18 years ago Reviewed by Ken.
sullivan [Mon, 15 Nov 2004 22:08:42 +0000 (22:08 +0000)]
    Reviewed by Ken.

        - fixed <rdar://problem/3880075> leak in CSSStyleDeclarationImpl::copyPropertiesInSet,
        seen often in Mail and Blot

        * khtml/css/css_valueimpl.cpp:
        delete temporary list after we're done using it

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8011 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed leak (3879883) that John found. Early return leaked
rjw [Mon, 15 Nov 2004 20:33:49 +0000 (20:33 +0000)]
Fixed leak (3879883) that John found.  Early return leaked
allocated instance.

        Reviewed by John.

        * khtml/css/css_computedstyle.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8010 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed missing retain of image property data.
rjw [Mon, 15 Nov 2004 20:29:17 +0000 (20:29 +0000)]
Fixed missing retain of image property data.

        Reviewed by John.

        * WebCoreSupport.subproj/WebImageData.h:
        * WebCoreSupport.subproj/WebImageData.m:
        (-[WebImageData dealloc]):
        (-[WebImageData _invalidateImages]):
        (-[WebImageData imageAtIndex:]):
        (-[WebImageData propertiesAtIndex:]):
        (-[WebImageData _frameDuration]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8009 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by John
kocienda [Mon, 15 Nov 2004 19:53:33 +0000 (19:53 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3879569> Many leaks in EditCommand mechanism, seen in Mail

        Fixed a couple of object lifetime issues. The EditCommand class used to hold an
        EditCommandPtr to its parent, but this caused a a reference cycle in composite
        commands as the children held a ref to their parent. Now, the parent variable
        is a non-retained reference to an EditCommand *. It would be nice to have a
        weak reference to the parent or even override deref in composite commands (but I
        can't since deref() is not virtual). However, this should be OK since any
        dangling parent pointer is a sign of a bigger object lifetime problem that
        would need to be addressed anyway.

        * khtml/css/css_valueimpl.cpp:
        (CSSStyleDeclarationImpl::CSSStyleDeclarationImpl): Fix bug in constructor that takes a
        QPtrList<CSSProperty> *. List values must be copied into newly-allocated list, rather than
        just assigning the list variable passed in to the local list variable, or the list will be
        * khtml/editing/htmlediting.cpp:
        (khtml::EditCommand::setStartingSelection): No longer call get(). m_parent is no longer a smart pointer.
        (khtml::EditCommand::setEndingSelection): Ditto.
        (khtml::EditCommand::assignTypingStyle): Short-circuit if passed in style is identical to current style.
        Unrelated to the change, but saves some ref's and deref's.
        (khtml::EditCommand::setTypingStyle): No longer call get(). m_parent is no longer a smart pointer.
        * khtml/editing/htmlediting.h: Change m_parent to a EditCommand *. Was an EditCommandPtr. Using an
        EditCommandPtr caused a reference cycle in composite commands as the children held a ref to their parent.
        (khtml::EditCommand::parent): No longer call get(). m_parent is no longer a smart pointer.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8008 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Kevin.
mjs [Mon, 15 Nov 2004 19:50:06 +0000 (19:50 +0000)]
    Reviewed by Kevin.

<rdar://problem/3807080> Safari so slow it seems like a hang accessing a page on an IBM website

        * khtml/xml/dom_nodeimpl.cpp:
        (NodeListImpl::NodeListImpl): Initialize isItemCacheValid, renamed isCacheValid to
        (NodeListImpl::recursiveLength): Adjusted for rename.
        (NodeListImpl::recursiveItem): Cache the last item accessed and its offset.
If the same offset is looked up again, just return it, otherwise, if looking up
a later offset, start at the last item and proceed from there.
        (NodeListImpl::itemById): Apply the special document optimization to all
nodes that are either a document or in a document - just walk up to make
sure the node found by ID has the root node as an ancestor.
        (NodeListImpl::rootNodeSubtreeModified): Adjust both cache bits.
        * khtml/xml/dom_nodeimpl.h: Prototype new stuff.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8007 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Cache image properties and frame durations.
rjw [Mon, 15 Nov 2004 19:16:30 +0000 (19:16 +0000)]
Cache image properties and frame durations.
Create NSImage and TIFF representations from CGImage, lazily, as needed for
dragging and element info dictionary.

        Reviewed by John.

        * WebCoreSupport.subproj/WebImageData.h:
        * WebCoreSupport.subproj/WebImageData.m:
        (-[WebImageData dealloc]):
        (-[WebImageData size]):
        (-[WebImageData propertiesAtIndex:]):
        (-[WebImageData _frameDurationAt:]):
        (-[WebImageData _frameDuration]):
        * WebCoreSupport.subproj/WebImageRenderer.h:
        * WebCoreSupport.subproj/WebImageRenderer.m:
        (-[WebImageRenderer dealloc]):
        (-[WebImageRenderer TIFFRepresentation]):
        (-[WebImageRenderer image]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8006 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by John.
mjs [Mon, 15 Nov 2004 18:53:42 +0000 (18:53 +0000)]
    Reviewed by John.

<rdar://problem/3879226> WebKit needlessly uses extra memory to store icon refcounts as NSNumbers

* Misc.subproj/WebIconDatabase.m:
        (-[WebIconDatabase init]):
        (-[WebIconDatabase _setIconURL:forURL:]):
        (-[WebIconDatabase _retainIconForIconURLString:]):
        (-[WebIconDatabase _releaseIconForIconURLString:]):
        (-[WebIconDatabase _retainFutureIconForURL:]):
        (-[WebIconDatabase _releaseFutureIconForURL:]):
        * Misc.subproj/WebIconDatabasePrivate.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8005 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Ken.
sullivan [Mon, 15 Nov 2004 17:20:58 +0000 (17:20 +0000)]
    Reviewed by Ken.

        - fixed <rdar://problem/3879539> leak of NSString after pasting into editable HTML (e.g. Mail)

        * kwq/KWQKHTMLPart.mm:
        release mutable copy of string after we're done using it

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8004 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Ken.
sullivan [Mon, 15 Nov 2004 16:54:14 +0000 (16:54 +0000)]
    Reviewed by Ken.

        - fixed <rdar://problem/3879513> leak in [WebArchive _propertyListRepresentation] copying HTML to pasteboard

        * WebView.subproj/WebArchive.m:
        (-[WebArchive _propertyListRepresentation]):
        the array holding the subresources was not released after use, oops!

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8003 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by mjs.
kdecker [Mon, 15 Nov 2004 05:18:01 +0000 (05:18 +0000)]
    Reviewed by mjs.

fixed: <rdar://problem/3823038> LEAK: huge leak in DOM::HTMLElementImpl::createContextualFragment(DOM::DOMString const&, bool)

        * khtml/html/html_elementimpl.cpp:
        (HTMLElementImpl::setInnerHTML): uses the ref counting system to deallocate fragments instead of explicitly invoking a destructor.
        (HTMLElementImpl::setOuterHTML): function is responsible for derefing the fragment prior to returning. Now it does.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8002 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Kevin.
mjs [Sun, 14 Nov 2004 02:07:49 +0000 (02:07 +0000)]
    Reviewed by Kevin.

<rdar://problem/3878766> VIP: Program listings pages at directv.com takes 75% of time traversing NodeLists

        * khtml/dom/dom_node.cpp:
        (NodeList::itemById): New method, just forward to impl.
        * khtml/dom/dom_node.h: Prototype it.
        * khtml/ecma/kjs_dom.cpp:
        (DOMNodeList::tryGet): Instead of looping over the whole list to do by-id access,
let the NodeList do it. The NodeList might be able to do it more efficiently.
        * khtml/xml/dom_nodeimpl.cpp:
(NodeListImpl::itemById): Optimize for the case where the NodeList
covers the whole document. In this case, just use getElementById,
then check that the element satisfies the list criteria.
        (ChildNodeListImpl::nodeMatches): Return true only if the node is our child.
        (TagNodeListImpl::TagNodeListImpl): Irrelevant change to reformat initializers.
        * khtml/xml/dom_nodeimpl.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8001 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Gramps.
mjs [Sat, 13 Nov 2004 02:54:31 +0000 (02:54 +0000)]
    Reviewed by Gramps.

- fixed another bug in the last checkin, isCacheValid was unitialized, resulting in
sometimes using a huge bogus length value.

        * khtml/xml/dom_nodeimpl.cpp:
        (NodeListImpl::NodeListImpl): Initialize isCacheValid.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@8000 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by NOBODY (OOPS!).
mjs [Sat, 13 Nov 2004 02:53:52 +0000 (02:53 +0000)]
    Reviewed by NOBODY (OOPS!).

- fixed another bug in the last checkin, isCacheValid was unitialized, resulting in
sometimes using a huge bogus length value.

        * khtml/xml/dom_nodeimpl.cpp:
        (NodeListImpl::NodeListImpl): Initialize isCacheValid.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7999 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Maciej.
darin [Sat, 13 Nov 2004 02:12:28 +0000 (02:12 +0000)]
    Reviewed by Maciej.

        - fixed an infinite loop in that last check-in

        * khtml/xml/dom_nodeimpl.cpp: (NodeImpl::notifyLocalNodeListsSubtreeModified):
        Added a ++i to the loop so it won't get stuck on the first element in the list.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7998 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Kevin.
mjs [Sat, 13 Nov 2004 01:12:18 +0000 (01:12 +0000)]
    Reviewed by Kevin.

- fixed <rdar://problem/3878183> Safari is 77% slower than it should be on a page on an IBM website due to NodeListImpl length

I fixed this by changing NodeLists to cache their length, but
invalidate it whenever there is a change in the DOM subtree at
which they are rooted. This makes NodeListImpl::recursiveLength()
drop completely off the profile, since we were repeatedly getting
a length for the same NodeList over and over.

        * khtml/xml/dom_nodeimpl.cpp:
        * khtml/xml/dom_nodeimpl.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7997 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Maciej.
darin [Fri, 12 Nov 2004 22:04:26 +0000 (22:04 +0000)]
    Reviewed by Maciej.

        - various small cleanups

        * khtml/xml/dom_docimpl.h: Added policyBaseURL and setPolicyBaseURL.
        * khtml/html/html_documentimpl.h: Removed policyBaseURL and setPolicyBaseURL.

        * khtml/xml/xml_tokenizer.h: Marked isWaitingForScripts const.
        * khtml/xml/xml_tokenizer.cpp: (khtml::XMLTokenizer::isWaitingForScripts): Marked const.
        * khtml/html/htmltokenizer.h: Marked isWaitingForScripts const.
        * khtml/html/htmltokenizer.cpp:
        (khtml::HTMLTokenizer::isWaitingForScripts): Marked const.
        (khtml::HTMLTokenizer::setOnHold): Took out extraneous line of code.

        * khtml/khtml_part.h: Removed docImpl function.
        * khtml/khtml_part.cpp: Ditto.

        * khtml/xml/dom_docimpl.cpp: (DocumentImpl::close): Simplified code that implements
        the "redirect during onload" optimization. Now uses isScheduledLocationChangePending.

        * kwq/KWQKHTMLPart.h: Removed now-unused _firstResponderAtMouseDownTime.
        * kwq/KWQKHTMLPart.mm: Removed _firstResponderAtMouseDownTime (forgot to land this
        part of the change last time, which is why the build broke).
        (KWQKHTMLPart::updatePolicyBaseURL): Use xmlDocImpl instead of docImpl.
        (KWQKHTMLPart::setPolicyBaseURL): Ditto.
        (KWQKHTMLPart::keyEvent): Ditto.
        (KWQKHTMLPart::dispatchCPPEvent): Ditto.
        (KWQKHTMLPart::bodyBackgroundColor): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7996 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoWebKit:
cblu [Fri, 12 Nov 2004 20:49:40 +0000 (20:49 +0000)]

Fixed: <rdar://problem/3874577> Opening restricted (parental) content in new window/tab reveals Safari's "Resources" folder

        Reviewed by john.

        * WebView.subproj/WebDefaultContextMenuDelegate.m:
        (-[WebDefaultUIDelegate openFrameInNewWindow:]): use the unreachable URL if there is one


Fixed: <rdar://problem/3874577> Opening restricted (parental) content in new window/tab reveals Safari's "Resources" folder

        Reviewed by john.

        * ContextMenuHandler.m:
        (-[BrowserWebView openFrameInNewTab:]): use the unreachable URL if there is one

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7995 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago <rdar://problem/3843312> REGRESSION: Tabbing into content area puts insertion point...
cblu [Fri, 12 Nov 2004 20:19:38 +0000 (20:19 +0000)]
<rdar://problem/3843312> REGRESSION: Tabbing into content area puts insertion point at start, should go to where it last was

        Reviewed by rjw.

        * kwq/KWQKHTMLPart.mm:
        (KWQKHTMLPart::nextKeyViewInFrameHierarchy): only blow away selection when another view is focused

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7994 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoversioning for TOT, Safari 2.0 (171+)
vicki [Fri, 12 Nov 2004 20:06:33 +0000 (20:06 +0000)]
versioning for TOT, Safari 2.0 (171+)

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7993 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoSafari-171 stamp
vicki [Fri, 12 Nov 2004 20:00:28 +0000 (20:00 +0000)]
Safari-171 stamp

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7991 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Kevin.
darin [Fri, 12 Nov 2004 18:35:37 +0000 (18:35 +0000)]
    Reviewed by Kevin.

        - fixed a couple places that would not work for XML documents

        * khtml/ecma/kjs_window.cpp:
        (Window::isSafeScript): Use xmlDocImpl instead of docImpl, since the function we're using
        is present on the base class.
        (WindowFunc::tryCall): More of the same.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7990 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago - land versions of these files generated by the newer gperf
darin [Fri, 12 Nov 2004 17:58:58 +0000 (17:58 +0000)]
    - land versions of these files generated by the newer gperf

        People building on Panther will continue to see these files modified.
        A workaround would be to install the newer gperf on our Tiger build machines.

        * khtml/css/cssproperties.c: Regenerated.
        * khtml/css/cssvalues.c: Regenerated.
        * khtml/html/doctypes.cpp: Regenerated.
        * khtml/html/kentities.c: Regenerated.
        * khtml/misc/htmlattrs.c: Regenerated.
        * khtml/misc/htmltags.c: Regenerated.
        * kwq/KWQColorData.c: Regenerated.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7989 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Report actual size (not partial size) but use partial size
rjw [Fri, 12 Nov 2004 02:55:18 +0000 (02:55 +0000)]
    Report actual size (not partial size) but use partial size
        when drawing.

        Reviewed by Maciej.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7988 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fix build horkage from previous checkin.
rjw [Fri, 12 Nov 2004 02:22:20 +0000 (02:22 +0000)]
    Fix build horkage from previous checkin.

        * kwq/KWQKHTMLPart.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7987 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by John.
darin [Fri, 12 Nov 2004 01:49:22 +0000 (01:49 +0000)]
    Reviewed by John.

        - fixed <rdar://problem/3846152> REGRESSION (125-166): can't drag text out of <input type=text> fields

        * kwq/WebCoreBridge.h: Added wasFirstResponderAtMouseDownTime: method.

        * kwq/KWQKHTMLPart.h: Removed _firstResponderAtMouseDownTime.
        * kwq/KWQKHTMLPart.mm:
        (KWQKHTMLPart::passWidgetMouseDownEventToWidget): Use the new wasFirstResponderAtMouseDownTime:
        method on the bridge instead of _firstResponderAtMouseDownTime. This will return YES for the case where
        we started with the NSTextField as first responder, and then took focus away and gave it back, which
        makes dragging text work again.
        (KWQKHTMLPart::mouseDown): Removed code to set _firstResponderAtMouseDownTime.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7986 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by John.
darin [Fri, 12 Nov 2004 01:29:05 +0000 (01:29 +0000)]
    Reviewed by John.

        - added _wasFirstResponderAtMouseDownTime method to bridge so we can fix
          <rdar://problem/3846152> REGRESSION (125-166): can't drag text out of <input type=text> fields
          with a subsequent change to WebCore.

        * WebCoreSupport.subproj/WebBridge.m:
        (wasFirstResponderAtMouseDownTime:): Added. Calls _wasFirstResponderAtMouseDownTime
        on the WebHTMLView.
        (_getPreSmartSet): Move global inside the function, add (void) for cleanliness.
        (_getPostSmartSet): Ditto.

        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLViewPrivate dealloc]): Release firstResponderAtMouseDownTime.
        (-[WebHTMLView _setMouseDownEvent:]): Early exit if event is not changing.
        Set firstResponderAtMouseDownTime to the first responder.
        (-[WebHTMLView mouseDown:]): Release firstResponderAtMouseDownTime after handling
        the mouseDown event.
        (-[WebHTMLView _wasFirstResponderAtMouseDownTime:]): Added. Uses the
        firstResponderAtMouseDownTime field.
        * WebView.subproj/WebHTMLViewInternal.h: Added firstResponderAtMouseDownTime field
        and _wasFirstResponderAtMouseDownTime method.

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

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7985 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Chris.
rjw [Fri, 12 Nov 2004 01:11:15 +0000 (01:11 +0000)]
    Reviewed by Chris.

Work-around to minimize impact of 3876764.  Cache frame durations
after first call.  So we'll still leak 1K for each animated
image, but that's better than 1K each time the frame is drawn!
        * WebCoreSupport.subproj/WebImageData.h:
        * WebCoreSupport.subproj/WebImageData.m:
        (-[WebImageData _frameDuration]):

Simplified animation cleanup code.  Fixed leak due to
incorrect key passed to CFDictionaryRemoveValue.

        (+[WebImageData stopAnimationsInView:]):
        (-[WebImageData addAnimatingRenderer:inView:]):
        (-[WebImageData removeAnimatingRenderer:]):
        (-[WebImageData _stopAnimation]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7984 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Disable the tokenizer deferral, since it hurts the PLT by 3% or so.
hyatt [Thu, 11 Nov 2004 23:55:39 +0000 (23:55 +0000)]
Disable the tokenizer deferral, since it hurts the PLT by 3% or so.

        Reviewed by vicki

        * khtml/html/htmltokenizer.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7983 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Maciej
kocienda [Thu, 11 Nov 2004 23:49:02 +0000 (23:49 +0000)]
    Reviewed by Maciej

        * khtml/editing/htmlediting.cpp:
        (khtml::InsertLineBreakCommand::doApply): Use new isLastVisiblePositionInBlock() helper instead
        of old isLastInBlock() member function on VisiblePosition. This is a cosmetic change in keeping
        with the prevailing style for the VisiblePosition class.
        * khtml/editing/htmlediting.h: Move isLastVisiblePositionInNode() function to visible_position.[cpp|h] files.
        * khtml/editing/visible_position.cpp: Removed isLastInBlock() helper. Renamed to isLastVisiblePositionInBlock().
        (khtml::visiblePositionsInDifferentBlocks): New helper method.
        (khtml::isLastVisiblePositionInBlock): Ditto.
        (khtml::isLastVisiblePositionInNode): Ditto.
        * khtml/editing/visible_position.h: Add declarations for new functions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7982 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Hyatt
kocienda [Thu, 11 Nov 2004 23:15:37 +0000 (23:15 +0000)]
    Reviewed by Hyatt

        * khtml/editing/htmlediting.cpp:
        (khtml::CompositeEditCommand::deleteInsignificantText): Call new compareBoundaryPoints convenience.
        (khtml::ApplyStyleCommand::removeBlockStyle): Ditto.
        (khtml::ApplyStyleCommand::removeInlineStyle): Ditto.
        (khtml::ApplyStyleCommand::nodeFullySelected): Ditto.
        (khtml::DeleteSelectionCommand::initializePositionData): Ditto.
        * khtml/xml/dom2_rangeimpl.cpp:
        (DOM::RangeImpl::compareBoundaryPoints): New convenience variant of this function which takes two Position objects.
        * khtml/xml/dom2_rangeimpl.h: Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7981 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Maciej.
darin [Thu, 11 Nov 2004 22:59:28 +0000 (22:59 +0000)]
    Reviewed by Maciej.

        - fixed <rdar://problem/3605906> Flash scrolled off the top and bottom cause CPU spin when combined with something dirty on the visible part of the page

        * WebView.subproj/WebHTMLView.m: (-[WebHTMLView drawRect:]):
        Work around AppKit bug by using rectangles from getRectsBeingDrawn:count: instead of
        using the passed-in rectangle.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7980 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Harrison
kocienda [Thu, 11 Nov 2004 22:23:23 +0000 (22:23 +0000)]
    Reviewed by Harrison

        Some improvements to deleting when complete lines are selected.

        * khtml/editing/htmlediting.cpp:
        (khtml::DeleteSelectionCommand::initializePositionData): Detect when the line containing
        the end of a selection is fully selected. Turn off block merging in this case.
        (khtml::DeleteSelectionCommand::handleSpecialCaseBRDelete): Fix a bug in the check for
        whether a BR immediately followed a block. The old code could erroneously skip nodes.
        (khtml::DeleteSelectionCommand::handleGeneralDelete): Add a case for when the entire
        start block is selected. This new code will now delete this block in one call, rather
        than iterating over each child.
        * khtml/editing/visible_position.cpp:
        (khtml::visiblePositionsOnDifferentLines): New helper called in initializePositionData()
        to do the work mentioned above in the comment for that function.
        (khtml::isFirstVisiblePositionOnLine): Ditto.
        (khtml::isLastVisiblePositionOnLine): Ditto.
        * khtml/editing/visible_position.h: Add new functions.
        * layout-tests/editing/deleting/delete-line-001-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-001.html: Added.
        * layout-tests/editing/deleting/delete-line-002-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-002.html: Added.
        * layout-tests/editing/deleting/delete-line-003-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-003.html: Added.
        * layout-tests/editing/deleting/delete-line-004-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-004.html: Added.
        * layout-tests/editing/deleting/delete-line-005-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-005.html: Added.
        * layout-tests/editing/deleting/delete-line-006-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-006.html: Added.
        * layout-tests/editing/deleting/delete-line-007-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-007.html: Added.
        * layout-tests/editing/deleting/delete-line-008-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-008.html: Added.
        * layout-tests/editing/deleting/delete-line-009-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-009.html: Added.
        * layout-tests/editing/deleting/delete-line-010-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-010.html: Added.
        * layout-tests/editing/deleting/delete-line-011-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-011.html: Added.
        * layout-tests/editing/deleting/delete-line-012-expected.txt: Added.
        * layout-tests/editing/deleting/delete-line-012.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7979 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoReplaced NOBODY with John.
rjw [Thu, 11 Nov 2004 21:35:51 +0000 (21:35 +0000)]
Replaced NOBODY with John.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7978 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Work-arounds to make new ImageIO code work correctly. Still
rjw [Thu, 11 Nov 2004 21:35:26 +0000 (21:35 +0000)]
    Work-arounds to make new ImageIO code work correctly.  Still
        disabled for now.  Requires at least Tiger 300.  Testing does
        show a 3% improvement in PLT tests!  That's huge!

        Reviewed by NOBODY (OOPS!).

        * WebCoreSupport.subproj/WebImageData.m:
        (-[WebImageData imageAtIndex:]):
        (-[WebImageData incrementalLoadWithBytes:length:complete:]):
        (-[WebImageData isNull]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7977 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Adele
kocienda [Thu, 11 Nov 2004 18:09:51 +0000 (18:09 +0000)]
    Reviewed by Adele

        * khtml/editing/htmlediting.cpp:
        (khtml::DeleteSelectionCommand::initializePositionData): Add some comments and a new piece of debugging output.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7976 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by John
kocienda [Thu, 11 Nov 2004 17:49:56 +0000 (17:49 +0000)]
    Reviewed by John

        Fix for these bugs:

        <rdar://problem/3875618> REGRESSION (Mail): Hitting down arrow with full line selected skips line (br case)
        <rdar://problem/3875641> REGRESSION (Mail): Hitting down arrow with full line selected skips line (div case)

        * khtml/editing/selection.cpp:
        (khtml::Selection::modifyMovingRightForward): Fixed by juggling the position as the starting point for
        the next line position when necessary.
        * layout-tests/editing/selection/move-3875618-fix-expected.txt: Added.
        * layout-tests/editing/selection/move-3875618-fix.html: Added.
        * layout-tests/editing/selection/move-3875641-fix-expected.txt: Added.
        * layout-tests/editing/selection/move-3875641-fix.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7975 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by John
kocienda [Thu, 11 Nov 2004 15:35:56 +0000 (15:35 +0000)]
    Reviewed by John

        Improved some function names, at John's urging. No changes to the
        functions themselves.

        canPerformSpecialCaseAllContentDelete() --> handleSpecialCaseAllContentDelete()
        canPerformSpecialCaseBRDelete() --> handleSpecialCaseBRDelete()
        performGeneralDelete() --> handleGeneralDelete()

        * khtml/editing/htmlediting.cpp:
        * khtml/editing/htmlediting.h

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7974 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by John
kocienda [Thu, 11 Nov 2004 15:28:56 +0000 (15:28 +0000)]
    Reviewed by John

        Updated some layout test results that changed as a result of my last checking.
        Added a new test that has been in my tree for a few days.

        * layout-tests/editing/deleting/delete-3775172-fix-expected.txt:
        * layout-tests/editing/deleting/delete-3800834-fix-expected.txt:
        * layout-tests/editing/inserting/insert-3851164-fix-expected.txt: Added.
        * layout-tests/editing/inserting/insert-3851164-fix.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7973 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by John
kocienda [Thu, 11 Nov 2004 15:26:01 +0000 (15:26 +0000)]
    Reviewed by John

        * khtml/editing/htmlediting.cpp:
        (khtml::debugNode): New debugging helper.
        (khtml::DeleteSelectionCommand::initializePositionData): No longer call obsoleted
        startPositionForDelete() and endPositionForDelete() functions. Just use the
        m_selectionToDelete object to determine start and end positions for the delete.
        (khtml::DeleteSelectionCommand::canPerformSpecialCaseAllContentDelete): New
        function that creates a special case for deleting all the content in a root
        editable element.
        (khtml::DeleteSelectionCommand::doApply): Call canPerformSpecialCaseAllContentDelete()
        function before BR special case and the general case delete functions.
        * khtml/editing/htmlediting.h: Updated for changed functions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7972 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by mjs.
kdecker [Thu, 11 Nov 2004 02:08:31 +0000 (02:08 +0000)]
    Reviewed by mjs.

Fixed <rdar://problem/3875011> DOMNodeList::tryGet() performs unnecessary (and expensive) dom tree traversals. Improved a loop from 2-n-squared to just n-squared.

        * khtml/ecma/kjs_dom.cpp:
        (DOMNodeList::tryGet): Got rid of an unnecessary node traversal.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7971 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed: <rdar://problem/3396872> ICONS: icon DB inconsistencies can cause slowness...
cblu [Wed, 10 Nov 2004 22:39:20 +0000 (22:39 +0000)]
Fixed: <rdar://problem/3396872> ICONS: icon DB inconsistencies can cause slowness during startup, idle and quit

        Reviewed by john.

        * Misc.subproj/WebFileDatabase.m:
        (-[WebFileDatabase objectForKey:]): added logging code
        * Misc.subproj/WebIconDatabase.m:
        (-[WebIconDatabase init]): use alloc, init rather than autorelease, retain
        (-[WebIconDatabase _loadIconDictionaries]): use 1 object for mapping icon URLs to site URLs and vice versa rather than 3. This avoids inconsistencies and is faster.
        (-[WebIconDatabase _updateFileDatabase]): write 1 object out

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7970 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Chris
kocienda [Wed, 10 Nov 2004 22:03:04 +0000 (22:03 +0000)]
    Reviewed by Chris

        * khtml/editing/htmlediting.cpp:
        (khtml::DeleteSelectionCommand::initializePositionData): Move position adjustments for
        smart delete from the two functions below to here. There was an unnecessary double
        calculation of the leading and trailing whitespace positions. Also refined the trailing
        case so it only acts when the leading position is null (which seems to match TextEdit in
        my tests). Also removed some unnecessary copying of Position objects.
        (khtml::DeleteSelectionCommand::startPositionForDelete): Move out smart delete adjustment
        from here.
        (khtml::DeleteSelectionCommand::endPositionForDelete): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7969 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Harrison
kocienda [Wed, 10 Nov 2004 19:04:35 +0000 (19:04 +0000)]
    Reviewed by Harrison

        (khtml::DeleteSelectionCommand::performGeneralDelete): Add some more comments to
        make things more clear.
        * khtml/editing/selection.cpp:
        (khtml::Selection::toRange): Fixed the upstream and downstream calls so that the
        resulting positions do not cross block boundaries. This was a bug and caused some
        delete problems when whole blocks were selected. I will be addressing that issue
        more fully in upcoming changes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7968 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Harrison
kocienda [Wed, 10 Nov 2004 16:58:30 +0000 (16:58 +0000)]
    Reviewed by Harrison

        Some cleanups and fixes in upstream and downstream functions.

        Removed redundant checks for isBlockFlow() when calling enclosingBlockFlowElement().
        Blocks do not need to skip the call to enclosingBlockFlowElement() for fear that the
        block's enclosing block will be returned.

        Remove code from upstream that confined the serach to block boundaries outside of
        the code which runs in the StayInBlock case. This code was redundant, and caused
        incorrect results to be returned in the DoNotStayInBlock case.

        In downstream, the check for crossing into a new block should use the equivalentDeepPosition()
        node, not the the this pointer's node.

        * khtml/xml/dom_position.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7967 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fix for 3873234, Safari UI is unresponsive when parsing multiple HTML docs and 38732...
hyatt [Wed, 10 Nov 2004 03:47:56 +0000 (03:47 +0000)]
Fix for 3873234, Safari UI is unresponsive when parsing multiple HTML docs and 3873233, Safari hangs when
        loading large local files.

        Reviewed by mjs

        * khtml/html/htmltokenizer.cpp:
        * khtml/html/htmltokenizer.h:
        * khtml/khtml_part.cpp:
        * khtml/khtml_part.h:
        * khtml/khtmlview.cpp:
        * khtml/xml/dom_docimpl.cpp:
        * khtml/xml/xml_tokenizer.h:
        * kwq/KWQDateTime.h:
        * kwq/KWQDateTime.mm:
        * kwq/KWQKHTMLPart.h:
        * kwq/KWQKHTMLPart.mm:
        * kwq/WebCoreBridge.h:
        * kwq/WebCoreBridge.mm:
        (-[WebCoreBridge stop]):
        (-[WebCoreBridge numPendingOrLoadingRequests]):
        (-[WebCoreBridge doneProcessingData]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7966 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fix for 3873234, Safari UI is unresponsive when parsing multiple HTML docs and 38732...
hyatt [Wed, 10 Nov 2004 03:46:05 +0000 (03:46 +0000)]
Fix for 3873234, Safari UI is unresponsive when parsing multiple HTML docs and 3873233, Safari hangs when
loading large local files.

        Reviewed by mjs

        * WebCoreSupport.subproj/WebBridge.m:
        (-[WebBridge tokenizerProcessedData]):
        * WebView.subproj/WebDataSource.m:
        (-[WebDataSource _receivedMainResourceError:complete:]):
        (-[WebDataSource isLoading]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7965 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Ken Kocienda.
harrison [Wed, 10 Nov 2004 01:04:41 +0000 (01:04 +0000)]
    Reviewed by Ken Kocienda.

<rdar://problem/3865837> Wrong text style after delete to start of document

        * khtml/editing/htmlediting.cpp:
        Sample computedStyle of m_selectionToDelete.start instead of m_downstreamStart.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7964 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed: <rdar://problem/3872724> soft link against JavaVM to save ~2MB RSHRD
cblu [Wed, 10 Nov 2004 00:28:07 +0000 (00:28 +0000)]
Fixed: <rdar://problem/3872724> soft link against JavaVM to save ~2MB RSHRD

        Reviewed by rjw.

        * ChangeLog:
        * JavaScriptCore.pbproj/project.pbxproj: don't link against JavaVM
        * bindings/softlinking.c: Added.
        (loadFramework): new
        (getFunctionPointer): new
        (JNI_GetCreatedJavaVMs): load JavaVM if not already loaded, get _JNI_GetCreatedJavaVMs symbol if we don't already have it, call JNI_GetCreatedJavaVMs

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7963 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed <rdar://problem/3872440> NSTimer prematurely released.
rjw [Tue, 9 Nov 2004 21:52:20 +0000 (21:52 +0000)]
    Fixed <rdar://problem/3872440> NSTimer prematurely released.

        Reviewed by Darin.

        * kwq/KWQTimer.mm:
        * kwq/KWQWidget.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7962 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed <rdar://problem/3870964> 8A300: Safari not recognizing a PDF link ...
rjw [Tue, 9 Nov 2004 20:57:09 +0000 (20:57 +0000)]
    Fixed <rdar://problem/3870964> 8A300: Safari not recognizing a PDF link (it displays raw data)

Add "text/pdf" as an acceptable PDF MIME type.

        Reviewed by Chris.

        * WebView.subproj/WebDataSource.m:
        (+[WebDataSource _repTypesAllowImageTypeOmission:]):
        * WebView.subproj/WebFrameView.m:
        (+[WebFrameView _viewTypesAllowImageTypeOmission:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7961 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by vicki.
cblu [Tue, 9 Nov 2004 20:52:46 +0000 (20:52 +0000)]
    Reviewed by vicki.

        * WebCore.pbproj/project.pbxproj: explicitly link against libxml2.2.6.14.dylib since the version number has been bumped

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7960 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Ken Kocienda.
harrison [Tue, 9 Nov 2004 03:46:04 +0000 (03:46 +0000)]
    Reviewed by Ken Kocienda.

<rdar://problem/3865854> Deleting first line deletes all lines

        * khtml/editing/htmlediting.cpp:
        Problem was that the code that deletes fully selected m_downstreamEnd.node() by deleting one
        of its ancestors, failed to end the loop that deletes all fully selected nodes.  Also,
        fixed this code to clear m_trailingWhitespaceValid.  Also removed dead m_endingPosition
        update because it is handled in calculateEndingPosition now.
        * layout-tests/editing/deleting/delete-3865854-fix-expected.txt: Added.
        * layout-tests/editing/deleting/delete-3865854-fix.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7959 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Hyatt
kocienda [Tue, 9 Nov 2004 00:32:21 +0000 (00:32 +0000)]
    Reviewed by Hyatt

        * khtml/html/html_elementimpl.cpp:
        (HTMLElementImpl::createContextualFragment): Now takes flag to control whether comments
        are added to the DOM.
        * khtml/html/html_elementimpl.h: Ditto.
        * khtml/html/htmlparser.cpp:
        (KHTMLParser::KHTMLParser): Ditto.
        (KHTMLParser::getElement): Remove ifdef for comment processing. Replace with flag check.
        * khtml/html/htmlparser.h: Add flag to constructor so callers can request comment nodes.
        * khtml/html/htmltokenizer.cpp:
        (khtml::HTMLTokenizer::HTMLTokenizer): Add flag to constructor so callers can request comment nodes.
        (khtml::HTMLTokenizer::parseComment): Fix code to handle parsing out comment text correctly.
        There were a couple of indexing errors that resulted in the comment text containing part of the
        comment markers.
        (khtml::HTMLTokenizer::processToken): Don't let token id be reset to ID_TEXT if token is a comment.
        * khtml/html/htmltokenizer.h:  Add flag to constructor so callers can request comment nodes.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7958 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed: <rdar://problem/3870907> WebCore unnecessary links against JavaVM and Security
cblu [Tue, 9 Nov 2004 00:22:02 +0000 (00:22 +0000)]
Fixed: <rdar://problem/3870907> WebCore unnecessary links against JavaVM and Security

        Reviewed by darin.

        * WebCore.pbproj/project.pbxproj: stop unnecessary linking
        * khtml/html/html_objectimpl.h: don't unnecessarily include JavaVM header
        * kwq/KWQKHTMLPart.h: ditto

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7957 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by John.
darin [Tue, 9 Nov 2004 00:02:11 +0000 (00:02 +0000)]
    Reviewed by John.

        - fixed <rdar://problem/3825966> 8A274 Safari crashes closing window: QTimer::fire() with MallocStackLogging and MallocScribble enabled

        * kwq/KWQTimer.mm: (QTimer::fire): Rearrange so we don't access the QTimer object after calling code
        that possibly deletes the QTimer.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7956 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoWebCore:
cblu [Mon, 8 Nov 2004 19:11:49 +0000 (19:11 +0000)]

Fixed: <rdar://problem/3783904> Return key behavior is confusingly different between popup menus and autofill menus

        Reviewed by john.

        * kwq/KWQTextField.mm:
        (-[KWQTextFieldController textView:shouldHandleEvent:]): let the bridge have a crack at the event so that it can swallow the newline if it wants to
        * kwq/WebCoreBridge.h:


Fixed: <rdar://problem/3783904> Return key behavior is confusingly different between popup menus and autofill menus

        Reviewed by john.

        * WebCoreSupport.subproj/WebBridge.m:
        (-[WebBridge control:textView:shouldHandleEvent:]): new
        * WebView.subproj/WebFormDelegate.h:
        * WebView.subproj/WebFormDelegate.m:
        (-[WebFormDelegate control:textView:shouldHandleEvent:inFrame:]): new


Fixed: <rdar://problem/3783904> Return key behavior is confusingly different between popup menus and autofill menus

        Reviewed by john.

        * BrowserWebController.m:
        (-[FormDelegate control:textView:shouldHandleEvent:inFrame:]): new
        * CompletionController.h:
        * CompletionController.m:
        (-[CompletionController handleInsertNewline]): new, factored out from next method
        (-[CompletionController control:textView:doCommandBySelector:]): call handleInsertNewline
        * FormCompletionController.h:
        * FormCompletionController.m:
        (-[FormCompletionController control:textView:shouldHandleEvent:]): new, call handleInsertNewline so that WebCore does not submit the form when a newline is entered

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7955 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Darin.
harrison [Mon, 8 Nov 2004 18:31:03 +0000 (18:31 +0000)]
    Reviewed by Darin.

        Renamed NodeImpl::enclosingNonBlockFlowElement to NodeImpl::enclosingInlineElement, per Hyatt.

        * khtml/editing/htmlediting.cpp:
        * khtml/xml/dom_nodeimpl.cpp:
        * khtml/xml/dom_nodeimpl.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7954 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed: <rdar://problem/3854218> Safari is sometimes really slow because of increased...
cblu [Sat, 6 Nov 2004 00:45:14 +0000 (00:45 +0000)]
Fixed: <rdar://problem/3854218> Safari is sometimes really slow because of increased null events to plug-ins

        * Plugins.subproj/WebBaseNetscapePluginView.m: reverted null event interval to 0.02

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7953 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed: <rdar://problem/3838413> REGRESSION (Mail): "Smart" word paste adds spaces...
cblu [Sat, 6 Nov 2004 00:36:52 +0000 (00:36 +0000)]
Fixed: <rdar://problem/3838413> REGRESSION (Mail): "Smart" word paste adds spaces before/after special characters

        Reviewed by rjw.

        * khtml/editing/htmlediting.cpp:
        (khtml::ReplaceSelectionCommand::doApply): call isCharacterSmartReplaceExempt on the part to see if a space should be inserted
        * khtml/editing/visible_position.cpp:
        (khtml::VisiblePosition::character): new, returns the character for the position
        * khtml/editing/visible_position.h:
        * kwq/KWQKHTMLPart.h:
        * kwq/KWQKHTMLPart.mm:
        (KWQKHTMLPart::isCharacterSmartReplaceExempt): new, calls the bridge
        * kwq/WebCoreBridge.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7952 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed: <rdar://problem/3838413> REGRESSION (Mail): "Smart" word paste adds spaces...
cblu [Sat, 6 Nov 2004 00:36:12 +0000 (00:36 +0000)]
Fixed: <rdar://problem/3838413> REGRESSION (Mail): "Smart" word paste adds spaces before/after special characters

        Reviewed by rjw.

        * WebCoreSupport.subproj/WebBridge.m:
        (_getPreSmartSet): copied from AppKit
        (_getPostSmartSet): ditto
        (-[WebBridge isCharacterSmartReplaceExempt:isPreviousCharacter:]): new

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7951 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed <rdar://problem/3810702> _checkNavigationPolicyForRequest:dataSource...
rjw [Fri, 5 Nov 2004 23:31:45 +0000 (23:31 +0000)]
    Fixed <rdar://problem/3810702> _checkNavigationPolicyForRequest:dataSource:formState:andCall:withSelector: ASSERTS when reentered from Xcode's man page viewer

        Reviewed by Maciej (a long time ago).

        * WebView.subproj/WebFrame.m:
        (-[WebFrame _loadDataSource:withLoadType:formState:]):

        Fixed <rdar://problem/3845307> WebKit needs to export _HIWebViewRegisterClass so HIWebViews can work in Carbon nib files

        As suggested in the bug, the fix is to actually call
        HIWebViewRegisterClass in WebKitInitForCarbon, rather than
        exporting the symbol.

        Reviewed by Chris.

        * Carbon.subproj/CarbonUtils.m:
        * Carbon.subproj/HIWebView.m:
        * WebKit.pbproj/project.pbxproj:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7950 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoversioning for TOT, Safari 2.0 (170+)
vicki [Fri, 5 Nov 2004 22:11:11 +0000 (22:11 +0000)]
versioning for TOT, Safari 2.0 (170+)

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7949 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoSafari-170 stamp
vicki [Fri, 5 Nov 2004 21:59:13 +0000 (21:59 +0000)]
Safari-170 stamp

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7947 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoBacking out change because tree is closed.
rjw [Fri, 5 Nov 2004 21:14:21 +0000 (21:14 +0000)]
Backing out change because tree is closed.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7946 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed <rdar://problem/3845307> WebKit needs to export _HIWebViewRegisterClass so...
rjw [Fri, 5 Nov 2004 20:38:41 +0000 (20:38 +0000)]
Fixed <rdar://problem/3845307> WebKit needs to export _HIWebViewRegisterClass so HIWebViews can work in Carbon nib files

        Reviewed by Chris.

        * Carbon.subproj/CarbonUtils.m:
        * Carbon.subproj/HIWebView.m:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7945 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Darin
adele [Fri, 5 Nov 2004 18:45:31 +0000 (18:45 +0000)]
    Reviewed by Darin

Fix for <rdar://problem/3854383> REGRESSION(166-168) input fields show black background when background color is set to transparent
and a workaround for displaying transparent backgrounds for textareas.

        * kwq/KWQLineEdit.mm: (QLineEdit::setPalette): If the background color is transparent (we check the alpha value) then we set the background to white
        * kwq/KWQTextEdit.mm: (QTextEdit::setPalette): If the background color is transparent then we don't draw the background
        * kwq/KWQTextArea.mm: (-[KWQTextArea setDrawsBackground:]): added setDrawsBackground function which calls setDrawsBackground on the super class,
          on the contentView, and on the textView.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7944 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Chris.
darin [Fri, 5 Nov 2004 18:00:25 +0000 (18:00 +0000)]
    Reviewed by Chris.

        - fixed <rdar://problem/3857151> Assertion failure in "trackingRectOwner" while moving mouse over Slashdot.org page

        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView _addTrackingRect:owner:userData:assumeInside:useTrackingNum:]): Changed to no longer call
        addTrackingRect to do the work for consistency with the new method below. Not too much copied and pasted code.
        (-[WebHTMLView _addTrackingRects:owner:userDataList:assumeInsideList:trackingNums:count:]): Added an override
        for this new method in Tiger. No harm in implementing it on Panther, although it won't be called.
        (-[WebHTMLView _removeTrackingRects:count:]): Ditto.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7943 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoland updated layout tests for the relpositioned change, the ex unit change, and the...
hyatt [Fri, 5 Nov 2004 03:09:18 +0000 (03:09 +0000)]
land updated layout tests for the relpositioned change, the ex unit change, and the text-top vertical-align change

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7942 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fix for relpositioned inlines. This was reviewed a long time ago, but I can't recal...
hyatt [Fri, 5 Nov 2004 02:59:48 +0000 (02:59 +0000)]
Fix for relpositioned inlines.  This was reviewed a long time ago, but I can't recall who reviewed it (either
darin or ken).

        Reviewed by darin or ken

        * khtml/rendering/bidi.cpp:
        * khtml/rendering/render_block.cpp:
        * khtml/rendering/render_box.cpp:
        * khtml/rendering/render_box.h:
        * khtml/rendering/render_layer.cpp:
        * khtml/rendering/render_line.cpp:
        * khtml/rendering/render_object.h:

Finish turning on XSLT.  Make sure child stylesheets can load.

        * khtml/xsl/xslt_processorimpl.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7941 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Make sure the dominant line direction is properly set for RTL runs so that spaces...
hyatt [Fri, 5 Nov 2004 01:19:35 +0000 (01:19 +0000)]
Make sure the dominant line direction is properly set for RTL runs so that spaces will reverse.

Change xHeight to measure the ascent of the x glyph, since the xHeight metrics appear to be
totally bogus in both CG and AppKit.

        Reviewed by darin

        * WebCoreSupport.subproj/WebTextRenderer.m:
        (-[WebTextRenderer xHeight]):
        (-[WebTextRenderer _createATSUTextLayoutForRun:style:]):
        (-[WebTextRenderer _trapezoidForRun:style:atPoint:]):
        (-[WebTextRenderer _ATSU_drawHighlightForRun:style:geometry:]):
        (-[WebTextRenderer _ATSU_drawRun:style:geometry:]):
        (-[WebTextRenderer _ATSU_pointToOffset:style:position:reversed:includePartialGlyphs:]):

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7940 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Implement CSS3 support for multiple backgrounds. Also fix a bug with background...
hyatt [Fri, 5 Nov 2004 01:18:23 +0000 (01:18 +0000)]
Implement CSS3 support for multiple backgrounds.  Also fix a bug with background propagation so that it only
happens (from the <body> to the root) for HTML documents.  Fixed background-position to handle a mixture of
keyword and length values.

        Reviewed by darin

        * khtml/css/cssparser.cpp:
        * khtml/css/cssparser.h:
        * khtml/css/cssstyleselector.cpp:
        * khtml/css/cssstyleselector.h:
        * khtml/rendering/render_box.cpp:
        * khtml/rendering/render_box.h:
        * khtml/rendering/render_form.cpp:
        * khtml/rendering/render_line.cpp:
        * khtml/rendering/render_line.h:
        * khtml/rendering/render_object.cpp:
        * khtml/rendering/render_object.h:
        * khtml/rendering/render_style.cpp:
        * khtml/rendering/render_style.h:
        * khtml/rendering/render_table.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7939 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Make sure the text decoder returns empty strings rather than null strings when the...
hyatt [Fri, 5 Nov 2004 00:32:01 +0000 (00:32 +0000)]
Make sure the text decoder returns empty strings rather than null strings when the utf8 char ptr is non-null.
Ensures that <a href=""> works with libxml (which returns data in utf-8 buffers).

        Reviewed by darin

        * kwq/KWQTextCodec.mm:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7938 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Make sure line-height returns the correct value for normal.
hyatt [Fri, 5 Nov 2004 00:30:34 +0000 (00:30 +0000)]
Make sure line-height returns the correct value for normal.

        Reviewed by darin

        * khtml/css/css_computedstyle.cpp:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7937 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoDon't ignore results file since it's not here any more.
darin [Thu, 4 Nov 2004 21:07:27 +0000 (21:07 +0000)]
Don't ignore results file since it's not here any more.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7936 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Ken Kocienda.
harrison [Thu, 4 Nov 2004 18:29:23 +0000 (18:29 +0000)]
    Reviewed by Ken Kocienda.

<rdar://problem/3857753> REGRESSION (Mail): Delete incorrectly causes text to take on new style

        * khtml/editing/htmlediting.cpp:
        (khtml::DeleteSelectionCommand::moveNodesAfterNode): Fixed to move entire source subtree (up
        to, but not including, the enclosingBlockFlowElement) rather than just the source element.
        Fixed to insert after the destination subtree, rather than the destination element.  Handles
        edge case of deleting back to the top of the tree, where there is nothing left to insert after.
        * khtml/xml/dom_nodeimpl.cpp:
        (NodeImpl::enclosingNonBlockFlowElement): New method to support moveNodesAfterNode changes.
        * khtml/xml/dom_nodeimpl.h: Declare NodeImpl::enclosingNonBlockFlowElement
        * layout-tests/editing/deleting/delete-3857753-fix-expected.txt: Added.
        * layout-tests/editing/deleting/delete-3857753-fix.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7935 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Ken.
darin [Thu, 4 Nov 2004 17:16:03 +0000 (17:16 +0000)]
    Reviewed by Ken.

        - fixed <rdar://problem/3865365> since -[WebScriptObject dealloc] does not call [super dealloc], the build will fail due to a warning
        - fixed behavior so that [[WebScriptObject alloc] initWithCoder:] doesn't leak WebUndefined instances
          and incidentally so that [[WebScriptObject alloc] init] returns the single shared instance rather
          than allocating a new one

        * bindings/objc/WebScriptObject.mm: Removed some stray semicolons.
        (+[WebUndefined allocWithZone:]): Made this the common bottleneck that returns the single instance
        of WebUndefined, since it's the single method that normally allocates new instances. Calls super to
        actually allocate only the very first time it's called.
        (-[WebUndefined initWithCoder:]): Simplified to just return self (no reason to re-lookup the single
        shared instance since there can be only one).
        (-[WebUndefined copyWithZone:]): Ditto.
        (-[WebUndefined retain]): Ditto.
        (-[WebUndefined retainCount]): Use UINT_MAX constant here (matches usage in NSObject.m for retain count
        of class).
        (-[WebUndefined autorelease]): Simplified to just return self (see above).
        (-[WebUndefined copy]): No need to override this since it just turns around and calls copyWithZone:.
        (-[WebUndefined dealloc]): Added an assertion since this method should never be called. Also added
        a call to [super dealloc] after return; to make the new -Wdealloc-check compiler happy (fixing the
        bug mentioned above).
        (+[WebUndefined undefined]): Reimplemented; calls allocWithZone:NULL to get to the shared instance.
        No need to call init, since that's a no-op for this class.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7934 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Darin.
harrison [Thu, 4 Nov 2004 00:40:31 +0000 (00:40 +0000)]
    Reviewed by Darin.

        Eliminate the use of a marker file to determine how to build.

        * .cvsignore:
        * Makefile.am:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7933 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by me
kocienda [Wed, 3 Nov 2004 15:58:45 +0000 (15:58 +0000)]
    Reviewed by me

        More layout tests.

        * layout-tests/editing/deleting/delete-br-008-expected.txt: Added.
        * layout-tests/editing/deleting/delete-br-008.html: Added.
        * layout-tests/editing/deleting/delete-br-009-expected.txt: Added.
        * layout-tests/editing/deleting/delete-br-009.html: Added.
        * layout-tests/editing/deleting/delete-br-010-expected.txt: Added.
        * layout-tests/editing/deleting/delete-br-010.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7932 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fix by Yasuo Kida, reviewed by me.
mjs [Wed, 3 Nov 2004 11:52:26 +0000 (11:52 +0000)]
    Fix by Yasuo Kida, reviewed by me.

<rdar://problem/3819004> REGRESSION (Mail): Can't move cursor / delete character after deleting the active input area

* kwq/KWQKHTMLPart.mm:
        (KWQKHTMLPart::setMarkedTextRange): Treat a collapsed range the
same as a nil range - setting an empty marked range should clear
the marked range entirely.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7931 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoWebCore:
mjs [Wed, 3 Nov 2004 11:19:25 +0000 (11:19 +0000)]

        Reviewed by Dave Hyatt (when I originally coded it).

WebCore part of fix for:

<rdar://problem/3759187> REGRESSION (Mail): implement firstRectForCharacterRange:

        * kwq/WebCoreBridge.h:
        * kwq/WebCoreBridge.mm:
        (-[WebCoreBridge firstRectForDOMRange:]): New method to compute the rect for a
DOMRange, or if the range is split into multiple lines, the rect for the part on
the first line only.

        * khtml/rendering/render_object.cpp:
        (RenderObject::caretRect): Added extraWidthToEndOfLine parameter and ditto
for the overrides below.
        * khtml/rendering/render_object.h:
        * khtml/rendering/render_box.cpp:
        * khtml/rendering/render_box.h:
        * khtml/rendering/render_br.cpp:
        * khtml/rendering/render_br.h:
        * khtml/rendering/render_flow.cpp:
        * khtml/rendering/render_flow.h:
        * khtml/rendering/render_text.cpp:


        Reviewed by Dave Hyatt (when I originally coded it).

        Redid WebKit part of fix for:

        <rdar://problem/3759187> REGRESSION (Mail): implement firstRectForCharacterRange:

        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView firstRectForCharacterRange:]): Call the appropriate new bridge method,
        and translate to screen coordinates.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7930 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Hyatt.
sullivan [Wed, 3 Nov 2004 01:07:17 +0000 (01:07 +0000)]
    Reviewed by Hyatt.

        - [NSFont menuFontOfSize:], called from WebStringTruncator, was taking > 9% of the time creating a
        very large bookmarks menu, so I cached this one NSFont object.

        * Misc.subproj/WebStringTruncator.m:
        new function, caches the font used when no font is specified
        (+[WebStringTruncator centerTruncateString:toWidth:]):
        call new function

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7929 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoWebCore:
kocienda [Tue, 2 Nov 2004 23:22:45 +0000 (23:22 +0000)]

        Reviewed by Hyatt

        Implemented command to insert a block in response to typing a return key (even though
        I am not turning that on by default with this patch....that will come later).

        This new command is called InsertParagraphSeparatorCommand.

        Reworked the command and function names associated with inserting content into a
        document. Before this patch, there were inputXXX and insertXXX variants, with the
        former used for more high-level actions and the latter used for lower-level stuff.
        However, this was confusing as the AppKit uses insertXXX for everything. This resulted
        in an insertXXX command going through an inputXXX WebCore step and then finally to an
        insertXXX WebCore step. To make this less confusing, I have changes all the names to
        be insertXXX, and modified the lower-level operations so that it is clear what they do.

        * khtml/editing/htmlediting.cpp:
        (khtml::EditCommandPtr::isInsertTextCommand): Name change.
        (khtml::EditCommand::isInsertTextCommand): Ditto.
        (khtml::CompositeEditCommand::inputText): Ditto.
        (khtml::CompositeEditCommand::insertTextIntoNode): Ditto.
        (khtml::CompositeEditCommand::deleteTextFromNode): Ditto.
        (khtml::CompositeEditCommand::replaceTextInNode): Ditto.
        (khtml::CompositeEditCommand::deleteInsignificantText): Name changes in implementation.
        (khtml::CompositeEditCommand::isLastVisiblePositionInNode): Ditto.
        (khtml::DeleteFromTextNodeCommand::DeleteFromTextNodeCommand): Class name change, was DeleteTextCommand.
        (khtml::DeleteFromTextNodeCommand::~DeleteFromTextNodeCommand): Ditto.
        (khtml::DeleteFromTextNodeCommand::doApply): Ditto.
        (khtml::DeleteFromTextNodeCommand::doUnapply): Ditto.
        (khtml::DeleteSelectionCommand::performGeneralDelete): Ditto.
        (khtml::DeleteSelectionCommand::fixupWhitespace): Ditto.
        (khtml::DeleteSelectionCommand::moveNodesAfterNode): Ditto.
        (khtml::InsertIntoTextNode::InsertIntoTextNode): Class name change.
        (khtml::InsertIntoTextNode::~InsertIntoTextNode): Ditto.
        (khtml::InsertIntoTextNode::doApply): Ditto.
        (khtml::InsertIntoTextNode::doUnapply): Ditto.
        (khtml::InsertLineBreakCommand::InsertLineBreakCommand): Class name change, was InsertNewlineCommand.
        (khtml::InsertNodeBeforeCommand::InsertNodeBeforeCommand): Code moved. No changes.
        (khtml::InsertNodeBeforeCommand::~InsertNodeBeforeCommand): Ditto.
        (khtml::InsertNodeBeforeCommand::doApply): Ditto.
        (khtml::InsertNodeBeforeCommand::doUnapply): Ditto.
        (khtml::InsertParagraphSeparatorCommand::InsertParagraphSeparatorCommand): New command.
        Class name change, was InsertNewlineCommandInQuotedContentCommand.
        (khtml::InsertParagraphSeparatorInQuotedContentCommand::~InsertParagraphSeparatorInQuotedContentCommand): Ditto.
        (khtml::InsertParagraphSeparatorInQuotedContentCommand::isMailBlockquote): Ditto.
        (khtml::InsertParagraphSeparatorInQuotedContentCommand::doApply): Ditto.
        (khtml::InsertTextCommand::InsertTextCommand): Class name change, was InputTextCommand.
        (khtml::InsertTextCommand::doApply): Ditto.
        (khtml::InsertTextCommand::deleteCharacter): Ditto.
        (khtml::InsertTextCommand::prepareForTextInsertion): Ditto.
        (khtml::InsertTextCommand::input): Ditto.
        (khtml::InsertTextCommand::insertSpace): Ditto.
        (khtml::InsertTextCommand::isInsertTextCommand): Ditto.
        (khtml::TypingCommand::insertLineBreak): Name change, was insertNewline.
        (khtml::TypingCommand::insertParagraphSeparatorInQuotedContent): Name change, was insertNewlineInQuotedContent.
        (khtml::TypingCommand::insertParagraphSeparator): New function.
        (khtml::TypingCommand::doApply): Name changes, as above.
        (khtml::TypingCommand::insertText): Ditto.
        (khtml::TypingCommand::deleteKeyPressed): Ditto.
        (khtml::TypingCommand::preservesTypingStyle): Ditto.
        * khtml/editing/htmlediting.h:
        (khtml::DeleteFromTextNodeCommand::node): Name change.
        (khtml::DeleteFromTextNodeCommand::offset): Ditto.
        (khtml::DeleteFromTextNodeCommand::count): Ditto.
        (khtml::InsertIntoTextNode::text): Ditto.
        (khtml::InsertNodeBeforeCommand::insertChild): Ditto.
        (khtml::InsertNodeBeforeCommand::refChild): Ditto.
        (khtml::TypingCommand::): Ditto.
        * khtml/editing/jsediting.cpp: Name changes, as above.
        * kwq/WebCoreBridge.h:
        * kwq/WebCoreBridge.mm:
        (-[WebCoreBridge insertLineBreak]): Name change, was insertNewline.
        (-[WebCoreBridge insertParagraphSeparator]): New function.
        (-[WebCoreBridge insertParagraphSeparatorInQuotedContent]): Name change, was insertNewlineInQuotedContent.


        Reviewed by Hyatt

        WebCore now implements a command to insert a block in response to typing a return key, and
        some names were improved in the course of this work.

        * WebView.subproj/WebHTMLView.m:
        (-[WebHTMLView insertNewline:]): Now calls insertLineBreak on bridge object.
        (-[WebHTMLView insertLineBreak:]): New method.
        (-[WebHTMLView insertParagraphSeparator:]): Now implemented.
        * WebView.subproj/WebView.m:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7928 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed <rdar://problem/3861469> Latest Real player crashes Safari on some...
rjw [Mon, 1 Nov 2004 23:46:31 +0000 (23:46 +0000)]
    Fixed <rdar://problem/3861469> Latest Real player crashes Safari on some sites.

        Reviewed by Ken.

        * bindings/c/c_instance.cpp:
        Initialize out parameters to void type.

        * bindings/c/c_runtime.cpp:
        Initialize out parameters to void type.
        Also added additional checks to protect against classes that
        don't implement all functions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7927 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed <rdar://problem/3861257> WebUndefined should be returned for undefined...
rjw [Mon, 1 Nov 2004 22:33:14 +0000 (22:33 +0000)]
    Fixed <rdar://problem/3861257> WebUndefined should be returned for undefined values

        Reviewed by John.

        * ChangeLog:
        * bindings/objc/WebScriptObject.mm:
        (+[WebScriptObject _convertValueToObjcValue:KJS::root:Bindings::]):
Added additional conversion Undefined -> WebUndefined.
        * bindings/objc/objc_utility.mm:
Added additional conversion WebUndefined -> Undefined.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7926 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by rjw.
kdecker [Mon, 1 Nov 2004 22:20:58 +0000 (22:20 +0000)]
    Reviewed by rjw.

fixed <rdar://problem/3681094> Crash in KJS::WindowFunc::tryCall with application/xhtml+xml Content-Type
once and for all.

        * khtml/ecma/kjs_window.cpp:
        (WindowFunc::tryCall): Added a nil check in the case of an empty document lacking a baseURL().

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7924 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago - fixed <rdar://problem/3855573> Remove reference to "WebScriptMethods" from...
darin [Mon, 1 Nov 2004 21:23:14 +0000 (21:23 +0000)]
    - fixed <rdar://problem/3855573> Remove reference to "WebScriptMethods" from WebScriptObject.h comments

        * bindings/objc/WebScriptObject.h: Removed unneeded #ifdef protection for multiple includes (since
        this is an Objective-C header and we use #import for those). Fixed comments as requested in the bug
        report to match the contents of the file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7923 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Hyatt.
darin [Mon, 1 Nov 2004 19:21:24 +0000 (19:21 +0000)]
    Reviewed by Hyatt.

        - fixed <rdar://problem/3859381> REGRESSION (167-168): text in form fields should not use body's text color

        * khtml/css/html4.css: Use color: initial for textarea and related ones.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7922 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by John
kocienda [Mon, 1 Nov 2004 17:56:33 +0000 (17:56 +0000)]
    Reviewed by John

        Fix for this bug:

        <rdar://problem/3775920> REGRESSION (Mail): Centering doesn't work in HTML mail

        * khtml/css/css_computedstyle.cpp:
        (DOM::CSSComputedStyleDeclarationImpl::copyInheritableProperties): Factor out the
        implementation here into new copyPropertiesInSet helper. This now calls the
        generalized copyPropertiesInSet function with the arguments needed to make copying
        inheritable work.
        * khtml/css/css_computedstyle.h:
        * khtml/css/css_valueimpl.cpp:
        (CSSStyleDeclarationImpl::diff): Move this function here from css_computedstyle.cpp.
        In order to do apply block properties, "regular" style declarations need to do style
        diffs as well.
        (CSSStyleDeclarationImpl::copyBlockProperties): New helper. Just like copyInheritableProperties
        except that it uses a different set of properties that apply only to blocks.
        (CSSStyleDeclarationImpl::copyPropertiesInSet): New helper that looks at a style declaration
        and copies out those properties listed in a pre-defined set.
        * khtml/css/css_valueimpl.h:
        * khtml/editing/htmlediting.cpp:
        (khtml::StyleChange::StyleChange): Modified to work with style changes that apply to a whole
        block, factoring out some of the special case code that should now only run in the inline case.
        (khtml::StyleChange::init): Factored out the code that now is in checkForLegacyHTMLStyleChange.
        (khtml::StyleChange::checkForLegacyHTMLStyleChange): New helper for case where we want
        special handling for "legacy" HTML styles like <B> and <I>.
        (khtml::ApplyStyleCommand::doApply): Much refactoring in this class to divide up the work of
        style changes into different kinds. CSS specifies certain properties only apply to certain
        element types. This set of changes now recognizes two such separate cases: styles that apply
        to blocks, and styles that apply to inlines.
        (khtml::ApplyStyleCommand::applyBlockStyle): New function to handle apply styles to whole blocks.
        (khtml::ApplyStyleCommand::applyInlineStyle): New function to handle apply styles to inlines.
        (khtml::ApplyStyleCommand::isHTMLStyleNode): Is now passed a CSSStyleDeclarationImpl to work
        with rather than working on the CSSStyleDeclarationImpl member variable of the class. This is
        done so that the function can be passed a portion of the styles being applied so that block styles
        and inline styles can be handled separately.
        (khtml::ApplyStyleCommand::removeCSSStyle): Ditto.
        (khtml::ApplyStyleCommand::removeBlockStyle): New function to handle removing styles from whole blocks.
        (khtml::ApplyStyleCommand::removeInlineStyle): New function to removing styles from inlines.
        (khtml::ApplyStyleCommand::addBlockStyleIfNeeded): New function to handle applying style to whole blocks.
        (khtml::ApplyStyleCommand::addInlineStyleIfNeeded): New function to handle applying style to inlines.
        * khtml/editing/htmlediting.h:
        (khtml::StyleChange::): Changed as described above.
        (khtml::EditCommand::setEndingSelectionNeedsLayout): New function to that tells the ending selection
        it needs to layout, even though it has not changed position in the DOM. For instance, this is needed
        when text align changes.
        * khtml/khtml_part.cpp:
        (KHTMLPart::setTypingStyle): Put in an early bail-out in the case where the current style matches
        the passed-in argument.
        (KHTMLPart::applyStyle): Modify this function so that block styles are applied when the selection
        is a caret. Formerly, this just set typing style and made no visible changes to the document.

        New tests.

        * layout-tests/editing/editing.js: Added some glue to change text align.
        * layout-tests/editing/style/block-style-001-expected.txt: Added.
        * layout-tests/editing/style/block-style-001.html: Added.
        * layout-tests/editing/style/block-style-002-expected.txt: Added.
        * layout-tests/editing/style/block-style-002.html: Added.
        * layout-tests/editing/style/block-style-003-expected.txt: Added.
        * layout-tests/editing/style/block-style-003.html: Added.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7921 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoversioning for TOT, Safari 2.0 (170u)
vicki [Fri, 29 Oct 2004 23:51:55 +0000 (23:51 +0000)]
versioning for TOT, Safari 2.0 (170u)

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7920 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years agoSafari-169 stamp
vicki [Fri, 29 Oct 2004 23:42:16 +0000 (23:42 +0000)]
Safari-169 stamp

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7918 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by NOBODY (OOPS!).
darin [Fri, 29 Oct 2004 21:53:52 +0000 (21:53 +0000)]
    Reviewed by NOBODY (OOPS!).

        - fixed <rdar://problem/3751619> Safari crash in khtml::CircularSearchBuffer::append(QChar const&)

        * khtml/editing/visible_text.cpp: (khtml::findPlainText): Fix exit condition to check for break
        before advancing one character; before it did it backwards.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7917 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed: <rdar://problem/3853262> REGRESSION(166-168) gmail gets blank page when loading
cblu [Fri, 29 Oct 2004 21:48:55 +0000 (21:48 +0000)]
Fixed: <rdar://problem/3853262> REGRESSION(166-168) gmail gets blank page when loading

        Reviewed by kocienda, adele.

        * khtml/rendering/render_frames.cpp:
        (RenderPartObject::updateWidget): remove infinite frame recursion check for iframes

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7916 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago * WebKit.exp: added _WebPlugInModeKey, forgot to add it earlier
cblu [Fri, 29 Oct 2004 20:16:58 +0000 (20:16 +0000)]
    * WebKit.exp: added _WebPlugInModeKey, forgot to add it earlier

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7915 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Chris.
darin [Fri, 29 Oct 2004 18:59:39 +0000 (18:59 +0000)]
    Reviewed by Chris.

        - fixed <rdar://problem/3857395> clicking on calendar in Apple Travel site crashes Safari in invalidateClick (getthere.net)

        * khtml/khtmlview.cpp:
        (KHTMLView::viewportMousePressEvent): Use a SharedPtr<KHTMLView> to make sure the KHTMLView is not
        deleted before this function finishes running.
        (KHTMLView::viewportMouseDoubleClickEvent): Ditto.
        (KHTMLView::viewportMouseReleaseEvent): Ditto.
        (KHTMLView::dispatchMouseEvent): Removed ref/deref pairs that aren't needed since dispatchEvent
        is guaranteed to do ref/deref as needed.

        * kwq/KWQObject.mm: (QObject::startTimer): Fixed a comment.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7914 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago - fixed <rdar://problem/3855573> Remove reference to "WebScriptMethods" from...
darin [Fri, 29 Oct 2004 18:11:20 +0000 (18:11 +0000)]
    - fixed <rdar://problem/3855573> Remove reference to "WebScriptMethods" from WebScriptObject.h comments

        * Plugins.subproj/WebScriptObject.h: Removed unneeded #ifdef protection for multiple includes (since
        this is an Objective-C header and we use #import for those). Fixed comments as requested in the bug
        report to match the contents of the file.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7913 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Enabled XSLT on Panther. See intrigue mail for compiling instructions.
cblu [Thu, 28 Oct 2004 23:14:40 +0000 (23:14 +0000)]
Enabled XSLT on Panther. See intrigue mail for compiling instructions.

        Reviewed by darin.

        * WebCore.pbproj/project.pbxproj: link against xslt unconditionally, link against specific version of libxml on Panther
        * WebCorePrefix.h: always use XSLT

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7912 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Chris
kocienda [Thu, 28 Oct 2004 22:35:00 +0000 (22:35 +0000)]
    Reviewed by Chris

        Fix for these bugs:

        <rdar://problem/3854848> Tiger Mail Crash in WebCore - khtml::CompositeEditCommand::insertNodeAfter
        <rdar://problem/3803832> REGRESSION (Mail): incorrect behavior after Return + Delete in quoted text

        * khtml/editing/htmlediting.cpp:
        (khtml::DeleteSelectionCommand::DeleteSelectionCommand): Added node pointer class members
        to initialization list, zeroing them out.
        (khtml::DeleteSelectionCommand::canPerformSpecialCaseBRDelete): New special-case helper to
        handle a delete of content in special cases where the only thing selected is a BR. This
        code path is much simpler than the newly-named performGeneralDelete, and detects when no
        content merging should be done between blocks. This aspect of the change fixes 3854848.
        One of the special cases added fixes 3803832.
        (khtml::DeleteSelectionCommand::performGeneralDelete): Renamed, from performDelete.
        (khtml::DeleteSelectionCommand::moveNodesAfterNode): Made this helper be a no-arg function, just
        like the other helpers in this class.
        (khtml::DeleteSelectionCommand::clearTransientState): Fix cut and paste error in deref code.
        (khtml::DeleteSelectionCommand::doApply): Updated for changed helpers.
        * khtml/editing/htmlediting.h: Added new helper and changed an old one.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7911 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Fixed: <rdar://problem/3856913> Panther-only crash in QString code copying front...
cblu [Thu, 28 Oct 2004 22:05:35 +0000 (22:05 +0000)]
Fixed: <rdar://problem/3856913> Panther-only crash in QString code copying front page of store.apple.com

        Reviewed by darin.

        * kwq/KWQKHTMLPart.mm:
        (KWQKHTMLPart::attributedString): check that the renderer is a list item before making list item calls on it

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7910 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by Harrison
kocienda [Thu, 28 Oct 2004 20:07:47 +0000 (20:07 +0000)]
    Reviewed by Harrison

        Reorganization of delete command functionality so that doApply is not
        several hundred lines long. This is not a squeaky-clean cleanup, but
        it is a step in the right direction. No functionality changes.

        * khtml/editing/htmlediting.cpp:
        (khtml::DeleteSelectionCommand::initializePositionData): New helper.
        (khtml::DeleteSelectionCommand::saveTypingStyleState): Ditto.
        (khtml::DeleteSelectionCommand::performDelete): Ditto.
        (khtml::DeleteSelectionCommand::fixupWhitespace): Ditto.
        (khtml::DeleteSelectionCommand::moveNodesAfterNode): Ditto.
        (khtml::DeleteSelectionCommand::calculateEndingPosition): Ditto.
        (khtml::DeleteSelectionCommand::calculateTypingStyleAfterDelete): Ditto.
        (khtml::DeleteSelectionCommand::clearTransientState): Ditto.
        (khtml::DeleteSelectionCommand::doApply): Factor out code into new helpers.
        * khtml/editing/htmlediting.h:

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7909 268f45cc-cd09-0410-ab3c-d52691b4dbfc

18 years ago Reviewed by me
kocienda [Thu, 28 Oct 2004 16:38:52 +0000 (16:38 +0000)]
    Reviewed by me

        * khtml/editing/htmlediting.cpp:
        (khtml::DeleteSelectionCommand::DeleteSelectionCommand): Typo in initializer caused
        new mergeBlocksAfterDelete flag to be set improperly, causing layout regressions.

git-svn-id: https://svn.webkit.org/repository/webkit/trunk@7908 268f45cc-cd09-0410-ab3c-d52691b4dbfc