WebKit-https.git
15 years ago Reviewed by John
kocienda [Tue, 25 Jan 2005 21:21:08 +0000 (21:21 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3954710> Mail crashed while editing signatures - NodeImpl::isBlockFlow

        The fix is more general than for this one bug, and may work to fix many crashers. The problem
        is that the ReplaceSelectionCommand never checked whether its starting selection is empty. If
        it is, then we need to bail before doing the work of the command, which we need to deref the
        start and end points of the selection in order to do its work. I think you can see the crash
        potential.

        * khtml/editing/htmlediting.cpp:
        (khtml::ReplaceSelectionCommand::doApply): Assert selection is not empty.
        * kwq/WebCoreBridge.mm:
        (partHasSelection): New helper function to test that bridge has a part with a selection.

        Use new helper function to test part and selection; return from these function if this test fails.

        (-[WebCoreBridge rangeByExpandingSelectionWithGranularity:])
        (-[WebCoreBridge rangeByAlteringCurrentSelection:direction:granularity:])
        (-[WebCoreBridge alterCurrentSelection:direction:granularity:])
        (-[WebCoreBridge rangeByAlteringCurrentSelection:verticalDistance:])
        (-[WebCoreBridge alterCurrentSelection:verticalDistance:])
        (-[WebCoreBridge documentFragmentWithText:])
        (-[WebCoreBridge replaceSelectionWithFragment:selectReplacement:smartReplace:])
        (-[WebCoreBridge insertLineBreak])
        (-[WebCoreBridge insertParagraphSeparator])
        (-[WebCoreBridge insertParagraphSeparatorInQuotedContent])
        (-[WebCoreBridge insertText:selectInsertedText:])
        (-[WebCoreBridge deleteSelectionWithSmartDelete:])
        (-[WebCoreBridge ensureSelectionVisible])

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8446 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
kdecker [Tue, 25 Jan 2005 03:56:38 +0000 (03:56 +0000)]
    Reviewed by Darin.

Fixed <rdar://problem/3932374> REGRESSION: 'ReferenceError - Can't find variable' JavaScript error at webxpress.fidelity.com

        * khtml/html/html_miscimpl.h: Missing virtual identifiers for namedItem() and nextNamedItem() exposed a flaw that prevented finding variables in form [HTMLFormCollectionImpl] scopes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8445 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed Panther build (missing ifdef)
mjs [Tue, 25 Jan 2005 01:18:06 +0000 (01:18 +0000)]
    Fixed Panther build (missing ifdef)

        * WebView.subproj/WebDefaultContextMenuDelegate.m:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8444 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
sullivan [Tue, 25 Jan 2005 00:46:15 +0000 (00:46 +0000)]
    Reviewed by Darin.

        - WebKit part of fix for <rdar://problem/3960231> Text context menu in WebKit needs Spotlight and Google items

        * WebView.subproj/WebDefaultContextMenuDelegate.m:
        (localizedMenuTitleFromAppKit):
        new function, gets localized string from AppKit so we can avoid adding localized
        strings to WebKit at this late date in Tiger. Returns the non-localized English
        string if we can't find the localized string in AppKit.
        (-[WebDefaultUIDelegate menuItemWithTag:]):
        handle the two new menu items by tag name
        (-[WebDefaultUIDelegate contextMenuItemsForElement:]):
        add menu items for Search in Google and Search in Spotlight to selected-text menu item
        for the noneditable case
        (-[WebDefaultUIDelegate editingContextMenuItemsForElement:]):
        add menu items for Search in Google and Search in Spotlight to selected-text menu item
        for the editable case

        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView _searchWithGoogleFromMenu:]):
        implement this menu item action method, using same method name and implementation as NSTextView
        (-[WebHTMLView _searchWithSpotlightFromMenu:]):
        implement this menu item action method, using same basic implementation as NSTextView
        (-[WebHTMLView validateUserInterfaceItem:]):
        validate new menu items

        * WebView.subproj/WebUIDelegatePrivate.h:
        define new tags for new menu items

        * English.lproj/StringsNotToBeLocalized.txt:
        updated for these changes

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8443 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by kocienda.
vicki [Tue, 25 Jan 2005 00:35:19 +0000 (00:35 +0000)]
    Reviewed by kocienda.

- fix <rdar://problem/3810661> computed style not handled for Apple extensions

        * khtml/css/css_computedstyle.cpp: (DOM::):
        (DOM::CSSComputedStyleDeclarationImpl::getPropertyCSSValue): add -apple-line-clamp and -apple-text-size-adjust

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8442 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Kevin.
darin [Mon, 24 Jan 2005 22:21:48 +0000 (22:21 +0000)]
    Reviewed by Kevin.

        - fixed <rdar://problem/3969884> REGRESSION (179-180): Typing password not echoed as "bullets" at paypal.com when field is first focused

        * kwq/KWQTextField.mm: (-[KWQTextFieldController initWithTextField:QLineEdit:]):
        Sizing up the form editor, which fixes things for normal fields and search fields,
        screws things up for password fields. Eventually, I'll need to figure out why this is
        and file a bug so that the AppKit team fixes it. In the mean time, just check for the
        case of the secure text field, and don't size up in that case.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8441 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Hyatt
kocienda [Mon, 24 Jan 2005 22:20:02 +0000 (22:20 +0000)]
    Reviewed by Hyatt

        Fix for this bug:

        <rdar://problem/3963560> 8A354: Inserting Japanese text to the line head causes line break

        * khtml/editing/htmlediting.cpp:
        (khtml::ReplaceSelectionCommand::doApply): Add one more case where we merge content into the
        existing line. This covers the situation see using the steps to reproduce this bug.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8440 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
mjs [Mon, 24 Jan 2005 22:06:20 +0000 (22:06 +0000)]
    Reviewed by Darin.

<rdar://problem/3967572> Editing should split elements before removing style

        * khtml/editing/htmlediting.cpp:
        (khtml::CompositeEditCommand::splitElement):
        (khtml::CompositeEditCommand::mergeIdenticalElements):
        (khtml::CompositeEditCommand::wrapContentsInDummySpan):
        (khtml::CompositeEditCommand::splitTextNodeContainingElement):
        (khtml::ApplyStyleCommand::applyInlineStyle):
        (khtml::ApplyStyleCommand::splitTextAtStartIfNeeded):
        (khtml::ApplyStyleCommand::splitTextAtEndIfNeeded):
        (khtml::ApplyStyleCommand::splitTextElementAtStartIfNeeded):
        (khtml::ApplyStyleCommand::splitTextElementAtEndIfNeeded):
        (khtml::areIdenticalElements):
        (khtml::ApplyStyleCommand::mergeStartWithPreviousIfIdentical):
        (khtml::ApplyStyleCommand::mergeEndWithNextIfIdentical):
        (khtml::ApplyStyleCommand::cleanUpEmptyStyleSpans):
        (khtml::ApplyStyleCommand::addInlineStyleIfNeeded):
        (khtml::InsertParagraphSeparatorCommand::doApply):
        (khtml::InsertParagraphSeparatorInQuotedContentCommand::doApply):
        (khtml::InsertTextCommand::prepareForTextInsertion):
        (khtml::SplitTextNodeCommand::doUnapply):
        (khtml::SplitElementCommand::SplitElementCommand):
        (khtml::SplitElementCommand::~SplitElementCommand):
        (khtml::SplitElementCommand::doApply):
        (khtml::SplitElementCommand::doUnapply):
        (khtml::MergeIdenticalElementsCommand::MergeIdenticalElementsCommand):
        (khtml::MergeIdenticalElementsCommand::~MergeIdenticalElementsCommand):
        (khtml::MergeIdenticalElementsCommand::doApply):
        (khtml::MergeIdenticalElementsCommand::doUnapply):
        (khtml::WrapContentsInDummySpanCommand::WrapContentsInDummySpanCommand):
        (khtml::WrapContentsInDummySpanCommand::~WrapContentsInDummySpanCommand):
        (khtml::WrapContentsInDummySpanCommand::doApply):
        (khtml::WrapContentsInDummySpanCommand::doUnapply):
        (khtml::SplitTextNodeContainingElementCommand::SplitTextNodeContainingElementCommand):
        (khtml::SplitTextNodeContainingElementCommand::~SplitTextNodeContainingElementCommand):
        (khtml::SplitTextNodeContainingElementCommand::doApply):
        * khtml/editing/htmlediting.h:

New layout test that shows the side benefit of this.

* layout-tests/editing/style/unbold-in-bold-expected.txt: Added.
        * layout-tests/editing/style/unbold-in-bold.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8439 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * WebView.subproj/WebDataSource.m: (+[WebDataSource _repTypesAllowImageTypeOm...
darin [Mon, 24 Jan 2005 21:07:10 +0000 (21:07 +0000)]
    * WebView.subproj/WebDataSource.m: (+[WebDataSource _repTypesAllowImageTypeOmission:]):
        Do the same check as for view types, so the representation types are consistent.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8438 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John.
darin [Mon, 24 Jan 2005 20:48:26 +0000 (20:48 +0000)]
    Reviewed by John.

        - fixed <rdar://problem/3933435> -[DOMRange cloneContents] sometimes alters the original DOMRange

        * khtml/xml/dom2_rangeimpl.cpp: (DOM::RangeImpl::processContents): Fixed code that was collapsing
        the range after processing the contents. That's right for extract and delete, but not clone.

        - fixed <rdar://problem/3714184> inline input works incorrectly when I type Korean slowly in textfield in form on www.google.co.kr
        - fixed <rdar://problem/3193848> Can't enter SSN to sign up for Chase visa online account (changing focus inside key press handler)
        - fixed <rdar://problem/3874683> REGRESSION (161-162): Crash after typing command-Z to undo after pasting text with newline in it into a text field <input type=text>

        * kwq/KWQLineEdit.mm:
        (QLineEdit::text): Changed to call a new string method on the controller instead of calling
        stringValue, since stringValue has a bad side effect of ending inline input.
        (QLineEdit::selectAll): Changed to only call selectText: when the field is already selected.
        When you are giving the field focus, it automatically gets all selected, and this change
        is required in the case where we defer the focus change.

        * kwq/KWQTextField.h: Added a new string method to the controller.
        * kwq/KWQTextField.mm:
        (-[KWQTextFieldController setMaximumLength:]): Call the new string method instead of using stringValue.
        (-[KWQTextFieldController controlTextDidChange:]): Removed the code to truncate at the first CR or LF.
        That's now handled in the shouldChangeTextInRange method instead.
        (-[KWQTextFieldController textView:shouldHandleEvent:]): Add new logic to defer responder changes until
        after the event is handled. This makes sure the key gets into the field that's already handling it at
        rather than ending up in the newly-focused field if the key press handler focuses a different element.
        (-[KWQTextFieldController textView:didHandleEvent:]): Turn off deferral here, now that the key press
        has been handled.
        (-[KWQTextFieldController string]): Added. Calls stringValue only if there's no editor. If there is
        an editor, calls string on that instead.
        (-[KWQTextFieldController textView:shouldChangeTextInRange:replacementString:]): Added. Called by
        the three field subclasses. Truncates incoming strings at the first CR or LF character. This works
        properly with Undo, and the old technique did not.
        (-[KWQTextFieldController preprocessString:]): Added. Helper used for all the setStringValue methods
        so they all truncate at the first CR or LF as well as truncating to the appropriate length.
        (-[KWQTextField setStringValue:]): Changed to call the controller.
        (-[KWQTextField textView:shouldChangeTextInRange:replacementString:]): Added. Calls the controller.
        (-[KWQSecureTextField setStringValue:]): More of the same.
        (-[KWQSecureTextField textView:shouldChangeTextInRange:replacementString:]): Ditto.
        (-[KWQSearchField setStringValue:]): Ditto.
        (-[KWQSearchField textView:shouldChangeTextInRange:replacementString:]): Ditto.

        * kwq/KWQWidget.h: Added the new setDeferFirstResponderChanges function.
        * kwq/KWQWidget.mm:
        (QWidget::~QWidget): Clear out the deferredFirstResponder global just in case we are destroyed
        while we "have the ball".
        (QWidget::hasFocus): If we are in the deferred mode, use the global to decide who to say is focused.
        (QWidget::setFocus): If we are in the deferred mode, set the global instead of making something be
        the first responder.
        (QWidget::setDeferFirstResponderChanges): Set the boolean. If the boolean is being cleared, then
        get the deferred first responder and call setFocus on it again, which will do the work we didn't
        want to do earlier.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8437 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John.
darin [Mon, 24 Jan 2005 20:48:16 +0000 (20:48 +0000)]
    Reviewed by John.

        - fixed <rdar://problem/3791158> REGRESSION (Mail): copyFont: and pasteFont: copy and paste only the NSFont, not other attributes

        * WebView.subproj/WebHTMLView.m: (-[WebHTMLView _styleFromFontAttributes:]): Added the last
        few loose ends here: strikethrough and underline.

        - fixed <rdar://problem/3967393> add a user default that lets you turn off WebKit PDF support

        * WebView.subproj/WebFrameView.m: (+[WebFrameView _viewTypesAllowImageTypeOmission:]):
        Leave the PDF-handling classes out of the dictionary if the secret default is set.

        * English.lproj/StringsNotToBeLocalized.txt: Updated for recent changes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8436 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Mon, 24 Jan 2005 18:28:43 +0000 (18:28 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3964350> Deleting all content and typing in a message gets one letter, then beeps

        This is a problem with the way we turn key events, when a user starts typing, into editing commands.
        Right now, that initial delete key in an empty window is turned into an editing command, and much
        editing code runs that need not run. What's more, this is confusing the selection machinery. The
        solution is to detect the "delete in an empty window" case and don't turn this into an editing
        command at all.

        * khtml/editing/htmlediting.cpp:
        (khtml::TypingCommand::deleteKeyPressed)

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8435 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3964286> REGRESSION (178-179): menus and "go to old...
adele [Mon, 24 Jan 2005 01:09:12 +0000 (01:09 +0000)]
    Fixed <rdar://problem/3964286> REGRESSION (178-179): menus and "go to old site" link at Ameritrade's new site don't work

        Removed check for document from checkCompleted.  In this case, WebCore never knew that a WebImageView was complete because
        it fell into this document check case in checkCompleted.
        Added check for document to tokenizerProcessedData (see <rdar://problem/3807144>)

        Reviewed by Darin.

        * khtml/khtml_part.cpp:
        (KHTMLPart::stop):
        (KHTMLPart::checkCompleted):
        * khtml/khtml_part.h:
        * kwq/KWQKHTMLPart.mm: (KWQKHTMLPart::tokenizerProcessedData):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8434 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3759399> Javascript / Liveconnect problems ((event handler...
rjw [Sat, 22 Jan 2005 00:48:28 +0000 (00:48 +0000)]
Fixed <rdar://problem/3759399> Javascript / Liveconnect problems ((event handler):Undefined value)

Java applets specified with <object> or <embed> weren't scriptable.  Now they are.

        Reviewed by Chris.

        * khtml/html/html_objectimpl.cpp:
        (HTMLEmbedElementImpl::getEmbedInstance):
        (HTMLObjectElementImpl::getObjectInstance):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8433 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Sat, 22 Jan 2005 00:16:54 +0000 (00:16 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3959464> REGRESSION (Mail): Insertion point goes back to beginning of document after deleting

        * khtml/editing/htmlediting.cpp:
        (khtml::DeleteSelectionCommand::handleGeneralDelete): Add special case to handle retaining a fully-selected block.
        This fixes the bug.
        * layout-tests/editing/deleting/delete-3959464-fix-expected.txt: Added.
        * layout-tests/editing/deleting/delete-3959464-fix.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8432 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3966998> REGRESSION(179-TOT) clicking on gmail message brings...
rjw [Fri, 21 Jan 2005 23:01:18 +0000 (23:01 +0000)]
Fixed <rdar://problem/3966998> REGRESSION(179-TOT) clicking on gmail message brings me to blank screen

getElementById() sometimes returns Undefined() instead of Null().

        Reviewed by Hyatt.

        * khtml/ecma/kjs_dom.cpp:
        (DOMDocumentProtoFunc::tryCall):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8430 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix for 3773809, make sure that overflow regions never end up at an invalid scroll...
hyatt [Fri, 21 Jan 2005 22:45:49 +0000 (22:45 +0000)]
Fix for 3773809, make sure that overflow regions never end up at an invalid scroll offset because of a layout change.

        Reviewed by kocienda

        * khtml/rendering/render_layer.cpp:
        (RenderLayer::updateScrollInfoAfterLayout):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8429 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Kristin Forster.
darin [Fri, 21 Jan 2005 22:43:54 +0000 (22:43 +0000)]
    Reviewed by Kristin Forster.

        - fixed <rdar://problem/3964972> update _initWithCGSEvent:eventRef: call in mouse moved workaround (breaks cursors in Carbon WebKit applications)

        * Carbon.subproj/HIWebView.m: (MouseMoved): Instead of munging the event record's window number directly
        before calling _initWithCGSEvent, on Tiger we call _eventRelativeToWindow on the event after creating it.
        Also added a check so that with Macromedia Contribute's workaround in place we don't do anything at all
        to the event. Tested with both Contribute and CarbonWeb.

        * English.lproj/StringsNotToBeLocalized.txt: Update for recent changes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8428 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix for 3966349, hang loading page. Make sure that loops using popOneBlock properly...
hyatt [Fri, 21 Jan 2005 21:45:20 +0000 (21:45 +0000)]
Fix for 3966349, hang loading page.  Make sure that loops using popOneBlock properly check for a null
blockStack to avoid an infinite loop.

        Reviewed by kocienda

        * khtml/html/htmlparser.cpp:
        (KHTMLParser::parseToken):
        (KHTMLParser::insertNode):
        (KHTMLParser::popInlineBlocks):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8427 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Fri, 21 Jan 2005 18:55:53 +0000 (18:55 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3957204> Mail crashed when sending a message in ~InsertParagraphSeparatorInQuotedContentCommand

        * khtml/editing/htmlediting.cpp:
        (khtml::InsertParagraphSeparatorInQuotedContentCommand::InsertParagraphSeparatorInQuotedContentCommand):
        InsertParagraphSeparatorInQuotedContentCommand has this member variable: DOM::ElementImpl *m_breakNode;
        I failed to initialize this variable to null in the constructor, and there is an uncommonly-traveled
        code path which does not set this variable to something good. In the destructor, we check for null,
        and deref if non-null. Obvious "BOOM" potential. Fixed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8426 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Fri, 21 Jan 2005 17:29:21 +0000 (17:29 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3966311> REGRESSION (Mail): Hitting return makes space character disappear

        * khtml/editing/htmlediting.cpp:
        (khtml::InsertParagraphSeparatorCommand::doApply): When writing the code to insert a block
        in response to the return key, I did not write code to cover the case described in the bug.
        Now I have.
        * layout-tests/editing/inserting/insert-div-025-expected.txt: Added.
        * layout-tests/editing/inserting/insert-div-025.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8425 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3815672> REGRESSION (Mail): Japanese text cannot be made bold
rjw [Fri, 21 Jan 2005 02:25:17 +0000 (02:25 +0000)]
Fixed <rdar://problem/3815672> REGRESSION (Mail): Japanese text cannot be made bold

The AppKit's font substitution API doesn't match font traits!
It only find fonts that contain the appropriate glyphs.  This
patch attempts to find the best variation within a family.

Also, made unrelated change to return printer/screen version of
substituted font.

        Reviewed by Maciej.

        * WebCoreSupport.subproj/WebTextRenderer.m:
        (-[WebTextRenderer _substituteFontForString:families:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8424 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Ken.
mjs [Fri, 21 Jan 2005 01:50:02 +0000 (01:50 +0000)]
    Reviewed by Ken.

<rdar://problem/3965196> security fix for javascript: exploit missed one case (already fixed in updates)

* khtml/ecma/kjs_window.cpp:
        (WindowFunc::tryCall): correct mistake in earlier fix for the following bug, caught by Adele:

<rdar://problem/3790449> REGRESSION (Mail): underline behavior is flaky because of how CSS handles text-decoration

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8423 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Ken.
mjs [Fri, 21 Jan 2005 01:36:41 +0000 (01:36 +0000)]
    Reviewed by Ken.

<rdar://problem/3965466> editing needs to insert text before applying typing style

        * khtml/editing/htmlediting.cpp:
        (khtml::ApplyStyleCommand::addInlineStyleIfNeeded): Remove a FIXME comment.
        (khtml::InsertLineBreakCommand::doApply): Apply style to the BR
node after inserting it.
        (khtml::InsertTextCommand::prepareForTextInsertion): Don't try to
apply style to any new nodes created before inserting.
        (khtml::InsertTextCommand::input): Apply style to the inserted range
after doing the text insert.
        * khtml/editing/htmlediting.h:

This change leads to improved results on one of the layout tests:

* layout-tests/editing/inserting/insert-div-024-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8422 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by me, changes by Han Ming Ong.
darin [Fri, 21 Jan 2005 01:22:08 +0000 (01:22 +0000)]
    Reviewed by me, changes by Han Ming Ong.

        - <rdar://problem/3964302> SWB: A few files need to be updated to be compilable under GCC 4.0

        * bindings/objc/WebScriptObjectPrivate.h: Make members public.
        * kjs/lookup.h: Change "value.h" to "object.h" because we need KJS::Object to compile a template.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8421 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agomove the Safari-180 release-marker in the ChangeLog, since the fix for 3786659 will...
vicki [Fri, 21 Jan 2005 00:58:23 +0000 (00:58 +0000)]
move the Safari-180 release-marker in the ChangeLog, since the fix for 3786659 will be included in Safari-180

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8420 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Fri, 21 Jan 2005 00:50:30 +0000 (00:50 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3786659> REGRESSION (Mail): editable WebViews don't work with "size up" and "size down" NSFontManager changes

        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView _addToStyle:fontA:fontB:]): This is the WebKit side of the fix. Replaced unimplemented code blocks
        with FIXME's in them for make bigger and make smaller with real working code.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8419 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3964634> undefined property value from binding seems to evalua...
rjw [Fri, 21 Jan 2005 00:41:59 +0000 (00:41 +0000)]
Fixed <rdar://problem/3964634> undefined property value from binding seems to evaluate to true in an if statement

The comprehensive fix for this problem requires new API, as described in 3965326.  However,
given that we can't add new API at this point, the 'ObjcFallbackObjectImp' will behave
like and Undefined object if  invokeUndefinedMethodFromWebScript:withArguments: isn't
implemented on the bound object.

        Reviewed by Chris.

        * bindings/objc/objc_runtime.h:
        * bindings/objc/objc_runtime.mm:
        (ObjcFallbackObjectImp::type):
        (ObjcFallbackObjectImp::implementsCall):
        (ObjcFallbackObjectImp::toBoolean):
        * bindings/testbindings.mm:
        (+[MyFirstInterface isSelectorExcludedFromWebScript:]):
        (+[MyFirstInterface isKeyExcludedFromWebScript:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8418 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFinished my thought in the changelog. I left a comment half-written earlier.
kocienda [Fri, 21 Jan 2005 00:01:07 +0000 (00:01 +0000)]
Finished my thought in the changelog. I left a comment half-written earlier.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8417 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoVersioning for TOT, Safari 2.0 (180+).
vicki [Thu, 20 Jan 2005 23:47:42 +0000 (23:47 +0000)]
Versioning for TOT, Safari 2.0 (180+).

The tree is open!

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8416 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoSafari-180 stamp
vicki [Thu, 20 Jan 2005 23:33:44 +0000 (23:33 +0000)]
Safari-180 stamp

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8415 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Thu, 20 Jan 2005 22:02:17 +0000 (22:02 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3964646> REGRESSION (179-180): Typing space at end of line makes following paragraph disappear

        * khtml/editing/htmlediting.cpp:
        (khtml::CompositeEditCommand::removeBlockPlaceholderIfNeeded): Recent change to block placeholder removal code
        caused this regression. The code became too aggressive in removing block placeholders, and would remove them
        from blocks other than the block containing the selection.
        * layout-tests/editing/inserting/insert-div-023-expected.txt: This file had a spur
        * layout-tests/editing/inserting/insert-div-024-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8414 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix for oddness on albertsons.com. Make sure not to crash when setting/removing...
hyatt [Thu, 20 Jan 2005 20:36:29 +0000 (20:36 +0000)]
Fix for oddness on albertsons.com.  Make sure not to crash when setting/removing style properties on a node
with no document.  The bug # is 3813900.

        Reviewed by john

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8413 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
harrison [Thu, 20 Jan 2005 19:34:48 +0000 (19:34 +0000)]
    Reviewed by Darin.

        PARTIAL fix for following bug.  Create attributed string with fonts and links.  Still need to add attachments.
        <rdar://problem/3942606> AX: Support kAXAttributedStringForTextMarkerRangeParameterizedAttribute

        * kwq/KWQAccObject.mm:
        (-[KWQAccObject textUnderElement]):
        Touched up previous checkin to match code review comments.

        (-[KWQAccObject value]):
        Use plainText for this one instead of obsolete attributedString.

        (-[KWQAccObject accessibilityParameterizedAttributeNames]):
        Advertise AXAttributedStringForTextMarkerRange.

        (AXAttributeStringAddFont):
        (AXAttributeStringAddElement):
        (-[KWQAccObject linkUIElementForNode:]):
        (-[KWQAccObject _addAttachmentForElement:URL:needsParagraph:usePlaceholder:]):
        (-[KWQAccObject accessibilityAttributedStringForRange:]):
        New routines to support AXAttributedStringForTextMarkerRange.

        (-[KWQAccObject doAXAttributedStringForTextMarkerRange:]):
        Use new routines instead of obsolete attributedString.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8412 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
harrison [Thu, 20 Jan 2005 19:01:44 +0000 (19:01 +0000)]
    Reviewed by Darin.

        <rdar://problem/3960196> AX Crash in DOM::Range::setStartBefore

        * khtml/xml/dom_docimpl.cpp:
        (DocumentImpl::getAccObjectCache):
        Adjust when detecting cache in non-top level WebArea.
        * kwq/KWQAccObject.mm:
        (-[KWQAccObject textUnderElement]):
        (-[KWQAccObject accessibilityAttributeValue:]):
        Make sure the elements document is the current one for the part.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8411 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John.
darin [Thu, 20 Jan 2005 18:05:19 +0000 (18:05 +0000)]
    Reviewed by John.

        - fixed <rdar://problem/3922980> Mail not crashing, just quitting itself suddenly in -[WebHTMLView(MailExtras) findString:options:]

        * khtml/dom/dom2_range.cpp: (DOM::operator==): Handle null and detached ranges without raising exceptions.
        The uncaught exception would make the entire program terminate.

        - fixed <rdar://problem/3885729> attempting to modify a computed style does nothing, but should raise a DOM exception

        * kwq/DOM-CSS.mm: (-[DOMCSSStyleDeclaration setProperty:::]): Uncommented this code which I had to disable
        back in November because Mail was stumbling over it.

        - fixed <rdar://problem/3943049> focus() called during onload handler results in square text field on <input type=search>

        * kwq/KWQLineEdit.mm: (QLineEdit::baselinePosition): Change computation so it will work even for fields
        that position their text in a way that depends on vertical size. In the case of a search field, the text
        is centered, so the old logic was broken.
        * kwq/KWQTextField.mm: (-[KWQTextFieldController initWithTextField:QLineEdit:]): Set the frame size to
        something large enough to accomodate the field editor. If we start the frame at size 0,0 we run into
        AppKit trouble when it insets the frame to figure out the frame for the field editor. In the case of this
        bug this happens because we become first responder before being sized and positioned by the HTML layout code.

        - improved debugging output when using "po" from gdb with Objective-C DOM

        * kwq/DOM.mm:
        (-[DOMNode description]): Added. Includes node name ("<tr>") and node value (e.g., string for text node).
        (-[DOMRange description]): Tweaked format.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8410 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago After further discussion with Real we have decided to NOT include the
rjw [Thu, 20 Jan 2005 02:09:23 +0000 (02:09 +0000)]
After further discussion with Real we have decided to NOT include the
additional CLSID for the real plugin.  See 3958601.

        Reviewed by Darin.

        * khtml/rendering/render_frames.cpp:
        (RenderPartObject::updateWidget):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8409 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3853676> Browser Crash when accessing CCWeb Progress Page...
rjw [Thu, 20 Jan 2005 01:39:19 +0000 (01:39 +0000)]
Fixed <rdar://problem/3853676> Browser Crash when accessing CCWeb Progress Page - KJS::Bindings::convertValueToJValue

Fixed the following problems with LiveConnect that are demonstrated by the application
described in 3853676.

1.  If a nil object is passed in an array from Java to JavaScript we will crash.
2.  We sometimes will incorrectly attempt to access a generic JavaScript as a Java runtime object wrapper.
3.  We will sometimes fail to find the correct static method ID.

        Reviewed by Maciej.

        * bindings/jni/jni_jsobject.cpp:
        (JSObject::convertJObjectToValue):
        (JSObject::listFromJArray):
        * bindings/jni/jni_runtime.cpp:
        (JavaField::valueFromInstance):
        (JavaField::setValueToInstance):
        * bindings/jni/jni_utility.cpp:
        (KJS::Bindings::getMethodID):
        (KJS::Bindings::convertValueToJValue):
        * bindings/runtime_array.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8408 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej.
harrison [Thu, 20 Jan 2005 00:45:57 +0000 (00:45 +0000)]
    Reviewed by Maciej.

        <rdar://problem/3949429> AX: word marker routines returns incorrect data for empty line

        Previous checkin for this bug was missing some of the patch.  Editing snafu with multiple changes in tree.

        * kwq/KWQAccObject.mm:
        (-[KWQAccObject doAXRightLineTextMarkerRangeForTextMarker:]):
        (-[KWQAccObject doAXNextWordEndTextMarkerForTextMarker:]):
        (-[KWQAccObject doAXPreviousWordStartTextMarkerForTextMarker:]):
        (-[KWQAccObject doAXNextLineEndTextMarkerForTextMarker:]):
        (-[KWQAccObject doAXPreviousLineStartTextMarkerForTextMarker:]):
        (-[KWQAccObject doAXPreviousSentenceStartTextMarkerForTextMarker:]):
        (-[KWQAccObject doAXPreviousParagraphStartTextMarkerForTextMarker:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8407 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix for 3513067, spaces being lost when RTL text is rendered. Make sure not to...
hyatt [Thu, 20 Jan 2005 00:05:14 +0000 (00:05 +0000)]
Fix for 3513067, spaces being lost when RTL text is rendered.  Make sure not to allow hangers or spaces
in the margin.

        Reviewed by john

        * WebCoreSupport.subproj/WebTextRenderer.m:
        (-[WebTextRenderer _createATSUTextLayoutForRun:style:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8406 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Thu, 20 Jan 2005 00:01:58 +0000 (00:01 +0000)]
    Reviewed by John

        * khtml/editing/htmlediting.cpp:
        (khtml::InsertParagraphSeparatorCommand::applyStyleAfterInsertion): Small, cosmetic change John and I decided
        to do on my last checkin, but I forgot to do before landing.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8405 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Wed, 19 Jan 2005 23:54:09 +0000 (23:54 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3959727> REGRESSION (Mail): Style not preserved on blank lines

        * khtml/editing/htmlediting.cpp:
        (khtml::CompositeEditCommand::applyStyle):
        (khtml::CompositeEditCommand::insertBlockPlaceholder): New function that unconditionally adds a block placeholder.
        (khtml::CompositeEditCommand::insertBlockPlaceholderIfNeeded): Now returns bool based on whether
        placeholder was added or not.
        (khtml::CompositeEditCommand::removeBlockPlaceholderIfNeeded): Now searches all the descendents of a block
        looking for a placeholder. The old code, which just looked at the last child of a node, started missing
        once block placeholders became styled (which started happening with this patch).
        (khtml::DeleteSelectionCommand::calculateTypingStyleAfterDelete): Now handles applying typing style
        to a block placeholder at call time, rather than setting the typing style as a latent style that
        might be applied later. This is an important part of the bug fix.
        (khtml::DeleteSelectionCommand::doApply): Now uses bool return value from insertBlockPlaceholderIfNeeded()
        and passes it along to calculateStyleBeforeInsertion, so the case where a block placeholder needs to
        be styled can be detected.
        (khtml::InsertParagraphSeparatorCommand::InsertParagraphSeparatorCommand): Changed the way this class
        managed style. Before it would calculate and set typing style for the block added. This is not
        sufficient. Added blocks need to styled immediately. Some name changes to instance variables in
        this class due to the change to accommodate this change.
        (khtml::InsertParagraphSeparatorCommand::~InsertParagraphSeparatorCommand): Name changes, as above.
        (khtml::InsertParagraphSeparatorCommand::calculateStyleBeforeInsertion): Ditto.
        (khtml::InsertParagraphSeparatorCommand::applyStyleAfterInsertion): Ditto.
        (khtml::InsertParagraphSeparatorCommand::doApply): Ditto.
        (khtml::ReplaceSelectionCommand::doApply): Improve check for testing when a placeholder
        block can be removed in its entirety after the insertion.
        * khtml/editing/htmlediting.h: Update header accordingly.
        * khtml/khtml_part.cpp:
        (KHTMLPart::selectionComputedStyle): Move position for computed style check downstream before
        doing check when the position is in an empty block (this makes sure any style on any block
        placeholder is accounted for).

        New layout tests to check bug fix.

        * layout-tests/editing/style/block-style-004-expected.txt: Added.
        * layout-tests/editing/style/block-style-004.html: Added.
        * layout-tests/editing/style/block-style-005-expected.txt: Added.
        * layout-tests/editing/style/block-style-005.html: Added.
        * layout-tests/editing/style/block-style-006-expected.txt: Added.
        * layout-tests/editing/style/block-style-006.html: Added.

        Results updated to reflect new block placeholder code.

        * layout-tests/editing/inserting/insert-div-004-expected.txt
        * layout-tests/editing/inserting/insert-div-005-expected.txt
        * layout-tests/editing/inserting/insert-div-006-expected.txt
        * layout-tests/editing/inserting/insert-div-008-expected.txt
        * layout-tests/editing/inserting/insert-div-011-expected.txt
        * layout-tests/editing/inserting/insert-div-012-expected.txt
        * layout-tests/editing/inserting/insert-div-013-expected.txt
        * layout-tests/editing/inserting/insert-div-014-expected.txt
        * layout-tests/editing/inserting/insert-div-015-expected.txt
        * layout-tests/editing/inserting/insert-div-016-expected.txt
        * layout-tests/editing/inserting/insert-div-017-expected.txt
        * layout-tests/editing/inserting/insert-div-018-expected.txt
        * layout-tests/editing/inserting/insert-div-019-expected.txt
        * layout-tests/editing/inserting/insert-div-021-expected.txt
        * layout-tests/editing/inserting/insert-div-022-expected.txt
        * layout-tests/editing/inserting/insert-div-023-expected.txt
        * layout-tests/editing/inserting/insert-div-024-expected.txt

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8404 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by vicki
vicki [Wed, 19 Jan 2005 22:57:03 +0000 (22:57 +0000)]
    Reviewed by vicki

- fixed <rdar://problem/3962559> stopAnimationsInView leaks after cvs-base

        * WebCoreSupport.subproj/WebImageData.m: (+[WebImageData stopAnimationsInView:]): add a release

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8403 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Dont null-check the renderer before submitting, since a script can set it to display...
hyatt [Wed, 19 Jan 2005 22:43:39 +0000 (22:43 +0000)]
Dont null-check the renderer before submitting, since a script can set it to display:none and still expect the
submission to occur.  Fixes bug #3477282.

        Reviewed by kdecker

        (DOM::HTMLInputElementImpl::defaultEventHandler):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8402 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed: <rdar://problem/3961809> plug-in code attempts to load empty URL
cblu [Wed, 19 Jan 2005 18:39:11 +0000 (18:39 +0000)]
Fixed: <rdar://problem/3961809> plug-in code attempts to load empty URL

        Reviewed by john.

        * Plugins.subproj/WebNetscapePluginEmbeddedView.m:
        (-[WebNetscapePluginEmbeddedView didStart]): don't start load if URL is empty

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8401 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoWebCore:
rjw [Wed, 19 Jan 2005 02:34:22 +0000 (02:34 +0000)]
WebCore:
Fixed several issues all arising from analysis of plugin detection code at ifilm.com:

Fixed <rdar://problem/3958592> can't script plug-ins if plug-in is invoked with <object> element instead of <embed>
Fixed <rdar://problem/3958597> <object> elements with IDs do not show up as named properties of the document
Fixed <rdar://problem/3960973> DOM objects for plugin elements are not accessible
Fixed <rdar://problem/3958601> need an additional class ID in WebCore for the Real plug-in

We now support accessing scriptable plugin objects that are specified with <applet>, <embed>, or <object>
tags.  Also, if any of these elements are named they can be accessed from the document or window objects.
Finally, DOM methods are properties will be forwarded appropriately for the plugin's root scriptable object.

        Reviewed by Chris.

        * khtml/dom/html_document.cpp:
        (HTMLDocument::objects):
        * khtml/dom/html_document.h:
        * khtml/ecma/kjs_dom.cpp:
        (DOMDocumentProtoFunc::tryCall):
        (DOMElementProtoFunc::tryCall):
        (KJS::getRuntimeObject):
        * khtml/ecma/kjs_dom.h:
        * khtml/ecma/kjs_html.cpp:
        (KJS::HTMLDocument::tryGet):
        (KJS::HTMLElement::tryGet):
        (KJS::HTMLCollection::tryGet):
        (KJS::HTMLCollection::getNamedItems):
        * khtml/ecma/kjs_window.cpp:
        (Window::get):
        * khtml/html/html_miscimpl.cpp:
        (HTMLCollectionImpl::traverseNextItem):
        * khtml/html/html_miscimpl.h:
        (DOM::HTMLCollectionImpl::):
        * khtml/html/html_objectimpl.cpp:
        (HTMLAppletElementImpl::getAppletInstance):
        (HTMLObjectElementImpl::HTMLObjectElementImpl):
        (HTMLObjectElementImpl::getObjectInstance):
        * khtml/html/html_objectimpl.h:
        * khtml/rendering/render_frames.cpp:
        (RenderPartObject::updateWidget):
        * kwq/KWQKHTMLPart.h:
        * kwq/KWQKHTMLPart.mm:
        (KWQKHTMLPart::getObjectInstanceForView):

JavaScriptCore:
Fixed several issues all arising from analysis of plugin detection code at ifilm.com:

Fixed <rdar://problem/3958592> can't script plug-ins if plug-in is invoked with <object> element instead of <embed>
Fixed <rdar://problem/3958597> <object> elements with IDs do not show up as named properties of the document
Fixed <rdar://problem/3960973> DOM objects for plugin elements are not accessible
Fixed <rdar://problem/3958601> need an additional class ID in WebCore for the Real plug-in

We now support accessing scriptable plugin objects that are specified with <applet>, <embed>, or <object>
tags.  Also, if any of these elements are named they can be accessed from the document or window objects.
Finally, DOM methods are properties will be forwarded appropriately for the plugin's root scriptable object.

        Reviewed by Chris.

        * bindings/objc/objc_instance.h:
        * bindings/objc/objc_instance.mm:
        (ObjcInstance::supportsSetValueOfUndefinedField):
        * bindings/runtime.h:
        (KJS::Bindings::Instance::supportsSetValueOfUndefinedField):
        * bindings/runtime_object.cpp:
        (RuntimeObjectImp::RuntimeObjectImp):
        (RuntimeObjectImp::get):
        (RuntimeObjectImp::put):
        (RuntimeObjectImp::canPut):
        (RuntimeObjectImp::hasProperty):
        (RuntimeObjectImp::defaultValue):
        * bindings/runtime_object.h:
        (KJS::RuntimeObjectImp::fallbackObject):
        * kjs/object.cpp:
        (KJS::ObjectImp::ObjectImp):
        * kjs/object.h:
        (KJS::ObjectImp::forwardingScriptMessage):
        (KJS::ObjectImp::setForwardingScriptMessage):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8400 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix for 3948123, rolling over link erases nearby text. The repaint rect check for...
hyatt [Tue, 18 Jan 2005 23:08:50 +0000 (23:08 +0000)]
Fix for 3948123, rolling over link erases nearby text.  The repaint rect check for lines was wrong whenever
two lines overlapped.

        Reviewed by kocienda

        * khtml/rendering/render_flow.cpp:
        (RenderFlow::paintLines):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8399 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Hyatt
kocienda [Tue, 18 Jan 2005 22:58:47 +0000 (22:58 +0000)]
    Reviewed by Hyatt

        Fix for this bug:

        <rdar://problem/3960116> Focus rings paint incorrectly for contenteditable blocks in web pages

        * khtml/rendering/render_flow.cpp:
        (RenderFlow::addFocusRingRects): Fix painting of focus rings so that ring only paints around
        outermost contenteditable elements.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8398 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Back out a change that was incorrectly committed yesterday.
rjw [Tue, 18 Jan 2005 22:52:52 +0000 (22:52 +0000)]
Back out a change that was incorrectly committed yesterday.

        Reviewed by Chris.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8397 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
harrison [Tue, 18 Jan 2005 21:03:46 +0000 (21:03 +0000)]
    Reviewed by Darin.

        <rdar://problem/3959668> accessibilityFocusedUIElement sometimes returns an ignored element; it must not

        * kwq/KWQAccObject.mm:
        (-[KWQAccObject accessibilityFocusedUIElement]):
        Return parentObjectUnignored if focused object is ignored.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8396 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Tue, 18 Jan 2005 18:11:21 +0000 (18:11 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3952877> REGRESSION (Mail): Command-left/right-arrows don't work with file attachment

        * khtml/editing/selection.cpp:
        (khtml::nodeForInlineBox): New helper function used in reimplementation of function below.
        (khtml::selectionForLine): Reimplemented using line box smarts. I originally wrote this code when
        I had a less than full understanding of line layout. I can do better now, and my new version no
        longer fails to notice attachments when doing the kind of navigation mentioned in the bug.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8395 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John Sullivan.
harrison [Tue, 18 Jan 2005 01:44:49 +0000 (01:44 +0000)]
    Reviewed by John Sullivan.

        <rdar://problem/3949429> AX: word marker routines returns incorrect data for empty line
        <rdar://problem/3949848> AX: paragraph marker routines do not work when given a paragraph end marker

        Also fixed sentence support in the same way.

        * kwq/KWQAccObject.mm:
        (-[KWQAccObject doAXNextLineEndTextMarkerForTextMarker:]):
        (-[KWQAccObject doAXNextSentenceEndTextMarkerForTextMarker:]):
        (-[KWQAccObject doAXPreviousSentenceStartTextMarkerForTextMarker:]):
        (-[KWQAccObject doAXNextParagraphEndTextMarkerForTextMarker:]):
        (-[KWQAccObject doAXPreviousParagraphStartTextMarkerForTextMarker:]):
        Move one position in desired direction before calling visible_units.cpp code.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8394 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Updated expected test results.
harrison [Tue, 18 Jan 2005 01:33:40 +0000 (01:33 +0000)]
    Updated expected test results.

        <rdar://problem/3945880> line-ending space seems not to be present

        * layout-tests/editing/deleting/delete-ws-fixup-001-expected.txt:
        * layout-tests/editing/deleting/delete-ws-fixup-002-expected.txt:
        * layout-tests/editing/inserting/insert-div-020-expected.txt:
        * layout-tests/editing/inserting/insert-div-021-expected.txt:
        * layout-tests/editing/inserting/insert-div-022-expected.txt:
        * layout-tests/editing/inserting/insert-div-023-expected.txt:
        * layout-tests/editing/inserting/insert-div-024-expected.txt:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8393 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Dave Hyatt (bidi.cpp) and Darin Adler (selection.cpp).
harrison [Tue, 18 Jan 2005 01:07:26 +0000 (01:07 +0000)]
    Reviewed by Dave Hyatt (bidi.cpp) and Darin Adler (selection.cpp).

        <rdar://problem/3945880> line-ending space seems not to be present

        * khtml/editing/selection.cpp:
        (khtml::Selection::validate):
        Extend AFTER_WHITE_SPACE code to support white spac in the middle of paragraphs, not just the end.
        * khtml/rendering/bidi.cpp:
        (khtml::RenderBlock::findNextLineBreak):
        Pick left/rightness of word selection based on being at the end of paragraph (i.e. after a hard line break).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8392 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * DOM.subproj/DOMPrivate.h: Check in generated file.
darin [Tue, 18 Jan 2005 01:06:27 +0000 (01:06 +0000)]
    * DOM.subproj/DOMPrivate.h: Check in generated file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8391 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John Louch.
darin [Tue, 18 Jan 2005 00:43:14 +0000 (00:43 +0000)]
    Reviewed by John Louch.

        - fixed <rdar://problem/3958503> need screenX and screenY to use WebKit windowFrame delegate

        * khtml/ecma/kjs_window.cpp: (Window::get): Change screenX and screenY to use frameGeometry instead
        of using mapToGlobal and screen in a complicated way.
        * kwq/KWQKHTMLView.mm: Removed unused mapToGlobal function.
        * kwq/KWQWindowWidget.h: Ditto.
        * kwq/KWQWindowWidget.mm: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8390 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoAdd layout test for overflow float bug.
hyatt [Mon, 17 Jan 2005 23:30:10 +0000 (23:30 +0000)]
Add layout test for overflow float bug.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8389 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix a screwup in rightmost/lowets position computation. 3955207. Make sure floats...
hyatt [Mon, 17 Jan 2005 23:27:03 +0000 (23:27 +0000)]
Fix a screwup in rightmost/lowets position computation.  3955207.  Make sure floats with layers are still checked.

        Reviewed by kocienda

* khtml/rendering/render_block.cpp

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8388 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John and Richard.
darin [Mon, 17 Jan 2005 23:23:21 +0000 (23:23 +0000)]
    Reviewed by John and Richard.

        - fixed <rdar://problem/3907453> printing a multi-page PDF document from Safari doesn't produce correct output

        * WebView.subproj/WebFrameViewPrivate.h: Added.
        * WebView.subproj/WebFrameView.m:
        (-[WebFrameView canPrintHeadersAndFooters]): Added. Returnes NO for documents that can't print headers or footers,
        and delegates to the document view to answer the question. Defaults to NO, since only a view that actively does
        the work is compatible with our header and footer code.
        (-[WebFrameView printOperationWithPrintInfo:]): Added. Returns an NSPrintOperation set up for printing. The reason
        we return this rather than an NSView is that in the PDFView case, the print info is changed around before creating
        the NSPrintOperation, and also the PDFKit SPI works this way.

        * WebView.subproj/WebHTMLView.m: (-[WebHTMLView canPrintHeadersAndFooters]): Added. Returns YES.
        * WebView.subproj/WebImageView.m: (-[WebImageView canPrintHeadersAndFooters]): Ditto.
        * WebView.subproj/WebTextView.m: (-[WebTextView canPrintHeadersAndFooters]): Ditto.

        * WebView.subproj/WebPDFView.m:
        (-[WebPDFView canPrintHeadersAndFooters]): Added. Returns NO.
        (-[WebPDFView printOperationWithPrintInfo:]): Added. Calls getPrintOperationForPrintInfo: autoRotate:YES on the PDFView.

        * WebKit.pbproj/project.pbxproj: Added WebFrameViewPrivate.h as a new private header.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8387 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed build snafu.
rjw [Mon, 17 Jan 2005 23:08:05 +0000 (23:08 +0000)]
Fixed build snafu.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8386 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Mon, 17 Jan 2005 22:54:43 +0000 (22:54 +0000)]
    Reviewed by John

        <rdar://problem/3953366> Problems with typing attributes in HTML compose

        * khtml/editing/htmlediting.cpp:
        (khtml::DeleteSelectionCommand::calculateTypingStyleAfterDelete): Remove misguided "optimization"
        that tried to sense when typing style could be cleared without actually doing a style diff between
        before-delete and after-delete positions. Removing this extra check and running the
        general-purpose code fixes the bug.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8385 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoWebCore:
rjw [Mon, 17 Jan 2005 22:41:22 +0000 (22:41 +0000)]
WebCore:
Fixed <rdar://problem/3753030> Need to ensure same origin for plugin binding invocations (origin security rules)

Keep track of originating execution context and target execution
context for native JS object wrappers, and perform appropriate
security checks.

        Reviewed by David Harrison.

        * khtml/ecma/kjs_binding.cpp:
        (ScriptInterpreter::isGlobalObject):
        (ScriptInterpreter::isSafeScript):
        (ScriptInterpreter::interpreterForGlobalObject):
        * khtml/ecma/kjs_binding.h:
        * khtml/ecma/kjs_window.cpp:
        (Window::interpreter):
        (Window::isSafeScript):
        * khtml/ecma/kjs_window.h:
        * kwq/DOMInternal.mm:
        (-[WebScriptObject _initializeScriptDOMNodeImp]):
        * kwq/KWQKHTMLPart.mm:
        (KWQKHTMLPart::windowScriptObject):
        (KWQKHTMLPart::windowScriptNPObject):
        * kwq/WebCoreBridge.mm:
        (rootForView):
        (-[WebCoreBridge init]):

JavaScriptCore:
Fixed <rdar://problem/3753030> Need to ensure same origin for plugin binding invocations (origin security rules)

Keep track of originating execution context and target execution
context for native JS object wrappers, and perform appropriate
security checks.

        Reviewed by David Harrison.

        * bindings/NP_jsobject.cpp:
        (_isSafeScript):
        (_NPN_CreateScriptObject):
        (_NPN_Invoke):
        (_NPN_Evaluate):
        (_NPN_GetProperty):
        (_NPN_SetProperty):
        (_NPN_RemoveProperty):
        (_NPN_HasProperty):
        (_NPN_HasMethod):
        (_NPN_SetException):
        * bindings/NP_jsobject.h:
        * bindings/c/c_instance.cpp:
        (CInstance::CInstance):
        (CInstance::stringValue):
        * bindings/c/c_instance.h:
        * bindings/c/c_utility.cpp:
        (convertValueToNPVariant):
        * bindings/jni/jni_instance.cpp:
        (JavaInstance::JavaInstance):
        (JavaInstance::valueOf):
        * bindings/jni/jni_instance.h:
        * bindings/objc/WebScriptObject.mm:
        (-[WebScriptObject _initializeWithObjectImp:KJS::originExecutionContext:Bindings::executionContext:Bindings::]):
        (-[WebScriptObject _initWithObjectImp:KJS::originExecutionContext:Bindings::executionContext:Bindings::]):
        (-[WebScriptObject KJS::Bindings::]):
        (-[WebScriptObject _setOriginExecutionContext:KJS::Bindings::]):
        (-[WebScriptObject _isSafeScript]):
        (-[WebScriptObject callWebScriptMethod:withArguments:]):
        (-[WebScriptObject evaluateWebScript:]):
        (-[WebScriptObject setValue:forKey:]):
        (-[WebScriptObject valueForKey:]):
        (-[WebScriptObject removeWebScriptKey:]):
        (-[WebScriptObject stringRepresentation]):
        (-[WebScriptObject webScriptValueAtIndex:]):
        (-[WebScriptObject setWebScriptValueAtIndex:value:]):
        (+[WebScriptObject _convertValueToObjcValue:KJS::originExecutionContext:Bindings::executionContext:Bindings::]):
        * bindings/objc/WebScriptObjectPrivate.h:
        * bindings/objc/objc_instance.h:
        * bindings/objc/objc_runtime.mm:
        (convertValueToObjcObject):
        * bindings/objc/objc_utility.mm:
        (KJS::Bindings::convertValueToObjcValue):
        * bindings/runtime.cpp:
        (Instance::Instance):
        (Instance::operator=):
        * bindings/runtime.h:
        (KJS::Bindings::Instance::Instance):
        (KJS::Bindings::Instance::setExecutionContext):
        (KJS::Bindings::Instance::executionContext):
        * bindings/runtime_root.cpp:
        (RootObject::setInterpreter):
        * bindings/runtime_root.h:
        * kjs/interpreter.h:
        (KJS::Interpreter::isGlobalObject):
        (KJS::Interpreter::interpreterForGlobalObject):
        (KJS::Interpreter::isSafeScript):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8384 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoRoll out unintended work-in-progress change
kocienda [Mon, 17 Jan 2005 22:20:50 +0000 (22:20 +0000)]
Roll out unintended work-in-progress change

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8383 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Mon, 17 Jan 2005 22:18:47 +0000 (22:18 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3786659> REGRESSION (Mail): editable WebViews don't work with
        "size up" and "size down" NSFontManager changes

        * khtml/css/cssparser.cpp:
        (CSSParser::parseValue): Add support for parsing new font size delta property.
        * khtml/css/cssproperties.c: Generated file.
        * khtml/css/cssproperties.h: Ditto.
        * khtml/css/cssproperties.in: Add support for parsing new font size delta property.
        * khtml/editing/htmlediting.cpp:
        (khtml::isEmptyStyleSpan): New helper function. Code existed before, but now factored out for convenient use.
        (khtml::CompositeEditCommand::removeNodeAttribute): Does check on value to see it exists before creating
        and running command to do the removal.
        (khtml::ApplyStyleCommand::doApply): Now calls new applyRelativeFontStyleChange() function as part of its work.
        (khtml::ApplyStyleCommand::applyRelativeFontStyleChange): New function that does the "heavy lifting" to handle
        relative font size changes.
        (khtml::ApplyStyleCommand::applyInlineStyle): Range check the start and end positions to make sure the start is
        before or equal to the end. Swap them if not true. I ran into this problem in some similar code in
        applyRelativeFontStyleChange(). Moving that goodness here too.
        (khtml::ApplyStyleCommand::splitTextAtEndIfNeeded): Uses a local variable to save a value used often.
        Code is the same, but reads better now. Function now returns bool as well, just like splitTextAtStartIfNeeded.
        I use the bool return value now (I obviously did not need it before).
        (khtml::ApplyStyleCommand::computedFontSize): New helper function.
        (khtml::ApplyStyleCommand::joinChildTextNodes): Ditto.
        (khtml::createStyleSpanElement): Ditto.
        * khtml/editing/htmlediting.h: Update header accordingly.
        * khtml/editing/jsediting.cpp: Add new command string to enable relative font size changes.
        * kwq/DOM-CSS.mm:
        (-[DOMCSSStyleDeclaration _fontSizeDelta]): New convenience.
        (-[DOMCSSStyleDeclaration _setFontSizeDelta:]): Ditto.
        * kwq/DOMPrivate.h: Declare new conveniences.
        * layout-tests/editing/editing.js: Add new JS to enable relative font size changes, as well as explicit font size setting.
        * layout-tests/editing/style/relative-font-size-change-001-expected.txt: Added.
        * layout-tests/editing/style/relative-font-size-change-001.html: Added.
        * layout-tests/editing/style/relative-font-size-change-002-expected.txt: Added.
        * layout-tests/editing/style/relative-font-size-change-002.html: Added.
        * layout-tests/editing/style/relative-font-size-change-003-expected.txt: Added.
        * layout-tests/editing/style/relative-font-size-change-003.html: Added.
        * layout-tests/editing/style/relative-font-size-change-004-expected.txt: Added.
        * layout-tests/editing/style/relative-font-size-change-004.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8382 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * khtml/css/cssproperties.c: Regenerated with gperf 3.0.1.
darin [Fri, 14 Jan 2005 23:55:37 +0000 (23:55 +0000)]
    * khtml/css/cssproperties.c: Regenerated with gperf 3.0.1.
        * khtml/css/cssvalues.c: Regenerated with gperf 3.0.1.
        * khtml/misc/htmlattrs.c: Regenerated with gperf 3.0.1.
        * khtml/misc/htmltags.c: Regenerated with gperf 3.0.1.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8381 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed: <rdar://problem/3886415> arrow keys don't work when the user hits Back to...
cblu [Fri, 14 Jan 2005 20:09:32 +0000 (20:09 +0000)]
Fixed: <rdar://problem/3886415> arrow keys don't work when the user hits Back to return to RSS page

        Reviewed by hyatt.

        (KWQKHTMLPart::openURLFromPageCache): restore mousePressNode
        (KWQKHTMLPart::mousePressNode): new
        * kwq/KWQPageState.h:
        * kwq/KWQPageState.mm:
        (-[KWQPageState initWithDocument:URL:windowProperties:locationProperties:interpreterBuiltins:]): save the mousePressNode
        (-[KWQPageState clear]): clear the mousePressNode
        (-[KWQPageState dealloc]): deref the mousePressNode
        (-[KWQPageState mousePressNode]): new

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8380 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed: <rdar://problem/3932107> Safari does not load QT Cocoa plug-in if the WebPlug...
cblu [Fri, 14 Jan 2005 01:01:03 +0000 (01:01 +0000)]
Fixed: <rdar://problem/3932107> Safari does not load QT Cocoa plug-in if the WebPluginMIMETypes key is not in the info.plist

Fixing this bug required that we allow WebKit plug-ins (as well as Netscape plug-ins) support BP_CreatePluginMIMETypesPreferences
which allows plug-ins create an auxiliary MIME types file.

        Reviewed by adele.

        * Plugins.subproj/WebBasePluginPackage.h:
        * Plugins.subproj/WebBasePluginPackage.m:
        (+[WebBasePluginPackage preferredLocalizationName]): moved from WebNetscapePluginPackage
        (-[WebBasePluginPackage pListForPath:createFile:]): ditto
        (-[WebBasePluginPackage getPluginInfoFromPLists]): ditto
        (-[WebBasePluginPackage isLoaded]): return isLoaded ivar
        (-[WebBasePluginPackage load]): if loaded, get BP_CreatePluginMIMETypesPreferences symbol
        * Plugins.subproj/WebNetscapePluginPackage.h:
        * Plugins.subproj/WebNetscapePluginPackage.m:
        (-[WebNetscapePluginPackage load]): call super when done so BP_CreatePluginMIMETypesPreferences can be initialized
        * Plugins.subproj/WebPluginPackage.m:
        (-[WebPluginPackage initWithPath:]): call getPluginInfoFromPLists
        (-[WebPluginPackage load]): call super when done so BP_CreatePluginMIMETypesPreferences can be initialized

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8373 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changed another place where the method name was incorrect in
rjw [Fri, 14 Jan 2005 00:17:57 +0000 (00:17 +0000)]
Changed another place where the method name was incorrect in
the comment.

Reviewed (and found) by Darin.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8371 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3952809> WebJavaPlugIn.h comments need method name corrected...
rjw [Fri, 14 Jan 2005 00:11:52 +0000 (00:11 +0000)]
Fixed <rdar://problem/3952809> WebJavaPlugIn.h comments need method name corrected (webPlugInCallJava)

        Reviewed by Maciej.

        * Plugins.subproj/WebJavaPlugIn.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8370 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoSafari-179+ for TOT. The tree is open!
vicki [Fri, 14 Jan 2005 00:05:35 +0000 (00:05 +0000)]
Safari-179+ for TOT.  The tree is open!

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8369 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoSafari-179 stamp
vicki [Thu, 13 Jan 2005 23:58:31 +0000 (23:58 +0000)]
Safari-179 stamp

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8367 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoReviewed by Adele.
vicki [Thu, 13 Jan 2005 23:54:30 +0000 (23:54 +0000)]
Reviewed by Adele.

- fix <rdar://problem/3946836> Safari about box lists 2004 instead of 2005

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8366 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3951911> REGRESSION: Animated GIF images with loop counts...
rjw [Thu, 13 Jan 2005 22:13:04 +0000 (22:13 +0000)]
Fixed <rdar://problem/3951911> REGRESSION: Animated GIF images with loop counts no longer update

Draw last image after animation loop terminates.  (We
were drawing the image at index+1, which doesn't exist!)

        Reviewed by Darin.

        * WebCoreSupport.subproj/WebImageData.m:
        (-[WebImageData _nextFrame:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8365 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3952084> REGRESSION: Links at projectseven.com now draw and...
rjw [Thu, 13 Jan 2005 22:02:58 +0000 (22:02 +0000)]
Fixed <rdar://problem/3952084> REGRESSION: Links at projectseven.com now draw and update incorrectly during hover

Turn off use of new CGContextStrokeLineSegments API.  We should
turn back on when 3952944 is fixed.

        Reviewed by Darin.

        * WebCoreSupport.subproj/WebTextRenderer.m:
        (-[WebTextRenderer drawLineForCharacters:yOffset:width:color:thickness:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8364 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agofixed incorrect check-in date
darin [Thu, 13 Jan 2005 21:17:34 +0000 (21:17 +0000)]
fixed incorrect check-in date

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8363 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Ken Kocienda.
harrison [Thu, 13 Jan 2005 19:45:33 +0000 (19:45 +0000)]
    Reviewed by Ken Kocienda.

        Better fix for 3905066.

        * khtml/editing/htmlediting.cpp:
        (khtml::InsertParagraphSeparatorCommand::doApply):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8362 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed: <rdar://problem/3937663> repro assertion failure and crash dragging image...
cblu [Thu, 13 Jan 2005 19:41:52 +0000 (19:41 +0000)]
Fixed: <rdar://problem/3937663> repro assertion failure and crash dragging image that has PDF source

        Reviewed by adele.

        * Misc.subproj/WebNSViewExtras.m:
        (-[NSView _web_dragImage:rect:event:pasteboard:source:offset:]): if [WebImageRenderer image] returns nil, fallback to code that uses a file icon as the drag image
        * WebCoreSupport.subproj/WebImageRendererFactory.m:
        (-[WebImageRendererFactory supportedMIMETypes]): removed code that omits PDF and PostScript from the list since this omission is only needed in WebImageView
        * WebView.subproj/WebImageView.m:
        (+[WebImageView supportedImageMIMETypes]): added code that omits PDF and PostScript since we don't want WebImageView to render these types

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8361 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoWebCore:
mjs [Thu, 13 Jan 2005 19:09:58 +0000 (19:09 +0000)]
WebCore:

        Reviewed by Darin.

<rdar://problem/3758033> REGRESSION (Mail): Support attributes in marked text (International input)

        * khtml/rendering/render_text.cpp:
        (InlineTextBox::paint): Support painting custom underline markers for
marked text in place of generic yellow.
        (InlineTextBox::paintMarkedTextUnderline): New method that handles this.
        * khtml/rendering/render_text.h:
        * kwq/KWQKHTMLPart.h: Declare new methods and structs.
        * kwq/KWQKHTMLPart.mm:
        (KWQKHTMLPart::clear): Clear marked test underlines.
        (KWQKHTMLPart::setMarkedTextRange): Takes attributes and ranges now.
        (convertAttributesToUnderlines): Converts NSAttributedString attributes
to simplified and C++-friendly form.
        (KWQKHTMLPart::markedTextUsesUnderlines): New method.
        (KWQKHTMLPart::markedTextUnderlines): New method.
        * kwq/KWQPainter.mm:
        (QPainter::drawLineForText): Handle pen width.
        * kwq/WebCoreBridge.h:
        * kwq/WebCoreBridge.mm:
        (-[WebCoreBridge setMarkedTextDOMRange:customAttributes:ranges:]): Take attributes
and ranges.
        * kwq/WebCoreTextRenderer.h:

WebKit:

        Reviewed by Darin.

<rdar://problem/3758033> REGRESSION (Mail): Support attributes in marked text (International input)

* WebCoreSupport.subproj/WebTextRenderer.m:
        (-[WebTextRenderer drawLineForCharacters:yOffset:width:color:thickness:]): Changed to support
underline thickness. Also added a bit of a hack here to move thickness 2 underlines down by
.5 pixels, since the rendering engine can't give a fractional pixel offset.
        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView validAttributesForMarkedText]): Support underline, underline color and marked
clause attributes. Others that NSText supports are unimplemented for now.
        (-[WebHTMLView firstRectForCharacterRange:]): Remove needless logging.
        (-[WebHTMLView unmarkText]): Updated for new WebCore SPI.
        (-[WebHTMLView _extractAttributes:ranges:fromAttributedString:]): New method to pull the attributes
and ranges out of an attributed string.
        (-[WebHTMLView setMarkedText:selectedRange:]): Extract attributes and pass to WebCore.
        (-[WebHTMLView insertText:]): Add comment noting that we don't really handle attributed strings
here.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8360 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Dave Hyatt.
harrison [Thu, 13 Jan 2005 01:31:43 +0000 (01:31 +0000)]
    Reviewed by Dave Hyatt.

        <rdar://problem/3888973> AX: Parent AXWebArea of nested AXWebAreas is messed up

        * kwq/KWQAccObject.mm:
        (-[KWQAccObject addChildrenToArray:]):
        Use the widget's outer view.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8353 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin Adler.
harrison [Thu, 13 Jan 2005 01:19:52 +0000 (01:19 +0000)]
    Reviewed by Darin Adler.

        * khtml/xml/dom_nodeimpl.cpp:
        (NodeImpl::displayNode):
        Add quotes around text node content.
        (NodeBaseImpl::childNode):
        Add nil check to return nil rather than crash when child node not found.
        * kwq/KWQAccObject.mm:
        (-[KWQAccObject doAXStringForTextMarkerRange:]):
        Pass range compliant positions to TextIterator.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8352 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Ken.
darin [Wed, 12 Jan 2005 23:55:08 +0000 (23:55 +0000)]
    Reviewed by Ken.

        - fixed <rdar://problem/3848257> WebView will draw more than AppKit asks it to, so views behind won't redraw enough (transparent WebView)

        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView _propagateDirtyRectsToOpaqueAncestors]): As recommended by Troy Stephens, do the
        layouts here in this call, since it's before propagating the dirty rects to our ancestors.
        This fixes the bug, but we only do it if the WebView is not opaque, because otherwise we can
        optimize by only doing layouts you really need, and doing them later on is safe because we
        know we don't need to draw any of the views behind us.
        (-[WebHTMLView _layoutIfNeeded]): Added. Factored out from the method below.
        (-[WebHTMLView _web_layoutIfNeededRecursive]): Added. Like the other "layout if needed" call,
        but unconditional.
        (-[WebHTMLView _web_layoutIfNeededRecursive:testDirtyRect:]): Factored out the guts into the
        _layoutIfNeeded method above. Otherwise unchanged.
        (-[NSView _web_layoutIfNeededRecursive]): Added.

        * WebView.subproj/WebFrame.m: (-[WebFrame _updateDrawsBackground]): Call setDrawsBackground:NO
        on the scroll view when changing the frame to no longer be in "draws background" mode. This
        is needed because the frame manages the "draws background" mode of the scroll view. It won't
        have any effect if you call setDrawsBackground:NO before starting to use a WebView, but without
        it calling setDrawsBackground:NO later won't have an immediate effect (easily visible in Safari).
        This was hidden before because the HTML view was filling with transparent color, which blew away
        the fill that was done by NSScrollView.

        - fixed <rdar://problem/3921129> reproducible crash at www.funnychristmas.com in CFSet manipulation in WebImageData

        * WebCoreSupport.subproj/WebImageData.m:
        (-[WebImageData _imageSourceOptions]): Changed types so we don't need a cast.
        (+[WebImageData stopAnimationsInView:]): Instead of building a set of sets, by putting in the sets with addObject,
        build a single set using unionSet, and then iterate the objects instead of having to iterate the sets and then the
        objects in each set. The old code ended up sharing the sets with the live code, when the whole idea was to gather
        all the renderers because the process of stopping modifies the active sets.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8348 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Avoid additional work on dealloc by adding early out to
rjw [Wed, 12 Jan 2005 23:41:00 +0000 (23:41 +0000)]
Avoid additional work on dealloc by adding early out to
removeNativeReference().  (This will save time on dealloc
for all ObjC DOM objects.)

        Reviewed by Darin.

        * bindings/runtime_root.cpp:
        (KJS::Bindings::removeNativeReference):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8347 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix for 3951203, CSS border style incorrectly clipped on inline elements. Make...
hyatt [Wed, 12 Jan 2005 23:33:21 +0000 (23:33 +0000)]
Fix for 3951203, CSS border style incorrectly clipped on inline elements.  Make sure the repaint rect set during
line layout is smarter about including the overflow for both old states and new states.

        Reviewed by john

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8346 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3923356> REGRESSION: Java/JavaScript security checks working...
rjw [Wed, 12 Jan 2005 23:24:18 +0000 (23:24 +0000)]
Fixed <rdar://problem/3923356> REGRESSION: Java/JavaScript security checks working incorrectly

We were always returning the first "root" object for all runtime
objects.  Changed 0 in loop to i, the index.

        Reviewed by David Harrison.

        * bindings/runtime_root.cpp:
        (KJS::Bindings::rootForImp):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8345 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3887930> Must use new Java plug-in API to get/set fields so...
rjw [Wed, 12 Jan 2005 22:03:55 +0000 (22:03 +0000)]
Fixed <rdar://problem/3887930> Must use new Java plug-in API to get/set fields so exception handling works (fixes many LiveConnect crashes)

Use the new dispatching API to invoke JNI, rather than calling JNI
directly.

        Reviewed by David Harrison.

        * bindings/jni/jni_instance.cpp:
        (JavaInstance::invokeMethod):
        * bindings/jni/jni_runtime.cpp:
        (JavaField::dispatchValueFromInstance):
        (JavaField::valueFromInstance):
        (JavaField::dispatchSetValueToInstance):
        (JavaField::setValueToInstance):
        * bindings/jni/jni_runtime.h:
        * bindings/jni/jni_utility.cpp:
        (KJS::Bindings::convertValueToJValue):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8344 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed <rdar://problem/3926825> Safari ignores GIF loop count
rjw [Wed, 12 Jan 2005 21:41:05 +0000 (21:41 +0000)]
Fixed <rdar://problem/3926825> Safari ignores GIF loop count

Get loop count from file properties, not image properties.

        Reviewed by Ken Kocienda.

        * WebCoreSupport.subproj/WebImageData.h:
        * WebCoreSupport.subproj/WebImageData.m:
        (-[WebImageData _commonTermination]):
        (-[WebImageData fileProperties]):
        (-[WebImageData _floatProperty:type:at:]):
        (-[WebImageData _floatFileProperty:type:]):
        (-[WebImageData _repetitionCount]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8343 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix for bug 3937608, versiontracker.com flashes and displays the right column below...
hyatt [Wed, 12 Jan 2005 19:13:35 +0000 (19:13 +0000)]
Fix for bug 3937608, versiontracker.com flashes and displays the right column below content.  Make sure to move
tables/overflows that dont fit within a block only in strict mode.

Fix for bug 3931049, characters dont show up when typing.  Make sure to dirty the right lines when this specific
case in editing is hit.

        Reviewed by darin (first one), kocienda (second one)

        * khtml/rendering/render_block.cpp:
        (khtml::RenderBlock::getClearDelta):
        * khtml/rendering/render_flow.cpp:
        (RenderFlow::dirtyLinesFromChangedChild):
        * khtml/rendering/render_line.cpp:
        (khtml::InlineFlowBox::verticallyAlignBoxes):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8342 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Wed, 12 Jan 2005 17:55:08 +0000 (17:55 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3926142> REGRESSION (Mail): Deleting text decreases quote level

        * khtml/editing/htmlediting.cpp:
        (khtml::DeleteSelectionCommand::initializePositionData): Change test that will prevent block
        merging. End-of-line test was just wrong. Call new start-of and end-of paragraph functions
        added to visible_position files.
        (khtml::DeleteSelectionCommand::moveNodesAfterNode): Add tests for BR elements, which makes
        tests to determine when to stop moving nodes more complete and correct. Also improved comments.
        * khtml/editing/visible_position.cpp:
        (khtml::isFirstVisiblePositionInParagraph): New function.
        (khtml::isLastVisiblePositionInParagraph): New function.
        * khtml/editing/visible_position.h: Update header accordingly.
        * layout-tests/editing/deleting/delete-block-merge-contents-002-expected.txt: New results, slightly different from former
        results but still correct.
        * layout-tests/editing/deleting/delete-block-merge-contents-007-expected.txt: Ditto.
        * layout-tests/editing/deleting/delete-block-merge-contents-018-expected.txt: Added.
        * layout-tests/editing/deleting/delete-block-merge-contents-018.html: Added.
        * layout-tests/editing/deleting/delete-block-merge-contents-019-expected.txt: Added.
        * layout-tests/editing/deleting/delete-block-merge-contents-019.html: Added.
        * layout-tests/editing/deleting/delete-block-merge-contents-020-expected.txt: Added.
        * layout-tests/editing/deleting/delete-block-merge-contents-020.html: Added.
        * layout-tests/editing/deleting/delete-block-merge-contents-021-expected.txt: Added.
        * layout-tests/editing/deleting/delete-block-merge-contents-021.html: Added.
        * layout-tests/editing/deleting/delete-line-014-expected.txt: These new results are actually better, and fix a bug.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8341 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed: <rdar://problem/3934749> assertion failure in WebBaseNetscapePluginView loadi...
cblu [Wed, 12 Jan 2005 00:49:51 +0000 (00:49 +0000)]
Fixed: <rdar://problem/3934749> assertion failure in WebBaseNetscapePluginView loading movie

        Reviewed by john.

        * Plugins.subproj/WebBaseNetscapePluginView.m:
        (-[WebBaseNetscapePluginView start]): call canStart before asserting about the webView

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8340 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed 3922875. Fall back to DOM object is EMBED element
rjw [Wed, 12 Jan 2005 00:30:16 +0000 (00:30 +0000)]
Fixed 3922875.  Fall back to DOM object is EMBED element
has no associated runtime object.

        Reviewed by Chris.

        * khtml/ecma/kjs_dom.cpp:
        (KJS::getRuntimeObject):
        * khtml/ecma/kjs_html.cpp:
        (KJS::HTMLDocument::tryGet):
        (KJS::HTMLElement::tryGet):
        (KJS::HTMLCollection::tryGet):
        (KJS::HTMLCollection::getNamedItems):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8339 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix for 3882299, missing content on gibson.com. Change our handling of ...
hyatt [Tue, 11 Jan 2005 23:46:03 +0000 (23:46 +0000)]
    Fix for 3882299, missing content on gibson.com.  Change our handling of " and ' in certain states of the parser to match
other browsers.

        Reviewed by Maciej

        * khtml/html/htmltokenizer.cpp:
        (khtml::HTMLTokenizer::parseTag):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8338 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
sullivan [Tue, 11 Jan 2005 23:43:39 +0000 (23:43 +0000)]
    Reviewed by Darin.

        - fixed <rdar://problem/3446838> REGRESSION (Mail): text decorations don't print
        (e.g. <strike>, underline)

        * WebCoreSupport.subproj/WebTextRenderer.m:
        (-[WebTextRenderer drawLineForCharacters:yOffset:withWidth:withColor:]):
        This bottleneck routine for drawing a line was setting the linewidth to 0 when
        the graphics context was not drawing to the screen. Thus, no lines. Now links
        are underlined when printing from Safari (as well as Mail).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8337 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed 3949145. CG has a much faster API for drawing lines.
rjw [Tue, 11 Jan 2005 22:50:01 +0000 (22:50 +0000)]
Fixed 3949145.  CG has a much faster API for drawing lines.
Switched over to that new API (CGContextStrokeLineSegments).

        Reviewed by John Sullivan.

        * WebCoreSupport.subproj/WebTextRenderer.m:
        (-[WebTextRenderer drawLineForCharacters:yOffset:withWidth:withColor:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8336 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed: <rdar://problem/3930733> Mail prints second page of email blank
cblu [Tue, 11 Jan 2005 22:44:53 +0000 (22:44 +0000)]
Fixed: <rdar://problem/3930733> Mail prints second page of email blank

        Reviewed by dave.

        * khtml/rendering/render_canvas.cpp:
        (RenderCanvas::paint): cache the print rect since the dirty rect can get changed during printing
        * khtml/rendering/render_flow.cpp:
        (RenderFlow::paintLines): removed null check since the print rect should never be null
        * khtml/rendering/render_list.cpp:
        (RenderListMarker::paint): ditto
        * kwq/KWQKHTMLPart.mm:
        (KWQKHTMLPart::adjustPageHeight): don't set the print rect here since this method is not called for the last page

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8335 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed: <rdar://problem/3948862> REGRESSION: missing images when RTFD is pasted into...
cblu [Tue, 11 Jan 2005 21:25:03 +0000 (21:25 +0000)]
Fixed: <rdar://problem/3948862> REGRESSION: missing images when RTFD is pasted into editable WebView

This problem occurred because we were creating image elements before creating corresponding image resources. The fix is to have AppKit call us back to create the resources before it creates the elements.

        Reviewed by john.

        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView _documentFragmentFromPasteboard:allowPlainText:]): don't deal with subresources since that's now done by the following method
        (-[WebHTMLView resourceForData:preferredFilename:]): new handler method called by AppKit

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8334 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Hyatt
kocienda [Mon, 10 Jan 2005 23:40:45 +0000 (23:40 +0000)]
    Reviewed by Hyatt

        Fix for this bug:

        <rdar://problem/3943648> extra line is inserted after pressing return within quoted text of reply

        * khtml/editing/htmlediting.cpp:
        (khtml::InsertParagraphSeparatorInQuotedContentCommand::doApply): This bug was the result of a
        simple coding mistake. A local variable was erroneously redefined in a deeper scope, and so the
        result calculated in that deeper scope was not available when tested.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8333 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John
kocienda [Mon, 10 Jan 2005 23:28:56 +0000 (23:28 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3946852> Option-e goes to next line

        * khtml/editing/htmlediting.cpp:
        (khtml::ReplaceSelectionCommand::doApply): Tweak case used to determine when merging content
        into the start line is done. We plan to change pretty substantially soon to better handle
        the problem described in <rdar://problem/3937352> Quote level not maintained when copied
        and pasted within a Mail message. In the meantime, this change does no harm, and fixes the bug.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8332 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin
kocienda [Mon, 10 Jan 2005 18:47:43 +0000 (18:47 +0000)]
    Reviewed by Darin

        Fix for this bug:

        <rdar://problem/3907005> Applying block styles to a line of text can unexpectedly affect other lines

        * khtml/editing/htmlediting.cpp:
        (khtml::CompositeEditCommand::moveParagraphContentsToNewBlockIfNecessary): New function which moves "paragraphs"
        to their own blocks if needed so that a block style can be applied.
        (khtml::CompositeEditCommand::isMailBlockquote): Moved this function to base class so it can be used more generally.
        (khtml::ApplyStyleCommand::applyBlockStyle): Pass a node instead of a block to addBlockStyleIfNeeded().
        (khtml::ApplyStyleCommand::addBlockStyleIfNeeded): Now accepts a node instead of a block for styling.
        This function also now calls moveParagraphContentsToNewBlockIfNecessary() to make sure that only the current
        paragraph is styled.
        (khtml::SetNodeAttributeCommand::doUnapply): Should not assert on undo if old value of attribute was null.
        Should remove attributue instead.
        * khtml/editing/htmlediting.h: Touch function declarations accordingly.
        * layout-tests/editing/style/create-block-for-style-001-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-001.html: Added.
        * layout-tests/editing/style/create-block-for-style-002-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-002.html: Added.
        * layout-tests/editing/style/create-block-for-style-003-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-003.html: Added.
        * layout-tests/editing/style/create-block-for-style-004-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-004.html: Added.
        * layout-tests/editing/style/create-block-for-style-005-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-005.html: Added.
        * layout-tests/editing/style/create-block-for-style-006-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-006.html: Added.
        * layout-tests/editing/style/create-block-for-style-007-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-007.html: Added.
        * layout-tests/editing/style/create-block-for-style-008-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-008.html: Added.
        * layout-tests/editing/style/create-block-for-style-009-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-009.html: Added.
        * layout-tests/editing/style/create-block-for-style-010-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-010.html: Added.
        * layout-tests/editing/style/create-block-for-style-011-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-011.html: Added.
        * layout-tests/editing/style/create-block-for-style-012-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-012.html: Added.
        * layout-tests/editing/style/create-block-for-style-013-expected.txt: Added.
        * layout-tests/editing/style/create-block-for-style-013.html: Added.

        Unrelated updates to these expected results.
        * layout-tests/editing/inserting/insert-div-007-expected.txt
        * layout-tests/editing/pasteboard/paste-text-013-expected.txt

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8331 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed: <rdar://problem/3936844> Mail: Messages with rich text do not print
cblu [Mon, 10 Jan 2005 18:38:39 +0000 (18:38 +0000)]
Fixed: <rdar://problem/3936844> Mail: Messages with rich text do not print

        Reviewed by john.

        * khtml/rendering/render_flow.cpp:
        (RenderFlow::paintLines): don't do pagination work if printRect is not set
        * khtml/rendering/render_list.cpp:
        (RenderListMarker::paint): ditto

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8330 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
harrison [Mon, 10 Jan 2005 18:22:06 +0000 (18:22 +0000)]
    Reviewed by Darin.

        * kwq/KWQTextUtilities.mm:
        (currentTextBreakLocaleID):
        Return empty string (AKA root locale) if locale pref can not be canonicalized.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8329 268f45cc-cd09-0410-ab3c-d52691b4dbfc