Timeline



Nov 3, 2003:

5:48 PM Changeset in webkit [5378] by cblu
  • 6 edits in trunk/WebKit

Fixed:
<rdar://problem/3472377>: Provide NSRTFDPboardType on pasteboard when copying or dragging images
<rdar://problem/3470809>: REGRESSION (111-112): Can't copy & paste image into Photoshop 7

Reviewed by hyatt.

  • Misc.subproj/WebNSPasteboardExtras.h:
  • Misc.subproj/WebNSPasteboardExtras.m: (-[NSPasteboard _web_writeFileDataAsRTFDAttachment:withFilename:]): renamed, now writes file data as an RTF attachment
  • Misc.subproj/WebNSViewExtras.m: (-[NSView _web_dragImage:originalData:rect:URL:title:event:]): call renamed _web_writeFileDataAsRTFDAttachment
  • WebView.subproj/WebDefaultContextMenuDelegate.m: (-[WebDefaultUIDelegate copyImageToClipboard:]): call renamed _web_writeFileDataAsRTFDAttachment
  • WebView.subproj/WebImageView.m: (-[WebImageView writeImageToPasteboard:]): call renamed _web_writeFileDataAsRTFDAttachment
4:50 PM Changeset in webkit [5377] by mjs
  • 5 edits in trunk/WebCore

Reviewed by Darin.

  • fixed 3370085 - new frames with javascript: URLs should be populated with the result of the JavaScript code
  • ChangeLog:
  • khtml/html/html_baseimpl.cpp: (HTMLFrameElementImpl::updateForNewURL): Skip special filtering of javascript:. (HTMLFrameElementImpl::attach): Ditto.
  • khtml/khtml_part.cpp: (KHTMLPart::replaceContentsWithScriptResult): New method to execute script and if the result is a string, set it as the new document contents. (KHTMLPart::requestFrame): Removed APPLE_CHANGES around javascript: handling. Use new replaceContentsWithScriptResult method. This executes the JS in the new child frame, not the parent frame, as in other browsers.
  • khtml/khtml_part.h: Prototype new method.
  • kwq/KWQKHTMLPartBrowserExtension.mm: (KHTMLPartBrowserExtension::openURLRequest): Use new part method.
4:49 PM Changeset in webkit [5376] by vicki
  • 5 edits in trunk

JavaScriptCore:

Reviewed by kocienda.

  • fixed <rdar://problem/3471096>: non-B&I builds should not use order files, because they cause false "regressions" in perf.
  • JavaScriptCore.pbproj/project.pbxproj: added empty SECTORDER_FLAGS variables to the Development and Deployment build styles

WebCore:

Reviewed by kocienda.

  • fixed <rdar://problem/3471096>: non-B&I builds should not use order files, because they cause false "regressions" in perf.
  • WebCore.pbproj/project.pbxproj: added empty SECTORDER_FLAGS variables to the Development and Deployment build styles

WebKit:

Reviewed by kocienda.

  • fixed <rdar://problem/3471096>: non-B&I builds should not use order files, because they cause false "regressions" in perf.
  • WebKit.pbproj/project.pbxproj: added empty SECTORDER_FLAGS variables to the Development and Deployment build styles

WebBrowser:

Reviewed by kocienda.

  • fixed <rdar://problem/3471096>: non-B&I builds should not use order files, because they cause false "regressions" in perf.
  • WebBrowser.pbproj/project.pbxproj: added empty SECTORDER_FLAGS variables to the Development and Deployment build styles
4:42 PM Changeset in webkit [5375] by kocienda
  • 3 edits in trunk/WebCore

Reviewed by Maciej

Fix for this bug:

<rdar://problem/3470338>: onclick is not executing when you hit enter on focused links

Now when we trap return or enter when a link is focused, we send
a click event to the node.

  • khtml/html/html_inlineimpl.cpp: (HTMLAnchorElementImpl::defaultEventHandler): Swallow key event when return or enter and send a mouse event. (HTMLAnchorElementImpl::performClick): New function to send simulated click.
  • khtml/html/html_inlineimpl.h: Declare performClick function.
4:40 PM Changeset in webkit [5374] by vicki
  • 2 edits in trunk/JavaScriptCore

JavaScriptCore:

Reviewed by kocienda.

  • fixed <rdar://problem/3471096>: non-B&I builds should not use order files, because they cause false "regressions" in perf.
  • JavaScriptCore.pbproj/project.pbxproj: added empty SECTORDER_FLAGS variables to the Development and Deployment build styles

WebCore:

Reviewed by kocienda.

  • fixed <rdar://problem/3471096>: non-B&I builds should not use order files, because they cause false "regressions" in perf.
  • WebCore.pbproj/project.pbxproj: added empty SECTORDER_FLAGS variables to the Development and Deployment build styles

WebKit:

Reviewed by kocienda.

  • fixed <rdar://problem/3471096>: non-B&I builds should not use order files, because they cause false "regressions" in perf.
  • WebKit.pbproj/project.pbxproj: added empty SECTORDER_FLAGS variables to the Development and Deployment build styles

WebBrowser:

Reviewed by kocienda.

  • fixed <rdar://problem/3471096>: non-B&I builds should not use order files, because they cause false "regressions" in perf.
  • WebBrowser.pbproj/project.pbxproj: added empty SECTORDER_FLAGS variables to the Development and Deployment build styles
3:44 PM Changeset in webkit [5373] by darin
  • 16 edits in trunk/WebCore

Reviewed by John.

  • fixed 3392543 -- incorrect base writing direction in RTL form controls
  • fixed baseline for list box <select> elements
  • moved style code from updateFromElement to setStyle
  • khtml/rendering/render_form.h: Added setStyle function to RenderFormElement, RenderLineEdit, RenderSelect, and RenderTextArea. Removed some unused stuff.
  • khtml/rendering/render_form.cpp: (RenderFormElement::setStyle): Set the font based on the style. (RenderFormElement::updateFromElement): Remove font setting code; moved to setStyle. Also ifdef'd out the color-setting code, which belongs in setStyle and is irrelevant for KWQ anyway. (RenderSubmitButton::setStyle): Set the writing direction based on the style. (RenderSubmitButton::updateFromElement): Tiny code cleanup, using a local variable. (RenderLineEdit::setStyle): Set the alignment and writing direction based on the style. (RenderLineEdit::updateFromElement): Remove alignment setting code; moved to setStyle. (RenderFileButton::RenderFileButton): Remove code to set m_focus, a data member that was never used. (RenderSelect::setWidgetWritingDirection): Added. Helper function. (RenderSelect::setStyle): Call setWidgetWritingDirection. (RenderSelect::updateFromElement): Call setWidgetWritingDirection if a new widget is created. (RenderSelect::baselinePosition): Added a FIXME comment. (RenderTextArea::setStyle): Set the alignment and writing direction based on the style. (RenderTextArea::updateFromElement): Remove alignment setting code; moved to setStyle.
  • kwq/KWQButton.h: Added setWritingDirection function.
  • kwq/KWQButton.mm: (+[KWQButton cellClass]): Use KWQButtonCell. (-[KWQButtonCell baseWritingDirection:]): Return base writing direction. (-[KWQButtonCell setBaseWritingDirection:]): Store base writing direction. (-[KWQButtonCell _textAttributes]): Set the base writing direction in the paragraph style if it's not already right. (QButton::setWritingDirection): Call setBaseWritingDirection: on the cell and do setNeedsDisplay: as needed.
  • kwq/KWQComboBox.h: Added setWritingDirection function.
  • kwq/KWQComboBox.mm: (QComboBox::setWritingDirection): Call setBaseWritingDirection: on the cell and do setNeedsDisplay: as needed. (-[KWQPopUpButtonCell setBaseWritingDirection:]): Store base writing direction. (-[KWQPopUpButtonCell baseWritingDirection:]): Return base writing direction. (-[KWQPopUpButtonCell _textAttributes]): Set the base writing direction in the paragraph style if it's not already right.
  • kwq/KWQLineEdit.h: Added setWritingDirection function.
  • kwq/KWQLineEdit.mm: (QLineEdit::setAlignment): Tweaked to match new setWritingDirection in style. (QLineEdit::setWritingDirection): Call setBaseWritingDirection: on the KWQTextField.
  • kwq/KWQTextField.h: Added setBaseWritingDirection method.
  • kwq/KWQTextField.mm: (-[KWQTextField setPasswordMode:]): Set the base writing direction of the secure field's cell based on the base writing direction of the parent field's cell, when creating the secure field. (-[KWQTextField setBaseWritingDirection:]): Added. Sets the base writing direction of the field's cell and the secure field's cell, if it exists, and does setNeedsDisplay as needed. (-[KWQTextFieldCell setBaseWritingDirection:]): Added. Stores the base writing direction. (-[KWQTextFieldCell baseWritingDirection]): Added. Returns the base writing direction. (-[KWQTextFieldCell _textAttributes]): Set the base writing direction in the paragraph style if it's not already right. (-[KWQSecureTextFieldCell setBaseWritingDirection:]): Added. Stores the base writing direction. (-[KWQSecureTextFieldCell _textAttributes]): Set the base writing direction in the paragraph style if it's not already right.
  • kwq/KWQListBox.h: Added itemIsGroupLabel and setWritingDirection functions, and made the private insertItem function take a QString instead of an NSObject. (QListBox::insertItem): Changed parameter from unsigned to int to match Qt, and made it use an inlne to call the private insertItem function shared with insertGroupLabel. (QListBox::insertGroupLabel): Changed parameter from unsigned to int and made it use an inline to call the private insertItem function.
  • kwq/KWQListBox.mm: (itemFont): Added. Returns a suitable font for individual items. (groupLabelFont): Added. Returns a suitable font for group labels. (paragraphStyle): Added. Returns a suitable paragraph style for a given writing direction. (stringAttributes): Added. Returns a string attributes dictionary for a given writing direction for either individual items or group labels. (QListBox::insertItem): Use an attributed string to give the item the appropriate font and writing direction, calling stringAttributes to get the attributes dictionary, and getting the writing direction from the table view. (QListBox::endBatchInsert): Tweak for consistency with other nearby methods. (QListBox::setSelected): Add an assertion and tweak a bit. (QListBox::isSelected): Add an assertion and tweak a bit. (QListBox::setWritingDirection): Update the styles for all the attributed strings and call reloadData if the writing direction changes. Also store the direction in the table view. (QListBox::itemIsGroupLabel): Added. Returns true if a given item is a group label; works by checking the font in the attributed string. (-[KWQTableView initWithListBox:items:]): Use itemFont() instead of computing the font. (-[KWQTableView tableView:shouldSelectRow:]): Use itemIsGroupLabel function instead of checking for NSString vs. NSttributedString in the items array; we now use attributed strings for all the items. (-[KWQTableView setBaseWritingDirection:]): Added. Stores a base writing direction. (-[KWQTableView baseWritingDirection]): Added. Returns the stored base writing direction.
  • kwq/KWQTextEdit.h: Added setWritingDirection function.
  • kwq/KWQTextEdit.mm: (QTextEdit::setAlignment): Removed code to set base writing direction. (QTextEdit::setWritingDirection): Added. Sets base writing direction.
  • kwq/KWQTextArea.mm: (-[KWQTextArea setBaseWritingDirection:]): Call setNeedsDisplay:YES.
3:37 PM Changeset in webkit [5372] by hyatt
  • 3 edits in trunk/WebCore

Fix 3472200, media tests fail to parse correctly. Merging a patch to handle exotic media types
from the KHTML trunk.

  • khtml/css/parser.cpp:
  • khtml/css/parser.y:
3:18 PM Changeset in webkit [5371] by hyatt
  • 2 edits in trunk/WebCore

Fix for 3471314, make sure to check the outline phase before painting backgrounds of iframes.

Reviewed by mjs

  • khtml/rendering/render_replaced.cpp: (RenderReplaced::paint):
3:04 PM Changeset in webkit [5370] by hyatt
  • 3 edits in trunk/WebCore

Fix for 3470855, links with script action stay focused when clicking.

Fix for 3470858, clicking and dragging a link and then releasing doesn't clear the glow ring.

Fix for 3472090, link should not be mouse focusable.

Reviewed by mjs

  • khtml/css/html4.css:
  • khtml/html/html_inlineimpl.h: (DOM::HTMLAnchorElementImpl::isMouseFocusable):
2:43 PM Changeset in webkit [5369] by hyatt
  • 2 edits in trunk/WebCore

Fix for 3472030, line-height not properly reset in all cases when the font shorthand is used.

Reviewed by mjs

  • khtml/css/cssparser.cpp: (CSSParser::parseFont):
2:12 PM Changeset in webkit [5368] by mjs
  • 2 edits in trunk/WebCore

Reviewed by Darin.

  • fixed 3472023 - REGRESSION: assertion on layout tests
  • kwq/KWQTextCodec.mm: (KWQTextDecoder::convertUTF16): Remove assert that length is not 0, since we can get zero length now when flushing.
2:01 PM Changeset in webkit [5367] by kocienda
  • 4 edits in trunk/WebCore

Reviewed by John

Fix for this bug:

<rdar://problem/3260780>: hitting return in a text field submits
form without running onclick handler for submit button

  • khtml/html/html_formimpl.cpp: (HTMLFormElementImpl::performSubmitClick): Refactored this function to be more generic, so that text fields can use it, too. (HTMLInputElementImpl::defaultEventHandler): Removed code to check for a submit button, and merged this code into new performSubmitClick. (HTMLSelectElementImpl::defaultEventHandler): Ditto.
  • khtml/html/html_formimpl.h: Removed declaration for simulateButtonClickForEvent. Added declaration for performSubmitClick.
  • khtml/rendering/render_form.cpp: (RenderLineEdit::slotReturnPressed): Now calls performSubmitClick rather than prepareSubmit on the form. This is the key piece for the bug fix.
11:45 AM Changeset in webkit [5366] by kocienda
  • 9 edits in trunk

WebCore:

Reviewed by John

Fix for this bug:

<rdar://problem/3470342>: focus rings are shown for links in
web pages even in non-frontmost windows

  • kwq/KWQKHTMLPart.h: Add setShowsFirstResponder, showsFirstResponder functions and _showsFirstResponder member variable. (KWQKHTMLPart::showsFirstResponder): Added.
  • kwq/KWQKHTMLPart.mm: (KWQKHTMLPart::KWQKHTMLPart): Initialize _showsFirstResponder to true (KWQKHTMLPart::setShowsFirstResponder): Sets whether the painter draws focus rings.
  • kwq/KWQPainter.h: Add _drawsFocusRing member variable. (QPainter::setDrawsFocusRing): Sets whether the painter draws focus rings.
  • kwq/KWQPainter.mm: (QPainter::QPainter): Initialize _drawsFocusRing to true. (QPainter::initFocusRing): Check _drawsFocusRing flag. Return if set to false. (QPainter::addFocusRingRect): Ditto. (QPainter::drawFocusRing): Ditto.
  • kwq/WebCoreBridge.h: Add setShowsFirstResponder method.
  • kwq/WebCoreBridge.mm: (-[WebCoreBridge drawRect:]): Sets the value for _drawsFocusRing on the painter using _showsFirstResponder value from the part. (-[WebCoreBridge setShowsFirstResponder:]): Set passed in value on the part. Redraw the focus ring if flag has changed.

WebKit:

Reviewed by John

Fix for this bug:

<rdar://problem/3470342>: focus rings are shown for links in
web pages even in non-frontmost windows

  • WebView.subproj/WebHTMLView.m: (-[WebHTMLView updateFocusRing]): New method. Uses the "keyness" of the view's window to toggle focus ring drawing. (-[WebHTMLView windowDidBecomeKey:]): Calls updateFocusRing. (-[WebHTMLView windowDidResignKey:]): Ditto.
10:46 AM Changeset in webkit [5365] by hyatt
  • 2 edits in trunk/WebCore

Remove unneeded style property additions for form controls. In the KHTML trunk these moved into the UA
sheet, but we don't even need them there, since we don't honor any of them.

  • khtml/html/html_formimpl.cpp: (HTMLInputElementImpl::attach): (HTMLSelectElementImpl::attach): (HTMLTextAreaElementImpl::attach):
10:35 AM Changeset in webkit [5364] by hyatt
  • 3 edits in trunk/WebCore

Patch from KHTML trunk. Move frameset positioning reset into adjustRenderStyle so as not to
pollute the render tree.

  • khtml/css/cssstyleselector.cpp: (khtml::CSSStyleSelector::adjustRenderStyle):
  • khtml/rendering/render_box.cpp: (RenderBox::setStyle):
10:22 AM Changeset in webkit [5363] by hyatt
  • 9 edits in trunk/WebCore

Fix for 3471609, 'transparent' should be an acceptable named color value instead of a special
keyword used only by background/border.

Reviewed by kocienda

  • khtml/css/cssparser.cpp: (CSSParser::parseValue):
  • khtml/css/cssstyleselector.cpp: (khtml::): (khtml::CSSStyleSelector::applyRule):
  • khtml/css/cssvalues.c:
  • khtml/css/cssvalues.h:
  • khtml/css/cssvalues.in:
  • khtml/rendering/render_box.cpp: (RenderBox::paintBackgroundExtended):
  • khtml/rendering/render_object.cpp: (RenderObject::setStyle):
  • khtml/rendering/render_style.h: (khtml::BorderValue::BorderValue): (khtml::BorderValue::isTransparent): (khtml::BorderValue::operator==): (khtml::CollapsedBorderValue::isTransparent): (khtml::RenderStyle::setBorderLeftColor): (khtml::RenderStyle::setBorderRightColor): (khtml::RenderStyle::setBorderTopColor): (khtml::RenderStyle::setBorderBottomColor):
12:20 AM Changeset in webkit [5362] by mjs
  • 2 edits in trunk/WebCore

Reviewed by Darin.

  • fixed 3471226 - HTML document that's less than 16 bytes comes out blank
  • kwq/KWQTextCodec.mm: (KWQTextDecoder::toUnicode): When flushing, don't bail out early if there are no new characters!

Nov 2, 2003:

10:34 PM Changeset in webkit [5361] by mjs
  • 1 edit in trunk/WebCore/kwq/KWQTextCodec.mm

Revert change I didn't mean to commit.

10:32 PM Changeset in webkit [5360] by mjs
  • 7 edits in trunk/WebCore

Reviewed by Darin.

  • don't resolve Null href or src attributes, since the distinction between empty and absent attribute is important.
  • khtml/dom/html_base.cpp: (HTMLIFrameElement::src):
  • khtml/dom/html_form.cpp: (HTMLInputElement::src):
  • khtml/dom/html_head.cpp: (HTMLBaseElement::href): (HTMLLinkElement::href): (HTMLScriptElement::src):
  • khtml/dom/html_image.cpp: (HTMLAreaElement::href): (HTMLImageElement::src):
  • khtml/dom/html_inline.cpp: (HTMLAnchorElement::href):
9:53 PM Changeset in webkit [5359]
  • 1 copy
    93 deletes in branches/unlabeled-1.95.4

This commit was manufactured by cvs2svn to create branch
'unlabeled-1.95.4'.

9:53 PM Changeset in webkit [5358] by hyatt
  • 15 edits in trunk/WebCore

Fix for 3468924, an implementation of outline-offset. We also give :focus links a built-in offset of
outline-width/2 by default for the Apple Aqua style.

Fix for 3470660, input type=image excluded from tab order.

Reviewed by darin

  • khtml/css/cssparser.cpp: (CSSParser::parseValue):
  • khtml/css/cssproperties.c: (hash_prop): (findProp):
  • khtml/css/cssproperties.h:
  • khtml/css/cssproperties.in:
  • khtml/css/cssstyleselector.cpp: (khtml::CSSStyleSelector::applyRule):
  • khtml/css/html4.css:
  • khtml/rendering/render_box.cpp: (RenderBox::getAbsoluteRepaintRect): (RenderBox::computeAbsoluteRepaintRect):
  • khtml/rendering/render_flow.cpp: (RenderFlow::getAbsoluteRepaintRect):
  • khtml/rendering/render_inline.cpp: (RenderInline::paintFocusRing): (RenderInline::paintOutlines): (RenderInline::paintOutline):
  • khtml/rendering/render_object.cpp: (RenderObject::paintOutline):
  • khtml/rendering/render_style.h: (khtml::OutlineValue::OutlineValue): (khtml::RenderStyle::outlineSize): (khtml::RenderStyle::outlineOffset): (khtml::RenderStyle::setOutlineOffset):
  • khtml/rendering/render_text.cpp: (RenderText::paintObject):
  • kwq/KWQPainter.h:
  • kwq/KWQPainter.mm: (QPainterPrivate::QPainterPrivate): (QPainter::initFocusRing): (QPainter::addFocusRingRect):
6:03 PM Changeset in webkit [5357] by mjs
  • 6 edits in trunk/WebKit

Reviewed by Darin.

  • remove use of obsolete HTTLCookiePolicyBaseURL SPI
  • WebCoreSupport.subproj/WebSubresourceClient.m: startLoadingResource:withURL:referrer:forDataSource:]): Use setMainDocumentURL, not setHTTPCookiePolicyBaseURL.
  • WebView.subproj/WebFrame.m: (-[WebFrame _addExtraFieldsToRequest:alwaysFromRequest:]): Likewise.
  • WebView.subproj/WebMainResourceClient.m: (-[WebMainResourceClient connection:willSendRequest:redirectResponse:]): Likewise.
2:51 PM Changeset in webkit [5356] by darin
  • 7 edits in trunk/JavaScriptCore

Reviewed by Maciej.

  • changed list manipulation to use Harri Porten's idea of a circular linked list that is built from head to tail rather than building the list backwards and reversing the list when done
  • kjs/grammar.y: Handle CatchNode and FinallyNode in a type-safe way. Change many places that passed 0L to pass nothing at all, or to pass 0.
  • kjs/nodes.h: (KJS::ElementNode::ElementNode): Build a circular list instead of a 0-terminated backwards list. (KJS::ArrayNode::ArrayNode): Break the circular list instead of reversing the list. (KJS::PropertyValueNode::PropertyValueNode): Moved before ObjectLiteralNode so the inline code in ObjectLiteralNode works. Build a circular list instead of a 0-terminated backwards list. Made the case for the first node separate so we don't need a nil check. (KJS::ObjectLiteralNode::ObjectLiteralNode): Break the circular list instead of reversing the list. (KJS::ArgumentListNode::ArgumentListNode): Build a circular list instead of a 0-terminated backwards list. Also, made the constructors inline (moved here from .cpp file). (KJS::ArgumentsNode::ArgumentsNode): Break the circular list instead of reversing the list. (KJS::NewExprNode::NewExprNode): Changed a 0L to 0. (KJS::StatListNode::StatListNode): Make this constructor no longer inline (moved into .cpp file). The one in the .cpp file builds a circular list instead of a 0-terminated backwards list. (KJS::VarDeclListNode::VarDeclListNode): Build a circular list instead of a 0-terminated backwards list. (KJS::VarStatementNode::VarStatementNode): Break the circular list instead of reversing the list. (KJS::BlockNode::BlockNode): Make this constructor no longer inline (moved into .cpp file). The one in the .cpp file breaks the list instead of reversing it. (KJS::ForNode::ForNode): Break the circular list instead of reversing the list. (KJS::CaseClauseNode::CaseClauseNode): Break the circular list instead of reversing the list. (KJS::ClauseListNode::ClauseListNode): Build a circular list instead of a 0-terminated backwards list. (KJS::CaseBlockNode::CaseBlockNode): Make this constructor no longer inline (moved into .cpp file). The one in the .cpp file breaks the list instead of reversing it. (KJS::TryNode::TryNode): Changed constructor to take typed parameters for the catch and finally nodes rather than just Node. (KJS::ParameterNode::ParameterNode): Build a circular list instead of a 0-terminated backwards list. (KJS::FuncDeclNode::FuncDeclNode): Break the circular list instead of reversing the list. (KJS::FuncExprNode::FuncExprNode): Break the circular list instead of reversing the list.
  • kjs/nodes.cpp: (StatListNode::StatListNode): Moved this constructor here, no longer inline. Did the "break circular list" thing instead of the "reverse list" thing. Added setLoc calls to match KJS in the KDE tree; since we don't currently use the JavaScript debugging support, it's unclear whether there's any benefit, but later we might be using it and it's good to be as close as possible. (BlockNode::BlockNode): Moved this constructor here, no longer inline. Did the "break circular list" thing instead of the "reverse list" thing. Added setLoc calls. (CaseBlockNode::CaseBlockNode): Moved this constructor here, no longer inline. Did the "break circular list" thing instead of the "reverse list" thing. (SourceElementsNode::SourceElementsNode): Moved this constructor here, no longer inline. Did the "break circular list" thing instead of the "reverse list" thing. Added setLoc calls.
  • kjs/grammar.cpp: Regenerated.
  • kjs/grammar.cpp.h: Regenerated.
  • kjs/grammar.h: Regenerated.

Nov 1, 2003:

9:37 PM Changeset in webkit [5355] by darin
  • 2 edits in trunk/WebKit

Reviewed by Maciej.

  • fixed 3470882 -- storage leaks in WebDownload code
  • fixed 3470884 -- download is always nil in downloadWindowForAuthenticationSheet: call from WebDownload
  • Misc.subproj/WebDownload.m: (-[WebDownloadInternal initWithDownload:]): Removed this method, which was never called. (-[WebDownloadInternal dealloc]): Added missing call to [super dealloc] to fix one cause of a leak of the WebDownloadInternal object itself. Removed the release of webDownload, which was always nil, and if it wasn't would end up causing a leak due to a reference cycle. (-[WebDownloadInternal download:didReceiveAuthenticationChallenge:]): Remove the use of webDownload, which was always nil, and instead use the download parameter passed to us, casting it to WebDownload, since it's guaranteed to be one. (-[WebDownload _setRealDelegate:]): Added. Shared by the methods below to set up the real delegate before calling init. The old code called init twice, causing an second call to the superclass's init method, which caused it to create an extra copy of its internal structure, as well as causing us to create two WebDownloadInternal objects. (-[WebDownload init]): Don't allocate a second WebDownloadInternal if _setRealDelegate already allocated it for us. Before we would allocate and leak an extra one each time. (-[WebDownload dealloc]): Added. Releases the WebDownloadInternal. This is the second cause of the leak of the WebDownloadInternal object. (-[WebDownload initWithRequest:delegate:]): Call [self _setRealDelegate:] instead of calling [self init] and then [_webInternal setRealDelegate:], avoiding the leaks caused by doing it the other way. (-[WebDownload _initWithLoadingConnection:request:response:delegate:proxy:]): Ditto. (-[WebDownload _initWithLoadingResource:request:response:delegate:proxy:]): Ditto. (-[WebDownload _initWithRequest:delegate:directory:]): Ditto.
7:53 PM Changeset in webkit [5354] by mjs
  • 2 edits in trunk/WebCore

Reviewed by Darin.

  • fixed 3469262 - REGRESSION (110-112): logitech page is very slow to load (due to style sheet updates?)
  • khtml/xml/dom_nodeimpl.cpp: (NodeImpl::dispatchGenericEvent): Don't do default handling for events that don't bubble. Not 100% sure why this caused the big slowdown but it was a somewhat arbitrary change in my earlier patch.
6:01 PM Changeset in webkit [5353] by darin
  • 2 edits in trunk/WebCore

Reviewed by John.

  • fixed 3468870 -- REGRESSION (108-109): cannot upload files (<input type=file> is broken)
  • kwq/KWQFileButton.mm: (KWQFileButton::filenameChanged): Pass the filename to the textChanged signal. The one without a parameter is a different signal.
3:46 PM Changeset in webkit [5352] by darin
  • 2 edits in trunk/WebCore

Reviewed by John.

  • fixed 3466714 -- page with <object> tag with text subtype and no URL causes hang
  • khtml/khtml_part.cpp: (KHTMLPart::requestObject): Added a check so that an empty URL string doesn't get completed, and results in an empty KURL object. Also made the change to allow empty URLs unconditional, because I don't think it's a WebCore-specific thing on further reflection.
10:49 AM Changeset in webkit [5351] by darin
  • 1 edit in trunk/WebCore/ChangeLog-2003-10-25

Fixed typo-type problems in some old log entries.

Oct 31, 2003:

6:23 PM Changeset in webkit [5350] by hyatt
  • 2 edits in trunk/WebCore

Fix for 3470489, I just forgot to commit a piece of my patch. This has actually even
been reviewed by ken already.

Reviewed by kocienda

  • khtml/rendering/render_replaced.cpp: (RenderWidget::eventFilter):
5:26 PM Changeset in webkit [5349] by darin
  • 2 edits in trunk/WebCore

Reviewed by John.

  • fixed 3457875 -- text disappears from fields in forms upon certain types of scrolling
  • kwq/KWQTextField.mm: (-[KWQTextField control:textShouldBeginEditing:]): Call releaseGState on the field editor and its clip view to prevent undesirable caching.
5:01 PM Changeset in webkit [5348] by kocienda
  • 2 edits in trunk/WebCore

Reviewed by Darin

Fix for this bug:

<rdar://problem/3440719>: Read-only text field form controls
do not keyboard focus and do not take part in tabbing

  • kwq/KWQTextField.mm: (-[KWQTextField acceptsFirstResponder]): We want our text fields to accept first responder even if not editable.

Once this is fixed, this one comes along for free:

<rdar://problem/3440710>: Read-only textarea form controls should select all text when focused

4:31 PM Changeset in webkit [5347] by darin
  • 2 edits in trunk/WebCore

Reviewed by Ken.

  • fixed 3469383 -- REGRESSION (100-111): if one line is selected on this page, too much gets copied (plain text)
  • khtml/khtml_part.cpp: (KHTMLPart::text): Range check the child node indices before using them to get at a child node. We don't want to set startNode or endNode to nil in any case. If the end node is set to nil, we end up copying the entire remainder of the page.
3:40 PM Changeset in webkit [5346]
  • 1 copy
    110 deletes in branches/unlabeled-1.16.4

This commit was manufactured by cvs2svn to create branch
'unlabeled-1.16.4'.

3:40 PM Changeset in webkit [5345] by hyatt
  • 13 edits in trunk/WebCore

Fix for 3470007, links don't get focus on mouse down. Fix the focus check on mouse down to actually
crawl up the content tree in order to find the nearest enclosing focusable node. Also clean up and
fix checks that improperly blurred the link after it got focused.

Fix for 3450335, AppKit widgets not obeying their desired focus policies. I renamed isSelectable to
isFocusable and then also added isMouseFocusable and isKeyboardFocusable methods.

Reviewed by kocienda

  • khtml/css/html4.css:
  • khtml/html/html_baseimpl.cpp: (HTMLFrameElementImpl::isFocusable):
  • khtml/html/html_baseimpl.h:
  • khtml/html/html_formimpl.cpp: (HTMLGenericFormElementImpl::isFocusable): (HTMLGenericFormElementImpl::isKeyboardFocusable): (HTMLGenericFormElementImpl::isMouseFocusable):
  • khtml/html/html_formimpl.h:
  • khtml/html/html_inlineimpl.cpp: (HTMLAnchorElementImpl::isFocusable):
  • khtml/html/html_inlineimpl.h:
  • khtml/khtmlview.cpp: (KHTMLView::dispatchMouseEvent):
  • khtml/xml/dom_docimpl.cpp: (DocumentImpl::nextFocusNode): (DocumentImpl::previousFocusNode):
  • khtml/xml/dom_nodeimpl.cpp: (NodeImpl::isFocusable): (NodeImpl::isKeyboardFocusable): (NodeImpl::isMouseFocusable):
  • khtml/xml/dom_nodeimpl.h:
  • kwq/KWQKHTMLPart.mm: (KWQKHTMLPart::khtmlMousePressEvent):
3:05 PM Changeset in webkit [5344] by kocienda
  • 2 edits in trunk/WebCore

Reviewed by David

Fix for this bug:

<rdar://problem/3470233>: solid outlines drawing incorrectly

The problem was that when I collected the array of line boxes for
render inlines, I put a QRect() at the start and the end of the list.
However, the border calculation code expected "invalid" rectangles
(rects with 0 for all four dimentions) in the first and last
positions, and the default constructor for QRect makes rects with 1,1
for width/height.

  • khtml/rendering/render_inline.cpp: (RenderInline::paintOutlines)
12:14 PM Changeset in webkit [5343] by darin
  • 2 edits in trunk/WebCore

Reviewed by Dave.

  • fixed 3464759 -- REGRESSION (100-107): Pressing on link loses mouse-down feedback almost instantly
  • khtml/khtmlview.cpp: (KHTMLView::viewportMouseMoveEvent): Pass true for the "read-only" flag in prepareMoveEvent to prevent us from updating :hover and :active while the mouse is down.
11:54 AM Changeset in webkit [5342] by hyatt
  • 4 edits in trunk/WebKit

Fix for 3466542, add a real minimum font size setting.

Reviewed by john

  • WebView.subproj/WebPreferences.h:
  • WebView.subproj/WebPreferences.m: (+[WebPreferences initialize]): (-[WebPreferences setMinimumFontSize:]): (-[WebPreferences minimumLogicalFontSize]): (-[WebPreferences setMinimumLogicalFontSize:]):
  • WebView.subproj/WebView.m: (-[WebView _updateWebCoreSettingsFromPreferences:]):
11:46 AM Changeset in webkit [5341] by kocienda
  • 3 edits in trunk/WebCore

Reviewed by David

<rdar://problem/3468910>: REGRESSION: other than focus rings, outlines styles are broken

Outlines now draw like they used to.

  • khtml/rendering/render_inline.cpp: (RenderInline::paintObject): Now branches to call focus ring or "regular" outlines. (RenderInline::paintFocusRing): Broke out drawing aqua focus rings into its own function. (RenderInline::paintOutlines): New function to draw "regular outlines. (RenderInline::paintOutline): Removed code which special-cased aqua outlines.
  • khtml/rendering/render_inline.h: Added paintOutlines declaration.
11:29 AM Changeset in webkit [5340] by hyatt
  • 9 edits in trunk/WebCore

Fix for 3467419, highlighted text in "search in books" Amazon feature is obscured. The
site is using the CSS3 opacity property. The CSS3 Color module is far enough along now
that opacity can safely be used without the "-khtml-" in front of it.

Fix for 3466542, the minimum font size bug. I am introducing a new minimum font size pref that
really is a hard override.

Reviewed by gramps (opacity) and john (min font size)

  • khtml/css/cssparser.cpp: (CSSParser::parseValue):
  • khtml/css/cssproperties.c: (hash_prop): (findProp):
  • khtml/css/cssproperties.h:
  • khtml/css/cssproperties.in:
  • khtml/css/cssstyleselector.cpp: (khtml::CSSStyleSelector::applyRule):
8:46 AM Changeset in webkit [5339] by kocienda
  • 5 edits in trunk

WebCore:

Reviewed by Darin

Fix for this bug:

<rdar://problem/3469088>: focus not removed from text link
when user hits cmd-L or clicks in window chrome

  • kwq/WebCoreBridge.h:
  • kwq/WebCoreBridge.mm: (-[WebCoreBridge deselectAll]): Now clears the focus ring as well. (-[WebCoreBridge deselectText]): Just clears text selection.

WebKit:

Reviewed by Darin

Fix for this bug:

<rdar://problem/3469088>: focus not removed from text link
when user hits cmd-L or clicks in window chrome

  • WebView.subproj/WebHTMLView.m: (-[WebHTMLView deselectText]): Added new method just to deselect text. (-[WebHTMLView resignFirstResponder]): Just deseclect text if we are doing a programmatic setting of focus. Deselect all otherwise.
1:29 AM Changeset in webkit [5338] by hyatt
  • 5 edits in trunk/WebCore

Three simple fixes:

(1) Fix for 3463777, crash on CSS3 box model draft. This was just a situation where a simple null-check
was needed for generated content.

(2) Added a != check when the old and new hover obj are the same to avoid a little bit of extra work.

(3) Removed the m_pressed member from DOM nodes and removed the calls to setPressed, since this code
is dead and not used by anyone.

  • khtml/khtmlview.cpp:
  • khtml/rendering/render_layer.cpp: (RenderLayer::addChild): (RenderLayer::updateHoverActiveState):
  • khtml/xml/dom_nodeimpl.cpp: (NodeImpl::NodeImpl):
  • khtml/xml/dom_nodeimpl.h: (DOM::NodeImpl::hasStyle): (DOM::NodeImpl::setHasStyle):

Oct 30, 2003:

5:51 PM Changeset in webkit [5337] by hyatt
  • 3 edits in trunk/WebCore

Fix for 3469330, the Marquee data in RenderStyle needs a copy constructor hack like all the other
objects do.

  • khtml/rendering/render_style.cpp: (direction):
  • khtml/rendering/render_style.h:
5:41 PM Changeset in webkit [5336] by hyatt
  • 2 edits in trunk/WebCore

Fix for 3469206, images didn't paint outlines any more. They were bailing early because I forgot to
add a check for my new PaintAction.

Reviewed by darin

  • khtml/rendering/render_replaced.cpp: (RenderReplaced::paint):
5:27 PM Changeset in webkit [5335] by hyatt
  • 6 edits in trunk/WebCore

Fixes for 3469057, outlines not drawn on aintitcool.com and also for 3469178, objects with width/height of 0
incorrectly excluded from tabbing.

Reviewed by darin

  • khtml/html/html_formimpl.cpp: (HTMLGenericFormElementImpl::isSelectable):
  • khtml/html/html_inlineimpl.cpp: (HTMLAnchorElementImpl::isSelectable):
  • khtml/rendering/render_flow.cpp: (RenderFlow::getAbsoluteRepaintRect):
  • khtml/rendering/render_inline.cpp: (RenderInline::addFocusRingRects): (RenderInline::paintOutline):
  • khtml/rendering/render_object.cpp: (RenderObject::addFocusRingRects): (RenderObject::getAbsoluteRepaintRectWithOutline):
4:47 PM Changeset in webkit [5334] by mjs
  • 2 edits in trunk/WebCore

Reviewed by Darin.

  • fixed 3426081 - empty cells HTMLCollection for <tr>
  • fixed 3367598 - "length" attribute for table row "cells" always returns 0
  • khtml/dom/html_table.cpp: (HTMLTableRowElement::cells): TR_CELLS, not TABLE_ROWS.
4:11 PM Changeset in webkit [5333] by hyatt
  • 6 edits in trunk/WebCore

Fix for 3468916, outline needs its own paint action so that focus rings don't draw under other content.

Reviewed by kocienda

  • khtml/rendering/render_block.cpp: (khtml::RenderBlock::paintObject): (khtml::RenderBlock::paintFloats):
  • khtml/rendering/render_image.cpp: (RenderImage::paintObject):
  • khtml/rendering/render_inline.cpp: (RenderInline::paintObject):
  • khtml/rendering/render_layer.cpp: (RenderLayer::paintLayer):
  • khtml/rendering/render_object.h:
4:09 PM Changeset in webkit [5332] by mjs
  • 2 edits in branches/Safari-1-1-branch

Correct bug number in ChangeLogs.

3:35 PM Changeset in webkit [5331] by cblu
  • 2 edits in trunk/WebKit

Fixed: <rdar://problem/3458368>: drawing to the screen while window hidden: ​http://www.bhphotovideo.com/

Reviewed by john.

  • Plugins.subproj/WebBaseNetscapePluginView.m: (-[WebBaseNetscapePluginView saveAndSetPortStateForUpdate:]): clip out the plug-in when the window is miniaturized or hidden (-[WebBaseNetscapePluginView restartNullEvents]): don't restart null events if the window is miniaturized, this allows restartNullEvents to be called in start and viewDidMoveToWindow without needing to make the check (-[WebBaseNetscapePluginView start]): just call restartNullEvents instead of checking if the window is miniaturized
3:31 PM Changeset in webkit [5330] by mjs
  • 2 edits in trunk/WebCore

Fixed Dave's fix to my fix to my fix:

  • khtml/html/html_imageimpl.cpp: (HTMLImageElementImpl::width): Put back m_render check, consider the case of image with no width/height attributes and display:none set. We don't want to crash on that. (HTMLImageElementImpl::height): Ditto.
3:28 PM Changeset in webkit [5329] by hyatt
  • 1 edit in trunk/WebCore/khtml/html/html_imageimpl.cpp

Fix build bustage.

3:13 PM Changeset in webkit [5328]
  • 1 copy
    93 deletes in branches/unlabeled-1.41.4

This commit was manufactured by cvs2svn to create branch
'unlabeled-1.41.4'.

3:13 PM Changeset in webkit [5327] by kocienda
  • 38 edits
    4 adds in trunk

WebCore:

Reviewed by Hyatt

Support for tabbing to links.

Fixes these bugs

<rdar://problem/3468395>: fix full keyboard access loop for buttons
<rdar://problem/3468397>: paint aqua focus rings around text links
<rdar://problem/3468406>: focused links do not activate by hitting

return key

<rdar://problem/3468410>: paintTextOutline in RenderText objects

should be done at a higher level

<rdar://problem/3468424>: improve function to determine if a node is

tab selectable

<rdar://problem/3468427>: add bridge function to access active

document view for a node

<rdar://problem/3468429>: links do not scroll into view when focusing
<rdar://problem/3468435>: when tabbing out of a frame, last focused

node is not cleared if it is a text link

<rdar://problem/3468436>: focus on text link does not clear when

clicking mouse in document body

<rdar://problem/3468440>: add setRect convenience to QRect
<rdar://problem/3468444>: elements paint their outlines even if they

are not visible

<rdar://problem/3468448>: setting focus on a node sets focus on all

its siblings

<rdar://problem/3468456>: cannot tab to form image input elements
<rdar://problem/3468461>: tab focus can seem to disappear from view on

pages with dhtml menus

<rdar://problem/3469021>: webcore bridge nextKeyView always starts
from the beginning of the document

  • WebCore-combined.exp: Export WebCoreGraphicsBridge class
  • WebCore.exp: Ditto
  • WebCore.pbproj/project.pbxproj: Add WebCoreGraphicsBridge class
  • khtml/css/cssparser.cpp: (CSSParser::parseValue): border style check now take apple aqua style into account
  • khtml/css/cssvalues.in: Add -apple-aqua border value
  • khtml/css/html4.css: Change focus, a:link:active, and a:visited:active to use new -apple-aqua focus rings.
  • khtml/html/html_formimpl.cpp: (HTMLGenericFormElementImpl::isSelectable): Improve function so that invisible elements are no longer considered selectable
  • khtml/html/html_inlineimpl.cpp: (HTMLAnchorElementImpl::isSelectable): Ditto. (HTMLAnchorElementImpl::defaultEventHandler): Make return key activate links as well as Enter.
  • khtml/html/html_inlineimpl.h: Move isSelectable function into implementation file.
  • khtml/rendering/render_block.cpp: (khtml::RenderBlock::paintObject): Do not paint outlines if content is invisible.
  • khtml/rendering/render_flow.cpp: (RenderFlow::getAbsoluteRepaintRect): Improve function to correctly calculate repaint rects including outlines.
  • khtml/rendering/render_image.cpp: (RenderImage::paintObject): Do not paint outlines if content is invisible.
  • khtml/rendering/render_inline.cpp: (RenderInline::paintObject): Ditto. (RenderInline::addFocusRingRects): New function to gather up rects to use for painting outlines. (RenderInline::paintOutline): Added code to paint aqua focus rings.
  • khtml/rendering/render_inline.h:
  • khtml/rendering/render_object.cpp: (RenderObject::drawBorder): Add case for apple aqua border style. (RenderObject::addFocusRingRects): New function to gather up rects to use for painting outlines. (RenderObject::paintOutline): Added code to paint aqua focus rings. (RenderObject::getAbsoluteRepaintRectWithOutline): Helper function to calculate rectangle to use for repainting when an object has an outline.
  • khtml/rendering/render_object.h: Add addFocusRingRects declaration.
  • khtml/rendering/render_style.h: Add APPLEAQUA to border enum
  • khtml/rendering/render_text.cpp: (RenderText::paintObject): Removed code to draw outlines.
  • khtml/rendering/render_text.h: Ditto.
  • khtml/xml/dom_elementimpl.cpp: Removed redundant isSelectable check.
  • khtml/xml/dom_elementimpl.h: Ditto.
  • khtml/xml/dom_nodeimpl.cpp: (NodeImpl::isSelectable): Added. (NodeBaseImpl::setFocus): Do not set focus on siblings.
  • khtml/xml/dom_nodeimpl.h: Moved implementation to .cpp file.
  • kwq/KWQButton.mm: (-[KWQButton nextKeyView]): Send a blur before going to next key view. This prevents focus from being lost altogether. (-[KWQButton previousKeyView]): Ditto.
  • kwq/KWQKHTMLPart.h: Added documentViewForNode function.
  • kwq/KWQKHTMLPart.mm: (KWQKHTMLPart::nextKeyViewInFrame): Add check for other focusables, like text links. (KWQKHTMLPart::nextKeyViewInFrameHierarchy): Remove focus when leaving a view. (KWQKHTMLPart::documentViewForNode): Added. (KWQKHTMLPart::khtmlMousePressEvent): Remove focus when you click in a non-focusable area, like the document body.
  • kwq/KWQPainter.h: Added declarations.
  • kwq/KWQPainter.mm: (QPainterPrivate::QPainterPrivate): Modified constructor for new fields. (QPainter::initFocusRing): Added. (QPainter::addFocusRingRect): Added. (QPainter::drawFocusRing): Added. (QPainter::clearFocusRing): Added.
  • kwq/KWQRect.h: (QRect::setRect): Added.
  • kwq/WebCoreBridge.h:
  • kwq/WebCoreBridge.mm: (-[WebCoreBridge nextKeyView]): Start from the currently-focused node, if there is one, not unconditionally from the start of the document. (-[WebCoreBridge previousKeyView]): Ditto
  • kwq/WebCoreGraphicsBridge.h: Added.
  • kwq/WebCoreGraphicsBridge.m: Added. (+[WebCoreGraphicsBridge sharedBridge]): Added. (-[WebCoreGraphicsBridge init]): Added. (-[WebCoreGraphicsBridge setFocusRingStyle:radius:color:]): Added.

WebKit:

Reviewed by Hyatt

  • WebCoreSupport.subproj/WebBridge.m: (-[WebBridge documentView]): Added.
  • WebCoreSupport.subproj/WebGraphicsBridge.h: Added.
  • WebCoreSupport.subproj/WebGraphicsBridge.m: Added. (+[WebGraphicsBridge createSharedBridge]): Added. (-[WebGraphicsBridge setFocusRingStyle:radius:color:]): Added.
  • WebKit.pbproj/project.pbxproj:
  • WebView.subproj/WebFrameView.m: Create a WebGraphicsBridge when creating a WebFrameView.
2:51 PM Changeset in webkit [5326] by mjs
  • 2 edits in trunk/WebCore

Fix mistake in my last change that I did not mean to commit.

  • khtml/html/html_imageimpl.cpp: (HTMLImageElementImpl::width): Remove extraneous second attempt to lay out.
2:50 PM Changeset in webkit [5325] by hyatt
  • 4 edits in trunk/WebCore

Put paged-media changes back without the wrongful removal of the style data copy constructors (which
were necessary after all).

  • ChangeLog:
  • khtml/css/cssstyleselector.cpp: (khtml::CSSStyleSelector::applyRule):
  • khtml/rendering/render_style.cpp: (StyleInheritedData::StyleInheritedData): (StyleInheritedData::operator==):
  • khtml/rendering/render_style.h: (khtml::): (khtml::RenderStyle::NonInheritedFlags::operator==): (khtml::RenderStyle::setBitDefaults): (khtml::RenderStyle::widows): (khtml::RenderStyle::orphans): (khtml::RenderStyle::pageBreakInside): (khtml::RenderStyle::pageBreakBefore): (khtml::RenderStyle::pageBreakAfter): (khtml::RenderStyle::setWidows): (khtml::RenderStyle::setOrphans): (khtml::RenderStyle::setPageBreakInside): (khtml::RenderStyle::setPageBreakBefore): (khtml::RenderStyle::setPageBreakAfter):
2:42 PM Changeset in webkit [5324] by mjs
  • 3 edits in trunk/WebCore

Reviewed by Dave.

  • fixed 3468129 - REGRESSION: FOUC occurs on Surfin' Safari
  • khtml/ecma/kjs_html.cpp: (KJS::HTMLElement::getValueProperty): Don't force layout for image width/height if you can determine it statically from the attribute.
2:40 PM Changeset in webkit [5323] by mjs
  • 5 edits in branches/Safari-1-1-branch

WebCore:

Merged fix for the following bug from HEAD:

  • fixed 3183754 - returning false from key press handlers does not prevent typing or form submission
  • kwq/KWQKHTMLPart.mm: (KWQKHTMLPart::keyEvent):

WebKit:

Merged fix for the following bug from HEAD:

  • fixed 3183754 - returning false from key press handlers does not prevent typing or form submission
  • WebView.subproj/WebHTMLView.m: (-[WebHTMLView keyDown:]): (-[WebHTMLView keyUp:]):
2:26 PM Changeset in webkit [5322] by vicki
  • 3 edits in trunk

fixed version number. tree is open.

2:23 PM Changeset in webkit [5321]
  • 3 copies in tags/Safari-112

This commit was manufactured by cvs2svn to create tag 'Safari-112'.

2:23 PM Changeset in webkit [5320] by vicki
  • 3 edits in trunk

forgot to tag before setting the version number back (doh!)

2:22 PM Changeset in webkit [5319] by vicki
  • 3 edits in trunk

tree is open for Safari-113

2:20 PM Changeset in webkit [5318] by vicki
  • 6 edits in trunk

Safari-112 stamp

2:00 PM Changeset in webkit [5317] by kocienda
  • 4 edits in trunk/WebCore

Reviewed by me

Rolling out this code since it caused a big performance
regression.

2003-10-29 David Hyatt <​hyatt@apple.com>

Add the paged media properties to RenderStyle.
They aren't used yet, but they should now be
parsed and interpreted correctly.

  • khtml/css/cssstyleselector.cpp: (khtml::CSSStyleSelector::applyRule):
  • khtml/rendering/render_style.cpp: (StyleSurroundData::StyleSurroundData): (StyleBoxData::StyleBoxData): (StyleBoxData::operator==): (StyleVisualData::~StyleVisualData): (StyleVisualData::StyleVisualData): (StyleBackgroundData::StyleBackgroundData): (StyleFlexibleBoxData::StyleFlexibleBoxData): (opacity): (textShadow): (StyleInheritedData::StyleInheritedData): (StyleInheritedData::~StyleInheritedData): (StyleInheritedData::operator==):
  • khtml/rendering/render_style.h: (khtml::StyleBackgroundData::~StyleBackgroundData): (khtml::StyleFlexibleBoxData::~StyleFlexibleBoxData): (khtml::RenderStyle::NonInheritedFlags::operator==): (khtml::RenderStyle::setBitDefaults):
12:59 PM Changeset in webkit [5316] by mjs
  • 1 edit in trunk/JavaScriptCore/ChangeLog

Corrected ChangeLog.

12:59 PM Changeset in webkit [5315] by mjs
  • 2 edits in trunk/JavaScriptCore

Reviewed by NOBODY (OOPS!).

  • fixed 3427069 - browsing mp3.com causes leaks (KJS)
  • kjs/string_object.cpp: (StringProtoFuncImp::call): Don't do an early return, since that could leak a temporary regexp.
10:42 AM Changeset in webkit [5314] by mjs
  • 6 edits in trunk/JavaScriptCore

Reviewed by Darin.

  • kjs/grammar.cpp: (yyerror): Updated the commented code.
  • kjs/grammar.y: Don't delete string and identifier tokens when done with them any more, they'll get cleaned up by the lexer now.
  • kjs/internal.cpp: (Parser::parse): Tell lexer when done parsing.
  • kjs/lexer.cpp: (Lexer::Lexer): Initialize new data members. (Lexer::lex): Use new methods to make strings and identifiers, and save them. (Lexer::makeIdentifier): Make a new Identifier and save it in an auto-growing array. (Lexer::makeUString): Likewise for UStrings. (Lexer::doneParsing): Clean up arrays of Ifentifiers and UStrings.
  • kjs/lexer.h:
12:26 AM Changeset in webkit [5313] by hyatt
  • 6 edits in trunk/WebCore

Fix for 3466628, widgets in overflow blocks don't move when the overflow block is scrolled. This
was a regression caused by my change to move widgets only during layout and not at paint time. The
problem with this is that scrolling an overflow block doesn't do a layout, just a repaint. I patched
the layer code to update widget positions on an overflow block scroll.

This then exposed a bug in absolutePosition, namely that the scroll offset was never factored in when
computing absolutePosition. This bug also explains why text selection and cursor display were wrong
inside scrolled overflow blocks.

Reviewed by mjs

  • khtml/rendering/render_box.cpp: (RenderBox::absolutePosition):
  • khtml/rendering/render_layer.cpp: (RenderLayer::scrollToOffset):
  • khtml/rendering/render_object.cpp: (RenderObject::absolutePosition):
  • khtml/rendering/render_text.cpp: (RenderText::cursorPos): (RenderText::posOfChar):
  • khtml/rendering/render_text.h:
Note: See TracTimeline for information about the timeline view.