1 2004-11-22 Maciej Stachowiak <mjs@apple.com>
5 <rdar://problem/3492044> performing JavaScript operations on form elements is slower than WinIE (HTMLFormCollection)
6 <rdar://problem/3489679> selecting an item on the Apache bugzilla query page is very slow (HTMLFormCollection)
7 <rdar://problem/3477810> checking 80 check boxes with JavaScript is 10x slower than in IE (HTMLFormCollection)
8 <rdar://problem/3760962> JavaScript that toggles checkboxes is slow (HTMLCollection,HTMLFormCollection)
10 * khtml/ecma/kjs_html.cpp:
11 (KJS::HTMLDocument::tryGet):
12 * khtml/html/html_formimpl.cpp:
13 (DOM::HTMLFormElementImpl::HTMLFormElementImpl):
14 (DOM::HTMLFormElementImpl::~HTMLFormElementImpl):
15 * khtml/html/html_formimpl.h:
16 * khtml/html/html_miscimpl.cpp:
17 (HTMLCollectionImpl::HTMLCollectionImpl):
18 (HTMLCollectionImpl::~HTMLCollectionImpl):
19 (HTMLCollectionImpl::CollectionInfo::CollectionInfo):
20 (HTMLCollectionImpl::CollectionInfo::reset):
21 (HTMLCollectionImpl::resetCollectionInfo):
22 (HTMLCollectionImpl::checkForNameMatch):
24 (HTMLCollectionImpl::updateNameCache):
25 (HTMLCollectionImpl::namedItems):
26 (HTMLFormCollectionImpl::HTMLFormCollectionImpl):
27 (HTMLFormCollectionImpl::~HTMLFormCollectionImpl):
28 (HTMLFormCollectionImpl::item):
29 (HTMLFormCollectionImpl::updateNameCache):
30 * khtml/html/html_miscimpl.h:
32 2004-11-22 David Hyatt <hyatt@apple.com>
34 Improve the WebCore cache so that the maximum cacheable object size is scaled based off the total cache
39 * khtml/misc/loader.cpp:
40 (CachedObject::finish):
43 * khtml/misc/loader.h:
44 (khtml::Cache::maxCacheableObjectSize):
46 2004-11-22 David Hyatt <hyatt@apple.com>
48 Fix for 3673381, huge directory listing so slow it seems like a hang. Rework painting and hit testing so that
49 it crawls the line box tree instead of the render tree. This allows more precise intersection/containment testing
50 that lets us short circuit earlier when painting and hit testing.
54 * khtml/khtml_part.cpp:
55 (KHTMLPart::isPointInsideSelection):
56 * khtml/rendering/render_block.cpp:
57 (khtml::RenderBlock::paint):
58 (khtml::RenderBlock::paintChildren):
59 (khtml::RenderBlock::paintObject):
60 (khtml::RenderBlock::paintFloats):
61 (khtml::RenderBlock::nodeAtPoint):
62 * khtml/rendering/render_block.h:
63 * khtml/rendering/render_box.cpp:
64 (RenderBox::nodeAtPoint):
65 * khtml/rendering/render_box.h:
66 * khtml/rendering/render_br.h:
67 * khtml/rendering/render_canvas.cpp:
68 (RenderCanvas::paint):
69 * khtml/rendering/render_flow.cpp:
70 (RenderFlow::paintLines):
71 (RenderFlow::hitTestLines):
72 (RenderFlow::caretRect):
73 (RenderFlow::addFocusRingRects):
74 (RenderFlow::paintFocusRing):
75 (RenderFlow::paintOutlines):
76 (RenderFlow::paintOutlineForLine):
77 * khtml/rendering/render_flow.h:
78 * khtml/rendering/render_frames.cpp:
79 (RenderFrameSet::nodeAtPoint):
80 * khtml/rendering/render_frames.h:
81 * khtml/rendering/render_image.cpp:
82 (RenderImage::nodeAtPoint):
83 * khtml/rendering/render_image.h:
84 * khtml/rendering/render_inline.cpp:
85 (RenderInline::paint):
86 (RenderInline::nodeAtPoint):
87 * khtml/rendering/render_inline.h:
88 * khtml/rendering/render_layer.cpp:
89 (RenderLayer::paintLayer):
90 (RenderLayer::hitTest):
91 (RenderLayer::hitTestLayer):
92 * khtml/rendering/render_layer.h:
93 * khtml/rendering/render_line.cpp:
94 (khtml::InlineBox::paint):
95 (khtml::InlineBox::nodeAtPoint):
96 (khtml::InlineFlowBox::flowObject):
97 (khtml::InlineFlowBox::nodeAtPoint):
98 (khtml::InlineFlowBox::paint):
99 (khtml::InlineFlowBox::paintBackgrounds):
100 (khtml::InlineFlowBox::paintBackground):
101 (khtml::InlineFlowBox::paintBackgroundAndBorder):
102 (khtml::InlineFlowBox::paintDecorations):
103 (khtml::EllipsisBox::paint):
104 (khtml::EllipsisBox::nodeAtPoint):
105 (khtml::RootInlineBox::paintEllipsisBox):
106 (khtml::RootInlineBox::paint):
107 (khtml::RootInlineBox::nodeAtPoint):
108 * khtml/rendering/render_line.h:
109 (khtml::InlineRunBox::paintBackgroundAndBorder):
110 * khtml/rendering/render_object.cpp:
111 (RenderObject::hitTest):
112 (RenderObject::setInnerNode):
113 (RenderObject::nodeAtPoint):
114 * khtml/rendering/render_object.h:
115 (khtml::RenderObject::PaintInfo::PaintInfo):
116 (khtml::RenderObject::PaintInfo::~PaintInfo):
117 (khtml::RenderObject::paintingRootForChildren):
118 (khtml::RenderObject::shouldPaintWithinRoot):
119 * khtml/rendering/render_table.cpp:
120 (RenderTable::layout):
121 (RenderTable::paint):
122 * khtml/rendering/render_text.cpp:
123 (simpleDifferenceBetweenColors):
124 (correctedTextColor):
125 (InlineTextBox::nodeAtPoint):
126 (InlineTextBox::paint):
127 (InlineTextBox::selectionStartEnd):
128 (InlineTextBox::paintSelection):
129 (InlineTextBox::paintMarkedTextBackground):
130 (InlineTextBox::paintDecoration):
131 (RenderText::posOfChar):
132 * khtml/rendering/render_text.h:
133 (khtml::RenderText::paint):
134 (khtml::RenderText::layout):
135 (khtml::RenderText::nodeAtPoint):
136 * khtml/xml/dom2_eventsimpl.cpp:
137 (MouseEventImpl::computeLayerPos):
138 * khtml/xml/dom_docimpl.cpp:
139 (DocumentImpl::prepareMouseEvent):
140 * kwq/KWQAccObject.mm:
141 (-[KWQAccObject accessibilityHitTest:]):
142 * kwq/KWQKHTMLPart.mm:
143 (KWQKHTMLPart::scrollOverflowWithScrollWheelEvent):
144 (KWQKHTMLPart::eventMayStartDrag):
145 (KWQKHTMLPart::khtmlMouseMoveEvent):
146 * kwq/WebCoreBridge.mm:
147 (-[WebCoreBridge elementAtPoint:]):
148 (-[WebCoreBridge _positionForPoint:]):
150 2004-11-22 Maciej Stachowiak <mjs@apple.com>
154 <rdar://problem/3890961> selecting an item on the Apache bugzilla query page can be sped up 10% (HTMLFormCollection)
155 <rdar://problem/3890958> JavaScript that toggles checkboxes can be improved 73% (HTMLCollection,HTMLFormCollection)
157 This avoids the O(N^2) penalty for named item traversal for form collections.
159 It also combines the item traversal logic for all non-form
160 collection operations into a single traverseNextItem
161 function. This avoids having 5 copies of the big switch statement
164 Also fixed a bug that prevented the last form element from being removed properly.
166 * khtml/html/html_formimpl.cpp:
167 (DOM::removeFromVector):
168 * khtml/dom/html_misc.cpp:
169 (HTMLCollection::namedItems):
170 * khtml/dom/html_misc.h:
171 * khtml/ecma/kjs_html.cpp:
172 (KJS::HTMLCollection::getNamedItems):
173 * khtml/html/html_miscimpl.cpp:
174 (HTMLCollectionImpl::traverseNextItem):
175 (HTMLCollectionImpl::calcLength):
176 (HTMLCollectionImpl::length):
177 (HTMLCollectionImpl::item):
178 (HTMLCollectionImpl::nextItem):
179 (HTMLCollectionImpl::checkForNameMatch):
180 (HTMLCollectionImpl::namedItem):
181 (HTMLCollectionImpl::namedItems):
182 (HTMLCollectionImpl::nextNamedItem):
183 (HTMLFormCollectionImpl::calcLength):
184 (HTMLFormCollectionImpl::namedItem):
185 (HTMLFormCollectionImpl::nextNamedItem):
186 (HTMLFormCollectionImpl::namedItems):
187 * khtml/html/html_miscimpl.h:
189 2004-11-22 Ken Kocienda <kocienda@apple.com>
193 Change around the way we block the Javascript "Paste" command identifier from
194 being available. Formerly, this was done with an ifdef we never compiled in.
195 Now, this is done with a couple of cheap runtime checks. The advantage is that
196 we can now compile this command into development builds, and still yet switch
197 on the command in deployment builds through the use of WebCore SPI so we can
198 write and run layout tests with all of our builds.
200 * khtml/editing/jsediting.cpp:
201 (DOM::JSEditor::queryCommandSupported): Checks state of paste command in case
202 command being queried is the paste command.
203 (DOM::JSEditor::setSupportsPasteCommand): New SPI to turn on paste command.
204 * khtml/editing/jsediting.h: Ditto.
205 * khtml/khtml_part.cpp:
206 (KHTMLPart::pasteFromPasteboard): Added.
207 (KHTMLPart::canPaste): Added.
208 * kwq/KWQKHTMLPart.mm:
209 (KHTMLPart::canPaste): Added.
210 * kwq/KWQRenderTreeDebug.cpp:
211 (externalRepresentation): Turn on paste command.
212 * kwq/WebCoreBridge.h: Add canPaste call so WebKit can fill in the answer.
214 2004-11-21 Maciej Stachowiak <mjs@apple.com>
218 <rdar://problem/3889655> HTMLCollectionImpl should use traverseNextNode to improve speed and save recursion
220 * khtml/html/html_miscimpl.cpp:
221 (HTMLCollectionImpl::calcLength):
222 (HTMLCollectionImpl::getItem):
223 (HTMLCollectionImpl::item):
224 (HTMLCollectionImpl::nextItem):
225 (HTMLCollectionImpl::getNamedItem):
226 (HTMLCollectionImpl::namedItem):
227 (HTMLCollectionImpl::nextNamedItemInternal):
228 (HTMLFormCollectionImpl::nextNamedItemInternal):
230 2004-11-19 Maciej Stachowiak <mjs@apple.com>
234 <rdar://problem/3482935> JavaScript so slow it seems like a hang (hrweb.apple.com) (HTMLCollection?)
235 <rdar://problem/3759149> PeopleSoft page in Safari twice as slow as Mozilla engine (HTMLFormCollection)
236 <rdar://problem/3888368> selecting an item on the Apache bugzilla query page can be improved 95% (HTMLFormCollection)
238 Many optimizations to HTMLFormCollection. Iterating it should not
239 be N^2 any more, though finding items by name could still be.
241 * khtml/html/html_formimpl.cpp:
242 (DOM::HTMLFormElementImpl::~HTMLFormElementImpl):
243 (DOM::HTMLFormElementImpl::length):
244 (DOM::HTMLFormElementImpl::submitClick):
245 (DOM::HTMLFormElementImpl::formData):
246 (DOM::HTMLFormElementImpl::submit):
247 (DOM::HTMLFormElementImpl::reset):
248 (DOM::HTMLFormElementImpl::radioClicked):
249 (DOM::appendToVector):
250 (DOM::removeFromVector):
251 (DOM::HTMLFormElementImpl::registerFormElement):
252 (DOM::HTMLFormElementImpl::removeFormElement):
253 (DOM::HTMLFormElementImpl::makeFormElementDormant):
254 (DOM::HTMLFormElementImpl::registerImgElement):
255 (DOM::HTMLFormElementImpl::removeImgElement):
256 * khtml/html/html_formimpl.h:
257 * khtml/html/html_miscimpl.cpp:
258 (HTMLFormCollectionImpl::FormCollectionInfo::FormCollectionInfo):
259 (void::HTMLFormCollectionImpl::FormCollectionInfo::reset):
260 (HTMLFormCollectionImpl::resetCollectionInfo):
261 (HTMLFormCollectionImpl::calcLength):
262 (HTMLFormCollectionImpl::item):
263 (HTMLFormCollectionImpl::getNamedItem):
264 (HTMLFormCollectionImpl::getNamedFormItem):
265 (HTMLFormCollectionImpl::firstItem):
266 (HTMLFormCollectionImpl::nextItem):
267 * khtml/html/html_miscimpl.h:
268 (DOM::HTMLFormCollectionImpl::~HTMLFormCollectionImpl):
269 * khtml/xml/dom_elementimpl.cpp:
270 (ElementImpl::setAttribute):
271 (ElementImpl::setAttributeMap):
272 * kwq/KWQPtrVector.h:
273 (QPtrVector::findRef):
274 * kwq/KWQVectorImpl.h:
275 * kwq/KWQVectorImpl.mm:
276 (KWQVectorImpl::findRef):
277 * kwq/WebCoreBridge.mm:
278 (-[WebCoreBridge elementWithName:inForm:]):
279 (-[WebCoreBridge controlsInForm:]):
281 2004-11-19 David Harrison <harrison@apple.com>
283 Reviewed by Ken and Darin.
285 <rdar://problem/3856215> Cannot remove bold from the beginning of a message
287 Problem is that KHTMLPart::computeAndSetTypingStyle always looked upstream
288 for the existing style, but in this case (hitting cmd-B with caret at top of
289 file) there is nothing upstream. Changed this to use the VisiblePosition
290 deepEquivalent instead.
292 * khtml/khtml_part.cpp:
293 (KHTMLPart::computeAndSetTypingStyle):
297 2004-11-19 Maciej Stachowiak <mjs@apple.com>
301 <rdar://problem/3864151> REGRESSION (125-167): Chrysler.com never stops loading
303 * khtml/xml/dom_docimpl.cpp:
304 (DocumentImpl::close): Don't fire the onload handler if there is a
305 redirect pending. This is a very long-standing bug that was masked
306 by our previously incorrect redirect logic. It used to be that an
307 older redirect would always win. Recently we changed things so
308 that a newer redirect would win, but a script that causes a
309 redirect would stop parsing once complete (so if there are two
310 redirects in the same script, the latter wins). However, we should
311 have also prevented onload in this case. Testing with other
312 browsers shows that onload handlers do not run at all when there
313 is a pending redirect.
315 2004-11-19 Ken Kocienda <kocienda@apple.com>
319 Fix some object lifetime issues in these two commands. This fixes some crashes
320 I am seeing in some new code I am working on, but have not yet reproduced otherwise.
322 * khtml/editing/htmlediting.cpp:
323 (khtml::InsertParagraphSeparatorCommand::~InsertParagraphSeparatorCommand): No longer deref nodes
324 in the ancestor list. They are not ref'ed when put on list. D'uh.
325 (khtml::InsertParagraphSeparatorCommand::doApply): Ref all cloned nodes that are created by the command
326 before putting them on the cloned nodes list. This are still deref'ed in the destructor.
327 (khtml::InsertParagraphSeparatorInQuotedContentCommand::~InsertParagraphSeparatorInQuotedContentCommand): Ditto
329 (khtml::InsertParagraphSeparatorInQuotedContentCommand::doApply): Ditto doApply comment.
331 2004-11-19 Ken Kocienda <kocienda@apple.com>
337 <rdar://problem/3655241> setTypingStyle: does not set the real typing style, and typingStyle does not return it
339 * khtml/khtml_part.cpp:
340 (KHTMLPart::computeAndSetTypingStyle): New helper that does the work of reducing a passed-in style
341 declaration given the current selection, and then sets the minimum necessary style as the typing
343 (KHTMLPart::applyStyle): Call new computeAndSetTypingStyle. The guts of computeAndSetTypingStyle used
344 to be here in the selection-as-caret case. But now [WebCoreBridge setTypingStyle:] needs this code
346 * khtml/khtml_part.h: Declare new computeAndSetTypingStyle() function.
347 * kwq/WebCoreBridge.h: Declare new typingStyle and setTypingStyle: methods.
348 * kwq/WebCoreBridge.mm:
349 (-[WebCoreBridge typingStyle]): Calls through to the part to retrieve the typing style.
350 (-[WebCoreBridge setTypingStyle:]): Calls through to the part to set the typing style.
352 2004-11-18 David Harrison <harrison@apple.com>
356 Back out Darin's fix for <rdar://problem/3885729>, because the new exception gets triggered by Mail.app.
357 Filed <rdar://problem/3886832> against Mail.app.
360 (-[DOMCSSStyleDeclaration setProperty:::]):
362 2004-11-18 Chris Blumenberg <cblu@apple.com>
364 Fixed: <rdar://problem/3587481> Bug Reporter Login Page: Password AutoFill does not work reliably
368 * kwq/KWQKHTMLPart.mm:
369 (KWQKHTMLPart::currentForm): just return the current form, don't scan the entire document looking for a form if there is no current form
371 2004-11-18 Maciej Stachowiak <mjs@apple.com>
375 - fix recursive item traversal, use traverseNextNode() instead of
376 the buggy hand-rolled traversal.
378 * khtml/xml/dom_nodeimpl.cpp:
379 (NodeListImpl::recursiveItem):
381 2004-11-17 Darin Adler <darin@apple.com>
385 - fixed <rdar://problem/3885744> crash with XMLHttpRequest test page (reported by KDE folks)
387 * khtml/ecma/xmlhttprequest.cpp: (KJS::XMLHttpRequest::slotFinished):
388 Rolled in fix from KDE; make sure to set job to 0 before calling changeState.
390 - fixed <rdar://problem/3885729> attempting to modify a computed style does nothing, but should raise a DOM exception
391 - fixed <rdar://problem/3885731> style declarations use too many malloc blocks; switch to QValueList
392 - fixed <rdar://problem/3885739> DOM::NodeImpl accessor in DOM::Node class is hot; should be inlined
393 - changed NodeImpl calls like replaceChild to always ref/deref the parameter; this is a better way to fix
394 an entire category of leaks we have been fixing one by one recently
395 - changed computed styles so they hold a reference to the DOM node; the old code could end up with a
396 stale RenderObject pointer, although I never saw it do that in practice
397 - implemented the length and item methods for computed styles
398 - implemented querying additional properties in computed styles (29 more)
400 * khtml/khtml_part.h: Update forward declaration of CSSMutableStyleDeclarationImpl since it's
401 now a separate class rather than a typedef. Changed the parameter type of setTypingStyle to
402 take a mutable style.
403 * khtml/khtml_part.cpp:
404 (KHTMLPart::setTypingStyle): Change parameter to take a mutable style.
405 (KHTMLPart::applyStyle): Add code to make a mutable style in case we are passed
406 a computed style; also change some types to mutable style.
407 (updateState): Update iteration of CSSProperty objects in a style declaration to use
408 the new valuesIterator interface.
409 (KHTMLPart::selectionHasStyle): Add a call to makeMutable.
410 (KHTMLPart::selectionStartHasStyle): Add call to makeMutable and update iteration.
411 (editingStyle): Change type to mutable style, and simplify the style-creation calls,
412 including accomodating the exception code that setCssText has now.
413 (KHTMLPart::applyEditingStyleToElement): Change types to mutable style.
414 (KHTMLPart::removeEditingStyleFromElement): Change code to call setChanged only if removing
415 the style attributes really was a change, although it's not an important optimization it's
418 * khtml/css/css_base.h: Remove unneeded setParsedValue method.
419 * khtml/css/css_base.cpp: Remove unneeded setParsedValue method. All the places that were
420 calling it were already removing the old property explicitly, so the code in here to remove
421 the property again was redundant.
423 * khtml/css/css_computedstyle.h: Updated virtual functions for changes to parameters in base class.
424 Moved all the "set"-type functions so they are private. Store a node pointer instead of a renderer.
425 * khtml/css/css_computedstyle.cpp:
426 (DOM::CSSComputedStyleDeclarationImpl::CSSComputedStyleDeclarationImpl): Hold a reference to
427 the node we compute style for, so we don't end up with a pointer to a deallocated RenderObject.
428 Before we had no guarantee the object would outlast us.
429 (DOM::CSSComputedStyleDeclarationImpl::setCssText): Add exception parameter, and set the
430 exception to NO_MODIFICATION_ALLOWED_ERR.
431 (DOM::CSSComputedStyleDeclarationImpl::getPropertyCSSValue): Update to use node pointer rather
432 than renderer pointer. Added implementation for box-align, box-direction, box-flex, box-flex-group,
433 box-lines, box-ordinal-group, box-orient, box-pack, caption-side, clear, cursor, direction,
434 list-style-image, list-style-position, list-style-type, marquee-direction, marquee-repetition,
435 marquee-style, user-modify, opacity, orphans, outline-style, page-break-after, page-break-before,
436 page-break-inside, position, unicode-bidi, widows, z-index.
437 (DOM::CSSComputedStyleDeclarationImpl::removeProperty): Add exception parameter, and set the
438 exception to NO_MODIFICATION_ALLOWED_ERR.
439 (DOM::CSSComputedStyleDeclarationImpl::setProperty): Ditto.
440 (DOM::CSSComputedStyleDeclarationImpl::length): Implemented.
441 (DOM::CSSComputedStyleDeclarationImpl::item): Implemented, calls getPropertyValue.
442 (DOM::CSSComputedStyleDeclarationImpl::copyInheritableProperties): Changed return type to
443 CSSMutableStyleDeclarationImpl.
444 (DOM::CSSComputedStyleDeclarationImpl::copy): Added.
445 (DOM::CSSComputedStyleDeclarationImpl::makeMutable): Added.
447 * khtml/css/css_ruleimpl.h: Update forward declaration of CSSMutableStyleDeclarationImpl since it's
448 now a separate class rather than a typedef.
449 * khtml/css/cssparser.h: Ditto.
451 * khtml/css/css_valueimpl.h: Refactor CSSStyleDeclarationImpl into two classes. New derived class
452 CSSMutableStyleDeclarationImpl has the guts, and the base class has only some virtual functions.
453 Removed a bunch of redundant stuff from other classes in this file too.
454 (DOM::DashboardRegionImpl::setNext): Ref new before deref'ing old to handle the set-to-same case.
455 (DOM::CSSProperty::CSSProperty): Added new overload so you can create a CSSProperty with initial values.
456 (DOM::CSSProperty::operator=): Added.
457 (DOM::CSSProperty::setValue): Use ref-before-deref pattern to simplify slightly.
459 * khtml/css/css_valueimpl.cpp:
460 (DOM::CSSStyleDeclarationImpl::CSSStyleDeclarationImpl): Remove uneeded things.
461 (DOM::CSSStyleDeclarationImpl::isStyleDeclaration): Put here now that it's no longer inline.
462 (DOM::CSSMutableStyleDeclarationImpl::CSSMutableStyleDeclarationImpl): Added.
463 (DOM::CSSMutableStyleDeclarationImpl::operator=): Added.
464 (DOM::CSSMutableStyleDeclarationImpl::~CSSMutableStyleDeclarationImpl): Updated.
465 (DOM::CSSMutableStyleDeclarationImpl::getPropertyValue): Removed now-uneeded check.
466 (DOM::CSSMutableStyleDeclarationImpl::get4Values): Moved here from base class.
467 (DOM::CSSMutableStyleDeclarationImpl::getShortHandValue): Ditto.
468 (DOM::CSSMutableStyleDeclarationImpl::getPropertyCSSValue): Update to use QValueList instead of QPtrList.
469 (DOM::CSSMutableStyleDeclarationImpl::removeProperty): Added exception parameter, updated for QValueList.
470 (DOM::CSSMutableStyleDeclarationImpl::setChanged): Moved here from base class.
471 (DOM::CSSMutableStyleDeclarationImpl::getPropertyPriority): Update to use QValueList.
472 (DOM::CSSMutableStyleDeclarationImpl::setProperty): Added more overloads to match new parameters.
473 (DOM::CSSMutableStyleDeclarationImpl::setStringProperty): Update to use QValueList.
474 (DOM::CSSMutableStyleDeclarationImpl::setImageProperty): Ditto.
475 (DOM::CSSMutableStyleDeclarationImpl::parseProperty): Remove unneeded initialization code due to QValueList.
476 (DOM::CSSMutableStyleDeclarationImpl::addParsedProperties): Added.
477 (DOM::CSSMutableStyleDeclarationImpl::setLengthProperty): Moved here from base class.
478 (DOM::CSSMutableStyleDeclarationImpl::length): Update to use QValueList.
479 (DOM::CSSMutableStyleDeclarationImpl::item): Moved here from base class.
480 (DOM::CSSMutableStyleDeclarationImpl::cssText): Return empty string rather than null string when there are
481 no styles in the list. Update to use QValueList.
482 (DOM::CSSMutableStyleDeclarationImpl::setCssText): Update to use QValueList and to take an exceptionCode
483 parameter and set it.
484 (DOM::CSSMutableStyleDeclarationImpl::merge): Update to use QValueList.
485 (DOM::CSSStyleDeclarationImpl::diff): Update to use QValueList.
486 (DOM::CSSMutableStyleDeclarationImpl::copyBlockProperties): Moved here from base class. Change return type.
487 (DOM::CSSStyleDeclarationImpl::copyPropertiesInSet): Update to use QValueList and use stack, not new/delete.
488 (DOM::CSSMutableStyleDeclarationImpl::makeMutable): Added.
489 (DOM::CSSMutableStyleDeclarationImpl::copy): Added.
491 * khtml/css/cssparser.cpp:
492 (CSSParser::parseValue): Changed to use addParsedProperties.
493 (CSSParser::parseDeclaration): Ditto.
494 (CSSParser::createStyleDeclaration): Use new constructor to create declaration in a more efficient manner.
496 * khtml/css/cssproperties.in: Removed unused font-size-adjust and -khtml-flow-mode.
497 * khtml/css/cssproperties.c: Regenerated.
498 * khtml/css/cssproperties.h: Regenerated.
500 * khtml/css/cssstyleselector.cpp: (khtml::CSSStyleSelector::applyDeclarations):
501 Updated to use QValueList interface to CSSMutableStyleDeclarationImpl.
503 * khtml/dom/css_value.cpp:
504 (DOM::CSSStyleDeclaration::cssText): Removed unneeded cast.
505 (DOM::CSSStyleDeclaration::setCssText): Added exception code handling.
506 (DOM::CSSStyleDeclaration::getPropertyValue): Changed to call getPropertyValue directly instead of
507 first doing getPropertyCSSValue and then doing cssText.
508 (DOM::CSSStyleDeclaration::getPropertyCSSValue): Removed unneeded cast.
509 (DOM::CSSStyleDeclaration::removeProperty): Added exception code handling.
510 (DOM::CSSStyleDeclaration::setProperty): Added exception code handling.
511 (DOM::CSSStyleDeclaration::length): Removed unneeded cast.
512 (DOM::CSSStyleDeclaration::item): Removed unneeded cast.
513 (DOM::CSSStyleDeclaration::parentRule): Removed unneeded cast.
514 (DOM::CSSValue::setCssText): Removed strange non-implementation (still not implemented).
516 * khtml/dom/dom_node.h: Made isNull and handle functions inline.
517 * khtml/dom/dom_node.cpp: Ditto.
519 * khtml/editing/htmlediting.h: Change some types to mutable style.
520 * khtml/editing/htmlediting.cpp:
521 (khtml::EditCommandPtr::typingStyle): Change return type to mutable style.
522 (khtml::EditCommandPtr::setTypingStyle): Change parameter to mutable style.
523 (khtml::StyleChange::init): Convert parameter to mutable style. Update to use QValueList.
524 (khtml::EditCommand::assignTypingStyle): Change parameter to mutable type.
525 (khtml::EditCommand::setTypingStyle): Ditto.
526 (khtml::ApplyStyleCommand::ApplyStyleCommand): Convert parameter to mutable style.
527 (khtml::ApplyStyleCommand::doApply): Change local variables to mutable style.
528 (khtml::ApplyStyleCommand::applyBlockStyle): Change parameter to mutable style.
529 (khtml::ApplyStyleCommand::applyInlineStyle): Ditto.
530 (khtml::ApplyStyleCommand::isHTMLStyleNode): Ditto.
531 (khtml::ApplyStyleCommand::removeCSSStyle): Ditto. Also update to use QValueList.
532 (khtml::ApplyStyleCommand::removeBlockStyle): Change parameter to mutable style.
533 (khtml::ApplyStyleCommand::removeInlineStyle): Ditto.
534 (khtml::ApplyStyleCommand::addBlockStyleIfNeeded): Ditto.
535 (khtml::ApplyStyleCommand::addInlineStyleIfNeeded): Ditto.
536 (khtml::InsertLineBreakCommand::doApply): Convert locals to mutable style.
537 (khtml::InsertTextCommand::prepareForTextInsertion): Ditto.
538 (khtml::RemoveCSSPropertyCommand::RemoveCSSPropertyCommand): Convert parameter to mutable style.
540 * khtml/editing/jsediting.cpp: Convert types to mutable styles where we create styles.
541 * khtml/html/html_baseimpl.h: Change type to mutable style.
542 * khtml/html/html_baseimpl.cpp: (HTMLBodyElementImpl::createLinkDecl): Ditto.
544 * khtml/html/html_elementimpl.h: Make CSSMappedAttributeDeclarationImpl use the mutable style
545 class as a base class, and change types to mutable style as needed.
546 * khtml/html/html_elementimpl.cpp:
547 (HTMLElementImpl::createInlineStyleDecl): Change type to mutable style.
548 (HTMLElementImpl::parseHTMLAttribute): Call parseProperty method.
549 (HTMLElementImpl::getInlineStyleDecl): Change type to mutable style.
550 (HTMLElementImpl::additionalAttributeStyleDecl): Ditto.
551 (HTMLElementImpl::createContextualFragment): Add ref/deref to fix potential node leak.
552 (HTMLElementImpl::setInnerHTML): Remove ref/deref pair because this leak is now fixed by changes
554 (HTMLElementImpl::setOuterHTML): Remove ref/deref pair because this leak is now fixed by changes
557 * khtml/html/html_tableimpl.h: Change types to mutable style.
558 * khtml/html/html_tableimpl.cpp:
559 (HTMLTableElementImpl::additionalAttributeStyleDecl): Change type to mutable style.
560 (HTMLTableElementImpl::getSharedCellDecl): Change type to mutable style.
561 (HTMLTableCellElementImpl::additionalAttributeStyleDecl): Change type to mutable style.
563 * khtml/html/htmlparser.cpp:
564 (KHTMLParser::parseToken): Use a local variable to protect the node by ref'ing it. This is better
565 than using an explicit delete to make the node go away, and is required for compatibility with the
566 changes to the NodeImpl functions.
567 (KHTMLParser::insertNode): Ditto.
568 (KHTMLParser::createHead): Get rid of explicit delete, no longer needed because of changes to
569 the NodeImpl functions.
571 * khtml/xml/dom_docimpl.cpp: (DocumentImpl::createCSSStyleDeclaration): Call simpler constructor
572 now that there's no need to make the property list explictly.
575 (-[DOMCSSStyleDeclaration setCssText:]): Raise exception when appropriate.
576 (-[DOMCSSStyleDeclaration removeProperty:]): Ditto.
577 (-[DOMCSSStyleDeclaration setProperty:::]): Dito.
579 * khtml/xml/dom_nodeimpl.cpp:
580 (NodeImpl::insertBefore): Always do a ref/deref, so callers don't have to worry about whether the
581 function succeeded or not for ownership purposes.
582 (NodeImpl::replaceChild): Ditto.
583 (NodeImpl::appendChild): Ditto.
584 (NodeBaseImpl::insertBefore): Ditto.
585 (NodeBaseImpl::replaceChild): Ditto.
586 (NodeBaseImpl::appendChild): Ditto.
587 (NodeBaseImpl::addChild): Ditto.
589 * WebCore-tests.exp: Removed CSSStyleDeclaration::length; not sure why it was in here.
590 * WebCore-combined.exp: Regenerated.
592 2004-11-18 Maciej Stachowiak <mjs@apple.com>
594 still even more build fixing
596 * khtml/html/html_miscimpl.cpp:
597 (HTMLCollectionImpl::resetCollectionInfo):
599 2004-11-18 Maciej Stachowiak <mjs@apple.com>
603 * khtml/html/html_miscimpl.cpp:
604 (HTMLCollectionImpl::resetCollectionInfo):
606 2004-11-18 Maciej Stachowiak <mjs@apple.com>
610 * khtml/html/html_miscimpl.h:
611 (DOM::HTMLCollectionImpl::CollectionInfo::CollectionInfo): it's haslength, not hasLenght.
613 2004-11-18 Maciej Stachowiak <mjs@apple.com>
617 - merged and cleaned up HTMLCollection and HTMLFormCollection speedups from konqueror
619 <rdar://problem/3822992> VIP: Program listings pages at directv.com take a really long time to load [HTMLCollection]
620 <rdar://problem/3701991> Safari unresponsive loading (www.maxim-ic.com) (HTMLCollection)
622 This is also a start on fixing 5 other bugs, but those need additional work to make
623 HTMLFormCollection fast.
625 * khtml/html/html_documentimpl.h:
626 (DOM::HTMLDocumentImpl::collectionInfo):
627 * khtml/html/html_formimpl.cpp:
628 (DOM::HTMLFormElementImpl::~HTMLFormElementImpl):
629 (DOM::HTMLFormElementImpl::isURLAttribute):
630 (DOM::HTMLFormElementImpl::registerImgElement):
631 (DOM::HTMLFormElementImpl::removeImgElement):
632 * khtml/html/html_formimpl.h:
633 * khtml/html/html_imageimpl.cpp:
634 (HTMLImageElementImpl::HTMLImageElementImpl):
635 (HTMLImageElementImpl::~HTMLImageElementImpl):
636 * khtml/html/html_imageimpl.h:
637 * khtml/html/html_miscimpl.cpp:
638 (HTMLCollectionImpl::HTMLCollectionImpl):
639 (HTMLCollectionImpl::~HTMLCollectionImpl):
640 (HTMLCollectionImpl::updateCollectionInfo):
641 (HTMLCollectionImpl::length):
642 (HTMLCollectionImpl::item):
643 (HTMLCollectionImpl::firstItem):
644 (HTMLCollectionImpl::nextItem):
645 (HTMLCollectionImpl::namedItem):
646 (HTMLCollectionImpl::nextNamedItemInternal):
647 (HTMLFormCollectionImpl::getNamedFormItem):
648 * khtml/html/html_miscimpl.h:
649 (DOM::HTMLCollectionImpl::):
650 (DOM::HTMLCollectionImpl::CollectionInfo::CollectionInfo):
651 * khtml/html/htmlparser.cpp:
652 (KHTMLParser::getElement):
653 * khtml/xml/dom_docimpl.cpp:
654 (DocumentImpl::DocumentImpl):
655 * khtml/xml/dom_docimpl.h:
656 (DOM::DocumentImpl::incDOMTreeVersion):
657 (DOM::DocumentImpl::domTreeVersion):
658 * khtml/xml/dom_nodeimpl.cpp:
662 2004-11-18 Kevin Decker <kdecker@apple.com>
666 fixed: <rdar://problem/3841842> getPropertyID expensive
669 (getPropertyID): avoid unnecessary memory allocations by using a fixed-sized stack based buffer.
671 2004-11-17 David Hyatt <hyatt@apple.com>
673 Improve responsiveness by being willing to break out of the tokenizer. (This patch was landed already
674 and subsequently backed out).
678 * khtml/html/html_baseimpl.cpp:
679 (HTMLBodyElementImpl::insertedIntoDocument):
680 * khtml/html/htmltokenizer.cpp:
681 (khtml::HTMLTokenizer::reset):
682 (khtml::HTMLTokenizer::scriptHandler):
683 (khtml::HTMLTokenizer::scriptExecution):
684 (khtml::HTMLTokenizer::write):
685 (khtml::HTMLTokenizer::continueProcessing):
686 (khtml::HTMLTokenizer::timerEvent):
687 (khtml::HTMLTokenizer::notifyFinished):
688 * khtml/html/htmltokenizer.h:
689 * khtml/khtmlview.cpp:
690 (KHTMLViewPrivate::KHTMLViewPrivate):
691 (KHTMLViewPrivate::reset):
694 (KHTMLView::timerEvent):
695 (KHTMLView::scheduleRelayout):
696 (KHTMLView::layoutPending):
697 (KHTMLView::haveDelayedLayoutScheduled):
698 (KHTMLView::unscheduleRelayout):
700 * khtml/xml/dom_docimpl.cpp:
701 (DocumentImpl::DocumentImpl):
702 (DocumentImpl::close):
703 (DocumentImpl::setParsing):
704 (DocumentImpl::shouldScheduleLayout):
705 (DocumentImpl::minimumLayoutDelay):
706 (DocumentImpl::write):
707 (DocumentImpl::finishParsing):
708 (DocumentImpl::stylesheetLoaded):
709 (DocumentImpl::updateStyleSelector):
710 * khtml/xml/dom_docimpl.h:
711 (DOM::DocumentImpl::parsing):
712 * kwq/KWQDateTime.mm:
713 (KWQUIEventTime::uiEventPending):
715 2004-11-17 David Harrison <harrison@apple.com>
717 Reviewed by Ken Kocienda.
719 Make sure previousLineStart is non-null before calling compareBoundaryPoints.
720 Treat null case as meaning no post-move merge is needed.
722 * khtml/editing/htmlediting.cpp:
723 (khtml::DeleteSelectionCommand::initializePositionData):
725 2004-11-17 David Harrison <harrison@apple.com>
727 Added displayNode and displayTree methods for debugging. Fixed comment typo in dispatchChildRemovalEvents.
728 * khtml/xml/dom_nodeimpl.cpp:
729 (NodeImpl::displayNode):
730 (NodeImpl::displayTree):
731 (NodeBaseImpl::dispatchChildRemovalEvents):
732 * khtml/xml/dom_nodeimpl.h:
734 2004-11-16 John Sullivan <sullivan@apple.com>
738 - fixed <rdar://problem/3881929> 32 byte leak in editingStyle() in KHTMLPart (one-time only)
740 * khtml/khtml_part.cpp:
742 delete the list we created when we're done with it
744 2004-11-16 Ken Kocienda <kocienda@apple.com>
748 It is unwise to use the QPtrList autodelete feature on shared objects like DOM nodes.
749 Instead, I replaced this with a helper function that derefs DOM nodes stored in a
750 QPtrList when the list goes out of scope.
752 * khtml/editing/htmlediting.cpp:
753 (khtml::derefNodesInList): New helper to deref DOM nodes stored in a QPtrList.
754 (khtml::InsertParagraphSeparatorCommand::InsertParagraphSeparatorCommand): No longer set lists to autodelete.
755 (khtml::InsertParagraphSeparatorCommand::~InsertParagraphSeparatorCommand): Call new derefNodesInList helper.
756 (khtml::InsertParagraphSeparatorInQuotedContentCommand::InsertParagraphSeparatorInQuotedContentCommand):
757 No longer set lists to autodelete.
758 (khtml::InsertParagraphSeparatorInQuotedContentCommand::~InsertParagraphSeparatorInQuotedContentCommand):
759 Call new derefNodesInList helper.
760 * khtml/editing/htmlediting.h: Add virtual destructor for InsertParagraphSeparatorCommand. It had no need
761 of one before, but now it does.
763 2004-11-15 David Harrison <harrison@apple.com>
765 Reviewed by Chris and Darin.
767 <rdar://problem/3880304> Non-linear performance hit for style changes
769 * khtml/xml/dom_nodeimpl.cpp:
770 (NodeImpl::traverseNextNode):
771 (NodeImpl::traverseNextSibling):
772 (NodeImpl::traversePreviousNodePostOrder):
773 Return 0 rather than traversing beyond stayWithin when this == stayWithin.
774 Add asserts that stayWithin is an ancestor of the returned node.
776 2004-11-15 Darin Adler <darin@apple.com>
780 - fixed <rdar://problem/3880036> Many leaks from CSSComputedStyleDeclarationImpl::getPropertyCSSValue, seen in Mail and Blot
782 * khtml/css/css_computedstyle.cpp:
783 (DOM::CSSComputedStyleDeclarationImpl::getPropertyValue): Ref and deref the value returned from getPropertyCSSValue,
784 since there's no guarantee it's already ref'd.
785 * khtml/css/css_valueimpl.cpp:
786 (CSSStyleDeclarationImpl::getPropertyValue): Wrap result in a CSSValue to ref/deref.
787 (CSSStyleDeclarationImpl::get4Values): Ref/deref explicitly.
788 (CSSStyleDeclarationImpl::getShortHandValue): Ditto.
789 (CSSStyleDeclarationImpl::merge): Ditto.
790 (CSSStyleDeclarationImpl::diff): Ditto.
791 * khtml/editing/htmlediting.cpp:
792 (khtml::StyleChange::currentlyHasStyle): Ditto.
793 (khtml::ApplyStyleCommand::removeCSSStyle): Ditto.
794 * khtml/html/html_baseimpl.cpp: (HTMLBodyElementImpl::parseHTMLAttribute): Ditto.
795 * khtml/html/html_tableimpl.cpp: (HTMLTableElementImpl::parseHTMLAttribute): Ditto.
797 2004-11-15 Darin Adler <darin@apple.com>
801 Use separate mutable style and computed style types as appropriate.
802 For now this should have no effect, but it prepares us for refactoring later.
803 Also remove some unnecessary "DOM::" prefixes and in one case factor out
806 * khtml/khtml_part.cpp:
807 (KHTMLPart::typingStyle):
808 (KHTMLPart::setTypingStyle):
810 (KHTMLPart::selectionHasStyle):
811 (KHTMLPart::selectionStartHasStyle):
812 (KHTMLPart::selectionComputedStyle):
813 * khtml/khtml_part.h:
814 * khtml/khtmlpart_p.h:
816 * khtml/css/css_base.h:
817 * khtml/css/css_ruleimpl.cpp:
818 (CSSStyleRuleImpl::setDeclaration):
819 * khtml/css/css_ruleimpl.h:
820 (DOM::CSSFontFaceRuleImpl::style):
821 (DOM::CSSPageRuleImpl::style):
822 (DOM::CSSStyleRuleImpl::style):
823 (DOM::CSSStyleRuleImpl::declaration):
824 * khtml/css/css_valueimpl.h:
825 (DOM::CSSPrimitiveValueImpl::):
826 * khtml/css/cssparser.cpp:
827 (CSSParser::parseValue):
828 (CSSParser::parseColor):
829 (CSSParser::parseDeclaration):
830 (CSSParser::createStyleDeclaration):
831 * khtml/css/cssparser.h:
832 * khtml/css/cssstyleselector.cpp:
833 (khtml::CSSStyleSelector::addMatchedDeclaration):
834 (khtml::CSSStyleSelector::matchRulesForList):
835 (khtml::CSSStyleSelector::styleForElement):
836 (khtml::CSSStyleSelector::applyDeclarations):
837 * khtml/css/cssstyleselector.h:
838 * khtml/css/parser.cpp:
839 * khtml/css/parser.y:
840 * khtml/dom/css_rule.h:
841 * khtml/dom/css_stylesheet.h:
842 * khtml/dom/css_value.h:
843 * khtml/dom/dom2_views.cpp:
844 * khtml/xml/dom2_viewsimpl.cpp:
845 (DOM::AbstractViewImpl::getComputedStyle):
846 * khtml/xml/dom_docimpl.cpp:
847 (DocumentImpl::importNode):
848 (DocumentImpl::setStyleSheet):
849 * khtml/xml/dom_docimpl.h:
850 * khtml/xml/dom_xmlimpl.cpp:
851 (DOM::ProcessingInstructionImpl::setStyleSheet):
852 * khtml/xml/dom_xmlimpl.h:
854 * khtml/dom/css_value.cpp:
855 (DOM::throwException): Added.
856 (DOM::CSSStyleDeclaration::setCssText): Call throwException, but always on 0 for now.
857 The real thing is coming with the next change to refactor.
858 (DOM::CSSPrimitiveValue::setFloatValue): Call throwException.
859 (DOM::CSSPrimitiveValue::setStringValue): Ditto.
861 2004-11-15 Darin Adler <darin@apple.com>
865 - fixed <rdar://problem/3878489> REGRESSION: modifying attribute of <textarea> blows away edited text (breaks simplemachines.org forum)
867 * khtml/xml/dom_nodeimpl.h: Added boolean "children changed" parameter to
868 dispatchSubtreeModifiedEvent, so it can be called in cases where only the
869 node's attributes changed without sending a misleading childrenChanged call,
870 but the childrenChanged call can happen at the exact right moment.
871 * khtml/xml/dom_nodeimpl.cpp: Removed some uneeded "DOM::".
872 (NodeImpl::dispatchSubtreeModifiedEvent): Only call "children changed" if
873 the boolean true is passed in.
875 * khtml/xml/dom_elementimpl.cpp:
876 (NamedAttrMapImpl::addAttribute): Pass false for "children changed".
877 (NamedAttrMapImpl::removeAttribute): Ditto.
879 2004-11-15 John Sullivan <sullivan@apple.com>
883 - fixed <rdar://problem/3880075> leak in CSSStyleDeclarationImpl::copyPropertiesInSet,
884 seen often in Mail and Blot
886 * khtml/css/css_valueimpl.cpp:
887 (CSSStyleDeclarationImpl::copyPropertiesInSet):
888 delete temporary list after we're done using it
890 2004-11-15 Richard Williamson <rjw@apple.com>
892 Fixed leak (3879883) that John found. Early return leaked
897 * khtml/css/css_computedstyle.cpp:
898 (DOM::CSSComputedStyleDeclarationImpl::getPropertyCSSValue):
900 2004-11-15 Ken Kocienda <kocienda@apple.com>
906 <rdar://problem/3879569> Many leaks in EditCommand mechanism, seen in Mail
908 Fixed a couple of object lifetime issues. The EditCommand class used to hold an
909 EditCommandPtr to its parent, but this caused a a reference cycle in composite
910 commands as the children held a ref to their parent. Now, the parent variable
911 is a non-retained reference to an EditCommand *. It would be nice to have a
912 weak reference to the parent or even override deref in composite commands (but I
913 can't since deref() is not virtual). However, this should be OK since any
914 dangling parent pointer is a sign of a bigger object lifetime problem that
915 would need to be addressed anyway.
917 * khtml/css/css_valueimpl.cpp:
918 (CSSStyleDeclarationImpl::CSSStyleDeclarationImpl): Fix bug in constructor that takes a
919 QPtrList<CSSProperty> *. List values must be copied into newly-allocated list, rather than
920 just assigning the list variable passed in to the local list variable, or the list will be
922 * khtml/editing/htmlediting.cpp:
923 (khtml::EditCommand::setStartingSelection): No longer call get(). m_parent is no longer a smart pointer.
924 (khtml::EditCommand::setEndingSelection): Ditto.
925 (khtml::EditCommand::assignTypingStyle): Short-circuit if passed in style is identical to current style.
926 Unrelated to the change, but saves some ref's and deref's.
927 (khtml::EditCommand::setTypingStyle): No longer call get(). m_parent is no longer a smart pointer.
928 * khtml/editing/htmlediting.h: Change m_parent to a EditCommand *. Was an EditCommandPtr. Using an
929 EditCommandPtr caused a reference cycle in composite commands as the children held a ref to their parent.
930 (khtml::EditCommand::parent): No longer call get(). m_parent is no longer a smart pointer.
932 2004-11-15 Maciej Stachowiak <mjs@apple.com>
936 <rdar://problem/3807080> Safari so slow it seems like a hang accessing a page on an IBM website
938 * khtml/xml/dom_nodeimpl.cpp:
939 (NodeListImpl::NodeListImpl): Initialize isItemCacheValid, renamed isCacheValid to
941 (NodeListImpl::recursiveLength): Adjusted for rename.
942 (NodeListImpl::recursiveItem): Cache the last item accessed and its offset.
943 If the same offset is looked up again, just return it, otherwise, if looking up
944 a later offset, start at the last item and proceed from there.
945 (NodeListImpl::itemById): Apply the special document optimization to all
946 nodes that are either a document or in a document - just walk up to make
947 sure the node found by ID has the root node as an ancestor.
948 (NodeListImpl::rootNodeSubtreeModified): Adjust both cache bits.
949 * khtml/xml/dom_nodeimpl.h: Prototype new stuff.
951 2004-11-15 John Sullivan <sullivan@apple.com>
955 - fixed <rdar://problem/3879539> leak of NSString after pasting into editable HTML (e.g. Mail)
957 * kwq/KWQKHTMLPart.mm:
958 (KWQKHTMLPart::documentFragmentWithText):
959 release mutable copy of string after we're done using it
961 2004-11-14 Kevin Decker <kdecker@apple.com>
965 fixed: <rdar://problem/3823038> LEAK: huge leak in DOM::HTMLElementImpl::createContextualFragment(DOM::DOMString const&, bool)
967 * khtml/html/html_elementimpl.cpp:
968 (HTMLElementImpl::setInnerHTML): uses the ref counting system to deallocate fragments instead of explicitly invoking a destructor.
969 (HTMLElementImpl::setOuterHTML): function is responsible for derefing the fragment prior to returning. Now it does.
971 2004-11-13 Maciej Stachowiak <mjs@apple.com>
975 <rdar://problem/3878766> VIP: Program listings pages at directv.com takes 75% of time traversing NodeLists
977 * khtml/dom/dom_node.cpp:
978 (NodeList::itemById): New method, just forward to impl.
979 * khtml/dom/dom_node.h: Prototype it.
980 * khtml/ecma/kjs_dom.cpp:
981 (DOMNodeList::tryGet): Instead of looping over the whole list to do by-id access,
982 let the NodeList do it. The NodeList might be able to do it more efficiently.
983 * khtml/xml/dom_nodeimpl.cpp:
984 (NodeListImpl::itemById): Optimize for the case where the NodeList
985 covers the whole document. In this case, just use getElementById,
986 then check that the element satisfies the list criteria.
987 (ChildNodeListImpl::nodeMatches): Return true only if the node is our child.
988 (TagNodeListImpl::TagNodeListImpl): Irrelevant change to reformat initializers.
989 * khtml/xml/dom_nodeimpl.h:
991 2004-11-12 Maciej Stachowiak <mjs@apple.com>
995 - fixed another bug in the last checkin, isCacheValid was unitialized, resulting in
996 sometimes using a huge bogus length value.
998 * khtml/xml/dom_nodeimpl.cpp:
999 (NodeListImpl::NodeListImpl): Initialize isCacheValid.
1001 2004-11-12 Darin Adler <darin@apple.com>
1005 - fixed an infinite loop in that last check-in
1007 * khtml/xml/dom_nodeimpl.cpp: (NodeImpl::notifyLocalNodeListsSubtreeModified):
1008 Added a ++i to the loop so it won't get stuck on the first element in the list.
1010 2004-11-12 Maciej Stachowiak <mjs@apple.com>
1014 - fixed <rdar://problem/3878183> Safari is 77% slower than it should be on a page on an IBM website due to NodeListImpl length
1016 I fixed this by changing NodeLists to cache their length, but
1017 invalidate it whenever there is a change in the DOM subtree at
1018 which they are rooted. This makes NodeListImpl::recursiveLength()
1019 drop completely off the profile, since we were repeatedly getting
1020 a length for the same NodeList over and over.
1022 * khtml/xml/dom_nodeimpl.cpp:
1023 (NodeImpl::NodeImpl):
1024 (NodeImpl::~NodeImpl):
1025 (NodeImpl::registerNodeList):
1026 (NodeImpl::unregisterNodeList):
1027 (NodeImpl::notifyLocalNodeListsSubtreeModified):
1028 (NodeImpl::notifyNodeListsSubtreeModified):
1029 (NodeImpl::dispatchSubtreeModifiedEvent):
1030 (NodeListImpl::NodeListImpl):
1031 (NodeListImpl::~NodeListImpl):
1032 (NodeListImpl::recursiveLength):
1033 (NodeListImpl::recursiveItem):
1034 (NodeListImpl::rootNodeSubtreeModified):
1035 (ChildNodeListImpl::ChildNodeListImpl):
1036 (ChildNodeListImpl::length):
1037 (ChildNodeListImpl::item):
1038 (TagNodeListImpl::TagNodeListImpl):
1039 (TagNodeListImpl::length):
1040 (TagNodeListImpl::item):
1041 (NameNodeListImpl::NameNodeListImpl):
1042 (NameNodeListImpl::length):
1043 (NameNodeListImpl::item):
1044 * khtml/xml/dom_nodeimpl.h:
1046 2004-11-12 Darin Adler <darin@apple.com>
1050 - various small cleanups
1052 * khtml/xml/dom_docimpl.h: Added policyBaseURL and setPolicyBaseURL.
1053 * khtml/html/html_documentimpl.h: Removed policyBaseURL and setPolicyBaseURL.
1055 * khtml/xml/xml_tokenizer.h: Marked isWaitingForScripts const.
1056 * khtml/xml/xml_tokenizer.cpp: (khtml::XMLTokenizer::isWaitingForScripts): Marked const.
1057 * khtml/html/htmltokenizer.h: Marked isWaitingForScripts const.
1058 * khtml/html/htmltokenizer.cpp:
1059 (khtml::HTMLTokenizer::isWaitingForScripts): Marked const.
1060 (khtml::HTMLTokenizer::setOnHold): Took out extraneous line of code.
1062 * khtml/khtml_part.h: Removed docImpl function.
1063 * khtml/khtml_part.cpp: Ditto.
1065 * khtml/xml/dom_docimpl.cpp: (DocumentImpl::close): Simplified code that implements
1066 the "redirect during onload" optimization. Now uses isScheduledLocationChangePending.
1068 * kwq/KWQKHTMLPart.h: Removed now-unused _firstResponderAtMouseDownTime.
1069 * kwq/KWQKHTMLPart.mm: Removed _firstResponderAtMouseDownTime (forgot to land this
1070 part of the change last time, which is why the build broke).
1071 (KWQKHTMLPart::updatePolicyBaseURL): Use xmlDocImpl instead of docImpl.
1072 (KWQKHTMLPart::setPolicyBaseURL): Ditto.
1073 (KWQKHTMLPart::keyEvent): Ditto.
1074 (KWQKHTMLPart::dispatchCPPEvent): Ditto.
1075 (KWQKHTMLPart::bodyBackgroundColor): Ditto.
1077 2004-11-12 Chris Blumenberg <cblu@apple.com>
1079 <rdar://problem/3843312> REGRESSION: Tabbing into content area puts insertion point at start, should go to where it last was
1083 * kwq/KWQKHTMLPart.mm:
1084 (KWQKHTMLPart::nextKeyViewInFrameHierarchy): only blow away selection when another view is focused
1088 2004-11-12 Darin Adler <darin@apple.com>
1092 - fixed a couple places that would not work for XML documents
1094 * khtml/ecma/kjs_window.cpp:
1095 (Window::isSafeScript): Use xmlDocImpl instead of docImpl, since the function we're using
1096 is present on the base class.
1097 (WindowFunc::tryCall): More of the same.
1099 2004-11-12 Darin Adler <darin@apple.com>
1101 - land versions of these files generated by the newer gperf
1103 People building on Panther will continue to see these files modified.
1104 A workaround would be to install the newer gperf on our Tiger build machines.
1106 * khtml/css/cssproperties.c: Regenerated.
1107 * khtml/css/cssvalues.c: Regenerated.
1108 * khtml/html/doctypes.cpp: Regenerated.
1109 * khtml/html/kentities.c: Regenerated.
1110 * khtml/misc/htmlattrs.c: Regenerated.
1111 * khtml/misc/htmltags.c: Regenerated.
1112 * kwq/KWQColorData.c: Regenerated.
1114 2004-11-11 Richard Williamson <rjw@apple.com>
1116 Fix build horkage from previous checkin.
1118 * kwq/KWQKHTMLPart.h:
1120 2004-11-11 Darin Adler <darin@apple.com>
1124 - fixed <rdar://problem/3846152> REGRESSION (125-166): can't drag text out of <input type=text> fields
1126 * kwq/WebCoreBridge.h: Added wasFirstResponderAtMouseDownTime: method.
1128 * kwq/KWQKHTMLPart.h: Removed _firstResponderAtMouseDownTime.
1129 * kwq/KWQKHTMLPart.mm:
1130 (KWQKHTMLPart::passWidgetMouseDownEventToWidget): Use the new wasFirstResponderAtMouseDownTime:
1131 method on the bridge instead of _firstResponderAtMouseDownTime. This will return YES for the case where
1132 we started with the NSTextField as first responder, and then took focus away and gave it back, which
1133 makes dragging text work again.
1134 (KWQKHTMLPart::mouseDown): Removed code to set _firstResponderAtMouseDownTime.
1136 2004-11-11 David Hyatt <hyatt@apple.com>
1138 Disable the tokenizer deferral, since it hurts the PLT by 3% or so.
1142 * khtml/html/htmltokenizer.cpp:
1143 (khtml::HTMLTokenizer::continueProcessing):
1145 2004-11-11 Ken Kocienda <kocienda@apple.com>
1149 * khtml/editing/htmlediting.cpp:
1150 (khtml::InsertLineBreakCommand::doApply): Use new isLastVisiblePositionInBlock() helper instead
1151 of old isLastInBlock() member function on VisiblePosition. This is a cosmetic change in keeping
1152 with the prevailing style for the VisiblePosition class.
1153 * khtml/editing/htmlediting.h: Move isLastVisiblePositionInNode() function to visible_position.[cpp|h] files.
1154 * khtml/editing/visible_position.cpp: Removed isLastInBlock() helper. Renamed to isLastVisiblePositionInBlock().
1155 (khtml::visiblePositionsInDifferentBlocks): New helper method.
1156 (khtml::isLastVisiblePositionInBlock): Ditto.
1157 (khtml::isLastVisiblePositionInNode): Ditto.
1158 * khtml/editing/visible_position.h: Add declarations for new functions.
1160 2004-11-11 Ken Kocienda <kocienda@apple.com>
1164 * khtml/editing/htmlediting.cpp:
1165 (khtml::CompositeEditCommand::deleteInsignificantText): Call new compareBoundaryPoints convenience.
1166 (khtml::ApplyStyleCommand::removeBlockStyle): Ditto.
1167 (khtml::ApplyStyleCommand::removeInlineStyle): Ditto.
1168 (khtml::ApplyStyleCommand::nodeFullySelected): Ditto.
1169 (khtml::DeleteSelectionCommand::initializePositionData): Ditto.
1170 * khtml/xml/dom2_rangeimpl.cpp:
1171 (DOM::RangeImpl::compareBoundaryPoints): New convenience variant of this function which takes two Position objects.
1172 * khtml/xml/dom2_rangeimpl.h: Ditto.
1174 2004-11-11 Ken Kocienda <kocienda@apple.com>
1176 Reviewed by Harrison
1178 Some improvements to deleting when complete lines are selected.
1180 * khtml/editing/htmlediting.cpp:
1181 (khtml::DeleteSelectionCommand::initializePositionData): Detect when the line containing
1182 the end of a selection is fully selected. Turn off block merging in this case.
1183 (khtml::DeleteSelectionCommand::handleSpecialCaseBRDelete): Fix a bug in the check for
1184 whether a BR immediately followed a block. The old code could erroneously skip nodes.
1185 (khtml::DeleteSelectionCommand::handleGeneralDelete): Add a case for when the entire
1186 start block is selected. This new code will now delete this block in one call, rather
1187 than iterating over each child.
1188 * khtml/editing/visible_position.cpp:
1189 (khtml::visiblePositionsOnDifferentLines): New helper called in initializePositionData()
1190 to do the work mentioned above in the comment for that function.
1191 (khtml::isFirstVisiblePositionOnLine): Ditto.
1192 (khtml::isLastVisiblePositionOnLine): Ditto.
1193 * khtml/editing/visible_position.h: Add new functions.
1194 * layout-tests/editing/deleting/delete-line-001-expected.txt: Added.
1195 * layout-tests/editing/deleting/delete-line-001.html: Added.
1196 * layout-tests/editing/deleting/delete-line-002-expected.txt: Added.
1197 * layout-tests/editing/deleting/delete-line-002.html: Added.
1198 * layout-tests/editing/deleting/delete-line-003-expected.txt: Added.
1199 * layout-tests/editing/deleting/delete-line-003.html: Added.
1200 * layout-tests/editing/deleting/delete-line-004-expected.txt: Added.
1201 * layout-tests/editing/deleting/delete-line-004.html: Added.
1202 * layout-tests/editing/deleting/delete-line-005-expected.txt: Added.
1203 * layout-tests/editing/deleting/delete-line-005.html: Added.
1204 * layout-tests/editing/deleting/delete-line-006-expected.txt: Added.
1205 * layout-tests/editing/deleting/delete-line-006.html: Added.
1206 * layout-tests/editing/deleting/delete-line-007-expected.txt: Added.
1207 * layout-tests/editing/deleting/delete-line-007.html: Added.
1208 * layout-tests/editing/deleting/delete-line-008-expected.txt: Added.
1209 * layout-tests/editing/deleting/delete-line-008.html: Added.
1210 * layout-tests/editing/deleting/delete-line-009-expected.txt: Added.
1211 * layout-tests/editing/deleting/delete-line-009.html: Added.
1212 * layout-tests/editing/deleting/delete-line-010-expected.txt: Added.
1213 * layout-tests/editing/deleting/delete-line-010.html: Added.
1214 * layout-tests/editing/deleting/delete-line-011-expected.txt: Added.
1215 * layout-tests/editing/deleting/delete-line-011.html: Added.
1216 * layout-tests/editing/deleting/delete-line-012-expected.txt: Added.
1217 * layout-tests/editing/deleting/delete-line-012.html: Added.
1219 2004-11-11 Ken Kocienda <kocienda@apple.com>
1223 * khtml/editing/htmlediting.cpp:
1224 (khtml::DeleteSelectionCommand::initializePositionData): Add some comments and a new piece of debugging output.
1226 2004-11-11 Ken Kocienda <kocienda@apple.com>
1232 <rdar://problem/3875618> REGRESSION (Mail): Hitting down arrow with full line selected skips line (br case)
1233 <rdar://problem/3875641> REGRESSION (Mail): Hitting down arrow with full line selected skips line (div case)
1235 * khtml/editing/selection.cpp:
1236 (khtml::Selection::modifyMovingRightForward): Fixed by juggling the position as the starting point for
1237 the next line position when necessary.
1238 * layout-tests/editing/selection/move-3875618-fix-expected.txt: Added.
1239 * layout-tests/editing/selection/move-3875618-fix.html: Added.
1240 * layout-tests/editing/selection/move-3875641-fix-expected.txt: Added.
1241 * layout-tests/editing/selection/move-3875641-fix.html: Added.
1243 2004-11-11 Ken Kocienda <kocienda@apple.com>
1247 Improved some function names, at John's urging. No changes to the
1248 functions themselves.
1250 canPerformSpecialCaseAllContentDelete() --> handleSpecialCaseAllContentDelete()
1251 canPerformSpecialCaseBRDelete() --> handleSpecialCaseBRDelete()
1252 performGeneralDelete() --> handleGeneralDelete()
1254 * khtml/editing/htmlediting.cpp:
1255 (khtml::DeleteSelectionCommand::handleSpecialCaseAllContentDelete)
1256 (khtml::DeleteSelectionCommand::handleSpecialCaseBRDelete)
1257 (khtml::DeleteSelectionCommand::handleGeneralDelete)
1258 (khtml::DeleteSelectionCommand::doApply)
1259 * khtml/editing/htmlediting.h
1261 2004-11-11 Ken Kocienda <kocienda@apple.com>
1265 Updated some layout test results that changed as a result of my last checking.
1266 Added a new test that has been in my tree for a few days.
1268 * layout-tests/editing/deleting/delete-3775172-fix-expected.txt:
1269 * layout-tests/editing/deleting/delete-3800834-fix-expected.txt:
1270 * layout-tests/editing/inserting/insert-3851164-fix-expected.txt: Added.
1271 * layout-tests/editing/inserting/insert-3851164-fix.html: Added.
1273 2004-11-11 Ken Kocienda <kocienda@apple.com>
1277 * khtml/editing/htmlediting.cpp:
1278 (khtml::debugNode): New debugging helper.
1279 (khtml::DeleteSelectionCommand::initializePositionData): No longer call obsoleted
1280 startPositionForDelete() and endPositionForDelete() functions. Just use the
1281 m_selectionToDelete object to determine start and end positions for the delete.
1282 (khtml::DeleteSelectionCommand::canPerformSpecialCaseAllContentDelete): New
1283 function that creates a special case for deleting all the content in a root
1285 (khtml::DeleteSelectionCommand::doApply): Call canPerformSpecialCaseAllContentDelete()
1286 function before BR special case and the general case delete functions.
1287 * khtml/editing/htmlediting.h: Updated for changed functions.
1289 2004-11-10 Kevin Decker <kdecker@apple.com>
1293 Fixed <rdar://problem/3875011> DOMNodeList::tryGet() performs unnecessary (and expensive) dom tree traversals. Improved a loop from 2-n-squared to just n-squared.
1295 * khtml/ecma/kjs_dom.cpp:
1296 (DOMNodeList::tryGet): Got rid of an unnecessary node traversal.
1298 2004-11-10 Ken Kocienda <kocienda@apple.com>
1302 * khtml/editing/htmlediting.cpp:
1303 (khtml::DeleteSelectionCommand::initializePositionData): Move position adjustments for
1304 smart delete from the two functions below to here. There was an unnecessary double
1305 calculation of the leading and trailing whitespace positions. Also refined the trailing
1306 case so it only acts when the leading position is null (which seems to match TextEdit in
1307 my tests). Also removed some unnecessary copying of Position objects.
1308 (khtml::DeleteSelectionCommand::startPositionForDelete): Move out smart delete adjustment
1310 (khtml::DeleteSelectionCommand::endPositionForDelete): Ditto.
1312 2004-11-10 Ken Kocienda <kocienda@apple.com>
1314 Reviewed by Harrison
1316 (khtml::DeleteSelectionCommand::performGeneralDelete): Add some more comments to
1317 make things more clear.
1318 * khtml/editing/selection.cpp:
1319 (khtml::Selection::toRange): Fixed the upstream and downstream calls so that the
1320 resulting positions do not cross block boundaries. This was a bug and caused some
1321 delete problems when whole blocks were selected. I will be addressing that issue
1322 more fully in upcoming changes.
1324 2004-11-10 Ken Kocienda <kocienda@apple.com>
1326 Reviewed by Harrison
1328 Some cleanups and fixes in upstream and downstream functions.
1330 Removed redundant checks for isBlockFlow() when calling enclosingBlockFlowElement().
1331 Blocks do not need to skip the call to enclosingBlockFlowElement() for fear that the
1332 block's enclosing block will be returned.
1334 Remove code from upstream that confined the serach to block boundaries outside of
1335 the code which runs in the StayInBlock case. This code was redundant, and caused
1336 incorrect results to be returned in the DoNotStayInBlock case.
1338 In downstream, the check for crossing into a new block should use the equivalentDeepPosition()
1339 node, not the the this pointer's node.
1341 * khtml/xml/dom_position.cpp:
1342 (DOM::Position::upstream)
1343 (DOM::Position::downstream)
1345 2004-11-09 David Hyatt <hyatt@apple.com>
1347 Fix for 3873234, Safari UI is unresponsive when parsing multiple HTML docs and 3873233, Safari hangs when
1348 loading large local files.
1352 * khtml/html/htmltokenizer.cpp:
1353 (khtml::HTMLTokenizer::HTMLTokenizer):
1354 (khtml::HTMLTokenizer::reset):
1355 (khtml::HTMLTokenizer::write):
1356 (khtml::HTMLTokenizer::stopped):
1357 (khtml::HTMLTokenizer::processingData):
1358 (khtml::HTMLTokenizer::continueProcessing):
1359 (khtml::HTMLTokenizer::timerEvent):
1360 (khtml::HTMLTokenizer::allDataProcessed):
1361 (khtml::HTMLTokenizer::end):
1362 (khtml::HTMLTokenizer::finish):
1363 (khtml::HTMLTokenizer::notifyFinished):
1364 * khtml/html/htmltokenizer.h:
1365 * khtml/khtml_part.cpp:
1366 (KHTMLPart::slotFinished):
1369 * khtml/khtml_part.h:
1370 (KHTMLPart::tokenizerProcessedData):
1371 * khtml/khtmlview.cpp:
1372 * khtml/xml/dom_docimpl.cpp:
1373 * khtml/xml/xml_tokenizer.h:
1374 (khtml::Tokenizer::stopped):
1375 (khtml::Tokenizer::processingData):
1376 * kwq/KWQDateTime.h:
1377 * kwq/KWQDateTime.mm:
1378 (QDateTime::secsTo):
1379 (KWQUIEventTime::uiEventPending):
1380 * kwq/KWQKHTMLPart.h:
1381 * kwq/KWQKHTMLPart.mm:
1382 (KWQKHTMLPart::tokenizerProcessedData):
1383 * kwq/WebCoreBridge.h:
1384 * kwq/WebCoreBridge.mm:
1385 (-[WebCoreBridge stop]):
1386 (-[WebCoreBridge numPendingOrLoadingRequests]):
1387 (-[WebCoreBridge doneProcessingData]):
1389 2004-11-09 David Harrison <harrison@apple.com>
1391 Reviewed by Ken Kocienda.
1393 <rdar://problem/3865837> Wrong text style after delete to start of document
1395 * khtml/editing/htmlediting.cpp:
1396 (khtml::DeleteSelectionCommand::saveTypingStyleState):
1397 Sample computedStyle of m_selectionToDelete.start instead of m_downstreamStart.
1399 2004-11-09 Richard Williamson <rjw@apple.com>
1401 Fixed <rdar://problem/3872440> NSTimer prematurely released.
1410 2004-11-09 Chris Blumenberg <cblu@apple.com>
1414 * WebCore.pbproj/project.pbxproj: explicitly link against libxml2.2.6.14.dylib since the version number has been bumped
1416 2004-11-08 David Harrison <harrison@apple.com>
1418 Reviewed by Ken Kocienda.
1420 <rdar://problem/3865854> Deleting first line deletes all lines
1422 * khtml/editing/htmlediting.cpp:
1423 (khtml::DeleteSelectionCommand::performGeneralDelete):
1424 Problem was that the code that deletes fully selected m_downstreamEnd.node() by deleting one
1425 of its ancestors, failed to end the loop that deletes all fully selected nodes. Also,
1426 fixed this code to clear m_trailingWhitespaceValid. Also removed dead m_endingPosition
1427 update because it is handled in calculateEndingPosition now.
1428 * layout-tests/editing/deleting/delete-3865854-fix-expected.txt: Added.
1429 * layout-tests/editing/deleting/delete-3865854-fix.html: Added.
1431 2004-11-08 Ken Kocienda <kocienda@apple.com>
1435 * khtml/html/html_elementimpl.cpp:
1436 (HTMLElementImpl::createContextualFragment): Now takes flag to control whether comments
1437 are added to the DOM.
1438 * khtml/html/html_elementimpl.h: Ditto.
1439 * khtml/html/htmlparser.cpp:
1440 (KHTMLParser::KHTMLParser): Ditto.
1441 (KHTMLParser::getElement): Remove ifdef for comment processing. Replace with flag check.
1442 * khtml/html/htmlparser.h: Add flag to constructor so callers can request comment nodes.
1443 * khtml/html/htmltokenizer.cpp:
1444 (khtml::HTMLTokenizer::HTMLTokenizer): Add flag to constructor so callers can request comment nodes.
1445 (khtml::HTMLTokenizer::parseComment): Fix code to handle parsing out comment text correctly.
1446 There were a couple of indexing errors that resulted in the comment text containing part of the
1448 (khtml::HTMLTokenizer::processToken): Don't let token id be reset to ID_TEXT if token is a comment.
1449 * khtml/html/htmltokenizer.h: Add flag to constructor so callers can request comment nodes.
1451 2004-11-08 Chris Blumenberg <cblu@apple.com>
1453 Fixed: <rdar://problem/3870907> WebCore unnecessary links against JavaVM and Security
1457 * WebCore.pbproj/project.pbxproj: stop unnecessary linking
1458 * khtml/html/html_objectimpl.h: don't unnecessarily include JavaVM header
1459 * kwq/KWQKHTMLPart.h: ditto
1461 2004-11-08 Darin Adler <darin@apple.com>
1465 - fixed <rdar://problem/3825966> 8A274 Safari crashes closing window: QTimer::fire() with MallocStackLogging and MallocScribble enabled
1467 * kwq/KWQTimer.mm: (QTimer::fire): Rearrange so we don't access the QTimer object after calling code
1468 that possibly deletes the QTimer.
1470 2004-11-08 Chris Blumenberg <cblu@apple.com>
1472 Fixed: <rdar://problem/3783904> Return key behavior is confusingly different between popup menus and autofill menus
1476 * kwq/KWQTextField.mm:
1477 (-[KWQTextFieldController textView:shouldHandleEvent:]): let the bridge have a crack at the event so that it can swallow the newline if it wants to
1478 * kwq/WebCoreBridge.h:
1480 2004-11-08 David Harrison <harrison@apple.com>
1484 Renamed NodeImpl::enclosingNonBlockFlowElement to NodeImpl::enclosingInlineElement, per Hyatt.
1486 * khtml/editing/htmlediting.cpp:
1487 (khtml::DeleteSelectionCommand::moveNodesAfterNode):
1488 * khtml/xml/dom_nodeimpl.cpp:
1489 (NodeImpl::enclosingInlineElement):
1490 * khtml/xml/dom_nodeimpl.h:
1492 2004-11-05 Chris Blumenberg <cblu@apple.com>
1494 Fixed: <rdar://problem/3838413> REGRESSION (Mail): "Smart" word paste adds spaces before/after special characters
1498 * khtml/editing/htmlediting.cpp:
1499 (khtml::ReplaceSelectionCommand::doApply): call isCharacterSmartReplaceExempt on the part to see if a space should be inserted
1500 * khtml/editing/visible_position.cpp:
1501 (khtml::VisiblePosition::character): new, returns the character for the position
1502 * khtml/editing/visible_position.h:
1503 * kwq/KWQKHTMLPart.h:
1504 * kwq/KWQKHTMLPart.mm:
1505 (KWQKHTMLPart::isCharacterSmartReplaceExempt): new, calls the bridge
1506 * kwq/WebCoreBridge.h:
1510 2004-11-05 Adele Amchan <adele@apple.com>
1514 Fix for <rdar://problem/3854383> REGRESSION(166-168) input fields show black background when background color is set to transparent
1515 and a workaround for displaying transparent backgrounds for textareas.
1517 * kwq/KWQLineEdit.mm: (QLineEdit::setPalette): If the background color is transparent (we check the alpha value) then we set the background to white
1518 * kwq/KWQTextEdit.mm: (QTextEdit::setPalette): If the background color is transparent, then we don't draw the background
1519 * kwq/KWQTextArea.mm: (-[KWQTextArea setDrawsBackground:]): added setDrawsBackground function which calls setDrawsBackground on the super class,
1520 on the contentView, and on the textView.
1522 2004-11-04 David Hyatt <hyatt@apple.com>
1524 Fix for relpositioned inlines. This was reviewed a long time ago, but I can't recall who reviewed it (either
1527 Reviewed by darin or ken
1529 * khtml/rendering/bidi.cpp:
1530 (khtml::appendRunsForObject):
1531 (khtml::RenderBlock::skipWhitespace):
1532 (khtml::RenderBlock::findNextLineBreak):
1533 * khtml/rendering/render_block.cpp:
1534 (khtml::RenderBlock::lowestPosition):
1535 (khtml::RenderBlock::rightmostPosition):
1536 (khtml::RenderBlock::leftmostPosition):
1537 * khtml/rendering/render_box.cpp:
1538 (RenderBox::position):
1539 * khtml/rendering/render_box.h:
1540 (khtml::RenderBox::staticX):
1541 (khtml::RenderBox::staticY):
1542 * khtml/rendering/render_layer.cpp:
1543 (RenderLayer::updateLayerPosition):
1544 (RenderLayer::convertToLayerCoords):
1545 * khtml/rendering/render_line.cpp:
1546 (khtml::InlineFlowBox::placeBoxesHorizontally):
1547 * khtml/rendering/render_object.h:
1548 (khtml::RenderObject::staticX):
1549 (khtml::RenderObject::staticY):
1551 Finish turning on XSLT. Make sure child stylesheets can load.
1553 * khtml/xsl/xslt_processorimpl.cpp:
1554 (DOM::stylesheetLoadFunc):
1555 (DOM::XSLTProcessorImpl::transformDocument):
1557 2004-11-04 David Hyatt <hyatt@apple.com>
1559 Implement CSS3 support for multiple backgrounds. Also fix a bug with background propagation so that it only
1560 happens (from the <body> to the root) for HTML documents. Fixed background-position to handle a mixture of
1561 keyword and length values.
1565 * khtml/css/cssparser.cpp:
1566 (CSSParser::parseValue):
1567 (CSSParser::addBackgroundValue):
1568 (CSSParser::parseBackgroundShorthand):
1569 (CSSParser::parseBackgroundColor):
1570 (CSSParser::parseBackgroundImage):
1571 (CSSParser::parseBackgroundPositionXY):
1572 (CSSParser::parseBackgroundPosition):
1573 (CSSParser::parseBackgroundProperty):
1574 (CSSParser::parseColorFromValue):
1575 * khtml/css/cssparser.h:
1576 * khtml/css/cssstyleselector.cpp:
1577 (khtml::CSSStyleSelector::adjustRenderStyle):
1578 (khtml::CSSStyleSelector::applyProperty):
1579 (khtml::CSSStyleSelector::mapBackgroundAttachment):
1580 (khtml::CSSStyleSelector::mapBackgroundImage):
1581 (khtml::CSSStyleSelector::mapBackgroundRepeat):
1582 (khtml::CSSStyleSelector::mapBackgroundXPosition):
1583 (khtml::CSSStyleSelector::mapBackgroundYPosition):
1584 * khtml/css/cssstyleselector.h:
1585 * khtml/rendering/render_box.cpp:
1586 (RenderBox::paintRootBoxDecorations):
1587 (RenderBox::paintBoxDecorations):
1588 (RenderBox::paintBackgrounds):
1589 (RenderBox::paintBackground):
1590 (RenderBox::paintBackgroundExtended):
1591 * khtml/rendering/render_box.h:
1592 * khtml/rendering/render_form.cpp:
1593 (RenderFieldset::paintBoxDecorations):
1594 * khtml/rendering/render_line.cpp:
1595 (khtml::InlineFlowBox::paintBackgrounds):
1596 (khtml::InlineFlowBox::paintBackground):
1597 (khtml::InlineFlowBox::paintBackgroundAndBorder):
1598 * khtml/rendering/render_line.h:
1599 * khtml/rendering/render_object.cpp:
1600 (RenderObject::setStyle):
1601 (RenderObject::updateBackgroundImages):
1602 (RenderObject::getVerticalPosition):
1603 * khtml/rendering/render_object.h:
1604 (khtml::RenderObject::paintBackgroundExtended):
1605 * khtml/rendering/render_style.cpp:
1607 (BackgroundLayer::BackgroundLayer):
1608 (BackgroundLayer::~BackgroundLayer):
1609 (BackgroundLayer::operator=):
1610 (BackgroundLayer::operator==):
1611 (BackgroundLayer::fillUnsetProperties):
1612 (BackgroundLayer::cullEmptyLayers):
1613 (StyleBackgroundData::StyleBackgroundData):
1614 (StyleBackgroundData::operator==):
1615 (RenderStyle::diff):
1616 (RenderStyle::adjustBackgroundLayers):
1617 * khtml/rendering/render_style.h:
1618 (khtml::OutlineValue::operator==):
1619 (khtml::OutlineValue::operator!=):
1620 (khtml::BackgroundLayer::backgroundImage):
1621 (khtml::BackgroundLayer::backgroundXPosition):
1622 (khtml::BackgroundLayer::backgroundYPosition):
1623 (khtml::BackgroundLayer::backgroundAttachment):
1624 (khtml::BackgroundLayer::backgroundRepeat):
1625 (khtml::BackgroundLayer::next):
1626 (khtml::BackgroundLayer::isBackgroundImageSet):
1627 (khtml::BackgroundLayer::isBackgroundXPositionSet):
1628 (khtml::BackgroundLayer::isBackgroundYPositionSet):
1629 (khtml::BackgroundLayer::isBackgroundAttachmentSet):
1630 (khtml::BackgroundLayer::isBackgroundRepeatSet):
1631 (khtml::BackgroundLayer::setBackgroundImage):
1632 (khtml::BackgroundLayer::setBackgroundXPosition):
1633 (khtml::BackgroundLayer::setBackgroundYPosition):
1634 (khtml::BackgroundLayer::setBackgroundAttachment):
1635 (khtml::BackgroundLayer::setBackgroundRepeat):
1636 (khtml::BackgroundLayer::clearBackgroundImage):
1637 (khtml::BackgroundLayer::clearBackgroundXPosition):
1638 (khtml::BackgroundLayer::clearBackgroundYPosition):
1639 (khtml::BackgroundLayer::clearBackgroundAttachment):
1640 (khtml::BackgroundLayer::clearBackgroundRepeat):
1641 (khtml::BackgroundLayer::setNext):
1642 (khtml::BackgroundLayer::operator!=):
1643 (khtml::BackgroundLayer::containsImage):
1644 (khtml::BackgroundLayer::hasImage):
1645 (khtml::BackgroundLayer::hasFixedImage):
1646 (khtml::RenderStyle::setBitDefaults):
1647 (khtml::RenderStyle::hasBackground):
1648 (khtml::RenderStyle::hasFixedBackgroundImage):
1649 (khtml::RenderStyle::outlineWidth):
1650 (khtml::RenderStyle::outlineStyle):
1651 (khtml::RenderStyle::outlineStyleIsAuto):
1652 (khtml::RenderStyle::outlineColor):
1653 (khtml::RenderStyle::backgroundColor):
1654 (khtml::RenderStyle::backgroundImage):
1655 (khtml::RenderStyle::backgroundRepeat):
1656 (khtml::RenderStyle::backgroundAttachment):
1657 (khtml::RenderStyle::backgroundXPosition):
1658 (khtml::RenderStyle::backgroundYPosition):
1659 (khtml::RenderStyle::accessBackgroundLayers):
1660 (khtml::RenderStyle::backgroundLayers):
1661 (khtml::RenderStyle::outlineOffset):
1662 (khtml::RenderStyle::resetOutline):
1663 (khtml::RenderStyle::setBackgroundColor):
1664 (khtml::RenderStyle::setOutlineWidth):
1665 (khtml::RenderStyle::setOutlineStyle):
1666 (khtml::RenderStyle::setOutlineColor):
1667 (khtml::RenderStyle::clearBackgroundLayers):
1668 (khtml::RenderStyle::inheritBackgroundLayers):
1669 (khtml::RenderStyle::setOutlineOffset):
1670 * khtml/rendering/render_table.cpp:
1671 (RenderTable::paintBoxDecorations):
1672 (RenderTableCell::paintBoxDecorations):
1674 2004-11-04 David Hyatt <hyatt@apple.com>
1676 Make sure the text decoder returns empty strings rather than null strings when the utf8 char ptr is non-null.
1677 Ensures that <a href=""> works with libxml (which returns data in utf-8 buffers).
1681 * kwq/KWQTextCodec.mm:
1682 (KWQTextDecoder::convertLatin1):
1683 (KWQTextDecoder::convertUTF16):
1684 (KWQTextDecoder::convertUsingTEC):
1685 (KWQTextDecoder::toUnicode):
1687 2004-11-04 David Hyatt <hyatt@apple.com>
1689 Make sure line-height returns the correct value for normal.
1693 * khtml/css/css_computedstyle.cpp:
1694 (DOM::CSSComputedStyleDeclarationImpl::getPropertyCSSValue):
1696 2004-11-04 David Harrison <harrison@apple.com>
1698 Reviewed by Ken Kocienda.
1700 <rdar://problem/3857753> REGRESSION (Mail): Delete incorrectly causes text to take on new style
1702 * khtml/editing/htmlediting.cpp:
1703 (khtml::DeleteSelectionCommand::moveNodesAfterNode): Fixed to move entire source subtree (up
1704 to, but not including, the enclosingBlockFlowElement) rather than just the source element.
1705 Fixed to insert after the destination subtree, rather than the destination element. Handles
1706 edge case of deleting back to the top of the tree, where there is nothing left to insert after.
1707 * khtml/xml/dom_nodeimpl.cpp:
1708 (NodeImpl::enclosingNonBlockFlowElement): New method to support moveNodesAfterNode changes.
1709 * khtml/xml/dom_nodeimpl.h: Declare NodeImpl::enclosingNonBlockFlowElement
1710 * layout-tests/editing/deleting/delete-3857753-fix-expected.txt: Added.
1711 * layout-tests/editing/deleting/delete-3857753-fix.html: Added.
1713 2004-11-03 Ken Kocienda <kocienda@apple.com>
1719 * layout-tests/editing/deleting/delete-br-008-expected.txt: Added.
1720 * layout-tests/editing/deleting/delete-br-008.html: Added.
1721 * layout-tests/editing/deleting/delete-br-009-expected.txt: Added.
1722 * layout-tests/editing/deleting/delete-br-009.html: Added.
1723 * layout-tests/editing/deleting/delete-br-010-expected.txt: Added.
1724 * layout-tests/editing/deleting/delete-br-010.html: Added.
1726 2004-11-03 Maciej Stachowiak <mjs@apple.com>
1728 Fix by Yasuo Kida, reviewed by me.
1730 <rdar://problem/3819004> REGRESSION (Mail): Can't move cursor / delete character after deleting the active input area
1732 * kwq/KWQKHTMLPart.mm:
1733 (KWQKHTMLPart::setMarkedTextRange): Treat a collapsed range the
1734 same as a nil range - setting an empty marked range should clear
1735 the marked range entirely.
1737 2004-11-02 Maciej Stachowiak <mjs@apple.com>
1739 Reviewed by Dave Hyatt (when I originally coded it).
1741 WebCore part of fix for:
1743 <rdar://problem/3759187> REGRESSION (Mail): implement firstRectForCharacterRange:
1745 * kwq/WebCoreBridge.h:
1746 * kwq/WebCoreBridge.mm:
1747 (-[WebCoreBridge firstRectForDOMRange:]): New method to compute the rect for a
1748 DOMRange, or if the range is split into multiple lines, the rect for the part on
1749 the first line only.
1751 * khtml/rendering/render_object.cpp:
1752 (RenderObject::caretRect): Added extraWidthToEndOfLine parameter and ditto
1753 for the overrides below.
1754 * khtml/rendering/render_object.h:
1755 * khtml/rendering/render_box.cpp:
1756 (RenderBox::caretRect):
1757 * khtml/rendering/render_box.h:
1758 * khtml/rendering/render_br.cpp:
1759 (RenderBR::caretRect):
1760 * khtml/rendering/render_br.h:
1761 * khtml/rendering/render_flow.cpp:
1762 (RenderFlow::caretRect):
1763 * khtml/rendering/render_flow.h:
1764 * khtml/rendering/render_text.cpp:
1765 (RenderText::caretRect):
1767 2004-11-02 Ken Kocienda <kocienda@apple.com>
1771 Implemented command to insert a block in response to typing a return key (even though
1772 I am not turning that on by default with this patch....that will come later).
1774 This new command is called InsertParagraphSeparatorCommand.
1776 Reworked the command and function names associated with inserting content into a
1777 document. Before this patch, there were inputXXX and insertXXX variants, with the
1778 former used for more high-level actions and the latter used for lower-level stuff.
1779 However, this was confusing as the AppKit uses insertXXX for everything. This resulted
1780 in an insertXXX command going through an inputXXX WebCore step and then finally to an
1781 insertXXX WebCore step. To make this less confusing, I have changes all the names to
1782 be insertXXX, and modified the lower-level operations so that it is clear what they do.
1784 * khtml/editing/htmlediting.cpp:
1785 (khtml::EditCommandPtr::isInsertTextCommand): Name change.
1786 (khtml::EditCommand::isInsertTextCommand): Ditto.
1787 (khtml::CompositeEditCommand::inputText): Ditto.
1788 (khtml::CompositeEditCommand::insertTextIntoNode): Ditto.
1789 (khtml::CompositeEditCommand::deleteTextFromNode): Ditto.
1790 (khtml::CompositeEditCommand::replaceTextInNode): Ditto.
1791 (khtml::CompositeEditCommand::deleteInsignificantText): Name changes in implementation.
1792 (khtml::CompositeEditCommand::isLastVisiblePositionInNode): Ditto.
1793 (khtml::DeleteFromTextNodeCommand::DeleteFromTextNodeCommand): Class name change, was DeleteTextCommand.
1794 (khtml::DeleteFromTextNodeCommand::~DeleteFromTextNodeCommand): Ditto.
1795 (khtml::DeleteFromTextNodeCommand::doApply): Ditto.
1796 (khtml::DeleteFromTextNodeCommand::doUnapply): Ditto.
1797 (khtml::DeleteSelectionCommand::performGeneralDelete): Ditto.
1798 (khtml::DeleteSelectionCommand::fixupWhitespace): Ditto.
1799 (khtml::DeleteSelectionCommand::moveNodesAfterNode): Ditto.
1800 (khtml::InsertIntoTextNode::InsertIntoTextNode): Class name change.
1801 (khtml::InsertIntoTextNode::~InsertIntoTextNode): Ditto.
1802 (khtml::InsertIntoTextNode::doApply): Ditto.
1803 (khtml::InsertIntoTextNode::doUnapply): Ditto.
1804 (khtml::InsertLineBreakCommand::InsertLineBreakCommand): Class name change, was InsertNewlineCommand.
1805 (khtml::InsertLineBreakCommand::insertNodeAfterPosition):
1806 (khtml::InsertLineBreakCommand::insertNodeBeforePosition):
1807 (khtml::InsertLineBreakCommand::doApply):
1808 (khtml::InsertNodeBeforeCommand::InsertNodeBeforeCommand): Code moved. No changes.
1809 (khtml::InsertNodeBeforeCommand::~InsertNodeBeforeCommand): Ditto.
1810 (khtml::InsertNodeBeforeCommand::doApply): Ditto.
1811 (khtml::InsertNodeBeforeCommand::doUnapply): Ditto.
1812 (khtml::InsertParagraphSeparatorCommand::InsertParagraphSeparatorCommand): New command.
1813 (khtml::InsertParagraphSeparatorCommand::doApply):
1814 (khtml::InsertParagraphSeparatorInQuotedContentCommand::InsertParagraphSeparatorInQuotedContentCommand):
1815 Class name change, was InsertNewlineCommandInQuotedContentCommand.
1816 (khtml::InsertParagraphSeparatorInQuotedContentCommand::~InsertParagraphSeparatorInQuotedContentCommand): Ditto.
1817 (khtml::InsertParagraphSeparatorInQuotedContentCommand::isMailBlockquote): Ditto.
1818 (khtml::InsertParagraphSeparatorInQuotedContentCommand::doApply): Ditto.
1819 (khtml::InsertTextCommand::InsertTextCommand): Class name change, was InputTextCommand.
1820 (khtml::InsertTextCommand::doApply): Ditto.
1821 (khtml::InsertTextCommand::deleteCharacter): Ditto.
1822 (khtml::InsertTextCommand::prepareForTextInsertion): Ditto.
1823 (khtml::InsertTextCommand::input): Ditto.
1824 (khtml::InsertTextCommand::insertSpace): Ditto.
1825 (khtml::InsertTextCommand::isInsertTextCommand): Ditto.
1826 (khtml::TypingCommand::insertLineBreak): Name change, was insertNewline.
1827 (khtml::TypingCommand::insertParagraphSeparatorInQuotedContent): Name change, was insertNewlineInQuotedContent.
1828 (khtml::TypingCommand::insertParagraphSeparator): New function.
1829 (khtml::TypingCommand::doApply): Name changes, as above.
1830 (khtml::TypingCommand::insertText): Ditto.
1831 (khtml::TypingCommand::deleteKeyPressed): Ditto.
1832 (khtml::TypingCommand::preservesTypingStyle): Ditto.
1833 * khtml/editing/htmlediting.h:
1834 (khtml::DeleteFromTextNodeCommand::node): Name change.
1835 (khtml::DeleteFromTextNodeCommand::offset): Ditto.
1836 (khtml::DeleteFromTextNodeCommand::count): Ditto.
1837 (khtml::InsertIntoTextNode::text): Ditto.
1838 (khtml::InsertNodeBeforeCommand::insertChild): Ditto.
1839 (khtml::InsertNodeBeforeCommand::refChild): Ditto.
1840 (khtml::TypingCommand::): Ditto.
1841 * khtml/editing/jsediting.cpp: Name changes, as above.
1842 * kwq/WebCoreBridge.h:
1843 * kwq/WebCoreBridge.mm:
1844 (-[WebCoreBridge insertLineBreak]): Name change, was insertNewline.
1845 (-[WebCoreBridge insertParagraphSeparator]): New function.
1846 (-[WebCoreBridge insertParagraphSeparatorInQuotedContent]): Name change, was insertNewlineInQuotedContent.
1848 2004-11-01 Kevin Decker <kdecker@apple.com>
1852 fixed <rdar://problem/3681094> Crash in KJS::WindowFunc::tryCall with application/xhtml+xml Content-Type
1855 * khtml/ecma/kjs_window.cpp:
1856 (WindowFunc::tryCall): Added a nil check in the case of an empty document lacking a baseURL().
1858 2004-11-01 Darin Adler <darin@apple.com>
1862 - fixed <rdar://problem/3859381> REGRESSION (167-168): text in form fields should not use body's text color
1864 * khtml/css/html4.css: Use color: initial for textarea and related ones.
1866 2004-11-01 Ken Kocienda <kocienda@apple.com>
1872 <rdar://problem/3775920> REGRESSION (Mail): Centering doesn't work in HTML mail
1874 * khtml/css/css_computedstyle.cpp:
1875 (DOM::CSSComputedStyleDeclarationImpl::copyInheritableProperties): Factor out the
1876 implementation here into new copyPropertiesInSet helper. This now calls the
1877 generalized copyPropertiesInSet function with the arguments needed to make copying
1879 * khtml/css/css_computedstyle.h:
1880 * khtml/css/css_valueimpl.cpp:
1881 (CSSStyleDeclarationImpl::diff): Move this function here from css_computedstyle.cpp.
1882 In order to do apply block properties, "regular" style declarations need to do style
1884 (CSSStyleDeclarationImpl::copyBlockProperties): New helper. Just like copyInheritableProperties
1885 except that it uses a different set of properties that apply only to blocks.
1886 (CSSStyleDeclarationImpl::copyPropertiesInSet): New helper that looks at a style declaration
1887 and copies out those properties listed in a pre-defined set.
1888 * khtml/css/css_valueimpl.h:
1889 * khtml/editing/htmlediting.cpp:
1890 (khtml::StyleChange::StyleChange): Modified to work with style changes that apply to a whole
1891 block, factoring out some of the special case code that should now only run in the inline case.
1892 (khtml::StyleChange::init): Factored out the code that now is in checkForLegacyHTMLStyleChange.
1893 (khtml::StyleChange::checkForLegacyHTMLStyleChange): New helper for case where we want
1894 special handling for "legacy" HTML styles like <B> and <I>.
1895 (khtml::ApplyStyleCommand::doApply): Much refactoring in this class to divide up the work of
1896 style changes into different kinds. CSS specifies certain properties only apply to certain
1897 element types. This set of changes now recognizes two such separate cases: styles that apply
1898 to blocks, and styles that apply to inlines.
1899 (khtml::ApplyStyleCommand::applyBlockStyle): New function to handle apply styles to whole blocks.
1900 (khtml::ApplyStyleCommand::applyInlineStyle): New function to handle apply styles to inlines.
1901 (khtml::ApplyStyleCommand::isHTMLStyleNode): Is now passed a CSSStyleDeclarationImpl to work
1902 with rather than working on the CSSStyleDeclarationImpl member variable of the class. This is
1903 done so that the function can be passed a portion of the styles being applied so that block styles
1904 and inline styles can be handled separately.
1905 (khtml::ApplyStyleCommand::removeCSSStyle): Ditto.
1906 (khtml::ApplyStyleCommand::removeBlockStyle): New function to handle removing styles from whole blocks.
1907 (khtml::ApplyStyleCommand::removeInlineStyle): New function to removing styles from inlines.
1908 (khtml::ApplyStyleCommand::addBlockStyleIfNeeded): New function to handle applying style to whole blocks.
1909 (khtml::ApplyStyleCommand::addInlineStyleIfNeeded): New function to handle applying style to inlines.
1910 * khtml/editing/htmlediting.h:
1911 (khtml::StyleChange::): Changed as described above.
1912 (khtml::StyleChange::usesLegacyStyles):
1913 (khtml::EditCommand::setEndingSelectionNeedsLayout): New function to that tells the ending selection
1914 it needs to layout, even though it has not changed position in the DOM. For instance, this is needed
1915 when text align changes.
1916 * khtml/khtml_part.cpp:
1917 (KHTMLPart::setTypingStyle): Put in an early bail-out in the case where the current style matches
1918 the passed-in argument.
1919 (KHTMLPart::applyStyle): Modify this function so that block styles are applied when the selection
1920 is a caret. Formerly, this just set typing style and made no visible changes to the document.
1924 * layout-tests/editing/editing.js: Added some glue to change text align.
1925 * layout-tests/editing/style/block-style-001-expected.txt: Added.
1926 * layout-tests/editing/style/block-style-001.html: Added.
1927 * layout-tests/editing/style/block-style-002-expected.txt: Added.
1928 * layout-tests/editing/style/block-style-002.html: Added.
1929 * layout-tests/editing/style/block-style-003-expected.txt: Added.
1930 * layout-tests/editing/style/block-style-003.html: Added.
1934 2004-10-29 Darin Adler <darin@apple.com>
1938 - fixed <rdar://problem/3751619> Safari crash in khtml::CircularSearchBuffer::append(QChar const&)
1940 * khtml/editing/visible_text.cpp: (khtml::findPlainText): Fix exit condition to check for break
1941 before advancing one character; before it did it backwards.
1943 2004-10-29 Chris Blumenberg <cblu@apple.com>
1945 Fixed: <rdar://problem/3853262> REGRESSION(166-168) gmail gets blank page when loading
1947 Reviewed by kocienda, adele.
1949 * khtml/rendering/render_frames.cpp:
1950 (RenderPartObject::updateWidget): remove infinite frame recursion check for iframes
1952 2004-10-29 Darin Adler <darin@apple.com>
1956 - fixed <rdar://problem/3857395> clicking on calendar in Apple Travel site crashes Safari in invalidateClick (getthere.net)
1958 * khtml/khtmlview.cpp:
1959 (KHTMLView::viewportMousePressEvent): Use a SharedPtr<KHTMLView> to make sure the KHTMLView is not
1960 deleted before this function finishes running.
1961 (KHTMLView::viewportMouseDoubleClickEvent): Ditto.
1962 (KHTMLView::viewportMouseReleaseEvent): Ditto.
1963 (KHTMLView::dispatchMouseEvent): Removed ref/deref pairs that aren't needed since dispatchEvent
1964 is guaranteed to do ref/deref as needed.
1966 * kwq/KWQObject.mm: (QObject::startTimer): Fixed a comment.
1968 2004-10-28 Chris Blumenberg <cblu@apple.com>
1970 Enabled XSLT on Panther. See intrigue mail for compiling instructions.
1974 * WebCore.pbproj/project.pbxproj: link against xslt unconditionally, link against specific version of libxml on Panther
1975 * WebCorePrefix.h: always use XSLT
1977 2004-10-28 Ken Kocienda <kocienda@apple.com>
1983 <rdar://problem/3854848> Tiger Mail Crash in WebCore - khtml::CompositeEditCommand::insertNodeAfter
1984 <rdar://problem/3803832> REGRESSION (Mail): incorrect behavior after Return + Delete in quoted text
1986 * khtml/editing/htmlediting.cpp:
1987 (khtml::DeleteSelectionCommand::DeleteSelectionCommand): Added node pointer class members
1988 to initialization list, zeroing them out.
1989 (khtml::DeleteSelectionCommand::canPerformSpecialCaseBRDelete): New special-case helper to
1990 handle a delete of content in special cases where the only thing selected is a BR. This
1991 code path is much simpler than the newly-named performGeneralDelete, and detects when no
1992 content merging should be done between blocks. This aspect of the change fixes 3854848.
1993 One of the special cases added fixes 3803832.
1994 (khtml::DeleteSelectionCommand::performGeneralDelete): Renamed, from performDelete.
1995 (khtml::DeleteSelectionCommand::moveNodesAfterNode): Made this helper be a no-arg function, just
1996 like the other helpers in this class.
1997 (khtml::DeleteSelectionCommand::clearTransientState): Fix cut and paste error in deref code.
1998 (khtml::DeleteSelectionCommand::doApply): Updated for changed helpers.
1999 * khtml/editing/htmlediting.h: Added new helper and changed an old one.
2001 2004-10-28 Chris Blumenberg <cblu@apple.com>
2003 Fixed: <rdar://problem/3856913> Panther-only crash in QString code copying front page of store.apple.com
2007 * kwq/KWQKHTMLPart.mm:
2008 (KWQKHTMLPart::attributedString): check that the renderer is a list item before making list item calls on it
2010 2004-10-28 Ken Kocienda <kocienda@apple.com>
2012 Reviewed by Harrison
2014 Reorganization of delete command functionality so that doApply is not
2015 several hundred lines long. This is not a squeaky-clean cleanup, but
2016 it is a step in the right direction. No functionality changes.
2018 * khtml/editing/htmlediting.cpp:
2019 (khtml::DeleteSelectionCommand::DeleteSelectionCommand):
2020 (khtml::DeleteSelectionCommand::initializePositionData): New helper.
2021 (khtml::DeleteSelectionCommand::saveTypingStyleState): Ditto.
2022 (khtml::DeleteSelectionCommand::performDelete): Ditto.
2023 (khtml::DeleteSelectionCommand::fixupWhitespace): Ditto.
2024 (khtml::DeleteSelectionCommand::moveNodesAfterNode): Ditto.
2025 (khtml::DeleteSelectionCommand::calculateEndingPosition): Ditto.
2026 (khtml::DeleteSelectionCommand::calculateTypingStyleAfterDelete): Ditto.
2027 (khtml::DeleteSelectionCommand::clearTransientState): Ditto.
2028 (khtml::DeleteSelectionCommand::doApply): Factor out code into new helpers.
2029 * khtml/editing/htmlediting.h:
2031 2004-10-28 Ken Kocienda <kocienda@apple.com>
2035 * khtml/editing/htmlediting.cpp:
2036 (khtml::DeleteSelectionCommand::DeleteSelectionCommand): Typo in initializer caused
2037 new mergeBlocksAfterDelete flag to be set improperly, causing layout regressions.
2039 2004-10-27 Ken Kocienda <kocienda@apple.com>
2043 * khtml/editing/htmlediting.cpp:
2044 (khtml::CompositeEditCommand::deleteSelection): Added new mergeBlocksAfterDelete flag to control
2045 whether content not in the block containing the start of the selection is moved to that block
2046 after the selection is deleted.
2047 (khtml::DeleteSelectionCommand::DeleteSelectionCommand): Ditto.
2048 (khtml::DeleteSelectionCommand::doApply): Ditto.
2049 (khtml::InputNewlineInQuotedContentCommand::InputNewlineInQuotedContentCommand): New command
2050 to handle the case of inserting a newline when in quoted content in Mail.
2051 (khtml::InputNewlineInQuotedContentCommand::~InputNewlineInQuotedContentCommand): Ditto.
2052 (khtml::InputNewlineInQuotedContentCommand::isMailBlockquote): Ditto.
2053 (khtml::InputNewlineInQuotedContentCommand::isLastVisiblePositionInBlockquote): Ditto.
2054 (khtml::InputNewlineInQuotedContentCommand::doApply): Ditto.
2055 (khtml::TypingCommand::insertNewlineInQuotedContent): Support for new newline command.
2056 (khtml::TypingCommand::doApply): Ditto.
2057 (khtml::TypingCommand::preservesTypingStyle): Ditto.
2058 * khtml/editing/htmlediting.h: Add new delclarations.
2059 (khtml::TypingCommand::): Ditto.
2060 * kwq/WebCoreBridge.h: Added new bridge method called from WebKit.
2061 * kwq/WebCoreBridge.mm:
2062 (-[WebCoreBridge insertNewlineInQuotedContent]): Ditto.
2064 2004-10-26 Chris Blumenberg <cblu@apple.com>
2066 Fixed: <rdar://problem/3774243> page up/down, arrow up/down, etc in Safari RSS should scroll main content
2070 * khtml/ecma/kjs_dom.cpp:
2071 (DOMElementProtoFunc::tryCall): added scrollByLines and scrollByPages to HTML element for Safari RSS
2072 * khtml/ecma/kjs_dom.h:
2073 (KJS::DOMElement::):
2074 * khtml/ecma/kjs_dom.lut.h:
2077 2004-10-26 David Hyatt <hyatt@apple.com>
2079 Fix for 3848214, deleting a partial word left a repaint artifact if the partial word was pulled back onto
2082 Reviewed by kocienda
2084 * khtml/rendering/bidi.cpp:
2085 (khtml::RenderBlock::layoutInlineChildren):
2087 2004-10-26 David Hyatt <hyatt@apple.com>
2089 Convert selectionRect() from using a list to a dict and patch it to be like setSelection. It was still trying
2090 to use the old dirty bit optimization (which had been removed), and so it was pathologically slow on large documents.
2092 Reviewed by kocienda
2094 * khtml/rendering/render_canvas.cpp:
2095 (RenderCanvas::selectionRect):
2096 * khtml/rendering/render_object.h:
2097 (khtml::RenderObject::hasSelectedChildren):
2099 2004-10-26 Ken Kocienda <kocienda@apple.com>
2105 <rdar://problem/3851164> mail crashed when I pasted a large amount of text into a reply
2107 * khtml/editing/htmlediting.cpp:
2108 (khtml::CompositeEditCommand::removeBlockPlaceholderIfNeeded): This now returns bool to
2109 let the caller know if a placeholder was removed.
2110 (khtml::ReplaceSelectionCommand::doApply): Use the bool return value from the call to
2111 removeBlockPlaceholderIfNeeded. If true, shift the selection to the now-empty block. In
2112 some cases, the selection was still set on the removed BR, and this was the cause of the
2114 * khtml/editing/htmlediting.h: Change removeBlockPlaceholderIfNeeded return type.
2116 2004-10-26 Darin Adler <darin@apple.com>
2120 - fixed <rdar://problem/3851301> leak of one NSCFDictionary for each XMLHttpRequest issued
2122 * kwq/KWQLoader.mm: (KWQServeSynchronousRequest): Add a release.
2124 2004-10-26 Ken Kocienda <kocienda@apple.com>
2128 * khtml/editing/htmlediting.cpp:
2129 (khtml::CompositeEditCommand::deleteInsignificantText): Do not call replaceText
2130 with a zero-length string. That triggers an assert. Call deleteText instead,
2131 using the same indices that are passed to replaceText.
2133 Cleaned up the asserts in these three functions below, making them
2134 more consistent. This is not needed for the fix, but I tripped over
2135 these in the course of debugging.
2137 (khtml::InsertTextCommand::InsertTextCommand):
2138 (khtml::InsertTextCommand::doApply):
2139 (khtml::InsertTextCommand::doUnapply):
2141 2004-10-25 Adele Amchan <adele@apple.com>
2145 * khtml/xml/dom_docimpl.cpp: (DocumentImpl::inDesignMode): made inDesignMode const
2146 * khtml/xml/dom_docimpl.h:
2148 2004-10-25 Adele Amchan <adele@apple.com>
2150 Reviewed by me, code change by Darin.
2152 * khtml/xml/dom_docimpl.cpp: Moved design mode code outside the XSLT ifdef.
2154 2004-10-25 Ken Kocienda <kocienda@apple.com>
2156 Oops. These two test results changed with my last checkin, for the better.
2158 * layout-tests/editing/deleting/delete-3800834-fix-expected.txt
2159 * layout-tests/editing/inserting/insert-3775316-fix-expected.txt
2161 2004-10-25 Ken Kocienda <kocienda@apple.com>
2167 <rdar://problem/3820349> REGRESSION (Mail): select all, delete does not always delete everything
2169 * khtml/editing/htmlediting.cpp:
2170 (khtml::DeleteSelectionCommand::startPositionForDelete): New helper that determines when to
2171 expand the selection outwards when the selection is on the visible boundary of a root
2172 editable element. This fixes the bug. Note that this function also contains a little code
2173 I factored out of doApply: it also takes care of adjusting the selection in the smart delete case.
2174 (khtml::DeleteSelectionCommand::endPositionForDelete): Ditto.
2175 (khtml::DeleteSelectionCommand::doApply): Call new helpers. Refactored out the code as described.
2176 * khtml/editing/htmlediting.h: Declare new helpers.
2177 * layout-tests/editing/deleting/delete-select-all-001-expected.txt: Added.
2178 * layout-tests/editing/deleting/delete-select-all-001.html: Added.
2179 * layout-tests/editing/deleting/delete-select-all-002-expected.txt: Added.
2180 * layout-tests/editing/deleting/delete-select-all-002.html: Added.
2181 * layout-tests/editing/deleting/delete-select-all-003-expected.txt: Added.
2182 * layout-tests/editing/deleting/delete-select-all-003.html: Added.
2184 2004-10-25 Ken Kocienda <kocienda@apple.com>
2186 Added some more editing layout tests.
2188 * layout-tests/editing/deleting/delete-ws-fixup-001-expected.txt: Added.
2189 * layout-tests/editing/deleting/delete-ws-fixup-001.html: Added.
2190 * layout-tests/editing/deleting/delete-ws-fixup-002-expected.txt: Added.
2191 * layout-tests/editing/deleting/delete-ws-fixup-002.html: Added.
2192 * layout-tests/editing/deleting/delete-ws-fixup-003-expected.txt: Added.
2193 * layout-tests/editing/deleting/delete-ws-fixup-003.html: Added.
2194 * layout-tests/editing/deleting/delete-ws-fixup-004-expected.txt: Added.
2195 * layout-tests/editing/deleting/delete-ws-fixup-004.html: Added.
2196 * layout-tests/editing/inserting/typing-003-expected.txt: Added.
2197 * layout-tests/editing/inserting/typing-003.html: Added.
2199 2004-10-25 Ken Kocienda <kocienda@apple.com>
2203 * khtml/rendering/bidi.cpp:
2204 (khtml::RenderBlock::findNextLineBreak): I did not get my fix for 3848343 and 3848224
2205 yesterday quite right: words that should have been placed on the next line were instead
2206 appearing on the line before, beyond the right margin. This was a one-word only error
2207 based on moving the line break object when it should have stayed put. Here is the rule:
2208 The line break object only moves to after the whitespace on the end of a line if that
2209 whitespace caused line overflow when its width is added in.
2211 2004-10-25 Adele Amchan <adele@apple.com>
2215 Fix for <rdar://problem/3619890> Feature request: designMode
2217 This change implements the designMode property of a document. This is an IE property that is also supported by Mozilla.
2218 This will enable more JS editing compatibility.
2220 * khtml/ecma/kjs_html.cpp:
2221 (KJS::HTMLDocument::tryGet): added case for designMode
2222 (KJS::HTMLDocument::putValue): added case for designMode
2223 * khtml/ecma/kjs_html.lut.h: (KJS::): regenerated
2224 * khtml/khtml_part.cpp: (KHTMLPart::isContentEditable): Now returns designMode value
2225 * khtml/xml/dom_docimpl.cpp:
2226 (DocumentImpl::DocumentImpl): initialize m_designMode member variable
2227 (DocumentImpl::setDesignMode): added function to assign m_designMode value
2228 (DocumentImpl::getDesignMode): return m_designMode value
2229 (DocumentImpl::inDesignMode): if designMode is inherited, this will find the appropriate parent document designMode and return that value.
2230 Otherwise, it will just return the m_designMode value.
2231 (DocumentImpl::parentDocument):
2232 * khtml/xml/dom_docimpl.h: (DOM::DocumentImpl::): added InheritedBool enum, prototypes, and m_designMode member variable.
2233 * kwq/KWQKHTMLPart.mm: (KWQKHTMLPart::isContentEditable): added check for isContentEditable function in KHTMLPart
2235 2004-10-22 Ken Kocienda <kocienda@apple.com>
2241 <rdar://problem/3844662> REGRESSION (Mail): Style changes can affect adjacent, unselected text
2243 * khtml/editing/htmlediting.cpp:
2244 (khtml::ApplyStyleCommand::doApply): Move end position downstream to be sure we remove style from
2245 everything that could be affected.
2246 (khtml::ApplyStyleCommand::removeCSSStyle): Comma in intended function call was outside the
2247 braces, making it act as a comma operator, with a zero value as the right value!!! This made
2248 an important check always fail!!! It turns out that we do not want the constant at all, since
2249 that constant is only needed when checking a computed style, not an inline style as is being
2251 (khtml::ApplyStyleCommand::removeStyle): Call nodeFullySelected with new interface.
2252 (khtml::ApplyStyleCommand::nodeFullySelected): Change interface and implementation to rely on
2253 RangeImpl::compareBoundaryPoints to perform the required check.
2254 * khtml/editing/htmlediting.h: Changed nodeFullySelected function interface.
2256 2004-10-22 Ken Kocienda <kocienda@apple.com>
2262 <rdar://problem/3848343> REGRESSION (Mail, 166-168u): Typed text after space at end of line before block quote does not appear
2263 <rdar://problem/3848224> REGRESSION (Mail): space typed at end of line vanishes after typing next character
2265 * khtml/rendering/bidi.cpp:
2266 (khtml::RenderBlock::findNextLineBreak): When the khtmlLineBreak is in AFTER_WHITE_SPACE mode, as
2267 it is when we are editing, add in the space of the current character when calculating the width
2268 of committed plus uncommitted characters. If this value exceeds the width of the line, move up
2269 the line break object and call skipWhitespace to move past the end of the whitespace.
2273 2004-10-22 Ken Kocienda <kocienda@apple.com>
2275 * WebCore.pbproj/project.pbxproj:
2276 Add GCC_ENABLE_OBJC_GC and GCC_FAST_OBJC_DISPATCH flags.
2278 2004-10-21 David Hyatt <hyatt@apple.com>
2282 Clean up the inline run function so that it doesn't return incorrect answers when making children non-inline.
2284 <rdar://problem/3848724> REGRESSION (166-168u): RenderText::layout called, firing assertion that kills Mail
2285 <rdar://problem/3848357> RenderText::layout called, firing assertion that kills Safari (www.apple.com/downloads/macosx)
2287 * khtml/rendering/render_block.cpp:
2288 (khtml::getInlineRun):
2289 (khtml::RenderBlock::makeChildrenNonInline):
2291 2004-10-21 David Hyatt <hyatt@apple.com>
2293 Fix for 3810389, crash because of continuation() craziness. Revert back to the old behavior of
2294 making sure that all line boxes get deleted and recreated when inlines are split because of a block.
2298 * khtml/rendering/render_inline.cpp:
2299 (RenderInline::splitFlow):
2301 2004-10-21 Ken Kocienda <kocienda@apple.com>
2305 Significant improvement to the way that whitespace is handled during editing.
2307 * khtml/editing/htmlediting.cpp:
2308 (khtml::CompositeEditCommand::deleteInsignificantText): New functions (there are actually
2309 two being added with this name) that delete "insignificant" unrendered text.
2310 (khtml::CompositeEditCommand::deleteInsignificantTextDownstream): Takes a position,
2311 calculates the downstream position to use as the endpoint for the deletion, and
2312 then calls deleteInsignificantText with this start and end.
2313 (khtml::DeleteSelectionCommand::doApply): Call new deleteInsignificantTextDownstream function.
2314 (khtml::InputNewlineCommand::doApply): Ditto.
2315 (khtml::InputTextCommand::input): Ditto.
2316 * khtml/editing/htmlediting.h: Add new declarations.
2318 Modified layout test results:
2319 * layout-tests/editing/deleting/delete-block-merge-contents-016-expected.txt:
2320 * layout-tests/editing/deleting/delete-block-merge-contents-017-expected.txt:
2321 * layout-tests/editing/deleting/delete-contiguous-ws-001-expected.txt:
2322 * layout-tests/editing/deleting/delete-selection-001-expected.txt:
2323 * layout-tests/editing/deleting/delete-tab-001-expected.txt:
2324 * layout-tests/editing/deleting/delete-tab-004-expected.txt:
2325 * layout-tests/editing/deleting/delete-trailing-ws-001-expected.txt:
2326 * layout-tests/editing/inserting/insert-3659587-fix-expected.txt:
2327 * layout-tests/editing/inserting/insert-3775316-fix-expected.txt:
2328 * layout-tests/editing/inserting/insert-3778059-fix-expected.txt:
2329 * layout-tests/editing/inserting/insert-br-001-expected.txt:
2330 * layout-tests/editing/inserting/insert-br-004-expected.txt:
2331 * layout-tests/editing/inserting/insert-br-005-expected.txt:
2332 * layout-tests/editing/inserting/insert-br-006-expected.txt:
2333 * layout-tests/editing/inserting/insert-tab-001-expected.txt:
2334 * layout-tests/editing/inserting/insert-tab-002-expected.txt:
2335 * layout-tests/editing/inserting/insert-tab-004-expected.txt:
2336 * layout-tests/editing/inserting/insert-text-with-newlines-expected.txt:
2337 * layout-tests/editing/inserting/typing-001-expected.txt:
2338 * layout-tests/editing/inserting/typing-around-br-001-expected.txt:
2339 * layout-tests/editing/inserting/typing-around-image-001-expected.txt:
2340 * layout-tests/editing/style/typing-style-003-expected.txt:
2341 * layout-tests/editing/undo/redo-typing-001-expected.txt:
2342 * layout-tests/editing/undo/undo-typing-001-expected.txt:
2344 2004-10-21 David Hyatt <hyatt@apple.com>
2346 Fix for 3847054, assertion failure in RenderText::layout() on news.com page. Fix getInlineRun so that
2347 it no longer breaks early (thus causing some children not to get properly wrapped by anonymous blocks).
2351 * khtml/rendering/render_block.cpp:
2352 (khtml::getInlineRun):
2354 2004-10-20 David Hyatt <hyatt@apple.com>
2356 Add better dumping of overflow information for scrolling regions.
2358 Fix for 3726524, crash in updateLayerPosition. Make sure anonymous elements properly remove themselves
2359 from the render tree so that layers and so forth are cleaned up.
2363 * khtml/rendering/render_container.cpp:
2364 (RenderContainer::detach):
2365 * khtml/rendering/render_layer.h:
2366 (khtml::RenderLayer::scrollXOffset):
2367 (khtml::RenderLayer::scrollYOffset):
2368 * kwq/KWQRenderTreeDebug.cpp:
2371 2004-10-20 David Hyatt <hyatt@apple.com>
2373 Fix for 3791146, make sure all lines are checked when computing overflow.
2375 Reviewed by kocienda
2377 * khtml/rendering/bidi.cpp:
2378 (khtml::RenderBlock::computeHorizontalPositionsForLine):
2379 (khtml::RenderBlock::layoutInlineChildren):
2380 (khtml::RenderBlock::findNextLineBreak):
2381 (khtml::RenderBlock::checkLinesForOverflow):
2382 * khtml/rendering/render_block.h:
2384 2004-10-20 David Hyatt <hyatt@apple.com>
2386 Fix for 3790936, make the unicode-breaking on the layout tests match Panther.
2388 Reviewed by kocienda
2390 * khtml/rendering/break_lines.cpp:
2391 (khtml::isBreakable):
2393 2004-10-20 Darin Adler <darin@apple.com>
2397 - fixed <rdar://problem/3317107> text input fields and text areas don't respect background color and text color CSS properties
2399 * khtml/rendering/render_form.cpp: (RenderFormElement::updateFromElement):
2400 Create a palette with the background and foreground colors in it and set it on the widget.
2402 * khtml/rendering/render_style.h: (khtml::StyleVisualData::operator==): No palette to compare
2403 with APPLE_CHANGES. Removed palette and palette-related function members.
2404 * khtml/rendering/render_style.cpp:
2405 (StyleVisualData::StyleVisualData): No palette to initialize with APPLE_CHANGES.
2406 (RenderStyle::diff): No palette to compare.
2408 * kwq/KWQLineEdit.h: Added setPalette override. Made text function const.
2409 * kwq/KWQLineEdit.mm:
2410 (QLineEdit::setPalette): Added. Sets foreground and background color based on palette.
2411 (QLineEdit::text): Made const.
2413 * kwq/KWQTextEdit.h: Added setPalette override.
2414 * kwq/KWQTextEdit.mm: (QTextEdit::setPalette): Added. Sets foreground and background color
2417 * kwq/KWQPalette.h: Removed most things, leaving only background and foreground colors
2418 per color group, and only a single color group per palette.
2419 * kwq/KWQColorGroup.mm: Removed.
2420 * kwq/KWQPalette.mm: Removed.
2421 * WebCore.pbproj/project.pbxproj: Removed KWQColorGroup.mm and KWQPalette.mm.
2423 * kwq/KWQApplication.h: Removed unused palette function.
2424 * kwq/KWQApplication.mm: Ditto.
2426 * kwq/KWQWidget.h: Removed unsetPalette.
2427 * kwq/KWQWidget.mm: Ditto.
2429 - fixed storage leak
2431 * khtml/html/html_formimpl.cpp: (DOM::HTMLGenericFormElementImpl::~HTMLGenericFormElementImpl):
2432 Roll in storage leak fix from KDE guys.
2434 2004-10-19 David Hyatt <hyatt@apple.com>
2436 Reviewed by kocienda
2438 More cleanup of block layout. Eliminates the separate step for tables that dont fit on a line with a float
2439 and consolidates it with clearing.
2441 Also patch dom_textimpl.cpp to reduce further the # of RenderTexts created.
2443 * khtml/rendering/render_block.cpp:
2444 (khtml::getInlineRun):
2445 (khtml::RenderBlock::layoutBlock):
2446 (khtml::RenderBlock::adjustFloatingBlock):
2447 (khtml::RenderBlock::collapseMargins):
2448 (khtml::RenderBlock::clearFloatsIfNeeded):
2449 (khtml::RenderBlock::estimateVerticalPosition):
2450 (khtml::RenderBlock::layoutBlockChildren):
2451 (khtml::RenderBlock::markAllDescendantsWithFloatsForLayout):
2452 (khtml::RenderBlock::getClearDelta):
2453 (khtml::RenderBlock::calcBlockMinMaxWidth):
2454 * khtml/rendering/render_block.h:
2455 * khtml/rendering/render_frames.cpp:
2456 (RenderFrameSet::layout):
2457 * khtml/xml/dom_textimpl.cpp:
2458 (TextImpl::rendererIsNeeded):
2460 Fix for 3841060, regression with * in frames. Reviewed by kocienda.
2462 * layout-tests/fast/frames/002-expected.txt: Added.
2463 * layout-tests/fast/frames/002.html: Added.
2465 2004-10-19 Darin Adler <darin@apple.com>
2469 - follow-on to my fix yesterday, which broke a layout test because I rolled out a fix that Maciej had done
2471 * khtml/html/htmltokenizer.cpp: (khtml::HTMLTokenizer::write): Need to check the actual queue of external
2472 scripts being loaded here. If the current code being run is the external script itself, then we don't want
2473 to defer parsing. But loadingExtScript has to stay true until after the script runs. The old code would
2474 assume that any time we're running a script there's no need to look at loadingExtScript, but that was also
2475 wrong since there can be a script loading in that case too. Layout tests check for both problems.
2477 * layout-tests/fast/tokenizer/external-script-document-write-expected.txt: Added.
2478 * layout-tests/fast/tokenizer/external-script-document-write.html: Added.
2479 * layout-tests/fast/tokenizer/resources/external-script-document-write.js: Added.
2481 * layout-tests/fast/tokenizer/004.html: Fixed line endings (were CR, should be LF).
2483 2004-10-18 Darin Adler <darin@apple.com>
2485 Reviewed by Dave Hyatt.
2487 - fixed <rdar://problem/3807234> REGRESSION (152-153): can't get element by ID that was just written with document.write in separate JS file (lacoccinelle.net)
2489 * khtml/html/htmltokenizer.cpp:
2490 (khtml::HTMLTokenizer::scriptHandler): Use !isEmpty instead of count != 0, since it's cheaper.
2491 (khtml::HTMLTokenizer::write): Roll back to the old version of the check here. The fix is that
2492 we only want to look at loadingExtScript if m_executingScript is 0.
2494 2004-10-18 Ken Kocienda <kocienda@apple.com>
2500 <rdar://problem/3840907> textedit doesn't render italic or bold text in html documents
2502 * khtml/css/css_valueimpl.cpp:
2503 (CSSPrimitiveValueImpl::getStringValue): This function did not return string values for idents.
2504 Also changed the return value to be DOMString, rather than DOMStringImpl, to deal with the
2505 lifecycle issues associated with creating a string to be returned in the ident case.
2506 * khtml/css/css_valueimpl.h: Change getStringValue to return DOMString rather than DOMStringImpl.
2507 * khtml/css/cssstyleselector.cpp:
2508 (khtml::CSSStyleSelector::applyProperty): Two calls of getStringValue needed updating.
2510 2004-10-18 Chris Blumenberg <cblu@apple.com>
2512 Fixed: <rdar://problem/3770135> hang loading page with EMBED tag pointing to same page (tridentantennas.co.uk)
2514 Reviewed by kocienda.
2516 * khtml/rendering/render_frames.cpp:
2517 (RenderPartObject::updateWidget): use completeURL before comparing the URL of the plug-in with the base URL of the document when avoiding frame recursion
2519 2004-10-15 Chris Blumenberg <cblu@apple.com>
2521 Fixed: <rdar://problem/3841774> would like to get NSColor from DOM-CSS
2526 (-[DOMRGBColor _color]): new, returns getNSColor on KWQColor
2529 2004-10-15 Ken Kocienda <kocienda@apple.com>
2533 * khtml/rendering/bidi.cpp:
2534 (khtml::RenderBlock::skipNonBreakingSpace): Also need to forego the
2535 skipping after a clean line break, in addition to the cases already
2537 * layout-tests/editing/inserting/insert-br-007-expected.txt: Added.
2538 * layout-tests/editing/inserting/insert-br-007.html: Added.
2539 * layout-tests/editing/inserting/insert-br-008-expected.txt: Added.
2540 * layout-tests/editing/inserting/insert-br-008.html: Added.
2544 2004-10-14 Ken Kocienda <kocienda@apple.com>
2550 <rdar://problem/3839989> REGRESSION (Mail): Left arrow does nothing after inserting attachment
2552 * khtml/editing/visible_position.cpp:
2553 (khtml::VisiblePosition::deepEquivalent): Remove code that attempted to bridge old-style
2554 position code to new-style VisiblePosition code. In retrospect, this code was misguided.
2555 Since we do a good job of insulating external code from the internal workings of
2556 VisiblePosition, the "hop ahead" being done here was not doing anyone any real good, and
2557 in the case of this bug, was doing harm. Simply removing this code makes the bug
2558 go away and does not cause any editing layout test regresssions.
2560 2004-10-14 Ken Kocienda <kocienda@apple.com>
2564 * khtml/rendering/bidi.cpp:
2565 (khtml::skipNonBreakingSpace): New helper.
2566 (khtml::RenderBlock::skipWhitespace): Do not skip non-breaking spaces that are
2567 at the start of a block. This was preventing users from typing spaces in empty
2569 * layout-tests/editing/inserting/insert-space-in-empty-doc-expected.txt: Added.
2570 * layout-tests/editing/inserting/insert-space-in-empty-doc.html: Added.
2572 2004-10-14 Adele Amchan <adele@apple.com>
2574 Reviewed by Darin and Ken.
2576 fix for <rdar://problem/3821070> null de-ref in DelectSelectionCommand::doApply()
2578 This change shifts some code around so that the code that determines what typing style
2579 is in effect is called before deleteUnrenderedText is called. Two asserts are also added
2580 to ensure that start and end nodes of the selection are in the document.
2582 * khtml/editing/htmlediting.cpp: (khtml::DeleteSelectionCommand::doApply):
2584 2004-10-14 Adele Amchan <adele@apple.com>
2588 This change makes these three functions virtual so that the work is being done in KWQHTMLPart
2589 instead of khtml_part, eliminating the need for the "#if APPLE_CHANGES" statements in the khtml code.
2591 * khtml/khtml_part.cpp:
2592 (KHTMLPart::shouldBeginEditing):
2593 (KHTMLPart::shouldEndEditing):
2594 (KHTMLPart::isContentEditable):
2595 * khtml/khtml_part.h:
2596 * kwq/KWQKHTMLPart.h:
2598 2004-10-14 Ken Kocienda <kocienda@apple.com>
2602 Final fix for these bugs:
2604 <rdar://problem/3806306> HTML editing puts spaces at start of line
2605 <rdar://problem/3814252> HTML editing groups space with word causing wrapping
2607 This change sets some new CSS properties that have been added to WebCore to
2608 enable whitespace-handling and line-breaking features that make WebView work
2609 more like a text editor.
2611 * khtml/css/cssstyleselector.cpp:
2612 (khtml::CSSStyleSelector::applyProperty): Add and remove special editing CSS properties
2613 based on property value.
2614 * khtml/html/html_elementimpl.cpp:
2615 (HTMLElementImpl::setContentEditable): Add and remove special editing CSS properties
2616 based on attribute value.
2617 * khtml/khtml_part.cpp:
2618 (KHTMLPart::applyEditingStyleToBodyElement): New helper. Calls applyEditingStyleToElement on
2620 (KHTMLPart::removeEditingStyleFromBodyElement): New helper. Calls removeEditingStyleFromElement on
2622 (KHTMLPart::applyEditingStyleToElement): Adds special editing CSS properties to passed in element.
2623 (KHTMLPart::removeEditingStyleFromElement): Removes special editing CSS properties from passed in element.
2624 * khtml/khtml_part.h: Add new declarations.
2625 * kwq/WebCoreBridge.h: Ditto.
2626 * kwq/WebCoreBridge.mm:
2627 (-[WebCoreBridge applyEditingStyleToBodyElement]): Call through to similarly-named function on KHTMLPart.
2628 (-[WebCoreBridge removeEditingStyleFromBodyElement]): Ditto.
2629 (-[WebCoreBridge applyEditingStyleToElement:]): Ditto.
2630 (-[WebCoreBridge removeEditingStyleFromElement:]): Ditto.
2632 2004-10-14 John Sullivan <sullivan@apple.com>
2636 - fixed <rdar://problem/3840052> Crash in removeBlockPlaceholderIfNeeded attaching file to empty document
2638 * khtml/editing/htmlediting.cpp:
2639 (khtml::CompositeEditCommand::removeBlockPlaceholderIfNeeded):
2640 needed a nil check to handle empty document case
2642 2004-10-13 Maciej Stachowiak <mjs@apple.com>
2646 <rdar://problem/3824626> Change to do colormatching for DeviceRGB colorspace causes ~11% Safari slowdown
2648 - I fixed this by turning off all colormatching for WebKit
2649 content. We might turn it back on later. For now, it's possible to
2650 turn it on temporarily by defining COLORMATCH_EVERYTHING.
2653 * khtml/ecma/kjs_html.cpp:
2654 (KJS::Context2DFunction::tryCall):
2655 (Context2D::colorRefFromValue):
2656 (Gradient::getShading):
2657 * khtml/rendering/render_canvasimage.cpp:
2658 (RenderCanvasImage::createDrawingContext):
2660 (QColor::getNSColor):
2662 * kwq/KWQPainter.mm:
2663 (CGColorFromNSColor):
2664 (QPainter::selectedTextBackgroundColor):
2665 (QPainter::rgbColorSpace):
2666 (QPainter::grayColorSpace):
2667 (QPainter::cmykColorSpace):
2668 * kwq/WebCoreGraphicsBridge.h:
2669 * kwq/WebCoreGraphicsBridge.m:
2670 (-[WebCoreGraphicsBridge createRGBColorSpace]):
2671 (-[WebCoreGraphicsBridge createGrayColorSpace]):
2672 (-[WebCoreGraphicsBridge createCMYKColorSpace]):
2674 2004-10-13 Ken Kocienda <kocienda@apple.com>
2678 * khtml/css/css_valueimpl.cpp:
2679 (CSSStyleDeclarationImpl::merge): A little cleanup. Also, make sure m_lstValues
2680 is non-null before appending.
2682 2004-10-13 Ken Kocienda <kocienda@apple.com>
2684 Update expected results for improved behavior as a result of fix to 3816768.
2686 * layout-tests/editing/deleting/delete-3775172-fix-expected.txt
2687 * layout-tests/editing/deleting/delete-3800834-fix-expected.txt
2688 * layout-tests/editing/deleting/delete-block-merge-contents-002-expected.txt
2690 2004-10-13 Ken Kocienda <kocienda@apple.com>
2694 * khtml/css/css_computedstyle.cpp:
2695 (DOM::CSSComputedStyleDeclarationImpl::getPropertyCSSValue): Add support to computed style
2696 for getting -khtml-line-break and -khml-nbsp-mode.
2698 2004-10-13 Ken Kocienda <kocienda@apple.com>
2704 <rdar://problem/3816768> REGRESSION (Mail): Deleting last character in block incorrectly
2705 moves caret out of block.
2707 The issue here is that an empty block with no explicit height set by style collapses
2708 to zero height, and does so immediately after the last bit of content is removed from
2709 it (as a result of deleting text with the delete key for instance). Since zero-height
2710 blocks are not eligible caret positions, the caret jumped to the closest eligible spot.
2712 The fix is to detect when a block has not been removed itself, but has had all its
2713 contents removed. In this case, a BR element is placed in the block, one that is
2714 specially marked as a placeholder. Later, if the block ever receives content, this
2715 placeholder is removed.
2717 * khtml/editing/htmlediting.cpp:
2718 (khtml::blockPlaceholerClassString): String which acts as a placeholder marker class.
2719 (khtml::CompositeEditCommand::insertBlockPlaceholderIfNeeded): Adds a placeholder BR if needed.
2720 (khtml::CompositeEditCommand::removeBlockPlaceholderIfNeeded): Removes a placeholder BR if needed.
2721 (khtml::DeleteSelectionCommand::moveNodesAfterNode): Call removeBlockPlaceholderIfNeeded.
2722 Also, do some cleanup on some old, crufty code in the move logic that is just so clearly wrong
2723 (it's very clear that we needs to be able to move more than just text nodes). This may expose
2724 bugs, but these bugs needs to be filed and fixed, not ducked. Besides, undoing this silliness
2725 made the test case in the bug work.
2726 (khtml::DeleteSelectionCommand::doApply): Call insertBlockPlaceholderIfNeeded and
2727 removeBlockPlaceholderIfNeeded.
2728 (khtml::InputTextCommand::input): Call removeBlockPlaceholderIfNeeded.
2729 (khtml::ReplaceSelectionCommand::doApply): Call removeBlockPlaceholderIfNeeded.
2730 * khtml/editing/htmlediting.h: Declare new functions.
2732 2004-10-13 Richard Williamson <rjw@apple.com>
2734 Added support for -apple-dashboard-region:none. And fixed
2735 a few computed style problems.
2737 Fixed <rdar://problem/3833532> -apple-dashboard-region: none; is needed
2740 * khtml/css/css_computedstyle.cpp:
2741 (DOM::CSSComputedStyleDeclarationImpl::getPropertyCSSValue):
2742 * khtml/css/css_valueimpl.cpp:
2743 (CSSPrimitiveValueImpl::cssText):
2744 * khtml/css/cssparser.cpp:
2745 (CSSParser::parseValue):
2746 (CSSParser::parseDashboardRegions):
2747 * khtml/css/cssstyleselector.cpp:
2748 (khtml::CSSStyleSelector::applyProperty):
2749 * khtml/rendering/render_style.cpp:
2750 (RenderStyle::noneDashboardRegions):
2751 * khtml/rendering/render_style.h:
2752 * kwq/KWQKHTMLPart.mm:
2753 (KWQKHTMLPart::dashboardRegionsDictionary):
2755 2004-10-13 David Hyatt <hyatt@apple.com>
2757 Rework block layout to clean it up and simplify it (r=kocienda).
2759 Also fixing the style sharing bug (r=mjs).
2761 * khtml/rendering/render_block.cpp:
2762 (khtml::RenderBlock::MarginInfo::MarginInfo):
2763 (khtml::RenderBlock::layoutBlock):
2764 (khtml::RenderBlock::adjustPositionedBlock):
2765 (khtml::RenderBlock::adjustFloatingBlock):
2766 (khtml::RenderBlock::handleSpecialChild):
2767 (khtml::RenderBlock::handleFloatingOrPositionedChild):
2768 (khtml::RenderBlock::handleCompactChild):
2769 (khtml::RenderBlock::insertCompactIfNeeded):
2770 (khtml::RenderBlock::handleRunInChild):
2771 (khtml::RenderBlock::collapseMargins):
2772 (khtml::RenderBlock::clearFloatsIfNeeded):
2773 (khtml::RenderBlock::estimateVerticalPosition):
2774 (khtml::RenderBlock::determineHorizontalPosition):
2775 (khtml::RenderBlock::setCollapsedBottomMargin):
2776 (khtml::RenderBlock::adjustChildIfOverhangingFloatsExist):
2777 (khtml::RenderBlock::handleBottomOfBlock):
2778 (khtml::RenderBlock::layoutBlockChildren):
2779 (khtml::RenderBlock::getAbsoluteRepaintRectIncludingFloats):
2780 (khtml::RenderBlock::addOverHangingFloats):
2781 * khtml/rendering/render_block.h:
2782 (khtml::RenderBlock::maxTopMargin):
2783 (khtml::RenderBlock::maxBottomMargin):
2784 (khtml::RenderBlock::CompactInfo::compact):
2785 (khtml::RenderBlock::CompactInfo::block):
2786 (khtml::RenderBlock::CompactInfo::matches):
2787 (khtml::RenderBlock::CompactInfo::clear):
2788 (khtml::RenderBlock::CompactInfo::set):
2789 (khtml::RenderBlock::CompactInfo::CompactInfo):
2790 (khtml::RenderBlock::MarginInfo::setAtTopOfBlock):
2791 (khtml::RenderBlock::MarginInfo::setAtBottomOfBlock):
2792 (khtml::RenderBlock::MarginInfo::clearMargin):
2793 (khtml::RenderBlock::MarginInfo::setSelfCollapsingBlockClearedFloat):
2794 (khtml::RenderBlock::MarginInfo::setTopQuirk):
2795 (khtml::RenderBlock::MarginInfo::setBottomQuirk):
2796 (khtml::RenderBlock::MarginInfo::setDeterminedTopQuirk):
2797 (khtml::RenderBlock::MarginInfo::setPosMargin):
2798 (khtml::RenderBlock::MarginInfo::setNegMargin):
2799 (khtml::RenderBlock::MarginInfo::setPosMarginIfLarger):
2800 (khtml::RenderBlock::MarginInfo::setNegMarginIfLarger):
2801 (khtml::RenderBlock::MarginInfo::setMargin):
2802 (khtml::RenderBlock::MarginInfo::atTopOfBlock):
2803 (khtml::RenderBlock::MarginInfo::canCollapseWithTop):
2804 (khtml::RenderBlock::MarginInfo::canCollapseWithBottom):
2805 (khtml::RenderBlock::MarginInfo::canCollapseTopWithChildren):
2806 (khtml::RenderBlock::MarginInfo::canCollapseBottomWithChildren):
2807 (khtml::RenderBlock::MarginInfo::selfCollapsingBlockClearedFloat):
2808 (khtml::RenderBlock::MarginInfo::quirkContainer):
2809 (khtml::RenderBlock::MarginInfo::determinedTopQuirk):
2810 (khtml::RenderBlock::MarginInfo::topQuirk):
2811 (khtml::RenderBlock::MarginInfo::bottomQuirk):
2812 (khtml::RenderBlock::MarginInfo::posMargin):
2813 (khtml::RenderBlock::MarginInfo::negMargin):
2814 (khtml::RenderBlock::MarginInfo::margin):
2815 * khtml/rendering/render_box.cpp:
2816 (RenderBox::calcAbsoluteVertical):
2817 * khtml/rendering/render_box.h:
2818 (khtml::RenderBox::marginTop):
2819 (khtml::RenderBox::marginBottom):
2820 (khtml::RenderBox::marginLeft):
2821 (khtml::RenderBox::marginRight):
2822 * khtml/rendering/render_image.cpp:
2823 (RenderImage::setImage):
2824 * khtml/rendering/render_object.cpp:
2825 (RenderObject::sizesToMaxWidth):
2826 * khtml/rendering/render_object.h:
2827 (khtml::RenderObject::collapsedMarginTop):
2828 (khtml::RenderObject::collapsedMarginBottom):
2829 (khtml::RenderObject::maxTopMargin):
2830 (khtml::RenderObject::maxBottomMargin):
2831 (khtml::RenderObject::marginTop):
2832 (khtml::RenderObject::marginBottom):
2833 (khtml::RenderObject::marginLeft):
2834 (khtml::RenderObject::marginRight):
2835 * khtml/rendering/render_text.h:
2836 (khtml::RenderText::marginLeft):
2837 (khtml::RenderText::marginRight):
2838 * khtml/xml/dom_elementimpl.cpp:
2839 (ElementImpl::recalcStyle):
2841 2004-10-12 Ken Kocienda <kocienda@apple.com>
2847 <rdar://problem/3836158> REGRESSION (Mail): command-right-arrow moves to beginning of next line
2849 * khtml/editing/selection.cpp:
2850 (khtml::endOfLastRunAt): Do not let the end of the last run on a line be a BR.
2851 This will make it seem like the run ends on the next line.
2853 2004-10-12 Ken Kocienda <kocienda@apple.com>
2859 <rdar://problem/3836986> Delete code removes elements of table structure; can result in very broken-looking web pages
2861 * khtml/editing/htmlediting.cpp:
2862 (khtml::isTableStructureNode): New helper. Determines whether a node is a table cell,
2863 row, section, or column.
2864 (khtml::CompositeEditCommand::removeFullySelectedNode): New helper that recurses into elements
2865 of table structure when doing deletes, rather than deleting the structure elements themselves.
2866 (khtml::DeleteSelectionCommand::moveNodesAfterNode): Do not move content between elements
2867 of table structure. We may want to revisit this some day, but this seems like the best behavior
2869 (khtml::DeleteSelectionCommand::doApply): Call removeFullySelectedNode instead of removeNode
2871 * khtml/editing/htmlediting.h: Add declarations for new functions.
2873 2004-10-12 Richard Williamson <rjw@apple.com>
2875 Fixed access to DOM object via WebScriptObject API.
2876 The execution context for DOM objects wasn't being found.
2877 <rdar://problem/3831372> The valueForKey method for @"offsetLeft" on a paragraph element causes a crash.
2880 Fixed <rdar://problem/3831063> regions use left offset instead of top offset
2883 * khtml/khtml_part.h:
2884 * khtml/rendering/render_object.cpp:
2885 (RenderObject::addDashboardRegions):
2887 (-[DOMNode isContentEditable]):
2888 (-[DOMNode KJS::Bindings::]):
2889 * kwq/KWQKHTMLPart.h:
2890 * kwq/KWQKHTMLPart.mm:
2891 (KWQKHTMLPart::executionContextForDOM):
2893 2004-10-12 Ken Kocienda <kocienda@apple.com>
2899 <rdar://problem/3834779> Mail crashes when editing HTML message - khtml::Selection::layout()
2901 * kwq/KWQKHTMLPart.mm:
2902 (KWQKHTMLPart::styleForSelectionStart): Table code seems to be more robust when the call to
2903 insert our style-checking node is done with an appendChild rather than an insertBefore. Note
2904 that this table-related problem was exposed by fixing Selection::layout(), which I did
2905 yesterday. This change simply improves things even more so that we do not crash in the
2906 scenario described in the bug.
2908 2004-10-11 Ken Kocienda <kocienda@apple.com>
2912 This is a partial fix to this bug:
2914 <rdar://problem/3832886> increase quote level on new mail document leads to immediate
2915 crash in caret painting code
2917 To eliminate the bad behavior for good, I have done some investigations in Mail code,
2918 and I have sent a suggested code change on to Grant. Basically, Mail can't add empty
2919 blocks (like blockquote elements used for quoting) to documents without giving those
2920 blocks some content (so they have a height).
2922 I added some other crash protections below.
2924 * khtml/editing/selection.cpp:
2925 (khtml::Selection::layout): Check for non-null position after calls to VisiblePosition, since
2926 the VisiblePosition constructors may fail to find a visible spot in the document. Also, add
2927 a couple position-has-renderer assertion checks.
2928 * kwq/KWQKHTMLPart.mm:
2929 (KWQKHTMLPart::styleForSelectionStart): Take out pos.isNotNull() assertion since VisiblePosition
2930 may not yield a position. This assertion is a holdover from before we had VisiblePosition.
2931 (KWQKHTMLPart::fontForSelection): Rearrange the code a little to deal with possible null results from
2934 2004-10-11 Darin Adler <darin@apple.com>
2938 - fixed <rdar://problem/3834230> empty table can result in division by 0
2940 * khtml/rendering/render_table.cpp: (RenderTableSection::layoutRows):
2941 Added 0 check; rolled in from KDE.
2943 2004-10-11 Darin Adler <darin@apple.com>
2947 - fixed <rdar://problem/3818712> form checkbox value property is read only
2949 The underlying problem was that we were storing two separate values for all
2950 form elements; one for the value property (JavaScript) and the other for the
2951 value attribute (DOM). This is a good idea for text input, but not for other types.
2953 * khtml/html/html_formimpl.h: Changed setValue to take a const DOMString reference.
2954 Added private storesValueSeparateFromAttribute function.
2955 * khtml/html/html_formimpl.cpp:
2956 (DOM::HTMLInputElementImpl::setType): Handle type changes, including detaching and re-attaching
2957 if type changed, and moving value from m_value to ATTR_VALUE and vice versa.
2958 (DOM::HTMLInputElementImpl::type): Added a case for ISINDEX and moved the default out of the
2959 switch so that we will get a warning if a type is left out.
2960 (DOM::HTMLInputElementImpl::parseHTMLAttribute): Tweaked comment format.
2961 (DOM::HTMLInputElementImpl::reset): Changed to only nuke the value if the value property is stored
2962 separately from the attribute. Otherwise, we just want to lave it alone
2963 (DOM::HTMLInputElementImpl::value): Changed to always use m_value if it's not null, then fall back
2964 on the attribute, and finally fall back to the "on" for the checkbox only if both are null.
2965 (DOM::HTMLInputElementImpl::setValue): Changed to set the attribute unless the value property is
2966 supposed to be stored separate from the attribute.
2967 (DOM::HTMLInputElementImpl::storesValueSeparateFromAttribute): Added. Returns true for text-type
2968 input elements, and false for the others.
2970 2004-10-11 Darin Adler <darin@apple.com>
2974 - fixed <rdar://problem/3296652> checkbox input type does not respond to onchange
2976 * khtml/rendering/render_form.cpp:
2977 (RenderFormElement::updateFromElement): Some new code, commented out, for form element colors.
2978 (RenderCheckBox::slotStateChanged): Added call to onChange.
2980 2004-10-11 Ken Kocienda <kocienda@apple.com>
2984 Finish selection affinity implementation. This includes code to set the
2985 affinity correctly when clicking with the mouse, and clearing the
2986 affinity when altering the selection using any of the Selection object
2989 Each instance of the positionForCoordinates, inlineBox and caretRect
2990 functions have been changed to include an EAffinity argument to give results
2991 which take this bit into account.
2993 * khtml/editing/selection.cpp:
2994 (khtml::Selection::init): Default affinity is now UPSTREAM, to match AppKit.
2995 (khtml::Selection::modifyAffinity): New function to compute affinity based on
2996 modification constants.
2997 (khtml::Selection::moveTo): Reset affinity to UPSTREAM.
2998 (khtml::Selection::modifyExtendingRightForward): Ditto.
2999 (khtml::Selection::modifyMovingRightForward): Ditto.
3000 (khtml::Selection::modifyExtendingLeftBackward): Ditto.
3001 (khtml::Selection::modifyMovingLeftBackward): Ditto.
3002 (khtml::Selection::modify): Support saving, restoring, and then calculating new
3003 affinity value as needed.
3004 (khtml::Selection::xPosForVerticalArrowNavigation):
3005 (khtml::Selection::clear): Reset affinity to UPSTREAM.
3006 (khtml::Selection::setBase): Ditto.
3007 (khtml::Selection::setExtent): Ditto.
3008 (khtml::Selection::setBaseAndExtent): Ditto.
3009 (khtml::Selection::layout): Pass affinity to caretRect().
3010 (khtml::Selection::validate): Pass along affinity parameter to new functions that
3012 (khtml::startOfFirstRunAt): Changed the way that the y-coordinate search is done, to
3013 keep this code working with changes made in selectionForLine().
3014 (khtml::endOfLastRunAt): Ditto.
3015 (khtml::selectionForLine): Make this function work for all renderers, not just text
3017 * khtml/editing/selection.h:
3018 (khtml::operator==): Consider affinity in equality check.
3019 * khtml/editing/visible_units.cpp:
3020 (khtml::previousLinePosition): Pass affinity argument to function so it can take this
3021 information into account while processing.
3022 (khtml::nextLinePosition): Ditto.
3023 (khtml::previousParagraphPosition): Ditto.
3024 (khtml::nextParagraphPosition): Ditto.
3025 * khtml/editing/visible_units.h: Ditto, for each of the functions listed.
3026 * khtml/khtml_events.cpp:
3027 (khtml::MouseEvent::offset): Rework code to remove dependence on NodeImpl::positionForCoordinates,
3028 as this function is being removed.
3029 * khtml/khtml_part.cpp:
3030 (KHTMLPart::isPointInsideSelection): Ditto.
3031 (KHTMLPart::selectClosestWordFromMouseEvent): Ditto.
3032 (KHTMLPart::handleMousePressEventTripleClick): Ditto.
3033 (KHTMLPart::handleMousePressEventSingleClick): Ditto. Plus, pass affinity argument in
3034 call to positionForCoordinates, and set resulting affinity on the selection.
3035 (KHTMLPart::handleMouseMoveEventSelection): Rework code to remove dependence on
3036 NodeImpl::positionForCoordinates, as this function is being removed.
3037 (KHTMLPart::khtmlMouseReleaseEvent): Ditto.
3038 * khtml/rendering/render_block.cpp:
3039 (khtml::RenderBlock::positionForCoordinates): Now takes an affinity argument.
3040 * khtml/rendering/render_block.h:
3041 * khtml/rendering/render_box.cpp:
3042 (RenderBox::caretRect): Ditto.
3043 * khtml/rendering/render_box.h:
3044 * khtml/rendering/render_br.cpp:
3045 (RenderBR::positionForCoordinates): Ditto.
3046 (RenderBR::caretRect): Ditto.
3047 (RenderBR::inlineBox): Ditto.
3048 * khtml/rendering/render_br.h:
3049 * khtml/rendering/render_container.cpp:
3050 (RenderContainer::positionForCoordinates): Ditto.
3051 * khtml/rendering/render_container.h:
3052 * khtml/rendering/render_flow.cpp:
3053 (RenderFlow::caretRect): Ditto.
3054 * khtml/rendering/render_flow.h:
3055 * khtml/rendering/render_inline.cpp:
3056 (RenderInline::positionForCoordinates): Ditto.
3057 * khtml/rendering/render_inline.h:
3058 * khtml/rendering/render_object.cpp:
3059 (RenderObject::caretRect): Ditto.
3060 (RenderObject::positionForCoordinates): Ditto.
3061 (RenderObject::inlineBox): Ditto.
3062 * khtml/rendering/render_object.h:
3063 * khtml/rendering/render_replaced.cpp:
3064 (RenderReplaced::positionForCoordinates): Ditto.
3065 * khtml/rendering/render_replaced.h:
3066 * khtml/rendering/render_text.cpp:
3067 (RenderText::positionForCoordinates): Ditto.
3068 (firstRendererOnNextLine): New helper used by caretRect().
3069 (RenderText::caretRect): Now takes an affinity argument.
3070 (RenderText::inlineBox): Ditto.
3071 * khtml/rendering/render_text.h:
3072 * khtml/xml/dom_nodeimpl.cpp: Remove positionForCoordinates helper.
3073 * khtml/xml/dom_nodeimpl.h: Ditto.
3074 * khtml/xml/dom_position.cpp:
3075 (DOM::Position::previousLinePosition): Now takes an affinity argument.
3076 (DOM::Position::nextLinePosition): Ditto.
3077 * khtml/xml/dom_position.h:
3078 * kwq/WebCoreBridge.h:
3079 * kwq/WebCoreBridge.mm:
3080 (-[WebCoreBridge caretRectAtNode:offset:affinity:]): Ditto.
3081 (-[WebCoreBridge setSelectedDOMRange:affinity:]): Ditto.
3082 (-[WebCoreBridge _positionForPoint:]): Rework code to remove dependence on
3083 NodeImpl::positionForCoordinates, as this function is being removed.
3085 2004-10-11 Darin Adler <darin@apple.com>
3089 - fixed <rdar://problem/3670280> scroll position on overflowed textareas resets when leaving the tab
3091 * kwq/KWQTextArea.mm: (-[KWQTextAreaTextView becomeFirstResponder]):
3092 Scroll to reveal the text area, don't scroll to reveal the text view itself.
3093 Scrolling the text view ended up putting it at the top left, regardless of
3094 where the insertion point is.
3096 2004-10-11 Darin Adler <darin@apple.com>
3100 - fixed <rdar://problem/3831546> More text is copied than is visually selected
3102 The bug here is that upstream was moving a position too far.
3104 * khtml/xml/dom_position.cpp:
3105 (DOM::Position::upstream): Use the "deep equivalent" node rather than the original node passed
3106 in for various checks. Also use local variables a bit more for slightly more efficiency.
3107 (DOM::Position::downstream): Ditto.
3109 2004-10-11 Darin Adler <darin@apple.com>
3113 - fixed <rdar://problem/3833841> adding an event listener for keypress events does not work
3115 * khtml/xml/dom2_eventsimpl.h: Added numEventIds and made typeToId take a const DOMString &.
3116 * khtml/xml/dom2_eventsimpl.cpp:
3117 (EventImpl::typeToId): Changed to use table. Added "keypress", otherwise, the same as before.
3118 (EventImpl::idToType): Changed to use table.
3120 2004-10-10 John Sullivan <sullivan@apple.com>
3122 - fixed <rdar://problem/3664375> repro crash in -[KWQAccObject accessibilityAttributeNames]
3124 (-[KWQAccObject accessibilityActionNames]):
3125 check for nil m_renderer
3127 2004-10-09 Darin Adler <darin@apple.com>
3131 - fixed <rdar://problem/3828147> REGRESSION: textareas with wrap="off" show their contents in a thin vertical line of text
3133 * kwq/KWQTextArea.h: Added setTextColor and setBackgroundColor methods. This is really for another fix
3134 I'm landing later, but it does no harm to add these now.
3135 * kwq/KWQTextArea.mm:
3136 (-[KWQTextArea _configureTextViewForWordWrapMode]): Set the container size after changing the flag that
3137 determines if the width tracks the text view. Otherwise, we won't successfully set the width in the case
3138 where we don't want it to track the text view. This caused the bug.
3139 (-[KWQTextArea _createTextView]): Remove unneeded call to setMaxSize. The above method already does that.
3140 (-[KWQTextArea setTextColor:]): Added.
3141 (-[KWQTextArea setBackgroundColor:]): Added.
3143 2004-10-09 Darin Adler <darin@apple.com>
3147 - fixed <rdar://problem/3829452> REGRESSION (156-157): onload handler doesn't run on page with meta refresh of 0 duration (new Apple start page)
3149 The fix for <rdar://problem/3773150> made it so <meta> redirects prevent tokenizing the rest of the page.
3150 This is incorrect; the reason the JavaScript-triggered loads prevent tokenizing is that they take place
3151 "right away" in other browsers, but that is not true of <meta> redirect. We fixed this by using a separate
3152 call for <meta> redirect and not preventing tokenizing when that's in effect.
3154 * khtml/khtml_part.h: Removed userGesture parameter from scheduleRedirection. Renamed
3155 isImmediateRedirectPending to isScheduledLocationChangePending. Added scheduleLocationChange.
3156 * khtml/khtml_part.cpp:
3157 (KHTMLPart::openURL): Updated for new constant name.
3158 (KHTMLPart::scheduleRedirection): Removed now-unneeded userGesture parameter, and removed code that
3159 does the special case for redirection during load; a <meta> refresh can never be one of those special
3160 redirects during a load because it redirects the frame itself, not another frame. Also tightened up
3161 the logic by always stopping the redirect timer even if we aren't restarting it.
3162 (KHTMLPart::scheduleLocationChange): Added. Like scheduleRedirection, but with a different constant
3163 so we can tell it apart and always a delay of 0. The "redirection during load" case was moved in here
3164 and renamed to locationChangeScheduledDuringLoad.
3165 (KHTMLPart::isScheduledLocationChangePending): Renamed from isImmediateRedirectPending. This now
3166 returns true only for location changes and history navigation, not <meta> redirects.
3167 (KHTMLPart::scheduleHistoryNavigation): Tightened up logic to do the stop() outside the if as above,
3168 and got rid of a silly timer delay computation that always resulted in 0.
3170 * khtml/khtmlpart_p.h: Added a new value to the RedirectionScheduled enum for scheduleLocationChange
3171 and also renamed one of the existing values.
3173 * khtml/html/htmltokenizer.cpp: (HTMLTokenizer::write): Changed to use isScheduledLocationChangePending instead
3174 of isImmediateRedirectPending, because we do want to continue tokenizing if it's actually a redirect.
3176 * khtml/ecma/kjs_html.cpp: (KJS::HTMLDocument::putValue): Changed to call the new scheduleLocationChange
3177 instead of calling scheduleRedirection with delay of 0.
3178 * khtml/ecma/kjs_window.cpp:
3179 (Window::put): Ditto.
3180 (WindowFunc::tryCall): Ditto.
3181 (Location::put): Ditto.
3182 (LocationFunc::tryCall): Ditto.
3184 2004-10-09 Darin Adler <darin@apple.com>
3188 - fixed <rdar://problem/3658277> REGRESSION (1.1-1.2): form submission should either not simulate a click at all or use (0,0) the way Mozilla does
3190 * khtml/html/html_elementimpl.cpp: (HTMLElementImpl::click): Use 0,0 for the coordinates.
3192 2004-10-09 Darin Adler <darin@apple.com>
3196 - fixed <rdar://problem/3804665> REGRESSION: WebCore framework now has many init routines
3198 * khtml/xml/dom_nodeimpl.h: Changed anyQName declaration to not use the inline function
3199 makeId. Surprisingly, the inline function was not "constant-folded" and we ended up with
3200 a copy of the function in each file as an init routine for the framework.
3202 * khtml/ecma/kjs_html.cpp: (Gradient::colorStops): Get rid of initialized ColorStop
3203 globals; their constructors were showing up as init routines for the framework.
3205 * khtml/rendering/render_style.h: Got rid of inline initialDashboardRegions function.
3206 * khtml/rendering/render_style.cpp: (RenderStyle::initialDashboardRegions): Made this
3207 a normal function. When it was an inline function, the constructors for the per-file
3208 copies of the globals were showing up as init routines for the framework.
3210 2004-10-09 Chris Blumenberg <cblu@apple.com>
3213 <rdar://problem/3625352> up and down arrow and page up/down keys don't work to scroll overflow:auto/scroll/overlay areas
3214 <rdar://problem/3397658> scroll wheel does not work to scroll overflow:auto/scroll/overlay areas (RSS)
3215 <rdar://problem/3822027> REGRESSION (Mail): When selection moves out of visible area, should center as NSText does
3217 Reviewed by hyatt, kocienda.
3219 * khtml/rendering/render_layer.cpp:
3220 (RenderLayer::scroll): new
3221 * khtml/rendering/render_layer.h:
3222 * khtml/rendering/render_object.cpp:
3223 (RenderObject::scroll): new
3224 * khtml/rendering/render_object.h:
3225 * kwq/KWQKHTMLPart.h:
3226 * kwq/KWQKHTMLPart.mm:
3227 (KWQKHTMLPart::scrollOverflow): new
3228 (KWQKHTMLPart::scrollOverflowWithScrollWheelEvent): new
3229 (KWQKHTMLPart::khtmlMousePressEvent): store pressed node so we know where the focus is
3230 * kwq/KWQScrollBar.h:
3231 * kwq/KWQScrollBar.mm:
3232 (QScrollBar::setValue): return a bool
3233 (QScrollBar::scrollbarHit): ditto
3234 (QScrollBar::scroll): new
3235 * kwq/WebCoreBridge.h:
3236 * kwq/WebCoreBridge.mm:
3237 (-[WebCoreBridge scrollOverflowInDirection:granularity:]): new
3238 (-[WebCoreBridge scrollOverflowWithScrollWheelEvent:]): new
3239 (-[WebCoreBridge ensureSelectionVisible]): visually center the extent of the selection
3241 2004-10-06 David Hyatt <hyatt@apple.com>
3243 Back out style sharing perf fix.
3245 * khtml/css/cssstyleselector.cpp:
3246 (khtml::CSSStyleSelector::locateCousinList):
3247 (khtml::CSSStyleSelector::canShareStyleWithElement):
3248 (khtml::CSSStyleSelector::locateSharedStyle):
3249 * khtml/css/cssstyleselector.h:
3250 * khtml/html/html_elementimpl.h:
3251 (DOM::HTMLElementImpl::inlineStyleDecl):
3252 * khtml/xml/dom_elementimpl.cpp:
3253 (ElementImpl::recalcStyle):
3254 * khtml/xml/dom_elementimpl.h:
3258 2004-10-05 David Hyatt <hyatt@apple.com>
3260 Fix a bug in the iteration of locateCousinList and clean up the style sharing stats code.
3262 * khtml/css/cssstyleselector.cpp:
3263 (khtml::CSSStyleSelector::locateCousinList):
3264 (khtml::CSSStyleSelector::elementsCanShareStyle):
3265 (khtml::CSSStyleSelector::locateSharedStyle):
3266 (khtml::CSSStyleSelector::styleForElement):
3268 2004-10-05 Ken Kocienda <kocienda@apple.com>
3272 * khtml/rendering/bidi.cpp:
3273 (khtml::RenderBlock::computeHorizontalPositionsForLine): Fix coding mistake that
3274 broke layout tests involving compacts.
3276 2004-10-05 Ken Kocienda <kocienda@apple.com>
3280 Finish selection affinity implementation. This includes code to set the
3281 affinity correctly when clicking with the mouse, and clearing the
3282 affinity when altering the selection using any of the Selection object
3285 Each instance of the positionForCoordinates function in the render tree
3286 has been changed to include an EAffinity argument. It is now the job of this
3287 function to set the selection affinity.
3289 * khtml/editing/selection.cpp:
3290 (khtml::Selection::moveTo): Set affinity to DOWNSTREAM.
3291 (khtml::Selection::modify): Ditto.
3292 (khtml::Selection::clear): Ditto.
3293 (khtml::Selection::setBase): Ditto.
3294 (khtml::Selection::setExtent): Ditto.
3295 (khtml::Selection::setBaseAndExtent): Ditto.
3296 * khtml/editing/selection.h:
3297 (khtml::operator==): Consider affinity in equality check.
3298 * khtml/khtml_events.cpp:
3299 (khtml::MouseEvent::offset): Rework code to remove dependence on NodeImpl::positionForCoordinates,
3300 as this function is being removed.
3301 * khtml/khtml_part.cpp:
3302 (KHTMLPart::isPointInsideSelection): Ditto.
3303 (KHTMLPart::selectClosestWordFromMouseEvent): Ditto.
3304 (KHTMLPart::handleMousePressEventTripleClick): Ditto.
3305 (KHTMLPart::handleMousePressEventSingleClick): Ditto. Plus, pass affinity argument in
3306 call to positionForCoordinates, and set resulting affinity on the selection.
3307 (KHTMLPart::handleMouseMoveEventSelection): Rework code to remove dependence on
3308 NodeImpl::positionForCoordinates, as this function is being removed.
3309 (KHTMLPart::khtmlMouseReleaseEvent):
3310 * khtml/rendering/render_block.cpp:
3311 (khtml::RenderBlock::positionForCoordinates): Changed, as described above.
3312 * khtml/rendering/render_block.h:
3313 * khtml/rendering/render_br.cpp:
3314 (RenderBR::positionForCoordinates): Ditto.
3315 * khtml/rendering/render_br.h: