported-content/list-delete-001-expected.png:
        * editing/unsupported-content/list-type-after-expected.checksum:
        * editing/unsupported-content/list-type-after-expected.png:
        * editing/unsupported-content/list-type-before-expected.checksum:
        * editing/unsupported-content/list-type-before-expected.png:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21484 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Hyatt.
andrew [Tue, 15 May 2007 10:00:12 +0000 (10:00 +0000)]
    Reviewed by Hyatt.

        - fix http://bugs.webkit.org/show_bug.cgi?id=13718
          REGRESSION (r21093): editing/execCommand/5136770.html failing pixel test

        * rendering/RenderContainer.cpp:
        (WebCore::RenderContainer::removeChildNode): Moved the code that clears the
        selection when moving a selection border into the fast path.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21483 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Hyatt.
rwlbuis [Tue, 15 May 2007 09:31:24 +0000 (09:31 +0000)]
    Reviewed by Hyatt.

        "Make Text Bigger/Smaller" affects SVGs when it probably shouldn't

        Ignore Make Text Bigger/Smaller for SVG text.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21482 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
andersca [Tue, 15 May 2007 09:28:42 +0000 (09:28 +0000)]

        Reviewed by Maciej.

        ASSERT or null value when trying to use a getter defined on a window object

        Add tests.

        * fast/dom/getter-on-window-object-expected.txt: Added.
        * fast/dom/getter-on-window-object.html: Added.


        Reviewed by Maciej.

        ASSERT or null value when trying to use a getter defined on a window object

        Handle the case where the value returned is a getter/setter object.

        * bindings/js/kjs_window.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21481 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Geoff.
andersca [Tue, 15 May 2007 08:27:42 +0000 (08:27 +0000)]
    Reviewed by Geoff.

        REGRESSION: With Shiira 1.2.2 , I can't open embedded link in flash object by clicking (http:/www.adobe.com )

        Shiira under some circumstances passes nil as the request to -[WebFrame loadRequest:]. ToT WebKit doesn't call any policy
        delegate methods in this case, which means that the page is actually being loaded.

        This patch makes ToT WebKit behave in the same way as Tiger WebKit when the request is nil.

        * loader/FrameLoader.cpp:
        Don't continue without calling the navigation policy method if the request is null.

        If the request can't be handled, don't continue the load.

        * platform/network/mac/ResourceRequestMac.mm:
        Don't create an NSURLRequest object if the request is null.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21480 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoWebCore:
mjs [Tue, 15 May 2007 08:24:53 +0000 (08:24 +0000)]

        Reviewed by Hyatt and Lars.

        - fixed <rdar://problem/5201758> REGRESSION: Stop button enabled and other problems caused by [WebView currentURL] returning non-nil for empty window

        * WebCore.exp: export new symbol
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::begin): move marker that we committed a real doc load from here...
        (WebCore::FrameLoader::transitionToCommitted): ...to here
        (WebCore::FrameLoader::frameHasLoaded): check that we have at least started loading something
        besides the initial empty doc.
        * loader/FrameLoader.h:


        Reviewed by Hyatt and Lars.

        - fixed <rdar://problem/5201758> REGRESSION: Stop button enabled and other problems caused by [WebView currentURL] returning non-nil for empty window

        * WebView/WebDataSource.mm:
        (-[WebDataSource request]): Return nil when we are still showing the initial empty doc

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21479 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago5319 -> 5879 passing test cases for the Qt build.
lars [Tue, 15 May 2007 07:37:22 +0000 (07:37 +0000)]
5319 -> 5879 passing test cases for the Qt build.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21478 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Spelling markers fixed:
justing [Tue, 15 May 2007 07:31:50 +0000 (07:31 +0000)]
    Spelling markers fixed:
        * editing/deleting/delete-block-table-expected.checksum:
        * editing/deleting/delete-block-table-expected.png:
        * editing/selection/move-by-line-001-expected.checksum:
        * editing/selection/move-by-line-001-expected.png:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21477 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
justing [Tue, 15 May 2007 06:15:21 +0000 (06:15 +0000)]

        Reviewed by ggaren

        Updated these expected results.  We now clear the
        selection inside a focused node *after* firing the
        mousedown event handler, which matches FF:
        * fast/forms/focus-selection-input-expected.txt:
        * fast/forms/focus-selection-textarea-expected.txt:


        Reviewed by ggaren

        REGRESSION: Three new layout test failures

        Two failures are correct.  Updated their expected results.

        In fast/events/frame-tab-focus.html, as we advance
        through focusable nodes, we descend into a subframe
        to focus a node and then ascend out of it into the
        main frame to focus the next.  When we focus the main
        frame, the node in that frame that was previously
        focused and contains an inactive selection is focused
        momentarily because setCaretVisible tries to focus the
        node containing the caret.

        * page/Frame.cpp:
        (WebCore::Frame::setCaretVisible): Don't focus the
        node containing the caret. FocusController will focus
        the previously focused node (which will contain the
        caret) if necessary when the frame gains focus.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21476 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by ggaren
justing [Tue, 15 May 2007 05:59:35 +0000 (05:59 +0000)]
    Reviewed by ggaren

        Spelling markers fixed:
        * editing/deleting/delete-to-select-table-expected.checksum:
        * editing/deleting/delete-to-select-table-expected.png:
        * editing/selection/end-of-document-expected.checksum:
        * editing/selection/end-of-document-expected.png:
        * editing/inserting/insert-paragraph-02-expected.checksum:
        * editing/inserting/insert-paragraph-02-expected.png:

        Focus halo now painted:
        * editing/execCommand/format-block-with-braces-expected.checksum:
        * editing/execCommand/format-block-with-braces-expected.png:

        Needed FIXMEs removed after fixing 4714993:
        * editing/pasteboard/merge-end-1-expected.checksum:
        * editing/pasteboard/merge-end-1-expected.png:
        * editing/pasteboard/merge-end-1-expected.txt:
        * editing/pasteboard/merge-end-1.html:
        * editing/pasteboard/merge-end-2-expected.checksum:
        * editing/pasteboard/merge-end-2-expected.png:
        * editing/pasteboard/merge-end-2-expected.txt:
        * editing/pasteboard/merge-end-2.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21475 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
mjs [Tue, 15 May 2007 05:53:17 +0000 (05:53 +0000)]

        Reviewed by Brady.

        - Test case for <rdar://problem/5203036> REGRESSION: Assertion failure in WebCore::FrameLoader::restoreScrollPositionAndViewState() going back from fark.com Photoshop contest (13684)

        * http/tests/incremental/slow-utf8-html.pl:
        * http/tests/navigation/back-to-slow-frame.html: Added.
        * http/tests/navigation/back-to-slow-frame-expected.txt: Added.
        * http/tests/navigation/resources/back-to-slow-frame-2.html: Added.
        * http/tests/navigation/resources/back-to-slow-frame-3.html: Added.
        * http/tests/navigation/resources/slow-resource-1-sec.pl: Added.


        Reviewed by Brady.

        - fixed <rdar://problem/5203036> REGRESSION: Assertion failure in WebCore::FrameLoader::restoreScrollPositionAndViewState() going back from fark.com Photoshop contest (13684)

        Test: http/tests/navigation/back-to-slow-frame.html

        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::restoreScrollPositionAndViewState): Add needed early return.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21474 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-05-14 Kimmo Kinnunen <kimmok@iki.fi>
bdash [Tue, 15 May 2007 01:53:54 +0000 (01:53 +0000)]
2007-05-14  Kimmo Kinnunen  <kimmok@iki.fi>

        Reviewed by Darin.

        - Fixes http://bugs.webkit.org/show_bug.cgi?id=13622 (Decompiler
          omits trailing comma in array literal)

         * kjs/nodes2string.cpp:
         (ArrayNode::streamTo): print extra ',' in case there was elision
         commas (check opt member var) and array elements present
         in the array expression

2007-05-14  Kimmo Kinnunen  <kimmok@iki.fi>

        Reviewed by Darin.

        - Tests for fix to http://bugs.webkit.org/show_bug.cgi?id=13622 (Decompiler omits
          trailing comma in array literal)

        * fast/js/resources/toString-elision-trailing-comma.js: Added.
        * fast/js/toString-elision-trailing-comma-expected.txt: Added.
        * fast/js/toString-elision-trailing-comma.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21473 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-05-14 Mitz Pettel <mitz@webkit.org>
bdash [Tue, 15 May 2007 01:41:53 +0000 (01:41 +0000)]
2007-05-14  Mitz Pettel  <mitz@webkit.org>

        Reviewed by Dave Hyatt.

        - fix http://bugs.webkit.org/show_bug.cgi?id=13603
          style leaks in washingtonpost.com
          also known as http://bugs.webkit.org/show_bug.cgi?id=8750
          w3c.org: incorrect recovery from unclosed <a> elements
          also known as http://bugs.webkit.org/show_bug.cgi?id=12808
          Handling of misnested tags results in entire page appearing as link
          also known as http://bugs.webkit.org/show_bug.cgi?id=12861
          Post text renders as link

        Test: fast/parser/residual-style-close-across-n-blocks.html

        * html/HTMLParser.cpp:
        (WebCore::HTMLParser::handleResidualStyleCloseTagAcrossBlocks): Changed to
        deal with multiple block crossings instead of bailing out, by doing multiple
        reparenting passes, starting from the outermost block.

2007-05-14  Mitz Pettel  <mitz@webkit.org>

        Reviewed by Dave Hyatt.

        - test for http://bugs.webkit.org/show_bug.cgi?id=13603
          style leaks in washingtonpost.com

        * fast/parser/residual-style-close-across-n-blocks-expected.txt: Added.
        * fast/parser/residual-style-close-across-n-blocks.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21472 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-05-14 Mitz Pettel <mitz@webkit.org>
bdash [Tue, 15 May 2007 01:39:27 +0000 (01:39 +0000)]
2007-05-14  Mitz Pettel  <mitz@webkit.org>

        Reviewed by Darin.

        - fix http://bugs.webkit.org/show_bug.cgi?id=13711
          REGRESSION - Bookmark Bar text rendering changed

        No test possible since this patch affects only Safari chrome.

        * platform/Font.cpp:
        (WebCore::Font::Font): Initialize m_isPlatformFont according to whether
        the Font object is derived from a FontPlatformData or not.
        * platform/Font.h:
        (WebCore::Font::isPlatformFont): Added. Returns whether this Font was
        derived from a platform font.
        * platform/mac/FontCacheMac.mm:
        (WebCore::FontCache::getFontDataForCharacters): Changed to not enable
        synthetic bold and italic in the fallback fonts if the original font
        was derived from a platform font.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21471 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
weinig [Tue, 15 May 2007 01:32:35 +0000 (01:32 +0000)]

        Reviewed by Alice.

        - Tests for http://bugs.webkit.org/show_bug.cgi?id=13541
          REGRESSION: Setting a DOMString attribute to JS null in the
          JS bindings should default to converting to the empty string

        * fast/dom/document-attribute-js-null-expected.txt: Added.
        * fast/dom/document-attribute-js-null.html: Added.
        * fast/dom/element-attribute-js-null-expected.txt: Added.
        * fast/dom/element-attribute-js-null.html: Added.
        * fast/forms/input-null-expected.txt: Removed.
        * fast/forms/input-null.html: Removed.


        Reviewed by Alice!

        - Patch for http://bugs.webkit.org/show_bug.cgi?id=13541
          REGRESSION: Setting a DOMString attribute to JS null in the
          JS bindings should default to converting to the empty string

          Convert JS null to the empty string when passing it to an attribute that takes
          a DOMString.

        Test: fast/dom/document-attribute-js-null.html
        Test: fast/dom/element-attribute-js-null.html

        * bindings/js/kjs_dom.cpp:
        * bindings/js/kjs_html.cpp:
        * bindings/js/kjs_html.h:
        * dom/Document.idl:
        * dom/Node.idl:
        * html/HTMLAnchorElement.idl:
        * html/HTMLAppletElement.idl:
        * html/HTMLAreaElement.idl:
        * html/HTMLBRElement.idl:
        * html/HTMLBaseElement.idl:
        * html/HTMLBaseFontElement.idl:
        * html/HTMLBlockquoteElement.idl:
        * html/HTMLBodyElement.idl:
        * html/HTMLButtonElement.idl:
        * html/HTMLDivElement.idl:
        * html/HTMLDocument.idl:
        * html/HTMLElement.idl:
        * html/HTMLEmbedElement.idl:
        * html/HTMLFontElement.idl:
        * html/HTMLFormElement.idl:
        * html/HTMLFrameElement.idl:
        * html/HTMLFrameSetElement.idl:
        * html/HTMLHRElement.idl:
        * html/HTMLHeadElement.idl:
        * html/HTMLHeadingElement.idl:
        * html/HTMLHtmlElement.idl:
        * html/HTMLIFrameElement.idl:
        * html/HTMLImageElement.idl:
        * html/HTMLInputElement.idl:
        * html/HTMLIsIndexElement.idl:
        * html/HTMLLIElement.idl:
        * html/HTMLLabelElement.idl:
        * html/HTMLLegendElement.idl:
        * html/HTMLLinkElement.idl:
        * html/HTMLMapElement.idl:
        * html/HTMLMetaElement.idl:
        * html/HTMLModElement.idl:
        * html/HTMLOListElement.idl:
        * html/HTMLObjectElement.idl:
        * html/HTMLOptGroupElement.idl:
        * html/HTMLOptionElement.idl:
        * html/HTMLParagraphElement.idl:
        * html/HTMLParamElement.idl:
        * html/HTMLQuoteElement.idl:
        * html/HTMLScriptElement.idl:
        * html/HTMLSelectElement.idl:
        * html/HTMLStyleElement.idl:
        * html/HTMLTableCaptionElement.idl:
        * html/HTMLTableCellElement.idl:
        * html/HTMLTableColElement.idl:
        * html/HTMLTableElement.idl:
        * html/HTMLTableRowElement.idl:
        * html/HTMLTableSectionElement.idl:
        * html/HTMLTextAreaElement.idl:
        * html/HTMLTitleElement.idl:
        * html/HTMLUListElement.idl:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21470 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Trying to fix the Qt build.
ggaren [Tue, 15 May 2007 00:24:54 +0000 (00:24 +0000)]
    Trying to fix the Qt build.

        * WebCore.pro:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21469 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoJavaScriptCore:
ggaren [Tue, 15 May 2007 00:20:16 +0000 (00:20 +0000)]

        Reviewed by Oliver Hunt.

        Added HashMap::swap and HashSet::swap. WebCore now uses HashSet::swap.
        I figured while I was in the neighborhood I might as well add HashMap::swap,

        * wtf/HashMap.h:
        * wtf/HashSet.h:


        Reviewed by Oliver Hunt.

        Some improvements on my last patch, suggested by Darin.

        * history/PageCache.cpp:
        (WebCore::PageCache::autoreleaseNow): Swap with an empty set instead of
        building up an array and then calling clear(). This is slightly more
        efficient and the code is cleaner.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21468 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoLayoutTests:
justing [Mon, 14 May 2007 21:37:04 +0000 (21:37 +0000)]

        Reviewed by darin

        <rdar://problem/5136696> Selection is removed when focusing a node in another frame

        Demonstrates bug:
        * editing/selection/5136696-expected.checksum: Added.
        * editing/selection/5136696-expected.png: Added.
        * editing/selection/5136696-expected.txt: Added.
        * editing/selection/5136696.html: Added.


        Reviewed by darin

        <rdar://problem/5136696> Selection is removed when focusing a node in another frame
        <rdar://problem/5192388> WordPress: Can't insert links/images

        Don't clear a selection when changing focus if the new
        focus node is in a different frame than the old selection.

        * dom/Document.cpp:
        (WebCore::Document::setFocusedNode): Moved code to clear
        selections to the FocusController, since it is Page level
        and will know about the frame/document that contained the
        old selection. Moved code to call shouldEndEditing to the
        FocusController because selections are now cleared before
        calling Document::setFocusedNode and shouldEndEditing must
        be called before selections are cleared.
        * dom/Element.cpp:
        (WebCore::Element::blur): Call FocusController::setFocusNode
        instead of Document::setFocusNode, if possible.
        * page/EventHandler.cpp:
        (WebCore::EventHandler::dispatchMouseEvent): Ditto.
        * page/FocusController.cpp:
        (WebCore::relinquishesEditingFocus): Moved from Document.cpp.
        (WebCore::clearSelectionIfNeeded): Ditto.  Don't clear if
        the old selection is in a frame that's different than the
        one that contains the new focus node.
        (WebCore::FocusController::setFocusedNode): Moved code
        here from Document::setFocusedNode.
        * page/Frame.cpp:
        (WebCore::Frame::setFocusedNodeIfNeeded): Call

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21467 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoWebCore:
ggaren [Mon, 14 May 2007 21:29:15 +0000 (21:29 +0000)]

        Reviewed by Darin Adler.

        Next step toward fixing <rdar://problem/5055182> The back cache has no
        global cap.

        Created a PageCache singleton. Right now, it only handles the timer-based
        deallocation of cached pages. I plan to add HistoryItem->CachedPage tracking
        and eviction logic in later patches.

        Updated for renames and file changes:

        * WebCore.exp:
        * WebCore.xcodeproj/project.pbxproj:
        * history/BackForwardList.cpp:

        Removed global timed release code, moved into new PageCache class, since
        it's eventually going to be in charge of deciding when things are released:

        * history/HistoryItem.cpp:
        * history/HistoryItem.h:

        Folded into PageCache:

        * history/HistoryItemTimer.cpp: Removed.
        * history/HistoryItemTimer.h: Removed.

        New singleton. Right now, it just handles timed release of cached pages:

        * history/PageCache.cpp: Added.
        * history/PageCache.h: Added.


        Reviewed by Darin Adler.

        Updated for WebCore move to PageCache singleton. Also removed some WebCore::
        prefixes and replaced them with 'using namespace WebCore'.

        * History/WebHistoryItem.mm:
        (+[WebHistoryItem _releaseAllPendingPageCaches]):
        (-[WebWindowWatcher windowWillClose:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21464 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
adele [Mon, 14 May 2007 21:16:38 +0000 (21:16 +0000)]
    Reviewed by Darin.

        Fix for <rdar://problem/5200081> Repro crash in PopupMenu::show clicking popup menu button after reload

        * platform/mac/PopupMenuMac.mm: (WebCore::PopupMenu::show): Added call to dismissPopup so the cell gets unattached properly.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21463 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
adele [Mon, 14 May 2007 20:57:15 +0000 (20:57 +0000)]
    Reviewed by Darin.

        Remove incorrect assertion.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21462 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoAdd another 1000 tests that we pass for the Qt build.
lars [Mon, 14 May 2007 20:43:34 +0000 (20:43 +0000)]
Add another 1000 tests that we pass for the Qt build.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21461 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Add another 300 test cases, that are ok on the Qt build.
lars [Mon, 14 May 2007 19:20:11 +0000 (19:20 +0000)]
    Add another 300 test cases, that are ok on the Qt build.
        Add a few more tests that need examination to the Skipped list.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21460 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Geoff.
andersca [Mon, 14 May 2007 18:53:14 +0000 (18:53 +0000)]
    Reviewed by Geoff.

        <rdar://problem/5172056> Regression: cannot open txt file, get broken image

        Never let plug-ins override text/plain.

        * dom/DOMImplementation.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21459 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFix dependcy tracking in project files.
zack [Mon, 14 May 2007 14:39:41 +0000 (14:39 +0000)]
Fix dependcy tracking in project files.

Add dependpath to project to get better dependency tracking for header
files in the qmake project files.


git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21456 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Zack
lars [Mon, 14 May 2007 13:41:06 +0000 (13:41 +0000)]
    Reviewed by Zack

        Updates after Maciej's frame change.

This gets things to run again. The #ifdef PLATFORM(QT)
in the FrameLoader should not be needed (and adding the line
inside on the mac doesn't seem to trigger any failures there).
I'm submitting with it to be on the safe side for now, since I
really need our tests to run again.

Btw, post-commit-linux-qt should show up green on with this submit :))

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21455 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Antti.
darin [Mon, 14 May 2007 13:30:41 +0000 (13:30 +0000)]
    Reviewed by Antti.

        - fix <rdar://problem/3109180> VIP: after go back, page jumps after loading even if I scrolled while loading

        * history/CachedPage.cpp: (WebCore::CachedPage::clear): Call clearFrame rather than
        clearPart -- the "part" term was leftover prehistoric terminology.

        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::open): Call setWasScrolledByUser(false) on the page, so that we
        can detect any user scrolling that happens during the "go back" process.
        (WebCore::FrameLoader::restoreScrollPositionAndViewState): Don't scroll if the view's
        wasScrolledByUser flag is set.
        (WebCore::FrameLoader::loadItem): Call setWasScrolledByUser(false) false when we do
        a new "load" (actually just a scroll) using the same frame. Also moved the line of code that
        sets m_currentHistoryItem to make this code path mroe similar to recursiveGoToItem.
        (WebCore::FrameLoader::recursiveGoToItem): Ditto.

        * page/Frame.cpp:
        (WebCore::Frame::~Frame): Call clearFrame rather than stting m_frame directly.
        (WebCore::Frame::sendScrollEvent): Call setWasScrolledByUser(true) on FrameView. The FrameView
        itself knows to ignore this if the scrolling is being done programmatically instead of by the user.

        * page/FrameView.h: Added wasScrolledByUser and setWasScrolledByUser. Also removed unneeded
        include and class forward declarations, renamed clearPart to clearFrame, removed uneeded function
        haveDelayedLayoutScheduled, and removed unneeded friend declarations.
        * page/FrameView.cpp:
        (WebCore::FrameViewPrivate::FrameViewPrivate): Added m_inProgrammaticScroll and m_wasScrolledByUser.
        (WebCore::FrameViewPrivate::reset): Set m_wasScrolledByUser to false.
        (WebCore::FrameView::clearFrame): Renamed from clearPart.
        (WebCore::FrameView::scrollRectIntoViewRecursively): Added code to set the m_inProgrammaticScroll flag.
        (WebCore::FrameView::setContentsPos): Ditto.
        (WebCore::FrameView::wasScrolledByUser): Added.
        (WebCore::FrameView::setWasScrolledByUser): Added. Does nothing if called when m_inProgrammaticScroll
        is true.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21454 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago2007-05-14 Kimmo Kinnunen <kimmok@iki.fi>
darin [Mon, 14 May 2007 13:26:44 +0000 (13:26 +0000)]
2007-05-14  Kimmo Kinnunen  <kimmok@iki.fi>

        Tests for http://bugs.webkit.org/show_bug.cgi?id=10878
        Bug 10878: Incorrect decompilation for "4..x"

        Test all the different forms of expressions that can contain
        a dot following a number literal.
        Includes new tests for PrefixDotNode, PostfixDotNode, DeleteDotNode,
        AssignDotNode and FunctionCallParenDotNode.

        * fast/js/resources/toString-number-dot-expr.js:
        * fast/js/toString-number-dot-expr-expected.txt: Updated.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21453 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Antti.
darin [Mon, 14 May 2007 13:12:54 +0000 (13:12 +0000)]
    Reviewed by Antti.

        - <rdar://problem/4865655> REGRESSION: find function does not work for the character ß

        The trick was to fold case consistently in the CircularSearchBuffer class.

        The problem was that we were folding the search target string with full case folding and
        the characters of the text we were searching with simple case folding. We now use full
        case folding in both cases.

        * editing/TextIterator.cpp:
        (WebCore::CircularSearchBuffer::CircularSearchBuffer): Changed the buffer to be a
        Vector<UChar>. Added a Vector<bool> to record which characters were starts of characters
        in the text we are searching. Changed the cursor to be an int rather than a pointer.
        (WebCore::CircularSearchBuffer::append): New function that knows how to append multiple
        characters when folding case turns one character into many -- uses the full foldCase
        rather than the simplified single-character foldCase. Also removed the append function
        that did multiple characters at once; that was part of an optimization we don't need.
        (WebCore::CircularSearchBuffer::isMatch): Instead of asserting that the buffer is full,
        changed this to return false if it's not full, which lets us fill up the buffer with
        the normal loop. Also return false if the start of the buffer is half-way through a
        case-folded character and updated this since m_cursor is now an integer.
        (WebCore::CircularSearchBuffer::length): This returns the length in non-case-folded
        characters of the characters that were appended, which may not be the same as the length
        of the original target string. To do that we have to total up the character start
        booleans for the current state of the circular buffer.
        (WebCore::findPlainText): Streamlined the logic in a few ways: 1) Moved more of the
        return value setup to the top of the function so it could be shared with the early
        error exit. 2) Removed the separate "fill the buffer" logic, which allowed me to
        remove some of the functions from CircularSearchBuffer and eliminated the need for
        two nested loops. 3) Replaced the CharacterIterator and bool that were used to record
        the location we found with integers matchStart and matchLength. 4) Got rid of the use
        of goto since we don't have two nested loops any more.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21452 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
andrew [Mon, 14 May 2007 12:16:41 +0000 (12:16 +0000)]
    Reviewed by Darin.

        Fix http://bugs.webkit.org/show_bug.cgi?id=13275
        Safari crashed when trying to login using the windows live contact control

        Check for cleared windowObj in JSAbstractEventListener::handleEvent

        * bindings/js/kjs_events.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21451 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoSkip a test that currently hangs
lars [Mon, 14 May 2007 07:17:04 +0000 (07:17 +0000)]
Skip a test that currently hangs

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21450 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
ap [Mon, 14 May 2007 04:45:44 +0000 (04:45 +0000)]
    Reviewed by Darin.

        JavaScript Form Validation issue on Banco do Brasil site

        Test: fast/forms/input-setvalue-selection.html

        * html/HTMLInputElement.cpp:
        (WebCore::HTMLInputElement::setValue): In other browsers, changing textInput.value
        always moves the insertion point past the text. In 10.4.9 WebKit, this only worked
        for inactive controls, and it never worked in ToT.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21448 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Mitz
beidson [Sun, 13 May 2007 15:28:01 +0000 (15:28 +0000)]
    Reviewed by Mitz


        Since HTMLInputElements register themselves with their Document for special handling, if their document
        ever changes then bad things can happen.

        Unfortunately, with adoptNode, the document can change.  This has been a long standing crasher with all
        input fields but was given recent high visibility by r21430

        The solution is to add functionality for a Node to react to its owner document changing via a pair of
        virtual callbacks.  Then we can tap into those callbacks in HTMLInputElement.

        No layout test possible until http://bugs.webkit.org/show_bug.cgi?id=13672 is resolved

        * dom/Node.cpp:
        (WebCore::Node::setDocument): Call willMoveToNewOwnerDocument and didMoveToNewOwnerDocument

        * dom/Node.h:
        (WebCore::Node::willMoveToNewOwnerDocument): Virtual callback to a node for before an owner document change
        (WebCore::Node::didMoveToNewOwnerDocument): Virtual callback to a node for after an owner document change

        * html/HTMLInputElement.cpp:
        (WebCore::HTMLInputElement::~HTMLInputElement): Unconditionally unregister both for both form state
          and the cache restoration callback
        (WebCore::HTMLInputElement::setInputType): Change registration with the document only if we're in one
        (WebCore::HTMLInputElement::attach): Don't bother registering for the callback here - that will be
          handled by setInputType() or by a change in owner document
        (WebCore::HTMLInputElement::willMoveToNewOwnerDocument): Unregister with the old owner document
        (WebCore::HTMLInputElement::didMoveToNewOwnerDocument): Register with the new owner document
        * html/HTMLInputElement.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21447 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * platform/mac/TextBreakIteratorInternalICUMac.mm: Fix comment typo.
darin [Sun, 13 May 2007 14:05:11 +0000 (14:05 +0000)]
    * platform/mac/TextBreakIteratorInternalICUMac.mm: Fix comment typo.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21446 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
rwlbuis [Sun, 13 May 2007 14:00:18 +0000 (14:00 +0000)]
    Reviewed by Darin.

        DOM3 keyIdentifier text is incorrect

        Use 4 digits for the Unicode based key identifiers.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21445 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - one more retain/release for a tiny bit more robustness
darin [Sun, 13 May 2007 13:21:18 +0000 (13:21 +0000)]
    - one more retain/release for a tiny bit more robustness

        * WebView/WebPDFView.mm:
        (-[WebPDFView _updatePreferences:]): [prefs release]
        (-[WebPDFView _updatePreferencesSoon]): [prefs retain]

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21444 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Geoff.
darin [Sun, 13 May 2007 13:16:43 +0000 (13:16 +0000)]
    Reviewed by Geoff.

        - fix <rdar://problem/5188400> Webkit crashes going back from PDF at perl.org site

        * WebView/WebPDFView.h: Replace _updatePreferencesTimer with _willUpdatePreferencesSoon BOOL.
        Also remove unneeded @public that gives other classes access to our dataSource member.

        * WebView/WebPDFView.mm: Rearrange top of file a bit, remove forward declaration of the
        _cancelUpdatePreferencesTimer method.
        (-[WebPDFView dealloc]): Removed call to _cancelUpdatePreferencesTimer.
        (-[WebPDFView _updatePreferencesNow:]): Added WebPreferences parameter. This sidesteps problems
        where the dataSource is no longer present by not looking at the dataSource field at all.
        Also removed the call to _cancelUpdatePreferencesTimer, added code to set _willUpdatePreferencesSoon
        to NO and added a release to balance a retain I did in _updatePreferencesSoon.
        (-[WebPDFView _updatePreferencesSoon]): Changed to use performSelectorAfterDelay instead of
        an NSTimer. Pass in the preferences object as a parameter, since we might not be able to get to
        the dataSource when the timer fires.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21443 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Antti.
darin [Sun, 13 May 2007 12:55:21 +0000 (12:55 +0000)]
    Reviewed by Antti.

        - fix <rdar://problem/5174903> Safari crashes opening the attached webarchive (made by OmniWeb)

        * loader/icon/IconLoader.cpp:
        (WebCore::IconLoader::didReceiveResponse): Check ResourceHandle for nil before calling url() on it.
        (WebCore::IconLoader::didFail): Ditto.
        (WebCore::IconLoader::didFinishLoading): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21442 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Oliver.
rwlbuis [Sun, 13 May 2007 07:25:59 +0000 (07:25 +0000)]
    Reviewed by Oliver.

        Cleanup, get rid of the SVGStyledElement::view() call.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21441 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoAdd layout tests for residual style fix.
hyatt [Sun, 13 May 2007 06:02:33 +0000 (06:02 +0000)]
Add layout tests for residual style fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21440 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Hyatt.
hyatt [Sun, 13 May 2007 06:01:41 +0000 (06:01 +0000)]
    Reviewed by Hyatt.

        - fix http://bugs.webkit.org/show_bug.cgi?id=13680
          Incomplete repaint of focus ring around continuation

        Test: fast/repaint/continuation-after-outline.html

        * rendering/RenderFlow.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@21439 268f45cc-cd09-0410-ab3c-d52691b4dbfc