15 years ago Reviewed by Chris Blumenberg.
sullivan [Wed, 22 Jun 2005 21:32:09 +0000 (21:32 +0000)]
    Reviewed by Chris Blumenberg.

        - fixed <rdar://problem/3764645> please add a way to allow WebKit clients to override the WebPDFView context menu

        * PublicHeaderChangesFromTiger.txt: Added.
        New file to keep track of changes made to public headers that haven't been through API review yet.
        Initially lists the WebMenuItem enum tags added to WebUIDelegate.h as part of this change.

        * WebView.subproj/WebDefaultContextMenuDelegate.m:
        (-[WebDefaultUIDelegate appendDefaultItems:toArray:]):
        new method, handles initial separator
        (-[WebDefaultUIDelegate contextMenuItemsForElement:defaultMenuItems:]):
        now has defaultMenuItems: parameter. Any menu items in this array are appended at the end of
        the standard set.
        (-[WebDefaultUIDelegate editingContextMenuItemsForElement:defaultMenuItems:]):
        (-[WebDefaultUIDelegate webView:contextMenuItemsForElement:defaultMenuItems:]):
        This had a defaultMenuItems parameter before but it was always nil. Now it passes the defaultMenuItems
        parameter on to the two methods that construct lists (one for editing, the other for viewing). Also
        tweaked variable name and type for clarity.

        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView menuForEvent:]):
        passes nil for new defaultItems parameter of _menuForElement:

        * WebView.subproj/WebImageView.m:
        (-[WebImageView menuForEvent:]):

        * WebView.subproj/WebTextView.m:
        (-[WebTextView menuForEvent:]):

        * WebView.subproj/WebPDFView.m:
        (-[WebPDFView elementAtPoint:]):
        new method to create the element dictionary needed for _menuForElement:defaultItems:. Only supplies the
        webFrame at this point.
        (-[WebPDFView _menuItemsFromPDFKitForEvent:]):
        new method to return copies of the menu items that PDFKit would include in the context menu, with
        WebKit tags applied
        (-[WebPDFView menuForEvent:]):
        now calls standard WebKit context menu mechanism, so clients' delegates can modify the context menu as
        desired. The initial set of items are the ones WebKit was already displaying for PDF context menus.

        * WebView.subproj/WebUIDelegate.h:
        added enum values for the menu items in the PDF context menu

        * WebView.subproj/WebViewPrivate.h:
        * WebView.subproj/WebView.m:
        (-[WebView _menuForElement:defaultItems:]):
        Added the defaultItems: parameter to this method, which is then passed along to WebDefaultUIDelegate.
        All callers pass nil except for WebPDFView, at least for now.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9471 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Vicki.
darin [Wed, 22 Jun 2005 21:05:59 +0000 (21:05 +0000)]
    Reviewed by Vicki.

        Test cases added:
        * layout-tests/fast/css/apple-prefix-expected.txt: Added.
        * layout-tests/fast/css/apple-prefix.html: Added.

        - fixed <rdar://problem/4141650> REGRESSION: Safari RSS slider broken with TOT WebCore

        * khtml/css/parser.y:
        (getPropertyID): Fix lifetime problem on the QString. Also construct the QString
        using length, since the passed-in char * is not necessarily null-terminated. Also do
        the mapping only if it's a "-apple-" prefix -- the old code would do it for, e.g.,
        "-applex-" too. The old code would not successfully map the "apple" in "appleLineClamp"
        to "khtml", and Dave changed things so that the internal name of all CSS attributes
        is "khtml" back in his 2005-05-03 check-in, breaking the use of appleLineClamp in RSS.
        (getValueID): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9470 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoAdded more bug numbers to the change log.
darin [Wed, 22 Jun 2005 18:23:46 +0000 (18:23 +0000)]
Added more bug numbers to the change log.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9467 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed and landed by Darin Adler.
darin [Wed, 22 Jun 2005 18:22:03 +0000 (18:22 +0000)]
    Reviewed and landed by Darin Adler.

        Test cases added:
        * layout-tests/fast/dom/title-text-property-expected.txt: Added
        * layout-tests/fast/dom/title-text-property.html: Added
        * layout-tests/fast/dom/title-text-property-expected-2.txt: Added
        * layout-tests/fast/dom/title-text-property-2.html: Added

        Fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3586>
        Setting and getting title's text attribute does not work

        * khtml/dom/html_head.cpp:
        Call HTMLTitleElementImpl::text() and ::setText().

        * khtml/html/html_headimpl.cpp:
        Pass the current title node to DocumentImpl::setTitle
        and ::removeTitle.

        Return the title.

        Set the new title.

        * khtml/xml/dom_docimpl.cpp:
        * khtml/xml/dom_docimpl.h:
        Only update the title if it's done by setting document.title
        or by changing the child nodes of the title element.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9466 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Change by Anders Carlsson.
darin [Wed, 22 Jun 2005 18:01:02 +0000 (18:01 +0000)]
    Change by Anders Carlsson.

        - added support for dumping title changes

        * DumpRenderTree/DumpRenderTree.m:
        (-[WaitUntilDoneDelegate webView:didReceiveTitle:forFrame:]): Added. Dump title change
        if requested by JavaScript.
        (+[LayoutTestController isSelectorExcludedFromWebScript:]): Added dumpTitleChanges.
        (-[LayoutTestController dumpTitleChanges]): Added, sets flag.
        (dumpRenderTree): Start flag as NO.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9465 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Mitz Pettel.
darin [Wed, 22 Jun 2005 17:39:53 +0000 (17:39 +0000)]
    Changes by Mitz Pettel.
        Reviewed by me.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3633>
          Layout order of bidi level runs is incorrect

        Test cases added:
        * layout-tests/fast/text/international/bidi-L2-run-reordering-expected.txt: Added.
        * layout-tests/fast/text/international/bidi-L2-run-reordering.html: Added.

        * khtml/rendering/bidi.cpp: (khtml::RenderBlock::bidiReorderLine): Fixed mistake
        where we'd advance too far after reversing a run.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9464 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - added test case for JavaScriptCore fix
darin [Wed, 22 Jun 2005 17:29:28 +0000 (17:29 +0000)]
    - added test case for JavaScriptCore fix

        * layout-tests/fast/js/string-replace-2-expected.txt: Added.
        * layout-tests/fast/js/string-replace-2.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9463 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Change by Anders Carlsson.
darin [Wed, 22 Jun 2005 17:26:17 +0000 (17:26 +0000)]
    Change by Anders Carlsson.
        Reviewed by me.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3294>
          String.prototype.replace() fails with function as second param

        * kjs/string_object.cpp: (replace): Added code to handle functions.

        * tests/mozilla/expected.html: Updated since ecma_3/RegExp/regress-209067.js is fixed now.

        * tests/mozilla/run-mozilla-tests: Fix a minor coding style issue that leads to a warning each
        time we run the tests.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9462 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Change by Mitz Pettel.
darin [Wed, 22 Jun 2005 16:44:08 +0000 (16:44 +0000)]
    Change by Mitz Pettel.
        Reviewed by me.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3618>
          RTL runs drawn by CG not reversed properly

        * WebCoreSupport.subproj/WebTextRenderer.m: (-[WebTextRenderer _CG_drawRun:style:geometry:]):
        Fix off-by-one mistake in order-swapping loops.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9461 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John Sullivan.
darin [Wed, 22 Jun 2005 16:38:03 +0000 (16:38 +0000)]
    Reviewed by John Sullivan.

        Test cases added:
        * layout-tests/fast/dom/css-set-property-exception-expected.txt: Added.
        * layout-tests/fast/dom/css-set-property-exception.html: Added.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3616>
          RSS search field, Dashboard widgets failing due to CSS exception

        * khtml/ecma/kjs_css.cpp: (KJS::DOMCSSStyleDeclaration::tryPut):
        Ignore exceptions when setting properties.

        * layout-tests/fast/js/exception-linenums.html: Changed test case that
        involved setting style to one that involved setting another property that
        raises an exception when you try to set it (setting prefix on a text node).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9460 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Change by Michael Gaiman.
darin [Wed, 22 Jun 2005 16:20:01 +0000 (16:20 +0000)]
    Change by Michael Gaiman.
        Reviewed by me.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3436>
          Missing implementation of -[NSData(WebNSDateExtras) _webkit_parseRFC822HeaderFields]

        * Misc.subproj/WebNSDataExtras.h: Fixed name of category say NSData, not NSDate.
        * Misc.subproj/WebNSDataExtras.m: (-[NSData _webkit_parseRFC822HeaderFields]): Fixed method name.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9459 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Change proposed by Eric Seidel.
darin [Wed, 22 Jun 2005 15:31:57 +0000 (15:31 +0000)]
    Change proposed by Eric Seidel.

        * WebCore.pbproj/project.pbxproj: Fix build rule to match the one in JavaScriptCore.
        The one here accidentally had hard-coded paths in it and also had the $() vs. ${}
        issue that broke building on Xcode 2.1 for no good reason (it didn't need $() or ${}).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9458 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Patch from KevinBallard <kevin@sb.org>, reviewed by me.
mjs [Wed, 22 Jun 2005 04:38:53 +0000 (04:38 +0000)]
    Patch from KevinBallard <kevin@sb.org>, reviewed by me.

- fix build for Xcode 2.1, use ${} for INPUT_FILE_BASE to fix build with Xcode 2.1

        * WebCore.pbproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9457 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - Fixes <rdar://problem/4114192>.
justing [Wed, 22 Jun 2005 01:15:43 +0000 (01:15 +0000)]
- Fixes <rdar://problem/4114192>.
- Fixes a crash in TOT Mail/Blot if you Edit -> Add Hyperlink on a Selection
   that ends at the end of the document
- Fixes a problem in TOT Mail/Blot where Edit -> Add Hyperlink creates
   a link one character beyond what is expected.

In replacement, during insertion, some special case code was being called inappropriately.

        Reviewed by Harrison

        Test cases added: None, requires manual interaction

        * khtml/editing/replace_selection_command.cpp:
        (khtml::ReplaceSelectionCommand::doApply): In "step 2" of a replacement, we should
        only next() the insertion point if the special pre-insertion-merge actually occured.
        mergeStart used to be the correct check for the occurence of a pre-insertion-merge, but
        a bug fix made this not the case.  A non-null m_lastNodeInserted correctly signals
        that some pre-insertion merging occured.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9456 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoJavaScriptCore:
adele [Wed, 22 Jun 2005 00:48:49 +0000 (00:48 +0000)]

        rolling out fix for http://bugzilla.opendarwin.org/show_bug.cgi?id=3293, since it caused layout test failures.

        * ChangeLog:
        * bindings/objc/objc_runtime.h:
        * bindings/objc/objc_runtime.mm:
        * bindings/runtime_array.cpp:
        * bindings/runtime_array.h:
        * bindings/runtime_object.cpp:
        * bindings/runtime_object.h:
        * kjs/array_instance.h:
        * kjs/array_object.cpp:
        * kjs/function.cpp:
        * kjs/function.h:
        * kjs/object.cpp:
        * kjs/object.h:
        * kjs/object_object.cpp:
        * kjs/object_object.h:
        * kjs/string_object.cpp:
        * kjs/string_object.h:
        * tests/mozilla/expected.html:


        rolling out fix for http://bugzilla.opendarwin.org/show_bug.cgi?id=3293, since it caused layout test failures.

        * ChangeLog:
        * khtml/ecma/kjs_binding.h:
        * khtml/ecma/kjs_css.cpp:
        * khtml/ecma/kjs_css.h:
        * khtml/ecma/kjs_dom.cpp:
        * khtml/ecma/kjs_dom.h:
        * khtml/ecma/kjs_html.cpp:
        * khtml/ecma/kjs_html.h:
        * khtml/ecma/kjs_window.cpp:
        * khtml/ecma/kjs_window.h:
        * layout-tests/fast/js/eval-var-decl-expected.txt: Removed. This test used hasOwnProperty
        * layout-tests/fast/js/eval-var-decl.html: Removed. This test used hasOwnProperty
        * layout-tests/fast/js/hasOwnProperty-expected.txt: Removed.
        * layout-tests/fast/js/hasOwnProperty.html: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9455 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * WebCore.pbproj/project.pbxproj: Switched to a build rule rather than a...
darin [Wed, 22 Jun 2005 00:23:09 +0000 (00:23 +0000)]
    * WebCore.pbproj/project.pbxproj: Switched to a build rule rather than a build phase for the
        .y files -- this gets rid of the problem where modifying the .y file would not cause sufficient

        * khtml/css/parser_wrapper.cpp: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9454 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * JavaScriptCore.pbproj/project.pbxproj: Switched to a build rule rather...
darin [Wed, 22 Jun 2005 00:22:55 +0000 (00:22 +0000)]
    * JavaScriptCore.pbproj/project.pbxproj: Switched to a build rule rather than a build phase for
        .y files -- this gets rid of the problem where modifying the .y file would not cause sufficient

        * kjs/grammar_wrapper.cpp: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9453 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Patch from Anders Carlsson <andersca@mac.com>, reviewed by Darin.
adele [Tue, 21 Jun 2005 23:02:19 +0000 (23:02 +0000)]
    Patch from Anders Carlsson <andersca@mac.com>, reviewed by Darin.

        Fixed: <http://bugzilla.opendarwin.org/show_bug.cgi?id=3450>
        <rdar://problem/3881901> String.replace() method not working when regex pattern contains {n, m}

        * pcre/pcre.c: (pcre_compile): Remember the last char length so it can be subtracted correctly if needed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9452 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Test cases for:
adele [Tue, 21 Jun 2005 23:02:07 +0000 (23:02 +0000)]
    Test cases for:
        <rdar://problem/3881901> String.replace() method not working when regex pattern contains {n, m}

        * layout-tests/fast/js/string_replace-expected.txt: Added.
        * layout-tests/fast/js/string_replace.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9451 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Patch contributed by Niels Leenheer <niels.leenheer@gmail.com>
ggaren [Tue, 21 Jun 2005 22:27:55 +0000 (22:27 +0000)]
    Patch contributed by Niels Leenheer <niels.leenheer@gmail.com>

        - fixed http://bugzilla.opendarwin.org/show_bug.cgi?id=3370
        (Safari can't handle strange frame dimensions)

        Reviewed by Darin.

        * khtml/dom/dom_string.cpp:
        * khtml/dom/dom_string.h:
        * khtml/html/html_imageimpl.cpp:
        * khtml/xml/dom_atomicstring.h:
        * khtml/xml/dom_stringimpl.cpp:
        * khtml/xml/dom_stringimpl.h:

        Test cases added:
        * layout-tests/fast/frames/invalid-expected.txt: Added.
        * layout-tests/fast/frames/invalid.html: Added.
        * layout-tests/fast/frames/valid-expected.txt: Added.
        * layout-tests/fast/frames/valid.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9450 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - fixed <rdar://problem/4155049> javascript function named as "opener" doesn...
ggaren [Tue, 21 Jun 2005 18:44:58 +0000 (18:44 +0000)]
    - fixed <rdar://problem/4155049> javascript function named as "opener" doesn't get called because of window.opener property

        Test cases only, fix is in JavaScriptCore

        Reviewed by cblu.

        Test cases added:

        * layout-tests/fast/js/global-opener-function-expected.txt: Added.
        * layout-tests/fast/js/global-opener-function.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9449 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - fixed <rdar://problem/4155532> 'delete' succeeds on functions
ggaren [Tue, 21 Jun 2005 18:42:30 +0000 (18:42 +0000)]
    - fixed <rdar://problem/4155532> 'delete' succeeds on functions
        - fixed <rdar://problem/4155049> javascript function named as "opener" doesn't get called because of window.opener property

        Reviewed by cblu.

        * kjs/nodes.cpp:
        (FuncDeclNode::processFuncDecl): Functions now have DontDelete and Internal attributes set when appropriate.

        Test cases:
        * tests/mozilla/expected.html: Updated for one new success.
        - see also test case added in WebCore.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9448 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Justin.
harrison [Tue, 21 Jun 2005 18:11:10 +0000 (18:11 +0000)]
    Reviewed by Justin.

        - show class and style attributes for element nodes
        - nil check the node being showTree'd
        - add Position::showTree()

        Test cases added: None needed for this debug-only utility code.

        * khtml/editing/selection.cpp:
        - nil check the node
        * khtml/editing/visible_position.cpp:
        - comments
        - nil check the node
        * khtml/xml/dom_nodeimpl.cpp:
        - show class and style attributes for element nodes
        * khtml/xml/dom_position.cpp:
        * khtml/xml/dom_position.h:
        - add Position::showTree()

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9447 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Vicki Murley
sullivan [Tue, 21 Jun 2005 18:05:57 +0000 (18:05 +0000)]
    Reviewed by Vicki Murley

        - fixed assertion failure Vicki ran into

        * Misc.subproj/WebIconDatabase.m:
        (-[WebIconDatabase _forgetIconForIconURLString:]):
        Handle the case where there are no associated page URLs for the icon URL

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9446 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoJavaScriptCore:
mjs [Tue, 21 Jun 2005 08:25:26 +0000 (08:25 +0000)]

        Reviewed by Darin(first pass) and Hyatt.

        - fixed http://bugzilla.opendarwin.org/show_bug.cgi?id=3576
        (roll in support for "const" keyword from KDE tree)
        - make processVarDecls handle deletability of variables declared
        in an eval block the same as evaluate would
        - make eval() call processVarDecls - needed to match mozilla and
        to make the second change testable

        I started with the KDE implementation of const but I ended up changing it a bit
        to avoid the use of a global variable. Now instead of the global variable it distinguishes
        const and var at the grammar level so the appropriate node can know the right kind of

        Test cases:
        * tests/mozilla/expected.html: Updated for one new test that is
        failing - we used to bail on it entirely because it checks for
        const support before starting.
        - see also test cases added in WebCore

        * kjs/grammar.y: Add rules for const declarations.
        * kjs/keywords.table: Add const keyword.
        * kjs/nodes.cpp:
        (VarDeclNode::VarDeclNode): Add parameter.
        (VarDeclNode::evaluate): Add const support.
        (VarDeclNode::processVarDecls): Add const support.
        (VarStatementNode::execute): Irrelevant change.
        (ForInNode::ForInNode): Tell our variable node that it's a variable.
        * kjs/nodes.h:
        (KJS::VarDeclNode::): Add declaration of type enum, extra constructor parameter.
        (KJS::VarStatementNode::VarStatementNode): Irrelevant change.
        * kjs/function.cpp:
        (KJS::GlobalFuncImp::call): Process var decls before evaluating.


        Reviewed by Darin(first pass) and Hyatt.

        - fixed http://bugzilla.opendarwin.org/show_bug.cgi?id=3576
        (roll in support for "const" keyword from KDE tree)
        - make processVarDecls handle deletability of variables declared
        in an eval block the same as evaluate would
        - make eval() call processVarDecls - needed to match mozilla and
        to make the second change testable

Test cases only, fix is in JavaScriptCore

        Test cases added:
        * layout-tests/fast/js/const-expected.txt: Added.
        * layout-tests/fast/js/const.html: Added.
        * layout-tests/fast/js/eval-var-decl-expected.txt: Added.
        * layout-tests/fast/js/eval-var-decl.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9445 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Chris Blumenberg.
sullivan [Tue, 21 Jun 2005 00:08:37 +0000 (00:08 +0000)]
    Reviewed by Chris Blumenberg.

        - fixed <rdar://problem/4155182> icon database retain counts can be incorrect after removing all icons

        Replaced the concept of "future retain count per page URL" with a simpler concept of "retain count per
        page URL", which is maintained even after the icon is actually loaded (unlike the future retain count).
        The total retain count for an icon is now the sum of the retain counts per page URL along with any retain
        count not associated with a page URL -- this is still needed for some internal housekeeping purposes.

        * Misc.subproj/WebIconDatabasePrivate.h:
        renamed iconURLToURLs -> iconURLToPageURLs for clarity
        renamed URLToIconURL -> pageURLToIconURL for clarity
        renamed futureURLToRetainCount -> pageURLToRetainCount (there's no more "future" aspect)
        renamed iconURLToRetainCount -> iconURLToExtraRetainCount (it now maintains only some of the retain count)

        * Misc.subproj/WebIconDatabase.m:
        (+[WebIconDatabase sharedIconDatabase]):
        updated for name changes only
        (-[WebIconDatabase init]):
        (-[WebIconDatabase iconForURL:withSize:cache:]):
        (-[WebIconDatabase iconURLForURL:]):
        (-[WebIconDatabase retainIconForURL:]):
        just bump the retain count in pageURLToRetainCount, instead of behaving differently based on whether
        an icon had been loaded for this URL; this let me delete the internal method _retainFutureIconForURL:
        (-[WebIconDatabase releaseIconForURL:]):
        decrement the retain count in pageURLToRetainCount, then handle the case where the retain count for
        this page has gone to zero. I deleted the internal method _releaseFutureIconForURL: formerly called here.
        (-[WebIconDatabase removeAllIcons]):
        remove all the code that dealt with retain counts; this operation no longer affects retain counts
        (-[WebIconDatabase _setIconURL:forURL:]):
        remove the code that consolidated multiple retain counts for different page URLs into a single retain
        count; the multiple retain counts are now maintained even after the icon is loaded
        (-[WebIconDatabase _clearDictionaries]):
        updated for name changes only
        (-[WebIconDatabase _loadIconDictionaries]):
        (-[WebIconDatabase _updateFileDatabase]):
        (-[WebIconDatabase _totalRetainCountForIconURLString:]):
        new method, sums the retain counts associated with specific page URLs and the extra retain count
        not associated with specific page URLs
        (-[WebIconDatabase _retainIconForIconURLString:]):
        updated for name changes
        (-[WebIconDatabase _forgetIconForIconURLString:]):
        no longer affects retain counts at all; this is up to callers
        (-[WebIconDatabase _releaseIconForIconURLString:]):
        this now distinguishes the case where the retain count not associated with any page URLs hits
        zero from the case where the total retain count hits zero, and handles both

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9444 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Justin Garcia.
darin [Mon, 20 Jun 2005 22:40:44 +0000 (22:40 +0000)]
    Reviewed by Justin Garcia.

        * Scripts/run-webkit-tests: Report number of tests when they succeed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9443 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Chris Blumenberg.
sullivan [Mon, 20 Jun 2005 22:00:41 +0000 (22:00 +0000)]
    Reviewed by Chris Blumenberg.

        - added support for emptying the icon database

        * Misc.subproj/WebIconDatabase.h: just fixed a typo
        * Misc.subproj/WebIconDatabasePrivate.h:
        added WebPendingPublic category with method removeAllIcons, and
        declared WebIconDatabaseDidRemoveAllIconsNotification string.

        * Misc.subproj/WebIconDatabase.m:
        (-[WebIconDatabase removeAllIcons]):
        new method, removes all known icons from memory and disk. There's one loose end,
        covered by radar bug 4155182, where it's possible for the icon database's retain
        counts to get off after this operation. I plan to fix this next.
        (-[WebIconDatabase _setIconURL:forURL:]):
        just fixed some extra whitespace
        (-[WebIconDatabase _forgetIconForIconURLString:]):
        new method, extracted from _releaseIconForIconURLString
        (-[WebIconDatabase _releaseIconForIconURLString:]):
        now calls extracted method

        * WebKit.exp:
        added _WebIconDatabaseDidRemoveAllIconsNotification

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9442 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoJavaScriptCore:
mjs [Mon, 20 Jun 2005 21:59:39 +0000 (21:59 +0000)]

        Patch from Mark Rowe <opendarwin.org@bdash.net.nz>, reviewed by me.

- fixed http://bugzilla.opendarwin.org/attachment.cgi?id=2483

        Test cases added:
        * tests/mozilla/expected.html: Updated for two fixed tests.
- also added a layout test

* bindings/objc/objc_runtime.h:
        * bindings/objc/objc_runtime.mm:
        * bindings/runtime_array.cpp:
        * bindings/runtime_array.h:
        * bindings/runtime_object.cpp:
        * bindings/runtime_object.h:
        * kjs/array_instance.h:
        * kjs/array_object.cpp:
        * kjs/function.cpp:
        * kjs/function.h:
        * kjs/object.cpp:
        * kjs/object.h:
        * kjs/object_object.cpp:
        * kjs/object_object.h:
        * kjs/string_object.cpp:
        * kjs/string_object.h:


        Patch from Mark Rowe <opendarwin.org@bdash.net.nz>, reviewed by me.

- fixed http://bugzilla.opendarwin.org/attachment.cgi?id=2483

        Test cases added:
        * layout-tests/fast/js/hasOwnProperty-expected.txt: Added.
        * layout-tests/fast/js/hasOwnProperty.html: Added.
- also fixed two mozilla JS tests

        * khtml/ecma/kjs_binding.h:
        * khtml/ecma/kjs_css.cpp:
        * khtml/ecma/kjs_css.h:
        * khtml/ecma/kjs_dom.cpp:
        * khtml/ecma/kjs_dom.h:
        * khtml/ecma/kjs_html.cpp:
        * khtml/ecma/kjs_html.h:
        * khtml/ecma/kjs_window.cpp:
        * khtml/ecma/kjs_window.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9441 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Darin. Layout tests by me. Reviewed by me.
ggaren [Mon, 20 Jun 2005 18:20:04 +0000 (18:20 +0000)]
    Changes by Darin. Layout tests by me. Reviewed by me.

Fixed <rdar://problem/4152454> crash in KJS::ExprStatementNode::execute(KJS::ExecState*) in World Clock widget with TOT.

        Test cases added:
        * layout-tests/fast/dom/gc-4-expected.txt: Added.
        * layout-tests/fast/dom/gc-4.html: Added.

        * khtml/ecma/kjs_html.cpp:
        (KJS::Context2D::mark): mark now (1) checks for NULL pointers and (2)
calls mark on its parent class (DOMObject::mark).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9440 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by both Anders Carlsson and me.
darin [Mon, 20 Jun 2005 05:52:16 +0000 (05:52 +0000)]
    Changes by both Anders Carlsson and me.
        Reviewed by Maciej and me.

        Test cases added:
        * layout-tests/fast/forms/element-order-expected.txt: Added.
        * layout-tests/fast/forms/element-order.html: Added.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3503>
          form.elements[] not order-preserving when elements added via DOM

        * khtml/html/html_formimpl.h: Added formElementIndex.
        * khtml/html/html_formimpl.cpp:
        (DOM::insertIntoVector): Added.
        (DOM::HTMLFormElementImpl::formElementIndex): Added. Computes an appropriate
        index for a form element, given the document position relative to the other elements.
        (DOM::HTMLFormElementImpl::registerFormElement): Use formElementIndex to determine
        where to insert the element in the formElements array.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9439 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Anders Carlsson and me.
darin [Mon, 20 Jun 2005 05:24:27 +0000 (05:24 +0000)]
    Changes by Anders Carlsson and me.
        Reviewed by Maciej and me.

        Test cases added: None. Could not figure out how to automate a test without
        making enhancements to DumpRenderTree.

        - fixed <rdar://problem/4141141> Onselect event is ignored with TEXTAREA and INPUT
          also <http://bugzilla.opendarwin.org/show_bug.cgi?id=3256>

        * khtml/html/html_formimpl.cpp: (DOM::HTMLTextAreaElementImpl::select):
        Remove unneeded explicit call to onSelect() here -- now handled by the renderer.

        * khtml/rendering/render_form.h: Added declarations as needed below.
        * khtml/rendering/render_form.cpp:
        (RenderFormElement::slotSelectionChanged): Add base class implementation that does nothing.
        (RenderLineEdit::RenderLineEdit): Connect the selection changed signal to our slot.
        (RenderLineEdit::slotSelectionChanged): Call onSelect when selection changes to something
        that's a selection (not a caret).
        (RenderTextArea::RenderTextArea): Connect as above.
        (RenderTextArea::slotSelectionChanged): Call as above.

        * kwq/KWQLineEdit.h: Added declarations as needed below.
        * kwq/KWQLineEdit.mm:
        (QLineEdit::QLineEdit): Added selectionChanged signal.
        (QLineEdit::hasSelectedText): Added.

        * kwq/KWQSlot.mm:
        (KWQSlot::KWQSlot): Make selectionChanged work on any form element.
        (KWQSlot::call): ditto.

        * kwq/KWQTextArea.h: Added hasSelection method.
        * kwq/KWQTextArea.mm:
        (-[KWQTextArea textViewDidChangeSelection:]): Added. Sends selectionChanged signal, except if
        we're in the middle of gaining or losing first responder.
        (-[KWQTextArea hasSelection]): Added.
        (-[KWQTextAreaTextView becomeFirstResponder]): Added. Sets flag to tell us we're in a responder
        change (using a count to avoid trouble if we reenter).
        (-[KWQTextAreaTextView resignFirstResponder]): Ditto.
        (-[KWQTextAreaTextView inResponderChange]): Added.

        * kwq/KWQTextEdit.h: Added declarations as needed below.
        * kwq/KWQTextEdit.mm:
        (QTextEdit::QTextEdit): Added selectionChanged signal.
        (QTextEdit::hasSelectedText): Added.

        * kwq/KWQTextField.h: Added declarations as needed below.
        * kwq/KWQTextField.mm:
        (-[KWQTextFieldController initWithTextField:QLineEdit:]): Fixed up to do the right
        thing with the result of the init function we call.
        (-[KWQTextFieldController controlTextDidEndEditing:]): Removed call to setHasFocus:NO
        since it's a little too late by the time the control is called; now handled by the
        text view delegate methods below.
        (-[KWQTextFieldController hasSelection]): Added.
        (-[KWQTextFieldController setHasFocus:]): Sets the new hasFocusAndSelectionSet after
        new focus and selection are both set up.
        (-[KWQTextFieldController textViewDidChangeSelection:]): Added. Sends the selection
        changed signal.
        (-[KWQTextField initWithQLineEdit:]): Fixed up to do the right thing with the result
        of the init function we call.
        (-[KWQTextField textViewDidChangeSelection:]): Added. Calls super and then calls the
        (-[KWQTextField textDidEndEditing:]): Added. Calls setHasFocus:NO and then calls super.
        (-[KWQSecureTextField initWithQLineEdit:]): Same as above, but for secure text fields.
        (-[KWQSecureTextField textViewDidChangeSelection:]): Ditto.
        (-[KWQSecureTextField textDidEndEditing:]): Ditto.
        (-[KWQSearchField initWithQLineEdit:]): Same as above, but for search fields.
        (-[KWQSearchField textViewDidChangeSelection:]): Ditto.
        (-[KWQSearchField textDidEndEditing:]): Ditto.
        (-[KWQTextFieldFormatter init]): Fixed up to do the right thing with the result of
        the init function we call.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9438 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Mitz Pettel
darin [Mon, 20 Jun 2005 01:20:01 +0000 (01:20 +0000)]
Changes by Mitz Pettel
        Reviewed by me.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3466>
          ATSUI text doesn't render at coordinates greater than 32K

        * WebCoreSupport.subproj/WebTextRenderer.m: (-[WebTextRenderer _ATSU_drawRun:style:geometry:]):
        Translate the matrix of the CGContext instead of passing the appropriate coordinates to ATSU.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9437 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * DumpRenderTree/DumpRenderTree.m: (dumpRenderTree): Add more auto-release...
darin [Sun, 19 Jun 2005 06:42:30 +0000 (06:42 +0000)]
    * DumpRenderTree/DumpRenderTree.m: (dumpRenderTree): Add more auto-release pools
        in the hope of making the tool use less memory and run faster.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9436 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Anders Carlsson.
darin [Sun, 19 Jun 2005 06:38:09 +0000 (06:38 +0000)]
    Changes by Anders Carlsson.
        Reviewed by me.

        Test cases added:
        * layout-tests/fast/dom/css-dom-read-expected.txt: Added.
        * layout-tests/fast/dom/css-dom-read.html: Added.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3534>
          CSSMutableStyleDeclarationImpl::item is unimplemented

        * khtml/css/css_valueimpl.cpp: (DOM::CSSMutableStyleDeclarationImpl::item):
        Get the style from the m_values list.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9435 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Eric Seidel.
darin [Sun, 19 Jun 2005 06:21:31 +0000 (06:21 +0000)]
    Reviewed by Eric Seidel.

        * pcre/get.c: (pcre_get_substring): Fix some computations so this works for UTF-16.
        This is unused in the current JavaScriptCore, but still good to fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9434 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - undid a last minute change I made that broke a patch I landed
darin [Sun, 19 Jun 2005 06:03:41 +0000 (06:03 +0000)]
    - undid a last minute change I made that broke a patch I landed
          (layout tests caught the mistake, all should work again now)

        * khtml/html/html_formimpl.cpp:
        (DOM::HTMLLabelElementImpl::formElement): Don't include this when iterating.
        (DOM::HTMLLegendElementImpl::formElement): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9433 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - disabling the applet tests, since they are causing other tests to fail
darin [Sun, 19 Jun 2005 05:50:24 +0000 (05:50 +0000)]
    - disabling the applet tests, since they are causing other tests to fail

        * layout-tests/fast/replaced/applet-hspace-expected.txt: Removed.
        * layout-tests/fast/replaced/applet-hspace.html: Removed.
        * layout-tests/fast/replaced/applet-vspace-expected.txt: Removed.
        * layout-tests/fast/replaced/applet-vspace.html: Removed.
        * layout-tests/fast/replaced/resources/applet.class: Removed.
        * layout-tests/fast/replaced/resources/applet.java: Removed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9432 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Anders Carlsson.
darin [Sun, 19 Jun 2005 05:44:29 +0000 (05:44 +0000)]
    Changes by Anders Carlsson.
        Reviewed by me.

        Test cases added:
        * layout-tests/fast/dom/resources/script3.js: Added.
        * layout-tests/fast/dom/resources/script4.js: Added.
        * layout-tests/fast/dom/resources/script5.js: Added.
        * layout-tests/fast/dom/script-add-expected.txt: Added.
        * layout-tests/fast/dom/script-add.html: Added.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3526>
          improve support for dynamically added <script> elements

        * khtml/html/html_headimpl.h: Added override of childrenChanged, new evaluateScript
        function, and m_evaluated boolean.
        * khtml/html/html_headimpl.cpp:
        (HTMLScriptElementImpl::HTMLScriptElementImpl): Init m_evaluated to false.
        (HTMLScriptElementImpl::childrenChanged): Added code to evaluate the script the first
        time a child is added (except for nodes created by parsing).
        (HTMLScriptElementImpl::insertedIntoDocument): Refactored to use the new text and
        evaluateScript functions. Also changed to not try to evaluate the script if the element
        is entirely empty. This allows you to declare an empty <script> element and fill the
        content in later using DHTML.
        (HTMLScriptElementImpl::notifyFinished): Refactored to use the new evaluateScript function.
        (HTMLScriptElementImpl::evaluateScript): Added. Checks and sets a new m_evaluated boolean
        to implement the "evaluate each script only once" rule.
        (HTMLScriptElementImpl::text): Replaced bogus implementation with one that works.
        (HTMLScriptElementImpl::setText): Ditto.

        * khtml/dom/html_head.cpp:
        (HTMLScriptElement::text): Replaced bogus implementation with one that calls through to
        the implementation class. This change doesn't really matter since we don't compile this file.
        (HTMLScriptElement::setText): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9431 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * khtml/html/html_formimpl.cpp: (DOM::HTMLLabelElementImpl::formElement):
darin [Sun, 19 Jun 2005 05:31:59 +0000 (05:31 +0000)]
    * khtml/html/html_formimpl.cpp: (DOM::HTMLLabelElementImpl::formElement):
        Oops, checked in a version that didn't even compile! Fixed now.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9430 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Anders Carlsson.
darin [Sun, 19 Jun 2005 05:00:25 +0000 (05:00 +0000)]
    Changes by Anders Carlsson.
        Reviewed by me.

        Test cases added:
        * layout-tests/fast/forms/focus-expected.txt: Added.
        * layout-tests/fast/forms/focus.html: Added.
        * layout-tests/fast/forms/legend-access-key-expected.txt: Added.
        * layout-tests/fast/forms/legend-access-key.html: Added.

        - fixed <rdar://problem/3481846> HTML4: accesskey not supported by HTML legend elements
          also <http://bugzilla.opendarwin.org/show_bug.cgi?id=3407>
        - added focus methods for <label> and <legend> elements
        - fixed crash when using JavaScript-created events that prevented tests from running

        * khtml/ecma/kjs_html.h: Added focus methods.
        * khtml/ecma/kjs_html.cpp: (KJS::KJS::HTMLElementFunction::tryCall): Added bindings for
        focus for <label> and <legend>. (Sorry about the merging, Dave.)

        * khtml/html/html_formimpl.h: Added new functions.
        * khtml/html/html_formimpl.cpp:
        (DOM::HTMLLabelElementImpl::focus): Added.
        (DOM::HTMLLegendElementImpl::formElement): Added.
        (DOM::HTMLLegendElementImpl::focus): Added.
        (DOM::HTMLLegendElementImpl::accessKeyAction): Added.

        * khtml/xml/dom_docimpl.cpp: (DocumentImpl::defaultEventHandler): If there's no QKeyEvent,
        then get the key name from keyIdentifier() instead of the QKeyEvent's unmodifiedText().
        The old code would just do a nil-deref.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9429 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Change by Niels Leenheer (also done by Dave Hyatt on stage at WWDC).
darin [Sun, 19 Jun 2005 03:39:56 +0000 (03:39 +0000)]
    Change by Niels Leenheer (also done by Dave Hyatt on stage at WWDC).
        Reviewed by me.

        Test cases added:
        * layout-tests/fast/replaced/object-align-hspace-vspace-expected.txt: Added.
        * layout-tests/fast/replaced/object-align-hspace-vspace.html: Added.
        * layout-tests/fast/replaced/resources/webkit-background.png: Added.

        * khtml/html/html_objectimpl.cpp:
        (DOM::HTMLObjectElementImpl::mapToEntry): Add cases for ATTR_ALIGN, ATTR_VSPACE and ATTR_HSPACE.
        (DOM::HTMLObjectElementImpl::parseMappedAttribute): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9428 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Change by Finlay Dobbie.
darin [Sun, 19 Jun 2005 02:57:29 +0000 (02:57 +0000)]
    Change by Finlay Dobbie.
        Reviewed by me.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3331>
          10.3.9 Build Failure: NSString may not respond to `+stringWithCString:encoding:'

        * bindings/objc/WebScriptObject.mm: (-[WebScriptObject stringRepresentation]):
        Undo change we did a while back to work around the gcc 3.3 compiler error.
        It no longer seems to happen, and the workaround code was 10.4-specific.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9427 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoOops, forgot these files.
darin [Sun, 19 Jun 2005 02:28:37 +0000 (02:28 +0000)]
Oops, forgot these files.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9426 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Change by Niels Leenheer (also done by Dave Hyatt on stage at WWDC).
darin [Sun, 19 Jun 2005 02:27:36 +0000 (02:27 +0000)]
    Change by Niels Leenheer (also done by Dave Hyatt on stage at WWDC).
        Reviewed by me.

        Test cases added:
        * layout-tests/fast/replaced/applet-hspace-expected.txt: Added.
        * layout-tests/fast/replaced/applet-hspace.html: Added.
        * layout-tests/fast/replaced/applet-vspace-expected.txt: Added.
        * layout-tests/fast/replaced/applet-vspace.html: Added.
        * layout-tests/fast/replaced/resources/applet.class: Added.
        * layout-tests/fast/replaced/resources/applet.java: Added.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3246>
          HSPACE and VSPACE attributes are ignored in APPLET element

        * khtml/html/html_objectimpl.cpp:
        (DOM::HTMLAppletElementImpl::mapToEntry): Add cases for ATTR_VSPACE and ATTR_HSPACE.
        (DOM::HTMLAppletElementImpl::parseMappedAttribute): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9425 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * Scripts/cvs-apply: Improve handling of patches with CR characters in them.
darin [Sun, 19 Jun 2005 02:08:10 +0000 (02:08 +0000)]
    * Scripts/cvs-apply: Improve handling of patches with CR characters in them.
        * Scripts/cvs-unapply: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9424 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - removed some unused code from one of the KWQ headers
darin [Sun, 19 Jun 2005 01:58:36 +0000 (01:58 +0000)]
    - removed some unused code from one of the KWQ headers

        * kwq/KWQKHTMLPartBrowserExtension.h:

        - added a layout test for a bug that's already fixed

        * layout-tests/fast/forms/element-by-name-expected.txt: Added.
        * layout-tests/fast/forms/element-by-name.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9423 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoTiny formatting tweak.
darin [Sun, 19 Jun 2005 01:49:17 +0000 (01:49 +0000)]
Tiny formatting tweak.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9422 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej.
vicki [Fri, 17 Jun 2005 23:02:48 +0000 (23:02 +0000)]
    Reviewed by Maciej.

        Test cases added: None, since this test requires manual interaction.

- fixed <rdar://problem/4086793> Selects wrapped by Span or Label elements
do not resize when options are added via javascript

* khtml/rendering/render_form.cpp:
        (RenderSelect::layout): remove call setting setNeedsLayout to true.  setNeedsLayout
is already true at this point, and this addtional call has the negative effect of
marking m_normalChildNeedsLayout on the element's parent, which keeps layout from
propagating all the way up to the root later on.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9421 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoBug #:
justing [Fri, 17 Jun 2005 21:55:39 +0000 (21:55 +0000)]
Bug #:
Submitted by:
Reviewed by:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9420 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej.
adele [Fri, 17 Jun 2005 20:00:11 +0000 (20:00 +0000)]
    Reviewed by Maciej.

        - fixed http://bugzilla.opendarwin.org/show_bug.cgi?id=3419
        <rdar://problem/4111752> REGRESSION (125-312/412): Safari doesn't display edmunds.com pages first time through

        * khtml/html/htmltokenizer.cpp: (khtml::HTMLTokenizer::parseTag): set 'script' flag when
          parsing <script src="foo" /> case.  In this case, parseTag calls the scriptHandler function which expects
          that flag to be set, and sets it back to false at the appropriate time.

        Test cases added: These still don't test the cached script cases, but they're good regression
        tests for the tokenizer.  We'll have to come up with a good way to test the cached cases later.

        * layout-tests/fast/tokenizer/external-script-document-write_2-expected.txt: Added.
        * layout-tests/fast/tokenizer/external-script-document-write_2.html: Added.
        * layout-tests/fast/tokenizer/resources/empty_script.js: Added.
        * layout-tests/fast/tokenizer/script_extra_close-expected.txt: Added.
        * layout-tests/fast/tokenizer/script_extra_close.html: Added.
        * layout-tests/fast/tokenizer/script_quirk-expected.txt: Added.
        * layout-tests/fast/tokenizer/script_quirk.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9414 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoBug #:
justing [Fri, 17 Jun 2005 18:50:29 +0000 (18:50 +0000)]
Bug #:
Submitted by:
Reviewed by:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9413 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Patch contributed by Anders Carlsson<andersca@mac.com>.
ggaren [Fri, 17 Jun 2005 18:42:48 +0000 (18:42 +0000)]
Patch contributed by Anders Carlsson<andersca@mac.com>.

Fixes: http://bugzilla.opendarwin.org/show_bug.cgi?id=3408

        Reviewed by Darin.

        Test cases added: None. Test requires manual interaction.
Manual test available at: http://bugzilla.opendarwin.org/attachment.cgi?id=2207

        * khtml/html/html_formimpl.cpp: Eliminated HTMLButtonElementImpl::click. HTMLButtonElementImpl now inherits HTMLElementImpl::click.
        * khtml/html/html_formimpl.h: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9412 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Patch contributed by Joos de Valk <webkit@joostdevalk.nl>.
ggaren [Fri, 17 Jun 2005 18:14:27 +0000 (18:14 +0000)]
Patch contributed by Joos de Valk <webkit@joostdevalk.nl>.

Fixes: http://bugzilla.opendarwin.org/show_bug.cgi?id=3408

        Reviewed by Darin.

        Test cases added: None. Test requires manual interaction.
Manual test available at: http://bugzilla.opendarwin.org/attachment.cgi?id=2207

        * khtml/html/html_formimpl.cpp: Eliminated HTMLButtonElementImpl::click. HTMLButtonElementImpl now inherits HTMLElementImpl::click.
        * khtml/html/html_formimpl.h: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9410 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoBug #:
rjw [Fri, 17 Jun 2005 17:09:36 +0000 (17:09 +0000)]
Bug #:
Submitted by:
Reviewed by:
Changes by Mitz Pettel
        Reviewed by Richard Williamson.

Fixed http://bugzilla.opendarwin.org/show_bug.cgi?id=3545

        * WebCoreSupport.subproj/WebTextRenderer.m:
        (-[WebTextRenderer _ATSU_pointToOffset:style:position:reversed:includePartialGlyphs:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9409 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by John.
darin [Fri, 17 Jun 2005 16:41:05 +0000 (16:41 +0000)]
    Reviewed by John.

        - fixed http://bugzilla.opendarwin.org/show_bug.cgi?id=3556
          black diamond question mark shown for invalid UTF-8 sequences

        Test cases added:
        * layout-tests/fast/encoding/invalid-UTF-8-expected.txt: Added.
        * layout-tests/fast/encoding/invalid-UTF-8.html: Added.

        * kwq/KWQTextCodec.mm:
        (unwanted): Added. Returns true for BOM, replacement, and null characters.
        (KWQTextDecoder::appendOmittingUnwanted): Renamed from appendOmittingNullsAndBOMs and
        changed to use the new "unwanted" function, which causes it to omit replacement characters.
        (KWQTextDecoder::convertUsingTEC): Call append function by its new name.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9408 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Chris.
sullivan [Fri, 17 Jun 2005 15:06:58 +0000 (15:06 +0000)]
    Reviewed by Chris.

        - fixed <rdar://problem/4151001> Reloading javascript-spawned window with no URL erases its contents

        * WebView.subproj/WebFrame.m:
        (-[WebFrame reload]):
        do nothing if URL is zero-length

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9407 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - added prepare-ChangeLog script which we use internally to make ChangeLogs...
mjs [Fri, 17 Jun 2005 07:49:31 +0000 (07:49 +0000)]
    - added prepare-ChangeLog script which we use internally to make ChangeLogs, for
benefit of all WebKit hackerdom

        * Scripts/prepare-ChangeLog: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9406 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoAdded a few debugging methods to highlight Selections, VisiblePositions and Nodes...
justing [Fri, 17 Jun 2005 02:13:12 +0000 (02:13 +0000)]
Added a few debugging methods to highlight Selections, VisiblePositions and Nodes in the DOM tree.

Reviewed by Dave Harrison

Test cases added: None, changes are only to aid debugging

* khtml/editing/selection.cpp:
* khtml/editing/selection.h:
* khtml/editing/visible_position.cpp:
* khtml/editing/visible_position.h:
* khtml/xml/dom_nodeimpl.cpp:
* khtml/xml/dom_nodeimpl.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9405 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed: <rdar://problem/4151759> 'delete' fails on variables declared inside 'eval...
ggaren [Fri, 17 Jun 2005 00:57:37 +0000 (00:57 +0000)]
Fixed: <rdar://problem/4151759> 'delete' fails on variables declared inside 'eval' statements.

        Reviewed by cblu.

        * kjs/context.h:
        (KJS::ContextImp::codeType): Added code type accessor for execution context objects.
        * kjs/internal.cpp:
        (ContextImp::ContextImp): Reflects change to ContextImp::codeType.
        * kjs/nodes.cpp:
        (VarDeclNode::evaluate): Added separate code path for variable declarations inside 'eval' statements.
        * tests/mozilla/expected.html: Updated expected test results to reflect fix.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9404 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFix build bustage
hyatt [Thu, 16 Jun 2005 21:06:46 +0000 (21:06 +0000)]
Fix build bustage

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9403 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoFix build bustage
hyatt [Thu, 16 Jun 2005 20:33:57 +0000 (20:33 +0000)]
Fix build bustage

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9402 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Add an initial implementation of QualifiedName and HTMLNames. These classes...
hyatt [Thu, 16 Jun 2005 19:42:20 +0000 (19:42 +0000)]
    Add an initial implementation of QualifiedName and HTMLNames.  These classes are not used by anyone yet.
Progress being recorded in bugzilla bug 3405.

        Reviewed by mjs

        Test cases added: (None, does not affect build yet)

        * WebCore.pbproj/project.pbxproj:
        * khtml/xml/dom_qname.h,.cpp
        * khtml/html/htmlnames.h,.cpp

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9401 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - recent changes to XcodeOptions made it depend on the current directory
darin [Thu, 16 Jun 2005 16:45:33 +0000 (16:45 +0000)]
    - recent changes to XcodeOptions made it depend on the current directory
          The intent was to have the WebKitBuild directory be next to WebKitTools,
          not inside the various build directories. Workaround for now is to call
          XcodeOptions when the directory is set to the WebKit directory.

        * Scripts/build-dumprendertree: Use a local variable for XcodeOptions and get it
        at the start of the script.
        * Scripts/build-webkit: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9400 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * Scripts/webkitdirs.pm: Recognize all 1.X versions of Xcode as old too,...
darin [Thu, 16 Jun 2005 03:41:24 +0000 (03:41 +0000)]
    * Scripts/webkitdirs.pm: Recognize all 1.X versions of Xcode as old too, not just 2.0.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9399 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago The behavior of <pre> is broken so that the newline
hyatt [Tue, 14 Jun 2005 22:37:11 +0000 (22:37 +0000)]
The behavior of <pre> is broken so that the newline
is not always properly being eaten after the open of the tag.  In addition we were eating the first newline
after all other tags.  This patch stops eating newlines after other tags and refines the behavior for pre
to catch all cases.

        Reviewed by sullivan

        Test cases added: fast/text/whitespace/030.html

        * khtml/html/htmltokenizer.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9396 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Anders Carlsson.
darin [Tue, 14 Jun 2005 18:45:48 +0000 (18:45 +0000)]
    Changes by Anders Carlsson.
        Reviewed by me.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3496>
          Add gdb-safari script to launch Safari under gdb

        * Scripts/gdb-safari: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9395 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Devin Lane.
sullivan [Tue, 14 Jun 2005 18:06:00 +0000 (18:06 +0000)]
    Changes by Devin Lane.
        Reviewed by me.

        - fixed <rdar://problem/3766909> PDF viewing could use a zoom control other than the one in the context menu

        * WebView.subproj/WebPDFView.h:
        now implements protocol _web_WebDocumentTextSizing
        * WebView.subproj/WebPDFView.m:
        (-[WebPDFView _updateScalingToReflectTextSize]):
        new method, sets the PDF scaling from the text size multiplier
        (-[WebPDFView setDataSource:]):
        call _updateScalingToReflectTextSize
        (-[WebPDFView _web_textSizeMultiplierChanged]):
        implementation of protocol _web_WebDocumentTextSizing, calls _updateScalingToReflectTextSize

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9393 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Updated expected.html to reflect fix to <rdar://problem/4147745>.
ggaren [Tue, 14 Jun 2005 17:59:32 +0000 (17:59 +0000)]
Updated expected.html to reflect fix to <rdar://problem/4147745>.

        Reviewed by cblu.

        * tests/mozilla/expected.html:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9392 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Dave Harrison.
sullivan [Tue, 14 Jun 2005 17:36:32 +0000 (17:36 +0000)]
    Reviewed by Dave Harrison.

        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView _delayedEndPrintMode:]):
        After discussing this with Dave, I made this method both more debugger-friendly with asserts for
        the cases we don't think could ever happen, and more paranoid by handling these cases in deployment

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9391 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fixed: <rdar://problem/4147745> JavaScript discards locally defined "arguments"...
ggaren [Tue, 14 Jun 2005 16:36:11 +0000 (16:36 +0000)]
Fixed: <rdar://problem/4147745> JavaScript discards locally defined "arguments" property

No layout tests added because this change fixes existing tests:

        Reviewed by cblu.

        * kjs/function.cpp:
        (KJS::ActivationImp::get): get now checks for an "arguments" property defined in the local variable object
before trying to return the built-in arguments array.

        * kjs/function.h: ActivationImp::put no longer overrides ObjectImp::put

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9390 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - fixed build for Xcode 2.1
darin [Tue, 14 Jun 2005 16:19:16 +0000 (16:19 +0000)]
    - fixed build for Xcode 2.1

        * WebKit.pbproj/project.pbxproj: Use BUILT_PRODUCTS_DIR instead of SYMROOT to search for the
        WebKitSystemInterface.h file. We could re-jigger this again later, but for now this is consistent
        with both the .a file's location and where build-webkit puts the file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9389 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Dave Harrison and Maciej.
sullivan [Tue, 14 Jun 2005 02:12:41 +0000 (02:12 +0000)]
    Reviewed by Dave Harrison and Maciej.

        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView _endPrintMode]):
        new method, extracted from identical code in beginDocument and endDocument. This method must be called
        once to counterbalance the code called from knowsPageRange that turns on "printing mode".
        (-[WebHTMLView _delayedEndPrintMode:]):
        new method, called from "perform after delay". Checks whether the same print operation is still underway
        and, if so, delays further. Otherwise calls _endPrintMode directly.
        (-[WebHTMLView knowsPageRange:]):
        after turning on "printing mode", queue up a delayed call to _delayedEndPrintMode:. If there's an early
        error in the print mechanism such that beginDocument is never called, this will cleanly end "printing
        mode" and make the webview usable again.
        (-[WebHTMLView beginDocument]):
        cancel any delayed call to _delayedEndPrintMode:. If we get this far along in printing, then we don't need
        the failsafe call to _delayedEndPrintMode: that was set up in knowsPageRange:. Also, call extracted method.
        (-[WebHTMLView endDocument]):
        call extracted method

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9388 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Darin.
vicki [Tue, 14 Jun 2005 00:29:40 +0000 (00:29 +0000)]
    Reviewed by Darin.

No test cases added since the steps to reproduce require use of showModalDialog.

- fix for <rdar://problem/4145910> parameters not working in showModalDialog
test page (dialogWidth, dialogHeight, etc.)

        * khtml/ecma/kjs_window.cpp:
        (KJS::parseFeatures): add a check for colon character in the feature string

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9380 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Chris Blumenberg and Adele.
mjs [Tue, 14 Jun 2005 00:25:18 +0000 (00:25 +0000)]
    Reviewed by Chris Blumenberg and Adele.

- better fix for <rdar://problem/4142247> REGRESSION: List to browse widgets at Apple website failed. Closing tab afterwards caused Safari crash

With this change and the matching WebKit change we'll still stop loading the moment you click
a download link, but the unload event and detaching of event handlers will not happen early any more.

        * WebView.subproj/WebDataSource.m:
        (-[WebDataSource _stopLoadingInternal]): call stopLoading on bridge instead of closeURL.
        * WebView.subproj/WebFrame.m:
        (-[WebFrame _transitionToCommitted:]): Revert previous attempt at fix.
        (-[WebFrame _continueLoadRequestAfterNavigationPolicy:formState:]): ditto
        (-[WebFrame stopLoading]): ditto

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9379 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Chris Blumenberg and Adele.
mjs [Tue, 14 Jun 2005 00:23:45 +0000 (00:23 +0000)]
    Reviewed by Chris Blumenberg and Adele.

- better fix for <rdar://problem/4142247> REGRESSION: List to browse widgets at Apple website failed. Closing tab afterwards caused Safari crash

With this change and the matching WebKit change we'll still stop loading the moment you click
a download link, but the unload event and detaching of event handlers will not happen early any more.

        * khtml/khtml_part.cpp:
        (KHTMLPart::stopLoading): Factored out from closeURL, make firing of unload
and detaching of event handlers optional.
        (KHTMLPart::closeURL): call stopLoading(true)
        * khtml/khtml_part.h:
        * kwq/WebCoreBridge.h:
        * kwq/WebCoreBridge.mm:
        (-[WebCoreBridge stopLoading]): Call stopLoading(false).

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9378 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Darin.
cpeterse [Tue, 14 Jun 2005 00:08:31 +0000 (00:08 +0000)]
    Changes by Darin.
        Reviewed by me.

        - fixed problems building deployment due to recent init change

        * WebView.subproj/WebArchive.m: (-[WebArchive initWithCoder:]):
        Put the [super init] call and check for nil outside the exception handler.
        * WebView.subproj/WebResource.m: (-[WebResource initWithCoder:]): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9377 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej.
darin [Mon, 13 Jun 2005 15:33:35 +0000 (15:33 +0000)]
    Reviewed by Maciej.

        - fixed <rdar://problem/4141569> REGRESSION (412+): printing fails on any page with images, and for standalone images
          also <http://bugzilla.opendarwin.org/show_bug.cgi?id=3318>

        * WebCoreSupport.subproj/WebImageData.m: Got rid of use of tabs instead of spaces throughout the file.
        (-[WebImageData _checkSolidColor:]): Wrap use of NSGraphicsContext with an autorelease pool.
        (-[WebImageData _fillSolidColorInRect:compositeOperation:context:]): Ditto.
        (-[WebImageData tileInRect:fromPoint:context:]): Ditto.
        (-[WebImageData _PDFDrawFromRect:toRect:operation:alpha:flipped:context:]): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9376 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Patch from Andrew Wellington <proton@wiretapped.net>
mjs [Mon, 13 Jun 2005 07:27:07 +0000 (07:27 +0000)]
Patch from Andrew Wellington <proton@wiretapped.net>
        Reviewed by me.

- fixed http://bugzilla.opendarwin.org/show_bug.cgi?id=3302
(Safari crashes when going to acid2 test page once it's in cache)

        No test cases addded since the repro steps require use of tabbed browsing.

        * khtml/khtml_part.cpp:
        (KHTMLPart::handleFallbackContent): check for nil

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9375 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Nick Zitzmann.
darin [Mon, 13 Jun 2005 03:28:08 +0000 (03:28 +0000)]
    Changes by Nick Zitzmann.
        Reviewed by me.

        - fixed init methods that don't handle return values from the init methods they call

        * WebView.subproj/WebArchive.m:
        (-[WebArchive init]): Use value returned by init, check it for nil too.
        (-[WebArchive initWithMainResource:subresources:subframeArchives:]): Ditto.
        (-[WebArchive _initWithPropertyList:]): Ditto.
        (-[WebArchive initWithCoder:]): Ditto.
        * WebView.subproj/WebClipView.m:
        (-[WebClipView initWithFrame:]): Ditto.
        * WebView.subproj/WebDebugDOMNode.m:
        (-[WebDebugDOMNode initWithName:value:source:children:]): Ditto.
        * WebView.subproj/WebFrame.m:
        (-[WebFormState initWithForm:values:sourceFrame:]): Ditto.
        (-[WebFrame initWithName:webFrameView:webView:]): Ditto.
        * WebView.subproj/WebFrameView.m:
        (-[WebFrameView initWithFrame:]): Ditto.
        * WebView.subproj/WebHTMLRepresentation.m:
        (-[WebHTMLRepresentation init]): Ditto.
        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView initWithFrame:]): Ditto.
        (-[WebTextCompleteController initWithHTMLView:]): Ditto.
        * WebView.subproj/WebImageView.m:
        (-[WebImageView initWithFrame:]): Ditto.
        * WebView.subproj/WebPreferences.m:
        (-[WebPreferences initWithIdentifier:]): Ditto.
        * WebView.subproj/WebRenderNode.m:
        (-[WebRenderNode initWithName:position:rect:view:children:]): Ditto.
        * WebView.subproj/WebResource.m:
        (-[WebResource init]): Ditto.
        (-[WebResource initWithCoder:]): Ditto.
        * WebView.subproj/WebView.m:
        (-[WebViewPrivate init]): Call super init.
        (-[_WebSafeForwarder initWithTarget:defaultTarget:templateClass:]): Use value returned by init, check it for nil too.
        (-[WebView initWithFrame:]): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9374 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Stuart Morgan.
darin [Mon, 13 Jun 2005 03:26:23 +0000 (03:26 +0000)]
    Changes by Stuart Morgan.
        Reviewed by me.

        * Scripts/cvs-abandon: Use cwd instead of `pwd`.
        * Scripts/cvs-apply: Ditto.
        * Scripts/cvs-create-patch: Ditto.
        * Scripts/cvs-unapply: Ditto.
        * Scripts/run-webkit-tests: Ditto.
        * Scripts/webkitdirs.pm: Ditto. Also improve handling when there's no "Configuration" file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9373 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * Scripts/cvs-apply: Handle case of an empty patch better.
darin [Mon, 13 Jun 2005 03:12:54 +0000 (03:12 +0000)]
    * Scripts/cvs-apply: Handle case of an empty patch better.
        * Scripts/cvs-unapply: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9372 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Changes by Michael Gaiman.
darin [Mon, 13 Jun 2005 02:45:26 +0000 (02:45 +0000)]
    Changes by Michael Gaiman.
        Reviewed by me.

        - fixed <http://bugzilla.opendarwin.org/show_bug.cgi?id=3487>
          WebKit no longer builds after configuration supporting build changes

        * Scripts/webkitdirs.pm: Chomp off the result of `pwd`, and don't die when no Configuration file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9371 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - added expected test results for new tests
darin [Mon, 13 Jun 2005 00:55:38 +0000 (00:55 +0000)]
    - added expected test results for new tests

        * layout-tests/fast/selectors/159-expected.txt: Added.
        * layout-tests/fast/selectors/177a-expected.txt: Added.
        * layout-tests/fast/selectors/177b-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9370 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * Scripts/update-webkit: Make this script work when called from any directory.
darin [Mon, 13 Jun 2005 00:10:33 +0000 (00:10 +0000)]
    * Scripts/update-webkit: Make this script work when called from any directory.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9367 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Maciej.
darin [Sun, 12 Jun 2005 22:48:34 +0000 (22:48 +0000)]
    Reviewed by Maciej.
        Includes changes by Stuart Morgan as well as my own.

        * Scripts/build-dumprendertree: Call the new setConfiguration function, and use XcodeOptions instead of
        symrootXcodeOptions to set the -buildstyle option.
        * Scripts/build-webkit: Ditto. Also remove the old way of supporting Xcode 2.1 and the old --debug option.
        The new --development option does the same thing.
        * Scripts/run-safari: Ditto.
        * Scripts/run-webkit-tests: Ditto.
        * Scripts/update-webkit: Ditto.
        * Scripts/set-webkit-configuration: Added. Sets the default configuration to Development or Deployment.

        * Scripts/webkitdirs.pm: use FindBin to find the WebKit directory; works no matter what the current
        directory is when invoking a script. Add code to determine the Xcode version so we can do the right
        thing for 2.0 and 2.1. Change the productDir function to return the appropriate per-configuration
        product directory. Read the default configuration from a file, overridable by a passed-in command-line

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9366 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Reviewed by Chris Blumenberg.
mjs [Sun, 12 Jun 2005 20:27:26 +0000 (20:27 +0000)]
    Reviewed by Chris Blumenberg.

- fixed <rdar://problem/4142247> REGRESSION: List to browse widgets at Apple website failed. Closing tab afterwards caused Safari crash

        * WebView.subproj/WebFrame.m:
        (-[WebFrame _transitionToCommitted:]): Stop loading the non-provisional data
source before swapping in the provisional.
        (-[WebFrame _continueLoadRequestAfterNavigationPolicy:formState:]): Stop only
the provisional load here, we would not want to stop loading if this navigation
later turns into a download or is cancelled before being committed.
        (-[WebFrame stopLoading]): Factored a bit.
        (-[WebFrame _cancelProvisionalLoad]): New method to stop only provisional load,
and cancel any pending policy deicions.
        (-[WebFrame _stopNonProvisionalLoadOnly]): New mthod that stops only the main

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9364 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago * Scripts/cvs-create-patch: Improve handling of directories with mixed CVS...
darin [Sun, 12 Jun 2005 05:56:32 +0000 (05:56 +0000)]
    * Scripts/cvs-create-patch: Improve handling of directories with mixed CVS roots by going into each directory
        to execute the cvs diff commands.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9363 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago - added first cuts at some cvs scripts
darin [Sun, 12 Jun 2005 05:41:20 +0000 (05:41 +0000)]
    - added first cuts at some cvs scripts

            cvs-abandon is for throwing away changes; discards any local changes, reverting to the state in CVS
            cvs-create-patch is for making patches; runs cvs diff with all the right options and handles added/deleted files
            cvs-apply is for applying patches; runs patch and cvs add and cvs rm
            cvs-unapply is for unapplying patches; does the opposite of cvs-apply

          An argument against these is "waste of time if we switch to Subversion", but they should be good for a while.

        * Scripts/cvs-abandon: Added.
        * Scripts/cvs-apply: Added.
        * Scripts/cvs-create-patch: Added.
        * Scripts/cvs-unapply: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9362 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Patch from Nicholas Shanks <contact@nickshanks.com>, by me.
mjs [Sun, 12 Jun 2005 04:07:31 +0000 (04:07 +0000)]
    Patch from Nicholas Shanks  <contact@nickshanks.com>, by me.

- Distinguish between pseudo-elements and pseudo-classes in CSS parsing.

Merge of svn log -v -r 399829 svn://anonsvn.kde.org/home/kde

        Test cases added: Added the following from CSS selector test suite,
unfortunately 159 and 117a do not show any difference with and without the
patch since layout tests do not dump selection style (yet).

        * layout-tests/fast/selectors/159.html: Added.
        * layout-tests/fast/selectors/177a.html: Added.
        * layout-tests/fast/selectors/177b.html: Added.

        * khtml/css/css_base.cpp:
        * khtml/css/css_base.h:
        * khtml/css/cssstyleselector.cpp:
        * khtml/css/parser.y:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9361 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Patch from Timothy Hatcher <timothy@colloquy.info>, reviewed by me.
mjs [Sat, 11 Jun 2005 09:51:01 +0000 (09:51 +0000)]
    Patch from Timothy Hatcher  <timothy@colloquy.info>, reviewed by me.

        Test cases added:
        * layout-tests/fast/dom/children-nodes-expected.txt: Added.
        * layout-tests/fast/dom/children-nodes.html: Added.

        * khtml/html/html_miscimpl.cpp:
        (DOM::HTMLCollectionImpl::traverseNextItem): For "children"
collection, use nextSibling() instead of traverseNextNext() to avoid
doing deep traversal.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9360 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Added missing expected results from the background image patch.
mjs [Sat, 11 Jun 2005 09:41:09 +0000 (09:41 +0000)]
    Added missing expected results from the background image patch.

        * layout-tests/fast/backgrounds/001-expected.txt: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9359 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix for bugzilla bug 3335, add support for the CSS3 indirect adjacent sibling select...
hyatt [Sat, 11 Jun 2005 05:31:17 +0000 (05:31 +0000)]
Fix for bugzilla bug 3335, add support for the CSS3 indirect adjacent sibling selector.  Patch merge
from KHTML tree by Nick Shanks.

        Reviewed by hyatt

        Test cases added: fast/selectors/046.html

        * khtml/css/css_base.cpp:
        * khtml/css/css_base.h:
        * khtml/css/cssstyleselector.cpp:
        * khtml/css/parser.y:
        * layout-tests/fast/selectors/046-expected.txt: Added.
        * layout-tests/fast/selectors/046.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9358 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix for 3237, background image repeats when it shouldn't. The Radar bug is 4005553...
hyatt [Sat, 11 Jun 2005 05:01:18 +0000 (05:01 +0000)]
Fix for 3237, background image repeats when it shouldn't.  The Radar bug is 4005553.  Patch from
Nate Cook.

        Reviewed by darin and hyatt

        Test cases added: fast/backgrounds/001.html

        * khtml/rendering/render_box.cpp:
        * layout-tests/fast/backgrounds/001.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9357 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Merge Allan Jensen's fix for 3236.
hyatt [Sat, 11 Jun 2005 01:58:02 +0000 (01:58 +0000)]
Merge Allan Jensen's fix for 3236.

        Reviewed by hyatt

        Test cases added: None, test case already added for the bug.

        * khtml/css/cssstyleselector.cpp:
        * khtml/rendering/render_style.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9356 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago reviewed by Dave Harrison (first & second drafts) and Darin Adler (third...
sullivan [Sat, 11 Jun 2005 00:42:53 +0000 (00:42 +0000)]
    reviewed by Dave Harrison (first & second drafts) and Darin Adler (third draft)

        - WebKit part of fix for
        <rdar://problem/4145214> REGRESSION (412+): Can't drag URLs from the location bar

        * Misc.subproj/WebNSPasteboardExtras.m:
        (+[NSPasteboard initialize]):
        Reinstate variation of old code that uses CreatePasteboardFlavorTypeName to set up our new pasteboard types.
        The newfangled way didn't work.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9355 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years agoUpdated copies of WebKitSystemInterface.h and libWebKitSystemInterface.a that include...
sullivan [Sat, 11 Jun 2005 00:40:24 +0000 (00:40 +0000)]
Updated copies of WebKitSystemInterface.h and libWebKitSystemInterface.a that include the new functions needed to fix 3446192.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9354 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Change by Mark Rowe <opendarwin.org@bdash.net.nz>.
darin [Fri, 10 Jun 2005 18:02:57 +0000 (18:02 +0000)]
    Change by Mark Rowe <opendarwin.org@bdash.net.nz>.
        Refined and reviewed by me.

        - test for exception file/line number fix

        * layout-tests/fast/js/exception-linenums-expected.txt: Added.
        * layout-tests/fast/js/exception-linenums.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9353 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Change by Mark Rowe <opendarwin.org@bdash.net.nz>.
darin [Fri, 10 Jun 2005 18:02:33 +0000 (18:02 +0000)]
    Change by Mark Rowe <opendarwin.org@bdash.net.nz>.
        Reviewed by me.

        - further improvements to exception file/line number fix

        * kjs/nodes.h: Added setExceptionDetailsIfNeeded function.
        * kjs/nodes.cpp: Updated macros to call the new setExceptionDetailsIfNeeded function.
        (Node::setExceptionDetailsIfNeeded): Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9352 268f45cc-cd09-0410-ab3c-d52691b4dbfc

15 years ago Fix from Carston Guenther for bugzilla bug 3236.
hyatt [Fri, 10 Jun 2005 06:47:36 +0000 (06:47 +0000)]
Fix from Carston Guenther for bugzilla bug 3236.

attr(x) should work even if x is not present on the element.

This is Radar bug #3584466

        Reviewed by hyatt

        Test cases added: fast/css-generated-content/013.html

        * khtml/css/cssstyleselector.cpp:
        * layout-tests/fast/css-generated-content/013-expected.txt: Added.
        * layout-tests/fast/css-generated-content/013.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@9351 268f45cc-cd09-0410-ab3c-d52691b4dbfc