Timeline



Oct 29, 2005:

10:26 PM Changeset in webkit [11008] by hyatt
  • 4 edits in trunk/WebCore

Fix the performance regression caused by doing too much copying of background/border values. This patch attempts to rectify things by doing the following:

(1) Don't initialize the border/background cached values on every single style resolution.
(2) Only cache the border/background values just after user agent styles have been resolved if the user agent set an appearance.
(3) Only check for appearance disabling if you had a UA appearance originally (and then continue to have an appearance after author/user styles have been resolved too)
(4) Make sure to patch the pseudoStyleForElement method too. With the removal of the initialization code for the values running over and over again, I needed to make sure to patch this function to match styleForElement.

Reviewed by sullivan

  • khtml/css/cssstyleselector.cpp: (khtml::CSSStyleSelector::initForStyleResolve): (khtml::CSSStyleSelector::styleForElement): (khtml::CSSStyleSelector::pseudoStyleForElement): (khtml::CSSStyleSelector::adjustRenderStyle):
  • khtml/css/cssstyleselector.h:

Oct 28, 2005:

3:29 PM Changeset in webkit [11007]
  • 71 copies
    5 deletes in tags/JavaScriptCore-417~4

This commit was manufactured by cvs2svn to create tag
'JavaScriptCore-417~4'.

3:29 PM Changeset in webkit [11006] by thatcher
  • 2 edits in branches/Safari-2-0-branch/JavaScriptCore

Versioning to JavaScriptCore-417.4

10:35 AM Changeset in webkit [11005] by bdakin
  • 6 edits in trunk/WebCore

Bug #:
Submitted by:
Reviewed by:

10:14 AM Changeset in webkit [11004] by adele
  • 2 edits in trunk/WebCore

Reviewed by Beth.

  • fixed <rdar://problem/4319232> finding text in overflow area doesn't always scroll to the right place
  • khtml/rendering/render_text.cpp: (RenderText::selectionRect): Subtracts scroll offset of containing block layer.
4:46 AM Changeset in webkit [11003] by hyatt
  • 3 edits in trunk/LayoutTests/editing/selection

Update editing tests for button changes

4:24 AM Changeset in webkit [11002] by hyatt
  • 1 edit in trunk/WebCore/khtml/rendering/render_theme_mac.mm

Adjust the padding of mini buttons down from 11px to 8px so that they fit better in small spaces.

1:44 AM Changeset in webkit [11001] by hyatt
  • 2 adds in trunk/LayoutTests/fast/forms

Add positioned test for buttons.

1:42 AM Changeset in webkit [11000] by hyatt
  • 2 edits in trunk/WebCore

Make sure positioned content works with width:intrinsic. Fixes positioned buttons. The bug
is bugzilla #5520.

Reviewed by mjs

fast/forms/positioned-button.html

  • khtml/rendering/render_box.cpp: (RenderBox::calcAbsoluteHorizontalValues):
1:01 AM Changeset in webkit [10999] by adele
  • 2 edits in trunk/WebCore

Reviewed by Hyatt.

-fixed <rdar://problem/4081091> focus() does not work properly on anchors
<rdar://problem/4317689> new form elements don't get revealed when focus() is called on them

  • khtml/xml/dom_docimpl.cpp: (DocumentImpl::setFocusNode): call scrollRectToVisible.
12:25 AM Changeset in webkit [10998] by hyatt
  • 11 edits in trunk/LayoutTests/fast

Check in updated button tests.

12:03 AM Changeset in webkit [10997] by hyatt
  • 9 edits in trunk/WebCore

Fix for bug 5519, buttons need to honor background/border: none and turn off styles.
This patch changes the default style rules so that buttons have both a border and
a background color.

If the author changes the border/background so that it no longer matches these chosen
defaults, we assume that the button has now been styled and we turn off the Aqua
appearance.

We also play more games with border and padding and juggle the values around so that
things look right with Aqua turned on and off.

Reviewed by mjs

  • khtml/css/cssstyleselector.cpp: (khtml::CSSStyleSelector::matchRules): (khtml::CSSStyleSelector::initForStyleResolve): (khtml::CSSStyleSelector::styleForElement): (khtml::CSSStyleSelector::adjustRenderStyle): (khtml::CSSStyleSelector::applyDeclarations):
  • khtml/css/cssstyleselector.h:
  • khtml/css/html4.css:
  • khtml/rendering/render_style.h: (khtml::BorderValue::BorderValue): (khtml::BorderData::operator!=): (khtml::RenderStyle::border):
  • khtml/rendering/render_theme.cpp: (khtml::RenderTheme::paint): (khtml::RenderTheme::isControlStyled):
  • khtml/rendering/render_theme.h:
  • khtml/rendering/render_theme_mac.h:
  • khtml/rendering/render_theme_mac.mm: (khtml::RenderThemeMac::adjustButtonStyle): (khtml::RenderThemeMac::setButtonCellState): (khtml::RenderThemeMac::paintButton):

Oct 27, 2005:

6:51 PM Changeset in webkit [10996] by vicki
  • 2 edits in trunk/WebCore

Reviewed by Hyatt.

Fix problems with link jumping. In the cases below, we were calculating the wrong position to scroll to.

<rdar://problem/4247537> link jumping should scroll to tallest object on line, not first object on line
<rdar://problem/3489554> when calculating position for link jumping, skip siblings that are unrendered whitespace
<rdar://problem/4244382> Safari - erratic behavior of empty anchor tags followed by whitespace
<rdar://problem/4256060> Link scrolling to last object on the page doesn't work if the link being scrolled to contains an empty inline
<rdar://problem/4276623> erratic link jumping when tables are involved

  • khtml/xml/dom_nodeimpl.cpp: (DOM::ContainerNodeImpl::getUpperLeftCorner):
5:49 PM Changeset in webkit [10995] by eseidel
  • 5 edits in trunk/WebCore

Bug #: none
Submitted by: eseidel
Reviewed by: none, build fix only.

Build fix, forgot to commit project file.
Also need to make XSLT #ifdef's work for portability.
​http://bugzilla.opendarwin.org/show_bug.cgi?id=3275

  • WebCore.xcodeproj/project.pbxproj:
  • khtml/ecma/XSLTProcessor.cpp:
  • khtml/ecma/XSLTProcessor.h:
  • khtml/ecma/kjs_window.cpp: (KJS::Window::getValueProperty):
5:06 PM Changeset in webkit [10994] by eseidel
  • 2 edits in trunk

Bug #: none
Submitted by: eseidel
Reviewed by: none needed, svg build fix only.

No review needed, svg build fix only.

  • kdom/xpointer/XPointerExpressionImpl.cpp: "Shared.h" to "kdom/Shared.h"
4:19 PM Changeset in webkit [10993] by harrison
  • 1 edit
    2 adds in trunk/LayoutTests

Reviewed by Justin Garcia and Dave Hyatt.

Test cases for <rdar://problem/4134884>.

  • fast/events/event-targets-expected.txt: Added.
  • fast/events/event-targets.html: Added.
4:15 PM Changeset in webkit [10992] by harrison
  • 11 edits in trunk/WebCore

Reviewed by Justin Garcia and Dave Hyatt.

<rdar://problem/4134884> crash trying to forward msg - khtml::RenderBlock::addChildToFlow

Problem was that onunload events were being fired while in the middle of trying to detach, which resulted in updateDocumentsRendering re-attaching the previously detached nodes. Similar problem with load events while attaching.

Solved by eliminating load and unload for object nodes, pre-sending unload before detach, and sending dom mutation and load events after attaching.

Also, added asserts to catch this problem more easily in the future.

Added test:

  • fast/events/event-targets.html Make sure that load and unload events do not fire on certain objects.
  • khtml/ecma/kjs_html.cpp: (KJS::Image::notifyFinished): Use constant string for "load" event name.
  • khtml/ecma/xmlhttprequest.cpp: (KJS::XMLHttpRequest::changeState): Use constant strings for "load" and "readystatechange" event names.
  • khtml/html/html_baseimpl.cpp: (HTMLFrameElementImpl::close): (HTMLFrameElementImpl::willRemove): (HTMLFrameElementImpl::detach): Add willRemove() function.
  • khtml/html/html_baseimpl.h: Add willRemove() function.
  • khtml/html/html_objectimpl.cpp: (DOM::HTMLObjectElementImpl::attach): (DOM::HTMLObjectElementImpl::detach): Stop needlessly sending load and unload events for OBJECT elements.
  • khtml/xml/dom2_traversalimpl.cpp: (DOM::NodeIteratorImpl::notifyBeforeNodeRemoval): Rename local variable from willRemove to removedNode for clarity.
  • khtml/xml/dom_docimpl.cpp: (DocumentImpl::DocumentImpl): (DocumentImpl::forbidEventDispatch): (DocumentImpl::allowEventDispatch): (DocumentImpl::eventDispatchForbidden): (DocumentImpl::createEvent): Add mechanism to prevent event dispatch.
  • khtml/xml/dom_docimpl.h:
  • khtml/xml/dom_nodeimpl.cpp: (DOM::NodeImpl::dispatchEvent): (DOM::NodeImpl::dispatchGenericEvent): (DOM::NodeImpl::dispatchHTMLEvent): (DOM::NodeImpl::dispatchWindowEvent): (DOM::NodeImpl::dispatchMouseEvent): (DOM::NodeImpl::dispatchSimulatedMouseEvent): (DOM::NodeImpl::dispatchUIEvent): (DOM::NodeImpl::dispatchSubtreeModifiedEvent): (DOM::NodeImpl::dispatchKeyEvent): (DOM::NodeImpl::dispatchWheelEvent): (DOM::NodeImpl::willRemove): (DOM::ContainerNodeImpl::insertBefore): (DOM::ContainerNodeImpl::replaceChild): (DOM::ContainerNodeImpl::willRemove): (DOM::ContainerNodeImpl::willRemoveChild): (DOM::ContainerNodeImpl::removeChild): (DOM::ContainerNodeImpl::removeChildren): (DOM::ContainerNodeImpl::appendChild): (DOM::ContainerNodeImpl::addChild): (DOM::ContainerNodeImpl::dispatchChildInsertedEvents): (DOM::ContainerNodeImpl::dispatchChildRemovalEvents): Prevent event dispatch during DOM node removals and additions.
  • khtml/xml/dom_nodeimpl.h:
3:49 PM Changeset in webkit [10991] by ggaren
  • 1 edit
    2 adds in trunk/LayoutTests

Layout test included with ​http://bugzilla.opendarwin.org/show_bug.cgi?id=5303
TextInputController should support attributed strings

  • fast/forms/attributed-strings-expected.txt: Added.
  • fast/forms/attributed-strings.html: Added.
3:42 PM Changeset in webkit [10990] by eseidel
  • 1 edit
    2 adds in trunk/LayoutTests

Bug #: 3275
Submitted by: eseidel
Reviewed by: darin

Test cases for the new XSLTProcessor JS object.
​http://bugzilla.opendarwin.org/show_bug.cgi?id=3275

  • fast/xsl/xslt-processer-expected.txt: Added.
  • fast/xsl/xslt-processer.html: Added.
3:35 PM Changeset in webkit [10989] by harrison
  • 2 edits in trunk/WebCore

Reviewed by Tim Omerick.

<rdar://problem/4251172> AX: Safari crashes on www.mozilla.org with Voice Over - khtml::RenderImage::imageMap

Test cases added: None. Manual testing is way to awkward, and automated testing
is not possible. See following bug...
<rdar://problem/4256882> Need automated testing support for accessibility APIs

  • ChangeLog:
  • khtml/rendering/render_image.cpp: (RenderImage::RenderImage): (RenderImage::imageMap): (RenderImage::updateAltText):
3:34 PM Changeset in webkit [10988] by eseidel
  • 10 edits
    2 adds in trunk/WebCore

Bug #: 3275
Submitted by: eseidel
Reviewed by: hyatt, darin

Add XSLTProcessor support to WebCore's JavaScript bindings.
<rdar://problem/3642402> add XSLTProcessor to WebCore's JavaScript support
​http://bugzilla.opendarwin.org/show_bug.cgi?id=3275

  • WebCore.xcodeproj/project.pbxproj:
  • khtml/ecma/XSLTProcessor.cpp: Added. (KJS::): (KJS::XSLTProcessor::XSLTProcessor): (KJS::XSLTProcessor::~XSLTProcessor): (KJS::XSLTProcessorProtoFunc::callAsFunction):
  • khtml/ecma/XSLTProcessor.h: Added. (KJS::XSLTProcessor::classInfo): (KJS::XSLTProcessor::): (KJS::XSLTProcessor::impl): (KJS::XSLTProcessorConstructorImp::XSLTProcessorConstructorImp): (KJS::XSLTProcessorConstructorImp::implementsConstruct): (KJS::XSLTProcessorConstructorImp::construct):
  • khtml/ecma/kjs_window.cpp: (KJS::Window::getValueProperty):
  • khtml/ecma/kjs_window.h: (KJS::Window::):
  • khtml/khtml_part.cpp: (KHTMLPart::replaceDocImpl):
  • khtml/xml/dom_docimpl.cpp: (DocumentImpl::DocumentImpl): (DocumentImpl::~DocumentImpl): (DocumentImpl::applyXSLTransform):
  • khtml/xml/dom_docimpl.h: (DOM::DocumentImpl::transformSourceDocument): (DOM::DocumentImpl::setTransformSourceDocument):
  • khtml/xml/xml_tokenizer.cpp: (khtml::xmlDocPtrForString): (khtml::XMLTokenizer::setTransformSource):
  • khtml/xml/xml_tokenizer.h:
  • khtml/xsl/xslt_processorimpl.cpp: (DOM::parseErrorFunc): (DOM::stylesheetLoadFunc): (DOM::setXSLTLoadCallBack): (DOM::writeToQString): (DOM::saveResultToString): (DOM::transformTextStringToXHTMLDocumentString): (DOM::xsltParamArrayFromQDict): (DOM::freeXsltParamArray): (DOM::XSLTProcessorImpl::createDocumentFromSource): (DOM::createFragmentFromSource): (DOM::xsltStylesheetPointer): (DOM::xmlDocPtrFromNode): (DOM::resultMIMEType): (DOM::XSLTProcessorImpl::transformToString): (DOM::XSLTProcessorImpl::transformToDocument): (DOM::XSLTProcessorImpl::transformToFragment): (DOM::XSLTProcessorImpl::setParameter): (DOM::XSLTProcessorImpl::getParameter): (DOM::XSLTProcessorImpl::removeParameter):
  • khtml/xsl/xslt_processorimpl.h: (DOM::XSLTProcessorImpl::XSLTProcessorImpl): (DOM::XSLTProcessorImpl::setXSLStylesheet): (DOM::XSLTProcessorImpl::importStylesheet): (DOM::XSLTProcessorImpl::clearParameters): (DOM::XSLTProcessorImpl::reset): (DOM::XSLTProcessorImpl::xslStylesheet):
3:15 PM Changeset in webkit [10987] by ggaren
  • 2 edits in trunk/WebKitTools

Patch by Alexey Proskuryakov.

Reviewed by Maciej.

Fixes ​http://bugzilla.opendarwin.org/show_bug.cgi?id=5303
TextInputController should support attributed strings

  • DumpRenderTree/TextInputController.m: (+[NSMutableAttributedString isSelectorExcludedFromWebScript:]): (+[NSMutableAttributedString webScriptNameForSelector:]): (-[NSMutableAttributedString getLength]): (-[NSMutableAttributedString attributeNamesAtIndex:]): (-[NSMutableAttributedString valueOfAttribute:atIndex:]): (-[NSMutableAttributedString addAttribute:value:]): (-[NSMutableAttributedString addAttribute:value:from:length:]): (-[NSMutableAttributedString addColorAttribute:red:green:blue:alpha:]): (-[NSMutableAttributedString addColorAttribute:red:green:blue:alpha:from:length:]): (-[NSMutableAttributedString addFontAttribute:fontName:size:]): (-[NSMutableAttributedString addFontAttribute:fontName:size:from:length:]): (+[TextInputController isSelectorExcludedFromWebScript:]): (+[TextInputController webScriptNameForSelector:]): (-[TextInputController insertText:]): (-[TextInputController attributedSubstringFrom:length:]): (-[TextInputController attributedStringWithString:]):
1:13 PM Changeset in webkit [10986] by adele
  • 2 edits in trunk/WebCore

Reviewed by Vicki.

Fixed bug that prevents overflow areas from scrolling to reveal anchor.
Part of <rdar://problem/3612121>.

  • khtml/khtml_part.cpp: (KHTMLPart::gotoAnchor):
10:39 AM Changeset in webkit [10985] by adele
  • 3 edits in trunk/WebCore

Reviewed by Vicki.

  • fixed <rdar://problem/4318167> REGRESSION: content doesn't scroll far enough to the left after clicking links at aplacecalledcommon.co.uk
  • khtml/rendering/render_layer.cpp: (khtml::RenderLayer::scrollRectToVisible): fixed a typo where verticalAlignment was being passed instead of horizontalAlignment.
  • khtml/khtml_part.cpp: (KHTMLPart::gotoAnchor): specified alignLeft as horizontal alignment.
2:30 AM Changeset in webkit [10984] by hyatt
  • 2 edits in trunk/WebCore

Fix form controls so that they can take advantage of
the style sharing optimization. Special-case the "type"
attribute to accomplish this.

Also add checks for missing pseudo-classes that the sharing
code now needs to check once form controls start sharing
style. These include enabled, checked and indeterminate.

Finally, :target should have been checked all along and was
broken by the original style sharing landing. Add it to the list
of things checked as well.

It's impossible to write a :target test. Other areas more than
covered by existing test cases.

Reviewed by mjs

  • khtml/css/cssstyleselector.cpp: (khtml::CSSStyleSelector::canShareStyleWithElement): (khtml::CSSStyleSelector::checkOneSelector):
2:00 AM Changeset in webkit [10983] by hyatt
  • 2 edits in trunk/WebCore

Fix for bug 5517, percentage height relative block inside a percentage height absolute
block did not stretch properly to fill the absolute block.

Also fixing box-sizing bugs in percentage height calculations. Recurrence in
calcPercentageHeight needed a content box adjustment.

Reviewed by mjs

fast/block/positioning/062.html
fast/box-sizing/percentage-test.html
fast/box-sizing/panels-one.html
fast/box-sizing/panels-two.html

  • khtml/rendering/render_box.cpp: (RenderBox::calcPercentageHeight):
1:57 AM Changeset in webkit [10982] by hyatt
  • 2 adds in trunk/LayoutTests/fast/block/positioning

Add new positioning test.

1:56 AM Changeset in webkit [10981] by hyatt
  • 6 adds in trunk/LayoutTests/fast/box-sizing

Add box-sizing tests.

1:07 AM Changeset in webkit [10980] by eseidel
  • 2 edits in trunk/WebCore

Bug #: none
Submitted by: eseidel
Reviewed by: none needed, svg build fix only.

No review, SVG build fix only.
Fixing fallout from <rdar://problem/4098450>.

  • WebCore+SVG/xml_kdomtokenizer.cpp: (KDOMTokenizer::write): now returns bool (KDOMTokenizer::finish): error check

Oct 26, 2005:

11:31 PM Changeset in webkit [10979] by eseidel
  • 2 edits
    4 adds in trunk/LayoutTests

Bug #: 5404
Submitted by: eseidel
Reviewed by: mjs & darin

Test case updates for XML serialization fixes.
​http://bugzilla.opendarwin.org/show_bug.cgi?id=5404

  • fast/dom/dom-parse-serialize-display-expected.txt:
  • fast/innerHTML/004-expected.txt: Added.
  • fast/innerHTML/004.xhtml: Added.
  • fast/innerHTML/005-expected.txt: Added.
  • fast/innerHTML/005.html: Added.
11:03 PM Changeset in webkit [10978] by eseidel
  • 7 edits in trunk/WebCore

Bug #: 5404
Submitted by: eseidel
Reviewed by: mjs & darin

Various fixes to createMarkup and toString code to properly
support serialization of XML.
​http://bugzilla.opendarwin.org/show_bug.cgi?id=5404

  • khtml/editing/markup.cpp: (khtml::startMarkup): (khtml::doesHTMLForbidEndTag): (khtml::shouldSelfClose): (khtml::endMarkup): (khtml::markup): (khtml::createFragmentFromMarkup): (khtml::createMarkup):
  • khtml/html/html_elementimpl.cpp: (HTMLElementImpl::createContextualFragment): (HTMLElementImpl::toString):
  • khtml/html/htmltokenizer.cpp: (khtml::parseHTMLDocumentFragment):
  • khtml/html/htmltokenizer.h:
  • khtml/xml/dom_xmlimpl.cpp: (DOM::ProcessingInstructionImpl::ProcessingInstructionImpl): (DOM::ProcessingInstructionImpl::~ProcessingInstructionImpl): (DOM::ProcessingInstructionImpl::toString):
  • khtml/xml/dom_xmlimpl.h: (DOM::ProcessingInstructionImpl::sheet): (DOM::ProcessingInstructionImpl::setStyleSheet):
3:13 PM Changeset in webkit [10977] by hyatt
  • 2 edits in trunk/WebCore

Don't allow position:relative to apply to table sections. Fixes the crash described in
Radar bug 4107882. (​http://cityoflakeforest.com/cs/pw/cs_pw2a3.htm)

Reviewed by bdakin

  • khtml/css/cssstyleselector.cpp: (khtml::CSSStyleSelector::adjustRenderStyle):
2:47 PM Changeset in webkit [10976] by adele
  • 5 edits in trunk/WebCore

Reviewed by John.

Fixes a few problems with <rdar://problem/3612121> that I noticed with code inspection.

  • khtml/rendering/render_layer.h: removed inline scrollToPoint
  • khtml/rendering/render_layer.cpp: (khtml::RenderLayer::getRectToExpose):
  • khtml/khtml_part.cpp: (KHTMLPart::gotoAnchor): Used to call scrollToPoint, now calls scrollRectToVisible
  • kwq/KWQKHTMLPart.mm: (KWQKHTMLPart::jumpToSelection): Now uses the enclosing layer of the selection start, instead of the outermost layer. (KWQKHTMLPart::centerSelectionInVisibleArea): ditto.
1:45 PM Changeset in webkit [10975] by andersca
  • 19 edits in trunk/LayoutTests

2005-10-26 Anders Carlsson <​andersca@mac.com>

Reviewed by Maciej.

  • dom/xhtml/level3/core/nodeisequalnode01-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode02-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode04-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode06-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode07-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode10-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode11-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode12-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode13-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode14-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode19-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode20-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode22-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode28-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode29-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode31-expected.txt:
  • dom/xhtml/level3/core/nodeisequalnode32-expected.txt: These now pass.
  • dom/xhtml/level3/core/nodeisequalnode21-expected.txt: This now fails because doctype is null.
1:44 PM Changeset in webkit [10974] by andersca
  • 8 edits in trunk/WebCore

2005-10-26 Anders Carlsson <​andersca@mac.com>

Reviewed by Maciej.

  • khtml/ecma/kjs_dom.cpp: (KJS::DOMNodeProtoFunc::callAsFunction):
  • khtml/ecma/kjs_dom.h: Add JS wrapper.

(KJS::DOMNode::):

  • khtml/xml/dom_nodeimpl.cpp: (DOM::NodeImpl::isEqualNode):
  • khtml/xml/dom_nodeimpl.h: Implement isEqualNode.
  • kwq/DOM.mm: (-[DOMNode isEqualNode:]): Add Objective C wrapper.
1:23 PM Changeset in webkit [10973] by thatcher
  • 2 edits
    2 adds in branches/Safari-2-0-branch/JavaScriptCore
  • Fixed <rdar://problem/4306000> Mail and Safari will launch faster if JavaScriptCore has no weak exports (typeinfo for subclasses in WebCore)

Reviewed by Darin.

  • as: Wrapper for the system assembler that supresses all type-info .weak_definition lines
  • gcc-path: small wrapper that sets the PATH before calling gcc, this makes gcc use our as
  • JavaScriptCore.xcodeproj/project.pbxproj: set CC and CPLUSPLUS to point to gcc-path only in Default configuration
12:53 PM Changeset in webkit [10972] by hyatt
  • 2 edits in trunk/LayoutTests/fast/css

Revise for overflow backout.

12:52 PM Changeset in webkit [10971] by hyatt
  • 1 edit in trunk/LayoutTests/fast/table/overflowHidden-expected.txt

Fix overflow test.

12:52 PM Changeset in webkit [10970] by hyatt
  • 1 edit in trunk/LayoutTests/fast/clip/008-expected.txt

Fix clip test now that overflow change was backed out.

12:48 PM Changeset in webkit [10969] by hyatt
  • 1 edit
    2 adds in trunk/LayoutTests/fast/forms

Add indeterminate test. Fix 001 test.

12:44 PM Changeset in webkit [10968] by hyatt
  • 12 edits in trunk/WebCore

Add support for the indeterminate boolean for placing checkboxes
into a mixed state. This is a feature of WinIE.

Along with this feature, add support for the CSS3 :indeterminate
selector, although one wonders how this could be part of the selectors spec
when no DOM standard covers the ability to make a mixed checkbox
in the first place.

This implementation may seem like it's overlooking some things, but it's not.
Specifically "indeterminate" has no effect on form submission, it does not
get cleared by a form reset, and the pre/post reversal code for undoing clicks
only resets the state that changed in WinIE. This is all bizarre behavior,
but this is a WinIE extension, so we're going to match.

Reviewed by mjs

fast/forms/indeterminate.html

  • khtml/css/css_base.cpp: (CSSSelector::extractPseudoType):
  • khtml/css/css_base.h: (DOM::CSSSelector::):
  • khtml/css/cssstyleselector.cpp: (khtml::CSSStyleSelector::checkOneSelector):
  • khtml/ecma/kjs_html.cpp: (KJS::HTMLElement::inputGetter): (KJS::HTMLElement::inputSetter):
  • khtml/ecma/kjs_html.h: (KJS::HTMLElement::):
  • khtml/html/html_formimpl.cpp: (DOM::HTMLInputElementImpl::init): (DOM::HTMLInputElementImpl::setIndeterminate): (DOM::HTMLInputElementImpl::preDispatchEventHandler): (DOM::HTMLInputElementImpl::postDispatchEventHandler):
  • khtml/html/html_formimpl.h: (DOM::HTMLInputElementImpl::isIndeterminate): (DOM::HTMLInputElementImpl::indeterminate):
  • khtml/rendering/render_theme.cpp: (khtml::RenderTheme::isIndeterminate):
  • khtml/rendering/render_theme.h:
  • khtml/rendering/render_theme_mac.mm: (khtml::RenderThemeMac::updateCheckedState): (khtml::RenderThemeMac::setCheckboxCellState):
  • khtml/xml/dom_nodeimpl.h: (DOM::NodeImpl::isIndeterminate):
12:23 AM Changeset in webkit [10967] by mjs
  • 2 edits in trunk/WebCore

Re-landed fix for the following bug:

<rdar://problem/4302874> crash repeatedly reloading www.supermanhomepage.com (fixed in Denver, broken on TOT)

  • khtml/ecma/kjs_events.cpp: (KJS::JSAbstractEventListener::handleEvent):

Oct 25, 2005:

11:10 PM Changeset in webkit [10966] by darin
  • 7 edits in trunk

LayoutTests:

  • dom/xhtml/level1/core/hc_documentgetdoctype-expected.txt: Expect success.

WebCore:

Reviewed by Eric.

  • khtml/html/html_documentimpl.cpp: (DOM::HTMLDocumentImpl::determineParseMode): Use the new setDocType function instead of modifying m_doctype directly.
  • khtml/xml/dom_docimpl.h: Changed to use SharedPtr more. (DOM::DocumentImpl::realDocType): Add get() since we're using SharedPtr. (DOM::DocumentImpl::setDocType): Added.
  • khtml/xml/dom_docimpl.cpp: (DOMImplementationImpl::createDocument): Use setDocType instead of using realDocType()->copyFrom to set up the document type in the new document. (DocumentImpl::DocumentImpl): Don't create an empty document type during constructor. (DocumentImpl::~DocumentImpl): Remove explicit deref since we're using SharedPtr. (DocumentImpl::doctype): Add get() since we're using SharedPtr. (DocumentTypeImpl::DocumentTypeImpl): Streamlined implementation since we're using SharedPtr.
  • khtml/xml/xml_tokenizer.cpp: (khtml::XMLTokenizer::internalSubset): Added. Calls setDocType. (khtml::internalSubsetHandler): Added. Calls xmlSAX2InternalSubset after calling our function. (khtml::XMLTokenizer::finish): Changed to pass our internalSubsetHandler function pointer.
11:08 PM Changeset in webkit [10965] by hyatt
  • 5 edits in trunk/WebCore

This patch makes a number of refinements to buttons to make
the custom <input>s on hotwire.com look correct.

Also back out my fix for bug 5283, since it causes some bad
regressions and basically needs to be rethought. (This part
not reviewed, since it's just a backout.)

Reviewed by darin

  • khtml/css/html4.css: Added rules to reset many more properties on form controls, including text-align, letter-spacing, word-spacing, line-height text-transform and others. Remove the !important from the line-height reset, since WinIE honors line-height on buttons. Mozilla does not, but we will match WinIE.
  • khtml/rendering/render_button.cpp: (khtml::RenderButton::setStyle): Make sure to preserve the box-flex of our generated interior block, so that it doesn't end up packing left inside its container when styles change.

(khtml::RenderButton::paintObject):
Change the clipping heuristic to only clip to the border box.
This does mean text can run right up to the edge of an Aqua
button, but without this change we just don't match other
browsers.

  • khtml/rendering/render_flexbox.cpp: (khtml::RenderFlexibleBox::layoutHorizontalBox): (khtml::RenderFlexibleBox::layoutVerticalBox): Change the behavior of box-align:center. If it would result in spillage out the top or left when centering, then don't move it.
11:05 PM Changeset in webkit [10964] by mjs
  • 8 edits in trunk/LayoutTests

Update isSameNode results back, now that the fix is landed.

  • dom/xhtml/level3/core/nodeissamenode01-expected.txt:
  • dom/xhtml/level3/core/nodeissamenode02-expected.txt:
  • dom/xhtml/level3/core/nodeissamenode03-expected.txt:
  • dom/xhtml/level3/core/nodeissamenode04-expected.txt:
  • dom/xhtml/level3/core/nodeissamenode05-expected.txt:
  • dom/xhtml/level3/core/nodeissamenode06-expected.txt:
  • dom/xhtml/level3/core/nodeissamenode10-expected.txt:
9:55 PM Changeset in webkit [10963] by mjs
  • 5 edits in trunk/WebCore

Re-landing the isSameNode fix from Anders, I verified there is no effect on performance:

<rdar://problem/4302880> Re-land isSameNode change

  • khtml/ecma/kjs_dom.cpp:
  • khtml/ecma/kjs_dom.h:
  • khtml/xml/dom_nodeimpl.h:
  • kwq/DOM.mm:
4:46 PM Changeset in webkit [10962] by mjs
  • 2 edits in trunk/WebCore

Re-land the following fix, I tested to verify there is no performance regression:

Makes <object> tags not use image mode for svg content:
​http://bugzilla.opendarwin.org/show_bug.cgi?id=5175

  • khtml/html/html_objectimpl.cpp: (DOM::HTMLObjectElementImpl::isImageType):
3:33 PM Changeset in webkit [10961] by adele
  • 2 edits in trunk/WebCore

Reviewed by Hyatt.

  • fixed <rdar://problem/3612121> setting focus, by tabbing, searching, or with JavaScript does not scroll overflow:auto/scroll/overlay to reveal focused element (3480)
  • khtml/rendering/render_layer.cpp: (khtml::RenderLayer::scrollRectToVisible): Added case for when the renderer has an overflow clip, so we can scroll overflow and views recursively.
3:32 PM Changeset in webkit [10960] by adele
  • 5 edits in trunk/WebCore

Reviewed by Vicki.

Fixed <rdar://problem/4098450> RoboHelp-generated html help system crashes in latest Safari -KWQValueListImpl::copyOnWrite

The tokenizer's timer was causing the tokenizer to be deleted twice.
In timerEvent, we'll now check to see if the write has destroyed the tokenizer before trying to do it again.

  • khtml/html/htmltokenizer.cpp: (khtml::HTMLTokenizer::write): Returns a boolean to indicate whether end() gets called (khtml::HTMLTokenizer::timerEvent): Moved code from allDataProcessed, and now, checks to see if write() called end() to notify WebKit that processing is done.
  • khtml/html/htmltokenizer.h: write() returns a bool.
  • khtml/xml/xml_tokenizer.cpp: (khtml::XMLTokenizer::write): returns a bool (always false in the XMLTokenizer case).
  • khtml/xml/xml_tokenizer.h: ditto.
3:26 PM Changeset in webkit [10959] by bdakin
  • 4 edits
    4 adds in trunk

* empty log message *

2:40 PM Changeset in webkit [10958] by hyatt
  • 1 edit in trunk/LayoutTests/fast/forms/001-expected.txt

Commit changed forms 001.html test to account for box-sizing and clipping.

2:14 PM Changeset in webkit [10957] by adele
  • 17 edits
    2 deletes in trunk/WebCore

Reviewed by Hyatt.

  • first step of fix for <rdar://problem/3612121> setting focus, by tabbing, searching, or with JavaScript does not scroll overflow:auto/scroll/overlay to reveal focused element (3480) This change moves the scrolling code for revealing elements to the RenderLayer, instead of being specific to the view.
  • khtml/rendering/render_layer.h: Added scrollRectToVisible. Added getRectToExpose. Added ScrollAlignment enum so callers can be more specific about exactly where they want to scroll. (khtml::RenderLayer::scrollToPoint): Added inline function to scroll to a point.
  • khtml/rendering/render_layer.cpp: (khtml::RenderLayer::scrollRectToVisible): Replaces recursive view scrolling functions. (khtml::RenderLayer::getRectToExpose): Calculates the rectangle to expose based on the alignment parameters.
  • khtml/khtml_part.cpp: (KHTMLPart::gotoAnchor): Changed use of setContentsPosRecursive to scrollToPoint. (KHTMLPart::setActiveNode): Changed use of ensureVisible to scrollRectToVisible.
  • khtml/khtmlview.cpp: (KHTMLView::doAutoScroll): Changed use of ensureVisible to scrollRectToVisible. (KHTMLView::focusNextPrevNode): ditto.
  • kwq/KWQKHTMLPart.mm: (KWQKHTMLPart::jumpToSelection): Changed use of _KWQ_scrollRectToVisible to scrollRectToVisible. (KWQKHTMLPart::nextKeyViewInFrame): Changed use of ensureRectVisibleCentered to scrollRectToVisible. (KWQKHTMLPart::centerSelectionInVisibleArea): ditto.
  • khtml/xml/dom_elementimpl.cpp: (ElementImpl::scrollIntoView): Changed use of setContentsPos and ensureVisible to scrollRectToVisible.
  • kwq/WebCoreBridge.mm: (-[WebCoreBridge ensureSelectionVisible]): Changed use of ensureRectVisibleCentered to scrollRectToVisible.
  • kwq/KWQComboBox.mm: (-[KWQPopUpButton becomeFirstResponder]): Changing use of _KWQ_scrollFrameToVisible to scrollRectToVisible.
  • kwq/KWQFileButton.mm: (KWQFileButton::focusChanged): ditto.
  • kwq/KWQListBox.mm: (-[KWQTableView becomeFirstResponder]): ditto.
  • kwq/KWQSlider.mm: (-[KWQSlider becomeFirstResponder]): ditto.
  • kwq/KWQTextArea.mm: (-[KWQTextAreaTextView becomeFirstResponder]): ditto.
  • kwq/KWQTextField.mm: (-[KWQTextFieldController setHasFocus:]): ditto.
  • kwq/KWQScrollView.h: Removed ensureVisible, and ensureRectVisibleCentered. Added ensureRectVisible which just calls across to AppKit to scroll the NSView.
  • kwq/KWQScrollView.mm: (QScrollView::ensureRectVisible): ditto.
  • WebCore.xcodeproj/project.pbxproj: Removed KWQNSViewExtras since it is no longer needed.
  • kwq/KWQNSViewExtras.h: Removed.
  • kwq/KWQNSViewExtras.m: Removed.
1:56 PM Changeset in webkit [10956] by vicki
  • 16 edits in trunk/LayoutTests
  • updated results for <rdar://problem/4288276> DOM tests expect hierarchy exception adding wrong types of nodes to parents (4568)
  • dom/html/level1/core/hc_attrappendchild2-expected.txt:
  • dom/html/level1/core/hc_nodeappendchildinvalidnodetype-expected.txt:
  • dom/html/level1/core/hc_nodeinsertbeforeinvalidnodetype-expected.txt:
  • dom/html/level1/core/hc_nodereplacechildinvalidnodetype-expected.txt:
  • dom/xhtml/level1/core/hc_attrappendchild2-expected.txt:
  • dom/xhtml/level1/core/hc_attrappendchild4-expected.txt:
  • dom/xhtml/level1/core/hc_attrinsertbefore5-expected.txt:
  • dom/xhtml/level1/core/hc_attrinsertbefore7-expected.txt:
  • dom/xhtml/level1/core/hc_nodeappendchildinvalidnodetype-expected.txt:
  • dom/xhtml/level1/core/hc_nodeinsertbeforeinvalidnodetype-expected.txt:
  • dom/xhtml/level1/core/hc_nodereplacechildinvalidnodetype-expected.txt:
  • dom/xhtml/level3/core/nodeinsertbefore03-expected.txt:
  • dom/xhtml/level3/core/nodeinsertbefore20-expected.txt:
  • dom/xhtml/level3/core/nodereplacechild20-expected.txt:
  • dom/xhtml/level3/core/nodereplacechild35-expected.txt:
1:54 PM Changeset in webkit [10955] by hyatt
  • 10 edits in trunk/WebCore

Land box-sizing support and change button to work correctly
with it.

Also stop honoring of font properties on <input> if
Aqua is enabled. If you allow bold, but don't honor size, then
you end up with poorly rendered buttons on cnn.com (box-sizing
made this apparent, since the buttons got smaller).

Fix the clipping code for button content to actually work. :)
Function needed to be designated virtual in order to get
called.

Reviewed by darin

fast/box-sizing/box-sizing.html is the new test case.

  • ChangeLog:
  • WebCore.xcodeproj/project.pbxproj:
  • khtml/rendering/render_block.cpp: (khtml::RenderBlock::calcMinMaxWidth):
  • khtml/rendering/render_block.h:
  • khtml/rendering/render_box.cpp: (RenderBox::calcBorderBoxWidth): (RenderBox::calcBorderBoxHeight): (RenderBox::calcContentBoxWidth): (RenderBox::calcContentBoxHeight): (RenderBox::calcWidth): (RenderBox::calcWidthUsing): (RenderBox::calcHeight): (RenderBox::calcHeightUsing): (RenderBox::calcPercentageHeight): (RenderBox::calcReplacedWidth): (RenderBox::calcReplacedWidthUsing): (RenderBox::calcReplacedHeightUsing): (RenderBox::availableHeightUsing): (RenderBox::calcAbsoluteHorizontalValues): (RenderBox::calcAbsoluteVerticalValues):
  • khtml/rendering/render_box.h:
  • khtml/rendering/render_button.cpp: (khtml::RenderButton::paintObject):
  • khtml/rendering/render_button.h:
  • khtml/rendering/render_flexbox.cpp: (khtml::RenderFlexibleBox::calcMinMaxWidth):
  • khtml/rendering/render_theme_mac.mm: (khtml::RenderThemeMac::setFontFromControlSize): (khtml::RenderThemeMac::adjustButtonStyle):
1:50 PM Changeset in webkit [10954] by hyatt
  • 5 adds in trunk/LayoutTests/fast/box-sizing

Commit new box-sizing test

1:49 PM Changeset in webkit [10953] by vicki
  • 2 edits in trunk/WebCore

Reviewed by Hyatt.

  • fix <rdar://problem/4288276> DOM tests expect hierarchy exception adding wrong types of nodes to parents (4568)

Check that the node being added is an allowable child. If inserting a document fragment node, check each child of
the node. Fixes 4 of the W3C DOM HTML tests and 11 of the XHTML tests.

  • khtml/xml/dom_nodeimpl.cpp: (DOM::NodeImpl::checkAddChild):
12:32 PM Changeset in webkit [10952] by thatcher
  • 2 edits in trunk/WebKit

Reviewed by Darin.

Remove the use of a stamp file in the Frameworks symlink build phase.
This lets us pass the build verification.

  • WebKit.xcodeproj/project.pbxproj:
9:39 AM Changeset in webkit [10951] by darin
  • 6 edits
    2 adds in trunk

LayoutTests:

  • fast/events/init-event-after-dispatch-expected.txt: Added.
  • fast/events/init-event-after-dispatch.html: Added.

WebCore:

Reviewed by Dave Hyatt.

  • fixed ​http://bugzilla.opendarwin.org/show_bug.cgi?id=4945 event init calls should not do anything if the event has already been dispatched
  • made a small performance improvement to event creation by simplifying the date code (seen on profiles)
  • changed DOMTimestamp to be a 64-bit integer on Mac OS X (requested by the DOM standard)
  • khtml/dom/dom_node.h: Made DOMTimestamp be unsigned long long on Mac OS X.
  • khtml/xml/dom_nodeimpl.cpp: (DOM::NodeImpl::dispatchGenericEvent): Assert that the event has a target. (DOM::NodeImpl::dispatchWindowEvent): Set the event target to the document.
  • khtml/xml/dom2_eventsimpl.h: Changed many data members to use SharedPtr instead of explicit ref/deref. Changed m_createTime to be a DOMTimeStamp instead of a QDateTime. (DOM::EventImpl::setTarget): Now inlined. (DOM::EventImpl::timeStamp): Now inlined. (DOM::EventImpl::preventDefault): Now inlined. (DOM::EventImpl::dispatched): Added. Returns true if the event was already dispatched, meaning it already has a target.
  • khtml/xml/dom2_eventsimpl.cpp: (DOM::currentTimeStamp): Added. Faster inline version on Mac OS X, since this shows up in profiles. (DOM::EventImpl::EventImpl): Changed to use contructor syntax and set m_createTime to currentTimeStamp(). (DOM::EventImpl::~EventImpl): Removed explicit derefs; no longer needed. (DOM::EventImpl::initEvent): Do nothing if dispatched() is true. Simplify since SharedPtr handles ref/deref. (DOM::UIEventImpl::UIEventImpl): Changed to use contructor syntax and take advantage of the fact that some data members are now SharedPtr. (DOM::UIEventImpl::initUIEvent): Do nothing if dispatched() is true. Simplify since SharedPtr handles ref/deref. (DOM::MouseEventImpl::MouseEventImpl): Changed to use contructor syntax and take advantage of the fact that some data members are now SharedPtr. (DOM::MouseEventImpl::~MouseEventImpl): Removed explicit derefs; no longer needed. (DOM::MouseEventImpl::initMouseEvent): Do nothing if dispatched() is true. Simplify since SharedPtr handles ref/deref. (DOM::MouseEventImpl::isDragEvent): Use a local variable. (DOM::KeyboardEventImpl::KeyboardEventImpl): Changed to use contructor syntax and take advantage of the fact that some data members are now SharedPtr. (DOM::KeyboardEventImpl::~KeyboardEventImpl): Removed explicit derefs; no longer needed. (DOM::KeyboardEventImpl::initKeyboardEvent): Do nothing if dispatched() is true. Simplify since SharedPtr handles ref/deref. (DOM::MutationEventImpl::MutationEventImpl): Changed to use contructor syntax and take advantage of the fact that some data members are now SharedPtr. (DOM::MutationEventImpl::initMutationEvent): Do nothing if dispatched() is true. Simplify since SharedPtr handles ref/deref. (DOM::ClipboardEventImpl::ClipboardEventImpl): Simplify since SharedPtr handles initialization and ref/deref. (DOM::RegisteredEventListener::RegisteredEventListener): Ditto.
Note: See TracTimeline for information about the timeline view.