Patch for https://bugs.webkit.org/show_bug.cgi?id=43013
[WebKit-https.git] / WebCore / ChangeLog
1 2010-07-26  Sam Weinig  <sam@webkit.org>
2
3         Reviewed by Anders Carlsson.
4
5         Patch for https://bugs.webkit.org/show_bug.cgi?id=43013
6         Part of <rdar://problem/8152434>
7         Add support for scrolling using the keyboard in WebKit2
8
9         Add a new export and make WindowsKeyboardCodes private.
10
11         * WebCore.exp.in:
12         * WebCore.xcodeproj/project.pbxproj:
13
14 2010-07-26  Darin Adler  <darin@apple.com>
15
16         Reviewed by Eric Carlson.
17
18         Don't handle empty URL attributes specially in reflection, even in cases such as <img src="">
19         https://bugs.webkit.org/show_bug.cgi?id=42087
20
21         * html/HTMLMediaElement.cpp:
22         (WebCore::HTMLMediaElement::insertedIntoDocument): Check if the src attribute is empty
23         by calling getAttribute instead of calling the src function. This keeps the behavior
24         correct with the reflection change, and has a side benefit of being more efficient as
25         well since there is no reason to resolve the URL.
26         (WebCore::HTMLMediaElement::loadResource): Call getNonEmptyURLAttribute directly here
27         instead of using the poster function.
28         (WebCore::HTMLMediaElement::getPluginProxyParams): Ditto.
29
30         * html/HTMLMediaElement.h: Removed the poster virtual function. The code that gets the
31         poster attribute is already video-specific and uses generic attribute fetching code, so
32         we no longer need a virtual function. Removed the src function. Since this would give
33         a different result from the JavaScript reflected src attribute, it's safer not to have
34         it, and it was easy to convert all the callers.
35
36         * html/HTMLMediaElement.idl: Removed the NonEmpty extended attribute from the src
37         attribute.
38
39         * html/HTMLSourceElement.cpp: Removed the src function for the same reason as above.
40         * html/HTMLSourceElement.h: Ditto.
41
42         * html/HTMLSourceElement.idl: Removed the NonEmpty extended attribute from the src
43         attribute.
44
45         * html/HTMLVideoElement.cpp:
46         (WebCore::HTMLVideoElement::parseMappedAttribute): Removed code to update m_posterURL,
47         which is no longer needed. Changed call site that used the poster() function to instead
48         use getNonEmptyURLAttribute.
49         (WebCore::HTMLVideoElement::updatePosterImage): Check if the poster attribute is empty
50         by calling getAttribute instead of calling the poster function. This keeps the behavior
51         correct with the reflection change, and has a side benefit of being more efficient as
52         well since there is no reason to resolve the URL.
53
54         * html/HTMLVideoElement.h: Removed the poster function for the same reason we removed
55         the src function above. Also removed now-unneeded m_posterURL data member.
56
57         * html/HTMLVideoElement.idl: Removed the NonEmpty extended attribute from the poster
58         attribute.
59
60 2010-07-26  Chris Fleizach  <cfleizach@apple.com>
61
62         Reviewed by Darin Adler.
63
64         AX: remove intValue() hasIntValue() from AccessibilityObject
65         https://bugs.webkit.org/show_bug.cgi?id=43006
66
67         These methods are very Mac centric. The mac centric code should be moved into the Mac wrapper.
68
69         No new tests, existing tests cover changes to logic.
70
71         * accessibility/AccessibilityObject.cpp:
72         (WebCore::AccessibilityObject::checkboxOrRadioValue):
73         * accessibility/AccessibilityObject.h:
74         (WebCore::):
75         * accessibility/AccessibilityRenderObject.cpp:
76         (WebCore::AccessibilityRenderObject::selectedRadioButton):
77         (WebCore::AccessibilityRenderObject::selectedTabItem):
78         (WebCore::AccessibilityRenderObject::checkboxOrRadioValue):
79         * accessibility/AccessibilityRenderObject.h:
80         * accessibility/mac/AccessibilityObjectWrapper.mm:
81         (-[AccessibilityObjectWrapper accessibilityAttributeValue:]):
82
83 2010-07-26  Ryosuke Niwa  <rniwa@webkit.org>
84
85         Reviewed by Darin Adler.
86
87         Applying inline style to a text node whose parent is an inline editable root causes crash
88         https://bugs.webkit.org/show_bug.cgi?id=39989
89
90         The crash was caused by splitTextElementAtStart and splitTextElementAtEnd assuming that the parent
91         and the grandparent of the specified text node is editable.
92
93         Modified splitTextElementAtStart and splitTextElementAtEnd so that they call splitTextAtStart
94         and splitTextAtEnd respectively when the grandparent is not editable.
95
96         Also modified SplitTextNodeContainingElement to exit early if the grandparent of m_text is not editable.
97
98         Test: editing/style/style-text-node-without-editable-parent.html
99
100         * editing/ApplyStyleCommand.cpp:
101         (WebCore::ApplyStyleCommand::splitTextElementAtStart):
102         (WebCore::ApplyStyleCommand::splitTextElementAtEnd):
103         * editing/SplitTextNodeContainingElementCommand.cpp:
104         (WebCore::SplitTextNodeContainingElementCommand::doApply):
105
106 2010-07-26  Simon Fraser  <simon.fraser@apple.com>
107
108         Reviewed by Anders Carlsson.
109
110         Gmail compose is upside-down after r64054
111         https://bugs.webkit.org/show_bug.cgi?id=43000
112         
113         We need to call setGeometryOrientation() whenever the iframe compositing layer attachment changes,
114         not just when requiresScrollLayer() returns false. This ensures that the orientation is updated
115         correctly.
116
117         Test: compositing/iframes/iframe-content-flipping.html
118
119         * rendering/RenderLayerCompositor.cpp:
120         (WebCore::RenderLayerCompositor::ensureRootPlatformLayer):
121
122 2010-07-26  Justin Schuh  <jschuh@chromium.org>
123
124         Reviewed by Darin Fisher.
125
126         Check history state against origin before setting
127         https://bugs.webkit.org/show_bug.cgi?id=42858
128
129         Tests: fast/loader/stateobjects/replacestate-base-illegal.html
130                fast/loader/stateobjects/replacestate-base-legal.html
131
132         * page/History.cpp:
133         (WebCore::History::urlForState):
134         (WebCore::History::stateObjectAdded):
135
136 2010-07-26  Martin Robinson  <mrobinson@igalia.com>
137
138         Reviewed by Xan Lopez.
139
140         [GTK] Get rid of libgdom
141         https://bugs.webkit.org/show_bug.cgi?id=42378
142
143         Remove all build steps for building GObject DOM bindings.
144
145         * GNUmakefile.am: Remove GObject DOM bindings build.
146
147 2010-07-26  Mark Rowe  <mrowe@apple.com>
148
149         Reviewed by Dan Bernstein.
150
151         <http://webkit.org/b/42990> REGRESSION (r63854): Safari RSS layout is broken due to changes in vendor prefix handling
152
153         In r63854 support for properties with the -khtml vendor prefix was removed. This breaks the layout of
154         Safari's RSS interface, several Dashboard widgets that are included with Mac OS X, and at least one
155         popular third-party Mac OS X application. This change will need to be revisited in a manner that has
156         a clearer backwards-compatibility strategy. <http://webkit.org/b/42093> will be reopened to track that.
157
158         * css/CSSParser.cpp:
159         (WebCore::cssPropertyID): Revert the WebCore change from r63854.
160
161 2010-07-26  Brady Eidson  <beidson@apple.com>
162
163         Reviewed by Sam Weinig.
164
165         Add ability for WK2 to register a scheme to always be displayed as an empty document
166         https://bugs.webkit.org/show_bug.cgi?id=42995
167
168         No new tests. (Currently not possible to test in WebKit1 DRT, can be testing in future WK2 tester)
169
170         * WebCore.exp.in:
171         
172         * loader/MainResourceLoader.cpp:
173         (WebCore::shouldLoadAsEmptyDocument):
174         
175         * platform/SchemeRegistry.cpp:
176         (WebCore::emptyDocumentSchemes):
177         (WebCore::SchemeRegistry::registerURLSchemeAsEmptyDocument):
178         (WebCore::SchemeRegistry::shouldLoadURLSchemeAsEmptyDocument):
179         * platform/SchemeRegistry.h:
180
181 2010-07-26  Patrick Gansterer  <paroga@paroga.com>
182
183         Reviewed by Kent Tamura.
184
185         [WINCE] Port ClipboardUtilities to WinCE
186         https://bugs.webkit.org/show_bug.cgi?id=42929
187
188         Make ClipboardUtilitiesWin more portable and remove the global namespace prefix
189         from GlobalLock/GlobalUnlock, since they are only macros on WinCE.
190
191         * platform/win/ClipboardUtilitiesWin.cpp:
192         (WebCore::urlFromPath):
193         (WebCore::getWebLocData):
194         (WebCore::createGlobalData):
195         (WebCore::getURL):
196         (WebCore::getPlainText):
197         (WebCore::fragmentFromHTML):
198
199 2010-07-26  Patrick Gansterer  <paroga@paroga.com>
200
201         Reviewed by Adam Roben.
202
203         [WINCE] Buildfix for BitmapImage.h
204         https://bugs.webkit.org/show_bug.cgi?id=42931
205
206         Use AffineTransform insted of TransformationMatrix.
207
208         * platform/graphics/BitmapImage.h:
209
210 2010-07-26  Dirk Schulze  <krit@webkit.org>
211
212         Unreviewed sort of XCode project file.
213
214         * WebCore.xcodeproj/project.pbxproj:
215
216 2010-07-26  David Kilzer  <ddkilzer@apple.com>
217
218         <http://webkit.org/b/42982> Streamline Inspector Source build phase script should use CONFIGURATION not BUILD_STYLE
219         <rdar://problem/6341764>
220
221         Reviewed by Mark Rowe.
222
223         * WebCore.xcodeproj/project.pbxproj: Switched Streamline
224         Inspector Source build phase script to use CONFIGURATION instead
225         of BUILD_STYLE.
226
227 2010-07-23  Ojan Vafai  <ojan@chromium.org>
228
229         Reviewed by Darin Adler.
230
231         crash in insertParagraph in a contentEditable list
232         https://bugs.webkit.org/show_bug.cgi?id=42919
233
234         Don't try to break out of a list item if the list it's in
235         is the rootEditableElement.
236
237         * editing/CompositeEditCommand.cpp:
238         (WebCore::CompositeEditCommand::breakOutOfEmptyListItem):
239
240 2010-07-26  Marcus Bulach  <bulach@chromium.org>
241
242         Reviewed by Jeremy Orlow.
243
244         Moves IDBKeyRange to use IDBKey rather than SerializedScriptValue / IDBAny.
245         https://bugs.webkit.org/show_bug.cgi?id=42583
246
247         Tests: existing storage/indexeddb/idb-keyrange.html must pass.
248
249         * storage/IDBKeyRange.cpp:
250         (WebCore::IDBKeyRange::IDBKeyRange):
251         * storage/IDBKeyRange.h:
252         (WebCore::IDBKeyRange::create):
253         (WebCore::IDBKeyRange::left):
254         (WebCore::IDBKeyRange::right):
255         * storage/IDBKeyRange.idl:
256         * storage/IndexedDatabaseRequest.cpp:
257         (WebCore::IndexedDatabaseRequest::makeSingleKeyRange):
258         (WebCore::IndexedDatabaseRequest::makeLeftBoundKeyRange):
259         (WebCore::IndexedDatabaseRequest::makeRightBoundKeyRange):
260         (WebCore::IndexedDatabaseRequest::makeBoundKeyRange):
261         * storage/IndexedDatabaseRequest.h:
262         * storage/IndexedDatabaseRequest.idl:
263
264 2010-07-26  Simon Fraser  <simon.fraser@apple.com>
265
266         Reviewed by Anders Carlsson.
267
268         Composited layers don't scroll in WebKit2
269         https://bugs.webkit.org/show_bug.cgi?id=42771
270
271         Rename two methods on RenderLayerCompositor to make their use more clear, and call them
272         when the FrameView gets resized and scrolled.
273
274         * page/FrameView.h:
275         * page/FrameView.cpp:
276         (WebCore::FrameView::setFrameRect): override so we know when the size of the FrameView changes, so that the
277         RenderLayerCompositor can update its clipping layer.
278         (WebCore::FrameView::scrollPositionChanged): Call compositor()->frameViewDidScroll().
279
280         * rendering/RenderLayerBacking.cpp:
281         (WebCore::RenderLayerBacking::updateAfterWidgetResize): updateContentLayerOffset() was renamed.
282
283         * rendering/RenderLayerCompositor.h:
284         * rendering/RenderLayerCompositor.cpp:
285         (WebCore::RenderLayerCompositor::frameViewDidChangeSize): Renamed from updateContentLayerOffset().
286         (WebCore::RenderLayerCompositor::frameViewDidScroll): Renamed from updateContentLayerScrollPosition().
287         (WebCore::RenderLayerCompositor::shouldPropagateCompositingToEnclosingIFrame): Propagate compositing on Mac
288         for any FrameView that doesn't have a native widget, which is the case with WebKit2.
289         (WebCore::RenderLayerCompositor::requiresScrollLayer): New method to tell us whether we need to handle
290         scrolling (and clipping) ourselves.
291         (WebCore::RenderLayerCompositor::ensureRootPlatformLayer): Do geometry flipping on the root platform
292         layer only if we're not doing our own scrolling, and use the new requiresScrollLayer() method.
293         (WebCore::RenderLayerCompositor::attachRootPlatformLayer): Send out the rootPlatformLayer(),
294         so that it returns the clipping layer if we have one.
295
296 2010-07-26  Tony Gentilcore  <tonyg@chromium.org>
297
298         Reviewed by Darin Fisher.
299
300         Move DocumentLoadTiming struct to a new file
301         https://bugs.webkit.org/show_bug.cgi?id=42917
302
303         Also makes DocumentLoadTiming Noncopyable.
304
305         No new tests because no new functionality.
306
307         * GNUmakefile.am:
308         * WebCore.gypi:
309         * WebCore.vcproj/WebCore.vcproj:
310         * WebCore.xcodeproj/project.pbxproj:
311         * loader/DocumentLoadTiming.h: Added.
312         (WebCore::DocumentLoadTiming::DocumentLoadTiming):
313         * loader/DocumentLoader.h:
314         * loader/FrameLoader.cpp:
315         * loader/FrameLoaderTypes.h:
316         * loader/MainResourceLoader.cpp:
317         * page/Timing.cpp:
318
319 2010-07-26  Mark Rowe  <mrowe@apple.com>
320
321         Fix an #if in WebCore.exp.in to correctly detect whether SVG filters are enabled.
322
323         * WebCore.exp.in:
324
325 2010-07-26  Steve Block  <steveblock@google.com>
326
327         Reviewed by Jeremy Orlow.
328
329         DeviceOrientation event listeners should never be called synchronously from addEventListener()
330         https://bugs.webkit.org/show_bug.cgi?id=42304
331
332         No new tests as there are currently no implementations to test. Tests will be
333         added once LayoutTestController methods for providing a mock implementation are
334         in place, see Bug 39589.
335
336         * dom/DeviceOrientationController.cpp:
337         (WebCore::DeviceOrientationController::DeviceOrientationController):
338         (WebCore::DeviceOrientationController::timerFired):
339         (WebCore::DeviceOrientationController::addListener):
340         (WebCore::DeviceOrientationController::removeListener):
341         (WebCore::DeviceOrientationController::removeAllListeners):
342         * dom/DeviceOrientationController.h:
343
344 2010-07-26  Andreas Kling  <andreas.kling@nokia.com>
345
346         Reviewed by Tor Arne Vestbø.
347
348         [Qt] Canvas: Optimize GraphicsContext::addPath()
349
350         Avoid "deep concatenating" the canvas path onto the GC path
351         if the GC path has no elements.
352
353         * platform/graphics/qt/GraphicsContextQt.cpp:
354         (WebCore::GraphicsContext::addPath):
355
356 2010-07-23  Stephen White  <senorblanco@chromium.org>
357
358         Reviewed by Darin Fisher.
359
360         Implement OpenGLES2 helper classes.
361         https://bugs.webkit.org/show_bug.cgi?id=42905
362
363         GLES2Canvas and GLES2Texture are some simple OpenGLES2 helper classes
364         for managing state, samplers and textures.
365
366         When this is hooked up, it will be covered by many layout tests.
367
368         * WebCore.gypi:
369         Add new files to chromium build (contents are protected by #if's for now).
370         * platform/graphics/chromium/GLES2Canvas.cpp: Added.
371         (WebCore::affineTo3x3):
372         (WebCore::GLES2Canvas::State::State):
373         (WebCore::GLES2Canvas::GLES2Canvas):
374         (WebCore::GLES2Canvas::~GLES2Canvas):
375         (WebCore::GLES2Canvas::clearRect):
376         (WebCore::GLES2Canvas::fillRect):
377         (WebCore::GLES2Canvas::setFillColor):
378         (WebCore::GLES2Canvas::setAlpha):
379         (WebCore::GLES2Canvas::translate):
380         (WebCore::GLES2Canvas::rotate):
381         (WebCore::GLES2Canvas::scale):
382         (WebCore::GLES2Canvas::concatCTM):
383         (WebCore::GLES2Canvas::save):
384         (WebCore::GLES2Canvas::restore):
385         (WebCore::GLES2Canvas::drawTexturedRect):
386         (WebCore::GLES2Canvas::setCompositeOperation):
387         (WebCore::GLES2Canvas::applyCompositeOperator):
388         (WebCore::GLES2Canvas::getQuadVertices):
389         (WebCore::GLES2Canvas::getQuadIndices):
390         (WebCore::loadShader):
391         (WebCore::GLES2Canvas::getSimpleProgram):
392         (WebCore::GLES2Canvas::getTexProgram):
393         (WebCore::GLES2Canvas::createTexture):
394         (WebCore::GLES2Canvas::getTexture):
395         (WebCore::GLES2Canvas::checkGLError):
396         * platform/graphics/chromium/GLES2Canvas.h: Added.
397         (WebCore::GLES2Canvas::gles2Context):
398         * platform/graphics/chromium/GLES2Texture.cpp: Added.
399         (WebCore::GLES2Texture::GLES2Texture):
400         (WebCore::GLES2Texture::~GLES2Texture):
401         (WebCore::GLES2Texture::create):
402         (WebCore::convertFormat):
403         (WebCore::GLES2Texture::load):
404         (WebCore::GLES2Texture::bind):
405         * platform/graphics/chromium/GLES2Texture.h: Added.
406         (WebCore::GLES2Texture::):
407         (WebCore::GLES2Texture::format):
408         (WebCore::GLES2Texture::width):
409         (WebCore::GLES2Texture::height):
410
411
412 2010-07-26  Jeremy Orlow  <jorlow@chromium.org>
413
414         Reviewed by Steve Block.
415
416         Fix IndexedDB release build on mac
417         https://bugs.webkit.org/show_bug.cgi?id=42377
418
419         * storage/IDBDatabaseImpl.cpp:
420         (WebCore::IDBDatabaseImpl::objectStore):
421
422 2010-07-26  Satish Sampath  <satish@chromium.org>
423
424         Reviewed by Steve Block.
425
426         Add WebKit plumbing to connect speech requests and callbacks between WebCore and the embedder.
427         https://bugs.webkit.org/show_bug.cgi?id=42367
428
429         No new tests, the relevant LayoutTestController bindings will be added in the next patch.
430
431         * page/SpeechInput.cpp: renamed a method.
432         * page/SpeechInput.h: renamed a method.
433         * page/SpeechInputClient.h: added an extra method to optionally let users stop recording once they have spoken.
434         * page/SpeechInputClientListener.h: renamed a method.
435         * page/SpeechInputListener.h: renamed a method.
436
437 2010-07-26  Andreas Kling  <andreas.kling@nokia.com>
438
439         Reviewed by Kenneth Rohde Christiansen.
440
441         [Qt] getImageData(): Single-pass RGB->BGR and un-premultiplication
442         https://bugs.webkit.org/show_bug.cgi?id=42945
443
444         Combine the two operations into a single loop over the pixel data.
445         Yields a considerable FPS gain on http://www.semantix.gr/statue.html
446
447         * platform/graphics/qt/ImageBufferQt.cpp:
448         (WebCore::getImageData):
449
450 2010-07-26  Pavel Feldman  <pfeldman@chromium.org>
451
452         Reviewed by Shinichiro Hamaji.
453
454         Web Inspector: make sure proxy objects are JSON-stringifiable.
455
456         https://bugs.webkit.org/show_bug.cgi?id=42961
457
458         * inspector/front-end/InjectedScript.js:
459         (injectedScriptConstructor):
460
461 2010-07-26  Alexander Pavlov  <apavlov@chromium.org>
462
463         Reviewed by Joseph Pecoraro.
464
465         Inspector should remember resources sorting set by the user
466         https://bugs.webkit.org/show_bug.cgi?id=19208
467
468         Sorting options both for time and size graphs are memorized in a single property,
469         which makes them possible to retrieve in a single message (e.g. useful for remote debugging).
470
471         * inspector/front-end/ResourcesPanel.js:
472         (WebInspector.ResourcesPanel.prototype.populateSidebar):
473         (WebInspector.ResourcesPanel.prototype._settingsLoaded):
474         (WebInspector.ResourcesPanel.prototype._loadSortOptions):
475         (WebInspector.ResourcesPanel.prototype._loadSortOptionForGraph):
476         (WebInspector.ResourcesPanel.prototype._graphSelected):
477         (WebInspector.ResourcesPanel.prototype._changeSortingFunction):
478         (WebInspector.ResourcesPanel.prototype._selectedOptionNameForGraph):
479         (WebInspector.ResourcesPanel.prototype._doChangeSortingFunction):
480         * inspector/front-end/Settings.js:
481         (WebInspector.populateApplicationSettings):
482
483 2010-07-25  Ryosuke Niwa  <rniwa@webkit.org>
484
485         Reviewed by Darin Adler.
486
487         Redo fails after text node is split by SplitTextNodeCommand
488         https://bugs.webkit.org/show_bug.cgi?id=42941
489
490         The bug was caused by SplitTextNodeCommand not implementing doReapply.
491         Implemented doReapply and added insertText1AndTrimText2 to share the code between doApply and doReapply.
492
493         Test: editing/undo/redo-split-text-node.html
494               editing/undo/redo-split-text-node-with-removal.html
495
496         * editing/SplitTextNodeCommand.cpp:
497         (WebCore::SplitTextNodeCommand::doApply): Calls executeApply.
498         (WebCore::SplitTextNodeCommand::doUnapply): Don't delete m_text1, which is later used in doReapply.
499         (WebCore::SplitTextNodeCommand::doReapply): Added.
500         (WebCore::SplitTextNodeCommand::insertText1AndTrimText2): Added.
501         * editing/SplitTextNodeCommand.h:
502
503 2010-07-23  Philippe Normand  <pnormand@igalia.com>
504
505         Reviewed by Eric Carlson.
506
507         [GTK] media/media-can-play-mpeg-audio.html fails
508         https://bugs.webkit.org/show_bug.cgi?id=42895
509
510         Fixed advertizing of mp3, mpeg and mp4 mime-types in the player.
511
512         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
513         (WebCore::mimeTypeCache):
514
515 2010-07-23  Philippe Normand  <pnormand@igalia.com>
516
517         Reviewed by Eric Carlson.
518
519         LayoutTests/media/media-document-audio-size.html failure
520         https://bugs.webkit.org/show_bug.cgi?id=38376
521
522         Advertize audio/x-wav along with audio/wav when appropriate,
523         otherwise WebKit attempts to use a plugin to play the media.
524
525         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
526         (WebCore::mimeTypeCache):
527
528 2010-07-25  Ryosuke Niwa  <rniwa@webkit.org>
529
530         Reviewed by Kent Tamura.
531
532         splitTextAt*IfNeed and splitTextElementAt*IfNeed need to be cleaned up
533         https://bugs.webkit.org/show_bug.cgi?id=42937
534
535         Isolated the code to decide whether or not text node should be split into isValidCaretPositionInTextNode.
536         Moved the condition check out of *IfNeeded methods to applyRelativeFontStyleChange and applyInlineStyle.
537
538         No new tests added since this is a clean up.
539
540         * editing/ApplyStyleCommand.cpp:
541         (WebCore::ApplyStyleCommand::applyRelativeFontStyleChange): Uses isValidCaretPositionInTextNode.
542         (WebCore::ApplyStyleCommand::applyInlineStyle): Uses isValidCaretPositionInTextNode.
543         (WebCore::ApplyStyleCommand::splitTextAtStart): Renamed from splitTextAtStartIfNeeded.
544         (WebCore::ApplyStyleCommand::splitTextAtEnd): Renamed from splitTextAtEndIfNeeded.
545         (WebCore::ApplyStyleCommand::splitTextElementAtStart): Renamed from splitTextElementAtStartIfNeeded.
546         (WebCore::ApplyStyleCommand::splitTextElementAtEnd): Renamed from splitTextElementAtEndIfNeeded.
547         (WebCore::ApplyStyleCommand::isValidCaretPositionInTextNode): Returns true if the position lies within a text node.
548         * editing/ApplyStyleCommand.h:
549
550 2010-07-25  Patrick Gansterer  <paroga@paroga.com>
551
552         Reviewed by Kent Tamura.
553
554         [WINCE] Buildfix for ImageBufferWince
555         https://bugs.webkit.org/show_bug.cgi?id=42928
556
557         * platform/graphics/wince/ImageBufferWince.cpp:
558         (WebCore::):
559         (WebCore::BufferedImage::draw): Pass through ColorSpace parameter.
560         (WebCore::BufferedImage::drawPattern): Ditto.
561         (WebCore::ImageBuffer::platformTransformColorSpace): Added.
562
563 2010-07-25  Patrick Gansterer  <paroga@paroga.com>
564
565         Reviewed by Kent Tamura.
566
567         [WINCE] Buildfix for PopupMenuWin.cpp
568         https://bugs.webkit.org/show_bug.cgi?id=42932
569
570         Add #if !OS(WINCE) around unsupported window messages.
571
572         * platform/win/PopupMenuWin.cpp:
573         (WebCore::PopupMenu::show):
574         (WebCore::PopupMenu::wndProc):
575
576 2010-07-25  Adam Barth  <abarth@webkit.org>
577
578         Reviewed by Eric Seidel.
579
580         Implement adjustMathMLAttributes
581         https://bugs.webkit.org/show_bug.cgi?id=42952
582
583         This is kind of mechanical overkill for adjusting one attribute name,
584         but it's nice to use the same machinery as SVG.
585
586         * html/HTMLTreeBuilder.cpp:
587         * mathml/mathattrs.in:
588
589 2010-07-25  Adam Barth  <abarth@webkit.org>
590
591         Reviewed by Eric Seidel.
592
593         Functions supporting the legacy tree builder are confusing
594         https://bugs.webkit.org/show_bug.cgi?id=42951
595
596         This patch just inlines these functions into their one call site and
597         removes the notImplemented() calls.  We thought we were going to grow
598         these to support the new tree builder, but we took a different path.
599         Previous to this patch, these functions were confusing because they
600         looked like part of the new tree builder.
601
602         * html/HTMLTreeBuilder.cpp:
603         (WebCore::HTMLTreeBuilder::passTokenToLegacyParser):
604         * html/HTMLTreeBuilder.h:
605
606 2010-07-25  Adam Barth  <abarth@webkit.org>
607
608         Reviewed by Eric Seidel.
609
610         Recent HTML5 TreeBuilder Regression?  All canvas tests ASSERT
611         https://bugs.webkit.org/show_bug.cgi?id=42948
612
613         Remove ASSERT and replace with correct branch.  I put this ASSERT in
614         to make sure we had test coverage for this branch.  I guess we do.  :)
615
616         * html/HTMLConstructionSite.cpp:
617         (WebCore::HTMLConstructionSite::attach):
618
619 2010-07-25  Pavel Feldman  <pfeldman@chromium.org>
620
621         Reviewed by Timothy Hatcher.
622
623         Web Inspector: check of jQuery object type prior to treating it as an array.
624
625         https://bugs.webkit.org/show_bug.cgi?id=42946
626
627         * inspector/front-end/InjectedScript.js:
628         (injectedScriptConstructor):
629
630 2010-07-24  Daniel Bates  <dbates@rim.com>
631
632         Unreviewed, build fix when building with WML support.
633
634         PageCache.cpp compile error when WML is enabled
635         https://bugs.webkit.org/show_bug.cgi?id=40765
636
637         Remove extraneous semicolon.
638
639         Note, building with WML support enabled is broken
640         even with this fix. See Bug #42943 for more details.
641
642         * history/PageCache.cpp:
643         (WebCore::PageCache::canCachePageContainingThisFrame):
644
645 2010-07-24  Adam Barth  <abarth@webkit.org>
646
647         Reviewed by Eric Seidel.
648
649         Atomize strings used by the HTML5 tree builder
650         https://bugs.webkit.org/show_bug.cgi?id=42875
651
652         This is about a 1% improvement to the parsing benchmark.
653
654         * html/HTMLTagNames.in:
655         * html/HTMLTreeBuilder.cpp:
656         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
657         (WebCore::HTMLTreeBuilder::processStartTag):
658         (WebCore::HTMLTreeBuilder::processEndTagForInBody):
659         (WebCore::HTMLTreeBuilder::processStartTagForInHead):
660
661 2010-07-24  Patrick Gansterer  <paroga@paroga.com>
662
663         Reviewed by Darin Adler.
664
665         Buildfix for !ENABLE(INSPECTOR) after r63891.
666         https://bugs.webkit.org/show_bug.cgi?id=42925
667
668         * bindings/js/ScriptValue.cpp: Added property svn:eol-style.
669         * bindings/js/ScriptValue.h: Added property svn:eol-style.
670         * inspector/ConsoleMessage.cpp: Added property svn:eol-style.
671         * inspector/ConsoleMessage.h: Added property svn:eol-style.
672
673 2010-07-24  Adam Barth  <abarth@webkit.org>
674
675         Reviewed by Eric Seidel.
676
677         Re-order whitespace checks in the tokenizer
678         https://bugs.webkit.org/show_bug.cgi?id=42923
679
680         This saves about 0.5% on the parsing benchmark.
681
682         * html/HTMLTokenizer.cpp:
683
684 2010-07-24  Andreas Kling  <andreas.kling@nokia.com>
685
686         Reviewed by Simon Hausmann.
687
688         [Qt] tst_QWebFrame::callQtInvokable() fails
689         https://bugs.webkit.org/show_bug.cgi?id=41065
690
691         Converting JS objects to QVariantMaps was broken.
692         This is a partial revert of <http://trac.webkit.org/changeset/61478>
693
694         * bridge/qt/qt_runtime.cpp:
695         (JSC::Bindings::convertValueToQVariant): Remove erroneous crash guard
696         around JSObject::getPropertyNames()
697
698 2010-07-24  Adam Barth  <abarth@webkit.org>
699
700         Reviewed by Eric Seidel.
701
702         Save a per-character branch in the HTML parser
703         https://bugs.webkit.org/show_bug.cgi?id=42921
704
705         The parsing benchmark claims this is a wash, but it makes me feel
706         better.
707
708         * html/HTMLToken.h:
709         (WebCore::HTMLToken::clear):
710         (WebCore::HTMLToken::beginStartTag):
711         (WebCore::HTMLToken::beginEndTag):
712         (WebCore::HTMLToken::ensureIsCharacterToken):
713         (WebCore::HTMLToken::beginComment):
714         (WebCore::HTMLToken::beginDOCTYPE):
715         * html/HTMLTokenizer.cpp:
716         (WebCore::HTMLTokenizer::bufferCharacter):
717
718 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
719
720         Reviewed by Darin Adler.
721
722         Buildfix for !ENABLE(FILTERS) after r63531.
723         https://bugs.webkit.org/show_bug.cgi?id=42914
724
725         * rendering/RenderSVGContainer.cpp: Add missing header.
726
727 2010-07-23  Kavita Kanetkar  <kkanetkar@chromium.org>
728
729         Reviewed by Joseph Pecoraro.
730
731         Implement remaining Inspector support for chrome's appcache
732         https://bugs.webkit.org/show_bug.cgi?id=42426
733
734         * inspector/InspectorApplicationCacheAgent.cpp:
735         (WebCore::InspectorApplicationCacheAgent::getApplicationCaches):
736         (WebCore::InspectorApplicationCacheAgent::buildObjectForApplicationCache):
737         (WebCore::InspectorApplicationCacheAgent::buildArrayForApplicationCacheResources):
738         (WebCore::InspectorApplicationCacheAgent::buildObjectForApplicationCacheResource):
739         * inspector/InspectorApplicationCacheAgent.h:
740         (WebCore::InspectorController::connectFrontend):
741         * loader/appcache/ApplicationCacheHost.cpp:
742         (WebCore::ApplicationCacheHost::fillResourceList):
743         (WebCore::ApplicationCacheHost::applicationCacheInfo):
744         * loader/appcache/ApplicationCacheHost.h: Changed creation + update times to double and moved appcache info here from InspectorApplicationCacheAgent.h
745         (WebCore::ApplicationCacheHost::CacheInfo::CacheInfo):
746         (WebCore::ApplicationCacheHost::ResourceInfo::ResourceInfo):
747
748 2010-07-23  Rafael Antognolli  <antognolli@profusion.mobi>
749
750         Reviewed by Antonio Gomes.
751
752         [EFL] Cleanup glib support (make it optional)
753         https://bugs.webkit.org/show_bug.cgi?id=42480
754
755         No new tests, no new features added.
756         Just add flags and directories for glib if necessary.
757         Put some ifdefs on glib calls.
758
759         * CMakeListsEfl.txt:
760         * platform/graphics/cairo/FontPlatformDataCairo.cpp:
761         (WebCore::FontPlatformData::FontPlatformData):
762
763 2010-07-23  James Robinson  <jamesr@chromium.org>
764
765         Reviewed by Dimitri Glazkov.
766
767         CanvasRenderingContext2D has lots of style errors
768         https://bugs.webkit.org/show_bug.cgi?id=42918
769
770         CanvasRenderingContext2D.h and .cpp have lots of style errors.  I'm planning to post patches
771         to both of these files soon, so to minimize merge conflicts and additional style violations
772         it seems best to fix all of these errors in bulk before posting any functional changes.  In
773         particular the header file is incorrectly indented so any correctly formatted addition will
774         be out of place.
775
776         * html/canvas/CanvasRenderingContext2D.cpp:
777         (WebCore::CanvasRenderingContext2D::CanvasRenderingContext2D):
778         (WebCore::CanvasRenderingContext2D::setFillStyle):
779         (WebCore::CanvasRenderingContext2D::transform):
780         (WebCore::CanvasRenderingContext2D::setTransform):
781         (WebCore::CanvasRenderingContext2D::arcTo):
782         (WebCore::CanvasRenderingContext2D::arc):
783         (WebCore::validateRectForCanvas):
784         (WebCore::CanvasRenderingContext2D::strokeRect):
785         (WebCore::CanvasRenderingContext2D::drawImage):
786         (WebCore::CanvasRenderingContext2D::createRadialGradient):
787         (WebCore::CanvasRenderingContext2D::createPattern):
788         (WebCore::CanvasRenderingContext2D::willDraw):
789         (WebCore::CanvasRenderingContext2D::createImageData):
790         (WebCore::CanvasRenderingContext2D::putImageData):
791         (WebCore::CanvasRenderingContext2D::setFont):
792         (WebCore::CanvasRenderingContext2D::drawTextInternal):
793         * html/canvas/CanvasRenderingContext2D.h:
794         (WebCore::CanvasRenderingContext2D::is2d):
795         (WebCore::CanvasRenderingContext2D::getLineCap):
796         (WebCore::CanvasRenderingContext2D::getLineJoin):
797         (WebCore::CanvasRenderingContext2D::state):
798         (WebCore::CanvasRenderingContext2D::):
799
800 2010-07-23  Adam Barth  <abarth@webkit.org>
801
802         Remove FIXME that we already fixed.
803
804         * html/HTMLTreeBuilder.cpp:
805         (WebCore::HTMLTreeBuilder::processCharacter):
806
807 2010-06-24  Dimitri Glazkov  <dglazkov@chromium.org>
808
809         Reviewed by Kent Tamura.
810
811         Move more form submission logic into FormSubmission.
812         https://bugs.webkit.org/show_bug.cgi?id=41110
813
814         Two major moves:
815         * Fold HTMLFormElement::prepareFormSubmission into FormSubmission::create
816         * Move non-static (non-FormData-building) methods of FormDataBuilder
817             to FormSubmission::Attributes.
818
819         Mechanical moving, no change in behavior, so no new tests.
820
821         * WebCore.xcodeproj/project.pbxproj: Gave FormDataBuilder.h "project" and
822             FormSubmission.h "private" visibility.
823         * html/HTMLFormElement.cpp:
824         (WebCore::HTMLFormElement::submit): Changed to use FormSubmission::create.
825         (WebCore::HTMLFormElement::parseMappedAttribute): Changed to use m_attributes.
826         * html/HTMLFormElement.h:
827         (WebCore::HTMLFormElement::enctype): Ditto.
828         (WebCore::HTMLFormElement::encoding): Ditto.
829         (WebCore::HTMLFormElement::acceptCharset): Ditto.
830         * loader/FormSubmission.cpp:
831         (WebCore::generateFormDataIdentifier): Moved from HTMLFormElement.
832         (WebCore::appendMailtoPostFormDataToURL): Ditto.
833         (WebCore::FormSubmission::Attributes::parseAction): Moved from FormDataBuilder.
834         (WebCore::FormSubmission::Attributes::parseEncodingType): Ditto.
835         (WebCore::FormSubmission::Attributes::parseMethodType): Ditto.
836         (WebCore::FormSubmission::FormSubmission): Made inline.
837         (WebCore::FormSubmission::create): Added logic that used to be in
838             HTMLFormElement::prepareFormSubmission.
839         * loader/FormSubmission.h:
840         (WebCore::FormSubmission::):
841         (WebCore::FormSubmission::Attributes::Attributes): Added new class.
842         (WebCore::FormSubmission::Attributes::method): Moved from FormDataBuilder.
843         (WebCore::FormSubmission::Attributes::isMultiPartForm): Ditto.
844         (WebCore::FormSubmission::Attributes::action): Ditto.
845         (WebCore::FormSubmission::Attributes::target): Ditto.
846         (WebCore::FormSubmission::Attributes::setTarget): Ditto.
847         (WebCore::FormSubmission::Attributes::encodingType): Ditto.
848         (WebCore::FormSubmission::Attributes::acceptCharset): Ditto.
849         (WebCore::FormSubmission::Attributes::setAcceptCharset): Ditto.
850         (WebCore::FormSubmission::target): Ditto.
851         (WebCore::FormSubmission::contentType): Const-ified return type.
852         (WebCore::FormSubmission::boundary): Ditto.
853         * platform/network/FormDataBuilder.cpp:
854         (WebCore::FormDataBuilder::encodingFromAcceptCharset): Renamed method for clarity,
855             made it use param instead of a class member.
856         * platform/network/FormDataBuilder.h:
857         (WebCore::FormDataBuilder::FormDataBuilder): Made constructor private.
858         * wml/WMLGoElement.cpp: Changed to use FormSubmision::Attributes.
859         * wml/WMLGoElement.h: Ditto.
860
861 2010-07-23  Adam Barth  <abarth@webkit.org>
862
863         Reviewed by Eric Seidel.
864
865         HTML5 Parser: document.write in a asynchronous script which is specified to load before page finish blows away document
866         https://bugs.webkit.org/show_bug.cgi?id=42365
867
868         This patch implements the changes Ian made to HTML5 to address these
869         issues.  Let's try it and see how well it works.
870
871         Test: http/tests/misc/write-from-dom-script.html
872
873         * dom/Document.cpp:
874         (WebCore::Document::Document):
875         (WebCore::Document::write):
876         * dom/Document.h:
877         (WebCore::Document::writeDisabled):
878         (WebCore::Document::setWriteDisabled):
879         * dom/DocumentParser.h:
880         (WebCore::DocumentParser::hasInsertionPoint):
881         * dom/ScriptElement.cpp:
882         (WebCore::ScriptElement::insertedIntoDocument):
883         (WebCore::ScriptElementData::ScriptElementData):
884         (WebCore::ScriptElementData::evaluateScript):
885         * dom/ScriptElement.h:
886         (WebCore::ScriptElementData::writeDisabled):
887         (WebCore::ScriptElementData::setWriteDisabled):
888         * html/HTMLDocumentParser.cpp:
889         (WebCore::HTMLDocumentParser::hasInsertionPoint):
890         (WebCore::HTMLDocumentParser::insert):
891         * html/HTMLDocumentParser.h:
892
893 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
894
895         Reviewed by Adam Roben.
896
897         [WINCE] Buildfix for PopupMenuWin.cpp
898         https://bugs.webkit.org/show_bug.cgi?id=42906
899
900         Use correct API for BitmapInfo landed in r63944.
901
902         * platform/win/PopupMenuWin.cpp:
903         (WebCore::PopupMenu::paint):
904
905 2010-07-23  fsamuel@chromium.org  <fsamuel@chromium.org>
906
907         Reviewed by David Hyatt.
908
909         Updated table painting to draw cell layers in the correct order.
910
911         Table Cell Layering
912         https://bugs.webkit.org/show_bug.cgi?id=40775
913
914         Tests: tables/layering/paint-test-layering-1.html
915                tables/layering/paint-test-layering-2.html
916
917         * accessibility/AccessibilityTable.cpp:
918         (WebCore::AccessibilityTable::isTableExposableThroughAccessibility):
919         (WebCore::AccessibilityTable::addChildren):
920         (WebCore::AccessibilityTable::cellForColumnAndRow):
921         * accessibility/AccessibilityTableCell.cpp:
922         (WebCore::AccessibilityTableCell::titleUIElement):
923         * accessibility/AccessibilityTableColumn.cpp:
924         (WebCore::AccessibilityTableColumn::headerObjectForSection):
925         * rendering/AutoTableLayout.cpp:
926         (WebCore::AutoTableLayout::recalcColumn):
927         * rendering/RenderTable.cpp:
928         (WebCore::RenderTable::splitColumn):
929         (WebCore::RenderTable::calcBorderLeft):
930         (WebCore::RenderTable::calcBorderRight):
931         (WebCore::RenderTable::cellAbove):
932         (WebCore::RenderTable::cellBelow):
933         (WebCore::RenderTable::cellBefore):
934         (WebCore::RenderTable::cellAfter):
935         * rendering/RenderTableSection.cpp:
936         (WebCore::RenderTableSection::RenderTableSection):
937         (WebCore::RenderTableSection::ensureRows):
938         (WebCore::RenderTableSection::addCell):
939         (WebCore::RenderTableSection::setCellWidths):
940         (WebCore::RenderTableSection::calcRowHeight):
941         (WebCore::RenderTableSection::layoutRows):
942         (WebCore::RenderTableSection::calcOuterBorderTop):
943         (WebCore::RenderTableSection::calcOuterBorderBottom):
944         (WebCore::RenderTableSection::calcOuterBorderLeft):
945         (WebCore::RenderTableSection::calcOuterBorderRight):
946         (WebCore::RenderTableSection::firstLineBoxBaseline):
947         (WebCore::compareCellPositions):
948         (WebCore::RenderTableSection::paintCell):
949         (WebCore::RenderTableSection::paintObject):
950         (WebCore::RenderTableSection::numColumns):
951         (WebCore::RenderTableSection::appendColumn):
952         (WebCore::RenderTableSection::splitColumn):
953         (WebCore::RenderTableSection::nodeAtPoint):
954         * rendering/RenderTableSection.h:
955         (WebCore::RenderTableSection::CellStruct::CellStruct):
956         (WebCore::RenderTableSection::CellStruct::primaryCell):
957         (WebCore::RenderTableSection::CellStruct::hasCells):
958         (WebCore::RenderTableSection::primaryCellAt):
959
960 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
961
962         Reviewed by Adam Roben.
963
964         [WINCE] Buildfix for Path.h
965         https://bugs.webkit.org/show_bug.cgi?id=42904
966
967         * platform/graphics/Path.h: Add missing typedef PlatformPath.
968
969 2010-07-23  Adam Barth  <abarth@webkit.org>
970
971         Reviewed by Eric Seidel.
972
973         Reserve capacity for attributes
974         https://bugs.webkit.org/show_bug.cgi?id=42877
975
976         This is about a 1.7% improvement to the parsing benchmark.
977
978         * html/HTMLToken.h:
979         (WebCore::AtomicHTMLToken::AtomicHTMLToken):
980
981 2010-07-23  Adam Barth  <abarth@webkit.org>
982
983         Reviewed by Eric Seidel.
984
985         Inline HTMLElementStack::top
986         https://bugs.webkit.org/show_bug.cgi?id=42879
987
988         According to sampling, HTMLElementStack::top is the hotest function in
989         the treebuilder.  Inlining it is about a 0.3% improvement to the
990         parsing benchmark.  This squares with how often it occurs in samples.
991
992         * html/HTMLElementStack.cpp:
993         * html/HTMLElementStack.h:
994         (WebCore::HTMLElementStack::top):
995
996 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
997
998         Reviewed by Adam Roben.
999
1000         [WINCE] Buildfix for GraphicsContext
1001         https://bugs.webkit.org/show_bug.cgi?id=42888
1002
1003         Use AffineTransform insted of TransformationMatrix and
1004         add ColorSpace parameter to the drawBitmap methods.
1005
1006         * platform/graphics/GraphicsContext.h:
1007         * platform/graphics/wince/GraphicsContextWince.cpp:
1008         (WebCore::GraphicsContext::affineTransform):
1009         (WebCore::GraphicsContext::fillRoundedRect):
1010         (WebCore::GraphicsContext::drawBitmap):
1011         (WebCore::GraphicsContext::drawBitmapPattern):
1012
1013 2010-07-23  Rafael Antognolli  <antognolli@profusion.mobi>
1014
1015         Reviewed by Antonio Gomes.
1016
1017         [EFL] Add support for using libcurl network backend.
1018         https://bugs.webkit.org/show_bug.cgi?id=42286
1019
1020         Add files and flags for soup/curl specific backend being
1021         used.
1022         Put some ifdefs around soup/glib calls.
1023
1024         No new tests, no new features implemented.
1025
1026         * CMakeListsEfl.txt:
1027         * platform/efl/FileSystemEfl.cpp:
1028         (WebCore::fileSystemRepresentation):
1029
1030 2010-07-23  Adam Barth  <abarth@webkit.org>
1031
1032         Reviewed by Eric Seidel.
1033
1034         Add fast-reject path for InputStreamPreprocessor::peek
1035         https://bugs.webkit.org/show_bug.cgi?id=42881
1036
1037         This is about a 2.5% speedup on the parsing benchmark.
1038
1039         * html/HTMLTokenizer.h:
1040         (WebCore::HTMLTokenizer::InputStreamPreprocessor::peek):
1041
1042 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
1043
1044         Reviewed by Adam Roben.
1045
1046         [WINCE] Merge PasteboardWince.cpp into PasteboardWin.cpp
1047         https://bugs.webkit.org/show_bug.cgi?id=42885
1048
1049         Make PasteboardWin more portable and remove the global namespace prefix
1050         from GlobalLock/GlobalUnlock, since they are only macros on WinCE.
1051
1052         * platform/win/PasteboardWin.cpp:
1053         (WebCore::PasteboardOwnerWndProc):
1054         (WebCore::Pasteboard::Pasteboard):
1055         (WebCore::Pasteboard::plainText):
1056         (WebCore::Pasteboard::documentFragment):
1057
1058 2010-07-23  Justin Schuh  <jschuh@chromium.org>
1059
1060         Unreviewed, rolling out r63925.
1061         http://trac.webkit.org/changeset/63925
1062         https://bugs.webkit.org/show_bug.cgi?id=42858
1063
1064         Introduced a regression in history state.
1065
1066         * page/History.cpp:
1067         (WebCore::History::urlForState):
1068
1069 2010-07-23  Martin Robinson  <mrobinson@igalia.com>
1070
1071         Unreviewed build fix.
1072
1073         Fix make dist by aligning the source list with currently used headers.
1074
1075         * GNUmakefile.am:
1076
1077 2010-07-23  Alex Milowski  <alex@milowski.com>
1078
1079         Reviewed by Darin Adler.
1080
1081         Moved the 'mathml' group back to where it belongs as
1082         siblings of the other major folders.
1083
1084         * WebCore.xcodeproj/project.pbxproj:
1085
1086 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
1087
1088         Reviewed by Adam Roben.
1089
1090         Buildfix for WinCE graphics after r56015.
1091         https://bugs.webkit.org/show_bug.cgi?id=41354
1092
1093         Use fabs() insetead of abs().
1094
1095         * platform/graphics/wince/GraphicsContextWince.cpp:
1096         (WebCore::GraphicsContext::fillRect):
1097         * platform/graphics/wince/PlatformPathWince.cpp:
1098         (WebCore::PlatformPath::addArcTo):
1099
1100 2010-07-23  ryuan choi  <ryuan.choi@samsung.com>
1101
1102         Reviewed by Antonio Gomes.
1103
1104         [EFL]build break because of r63927
1105         https://bugs.webkit.org/show_bug.cgi?id=42864
1106
1107         * CMakeLists.txt:
1108
1109 2010-07-23  Adam Barth  <abarth@webkit.org>
1110
1111         Reviewed by Eric Seidel.
1112
1113         Inline HTMLTokenizer::InputStreamPreprocessor::peek
1114         https://bugs.webkit.org/show_bug.cgi?id=42880
1115
1116         According to sampling HTMLTokenizer::InputStreamPreprocessor::peek is
1117         the hotest function in the tokenizer.  Inlining this function is about
1118         a 2.2% speedup on the parser benchmark.  Turns out the tokenizer is
1119         much hotter than the tree builder.  :)
1120
1121         * html/HTMLTokenizer.h:
1122         (WebCore::HTMLTokenizer::InputStreamPreprocessor::peek):
1123
1124 2010-07-23  Andras Becsi  <abecsi@webkit.org>
1125
1126         Unreviewed build fix.
1127
1128         [Qt] Fix WebKit2 linkage.
1129
1130         No new tests needed.
1131
1132         * WebCore.pro: add InjectedBundlePageEditorClient.cpp to WK2 sources
1133
1134 2010-07-23  Balazs Kelemen  <kb@inf.u-szeged.hu>
1135
1136         Reviewed by Kenneth Rohde Christiansen.
1137
1138         [Qt] Setup the QtWebProcess
1139
1140         https://bugs.webkit.org/show_bug.cgi?id=42623
1141
1142         * WebCore.pro: Added WebProcessMainQt.cpp to SOURCES.
1143
1144 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1145
1146         Not reviewed: Chromium tests fix.
1147
1148         * inspector/CodeGeneratorInspector.pm:
1149
1150 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
1151
1152         Reviewed by Adam Roben.
1153
1154         [WINCE] Buildfix for PluginDatabaseWin.cpp
1155         https://bugs.webkit.org/show_bug.cgi?id=41887
1156
1157         Define KEY_ENUMERATE_SUB_KEYS only if not defined.
1158
1159         * plugins/win/PluginDatabaseWin.cpp:
1160
1161 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1162
1163         Reviewed by Yury Semikhatsky.
1164
1165         Chromium DevTools: migrate from utility context-based injected
1166         script dispatch to upstream schema.
1167
1168         https://bugs.webkit.org/show_bug.cgi?id=42883
1169
1170         * bindings/v8/ScriptProfile.cpp:
1171         (WebCore::ScriptProfile::title):
1172         * inspector/InspectorController.cpp:
1173         (WebCore::InspectorController::getProfile):
1174         * inspector/front-end/InjectedScriptAccess.js:
1175
1176 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1177
1178         Not reviewed: rolling out r63963.
1179
1180         * bindings/v8/ScriptProfile.cpp:
1181         (WebCore::ScriptProfile::title):
1182         * inspector/InspectorController.cpp:
1183         (WebCore::InspectorController::getProfile):
1184         * inspector/front-end/InjectedScriptAccess.js:
1185
1186 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1187
1188         Reviewed by Yury Semikhatsky.
1189
1190         Chromium DevTools: migrate from utility context-based injected
1191         script dispatch to upstream schema.
1192
1193         https://bugs.webkit.org/show_bug.cgi?id=42883
1194
1195         * bindings/v8/ScriptProfile.cpp:
1196         (WebCore::ScriptProfile::title):
1197         * inspector/InspectorController.cpp:
1198         (WebCore::InspectorController::getProfile):
1199         * inspector/front-end/InjectedScriptAccess.js:
1200         (InjectedScriptAccess.get if):
1201         (InjectedScriptAccess):
1202
1203 2010-07-23  Yury Semikhatsky  <yurys@chromium.org>
1204
1205         Reviewed by Pavel Feldman.
1206
1207         Web Inspector: error info propagation in InspectorBackendDispatch should be improved
1208         https://bugs.webkit.org/show_bug.cgi?id=42873
1209
1210         * inspector/CodeGeneratorInspector.pm: report wrong parameter count and wrong
1211         parametere type errror to the dispatch method caller
1212
1213 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1214
1215         Reviewed by Yury Semikhatsky.
1216
1217         Web Inspector: injectedScriptId is missing in the stack frame wrappers.
1218
1219         https://bugs.webkit.org/show_bug.cgi?id=42889
1220
1221         * inspector/front-end/InjectedScript.js:
1222         (injectedScriptConstructor):
1223
1224 2010-07-23  Shinichiro Hamaji  <hamaji@chromium.org>
1225
1226         Reviewed by Kent Tamura.
1227
1228         Backslash isn't transcoded in <input>
1229         https://bugs.webkit.org/show_bug.cgi?id=42878
1230
1231         * css/CSSStyleSelector.cpp:
1232         (WebCore::CSSStyleSelector::applyProperty):
1233
1234 2010-07-22  Ilya Tikhonovsky  <loislo@chromium.org>
1235
1236         Reviewed by Yury Semikhatsky.
1237
1238         WebInspector: CodeGeneratorInspector was extended and now it can
1239         generate InspectorBackedDispatcher's code. This dispatcher will
1240         validate and dispatch incoming inspector commands only by native
1241         code without using javaScript. That is necessary step for
1242         RemoteDebugging support.
1243         https://bugs.webkit.org/show_bug.cgi?id=42588
1244
1245         * GNUmakefile.am:
1246         * WebCore.gyp/WebCore.gyp:
1247         * WebCore.gypi:
1248         * WebCore.pri:
1249         * WebCore.vcproj/WebCore.vcproj:
1250         * WebCore.xcodeproj/project.pbxproj:
1251         * inspector/CodeGeneratorInspector.pm:
1252         * inspector/InspectorController.cpp:
1253         (WebCore::InspectorController::InspectorController):
1254         * inspector/InspectorController.h:
1255         (WebCore::InspectorController::inspectorBackendDispatcher):
1256         * inspector/front-end/ElementsPanel.js:
1257         (WebInspector.ElementsPanel.prototype.setDocument):
1258
1259 2010-07-22  Shinichiro Hamaji  <hamaji@chromium.org>
1260
1261         Reviewed by Kent Tamura.
1262
1263         Backslash is transcoded into yen sign even when non japanese font is specified
1264         https://bugs.webkit.org/show_bug.cgi?id=39654
1265
1266         Test: fast/text/backslash-to-yen-sign-euc.html
1267
1268         * css/CSSStyleSelector.cpp:
1269         (WebCore::CSSStyleSelector::applyProperty): Call FontDescriptor::setIsSpecifiedFont.
1270         * platform/graphics/Font.cpp:
1271         (WebCore::Font::Font):
1272         * platform/graphics/FontDescription.h: Added m_isSpecifiedFont and its accessors.
1273         (WebCore::FontDescription::FontDescription):
1274         (WebCore::FontDescription::isSpecifiedFont):
1275         (WebCore::FontDescription::setIsSpecifiedFont):
1276         (WebCore::FontDescription::operator==):
1277         * platform/text/transcoder/FontTranscoder.cpp:
1278         (WebCore::FontTranscoder::converterType): Don't transcode backslashes if a font family is explicitly specified.
1279         (WebCore::FontTranscoder::convert):
1280         (WebCore::FontTranscoder::needsTranscoding):
1281         * platform/text/transcoder/FontTranscoder.h: Change the arguments from AtomicString fontFamily to Font.
1282         * rendering/RenderText.cpp:
1283         (WebCore::RenderText::updateNeedsTranscoding):
1284         (WebCore::RenderText::setTextInternal):
1285
1286 2010-07-22  Jon Honeycutt  <jhoneycutt@apple.com>
1287
1288         Windows build fix.
1289
1290         Commit some changes missed by the commit-queue in r63944.
1291
1292         * platform/win/BitmapInfo.h:
1293         (WebCore::BitmapInfo::is16bit):
1294         (WebCore::BitmapInfo::is32bit):
1295         (WebCore::BitmapInfo::width):
1296         (WebCore::BitmapInfo::height):
1297         (WebCore::BitmapInfo::size):
1298         (WebCore::BitmapInfo::paddedWidth):
1299         (WebCore::BitmapInfo::numPixels):
1300         (WebCore::BitmapInfo::paddedBytesPerLine):
1301         (WebCore::BitmapInfo::bytesPerLine):
1302
1303 2010-07-22  Patrick Gansterer  <paroga@paroga.com>
1304
1305         Reviewed by Adam Roben.
1306
1307         [WINCE] Add additonal methods to BitmapInfo.
1308         https://bugs.webkit.org/show_bug.cgi?id=42071
1309
1310         Added a parameter to create 16bit BitmapInfo and
1311         some additional accessors for the WinCE port.
1312
1313         Set bmiHeader.biSize only at constructor.
1314
1315         * platform/win/BitmapInfo.cpp: Added property svn:eol-style.
1316         (WebCore::bitmapInfoForSize):
1317         (WebCore::BitmapInfo::create):
1318         (WebCore::BitmapInfo::createBottomUp):
1319         * platform/win/BitmapInfo.h: Added property svn:eol-style.
1320         (WebCore::BitmapInfo::is16bit):
1321         (WebCore::BitmapInfo::is32bit):
1322         (WebCore::BitmapInfo::width):
1323         (WebCore::BitmapInfo::height):
1324         (WebCore::BitmapInfo::size):
1325         (WebCore::BitmapInfo::paddedWidth):
1326         (WebCore::BitmapInfo::numPixels):
1327         (WebCore::BitmapInfo::paddedBytesPerLine):
1328         (WebCore::BitmapInfo::bytesPerLine):
1329
1330 2010-07-22  MORITA Hajime  <morrita@google.com>
1331
1332         Reviewed by Ojan Vafai.
1333
1334         Autoscroll on selection should take scrollbars into account as outside of the content.
1335         https://bugs.webkit.org/show_bug.cgi?id=40403
1336
1337         ScrollView's scroll-bar was considered as a part of content rect,
1338         that prevents autoscroll to happen when the pointer is over the
1339         scroll-bar.  This change excluded scroll-bar region from scrolling
1340         delta calculation.
1341         
1342         Test: WebCore/manual-tests/autoscroll-over-scrollbar.html
1343
1344         * rendering/RenderLayer.cpp:
1345         (WebCore::RenderLayer::scrollRectToVisible):
1346
1347 2010-07-22  Kent Tamura  <tkent@chromium.org>
1348
1349         Reviewed by Darin Adler.
1350
1351         maxLength should not be applied to non-text types
1352         https://bugs.webkit.org/show_bug.cgi?id=42814
1353
1354         According to the HTML5 specification, maxLength attribute should
1355         be applied to text, search, url, tel, email, and password types.
1356
1357         Test: fast/forms/input-maxlength-unsupported.html
1358
1359         * dom/InputElement.cpp:
1360         (WebCore::InputElement::handleBeforeTextInsertedEvent):
1361          Apply maxLength() only if supportsMaxLength() is true.
1362         * dom/InputElement.h:
1363         * html/HTMLInputElement.cpp:
1364         (WebCore::HTMLInputElement::patternMismatch):
1365          Code cleanup by isTextType().
1366         (WebCore::HTMLInputElement::tooLong):
1367          ditto.
1368         (WebCore::HTMLInputElement::isTextType):
1369          Added. This returns true for types with maxLength/pattern/placeholder.
1370         * html/HTMLInputElement.h:
1371         (WebCore::HTMLInputElement::supportsMaxLength):
1372          Added. Just calls isTextType().
1373         (WebCore::HTMLInputElement::supportsPlaceholder):
1374          Code cleanup by isTextType().
1375         * wml/WMLInputElement.h:
1376         (WebCore::WMLInputElement::supportsMaxLength):
1377          Added. Always returns true.
1378
1379 2010-07-22  Ryosuke Niwa  <rniwa@webkit.org>
1380
1381         Reviewed by Kent Tamura.
1382
1383         [cleanup] logic to extract adjacent lists and list children in listifyParagraph and unlistifyParagraph should be isolated
1384         https://bugs.webkit.org/show_bug.cgi?id=42841
1385
1386         Added enclosingListChild that takes the list node containing the node for the second argument
1387         to ensure that it only returns the enclosing list child of the specified list node.
1388
1389         Added adjacentEnclosingList that finds adjacent lists of the specified position.
1390         It takes the current position, the adjacent position that may or may not have a list, and the list type,
1391         and avoids a list that belongs to a different table cell or an outer list that contains the current position.
1392
1393         No new tests needed.
1394
1395         * editing/InsertListCommand.cpp:
1396         (WebCore::enclosingListChild): Added. See above.
1397         (WebCore::InsertListCommand::unlistifyParagraph): Uses enclosingListChild.
1398         (WebCore::adjacentEnclosingList): Added. See above.
1399         (WebCore::InsertListCommand::listifyParagraph): Uses adjacentEnclosingList.
1400
1401 2010-07-22  Patrick Gansterer  <paroga@paroga.com>
1402
1403         Reviewed by Adam Roben.
1404
1405         [WIN] Correct DeleteObject usage in PasteboardWin
1406         https://bugs.webkit.org/show_bug.cgi?id=41956
1407
1408         Call DeleteObject() if OpenClipboard() fails and don't
1409         call DeleteObject() on objects returned from SelectObject().
1410
1411         * platform/win/PasteboardWin.cpp:
1412         (WebCore::Pasteboard::writeImage):
1413
1414 2010-07-22  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
1415
1416         Reviewed by Antonio Gomes.
1417
1418         [EFL] Log function is incorrectly used.
1419         r63804 used LOG function, but it should receive one more parameter.
1420         Used LOG_ERROR instead, since this is an error message.
1421         https://bugs.webkit.org/show_bug.cgi?id=42842
1422
1423         * platform/efl/WidgetEfl.cpp:
1424         (WebCore::Widget::applyFallbackCursor): use LOG_ERROR in place of LOG
1425
1426 2010-07-22  Kent Tamura  <tkent@chromium.org>
1427
1428         Reviewed by Ojan Vafai.
1429
1430         Cancel button of search input field should fire input/change events
1431         https://bugs.webkit.org/show_bug.cgi?id=34393
1432
1433         Test: fast/forms/search-cancel-button-events.html
1434
1435         * rendering/TextControlInnerElements.cpp:
1436         (WebCore::SearchFieldCancelButtonElement::defaultEventHandler):
1437
1438 2010-07-22  Sam Weinig  <sam@webkit.org>
1439
1440         Reviewed by Maciej Stachowiak.
1441
1442         Fix for <rdar://problem/8222626>
1443         Send textDidChangeInTextField delegate callback only in response to typing or other forms of user text input.
1444
1445         * GNUmakefile.am:
1446         * WebCore.gypi:
1447         * WebCore.pro:
1448         * WebCore.vcproj/WebCore.vcproj:
1449         * WebCore.xcodeproj/project.pbxproj:
1450         Update project files.
1451         
1452         * dom/UserTypingGestureIndicator.cpp: Copied from dom/UserGestureIndicator.cpp.
1453         (WebCore::UserTypingGestureIndicator::processingUserTypingGesture):
1454         (WebCore::focusedNode):
1455         (WebCore::UserTypingGestureIndicator::focusedElementAtGestureStart):
1456         (WebCore::UserTypingGestureIndicator::UserTypingGestureIndicator):
1457         (WebCore::UserTypingGestureIndicator::~UserTypingGestureIndicator):
1458         * dom/UserTypingGestureIndicator.h: Copied from dom/UserGestureIndicator.h.
1459         Version of UserGestureIndicator that just tracks keyboard gestures and the focused
1460         node they were targeting.
1461         
1462         * editing/Editor.cpp:
1463         (WebCore::Editor::confirmComposition):
1464         (WebCore::Editor::setComposition):
1465         * page/EventHandler.cpp:
1466         (WebCore::EventHandler::keyEvent):
1467         Put UserTypingGestureIndicator on the stack for user generated keyboard events.
1468
1469 2010-07-22  Kent Tamura  <tkent@chromium.org>
1470
1471         Reviewed by Ojan Vafai.
1472
1473         <input type=number> stepper buttons should dispatch input/change events
1474         https://bugs.webkit.org/show_bug.cgi?id=42440
1475
1476         Test: fast/forms/input-number-events.html
1477
1478         * html/HTMLInputElement.cpp:
1479         (WebCore::HTMLInputElement::stepUpFromRenderer):
1480          Sets a flag to dispatch 'change' event and dispatches 'input' event
1481          if the value is changed.
1482
1483 2010-07-22  Justin Schuh  <jschuh@chromium.org>
1484
1485         Reviewed by Adam Barth.
1486
1487         Make history state methods validate against document URL
1488         https://bugs.webkit.org/show_bug.cgi?id=42858
1489
1490         Test: fast/history/history-replace-illegal.html
1491
1492         * page/History.cpp:
1493         (WebCore::History::urlForState):
1494
1495 2010-07-22  Andy Estes  <aestes@apple.com>
1496
1497         Reviewed by Maciej Stachowiak.
1498
1499         When there are no beforeload listeners on a document,
1500         HTMLLinkElement::process() should be called immediately when the node is
1501         inserted into the document, rather than waiting until after attach is
1502         performed.
1503         <https://bugs.webkit.org/show_bug.cgi?id=42859>
1504         <rdar://problem/8194528>
1505
1506         No new tests.
1507
1508         * html/HTMLLinkElement.cpp:
1509         (WebCore::HTMLLinkElement::insertedIntoDocument): If there are no
1510         beforeload listeners, call process() immediately.  Otherwise, add it as
1511         a post-attach callback.
1512
1513 2010-07-22  James Robinson  <jamesr@chromium.org>
1514
1515         Reviewed by Simon Fraser.
1516
1517         REGRESSION: box shadows on WebGL canvas elements repainted every frame
1518         https://bugs.webkit.org/show_bug.cgi?id=42852
1519
1520         Avoid calling HTMLCanvasElement::willDraw() every time a WebGL canvas'
1521         contents update.  This call is only needed in the non-compositing
1522         rendering path.
1523
1524         Covered by fast/canvas/webgl/canvas-test.html
1525
1526         * html/canvas/WebGLRenderingContext.cpp:
1527         (WebCore::WebGLRenderingContext::markContextChanged):
1528         (WebCore::WebGLRenderingContext::paintRenderingResultsToCanvas):
1529
1530 2010-07-22  Patrick Gansterer  <paroga@paroga.com>
1531
1532         Reviewed by Ojan Vafai.
1533
1534         Remove useless PlatformScrollBar
1535         https://bugs.webkit.org/show_bug.cgi?id=41911
1536
1537         * WebCore.gypi:
1538         * platform/win/PlatformScrollBar.h: Removed.
1539         * platform/win/PlatformScrollBarWin.cpp: Removed.
1540
1541 2010-07-22  Ryosuke Niwa  <rniwa@webkit.org>
1542
1543         Reviewed by Ojan Vafai.
1544
1545         [Windows] Home hey doesn't work in first DIV inside a TABLE
1546         https://bugs.webkit.org/show_bug.cgi?id=42767
1547
1548         The bug was caused by positionAvoidingFirstPositionInTable moving outside of the editing region.
1549         Added a check so that it avoids the first position in the table only if the table is editable.
1550
1551         Test: editing/selection/home-inside-noneditable-table.html
1552
1553         * editing/visible_units.cpp:
1554         (WebCore::positionAvoidingFirstPositionInTable):
1555
1556 2010-07-22  Antonio Gomes  <tonikitoo@webkit.org>
1557
1558         Reviewed by Laszlo Gombos.
1559
1560         Fix build warning for missing parentheses.
1561
1562         WebCore/html/HTMLInputElement.cpp:113: warning: suggest parentheses around '&&' within '||'
1563
1564         * html/HTMLInputElement.cpp:
1565         (WebCore::isNumberCharacter):
1566
1567 2010-07-22  John Gregg  <johnnyg@google.com>
1568
1569         Reviewed by Kent Tamura.
1570
1571         [chromium] fast/forms/input-file-directory-upload.html fails on win after r63454
1572         https://bugs.webkit.org/show_bug.cgi?id=42768
1573
1574         Normalize path separators for the script API by convering \'s to /'s.
1575
1576         Covered by input-file-directory-upload.html
1577
1578         * html/HTMLInputElement.cpp:
1579         (WebCore::HTMLInputElement::setFileListFromRenderer):
1580
1581 2010-07-22  Dmitry Titov  <dimich@chromium.org>
1582
1583         Rubberstamped by Simon Fraser.
1584
1585         * platform/PlatformStrategies.h:
1586         (WebCore::PlatformStrategies::PlatformStrategies): Add initialization of m_localizationStrategy.
1587
1588 2010-07-22  Dumitru Daniliuc  <dumi@chromium.org>
1589
1590         Reviewed by Ojan Vafai.
1591
1592         Make db.changeVersion()'s third parameter optional, as required by the spec.
1593         https://bugs.webkit.org/show_bug.cgi?id=42787
1594
1595         * bindings/js/JSDatabaseCustom.cpp:
1596         (WebCore::JSDatabase::changeVersion):
1597         * bindings/js/JSDatabaseSyncCustom.cpp:
1598         (WebCore::JSDatabaseSync::changeVersion):
1599         * bindings/v8/custom/V8DatabaseCustom.cpp:
1600         (WebCore::V8Database::changeVersionCallback):
1601         * bindings/v8/custom/V8DatabaseSyncCustom.cpp:
1602         (WebCore::V8DatabaseSync::changeVersionCallback):
1603         * storage/SQLTransaction.cpp:
1604         (WebCore::SQLTransaction::deliverTransactionCallback):
1605         * storage/SQLTransactionSync.cpp:
1606         (WebCore::SQLTransactionSync::execute):
1607
1608 2010-07-20  Jessie Berlin  <jberlin@apple.com>
1609
1610         Reviewed by Darin Adler.
1611
1612         Code Generator: Allow negative and string constants.
1613         Also add tests for hexadecimal numbers.
1614
1615         * bindings/scripts/CodeGeneratorJS.pm:
1616         If it is a string, do not try to cast it to an int.
1617
1618         * bindings/scripts/IDLStructure.pm:
1619         Add a regex just for constant values that allows strings, hexadecimal numbers, and integral numbers.
1620
1621         * bindings/scripts/test/CPP/WebDOMTestObj.h:
1622         Update the tests.
1623
1624         * bindings/scripts/test/JS/JSTestObj.cpp:
1625         Ditto.
1626
1627         * bindings/scripts/test/JS/JSTestObj.h:
1628         Ditto.
1629
1630         * bindings/scripts/test/ObjC/DOMTestObj.h:
1631         Ditto.
1632
1633         * bindings/scripts/test/TestObj.idl:
1634         Add tests for negative numbers, strings, and hexadecimal numbers.
1635
1636         * bindings/scripts/test/V8/V8TestObj.cpp:
1637         Update the tests.
1638
1639 2010-07-22  Simon Fraser  <simon.fraser@apple.com>
1640
1641         Reviewed by Darin Adler.
1642
1643         Composited layers don't scroll in WebKit2
1644         https://bugs.webkit.org/show_bug.cgi?id=42771
1645
1646         More prep work: make a single scrollPositionChanged() bottleneck method, and use it to 
1647         send the scroll event, repaint fixed elements, and update compositing layers.
1648
1649         * page/FrameView.cpp:
1650         (WebCore::FrameView::scrollPositionChangedViaPlatformWidget):
1651         (WebCore::FrameView::scrollPositionChanged):
1652         (WebCore::FrameView::repaintFixedElementsAfterScrolling):
1653         (WebCore::FrameView::valueChanged):
1654         * page/FrameView.h:
1655         * platform/ScrollView.cpp:
1656         (WebCore::ScrollView::valueChanged):
1657
1658 2010-07-22  Sam Weinig  <sam@webkit.org>
1659
1660         Reviewed by Anders Carlsson.
1661
1662         Patch for https://bugs.webkit.org/show_bug.cgi?id=42836
1663         Add localized strings stubs for WebKit2.
1664
1665         - Convert WebKit (mac) to use a LocalizationStrategy rather than the WebCoreViewFactory
1666           for localized strings.
1667
1668         Gets us below 200 tests failing in WebKitTestRunner.
1669
1670         * WebCore.exp.in:
1671         * WebCore.xcodeproj/project.pbxproj:
1672         * page/mac/WebCoreViewFactory.h:
1673         * platform/LocalizationStrategy.h: Added.
1674         (WebCore::LocalizationStrategy::~LocalizationStrategy):
1675         * platform/PlatformStrategies.h:
1676         (WebCore::PlatformStrategies::pluginStrategy):
1677         (WebCore::PlatformStrategies::localizationStrategy):
1678         (WebCore::PlatformStrategies::PlatformStrategies):
1679         (WebCore::PlatformStrategies::~PlatformStrategies):
1680         * platform/mac/LocalizedStringsMac.mm:
1681
1682 2010-07-22  Andras Becsi  <abecsi@webkit.org>
1683
1684         Reviewed by Antonio Gomes.
1685
1686         [Qt] Implement WebContext::applicationCacheDirectory() for the Qt port of WebKit2.
1687         https://bugs.webkit.org/show_bug.cgi?id=42830
1688
1689         No new tests needed.
1690
1691         * WebCore.pri: add ApplicationCacheStorage.h to wk2 headers
1692         * WebCore.pro: add WebContextQt.cpp to sources
1693
1694 2010-07-22  Joone Hur  <joone.hur@samsung.com>
1695
1696         Reviewed by Antonio Gomes.
1697
1698         [EFL] add support for logging
1699         https://bugs.webkit.org/show_bug.cgi?id=40807
1700
1701         Enable to turn log channels on for EFL port
1702
1703         * platform/efl/LoggingEfl.cpp:
1704         (WebCore::InitializeLoggingChannelsIfNecessary):
1705
1706 2010-07-22  Balazs Kelemen  <kb@inf.u-szeged.hu>
1707
1708         Ureviewed build fix.
1709
1710         [Qt] Fix WebKit2 build
1711
1712         https://bugs.webkit.org/show_bug.cgi?id=42638
1713
1714         * WebCore.pri:
1715         * WebCore.pro:
1716
1717 2010-07-22  Andrey Kosyakov  <caseq@chromium.org>
1718
1719         Reviewed by Yury Semikhatsky.
1720
1721         Web Inspector: Stop granting universal access to inspector front-end page.
1722         https://bugs.webkit.org/show_bug.cgi?id=42817
1723
1724         * inspector/InspectorFrontendClientLocal.cpp:
1725         (WebCore::InspectorFrontendClientLocal::windowObjectCleared):
1726
1727 2010-07-22  Yury Semikhatsky  <yurys@chromium.org>
1728
1729         Unreviewed. Fix inspector console tests crashes.
1730
1731         * bindings/v8/ScriptValue.cpp:
1732         (WebCore::v8ToInspectorValue):
1733         (WebCore::ScriptValue::toInspectorValue): enter v8 context before calling GetPropertyNames
1734
1735 2010-07-22  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1736
1737         Unreviewed build fix.
1738
1739         [EFL] Build break for SVN r63881
1740         https://bugs.webkit.org/show_bug.cgi?id=42813
1741
1742         * CMakeLists.txt: Add platform/SchemeRegistry.cpp
1743
1744 2010-07-22  Hayato Ito  <hayato@chromium.org>
1745
1746         Reviewed by Darin Adler.
1747
1748         Refactor CSSSelector's destructor and make it inline.
1749
1750         https://bugs.webkit.org/show_bug.cgi?id=42726
1751
1752         Refactor with no behavior change, thus no new tests.
1753
1754         * css/CSSSelector.cpp:
1755         (WebCore::CSSSelectorBag::~CSSSelectorBag): Make sure the bag is empty.
1756         (WebCore::CSSSelectorBag::add): Renamed from append().
1757         (WebCore::CSSSelector::releaseOwnedSelectorsToBag):
1758         (WebCore::CSSSelector::deleteReachableSelectors):
1759         * css/CSSSelector.h:
1760         (WebCore::CSSSelector::~CSSSelector):
1761
1762 2010-07-22  Yury Semikhatsky  <yurys@chromium.org>
1763
1764         Reviewed by Pavel Feldman.
1765
1766         Web Inspector: should be possible to convert console message arguments to InspectorValues
1767         https://bugs.webkit.org/show_bug.cgi?id=42457
1768
1769         Now it is possible to convert simple JavaScript objects to
1770         InspectorValues.
1771
1772         * bindings/js/ScriptValue.cpp:
1773         (WebCore::jsToInspectorValue):
1774         (WebCore::ScriptValue::toInspectorValue):
1775         * bindings/js/ScriptValue.h:
1776         * bindings/v8/ScriptValue.cpp:
1777         (WebCore::v8ToInspectorValue):
1778         (WebCore::ScriptValue::toInspectorValue):
1779         * bindings/v8/ScriptValue.h:
1780         * inspector/ConsoleMessage.cpp: console notifications are pushed to RemoteInspectorFrontend instead of InspectorFrontend.
1781         (WebCore::ConsoleMessage::CallFrame::buildInspectorObject):
1782         (WebCore::ConsoleMessage::addToFrontend):
1783         (WebCore::ConsoleMessage::updateRepeatCountInConsole):
1784         * inspector/ConsoleMessage.h:
1785         * inspector/InjectedScript.cpp:
1786         (WebCore::InjectedScript::wrapForConsole): return InspectorValue instead of SerializedScriptValue
1787         * inspector/InjectedScript.h:
1788         * inspector/Inspector.idl:
1789         * inspector/InspectorBackend.cpp:
1790         (WebCore::InspectorBackend::clearConsoleMessages): send response directly from the backend
1791         (WebCore::InspectorBackend::remoteFrontend):
1792         * inspector/InspectorBackend.h:
1793         * inspector/InspectorController.cpp:
1794         (WebCore::InspectorController::addConsoleMessage):
1795         (WebCore::InspectorController::clearConsoleMessages): don't send notification to the front end,
1796         it will be send either from InspectorBackend.cpp if it was a user request or reset message will
1797         be send to the front end if the messages are cleared due to page navigation.
1798         (WebCore::InspectorController::populateScriptObjects):
1799         * inspector/front-end/ConsoleView.js:
1800         * inspector/front-end/inspector.js:
1801
1802 2010-07-22  Alexander Pavlov  <apavlov@chromium.org>
1803
1804         Reviewed by Pavel Feldman.
1805
1806         Web Inspector: Modifying settings actually modifies defaultValues
1807         https://bugs.webkit.org/show_bug.cgi?id=42816
1808
1809         Avoid using live modifiable objects as property default values.
1810
1811         * inspector/front-end/Settings.js:
1812         (WebInspector.Settings):
1813         (WebInspector.Settings.prototype.reset):
1814         (WebInspector.Settings.prototype._load):
1815         (WebInspector.Settings.prototype.installSetting):
1816         (WebInspector.Settings.prototype._get):
1817
1818 2010-07-22  Ben Murdoch  <benm@google.com>
1819
1820         Reviewed by Simon Fraser.
1821
1822         Touch events do not affect the :active CSS state
1823         https://bugs.webkit.org/show_bug.cgi?id=39493
1824
1825         Clean the code up by using a typedef based on comments
1826         post landing of the original patch.
1827
1828         No change in functionality so no new tests.
1829
1830         * rendering/HitTestRequest.h: Typedef the HitTestRequestType to unsigned.
1831         * WebCore.exp.in: Update export signature.
1832         * WebCore.order: Ditto.
1833         * page/EventHandler.cpp: use the new typedef.
1834         * page/EventHandler.h: ditto.
1835
1836
1837 2010-07-22  Alexander Pavlov  <apavlov@chromium.org>
1838
1839         Reviewed by Pavel Feldman.
1840
1841         Web Inspector: Rename method for CSS rule source range retrieval and fix return object format
1842         https://bugs.webkit.org/show_bug.cgi?id=42379
1843
1844         * inspector/Inspector.idl:
1845         * inspector/InspectorBackend.cpp:
1846         (WebCore::InspectorBackend::getRuleRanges):
1847         * inspector/InspectorBackend.h:
1848         * inspector/InspectorBackend.idl:
1849         * inspector/InspectorCSSStore.cpp:
1850         (WebCore::InspectorCSSStore::getRuleRanges):
1851         * inspector/InspectorCSSStore.h:
1852         * inspector/InspectorDOMAgent.cpp:
1853         (WebCore::InspectorDOMAgent::getRuleRanges):
1854         * inspector/InspectorDOMAgent.h:
1855         * inspector/front-end/DOMAgent.js:
1856
1857 2010-07-21  Pavel Feldman  <pfeldman@chromium.org>
1858
1859         Reviewed by Yury Semikhatsky.
1860
1861         Chromium Dev Tools: Separators are missing in the context menu.
1862
1863         https://bugs.webkit.org/show_bug.cgi?id=42761
1864
1865         * bindings/js/JSInspectorFrontendHostCustom.cpp:
1866         (WebCore::JSInspectorFrontendHost::showContextMenu):
1867         * bindings/v8/custom/V8InspectorFrontendHostCustom.cpp:
1868         (WebCore::V8InspectorFrontendHost::showContextMenuCallback):
1869         * platform/ContextMenuItem.h:
1870         (WebCore::):
1871
1872 2010-07-22  Zoltan Herczeg  <zherczeg@webkit.org>
1873
1874         Reviewed by Nikolas Zimmermann.
1875
1876         SVGFilterElement & SVGFE*Element don't support dynamic invalidation, when attributes change
1877         https://bugs.webkit.org/show_bug.cgi?id=42244
1878
1879         Implementing svgAttributeChanged and childrenChanged for the base class of all filters,
1880         called SVGFilterPrimitiveStandardAttributes.
1881
1882         Tests: svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-height-attr.html
1883                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-result-attr.html
1884                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-width-attr.html
1885                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-x-attr.html
1886                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-y-attr.html
1887                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-height-prop.html
1888                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-result-prop.html
1889                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-width-prop.html
1890                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-x-prop.html
1891                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-y-prop.html
1892
1893         * svg/SVGFilterPrimitiveStandardAttributes.cpp:
1894         (WebCore::SVGFilterPrimitiveStandardAttributes::svgAttributeChanged):
1895         (WebCore::SVGFilterPrimitiveStandardAttributes::childrenChanged):
1896         * svg/SVGFilterPrimitiveStandardAttributes.h:
1897         (WebCore::SVGFilterPrimitiveStandardAttributes::invalidateFilter):
1898
1899 2010-07-22  Kent Tamura  <tkent@chromium.org>
1900
1901         Reviewed by Darin Adler.
1902
1903         Small refactoring for input value sanitization
1904         https://bugs.webkit.org/show_bug.cgi?id=42807
1905
1906         Rename some functions to clarify their roles.
1907
1908         * dom/InputElement.cpp:
1909         (WebCore::replaceEOLAndLimitLength):
1910           Renamed from sanitizeUserInputValue().
1911         (WebCore::InputElement::sanitizeValueForTextField):
1912           Renamed from sanitizeValue(), and call replaceEOLAndLimitLength()
1913           instead of sanitizeUserInputValue().
1914         (WebCore::InputElement::sanitizeUserInputValue):
1915           Just call replaceEOLAndLimitLength().
1916         (WebCore::InputElement::updateValueIfNeeded):
1917           Use non-static sanitizeValue() for consistency.
1918         * dom/InputElement.h:
1919         * html/HTMLInputElement.cpp:
1920         (WebCore::HTMLInputElement::setInputType):
1921           Revert the change of r63876. It is not needed because of the
1922           updateValueIfNeeded() change.
1923         (WebCore::HTMLInputElement::sanitizeValue):
1924           Apply the sanitizeValue() renaming.
1925
1926 2010-07-22  Tony Gentilcore  <tonyg@chromium.org>
1927
1928         Reviewed by Darin Fisher.
1929
1930         webkitPerformance.timing.responseEnd should not include document parse time
1931         https://bugs.webkit.org/show_bug.cgi?id=42797
1932
1933         No new tests because timing based test would be flaky.
1934
1935         * loader/FrameLoader.cpp:
1936         (WebCore::FrameLoader::finishedLoading): finishedLoading() is called by the platform at the right time, but didReceiveData() synchronously invokes parsing without returning to the event loop prior to this. So by the time the didFinishLoading() method is executed, parsing is finished. The solution is to move this time to didReceiveData() prior to parsing.
1937         * loader/MainResourceLoader.cpp:
1938         (WebCore::MainResourceLoader::didReceiveData): Update the time each time didReceiveData() is called.
1939         (WebCore::MainResourceLoader::didFinishLoading): When finished, and after parsing, set responseEnd appropriately.
1940         * loader/MainResourceLoader.h:
1941         * page/Timing.cpp:
1942         (WebCore::Timing::resourceLoadTimeRelativeToAbsolute): Now that responseEnd is set correctly, it can be much shorter for very fast loads (like local files). This exposed the fact that this skew check was not tight enough. We need to make sure that the whole range of values from the ResourceLoadTiming API fit within fetchStart-responseEnd, not just that requestTime fits in this range.
1943
1944 2010-07-21  Kent Tamura  <tkent@chromium.org>
1945
1946         Reviewed by Darin Adler.
1947
1948         Assertion failure by changing the type of an input element with a
1949         non-number value to 'range'.
1950         https://bugs.webkit.org/show_bug.cgi?id=42643
1951
1952         Test: fast/forms/input-value-sanitization.html
1953
1954         * html/HTMLInputElement.cpp:
1955         (WebCore::HTMLInputElement::setInputType):
1956         Update the value by HTMLInputElement::sanitizeValue() in a case that
1957         storesValueSeparateFromAttribute() state is not changed.
1958
1959 2010-07-21  Adam Barth  <abarth@webkit.org>
1960
1961         Reviewed by Eric Seidel.
1962
1963         Implement spec changes for basefont and bgsound
1964         https://bugs.webkit.org/show_bug.cgi?id=42792
1965
1966         * html/HTMLTreeBuilder.cpp:
1967         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
1968         (WebCore::HTMLTreeBuilder::processStartTag):
1969         (WebCore::HTMLTreeBuilder::processStartTagForInHead):
1970
1971 2010-07-21  Adam Barth  <abarth@webkit.org>
1972
1973         Reviewed by Eric Seidel.
1974
1975         Update a list of tags to match changes in the HTML5 spec
1976         https://bugs.webkit.org/show_bug.cgi?id=42791
1977
1978         We need to atomize these strings at some point.
1979
1980         * html/HTMLTreeBuilder.cpp:
1981         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
1982
1983 2010-07-21  Adam Barth  <abarth@webkit.org>
1984
1985         Reviewed by Eric Seidel.
1986
1987         HTMLTreeBuilder needs to update to match new spec behavior for <button>
1988         https://bugs.webkit.org/show_bug.cgi?id=42233
1989
1990         The spec has some typos in this area, but this is my best guess for
1991         what Ian means.  I've filed bugs against the spec for each typo.
1992
1993         * html/HTMLElementStack.cpp:
1994         (WebCore::HTMLNames::isScopeMarker):
1995         (WebCore::HTMLNames::isButtonScopeMarker):
1996         (WebCore::HTMLElementStack::inButtonScope):
1997         * html/HTMLElementStack.h:
1998         * html/HTMLTreeBuilder.cpp:
1999         (WebCore::HTMLTreeBuilder::processFakePEndTagIfPInButtonScope):
2000         (WebCore::HTMLTreeBuilder::processCloseWhenNestedTag):
2001         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
2002         (WebCore::HTMLTreeBuilder::processEndTagForInBody):
2003         * html/HTMLTreeBuilder.h:
2004
2005 2010-07-21  Justin Schuh  <jschuh@chromium.org>
2006
2007         Reviewed by Oliver Hunt.
2008
2009         Prevent DeleteButtonController enable state from changing when not editing
2010         https://bugs.webkit.org/show_bug.cgi?id=42659
2011
2012         Test: svg/custom/use-invalid-html.xhtml
2013
2014         * dom/ContainerNode.cpp:
2015         (WebCore::ContainerNode::cloneChildNodes):
2016
2017 2010-07-21  Beth Dakin  <bdakin@apple.com>
2018
2019         Reviewed by Dan Bernstein.
2020
2021         Fix for https://bugs.webkit.org/show_bug.cgi?id=42605 New border-
2022         radius path-based drawing code has some issues with corner-joins
2023
2024         GraphicsContext::clipConvexPolygon() now takes a boolean parameter 
2025         indicating whether or not the clip should be antialiased.
2026         * platform/graphics/GraphicsContext.h:
2027         * platform/graphics/cairo/GraphicsContextCairo.cpp:
2028         (WebCore::GraphicsContext::clipConvexPolygon):
2029         * platform/graphics/cg/GraphicsContextCG.cpp:
2030         (WebCore::GraphicsContext::clipConvexPolygon):
2031         * platform/graphics/haiku/GraphicsContextHaiku.cpp:
2032         (WebCore::GraphicsContext::clipConvexPolygon):
2033         * platform/graphics/qt/GraphicsContextQt.cpp:
2034         (WebCore::GraphicsContext::clipConvexPolygon):
2035         * platform/graphics/skia/GraphicsContextSkia.cpp:
2036         (WebCore::GraphicsContext::clipConvexPolygon):
2037         * platform/graphics/wince/GraphicsContextWince.cpp:
2038         (WebCore::GraphicsContext::clipConvexPolygon):
2039         * platform/graphics/wx/GraphicsContextWx.cpp:
2040         (WebCore::GraphicsContext::clipConvexPolygon):
2041
2042         Compute 4 booleans determining if each edge is a "matching" edge of 
2043         not. Send the two appropriate values to clipBorderSidePolygon() so 
2044         we can decide there whether or not to antialias the clip based on 
2045         matchy-ness.
2046         * rendering/RenderBoxModelObject.cpp:
2047         (WebCore::RenderBoxModelObject::paintBorder):
2048
2049         If both edges are matching, don't antialias. If neither edge 
2050         matches, never antialias. If one edge matches and one does not, 
2051         apply two clips to get the appropriate antialiasing in each corner.
2052         (WebCore::RenderBoxModelObject::clipBorderSidePolygon):
2053         * rendering/RenderBoxModelObject.h:
2054
2055 2010-07-21  Brady Eidson  <beidson@apple.com>
2056
2057         Reviewed by Geoffrey Garen.
2058
2059         Break out "scheme registration" functionality from SecurityOrigin to a SchemeRegistry
2060         https://bugs.webkit.org/show_bug.cgi?id=42783
2061
2062         No new tests. (No change in behavior)
2063
2064         * GNUmakefile.am:
2065         * WebCore.exp.in:
2066         * WebCore.gypi:
2067         * WebCore.pro:
2068         * WebCore.vcproj/WebCore.vcproj:
2069         * WebCore.xcodeproj/project.pbxproj:
2070         
2071         * loader/FrameLoader.cpp:
2072         (WebCore::FrameLoader::isMixedContent):
2073         (WebCore::FrameLoader::loadFrameRequest):
2074         
2075         * page/SecurityOrigin.cpp:
2076         (WebCore::SecurityOrigin::SecurityOrigin):
2077         (WebCore::SecurityOrigin::canLoad):
2078         (WebCore::SecurityOrigin::isLocal):
2079         * page/SecurityOrigin.h:
2080         
2081         * platform/SchemeRegistry.cpp: Added.
2082         (WebCore::localSchemes):
2083         (WebCore::secureSchemes):
2084         (WebCore::schemesWithUniqueOrigins):
2085         (WebCore::SchemeRegistry::registerURLSchemeAsLocal):
2086         (WebCore::SchemeRegistry::removeURLSchemeRegisteredAsLocal):
2087         (WebCore::SchemeRegistry::localURLSchemes):
2088         (WebCore::SchemeRegistry::shouldTreatURLAsLocal):
2089         (WebCore::SchemeRegistry::shouldTreatURLSchemeAsLocal):
2090         (WebCore::SchemeRegistry::registerURLSchemeAsNoAccess):
2091         (WebCore::SchemeRegistry::shouldTreatURLSchemeAsNoAccess):
2092         (WebCore::SchemeRegistry::registerURLSchemeAsSecure):
2093         (WebCore::SchemeRegistry::shouldTreatURLSchemeAsSecure):
2094         * platform/SchemeRegistry.h: Added.
2095
2096 2010-07-21  Chris Marrin  <cmarrin@apple.com>
2097
2098         Reviewed by Simon Fraser.
2099
2100         Assertion failure in AnimationBase::updateStateMachine() coming out of paused state
2101         https://bugs.webkit.org/show_bug.cgi?id=37993
2102         
2103         Added logic to properly handle pausing and resuming when in the 
2104         AnimationStateStartWaitStyleAvailable state. This was causing an
2105         assert when going out of the pause state because the paused flag
2106         was not set.
2107         
2108         The fix is a straightforward implementation, going into a new
2109         AnimationStatePausedWaitStyleAvailable state and setting the paused
2110         flag (actually setting the m_pauseTime variable to something other
2111         than -1). Also added handling of the new state, both when the 
2112         "style available" callback comes in while in this state and when
2113         unpausing while in this state. 
2114         
2115         For now a LayoutTest is not possible since there's no way to go in
2116         and out of the pause state. I've opened https://bugs.webkit.org/show_bug.cgi?id=42790
2117         to track this.
2118
2119         * page/animation/AnimationBase.cpp:
2120         (WebCore::AnimationBase::updateStateMachine):
2121         * page/animation/AnimationBase.h:
2122         (WebCore::AnimationBase::):
2123
2124 2010-07-21  Bo Liu  <boliu@chromium.org>
2125
2126         Reviewed by Darin Fisher.
2127
2128         [Chromium] Add chromium WebMediaPlayer to PlatformMedia
2129         https://bugs.webkit.org/show_bug.cgi?id=41295
2130
2131         * platform/graphics/MediaPlayer.h:
2132         (WebCore::PlatformMedia::):
2133
2134 2010-07-21  Simon Fraser  <simon.fraser@apple.com>
2135
2136         Reviewed by Anders Carlsson.
2137
2138         Composited layers don't scroll in WebKit2
2139         https://bugs.webkit.org/show_bug.cgi?id=42771
2140
2141         Prep work: FrameView::scrollPositionChanged() sounds like a generic "did scroll" bottleneck,
2142         but this is deceiving. It's only every called on one platform (Mac) when the NSScrollView gets
2143         scrolled, so rename it to FrameView::scrollPositionChangedViaPlatformWidget().
2144
2145         * WebCore.exp.in:
2146         * page/FrameView.cpp:
2147         (WebCore::FrameView::scrollPositionChangedViaPlatformWidget):
2148         * page/FrameView.h:
2149
2150 2010-07-21  Peter Beverloo  <peter@lvp-media.com>
2151
2152         Reviewed by Eric Carlson.
2153
2154         Removed support for the -khtml CSS vendor prefix and limit the
2155         -apple prefix to two properties (dashboard-region and line-clamp).
2156
2157         Test: fast/css/limited-vendor-prefix-behavior.html
2158
2159         * css/CSSParser.cpp:
2160         (WebCore::cssPropertyID):
2161
2162 2010-07-21  Adam Barth  <abarth@webkit.org>
2163
2164         Reviewed by Eric Seidel.
2165
2166         Fix the last tree HTML5 tree builder crashes
2167         https://bugs.webkit.org/show_bug.cgi?id=42773
2168
2169         This patch changes the internal representation of a bookmark to handle
2170         the case where one of the adjecent entries in the list of active
2171         formatting elements is actually a marker.
2172
2173         After this patch, the bookmarking mechanism isn't as general, but it
2174         works for the cases we need in the adoption agency.
2175
2176         Also, after this patch, there aren't any more known crashers in the
2177         HTML5 tree builder.  :)
2178
2179         * html/HTMLFormattingElementList.cpp:
2180         (WebCore::HTMLFormattingElementList::bookmarkFor):
2181         (WebCore::HTMLFormattingElementList::swapTo):
2182         * html/HTMLFormattingElementList.h:
2183         (WebCore::HTMLFormattingElementList::Bookmark::Bookmark):
2184         (WebCore::HTMLFormattingElementList::Bookmark::moveToAfter):
2185         (WebCore::HTMLFormattingElementList::Bookmark::hasBeenMoved):
2186         (WebCore::HTMLFormattingElementList::Bookmark::mark):
2187         (WebCore::HTMLFormattingElementList::first):
2188         * html/HTMLTreeBuilder.cpp:
2189         (WebCore::HTMLTreeBuilder::callTheAdoptionAgency):
2190
2191 2010-07-21  Tony Gentilcore  <tonyg@chromium.org>
2192
2193         Unreviewed build fix.
2194
2195         Disable overzealous ASSERT
2196         https://bugs.webkit.org/show_bug.cgi?id=42775
2197
2198         No new tests because no changed functionality.
2199
2200         * loader/FrameLoader.cpp:
2201         (WebCore::FrameLoader::stopLoading):
2202
2203 2010-07-21  Yael Aharon  <yael.aharon@nokia.com>
2204
2205         Reviewed by Darin Adler.
2206
2207         Crash in Notification::disconnectFrame() triggered by Frame::lifeSupportTimerFired()
2208         https://bugs.webkit.org/show_bug.cgi?id=42534
2209
2210         Call NotificationsCenter::disconnectFrame() when the frame is disconnected from the page.
2211         Calling it from the destructor of Frame is too late and sometimes causes access violation.
2212         I was not able to reproduce this crash, so did not add new tests.
2213         This patch is based on the error reported in 
2214         http://code.google.com/p/chromium/issues/detail?id=49323.
2215
2216         * page/DOMWindow.cpp:
2217         (WebCore::DOMWindow::pageDestroyed):
2218         * page/DOMWindow.h:
2219         * page/Frame.cpp:
2220         (WebCore::Frame::pageDestroyed):
2221
2222 2010-07-21  Anders Carlsson  <andersca@apple.com>
2223
2224         Reviewed by Sam Weinig.
2225
2226         Don't assert when clicking on a plug-in in WebKit2
2227         https://bugs.webkit.org/show_bug.cgi?id=42762
2228
2229         Remove a now invalid assert and return false instead.
2230
2231         * page/mac/EventHandlerMac.mm:
2232         (WebCore::EventHandler::passMouseDownEventToWidget):
2233
2234 2010-07-21  Anton Muhin  <antonm@chromium.org>
2235
2236         Reviewed by Adam Barth.
2237
2238         [v8] Revert r60670 as it introduced a regression: in some cases named children couldn't be retrieved.
2239         https://bugs.webkit.org/show_bug.cgi?id=42766
2240
2241         See http://code.google.com/p/chromium/issues/detail?id=48804 for more details.
2242
2243         * bindings/scripts/CodeGeneratorV8.pm:
2244         * bindings/v8/ScriptController.cpp:
2245         (WebCore::ScriptController::namedItemAdded):
2246         (WebCore::ScriptController::namedItemRemoved):
2247         * bindings/v8/V8DOMWindowShell.cpp:
2248         (WebCore::V8DOMWindowShell::updateDocumentWrapperCache):
2249         * bindings/v8/V8DOMWindowShell.h:
2250         * bindings/v8/V8DOMWrapper.cpp:
2251         (WebCore::V8DOMWrapper::instantiateV8Object):
2252         * bindings/v8/custom/V8HTMLDocumentCustom.cpp:
2253         (WebCore::V8HTMLDocument::namedPropertyDeleter):
2254         (WebCore::V8HTMLDocument::namedPropertyGetter):
2255         (WebCore::V8HTMLDocument::indexedPropertyGetter):
2256         (WebCore::V8HTMLDocument::allAccessorSetter):
2257         (WebCore::toV8):
2258
2259 2010-07-21  Eric Carlson  <eric.carlson@apple.com>
2260
2261         Reviewed by Simon Fraser.
2262
2263         Update plug-in proxy backed <video> and <audio> elements.
2264         https://bugs.webkit.org/show_bug.cgi?id=42770
2265         <rdar://problem/7963467>
2266
2267         * html/HTMLMediaElement.cpp:
2268         (WebCore::HTMLMediaElement::attributeChanged): Only call m_player->setControls when
2269         the 'controls' attribute changes.
2270         (WebCore::HTMLMediaElement::createRenderer): 
2271         (WebCore::HTMLMediaElement::attach): If we have a proxy widget, call the WebFrameLoaderClient's 
2272         'hide' and 'show' methods for the media proxy plug-in to make sure we remove and re-add the 
2273         plug-in's view to the global list of plugin views.
2274         (WebCore::HTMLMediaElement::createMediaPlayerProxy): Bail if m_proxyWidget is non-NULL
2275
2276         * loader/EmptyClients.h:
2277         (WebCore::EmptyFrameLoaderClient::hideMediaPlayerProxyPlugin): Empty stub method.
2278         (WebCore::EmptyFrameLoaderClient::showMediaPlayerProxyPlugin): Ditto.
2279
2280         * loader/FrameLoader.cpp:
2281         * loader/FrameLoaderClient.h:
2282         * loader/SubframeLoader.cpp:
2283         (WebCore::FrameLoader::loadMediaPlayerProxyPlugin): Always set m_containsPlugIns in loadMediaPlayerProxyPlugin(),
2284          even when we don't have a renderer.
2285         (WebCore::FrameLoader::hideMediaPlayerProxyPlugin): New, call hideMediaPlayerProxyPlugin.
2286         (WebCore::FrameLoader::showMediaPlayerProxyPlugin): New, call showMediaPlayerProxyPlugin.
2287
2288         * platform/graphics/MediaPlayer.cpp:
2289         (WebCore::NullMediaPlayerPrivate::setControls):
2290         (WebCore::MediaPlayer::setControls): New, pass through to media engine.
2291         (WebCore::MediaPlayer::enterFullscreen): Ditto.
2292         (WebCore::MediaPlayer::exitFullscreen): Ditto.
2293         * platform/graphics/MediaPlayer.h:
2294         * platform/graphics/MediaPlayerPrivate.h:
2295         (WebCore::MediaPlayerPrivateInterface::setControls):
2296         (WebCore::MediaPlayerPrivateInterface::enterFullscreen):
2297         (WebCore::MediaPlayerPrivateInterface::exitFullscreen):
2298
2299 2010-07-21  Dimitri Glazkov  <dglazkov@chromium.org>
2300
2301         Reviewed by Darin Adler.
2302
2303         Make more members of CSSStyleSelector private.
2304         https://bugs.webkit.org/show_bug.cgi?id=42757
2305
2306         No change in behavior, so no new tests.
2307
2308         * css/CSSStyleSelector.h: Made a publicly-unused methods private.
2309
2310 2010-07-21  Pavel Feldman  <pfeldman@chromium.org>
2311
2312         Reviewed by Timothy Hatcher.
2313
2314         Web Inspector: consider jQuery object to be of array nature.
2315
2316         https://bugs.webkit.org/show_bug.cgi?id=42758
2317
2318         * inspector/front-end/InjectedScript.js:
2319         (injectedScriptConstructor):
2320
2321 2010-07-21  Alexey Proskuryakov  <ap@apple.com>
2322
2323         Unreviewed Windows build fix.
2324
2325         https://bugs.webkit.org/show_bug.cgi?id=42717
2326         <rdar://problem/7062824> A wrong password entered for site or proxy auth remains in WebCore
2327         credential storage, and is sent with subsequent requests
2328
2329         * platform/network/cf/ResourceHandleCFNet.cpp:
2330         (WebCore::ResourceHandle::didReceiveAuthenticationChallenge): Don't use that direct a
2331         copy/paste, oops!
2332
2333 2010-07-21  Simon Fraser  <simon.fraser@apple.com>
2334
2335         Reviewed by Darin Adler.
2336
2337         When GraphicsLayer::contentsOrientation() is CompositingCoordinatesBottomUp, need to flip repaint rects
2338         https://bugs.webkit.org/show_bug.cgi?id=42662
2339         
2340         If the contentsOrientation() on a GraphicsLayer is CompositingCoordinatesBottomUp, then we
2341         need to flip the rects passed to setNeedsDisplayInRect:.
2342         
2343         Avoid writing this code twice by making a bare function to share code
2344         between WebLayer and WebTiledLayer. Convert the existing +drawContents:ofLayer:intoContext:
2345         class method into a bare function, because the Obj-C calling overhead doesn't buy us anything.
2346         
2347         Take out an assertion in GraphicsLayerCA::updateContentsTransform() that is not
2348         correct.
2349         
2350         No new tests because no layers use bottom-up contents at present.
2351
2352         * platform/graphics/mac/GraphicsLayerCA.mm:
2353         (WebCore::GraphicsLayerCA::updateContentsTransform):
2354         * platform/graphics/mac/WebLayer.h:
2355         * platform/graphics/mac/WebLayer.mm:
2356         (drawLayerContents):
2357         (setLayerNeedsDisplayInRect):
2358         (-[WebLayer setNeedsDisplayInRect:]):
2359         (-[WebLayer drawInContext:]):
2360         * platform/graphics/mac/WebTiledLayer.mm:
2361         (-[WebTiledLayer setNeedsDisplayInRect:]):
2362         (-[WebTiledLayer drawInContext:]):
2363
2364 2010-07-20  Alexey Proskuryakov  <ap@apple.com>
2365
2366         Reviewed by Darin Adler.
2367
2368         https://bugs.webkit.org/show_bug.cgi?id=42717
2369         <rdar://problem/7062824> A wrong password entered for site or proxy auth remains in WebCore
2370         credential storage, and is sent with subsequent requests
2371
2372         Tests: http/tests/security/401-logout/401-logout.php
2373                http/tests/xmlhttprequest/remember-bad-password.html
2374
2375         * platform/network/CredentialStorage.cpp: (WebCore::CredentialStorage::remove):
2376         * platform/network/CredentialStorage.h:
2377         Added a way to remove stored credentials for a given protection space.
2378
2379         * platform/network/cf/ResourceHandleCFNet.cpp: (WebCore::ResourceHandle::didReceiveAuthenticationChallenge):
2380         * platform/network/mac/ResourceHandleMac.mm: (WebCore::ResourceHandle::didReceiveAuthenticationChallenge):
2381         Remove stored credentials if they didn't work the first time.
2382
2383 2010-07-20  Steve Falkenburg  <sfalken@apple.com>
2384
2385         Reviewed by Adam Roben.
2386
2387         WebKit on Windows should build optionally with an unversioned ICU DLL
2388         https://bugs.webkit.org/show_bug.cgi?id=42722
2389         <rdar://problem/8211767> WebKit needs to link against unversioned ICU
2390
2391         To get the proper value for U_DISABLE_RENAMING into all source files, we force
2392         the include of ICUVersion.h (our generated header) via the compiler options.
2393         
2394         * WebCore.vcproj/WebCore.vcproj:
2395         Add forced include of ICUVersion.h.
2396         * WebCore.vcproj/WebCoreCommon.vsprops: Add forced include of ICUVersion.h.
2397
2398 2010-07-21  Ilya Tikhonovsky  <loislo@chromium.org>
2399
2400         Reviewed by Pavel Feldman.
2401
2402         WebInspector: Serialization to JSON in InspectorValue works incorrect
2403         if comma is assigned as decimal separator. Windows with Russian locale.
2404         https://bugs.webkit.org/show_bug.cgi?id=42755
2405
2406         * inspector/InspectorValues.cpp:
2407         (WebCore::InspectorBasicValue::writeJSON):
2408
2409 2010-07-20  Simon Fraser  <simon.fraser@apple.com>
2410
2411         Reviewed by Dan Bernstein.
2412
2413         CSS rotation transform can cause elements with certain styles to vanish during rotation.
2414         https://bugs.webkit.org/show_bug.cgi?id=42579
2415         
2416         Remove some untested, broken code that, as far as I can tell, could never get hit.
2417         Use the correct rect for the PaintInfo so that zero-sized elements with overflow:hidden
2418         but a border paint correctly.
2419
2420         Test: compositing/overflow/zero-size-overflow.html
2421
2422         * rendering/RenderLayerBacking.cpp:
2423         (WebCore::RenderLayerBacking::paintIntoLayer):
2424
2425 2010-07-21  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
2426
2427         Reviewed by Laszlo Gombos.
2428
2429         [QT] Leak and few more fixes to qt port of geolocation
2430         https://bugs.webkit.org/show_bug.cgi?id=42753
2431
2432         Fixes memory leak and default values set to attributes. 
2433         
2434         Memory leak fix no new test case added. 
2435
2436         * platform/qt/GeolocationServiceQt.cpp: Added.
2437         (WebCore::GeolocationServiceQt::~GeolocationServiceQt):
2438         (WebCore::GeolocationServiceQt::positionUpdated):
2439
2440 2010-07-21  Kristian Amlie <kristian.amlie@nokia.com>
2441
2442         Reviewed by Kenneth Rohde Christiansen.
2443
2444         Added automatic sqlite extraction for Symbian to QtWebKit.
2445
2446         Also added sqlite detection in case sqlite is not present in the SDK.
2447         This is possible if WebKit is compiled standalone.
2448
2449         The inclusion part is a consequence of commit c578c6c1d6d in the Qt
2450         repository. It will not work on Qt versions < 4.7.1, but that is ok,
2451         since the only build system it will affect is marked as experimental
2452         in the whole 4.7 series.
2453
2454         * WebCore.pro:
2455
2456 2010-07-21  Adam Barth  <abarth@webkit.org>
2457
2458         Reviewed by Eric Seidel.
2459
2460         Associate elements with the active form
2461         https://bugs.webkit.org/show_bug.cgi?id=42728
2462
2463         This patch fixes fast/forms/formmove3.html.  The test still doesn't
2464         pass due to some render tree differences, but it works as intended now.
2465
2466         To fix this test, I needed to deviate from the spec slight.  Minefield
2467         seems to have the same deviation:
2468           - http://www.w3.org/Bugs/Public/show_bug.cgi?id=10216
2469
2470         * html/HTMLConstructionSite.cpp:
2471         (WebCore::HTMLConstructionSite::takeForm):
2472         (WebCore::HTMLConstructionSite::setForm):
2473         (WebCore::HTMLConstructionSite::createHTMLElement):
2474         * html/HTMLConstructionSite.h:
2475         (WebCore::HTMLConstructionSite::form):
2476         * html/HTMLTreeBuilder.cpp:
2477         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
2478
2479 2010-07-21  Adam Barth  <abarth@webkit.org>
2480
2481         Reviewed by Eric Seidel.
2482
2483         The adoption agency doesn't properly attach()
2484         https://bugs.webkit.org/show_bug.cgi?id=42727
2485
2486         The adoption agency is transliterated rather directly from the spec,
2487         but it misses some of the WebKit-specific machinations, such as
2488         attaching to the render tree.
2489
2490         The algorithm, as written, is a minor layer violation.  I've added to
2491         the problem by calling attach() from HTMLTreeBuilder (even though
2492         that's the job of the HTMLConstructionSite).  We'll need to clean up
2493         the layering a bit at some point.
2494
2495         This patch half fixes fast/forms/formmove3.html.  Hopefully I'll finish
2496         fixing it in the next patch.
2497
2498         * html/HTMLConstructionSite.cpp:
2499         (WebCore::HTMLConstructionSite::createHTMLElementFromSavedElement):
2500         * html/HTMLTreeBuilder.cpp:
2501         (WebCore::HTMLTreeBuilder::callTheAdoptionAgency):
2502
2503 2010-07-21  Adam Barth  <abarth@webkit.org>
2504
2505         Reviewed by Eric Seidel.
2506
2507         <input> elements with no type attribute should be foster parented
2508         https://bugs.webkit.org/show_bug.cgi?id=42725
2509
2510         Fix spec transcription error.
2511
2512         * html/HTMLTreeBuilder.cpp:
2513         (WebCore::HTMLTreeBuilder::processStartTagForInTable):
2514
2515 2010-07-21  Adam Barth  <abarth@webkit.org>
2516
2517         Reviewed by Eric Seidel.
2518
2519         Fix fast/css/last-child-style-sharing.html
2520         https://bugs.webkit.org/show_bug.cgi?id=42731
2521
2522         Prior to this patch, we weren't calling finishParsingChildren on the
2523         body element.  We need a more systematic way of catching these bugs.
2524
2525         * html/HTMLElementStack.cpp:
2526         (WebCore::HTMLElementStack::popAll):
2527         * html/HTMLElementStack.h:
2528         * html/HTMLTreeBuilder.cpp:
2529         (WebCore::HTMLTreeBuilder::processEndOfFile):
2530
2531 2010-07-21  Hans Wennborg  <hans@chromium.org>
2532
2533         Reviewed by Steve Block.
2534
2535         Runtime feature switch for device orientation
2536         https://bugs.webkit.org/show_bug.cgi?id=42265
2537
2538         Add a runtime feature switch that decides whether device orientation
2539         events are available or not. Defaults to true.
2540
2541         * WebCore.vcproj/WebCoreCommon.vsprops: Add bindings/generic to include path.
2542         * bindings/generic/RuntimeEnabledFeatures.cpp:
2543         * bindings/generic/RuntimeEnabledFeatures.h:
2544         (WebCore::RuntimeEnabledFeatures::setDeviceOrientationEnabled):
2545         (WebCore::RuntimeEnabledFeatures::deviceOrientationEnabled):
2546         (WebCore::RuntimeEnabledFeatures::deviceOrientationEventEnabled):
2547         (WebCore::RuntimeEnabledFeatures::ondeviceorientationEnabled):
2548         * page/DOMWindow.cpp:
2549         (WebCore::DOMWindow::addEventListener):
2550         (WebCore::DOMWindow::removeEventListener):
2551         (WebCore::DOMWindow::removeAllEventListeners):
2552         * page/DOMWindow.idl:
2553         * page/Page.cpp:
2554         (WebCore::Page::Page):
2555
2556 2010-07-21  Zoltan Herczeg  <zherczeg@webkit.org>
2557
2558         Reviewed Nikolas Zimmermann.
2559
2560         SVGFilterElement & SVGFE*Element don't support dynamic invalidation, when attributes change
2561         https://bugs.webkit.org/show_bug.cgi?id=42244
2562
2563         The implementation was done by Nikolas Zimmermann before, but
2564         there was no tests for it. The patch also implements the simple
2565         setFilterRes method.
2566
2567         Tests: svg/dynamic-updates/SVGFilterElement-dom-filterRes-attr.html
2568                svg/dynamic-updates/SVGFilterElement-dom-filterUnits-attr.html
2569                svg/dynamic-updates/SVGFilterElement-dom-height-attr.html
2570                svg/dynamic-updates/SVGFilterElement-dom-primitiveUnits-attr.html
2571                svg/dynamic-updates/SVGFilterElement-dom-width-attr.html
2572                svg/dynamic-updates/SVGFilterElement-dom-x-attr.html
2573                svg/dynamic-updates/SVGFilterElement-dom-y-attr.html
2574                svg/dynamic-updates/SVGFilterElement-svgdom-filterRes-call.html
2575                svg/dynamic-updates/SVGFilterElement-svgdom-filterResX-prop.html
2576                svg/dynamic-updates/SVGFilterElement-svgdom-filterResY-prop.html
2577                svg/dynamic-updates/SVGFilterElement-svgdom-filterUnits-prop.html
2578                svg/dynamic-updates/SVGFilterElement-svgdom-height-prop.html
2579                svg/dynamic-updates/SVGFilterElement-svgdom-primitiveUnits-prop.html
2580                svg/dynamic-updates/SVGFilterElement-svgdom-width-prop.html
2581                svg/dynamic-updates/SVGFilterElement-svgdom-x-prop.html
2582                svg/dynamic-updates/SVGFilterElement-svgdom-y-prop.html
2583
2584         * svg/SVGFilterElement.cpp:
2585         (WebCore::SVGFilterElement::setFilterRes):
2586         (WebCore::SVGFilterElement::svgAttributeChanged):
2587         (WebCore::SVGFilterElement::childrenChanged):
2588         * svg/SVGFilterElement.h:
2589
2590 2010-07-21  Ben Murdoch  <benm@google.com>
2591
2592         Unreviewed, build fix.
2593
2594         Forgot to update the role of HitTestRequest.h in
2595         the xcode project after landing patch from
2596         https://bugs.webkit.org/show_bug.cgi?id=39493
2597
2598         * WebCore.xcodeproj/project.pbxproj: Make HitTestRequest.h
2599             private so it can be included in WebKit by EventHandler.h
2600             which is already private..
2601
2602 2010-07-20  Ben Murdoch  <benm@google.com>
2603
2604         Reviewed by Steve Block.
2605
2606         Touch events do not affect the :active CSS state
2607         https://bugs.webkit.org/show_bug.cgi?id=39493
2608
2609         Test: fast/events/touch/touch-active-state.html
2610
2611         * WebCore.exp.in: Update exports for new signature of
2612             hitTestResultAtPoint.
2613         * WebCore.order: ditto.
2614         * page/EventHandler.cpp:
2615         (WebCore::EventHandler::EventHandler):
2616         (WebCore::EventHandler::hitTestResultAtPoint): Pass the type
2617             of the hit test to perform as a parameter with a default
2618             value rather than harcoding it in the function body.
2619         (WebCore::EventHandler::handleMouseMoveEvent): Do not modiify
2620             the active element during a mouse move if the user is
2621             touching the screen.
2622         (WebCore::EventHandler::handleTouchEvent): Set the correct
2623             type of hit test to perform depending on the type of the
2624             touch event we are handling.
2625         * page/EventHandler.h: Update the signature of hitTestResultAtPoint.
2626
2627 2010-07-20  Yury Semikhatsky  <yurys@chromium.org>
2628
2629         Reviewed by Pavel Feldman.
2630
2631         console.assert should include stack trace with line numbers.
2632         https://bugs.webkit.org/show_bug.cgi?id=22556
2633
2634         Test: inspector/console-assert.html
2635
2636         * bindings/v8/ScriptCallStack.cpp:
2637         (WebCore::getTopFrameLocation):
2638         (WebCore::toScriptCallFramesVector):
2639         (WebCore::ScriptCallStack::create):
2640         (WebCore::ScriptCallStack::ScriptCallStack):
2641         * bindings/v8/ScriptCallStack.h:
2642         * bindings/v8/ScriptController.cpp:
2643         (WebCore::ScriptController::setCaptureCallStackForUncaughtExceptions):
2644         * bindings/v8/custom/V8ConsoleCustom.cpp:
2645         (WebCore::V8Console::traceCallback):
2646         (WebCore::V8Console::assertCallback):
2647         * inspector/ConsoleMessage.cpp:
2648         (WebCore::ConsoleMessage::ConsoleMessage):
2649         * inspector/InspectorController.cpp:
2650         (WebCore::InspectorController::addMessageToConsole):
2651         * inspector/front-end/ConsoleView.js:
2652         (WebInspector.ConsoleMessage.prototype._formatMessage):
2653         (WebInspector.ConsoleMessage.prototype.toMessageElement):
2654         * page/Console.cpp:
2655         (WebCore::Console::lastWMLErrorMessage):
2656         * page/Console.idl:
2657
2658 2010-07-20  Rafael Antognolli  <antognolli@profusion.mobi>
2659
2660         Reviewed by Kent Tamura.
2661
2662         [EFL] Use log functions instead of fprintf
2663         https://bugs.webkit.org/show_bug.cgi?id=42576
2664
2665         Use LOG and LOG_ERROR instead of fprintf.
2666
2667         No new tests, no new functionality.
2668
2669         * platform/efl/WidgetEfl.cpp:
2670         (WebCore::Widget::applyFallbackCursor):
2671
2672 2010-07-20  Ilya Tikhonovsky  <loislo@chromium.org>
2673
2674         Reviewed by Yury Semikhatsky.
2675
2676         WebInspector: the bindings generation helper script can be
2677         simplified a bit. Chromium.
2678         https://bugs.webkit.org/show_bug.cgi?id=42523
2679
2680         * WebCore.gyp/WebCore.gyp:
2681         * WebCore.gyp/scripts/rule_binding.py:
2682         * bindings/scripts/CodeGenerator.pm:
2683         * bindings/scripts/CodeGeneratorCPP.pm:
2684         * bindings/scripts/CodeGeneratorJS.pm:
2685         * bindings/scripts/CodeGeneratorObjC.pm:
2686         * bindings/scripts/CodeGeneratorV8.pm:
2687         * bindings/scripts/generate-bindings.pl:
2688         * inspector/CodeGeneratorInspector.pm:
2689
2690 2010-07-20  Steve Falkenburg  <sfalken@apple.com>
2691
2692         Re-save vsprops file after no-op edit in Visual Studio
2693         to fix manual edit issues.
2694
2695         * WebCore.vcproj/WebCoreCommon.vsprops:
2696
2697 2010-07-20  Steve Falkenburg  <sfalken@apple.com>
2698
2699         Re-save vcproj file after no-op edit in Visual Studio
2700         to fix manual edit issues.
2701
2702         * WebCore.vcproj/WebCore.vcproj:
2703
2704 2010-07-20  Adam Barth  <abarth@webkit.org>
2705
2706         Reviewed by Eric Seidel.
2707
2708         Include attributes when reconstructing elements in HTML5 tree builder
2709         https://bugs.webkit.org/show_bug.cgi?id=42594
2710
2711         Remove FIXME now that we have a test.
2712
2713         * html/HTMLConstructionSite.cpp:
2714         (WebCore::HTMLConstructionSite::createHTMLElementFromElementRecord):
2715
2716 2010-07-20  Sam Weinig  <sam@webkit.org>
2717
2718         Reviewed by Brady Eidson.
2719
2720         Patch for https://bugs.webkit.org/show_bug.cgi?id=42719
2721         Make Acid2 pass in WebKit2
2722
2723         * WebCore.exp.in: Update exported functions.
2724
2725 2010-07-20  Gavin Peters  <gavinp@chromium.org>
2726
2727         Reviewed by Adam Barth.
2728
2729         prefetch categorization is exactly wrong
2730         https://bugs.webkit.org/show_bug.cgi?id=42651
2731
2732         While I was in the area, I saw that CachedResource::schedule() was dead, so I have removed it.
2733
2734         Test: fast/dom/HTMLLinkElement/onload-completion-test.html
2735
2736         * loader/CachedCSSStyleSheet.h:
2737         * loader/CachedFont.h:
2738         * loader/CachedImage.h:
2739         * loader/CachedResource.cpp:
2740         * loader/CachedResource.h:
2741         (WebCore::CachedResource::isPrefetch):
2742         * loader/CachedScript.h:
2743         * loader/CachedXBLDocument.h:
2744         * loader/CachedXSLStyleSheet.h:
2745
2746 2010-07-20  James Hawkins  <jhawkins@chromium.org>
2747
2748         Reviewed by Darin Fisher.
2749
2750         Expose the form submission trigger on the HTMLFormElement object. This
2751         is used to verify that the user submitted the form instead of JS when
2752         saving form data in Chrome AutoFill.
2753         https://bugs.webkit.org/show_bug.cgi?id=42479
2754
2755         No new tests as this is only used by the Chromium WebKit API.
2756
2757         * html/HTMLFormElement.cpp:
2758         (WebCore::HTMLFormElement::HTMLFormElement):
2759         (WebCore::HTMLFormElement::submit):
2760         (WebCore::HTMLFormElement::reset):
2761         (WebCore::HTMLFormElement::submissionTrigger):
2762         * html/HTMLFormElement.h:
2763
2764 2010-07-20  Adam Barth  <abarth@webkit.org>
2765
2766         Unreviewed.
2767
2768         Fix fast/dom/title-content-write-set.html for HTML5 tree builder
2769         https://bugs.webkit.org/show_bug.cgi?id=42668
2770
2771         Address some late-breaking review comments.
2772
2773         * dom/CharacterData.cpp:
2774         (WebCore::CharacterData::appendDataCommon):
2775         (WebCore::CharacterData::parserAppendData):
2776         (WebCore::CharacterData::appendData):
2777         (WebCore::CharacterData::insertData):
2778         (WebCore::CharacterData::replaceData):
2779         * dom/CharacterData.h:
2780
2781 2010-07-20  Daniel Erat  <derat@chromium.org>
2782
2783         Reviewed by Ojan Vafai.
2784
2785         Subpixel rendering always disabled for Chromium Linux
2786         https://bugs.webkit.org/show_bug.cgi?id=42220
2787
2788         Explicitly initialize global Chromium Linux isSkiaSubpixelGlyphs
2789         flag to false for clarity.
2790
2791         * platform/graphics/chromium/FontPlatformDataLinux.cpp:
2792
2793 2010-07-20  Adam Barth  <abarth@webkit.org>
2794
2795         Reviewed by Darin Adler.
2796
2797         Fix fast/dom/title-content-write-set.html for HTML5 tree builder
2798         https://bugs.webkit.org/show_bug.cgi?id=42668
2799
2800         When I split parserAppendData from appendData, it was tempting to cut
2801         the function at dispatchModifiedEvent, but that's not quite right.  We
2802         still need to notify the parent that it's children have changed.
2803
2804         * dom/CharacterData.cpp:
2805         (WebCore::CharacterData::appendDataCommon):
2806         (WebCore::CharacterData::parserAppendData):
2807         (WebCore::CharacterData::appendData):
2808         * dom/CharacterData.h:
2809
2810 2010-07-20  Chris Fleizach  <cfleizach@apple.com>
2811
2812         Reviewed by Beth Dakin.
2813
2814         CrashTracer: [USER] 300 crashes in Safari at com.apple.WebCore: WebCore::AccessibilityTable::isTableExposableThroughAccessibility + 573
2815         https://bugs.webkit.org/show_bug.cgi?id=42652
2816
2817         When a table cell accesses its parent table, we should not use getOrCreate, because creating an AXTable inspects its render tree state
2818         which may be out of date, leading to a crash.
2819         By using only get(), it implies that the AXTable must be created before AXTableCells. This should
2820         always be the case when AT clients access a table.
2821
2822         Test: accessibility/updating-attribute-in-table-causes-crash.html
2823
2824         * accessibility/AccessibilityTableCell.cpp:
2825         (WebCore::AccessibilityTableCell::parentTable):
2826
2827 2010-07-20  Abhishek Arya  <inferno@chromium.org>
2828
2829         Reviewed by David Hyatt.
2830
2831         Check the node is a text node before doing the static cast
2832         for editing commands.
2833         https://bugs.webkit.org/show_bug.cgi?id=42655
2834
2835         Test: editing/execCommand/editing-nontext-node-crash.xhtml
2836
2837         * editing/DeleteSelectionCommand.cpp:
2838         (WebCore::DeleteSelectionCommand::fixupWhitespace):
2839         * editing/InsertLineBreakCommand.cpp:
2840         (WebCore::InsertLineBreakCommand::doApply):
2841         * editing/InsertParagraphSeparatorCommand.cpp:
2842         (WebCore::InsertParagraphSeparatorCommand::doApply):
2843
2844 2010-07-20  Leo Yang  <leo.yang@torchmobile.com.cn>
2845
2846         Reviewed by David Hyatt.
2847
2848         Don't merge Anonymous block whose first child is inline run-in.
2849         Make run-in recalculate its style after its renderer is destroyed.
2850         https://bugs.webkit.org/show_bug.cgi?id=41375.
2851
2852         Test: fast/runin/crash-when-reparent-sibling.html
2853
2854         * rendering/RenderBlock.cpp:
2855         (WebCore::canMergeContiguousAnonymousBlocks):
2856         * rendering/RenderObjectChildList.cpp:
2857         (WebCore::RenderObjectChildList::destroyLeftoverChildren):
2858
2859 2010-07-20  Steve Block  <steveblock@google.com>
2860
2861         Unreviewed Qt test fix.
2862
2863         Qt should not use PREEMPT_GEOLOCATION_PERMISSION until Bug 42027 is fixed.
2864         https://bugs.webkit.org/show_bug.cgi?id=42068
2865
2866         No new tests, test fix only.
2867
2868         * WebCore.pro:
2869
2870 2010-07-20  Sheriff Bot  <webkit.review.bot@gmail.com>
2871
2872         Unreviewed, rolling out r63764.
2873         http://trac.webkit.org/changeset/63764
2874         https://bugs.webkit.org/show_bug.cgi?id=42658
2875
2876         have broken Chromium linux build (Requested by loislo on
2877         #webkit).
2878
2879         * WebCore.gyp/WebCore.gyp:
2880         * WebCore.gyp/scripts/rule_binding.py:
2881         * bindings/scripts/CodeGenerator.pm:
2882         * bindings/scripts/CodeGeneratorCPP.pm:
2883         * bindings/scripts/CodeGeneratorJS.pm:
2884         * bindings/scripts/CodeGeneratorObjC.pm:
2885         * bindings/scripts/CodeGeneratorV8.pm:
2886         * bindings/scripts/generate-bindings.pl:
2887         * inspector/CodeGeneratorInspector.pm:
2888
2889 2010-07-20  Alexey Proskuryakov  <ap@apple.com>
2890
2891         Reviewed by Brady Eidson.
2892
2893         https://bugs.webkit.org/show_bug.cgi?id=41531
2894         Asynchronous cross origin XMLHttpRequest doesn't expose 401 response when withCredentials is false
2895
2896         This doesn't match Firefox, but it matches our sync case, XHR2 spec and common sense.
2897
2898         Test: http/tests/xmlhttprequest/cross-origin-no-authorization.html (updated results).
2899
2900         * loader/DocumentThreadableLoader.cpp:
2901         (WebCore::DocumentThreadableLoader::didReceiveAuthenticationChallenge): Instead of canceling
2902         the request, continue without credentials - if the platform has a necessary method on
2903         ResourceHandle.
2904
2905         * loader/SubresourceLoader.cpp:
2906         (WebCore::SubresourceLoader::didReceiveAuthenticationChallenge): Don't ask resource loader
2907         client for credentials if subresource loader client already took care of those.
2908
2909         * platform/network/ResourceHandle.cpp: (WebCore::ResourceHandle::hasAuthenticationChallenge):
2910         * platform/network/ResourceHandle.h:
2911         Added an accessor to check whether ResourceHandle is currently waiting for credentials.
2912
2913 2010-07-20  Leandro Pereira  <leandro@profusion.mobi>
2914
2915         [EFL] Unreviewed build fix.
2916
2917         Build SVGPathBuilder, SVGPathParser and SVGPathSegListBuilder.
2918         Add websockets include directory.
2919
2920         The EFL port doesn't yet support automated tests.
2921
2922         * CMakeLists.txt:
2923
2924 2010-07-20  Ilya Tikhonovsky  <loislo@chromium.org>
2925
2926         Reviewed by Yury Semikhatsky.
2927
2928         WebInspector: the bindings generation helper script can be
2929         simplified a bit. Chromium.
2930         https://bugs.webkit.org/show_bug.cgi?id=42523
2931
2932         * WebCore.gyp/WebCore.gyp:
2933         * WebCore.gyp/scripts/rule_binding.py:
2934         * bindings/scripts/CodeGenerator.pm:
2935         * bindings/scripts/CodeGeneratorCPP.pm:
2936         * bindings/scripts/CodeGeneratorJS.pm:
2937         * bindings/scripts/CodeGeneratorObjC.pm:
2938         * bindings/scripts/CodeGeneratorV8.pm:
2939         * bindings/scripts/generate-bindings.pl:
2940         * inspector/CodeGeneratorInspector.pm:
2941
2942 2010-07-20  Joseph Pecoraro  <joepeck@webkit.org>
2943
2944         Reviewed by Geoffrey Garen.
2945
2946         WebScriptObject Should Allow Safely Checking For Key Existence
2947         https://bugs.webkit.org/show_bug.cgi?id=42613
2948
2949         Test: platform/mac/fast/objc/webScriptObject-hasWebScriptKey.html
2950
2951         Add private API "hasWebScriptKey" to check for key existence in
2952         a WebScriptObject. Like JavaScript's `in` syntax. This is intended
2953         to be made public eventually.
2954
2955         * bindings/objc/WebScriptObject.mm:
2956         (-[WebScriptObject hasWebScriptKey:]):
2957         * bindings/objc/WebScriptObjectPrivate.h:
2958
2959 2010-07-20  Adam Barth  <abarth@webkit.org>
2960
2961         Reviewed by Eric Seidel.
2962
2963         HTML5 tree builder needs to call dispatchDocumentElementAvailable
2964         https://bugs.webkit.org/show_bug.cgi?id=42654
2965
2966         This patch fixes the follout LayoutTests with --html5-treebuilder:
2967           - userscripts/script-not-run-for-fragments.html
2968           - userscripts/script-run-at-start.html
2969
2970         * html/HTMLConstructionSite.cpp:
2971         (WebCore::HTMLConstructionSite::HTMLConstructionSite):
2972         (WebCore::HTMLConstructionSite::dispatchDocumentElementAvailableIfNeeded):
2973         (WebCore::HTMLConstructionSite::insertHTMLHtmlStartTagBeforeHTML):
2974         (WebCore::HTMLConstructionSite::insertHTMLHtmlElement):
2975         * html/HTMLConstructionSite.h:
2976         * html/HTMLTreeBuilder.cpp:
2977         (WebCore::HTMLTreeBuilder::HTMLTreeBuilder):
2978
2979 2010-07-20  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
2980
2981         Reviewed by Steve Block.
2982
2983         Need to be able to configure Geolocation policy regarding user permissions
2984         https://bugs.webkit.org/show_bug.cgi?id=42068
2985
2986         Introducing new USE() flag PREEMPT_GEOLOCATION_PREMISSION using which
2987         acquires user permission first before starting location service
2988
2989         This change does not introduce any change in behavior for any platform. So there are no new tests
2990         added.
2991
2992         * WebCore.pro:
2993         * page/Geolocation.cpp:
2994         (WebCore::Geolocation::startRequest):
2995         (WebCore::Geolocation::setIsAllowed):
2996         (WebCore::Geolocation::startUpdating):
2997         * page/Geolocation.h:
2998
2999 2010-07-20  Adam Barth  <abarth@webkit.org>
3000
3001         Reviewed by Darin Adler.
3002
3003         HTML5 tree builder should restore form state
3004         https://bugs.webkit.org/show_bug.cgi?id=42644
3005
3006         We need to tell self-closing tags that we're done parsing their
3007         children.  This patch fixes the following LayoutTests when run with
3008         --html5-treebuilder:
3009
3010           fast/forms/button-state-restore.html
3011           fast/forms/state-restore-to-non-autocomplete-form.html
3012           fast/forms/state-restore-to-non-edited-controls.html
3013           fast/history/saves-state-after-fragment-nav.html
3014           http/tests/navigation/restore-form-state-https.html
3015
3016         * html/HTMLConstructionSite.cpp:
3017         (WebCore::HTMLConstructionSite::insertSelfClosingHTMLElement):
3018
3019 2010-07-20  Sheriff Bot  <webkit.review.bot@gmail.com>
3020
3021         Unreviewed, rolling out r63750.
3022         http://trac.webkit.org/changeset/63750
3023         https://bugs.webkit.org/show_bug.cgi?id=42648
3024
3025         This revision breaks the windows builds (Requested by
3026         hwennborg on #webkit).
3027
3028         * bindings/generic/RuntimeEnabledFeatures.cpp:
3029         * bindings/generic/RuntimeEnabledFeatures.h:
3030         * page/DOMWindow.cpp:
3031         (WebCore::DOMWindow::addEventListener):
3032         (WebCore::DOMWindow::removeEventListener):
3033         (WebCore::DOMWindow::removeAllEventListeners):
3034         * page/DOMWindow.idl:
3035         * page/Page.cpp:
3036         (WebCore::Page::Page):
3037
3038 2010-07-20  Ilya Tikhonovsky  <loislo@chromium.org>
3039
3040         Reviewed by Pavel Feldman.
3041
3042         WebInspector: It is possible to show full call stack instead of top frame for Caller
3043         and Call Site properties in Timeline panel.
3044         https://bugs.webkit.org/show_bug.cgi?id=42620
3045
3046         * English.lproj/localizedStrings.js:
3047         * inspector/front-end/TimelinePanel.js:
3048         (WebInspector.TimelinePanel.FormattedRecord):
3049         (WebInspector.TimelinePanel.FormattedRecord.prototype._generatePopupContent):
3050         (WebInspector.TimelinePanel.FormattedRecord.prototype._getRecordDetails):
3051         (WebInspector.TimelinePanel.PopupContentHelper.prototype._appendElementRow):
3052         (WebInspector.TimelinePanel.PopupContentHelper.prototype._appendLinkRow):
3053         (WebInspector.TimelinePanel.PopupContentHelper.prototype._appendStackTrace):
3054         * inspector/front-end/inspector.css:
3055         (.timeline-details):
3056         (.timeline-function-name):
3057         (.timeline-stacktrace-title):
3058
3059 2010-07-20  Leon Clarke  <leonclarke@google.com>
3060
3061         Reviewed by Pavel Feldman.
3062
3063         Make things compile again when the inspector is disabled, following
3064         recent inspector improvements.
3065         https://bugs.webkit.org/show_bug.cgi?id=42632
3066
3067         No new tests. Fixing a build break.
3068
3069
3070         * bindings/v8/ScriptCallStack.cpp:
3071         (WebCore::ScriptCallStack::stackTrace):
3072         * inspector/InspectorController.h:
3073         (WebCore::InspectorController::didInsertDOMNode):
3074         (WebCore::InspectorController::didRemoveDOMNode):
3075         (WebCore::InspectorController::didModifyDOMAttr):
3076         * inspector/InspectorDOMAgent.h:
3077         * loader/appcache/ApplicationCacheGroup.cpp:
3078
3079 2010-07-20  Anton Muhin  <antonm@chromium.org>
3080
3081         Reviewed by Adam Barth.
3082
3083         [v8] Allow handles to be disposed and WebKit objects to be dereferenced even if their map is already destroyed.
3084         https://bugs.webkit.org/show_bug.cgi?id=42634
3085
3086         Currently DOMDataStore could be destroyed even if it has some mappings (it gets destroyed
3087         when its isolated context gets GCed).  However in this case, handles allocated for
3088         such objects would never be disposed as we require presence of mapping from wrapped
3089         WebKit object to handle being collected in the map and now map is gone.  That leads to
3090         zombie objects in both WebKit (wrapped WebKit object doesn't get dereferenced) and V8
3091         (both handle and V8 wrapper object could not be destroyed).
3092
3093         See http://code.google.com/p/chromium/issues/detail?id=47125 for further discussion.
3094
3095         * bindings/v8/DOMData.h:
3096         (WebCore::DOMData::handleWeakObject):
3097         * bindings/v8/DOMDataStore.cpp:
3098         (WebCore::DOMDataStore::weakNodeCallback):
3099         * bindings/v8/V8DOMMap.h:
3100         (WebCore::WeakReferenceMap::~WeakReferenceMap):
3101
3102 2010-07-20  Hans Wennborg  <hans@chromium.org>
3103
3104         Reviewed by Steve Block.
3105
3106         Runtime feature switch for device orientation
3107         https://bugs.webkit.org/show_bug.cgi?id=42265
3108
3109         Add a runtime feature switch that decides whether device orientation
3110         events are available or not. Defaults to true.
3111
3112         * bindings/generic/RuntimeEnabledFeatures.cpp:
3113         * bindings/generic/RuntimeEnabledFeatures.h:
3114         (WebCore::RuntimeEnabledFeatures::setDeviceOrientationEnabled):
3115         (WebCore::RuntimeEnabledFeatures::deviceOrientationEnabled):
3116         (WebCore::RuntimeEnabledFeatures::deviceOrientationEventEnabled):
3117         (WebCore::RuntimeEnabledFeatures::ondeviceorientationEnabled):
3118         * page/DOMWindow.cpp:
3119         (WebCore::DOMWindow::addEventListener):
3120         (WebCore::DOMWindow::removeEventListener):
3121         (WebCore::DOMWindow::removeAllEventListeners):
3122         * page/DOMWindow.idl:
3123         * page/Page.cpp:
3124         (WebCore::Page::Page):
3125
3126 2010-07-20  Hayato Ito  <hayato@chromium.org>
3127
3128         Reviewed by Darin Adler.
3129
3130         Fixed a crash when deeply nested CSS selector is used.
3131         https://bugs.webkit.org/show_bug.cgi?id=41129
3132
3133         This patch deletes CSSSelectors iteratively so that it doesn't cause stack overflow.
3134
3135         Test: fast/css/css-selector-deeply-nested.html
3136
3137         * css/CSSSelector.cpp:
3138         (WebCore::CSSSelectorBag::~CSSSelectorBag):
3139         (WebCore::CSSSelectorBag::isEmpty):
3140         (WebCore::CSSSelectorBag::append):
3141         (WebCore::CSSSelectorBag::takeAny):
3142         (WebCore::CSSSelector::~CSSSelector):
3143         (WebCore::CSSSelector::specificity):
3144         * css/CSSSelector.h:
3145
3146 2010-07-20  Yury Semikhatsky  <yurys@chromium.org>
3147
3148         Reviewed by Pavel Feldman.
3149
3150         [v8] Web Inspector: inspected page renderer crashes when inspected page has JS syntax error
3151         https://bugs.webkit.org/show_bug.cgi?id=42642
3152
3153         * bindings/v8/V8ConsoleMessage.cpp:
3154         (WebCore::V8ConsoleMessage::handler):
3155
3156 2010-07-20  Sheriff Bot  <webkit.review.bot@gmail.com>
3157
3158         Unreviewed, rolling out r63742.
3159         http://trac.webkit.org/changeset/63742
3160         https://bugs.webkit.org/show_bug.cgi?id=42641
3161
3162         Broke Leopard Intel build. (Requested by bbandix on #webkit).
3163
3164         * WebCore.pro:
3165         * page/Geolocation.cpp:
3166         (WebCore::Geolocation::startRequest):
3167         (WebCore::Geolocation::setIsAllowed):
3168         (WebCore::Geolocation::startUpdating):
3169         * page/Geolocation.h:
3170
3171 2010-07-20  Adam Roben  <aroben@apple.com>
3172
3173         Windows build fix
3174
3175         * WebCore.vcproj/WebCore.vcproj: Fixed case of "Name" attribute.
3176
3177 2010-07-20  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
3178
3179         Reviewed by Steve Block.
3180
3181         Need to be able to configure Geolocation policy regarding user permissions
3182         https://bugs.webkit.org/show_bug.cgi?id=42068
3183
3184         Introducing new USE() flag PREEMPT_GEOLOCATION_PREMISSION using which
3185         acquires user permission first before starting location service
3186
3187         This change does not introduce any change in behavior for any platform. So there are no new tests
3188         added.
3189
3190         * WebCore.pro:
3191         * page/Geolocation.cpp:
3192         (WebCore::Geolocation::startRequest):
3193         (WebCore::Geolocation::setIsAllowed):
3194         (WebCore::Geolocation::startUpdating):
3195         * page/Geolocation.h:
3196
3197 2010-07-20  Alexander Pavlov  <apavlov@chromium.org>
3198
3199         Reviewed by Pavel Feldman.
3200
3201         Inspector: Resources Search Should Search Only Filtered Resources
3202         https://bugs.webkit.org/show_bug.cgi?id=28290
3203
3204         * inspector/front-end/AbstractTimelinePanel.js:
3205         (WebInspector.AbstractTimelinePanel.prototype._updateFilter):
3206         * inspector/front-end/ResourcesPanel.js:
3207         (WebInspector.ResourcesPanel.prototype.get searchableViews):
3208         * inspector/front-end/inspector.js:
3209         (WebInspector.performSearch):
3210         (WebInspector.doPerformSearch):
3211
3212 2010-07-20  Alexander Pavlov  <apavlov@chromium.org>
3213
3214         Reviewed by Pavel Feldman.
3215
3216         Web Inspector: Incorrect absolute URLs in tooltips of links in the ElementsTreeOutline
3217         https://bugs.webkit.org/show_bug.cgi?id=42626
3218
3219         * inspector/front-end/ElementsTreeOutline.js:
3220         (WebInspector.ElementsTreeElement.prototype._attributeHTML):
3221
3222 2010-07-20  Shinichiro Hamaji  <hamaji@chromium.org>
3223
3224         Unreviewed comment fix. The comment in IDL should be //, not # .
3225
3226         Code generator: ensure generated constants match their corresponding enums.
3227         https://bugs.webkit.org/show_bug.cgi?id=42250
3228
3229         * dom/OverflowEvent.idl:
3230
3231 2010-07-20  Hans Wennborg  <hans@chromium.org>
3232
3233         Reviewed by Steve Block.
3234
3235         Add WebCore/bindings/generic/RuntimeEnabledFeatures.cpp to build files
3236         https://bugs.webkit.org/show_bug.cgi?id=42380
3237
3238         RuntimeEnabledFeatures.cpp and .h were moved from bindings/v8 to
3239         bindings/generic a while a go (in r54593), but need to
3240         be added to the build in order to be used.
3241
3242         No new functionality so no new tests.
3243
3244         * Android.mk:
3245         * CMakeLists.txt:
3246         * GNUmakefile.am:
3247         * WebCore.pro:
3248         * WebCore.vcproj/WebCore.vcproj:
3249         * WebCore.xcodeproj/project.pbxproj:
3250         * websockets/WebSocket.cpp: Remove #if USE(V8), as RuntimeEnabledFeatures.cpp is no longer V8-only.
3251         * websockets/WebSocket.h: Ditto.
3252
3253 2010-07-20  Tony Chang  <tony@chromium.org>
3254
3255         Reviewed by Kent Tamura.
3256
3257         clean up style in ClipboardWin and PasteboardWin
3258         https://bugs.webkit.org/show_bug.cgi?id=42609
3259
3260         No new tests since this is just a style cleanup.
3261
3262         * platform/win/ClipboardUtilitiesWin.cpp:
3263         (WebCore::createGlobalData):
3264         (WebCore::markupToCFHTML):
3265         (WebCore::getURL):
3266         (WebCore::getPlainText):
3267         (WebCore::getTextHTML):
3268         (WebCore::fragmentFromFilenames):
3269         (WebCore::containsFilenames):
3270         (WebCore::fragmentFromCFHTML):
3271         (WebCore::fragmentFromHTML):
3272         * platform/win/ClipboardUtilitiesWin.h:
3273         * platform/win/ClipboardWin.cpp:
3274         (WebCore::pathRemoveBadFSCharacters):
3275         (WebCore::createGlobalHDropContent):
3276         (WebCore::createGlobalUrlFileDescriptor):
3277         (WebCore::writeURL):
3278         (WebCore::ClipboardWin::clearData):
3279         (WebCore::ClipboardWin::clearAllData):
3280         (WebCore::ClipboardWin::getData):
3281         (WebCore::ClipboardWin::types):
3282         (WebCore::ClipboardWin::declareAndWriteDragImage):
3283         (WebCore::ClipboardWin::writeRange):
3284         * platform/win/ClipboardWin.h:
3285         (WebCore::ClipboardWin::create):
3286         (WebCore::ClipboardWin::dataObject):
3287         * platform/win/PasteboardWin.cpp:
3288         (WebCore::PasteboardOwnerWndProc):
3289         (WebCore::Pasteboard::writeSelection):
3290         (WebCore::Pasteboard::writeURL):
3291         (WebCore::Pasteboard::plainText):
3292         (WebCore::Pasteboard::documentFragment):
3293
3294 2010-07-20  Yury Semikhatsky  <yurys@chromium.org>
3295
3296         Reviewed by Pavel Feldman.
3297
3298         Web Inspector: jump to source is broken for call frames logged by console.trace
3299         https://bugs.webkit.org/show_bug.cgi?id=42619
3300
3301         * inspector/front-end/ConsoleView.js:
3302         (WebInspector.ConsoleMessage.prototype._populateStackTraceTreeElement):
3303         * inspector/front-end/treeoutline.js: allow passing HTML element as a title.
3304         (TreeElement.prototype.set title):
3305         (TreeElement.prototype.set shouldRefreshChildren):
3306         (TreeElement.prototype._setListItemNodeContent):
3307         (TreeElement.prototype._attach):
3308
3309 2010-07-20  Tony Chang  <tony@chromium.org>
3310
3311         Reviewed by Dan Bernstein.
3312
3313         [chromium] crash in Position::getInlineBoxAndOffset
3314         https://bugs.webkit.org/show_bug.cgi?id=42202
3315
3316         Test: editing/selection/firstRect-crash.html
3317         Manual test: win/contextmenu-key2.html
3318
3319         * page/EventHandler.cpp: Code no longer needed
3320         * page/Frame.cpp:
3321         (WebCore::Frame::firstRectForRange): Normalize Positions to VisiblePositions because
3322             Positions may be pointing to nodes that have no renderer.  If there's no renderer,
3323             getInlineBoxAndOffset will crash.
3324
3325 2010-07-20  Adam Barth  <abarth@webkit.org>
3326
3327         Reviewed by Darin Adler.
3328
3329         Foster parenting depends on the current element at insertion time
3330         https://bugs.webkit.org/show_bug.cgi?id=42599
3331
3332         We need to consider the current element at insertion-time when deciding
3333         whether to redirect insertion to the foster parent.  Previously, we
3334         were considering the current element *both* at insertion-time and at
3335         the time we created the guard.
3336
3337         * html/HTMLConstructionSite.h:
3338         (WebCore::HTMLConstructionSite::RedirectToFosterParentGuard::RedirectToFosterParentGuard):
3339         * html/HTMLTreeBuilder.cpp:
3340         (WebCore::HTMLTreeBuilder::processStartTagForInTable):
3341         (WebCore::HTMLTreeBuilder::processEndTagForInTable):
3342         (WebCore::HTMLTreeBuilder::defaultForInTableText):
3343
3344 2010-07-20  Matthew Delaney  <mdelaney@apple.com>
3345
3346         Reviewed by Kenneth Rohde Christiansen.
3347
3348         Failing 2d.path.stroke.prune.curve philip canvas test
3349         https://bugs.webkit.org/show_bug.cgi?id=42190
3350
3351         * html/canvas/CanvasRenderingContext2D.cpp:
3352         (WebCore::CanvasRenderingContext2D::closePath): Added check to make sure there's a non-trivial path to close. Since there is currently no way to check if the current point is the start point, or similarly if there is only 1 point in the current subpath (since these are both sufficient conditions for a trivial subpath), then checking that the bounding rectangle has both zero width and height proves also to be a sufficient condition for a trivial path.
3353         (WebCore::CanvasRenderingContext2D::quadraticCurveTo): Added in simple bounds as per the spec.
3354         (WebCore::CanvasRenderingContext2D::bezierCurveTo): Added in simple bounds as per the spec.
3355         * platform/graphics/cg/PathCG.cpp:
3356         (WebCore::Path::closeSubpath): Moved the check for an empty path up on level to make it platform independent and remove redundancy.
3357
3358 2010-07-19  Victoria Kirst  <vrk@google.com>
3359
3360         Reviewed by David Levin.
3361
3362         Added a simple implementation of VideoLayerChromium. Uses the
3363         LayerChromium::updateTextureRect() to send video frames to the
3364         GPU.
3365         https://bugs.webkit.org/show_bug.cgi?id=42234
3366
3367         * WebCore.gypi: Added include for VideoLayerChromium.
3368         * platform/graphics/chromium/GraphicsLayerChromium.cpp:
3369         (WebCore::GraphicsLayerChromium::setContentsToMedia): Implemented
3370         setContentsToMedia, though it does not seem to trigger a repaint
3371         correctly.
3372         * platform/graphics/chromium/GraphicsLayerChromium.h:
3373         * platform/graphics/chromium/VideoLayerChromium.cpp: Added.
3374         (WebCore::VideoLayerChromium::create):
3375         (WebCore::VideoLayerChromium::VideoLayerChromium):
3376         (WebCore::VideoLayerChromium::updateTextureContents):
3377         * platform/graphics/chromium/VideoLayerChromium.h: Added.
3378         (WebCore::VideoLayerChromium::drawsContent):
3379
3380 2010-07-19  Dirk Schulze  <krit@webkit.org>
3381
3382         Reviewed by Nikolas Zimmermann.
3383
3384         SVG CleanUp of SVGPathData parsing
3385         https://bugs.webkit.org/show_bug.cgi?id=41410
3386
3387         Cleanup the parsing code for SVG Paths. Move classes out of the convoluted SVGParserUtilities.cpp
3388         in their own files. Make use of WebCore specific objects in parsing code like FloatPoint, AffineTransform,
3389         etc. instead of using home-brewn solutions.
3390         The SVGPathParser parses a path data string and delivers the parsed segments and values to the SVGPathConsumer.
3391         SVGPathConsumer is the base class for SVGPathBuilder and SVGPathSegListBuilder, that either build the platform
3392         Path object or a SVGPathSegList out of the segments.
3393         We're now directly parsing floats instead of truncating precision to float afterwards.
3394
3395         SVG Path with an arc with radius of 0 does not render
3396         https://bugs.webkit.org/show_bug.cgi?id=40448
3397
3398         If one of the radii on the elliptic arc are zero, we should draw a line from the starting point to
3399         the end point according to the spec. Fixed this bug with this patch, because an is zero check and 
3400         an DRT check was neccessary with the current clean-up.
3401         Extended svg/dom/path-parser.xhml to check the correct behavior.
3402
3403         * Android.mk:
3404         * GNUmakefile.am:
3405         * WebCore.gypi:
3406         * WebCore.pro:
3407         * WebCore.vcproj/WebCore.vcproj:
3408         * WebCore.xcodeproj/project.pbxproj:
3409         * platform/graphics/FloatPoint.h:
3410         (WebCore::FloatPoint::move): Follow WebKit style and use multiple lines for the function.
3411         (WebCore::FloatPoint::scale): Scale FloatPoint.
3412         (WebCore::operator+=):
3413         (WebCore::operator+): Add two FloatPoints and give back the sum as FloatPoint.
3414         * svg/SVGAllInOne.cpp: Added new created files.
3415         * svg/SVGAnimateElement.cpp:
3416         (WebCore::SVGAnimateElement::calculateFromAndToValues): Use new PathParser to create a PathSegList.
3417         * svg/SVGAnimateMotionElement.cpp:
3418         (WebCore::SVGAnimateMotionElement::parseMappedAttribute): Use new PathParser to create a Path.
3419         * svg/SVGGlyphElement.cpp:
3420         (WebCore::parsePathData): Use new PathParser to create a Path.
3421         * svg/SVGParserUtilities.cpp:
3422         (WebCore::parseArcFlag): Removed parseArcFlag(double&), we parse in float now.
3423         * svg/SVGParserUtilities.h: Removed path parsing code and a lot of unneccessary includes.
3424         * svg/SVGPathBuilder.cpp: Added.
3425         (WebCore::SVGPathBuilder::SVGPathBuilder):
3426         (WebCore::SVGPathBuilder::build): Builds a normalized Path.
3427         (WebCore::SVGPathBuilder::moveTo):
3428         (WebCore::SVGPathBuilder::lineTo):
3429         (WebCore::SVGPathBuilder::curveToCubic):
3430         (WebCore::SVGPathBuilder::closePath):
3431         * svg/SVGPathBuilder.h: Added. We just create normalized Paths, so this functions should never be reached.
3432         (WebCore::SVGPathBuilder::lineToHorizontal):
3433         (WebCore::SVGPathBuilder::lineToVertical):
3434         (WebCore::SVGPathBuilder::curveToCubicSmooth):
3435         (WebCore::SVGPathBuilder::curveToQuadratic):
3436         (WebCore::SVGPathBuilder::curveToQuadraticSmooth):
3437         (WebCore::SVGPathBuilder::arcTo):
3438         * svg/SVGPathConsumer.h: Added. Base class of SVGPathBuilder and SVGPathSegListBuilder.
3439         (WebCore::):
3440         (WebCore::SVGPathConsumer::SVGPathConsumer):
3441         (WebCore::SVGPathConsumer::~SVGPathConsumer):
3442         * svg/SVGPathElement.cpp:
3443         (WebCore::SVGPathElement::parseMappedAttribute): Use new PathParser to create a PathSegList.
3444         * svg/SVGPathParser.cpp: Added.
3445         (WebCore::SVGPathParser::SVGPathParser):
3446         (WebCore::SVGPathParser::~SVGPathParser):
3447         (WebCore::SVGPathParser::parseClosePathSegment):
3448         (WebCore::SVGPathParser::parseMoveToSegment):
3449         (WebCore::SVGPathParser::parseLineToSegment):
3450         (WebCore::SVGPathParser::parseLineToHorizontalSegment):
3451         (WebCore::SVGPathParser::parseLineToVerticalSegment):
3452         (WebCore::SVGPathParser::parseCurveToCubicSegment):
3453         (WebCore::SVGPathParser::parseCurveToCubicSmoothSegment):
3454         (WebCore::SVGPathParser::parseCurveToQuadraticSegment):
3455         (WebCore::SVGPathParser::parseCurveToQuadraticSmoothSegment):
3456         (WebCore::SVGPathParser::parseArcToSegment):
3457         (WebCore::SVGPathParser::parsePathDataString):
3458         (WebCore::SVGPathParser::decomposeArcToCubic): Normalizes an arc to multiple cubic curves.
3459         * svg/SVGPathParser.h: Added.
3460         * svg/SVGPathSegListBuilder.cpp: Added.
3461         (WebCore::SVGPathSegListBuilder::SVGPathSegListBuilder):
3462         (WebCore::SVGPathSegListBuilder::build):
3463         (WebCore::SVGPathSegListBuilder::moveTo):
3464         (WebCore::SVGPathSegListBuilder::lineTo):
3465         (WebCore::SVGPathSegListBuilder::lineToHorizontal):
3466         (WebCore::SVGPathSegListBuilder::lineToVertical):
3467         (WebCore::SVGPathSegListBuilder::curveToCubic):
3468         (WebCore::SVGPathSegListBuilder::curveToCubicSmooth):
3469         (WebCore::SVGPathSegListBuilder::curveToQuadratic):
3470         (WebCore::SVGPathSegListBuilder::curveToQuadraticSmooth):
3471         (WebCore::SVGPathSegListBuilder::arcTo):
3472         (WebCore::SVGPathSegListBuilder::closePath):
3473         * svg/SVGPathSegListBuilder.h: Added.
3474
3475 2010-07-15  Yuzo Fujishima  <yuzo@google.com>
3476
3477         Reviewed by Dan Bernstein.
3478
3479         Fix for Bug 42342 - Font download error for an @font-face rule invalidates other @font-face rules for the same font-family
3480         https://bugs.webkit.org/show_bug.cgi?id=42342
3481
3482         Test: fast/css/font-face-download-error.html
3483
3484         * css/CSSSegmentedFontFace.cpp:
3485         (WebCore::CSSSegmentedFontFace::isValid): Valid if at least one font
3486         face is valid.
3487         (WebCore::CSSSegmentedFontFace::getFontData): Check validity for each
3488         font face.
3489         * css/CSSSegmentedFontFace.h: Make isValid private.
3490
3491 2010-07-19  Kent Tamura  <tkent@chromium.org>
3492
3493         Reviewed by Shinichiro Hamaji.
3494
3495         [Chromium] Fix style errors of RenderThemeChromiumWin.cpp
3496         https://bugs.webkit.org/show_bug.cgi?id=42568
3497
3498         * rendering/RenderThemeChromiumWin.cpp:
3499         (WebCore::):
3500         (WebCore::getNonClientMetrics):
3501         (WebCore::systemFontSize):
3502         (WebCore::pointsToPixels):
3503         (WebCore::querySystemBlinkInterval):
3504         (WebCore::RenderThemeChromiumWin::platformActiveSelectionBackgroundColor):
3505         (WebCore::RenderThemeChromiumWin::platformInactiveSelectionBackgroundColor):
3506         (WebCore::RenderThemeChromiumWin::platformActiveSelectionForegroundColor):
3507         (WebCore::RenderThemeChromiumWin::platformActiveTextSearchHighlightColor):
3508         (WebCore::RenderThemeChromiumWin::paintButton):
3509         (WebCore::RenderThemeChromiumWin::paintSliderTrack):
3510         (WebCore::RenderThemeChromiumWin::paintMenuList):
3511         (WebCore::RenderThemeChromiumWin::paintTextFieldInternal):
3512         (WebCore::RenderThemeChromiumWin::paintInnerSpinButton):
3513
3514 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3515
3516         (Hopefully the last) Windows build fix.
3517         
3518         Version of CoreServices.h in WebKit Support Libraries uses
3519         pragma once, unlike the version I have locally. Switch based
3520         on __COLORSYNCDEPRECATED__ instead.
3521
3522         * WebCorePrefix.h:
3523
3524 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3525
3526         Windows build fix.
3527
3528         * WebCorePrefix.h:
3529
3530 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3531
3532         Build fix.
3533
3534         * WebCorePrefix.h:
3535
3536 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3537
3538         Build fix.
3539
3540         * WebCorePrefix.h:
3541
3542 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3543
3544         Windows Build fixes for new ColorSync API.
3545         We support both new and old APIs, since the newer headers aren't in the tree yet.
3546
3547         * WebCorePrefix.h: Removed include of CoreServices.h. Included via ColorSyncPriv.h instead, since header may not be present.
3548         * platform/graphics/cg/ColorCG.cpp:
3549         (WebCore::createCGColor): Conditionally use new ColorSync API.
3550         * platform/graphics/opentype/OpenTypeUtilities.cpp: Define Fixed if CoreServices.h doesn't.
3551         * platform/graphics/win/GraphicsLayerCACF.cpp:
3552         (WebCore::GraphicsLayerCACF::updateLayerDrawsContent): Use 0 instead of nil, since nil is no longer present via CoreServices.h.
3553         * platform/graphics/win/MediaPlayerPrivateQuickTimeVisualContext.cpp: Include AssertMacros.h.
3554         * platform/network/cf/ResourceErrorCF.cpp:
3555         (WebCore::ResourceError::operator CFErrorRef): Use 0 instead of nil, since nil is no longer present via CoreServices.h.
3556
3557 2010-07-19  Tony Gentilcore  <tonyg@chromium.org>
3558
3559         Reviewed by Dimitri Glazkov.
3560
3561         Only set unloadEventEnd when the unload event is actually fired
3562         https://bugs.webkit.org/show_bug.cgi?id=42607
3563
3564         r63689 introduced this ASSERT and it began failing on Qt and Mac debug.
3565         For some reason, it did not fail on Windows (which is where I was
3566         testing). The ASSERT was disabled in r63699.
3567
3568         Test: page-cache related layout tests don't crash in debug mode on Mac.
3569
3570         * loader/FrameLoader.cpp:
3571         (WebCore::FrameLoader::stopLoading):
3572
3573 2010-07-19  Andreas Kling  <andreas.kling@nokia.com>
3574
3575         Reviewed by Kenneth Rohde Christiansen.
3576
3577         [Qt] Don't unnecessarily copy QPainterPath in fillPath() and strokePath()
3578         https://bugs.webkit.org/show_bug.cgi?id=42513
3579
3580         Avoid making unnecessary deep-copies of QPainterPaths that will
3581         be discarded after use.
3582
3583         * platform/graphics/qt/GraphicsContextQt.cpp:
3584         (WebCore::GraphicsContext::fillPath):
3585         (WebCore::GraphicsContext::strokePath):
3586
3587 2010-07-19  Kenneth Russell  <kbr@google.com>
3588
3589         Reviewed by Nate Chapin.
3590
3591         WebGL rendering results must be made available to Canvas.toDataURL and 2D drawImage
3592         https://bugs.webkit.org/show_bug.cgi?id=34719
3593
3594         Fixed compiler warning introduced by original patch. No new tests;
3595         covered by existing tests.
3596
3597         * platform/graphics/mac/GraphicsContext3DMac.mm:
3598         (WebCore::GraphicsContext3D::paintRenderingResultsToCanvas):
3599
3600 2010-07-19  Anders Carlsson  <andersca@apple.com>
3601
3602         Reviewed by Darin Adler, Adam Roben, Dan Bernstein and Sam Weinig.
3603
3604         Handle NP_ASFILE and NP_ASFILEONLY transfer modes
3605         https://bugs.webkit.org/show_bug.cgi?id=42587
3606
3607         * WebCore.exp.in:
3608         Export functions from FileSystem.h
3609         
3610         * platform/mac/FileSystemMac.mm:
3611         (WebCore::openTemporaryFile):
3612         Try to create a temporary file using mkstemp.
3613
3614 2010-07-19  Anders Carlsson  <andersca@apple.com>
3615
3616         Reviewed by Sam Weinig.
3617
3618         Implement NPN_PostURLNotify
3619         https://bugs.webkit.org/show_bug.cgi?id=42602
3620
3621         Export symbols needed by WebKit2.
3622
3623         * WebCore.exp.in:
3624
3625 2010-07-19  Tony Gentilcore  <tonyg@chromium.org>
3626
3627         Unreviewed build fix.
3628
3629         Remove overzealous ASSERT from r63689
3630         https://bugs.webkit.org/show_bug.cgi?id=42606
3631
3632         No new tests because no new functionality.
3633
3634         * loader/FrameLoader.cpp:
3635         (WebCore::FrameLoader::stopLoading):
3636
3637 2010-07-19  Joseph Pecoraro  <joepeck@webkit.org>
3638
3639         Reviewed by Mark Rowe.
3640
3641         Web Inspector: Do Not Copy *.re2js Inspector Resources in XCode Build Phase
3642         https://bugs.webkit.org/show_bug.cgi?id=42601
3643
3644         Remove *.re2js files after copyin