2010-07-26 Shimeng (Simon) Wang <swang@google.com>
[WebKit-https.git] / WebCore / ChangeLog
1 2010-07-26  Shimeng (Simon) Wang  <swang@google.com>
2
3         Reviewed by Darin Adler.
4
5         Add auto hyphenation support for Android.
6         https://bugs.webkit.org/show_bug.cgi?id=42800
7
8         * platform/text/android/HyphenationAndroid.cpp: Added.
9         (WebCore::loadHyphenationDictionary):
10         (WebCore::lastHyphenLocation):
11
12 2010-07-26  Sam Weinig  <sam@webkit.org>
13
14         Reviewed by Anders Carlsson.
15
16         Patch for https://bugs.webkit.org/show_bug.cgi?id=43013
17         Part of <rdar://problem/8152434>
18         Add support for scrolling using the keyboard in WebKit2
19
20         Add a new export and make WindowsKeyboardCodes private.
21
22         * WebCore.exp.in:
23         * WebCore.xcodeproj/project.pbxproj:
24
25 2010-07-26  Darin Adler  <darin@apple.com>
26
27         Reviewed by Eric Carlson.
28
29         Don't handle empty URL attributes specially in reflection, even in cases such as <img src="">
30         https://bugs.webkit.org/show_bug.cgi?id=42087
31
32         * html/HTMLMediaElement.cpp:
33         (WebCore::HTMLMediaElement::insertedIntoDocument): Check if the src attribute is empty
34         by calling getAttribute instead of calling the src function. This keeps the behavior
35         correct with the reflection change, and has a side benefit of being more efficient as
36         well since there is no reason to resolve the URL.
37         (WebCore::HTMLMediaElement::loadResource): Call getNonEmptyURLAttribute directly here
38         instead of using the poster function.
39         (WebCore::HTMLMediaElement::getPluginProxyParams): Ditto.
40
41         * html/HTMLMediaElement.h: Removed the poster virtual function. The code that gets the
42         poster attribute is already video-specific and uses generic attribute fetching code, so
43         we no longer need a virtual function. Removed the src function. Since this would give
44         a different result from the JavaScript reflected src attribute, it's safer not to have
45         it, and it was easy to convert all the callers.
46
47         * html/HTMLMediaElement.idl: Removed the NonEmpty extended attribute from the src
48         attribute.
49
50         * html/HTMLSourceElement.cpp: Removed the src function for the same reason as above.
51         * html/HTMLSourceElement.h: Ditto.
52
53         * html/HTMLSourceElement.idl: Removed the NonEmpty extended attribute from the src
54         attribute.
55
56         * html/HTMLVideoElement.cpp:
57         (WebCore::HTMLVideoElement::parseMappedAttribute): Removed code to update m_posterURL,
58         which is no longer needed. Changed call site that used the poster() function to instead
59         use getNonEmptyURLAttribute.
60         (WebCore::HTMLVideoElement::updatePosterImage): Check if the poster attribute is empty
61         by calling getAttribute instead of calling the poster function. This keeps the behavior
62         correct with the reflection change, and has a side benefit of being more efficient as
63         well since there is no reason to resolve the URL.
64
65         * html/HTMLVideoElement.h: Removed the poster function for the same reason we removed
66         the src function above. Also removed now-unneeded m_posterURL data member.
67
68         * html/HTMLVideoElement.idl: Removed the NonEmpty extended attribute from the poster
69         attribute.
70
71 2010-07-26  Chris Fleizach  <cfleizach@apple.com>
72
73         Reviewed by Darin Adler.
74
75         AX: remove intValue() hasIntValue() from AccessibilityObject
76         https://bugs.webkit.org/show_bug.cgi?id=43006
77
78         These methods are very Mac centric. The mac centric code should be moved into the Mac wrapper.
79
80         No new tests, existing tests cover changes to logic.
81
82         * accessibility/AccessibilityObject.cpp:
83         (WebCore::AccessibilityObject::checkboxOrRadioValue):
84         * accessibility/AccessibilityObject.h:
85         (WebCore::):
86         * accessibility/AccessibilityRenderObject.cpp:
87         (WebCore::AccessibilityRenderObject::selectedRadioButton):
88         (WebCore::AccessibilityRenderObject::selectedTabItem):
89         (WebCore::AccessibilityRenderObject::checkboxOrRadioValue):
90         * accessibility/AccessibilityRenderObject.h:
91         * accessibility/mac/AccessibilityObjectWrapper.mm:
92         (-[AccessibilityObjectWrapper accessibilityAttributeValue:]):
93
94 2010-07-26  Ryosuke Niwa  <rniwa@webkit.org>
95
96         Reviewed by Darin Adler.
97
98         Applying inline style to a text node whose parent is an inline editable root causes crash
99         https://bugs.webkit.org/show_bug.cgi?id=39989
100
101         The crash was caused by splitTextElementAtStart and splitTextElementAtEnd assuming that the parent
102         and the grandparent of the specified text node is editable.
103
104         Modified splitTextElementAtStart and splitTextElementAtEnd so that they call splitTextAtStart
105         and splitTextAtEnd respectively when the grandparent is not editable.
106
107         Also modified SplitTextNodeContainingElement to exit early if the grandparent of m_text is not editable.
108
109         Test: editing/style/style-text-node-without-editable-parent.html
110
111         * editing/ApplyStyleCommand.cpp:
112         (WebCore::ApplyStyleCommand::splitTextElementAtStart):
113         (WebCore::ApplyStyleCommand::splitTextElementAtEnd):
114         * editing/SplitTextNodeContainingElementCommand.cpp:
115         (WebCore::SplitTextNodeContainingElementCommand::doApply):
116
117 2010-07-26  Simon Fraser  <simon.fraser@apple.com>
118
119         Reviewed by Anders Carlsson.
120
121         Gmail compose is upside-down after r64054
122         https://bugs.webkit.org/show_bug.cgi?id=43000
123         
124         We need to call setGeometryOrientation() whenever the iframe compositing layer attachment changes,
125         not just when requiresScrollLayer() returns false. This ensures that the orientation is updated
126         correctly.
127
128         Test: compositing/iframes/iframe-content-flipping.html
129
130         * rendering/RenderLayerCompositor.cpp:
131         (WebCore::RenderLayerCompositor::ensureRootPlatformLayer):
132
133 2010-07-26  Justin Schuh  <jschuh@chromium.org>
134
135         Reviewed by Darin Fisher.
136
137         Check history state against origin before setting
138         https://bugs.webkit.org/show_bug.cgi?id=42858
139
140         Tests: fast/loader/stateobjects/replacestate-base-illegal.html
141                fast/loader/stateobjects/replacestate-base-legal.html
142
143         * page/History.cpp:
144         (WebCore::History::urlForState):
145         (WebCore::History::stateObjectAdded):
146
147 2010-07-26  Martin Robinson  <mrobinson@igalia.com>
148
149         Reviewed by Xan Lopez.
150
151         [GTK] Get rid of libgdom
152         https://bugs.webkit.org/show_bug.cgi?id=42378
153
154         Remove all build steps for building GObject DOM bindings.
155
156         * GNUmakefile.am: Remove GObject DOM bindings build.
157
158 2010-07-26  Mark Rowe  <mrowe@apple.com>
159
160         Reviewed by Dan Bernstein.
161
162         <http://webkit.org/b/42990> REGRESSION (r63854): Safari RSS layout is broken due to changes in vendor prefix handling
163
164         In r63854 support for properties with the -khtml vendor prefix was removed. This breaks the layout of
165         Safari's RSS interface, several Dashboard widgets that are included with Mac OS X, and at least one
166         popular third-party Mac OS X application. This change will need to be revisited in a manner that has
167         a clearer backwards-compatibility strategy. <http://webkit.org/b/42093> will be reopened to track that.
168
169         * css/CSSParser.cpp:
170         (WebCore::cssPropertyID): Revert the WebCore change from r63854.
171
172 2010-07-26  Brady Eidson  <beidson@apple.com>
173
174         Reviewed by Sam Weinig.
175
176         Add ability for WK2 to register a scheme to always be displayed as an empty document
177         https://bugs.webkit.org/show_bug.cgi?id=42995
178
179         No new tests. (Currently not possible to test in WebKit1 DRT, can be testing in future WK2 tester)
180
181         * WebCore.exp.in:
182         
183         * loader/MainResourceLoader.cpp:
184         (WebCore::shouldLoadAsEmptyDocument):
185         
186         * platform/SchemeRegistry.cpp:
187         (WebCore::emptyDocumentSchemes):
188         (WebCore::SchemeRegistry::registerURLSchemeAsEmptyDocument):
189         (WebCore::SchemeRegistry::shouldLoadURLSchemeAsEmptyDocument):
190         * platform/SchemeRegistry.h:
191
192 2010-07-26  Patrick Gansterer  <paroga@paroga.com>
193
194         Reviewed by Kent Tamura.
195
196         [WINCE] Port ClipboardUtilities to WinCE
197         https://bugs.webkit.org/show_bug.cgi?id=42929
198
199         Make ClipboardUtilitiesWin more portable and remove the global namespace prefix
200         from GlobalLock/GlobalUnlock, since they are only macros on WinCE.
201
202         * platform/win/ClipboardUtilitiesWin.cpp:
203         (WebCore::urlFromPath):
204         (WebCore::getWebLocData):
205         (WebCore::createGlobalData):
206         (WebCore::getURL):
207         (WebCore::getPlainText):
208         (WebCore::fragmentFromHTML):
209
210 2010-07-26  Patrick Gansterer  <paroga@paroga.com>
211
212         Reviewed by Adam Roben.
213
214         [WINCE] Buildfix for BitmapImage.h
215         https://bugs.webkit.org/show_bug.cgi?id=42931
216
217         Use AffineTransform insted of TransformationMatrix.
218
219         * platform/graphics/BitmapImage.h:
220
221 2010-07-26  Dirk Schulze  <krit@webkit.org>
222
223         Unreviewed sort of XCode project file.
224
225         * WebCore.xcodeproj/project.pbxproj:
226
227 2010-07-26  David Kilzer  <ddkilzer@apple.com>
228
229         <http://webkit.org/b/42982> Streamline Inspector Source build phase script should use CONFIGURATION not BUILD_STYLE
230         <rdar://problem/6341764>
231
232         Reviewed by Mark Rowe.
233
234         * WebCore.xcodeproj/project.pbxproj: Switched Streamline
235         Inspector Source build phase script to use CONFIGURATION instead
236         of BUILD_STYLE.
237
238 2010-07-23  Ojan Vafai  <ojan@chromium.org>
239
240         Reviewed by Darin Adler.
241
242         crash in insertParagraph in a contentEditable list
243         https://bugs.webkit.org/show_bug.cgi?id=42919
244
245         Don't try to break out of a list item if the list it's in
246         is the rootEditableElement.
247
248         * editing/CompositeEditCommand.cpp:
249         (WebCore::CompositeEditCommand::breakOutOfEmptyListItem):
250
251 2010-07-26  Marcus Bulach  <bulach@chromium.org>
252
253         Reviewed by Jeremy Orlow.
254
255         Moves IDBKeyRange to use IDBKey rather than SerializedScriptValue / IDBAny.
256         https://bugs.webkit.org/show_bug.cgi?id=42583
257
258         Tests: existing storage/indexeddb/idb-keyrange.html must pass.
259
260         * storage/IDBKeyRange.cpp:
261         (WebCore::IDBKeyRange::IDBKeyRange):
262         * storage/IDBKeyRange.h:
263         (WebCore::IDBKeyRange::create):
264         (WebCore::IDBKeyRange::left):
265         (WebCore::IDBKeyRange::right):
266         * storage/IDBKeyRange.idl:
267         * storage/IndexedDatabaseRequest.cpp:
268         (WebCore::IndexedDatabaseRequest::makeSingleKeyRange):
269         (WebCore::IndexedDatabaseRequest::makeLeftBoundKeyRange):
270         (WebCore::IndexedDatabaseRequest::makeRightBoundKeyRange):
271         (WebCore::IndexedDatabaseRequest::makeBoundKeyRange):
272         * storage/IndexedDatabaseRequest.h:
273         * storage/IndexedDatabaseRequest.idl:
274
275 2010-07-26  Simon Fraser  <simon.fraser@apple.com>
276
277         Reviewed by Anders Carlsson.
278
279         Composited layers don't scroll in WebKit2
280         https://bugs.webkit.org/show_bug.cgi?id=42771
281
282         Rename two methods on RenderLayerCompositor to make their use more clear, and call them
283         when the FrameView gets resized and scrolled.
284
285         * page/FrameView.h:
286         * page/FrameView.cpp:
287         (WebCore::FrameView::setFrameRect): override so we know when the size of the FrameView changes, so that the
288         RenderLayerCompositor can update its clipping layer.
289         (WebCore::FrameView::scrollPositionChanged): Call compositor()->frameViewDidScroll().
290
291         * rendering/RenderLayerBacking.cpp:
292         (WebCore::RenderLayerBacking::updateAfterWidgetResize): updateContentLayerOffset() was renamed.
293
294         * rendering/RenderLayerCompositor.h:
295         * rendering/RenderLayerCompositor.cpp:
296         (WebCore::RenderLayerCompositor::frameViewDidChangeSize): Renamed from updateContentLayerOffset().
297         (WebCore::RenderLayerCompositor::frameViewDidScroll): Renamed from updateContentLayerScrollPosition().
298         (WebCore::RenderLayerCompositor::shouldPropagateCompositingToEnclosingIFrame): Propagate compositing on Mac
299         for any FrameView that doesn't have a native widget, which is the case with WebKit2.
300         (WebCore::RenderLayerCompositor::requiresScrollLayer): New method to tell us whether we need to handle
301         scrolling (and clipping) ourselves.
302         (WebCore::RenderLayerCompositor::ensureRootPlatformLayer): Do geometry flipping on the root platform
303         layer only if we're not doing our own scrolling, and use the new requiresScrollLayer() method.
304         (WebCore::RenderLayerCompositor::attachRootPlatformLayer): Send out the rootPlatformLayer(),
305         so that it returns the clipping layer if we have one.
306
307 2010-07-26  Tony Gentilcore  <tonyg@chromium.org>
308
309         Reviewed by Darin Fisher.
310
311         Move DocumentLoadTiming struct to a new file
312         https://bugs.webkit.org/show_bug.cgi?id=42917
313
314         Also makes DocumentLoadTiming Noncopyable.
315
316         No new tests because no new functionality.
317
318         * GNUmakefile.am:
319         * WebCore.gypi:
320         * WebCore.vcproj/WebCore.vcproj:
321         * WebCore.xcodeproj/project.pbxproj:
322         * loader/DocumentLoadTiming.h: Added.
323         (WebCore::DocumentLoadTiming::DocumentLoadTiming):
324         * loader/DocumentLoader.h:
325         * loader/FrameLoader.cpp:
326         * loader/FrameLoaderTypes.h:
327         * loader/MainResourceLoader.cpp:
328         * page/Timing.cpp:
329
330 2010-07-26  Mark Rowe  <mrowe@apple.com>
331
332         Fix an #if in WebCore.exp.in to correctly detect whether SVG filters are enabled.
333
334         * WebCore.exp.in:
335
336 2010-07-26  Steve Block  <steveblock@google.com>
337
338         Reviewed by Jeremy Orlow.
339
340         DeviceOrientation event listeners should never be called synchronously from addEventListener()
341         https://bugs.webkit.org/show_bug.cgi?id=42304
342
343         No new tests as there are currently no implementations to test. Tests will be
344         added once LayoutTestController methods for providing a mock implementation are
345         in place, see Bug 39589.
346
347         * dom/DeviceOrientationController.cpp:
348         (WebCore::DeviceOrientationController::DeviceOrientationController):
349         (WebCore::DeviceOrientationController::timerFired):
350         (WebCore::DeviceOrientationController::addListener):
351         (WebCore::DeviceOrientationController::removeListener):
352         (WebCore::DeviceOrientationController::removeAllListeners):
353         * dom/DeviceOrientationController.h:
354
355 2010-07-26  Andreas Kling  <andreas.kling@nokia.com>
356
357         Reviewed by Tor Arne Vestbø.
358
359         [Qt] Canvas: Optimize GraphicsContext::addPath()
360
361         Avoid "deep concatenating" the canvas path onto the GC path
362         if the GC path has no elements.
363
364         * platform/graphics/qt/GraphicsContextQt.cpp:
365         (WebCore::GraphicsContext::addPath):
366
367 2010-07-23  Stephen White  <senorblanco@chromium.org>
368
369         Reviewed by Darin Fisher.
370
371         Implement OpenGLES2 helper classes.
372         https://bugs.webkit.org/show_bug.cgi?id=42905
373
374         GLES2Canvas and GLES2Texture are some simple OpenGLES2 helper classes
375         for managing state, samplers and textures.
376
377         When this is hooked up, it will be covered by many layout tests.
378
379         * WebCore.gypi:
380         Add new files to chromium build (contents are protected by #if's for now).
381         * platform/graphics/chromium/GLES2Canvas.cpp: Added.
382         (WebCore::affineTo3x3):
383         (WebCore::GLES2Canvas::State::State):
384         (WebCore::GLES2Canvas::GLES2Canvas):
385         (WebCore::GLES2Canvas::~GLES2Canvas):
386         (WebCore::GLES2Canvas::clearRect):
387         (WebCore::GLES2Canvas::fillRect):
388         (WebCore::GLES2Canvas::setFillColor):
389         (WebCore::GLES2Canvas::setAlpha):
390         (WebCore::GLES2Canvas::translate):
391         (WebCore::GLES2Canvas::rotate):
392         (WebCore::GLES2Canvas::scale):
393         (WebCore::GLES2Canvas::concatCTM):
394         (WebCore::GLES2Canvas::save):
395         (WebCore::GLES2Canvas::restore):
396         (WebCore::GLES2Canvas::drawTexturedRect):
397         (WebCore::GLES2Canvas::setCompositeOperation):
398         (WebCore::GLES2Canvas::applyCompositeOperator):
399         (WebCore::GLES2Canvas::getQuadVertices):
400         (WebCore::GLES2Canvas::getQuadIndices):
401         (WebCore::loadShader):
402         (WebCore::GLES2Canvas::getSimpleProgram):
403         (WebCore::GLES2Canvas::getTexProgram):
404         (WebCore::GLES2Canvas::createTexture):
405         (WebCore::GLES2Canvas::getTexture):
406         (WebCore::GLES2Canvas::checkGLError):
407         * platform/graphics/chromium/GLES2Canvas.h: Added.
408         (WebCore::GLES2Canvas::gles2Context):
409         * platform/graphics/chromium/GLES2Texture.cpp: Added.
410         (WebCore::GLES2Texture::GLES2Texture):
411         (WebCore::GLES2Texture::~GLES2Texture):
412         (WebCore::GLES2Texture::create):
413         (WebCore::convertFormat):
414         (WebCore::GLES2Texture::load):
415         (WebCore::GLES2Texture::bind):
416         * platform/graphics/chromium/GLES2Texture.h: Added.
417         (WebCore::GLES2Texture::):
418         (WebCore::GLES2Texture::format):
419         (WebCore::GLES2Texture::width):
420         (WebCore::GLES2Texture::height):
421
422
423 2010-07-26  Jeremy Orlow  <jorlow@chromium.org>
424
425         Reviewed by Steve Block.
426
427         Fix IndexedDB release build on mac
428         https://bugs.webkit.org/show_bug.cgi?id=42377
429
430         * storage/IDBDatabaseImpl.cpp:
431         (WebCore::IDBDatabaseImpl::objectStore):
432
433 2010-07-26  Satish Sampath  <satish@chromium.org>
434
435         Reviewed by Steve Block.
436
437         Add WebKit plumbing to connect speech requests and callbacks between WebCore and the embedder.
438         https://bugs.webkit.org/show_bug.cgi?id=42367
439
440         No new tests, the relevant LayoutTestController bindings will be added in the next patch.
441
442         * page/SpeechInput.cpp: renamed a method.
443         * page/SpeechInput.h: renamed a method.
444         * page/SpeechInputClient.h: added an extra method to optionally let users stop recording once they have spoken.
445         * page/SpeechInputClientListener.h: renamed a method.
446         * page/SpeechInputListener.h: renamed a method.
447
448 2010-07-26  Andreas Kling  <andreas.kling@nokia.com>
449
450         Reviewed by Kenneth Rohde Christiansen.
451
452         [Qt] getImageData(): Single-pass RGB->BGR and un-premultiplication
453         https://bugs.webkit.org/show_bug.cgi?id=42945
454
455         Combine the two operations into a single loop over the pixel data.
456         Yields a considerable FPS gain on http://www.semantix.gr/statue.html
457
458         * platform/graphics/qt/ImageBufferQt.cpp:
459         (WebCore::getImageData):
460
461 2010-07-26  Pavel Feldman  <pfeldman@chromium.org>
462
463         Reviewed by Shinichiro Hamaji.
464
465         Web Inspector: make sure proxy objects are JSON-stringifiable.
466
467         https://bugs.webkit.org/show_bug.cgi?id=42961
468
469         * inspector/front-end/InjectedScript.js:
470         (injectedScriptConstructor):
471
472 2010-07-26  Alexander Pavlov  <apavlov@chromium.org>
473
474         Reviewed by Joseph Pecoraro.
475
476         Inspector should remember resources sorting set by the user
477         https://bugs.webkit.org/show_bug.cgi?id=19208
478
479         Sorting options both for time and size graphs are memorized in a single property,
480         which makes them possible to retrieve in a single message (e.g. useful for remote debugging).
481
482         * inspector/front-end/ResourcesPanel.js:
483         (WebInspector.ResourcesPanel.prototype.populateSidebar):
484         (WebInspector.ResourcesPanel.prototype._settingsLoaded):
485         (WebInspector.ResourcesPanel.prototype._loadSortOptions):
486         (WebInspector.ResourcesPanel.prototype._loadSortOptionForGraph):
487         (WebInspector.ResourcesPanel.prototype._graphSelected):
488         (WebInspector.ResourcesPanel.prototype._changeSortingFunction):
489         (WebInspector.ResourcesPanel.prototype._selectedOptionNameForGraph):
490         (WebInspector.ResourcesPanel.prototype._doChangeSortingFunction):
491         * inspector/front-end/Settings.js:
492         (WebInspector.populateApplicationSettings):
493
494 2010-07-25  Ryosuke Niwa  <rniwa@webkit.org>
495
496         Reviewed by Darin Adler.
497
498         Redo fails after text node is split by SplitTextNodeCommand
499         https://bugs.webkit.org/show_bug.cgi?id=42941
500
501         The bug was caused by SplitTextNodeCommand not implementing doReapply.
502         Implemented doReapply and added insertText1AndTrimText2 to share the code between doApply and doReapply.
503
504         Test: editing/undo/redo-split-text-node.html
505               editing/undo/redo-split-text-node-with-removal.html
506
507         * editing/SplitTextNodeCommand.cpp:
508         (WebCore::SplitTextNodeCommand::doApply): Calls executeApply.
509         (WebCore::SplitTextNodeCommand::doUnapply): Don't delete m_text1, which is later used in doReapply.
510         (WebCore::SplitTextNodeCommand::doReapply): Added.
511         (WebCore::SplitTextNodeCommand::insertText1AndTrimText2): Added.
512         * editing/SplitTextNodeCommand.h:
513
514 2010-07-23  Philippe Normand  <pnormand@igalia.com>
515
516         Reviewed by Eric Carlson.
517
518         [GTK] media/media-can-play-mpeg-audio.html fails
519         https://bugs.webkit.org/show_bug.cgi?id=42895
520
521         Fixed advertizing of mp3, mpeg and mp4 mime-types in the player.
522
523         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
524         (WebCore::mimeTypeCache):
525
526 2010-07-23  Philippe Normand  <pnormand@igalia.com>
527
528         Reviewed by Eric Carlson.
529
530         LayoutTests/media/media-document-audio-size.html failure
531         https://bugs.webkit.org/show_bug.cgi?id=38376
532
533         Advertize audio/x-wav along with audio/wav when appropriate,
534         otherwise WebKit attempts to use a plugin to play the media.
535
536         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
537         (WebCore::mimeTypeCache):
538
539 2010-07-25  Ryosuke Niwa  <rniwa@webkit.org>
540
541         Reviewed by Kent Tamura.
542
543         splitTextAt*IfNeed and splitTextElementAt*IfNeed need to be cleaned up
544         https://bugs.webkit.org/show_bug.cgi?id=42937
545
546         Isolated the code to decide whether or not text node should be split into isValidCaretPositionInTextNode.
547         Moved the condition check out of *IfNeeded methods to applyRelativeFontStyleChange and applyInlineStyle.
548
549         No new tests added since this is a clean up.
550
551         * editing/ApplyStyleCommand.cpp:
552         (WebCore::ApplyStyleCommand::applyRelativeFontStyleChange): Uses isValidCaretPositionInTextNode.
553         (WebCore::ApplyStyleCommand::applyInlineStyle): Uses isValidCaretPositionInTextNode.
554         (WebCore::ApplyStyleCommand::splitTextAtStart): Renamed from splitTextAtStartIfNeeded.
555         (WebCore::ApplyStyleCommand::splitTextAtEnd): Renamed from splitTextAtEndIfNeeded.
556         (WebCore::ApplyStyleCommand::splitTextElementAtStart): Renamed from splitTextElementAtStartIfNeeded.
557         (WebCore::ApplyStyleCommand::splitTextElementAtEnd): Renamed from splitTextElementAtEndIfNeeded.
558         (WebCore::ApplyStyleCommand::isValidCaretPositionInTextNode): Returns true if the position lies within a text node.
559         * editing/ApplyStyleCommand.h:
560
561 2010-07-25  Patrick Gansterer  <paroga@paroga.com>
562
563         Reviewed by Kent Tamura.
564
565         [WINCE] Buildfix for ImageBufferWince
566         https://bugs.webkit.org/show_bug.cgi?id=42928
567
568         * platform/graphics/wince/ImageBufferWince.cpp:
569         (WebCore::):
570         (WebCore::BufferedImage::draw): Pass through ColorSpace parameter.
571         (WebCore::BufferedImage::drawPattern): Ditto.
572         (WebCore::ImageBuffer::platformTransformColorSpace): Added.
573
574 2010-07-25  Patrick Gansterer  <paroga@paroga.com>
575
576         Reviewed by Kent Tamura.
577
578         [WINCE] Buildfix for PopupMenuWin.cpp
579         https://bugs.webkit.org/show_bug.cgi?id=42932
580
581         Add #if !OS(WINCE) around unsupported window messages.
582
583         * platform/win/PopupMenuWin.cpp:
584         (WebCore::PopupMenu::show):
585         (WebCore::PopupMenu::wndProc):
586
587 2010-07-25  Adam Barth  <abarth@webkit.org>
588
589         Reviewed by Eric Seidel.
590
591         Implement adjustMathMLAttributes
592         https://bugs.webkit.org/show_bug.cgi?id=42952
593
594         This is kind of mechanical overkill for adjusting one attribute name,
595         but it's nice to use the same machinery as SVG.
596
597         * html/HTMLTreeBuilder.cpp:
598         * mathml/mathattrs.in:
599
600 2010-07-25  Adam Barth  <abarth@webkit.org>
601
602         Reviewed by Eric Seidel.
603
604         Functions supporting the legacy tree builder are confusing
605         https://bugs.webkit.org/show_bug.cgi?id=42951
606
607         This patch just inlines these functions into their one call site and
608         removes the notImplemented() calls.  We thought we were going to grow
609         these to support the new tree builder, but we took a different path.
610         Previous to this patch, these functions were confusing because they
611         looked like part of the new tree builder.
612
613         * html/HTMLTreeBuilder.cpp:
614         (WebCore::HTMLTreeBuilder::passTokenToLegacyParser):
615         * html/HTMLTreeBuilder.h:
616
617 2010-07-25  Adam Barth  <abarth@webkit.org>
618
619         Reviewed by Eric Seidel.
620
621         Recent HTML5 TreeBuilder Regression?  All canvas tests ASSERT
622         https://bugs.webkit.org/show_bug.cgi?id=42948
623
624         Remove ASSERT and replace with correct branch.  I put this ASSERT in
625         to make sure we had test coverage for this branch.  I guess we do.  :)
626
627         * html/HTMLConstructionSite.cpp:
628         (WebCore::HTMLConstructionSite::attach):
629
630 2010-07-25  Pavel Feldman  <pfeldman@chromium.org>
631
632         Reviewed by Timothy Hatcher.
633
634         Web Inspector: check of jQuery object type prior to treating it as an array.
635
636         https://bugs.webkit.org/show_bug.cgi?id=42946
637
638         * inspector/front-end/InjectedScript.js:
639         (injectedScriptConstructor):
640
641 2010-07-24  Daniel Bates  <dbates@rim.com>
642
643         Unreviewed, build fix when building with WML support.
644
645         PageCache.cpp compile error when WML is enabled
646         https://bugs.webkit.org/show_bug.cgi?id=40765
647
648         Remove extraneous semicolon.
649
650         Note, building with WML support enabled is broken
651         even with this fix. See Bug #42943 for more details.
652
653         * history/PageCache.cpp:
654         (WebCore::PageCache::canCachePageContainingThisFrame):
655
656 2010-07-24  Adam Barth  <abarth@webkit.org>
657
658         Reviewed by Eric Seidel.
659
660         Atomize strings used by the HTML5 tree builder
661         https://bugs.webkit.org/show_bug.cgi?id=42875
662
663         This is about a 1% improvement to the parsing benchmark.
664
665         * html/HTMLTagNames.in:
666         * html/HTMLTreeBuilder.cpp:
667         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
668         (WebCore::HTMLTreeBuilder::processStartTag):
669         (WebCore::HTMLTreeBuilder::processEndTagForInBody):
670         (WebCore::HTMLTreeBuilder::processStartTagForInHead):
671
672 2010-07-24  Patrick Gansterer  <paroga@paroga.com>
673
674         Reviewed by Darin Adler.
675
676         Buildfix for !ENABLE(INSPECTOR) after r63891.
677         https://bugs.webkit.org/show_bug.cgi?id=42925
678
679         * bindings/js/ScriptValue.cpp: Added property svn:eol-style.
680         * bindings/js/ScriptValue.h: Added property svn:eol-style.
681         * inspector/ConsoleMessage.cpp: Added property svn:eol-style.
682         * inspector/ConsoleMessage.h: Added property svn:eol-style.
683
684 2010-07-24  Adam Barth  <abarth@webkit.org>
685
686         Reviewed by Eric Seidel.
687
688         Re-order whitespace checks in the tokenizer
689         https://bugs.webkit.org/show_bug.cgi?id=42923
690
691         This saves about 0.5% on the parsing benchmark.
692
693         * html/HTMLTokenizer.cpp:
694
695 2010-07-24  Andreas Kling  <andreas.kling@nokia.com>
696
697         Reviewed by Simon Hausmann.
698
699         [Qt] tst_QWebFrame::callQtInvokable() fails
700         https://bugs.webkit.org/show_bug.cgi?id=41065
701
702         Converting JS objects to QVariantMaps was broken.
703         This is a partial revert of <http://trac.webkit.org/changeset/61478>
704
705         * bridge/qt/qt_runtime.cpp:
706         (JSC::Bindings::convertValueToQVariant): Remove erroneous crash guard
707         around JSObject::getPropertyNames()
708
709 2010-07-24  Adam Barth  <abarth@webkit.org>
710
711         Reviewed by Eric Seidel.
712
713         Save a per-character branch in the HTML parser
714         https://bugs.webkit.org/show_bug.cgi?id=42921
715
716         The parsing benchmark claims this is a wash, but it makes me feel
717         better.
718
719         * html/HTMLToken.h:
720         (WebCore::HTMLToken::clear):
721         (WebCore::HTMLToken::beginStartTag):
722         (WebCore::HTMLToken::beginEndTag):
723         (WebCore::HTMLToken::ensureIsCharacterToken):
724         (WebCore::HTMLToken::beginComment):
725         (WebCore::HTMLToken::beginDOCTYPE):
726         * html/HTMLTokenizer.cpp:
727         (WebCore::HTMLTokenizer::bufferCharacter):
728
729 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
730
731         Reviewed by Darin Adler.
732
733         Buildfix for !ENABLE(FILTERS) after r63531.
734         https://bugs.webkit.org/show_bug.cgi?id=42914
735
736         * rendering/RenderSVGContainer.cpp: Add missing header.
737
738 2010-07-23  Kavita Kanetkar  <kkanetkar@chromium.org>
739
740         Reviewed by Joseph Pecoraro.
741
742         Implement remaining Inspector support for chrome's appcache
743         https://bugs.webkit.org/show_bug.cgi?id=42426
744
745         * inspector/InspectorApplicationCacheAgent.cpp:
746         (WebCore::InspectorApplicationCacheAgent::getApplicationCaches):
747         (WebCore::InspectorApplicationCacheAgent::buildObjectForApplicationCache):
748         (WebCore::InspectorApplicationCacheAgent::buildArrayForApplicationCacheResources):
749         (WebCore::InspectorApplicationCacheAgent::buildObjectForApplicationCacheResource):
750         * inspector/InspectorApplicationCacheAgent.h:
751         (WebCore::InspectorController::connectFrontend):
752         * loader/appcache/ApplicationCacheHost.cpp:
753         (WebCore::ApplicationCacheHost::fillResourceList):
754         (WebCore::ApplicationCacheHost::applicationCacheInfo):
755         * loader/appcache/ApplicationCacheHost.h: Changed creation + update times to double and moved appcache info here from InspectorApplicationCacheAgent.h
756         (WebCore::ApplicationCacheHost::CacheInfo::CacheInfo):
757         (WebCore::ApplicationCacheHost::ResourceInfo::ResourceInfo):
758
759 2010-07-23  Rafael Antognolli  <antognolli@profusion.mobi>
760
761         Reviewed by Antonio Gomes.
762
763         [EFL] Cleanup glib support (make it optional)
764         https://bugs.webkit.org/show_bug.cgi?id=42480
765
766         No new tests, no new features added.
767         Just add flags and directories for glib if necessary.
768         Put some ifdefs on glib calls.
769
770         * CMakeListsEfl.txt:
771         * platform/graphics/cairo/FontPlatformDataCairo.cpp:
772         (WebCore::FontPlatformData::FontPlatformData):
773
774 2010-07-23  James Robinson  <jamesr@chromium.org>
775
776         Reviewed by Dimitri Glazkov.
777
778         CanvasRenderingContext2D has lots of style errors
779         https://bugs.webkit.org/show_bug.cgi?id=42918
780
781         CanvasRenderingContext2D.h and .cpp have lots of style errors.  I'm planning to post patches
782         to both of these files soon, so to minimize merge conflicts and additional style violations
783         it seems best to fix all of these errors in bulk before posting any functional changes.  In
784         particular the header file is incorrectly indented so any correctly formatted addition will
785         be out of place.
786
787         * html/canvas/CanvasRenderingContext2D.cpp:
788         (WebCore::CanvasRenderingContext2D::CanvasRenderingContext2D):
789         (WebCore::CanvasRenderingContext2D::setFillStyle):
790         (WebCore::CanvasRenderingContext2D::transform):
791         (WebCore::CanvasRenderingContext2D::setTransform):
792         (WebCore::CanvasRenderingContext2D::arcTo):
793         (WebCore::CanvasRenderingContext2D::arc):
794         (WebCore::validateRectForCanvas):
795         (WebCore::CanvasRenderingContext2D::strokeRect):
796         (WebCore::CanvasRenderingContext2D::drawImage):
797         (WebCore::CanvasRenderingContext2D::createRadialGradient):
798         (WebCore::CanvasRenderingContext2D::createPattern):
799         (WebCore::CanvasRenderingContext2D::willDraw):
800         (WebCore::CanvasRenderingContext2D::createImageData):
801         (WebCore::CanvasRenderingContext2D::putImageData):
802         (WebCore::CanvasRenderingContext2D::setFont):
803         (WebCore::CanvasRenderingContext2D::drawTextInternal):
804         * html/canvas/CanvasRenderingContext2D.h:
805         (WebCore::CanvasRenderingContext2D::is2d):
806         (WebCore::CanvasRenderingContext2D::getLineCap):
807         (WebCore::CanvasRenderingContext2D::getLineJoin):
808         (WebCore::CanvasRenderingContext2D::state):
809         (WebCore::CanvasRenderingContext2D::):
810
811 2010-07-23  Adam Barth  <abarth@webkit.org>
812
813         Remove FIXME that we already fixed.
814
815         * html/HTMLTreeBuilder.cpp:
816         (WebCore::HTMLTreeBuilder::processCharacter):
817
818 2010-06-24  Dimitri Glazkov  <dglazkov@chromium.org>
819
820         Reviewed by Kent Tamura.
821
822         Move more form submission logic into FormSubmission.
823         https://bugs.webkit.org/show_bug.cgi?id=41110
824
825         Two major moves:
826         * Fold HTMLFormElement::prepareFormSubmission into FormSubmission::create
827         * Move non-static (non-FormData-building) methods of FormDataBuilder
828             to FormSubmission::Attributes.
829
830         Mechanical moving, no change in behavior, so no new tests.
831
832         * WebCore.xcodeproj/project.pbxproj: Gave FormDataBuilder.h "project" and
833             FormSubmission.h "private" visibility.
834         * html/HTMLFormElement.cpp:
835         (WebCore::HTMLFormElement::submit): Changed to use FormSubmission::create.
836         (WebCore::HTMLFormElement::parseMappedAttribute): Changed to use m_attributes.
837         * html/HTMLFormElement.h:
838         (WebCore::HTMLFormElement::enctype): Ditto.
839         (WebCore::HTMLFormElement::encoding): Ditto.
840         (WebCore::HTMLFormElement::acceptCharset): Ditto.
841         * loader/FormSubmission.cpp:
842         (WebCore::generateFormDataIdentifier): Moved from HTMLFormElement.
843         (WebCore::appendMailtoPostFormDataToURL): Ditto.
844         (WebCore::FormSubmission::Attributes::parseAction): Moved from FormDataBuilder.
845         (WebCore::FormSubmission::Attributes::parseEncodingType): Ditto.
846         (WebCore::FormSubmission::Attributes::parseMethodType): Ditto.
847         (WebCore::FormSubmission::FormSubmission): Made inline.
848         (WebCore::FormSubmission::create): Added logic that used to be in
849             HTMLFormElement::prepareFormSubmission.
850         * loader/FormSubmission.h:
851         (WebCore::FormSubmission::):
852         (WebCore::FormSubmission::Attributes::Attributes): Added new class.
853         (WebCore::FormSubmission::Attributes::method): Moved from FormDataBuilder.
854         (WebCore::FormSubmission::Attributes::isMultiPartForm): Ditto.
855         (WebCore::FormSubmission::Attributes::action): Ditto.
856         (WebCore::FormSubmission::Attributes::target): Ditto.
857         (WebCore::FormSubmission::Attributes::setTarget): Ditto.
858         (WebCore::FormSubmission::Attributes::encodingType): Ditto.
859         (WebCore::FormSubmission::Attributes::acceptCharset): Ditto.
860         (WebCore::FormSubmission::Attributes::setAcceptCharset): Ditto.
861         (WebCore::FormSubmission::target): Ditto.
862         (WebCore::FormSubmission::contentType): Const-ified return type.
863         (WebCore::FormSubmission::boundary): Ditto.
864         * platform/network/FormDataBuilder.cpp:
865         (WebCore::FormDataBuilder::encodingFromAcceptCharset): Renamed method for clarity,
866             made it use param instead of a class member.
867         * platform/network/FormDataBuilder.h:
868         (WebCore::FormDataBuilder::FormDataBuilder): Made constructor private.
869         * wml/WMLGoElement.cpp: Changed to use FormSubmision::Attributes.
870         * wml/WMLGoElement.h: Ditto.
871
872 2010-07-23  Adam Barth  <abarth@webkit.org>
873
874         Reviewed by Eric Seidel.
875
876         HTML5 Parser: document.write in a asynchronous script which is specified to load before page finish blows away document
877         https://bugs.webkit.org/show_bug.cgi?id=42365
878
879         This patch implements the changes Ian made to HTML5 to address these
880         issues.  Let's try it and see how well it works.
881
882         Test: http/tests/misc/write-from-dom-script.html
883
884         * dom/Document.cpp:
885         (WebCore::Document::Document):
886         (WebCore::Document::write):
887         * dom/Document.h:
888         (WebCore::Document::writeDisabled):
889         (WebCore::Document::setWriteDisabled):
890         * dom/DocumentParser.h:
891         (WebCore::DocumentParser::hasInsertionPoint):
892         * dom/ScriptElement.cpp:
893         (WebCore::ScriptElement::insertedIntoDocument):
894         (WebCore::ScriptElementData::ScriptElementData):
895         (WebCore::ScriptElementData::evaluateScript):
896         * dom/ScriptElement.h:
897         (WebCore::ScriptElementData::writeDisabled):
898         (WebCore::ScriptElementData::setWriteDisabled):
899         * html/HTMLDocumentParser.cpp:
900         (WebCore::HTMLDocumentParser::hasInsertionPoint):
901         (WebCore::HTMLDocumentParser::insert):
902         * html/HTMLDocumentParser.h:
903
904 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
905
906         Reviewed by Adam Roben.
907
908         [WINCE] Buildfix for PopupMenuWin.cpp
909         https://bugs.webkit.org/show_bug.cgi?id=42906
910
911         Use correct API for BitmapInfo landed in r63944.
912
913         * platform/win/PopupMenuWin.cpp:
914         (WebCore::PopupMenu::paint):
915
916 2010-07-23  fsamuel@chromium.org  <fsamuel@chromium.org>
917
918         Reviewed by David Hyatt.
919
920         Updated table painting to draw cell layers in the correct order.
921
922         Table Cell Layering
923         https://bugs.webkit.org/show_bug.cgi?id=40775
924
925         Tests: tables/layering/paint-test-layering-1.html
926                tables/layering/paint-test-layering-2.html
927
928         * accessibility/AccessibilityTable.cpp:
929         (WebCore::AccessibilityTable::isTableExposableThroughAccessibility):
930         (WebCore::AccessibilityTable::addChildren):
931         (WebCore::AccessibilityTable::cellForColumnAndRow):
932         * accessibility/AccessibilityTableCell.cpp:
933         (WebCore::AccessibilityTableCell::titleUIElement):
934         * accessibility/AccessibilityTableColumn.cpp:
935         (WebCore::AccessibilityTableColumn::headerObjectForSection):
936         * rendering/AutoTableLayout.cpp:
937         (WebCore::AutoTableLayout::recalcColumn):
938         * rendering/RenderTable.cpp:
939         (WebCore::RenderTable::splitColumn):
940         (WebCore::RenderTable::calcBorderLeft):
941         (WebCore::RenderTable::calcBorderRight):
942         (WebCore::RenderTable::cellAbove):
943         (WebCore::RenderTable::cellBelow):
944         (WebCore::RenderTable::cellBefore):
945         (WebCore::RenderTable::cellAfter):
946         * rendering/RenderTableSection.cpp:
947         (WebCore::RenderTableSection::RenderTableSection):
948         (WebCore::RenderTableSection::ensureRows):
949         (WebCore::RenderTableSection::addCell):
950         (WebCore::RenderTableSection::setCellWidths):
951         (WebCore::RenderTableSection::calcRowHeight):
952         (WebCore::RenderTableSection::layoutRows):
953         (WebCore::RenderTableSection::calcOuterBorderTop):
954         (WebCore::RenderTableSection::calcOuterBorderBottom):
955         (WebCore::RenderTableSection::calcOuterBorderLeft):
956         (WebCore::RenderTableSection::calcOuterBorderRight):
957         (WebCore::RenderTableSection::firstLineBoxBaseline):
958         (WebCore::compareCellPositions):
959         (WebCore::RenderTableSection::paintCell):
960         (WebCore::RenderTableSection::paintObject):
961         (WebCore::RenderTableSection::numColumns):
962         (WebCore::RenderTableSection::appendColumn):
963         (WebCore::RenderTableSection::splitColumn):
964         (WebCore::RenderTableSection::nodeAtPoint):
965         * rendering/RenderTableSection.h:
966         (WebCore::RenderTableSection::CellStruct::CellStruct):
967         (WebCore::RenderTableSection::CellStruct::primaryCell):
968         (WebCore::RenderTableSection::CellStruct::hasCells):
969         (WebCore::RenderTableSection::primaryCellAt):
970
971 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
972
973         Reviewed by Adam Roben.
974
975         [WINCE] Buildfix for Path.h
976         https://bugs.webkit.org/show_bug.cgi?id=42904
977
978         * platform/graphics/Path.h: Add missing typedef PlatformPath.
979
980 2010-07-23  Adam Barth  <abarth@webkit.org>
981
982         Reviewed by Eric Seidel.
983
984         Reserve capacity for attributes
985         https://bugs.webkit.org/show_bug.cgi?id=42877
986
987         This is about a 1.7% improvement to the parsing benchmark.
988
989         * html/HTMLToken.h:
990         (WebCore::AtomicHTMLToken::AtomicHTMLToken):
991
992 2010-07-23  Adam Barth  <abarth@webkit.org>
993
994         Reviewed by Eric Seidel.
995
996         Inline HTMLElementStack::top
997         https://bugs.webkit.org/show_bug.cgi?id=42879
998
999         According to sampling, HTMLElementStack::top is the hotest function in
1000         the treebuilder.  Inlining it is about a 0.3% improvement to the
1001         parsing benchmark.  This squares with how often it occurs in samples.
1002
1003         * html/HTMLElementStack.cpp:
1004         * html/HTMLElementStack.h:
1005         (WebCore::HTMLElementStack::top):
1006
1007 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
1008
1009         Reviewed by Adam Roben.
1010
1011         [WINCE] Buildfix for GraphicsContext
1012         https://bugs.webkit.org/show_bug.cgi?id=42888
1013
1014         Use AffineTransform insted of TransformationMatrix and
1015         add ColorSpace parameter to the drawBitmap methods.
1016
1017         * platform/graphics/GraphicsContext.h:
1018         * platform/graphics/wince/GraphicsContextWince.cpp:
1019         (WebCore::GraphicsContext::affineTransform):
1020         (WebCore::GraphicsContext::fillRoundedRect):
1021         (WebCore::GraphicsContext::drawBitmap):
1022         (WebCore::GraphicsContext::drawBitmapPattern):
1023
1024 2010-07-23  Rafael Antognolli  <antognolli@profusion.mobi>
1025
1026         Reviewed by Antonio Gomes.
1027
1028         [EFL] Add support for using libcurl network backend.
1029         https://bugs.webkit.org/show_bug.cgi?id=42286
1030
1031         Add files and flags for soup/curl specific backend being
1032         used.
1033         Put some ifdefs around soup/glib calls.
1034
1035         No new tests, no new features implemented.
1036
1037         * CMakeListsEfl.txt:
1038         * platform/efl/FileSystemEfl.cpp:
1039         (WebCore::fileSystemRepresentation):
1040
1041 2010-07-23  Adam Barth  <abarth@webkit.org>
1042
1043         Reviewed by Eric Seidel.
1044
1045         Add fast-reject path for InputStreamPreprocessor::peek
1046         https://bugs.webkit.org/show_bug.cgi?id=42881
1047
1048         This is about a 2.5% speedup on the parsing benchmark.
1049
1050         * html/HTMLTokenizer.h:
1051         (WebCore::HTMLTokenizer::InputStreamPreprocessor::peek):
1052
1053 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
1054
1055         Reviewed by Adam Roben.
1056
1057         [WINCE] Merge PasteboardWince.cpp into PasteboardWin.cpp
1058         https://bugs.webkit.org/show_bug.cgi?id=42885
1059
1060         Make PasteboardWin more portable and remove the global namespace prefix
1061         from GlobalLock/GlobalUnlock, since they are only macros on WinCE.
1062
1063         * platform/win/PasteboardWin.cpp:
1064         (WebCore::PasteboardOwnerWndProc):
1065         (WebCore::Pasteboard::Pasteboard):
1066         (WebCore::Pasteboard::plainText):
1067         (WebCore::Pasteboard::documentFragment):
1068
1069 2010-07-23  Justin Schuh  <jschuh@chromium.org>
1070
1071         Unreviewed, rolling out r63925.
1072         http://trac.webkit.org/changeset/63925
1073         https://bugs.webkit.org/show_bug.cgi?id=42858
1074
1075         Introduced a regression in history state.
1076
1077         * page/History.cpp:
1078         (WebCore::History::urlForState):
1079
1080 2010-07-23  Martin Robinson  <mrobinson@igalia.com>
1081
1082         Unreviewed build fix.
1083
1084         Fix make dist by aligning the source list with currently used headers.
1085
1086         * GNUmakefile.am:
1087
1088 2010-07-23  Alex Milowski  <alex@milowski.com>
1089
1090         Reviewed by Darin Adler.
1091
1092         Moved the 'mathml' group back to where it belongs as
1093         siblings of the other major folders.
1094
1095         * WebCore.xcodeproj/project.pbxproj:
1096
1097 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
1098
1099         Reviewed by Adam Roben.
1100
1101         Buildfix for WinCE graphics after r56015.
1102         https://bugs.webkit.org/show_bug.cgi?id=41354
1103
1104         Use fabs() insetead of abs().
1105
1106         * platform/graphics/wince/GraphicsContextWince.cpp:
1107         (WebCore::GraphicsContext::fillRect):
1108         * platform/graphics/wince/PlatformPathWince.cpp:
1109         (WebCore::PlatformPath::addArcTo):
1110
1111 2010-07-23  ryuan choi  <ryuan.choi@samsung.com>
1112
1113         Reviewed by Antonio Gomes.
1114
1115         [EFL]build break because of r63927
1116         https://bugs.webkit.org/show_bug.cgi?id=42864
1117
1118         * CMakeLists.txt:
1119
1120 2010-07-23  Adam Barth  <abarth@webkit.org>
1121
1122         Reviewed by Eric Seidel.
1123
1124         Inline HTMLTokenizer::InputStreamPreprocessor::peek
1125         https://bugs.webkit.org/show_bug.cgi?id=42880
1126
1127         According to sampling HTMLTokenizer::InputStreamPreprocessor::peek is
1128         the hotest function in the tokenizer.  Inlining this function is about
1129         a 2.2% speedup on the parser benchmark.  Turns out the tokenizer is
1130         much hotter than the tree builder.  :)
1131
1132         * html/HTMLTokenizer.h:
1133         (WebCore::HTMLTokenizer::InputStreamPreprocessor::peek):
1134
1135 2010-07-23  Andras Becsi  <abecsi@webkit.org>
1136
1137         Unreviewed build fix.
1138
1139         [Qt] Fix WebKit2 linkage.
1140
1141         No new tests needed.
1142
1143         * WebCore.pro: add InjectedBundlePageEditorClient.cpp to WK2 sources
1144
1145 2010-07-23  Balazs Kelemen  <kb@inf.u-szeged.hu>
1146
1147         Reviewed by Kenneth Rohde Christiansen.
1148
1149         [Qt] Setup the QtWebProcess
1150
1151         https://bugs.webkit.org/show_bug.cgi?id=42623
1152
1153         * WebCore.pro: Added WebProcessMainQt.cpp to SOURCES.
1154
1155 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1156
1157         Not reviewed: Chromium tests fix.
1158
1159         * inspector/CodeGeneratorInspector.pm:
1160
1161 2010-07-23  Patrick Gansterer  <paroga@paroga.com>
1162
1163         Reviewed by Adam Roben.
1164
1165         [WINCE] Buildfix for PluginDatabaseWin.cpp
1166         https://bugs.webkit.org/show_bug.cgi?id=41887
1167
1168         Define KEY_ENUMERATE_SUB_KEYS only if not defined.
1169
1170         * plugins/win/PluginDatabaseWin.cpp:
1171
1172 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1173
1174         Reviewed by Yury Semikhatsky.
1175
1176         Chromium DevTools: migrate from utility context-based injected
1177         script dispatch to upstream schema.
1178
1179         https://bugs.webkit.org/show_bug.cgi?id=42883
1180
1181         * bindings/v8/ScriptProfile.cpp:
1182         (WebCore::ScriptProfile::title):
1183         * inspector/InspectorController.cpp:
1184         (WebCore::InspectorController::getProfile):
1185         * inspector/front-end/InjectedScriptAccess.js:
1186
1187 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1188
1189         Not reviewed: rolling out r63963.
1190
1191         * bindings/v8/ScriptProfile.cpp:
1192         (WebCore::ScriptProfile::title):
1193         * inspector/InspectorController.cpp:
1194         (WebCore::InspectorController::getProfile):
1195         * inspector/front-end/InjectedScriptAccess.js:
1196
1197 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1198
1199         Reviewed by Yury Semikhatsky.
1200
1201         Chromium DevTools: migrate from utility context-based injected
1202         script dispatch to upstream schema.
1203
1204         https://bugs.webkit.org/show_bug.cgi?id=42883
1205
1206         * bindings/v8/ScriptProfile.cpp:
1207         (WebCore::ScriptProfile::title):
1208         * inspector/InspectorController.cpp:
1209         (WebCore::InspectorController::getProfile):
1210         * inspector/front-end/InjectedScriptAccess.js:
1211         (InjectedScriptAccess.get if):
1212         (InjectedScriptAccess):
1213
1214 2010-07-23  Yury Semikhatsky  <yurys@chromium.org>
1215
1216         Reviewed by Pavel Feldman.
1217
1218         Web Inspector: error info propagation in InspectorBackendDispatch should be improved
1219         https://bugs.webkit.org/show_bug.cgi?id=42873
1220
1221         * inspector/CodeGeneratorInspector.pm: report wrong parameter count and wrong
1222         parametere type errror to the dispatch method caller
1223
1224 2010-07-23  Pavel Feldman  <pfeldman@chromium.org>
1225
1226         Reviewed by Yury Semikhatsky.
1227
1228         Web Inspector: injectedScriptId is missing in the stack frame wrappers.
1229
1230         https://bugs.webkit.org/show_bug.cgi?id=42889
1231
1232         * inspector/front-end/InjectedScript.js:
1233         (injectedScriptConstructor):
1234
1235 2010-07-23  Shinichiro Hamaji  <hamaji@chromium.org>
1236
1237         Reviewed by Kent Tamura.
1238
1239         Backslash isn't transcoded in <input>
1240         https://bugs.webkit.org/show_bug.cgi?id=42878
1241
1242         * css/CSSStyleSelector.cpp:
1243         (WebCore::CSSStyleSelector::applyProperty):
1244
1245 2010-07-22  Ilya Tikhonovsky  <loislo@chromium.org>
1246
1247         Reviewed by Yury Semikhatsky.
1248
1249         WebInspector: CodeGeneratorInspector was extended and now it can
1250         generate InspectorBackedDispatcher's code. This dispatcher will
1251         validate and dispatch incoming inspector commands only by native
1252         code without using javaScript. That is necessary step for
1253         RemoteDebugging support.
1254         https://bugs.webkit.org/show_bug.cgi?id=42588
1255
1256         * GNUmakefile.am:
1257         * WebCore.gyp/WebCore.gyp:
1258         * WebCore.gypi:
1259         * WebCore.pri:
1260         * WebCore.vcproj/WebCore.vcproj:
1261         * WebCore.xcodeproj/project.pbxproj:
1262         * inspector/CodeGeneratorInspector.pm:
1263         * inspector/InspectorController.cpp:
1264         (WebCore::InspectorController::InspectorController):
1265         * inspector/InspectorController.h:
1266         (WebCore::InspectorController::inspectorBackendDispatcher):
1267         * inspector/front-end/ElementsPanel.js:
1268         (WebInspector.ElementsPanel.prototype.setDocument):
1269
1270 2010-07-22  Shinichiro Hamaji  <hamaji@chromium.org>
1271
1272         Reviewed by Kent Tamura.
1273
1274         Backslash is transcoded into yen sign even when non japanese font is specified
1275         https://bugs.webkit.org/show_bug.cgi?id=39654
1276
1277         Test: fast/text/backslash-to-yen-sign-euc.html
1278
1279         * css/CSSStyleSelector.cpp:
1280         (WebCore::CSSStyleSelector::applyProperty): Call FontDescriptor::setIsSpecifiedFont.
1281         * platform/graphics/Font.cpp:
1282         (WebCore::Font::Font):
1283         * platform/graphics/FontDescription.h: Added m_isSpecifiedFont and its accessors.
1284         (WebCore::FontDescription::FontDescription):
1285         (WebCore::FontDescription::isSpecifiedFont):
1286         (WebCore::FontDescription::setIsSpecifiedFont):
1287         (WebCore::FontDescription::operator==):
1288         * platform/text/transcoder/FontTranscoder.cpp:
1289         (WebCore::FontTranscoder::converterType): Don't transcode backslashes if a font family is explicitly specified.
1290         (WebCore::FontTranscoder::convert):
1291         (WebCore::FontTranscoder::needsTranscoding):
1292         * platform/text/transcoder/FontTranscoder.h: Change the arguments from AtomicString fontFamily to Font.
1293         * rendering/RenderText.cpp:
1294         (WebCore::RenderText::updateNeedsTranscoding):
1295         (WebCore::RenderText::setTextInternal):
1296
1297 2010-07-22  Jon Honeycutt  <jhoneycutt@apple.com>
1298
1299         Windows build fix.
1300
1301         Commit some changes missed by the commit-queue in r63944.
1302
1303         * platform/win/BitmapInfo.h:
1304         (WebCore::BitmapInfo::is16bit):
1305         (WebCore::BitmapInfo::is32bit):
1306         (WebCore::BitmapInfo::width):
1307         (WebCore::BitmapInfo::height):
1308         (WebCore::BitmapInfo::size):
1309         (WebCore::BitmapInfo::paddedWidth):
1310         (WebCore::BitmapInfo::numPixels):
1311         (WebCore::BitmapInfo::paddedBytesPerLine):
1312         (WebCore::BitmapInfo::bytesPerLine):
1313
1314 2010-07-22  Patrick Gansterer  <paroga@paroga.com>
1315
1316         Reviewed by Adam Roben.
1317
1318         [WINCE] Add additonal methods to BitmapInfo.
1319         https://bugs.webkit.org/show_bug.cgi?id=42071
1320
1321         Added a parameter to create 16bit BitmapInfo and
1322         some additional accessors for the WinCE port.
1323
1324         Set bmiHeader.biSize only at constructor.
1325
1326         * platform/win/BitmapInfo.cpp: Added property svn:eol-style.
1327         (WebCore::bitmapInfoForSize):
1328         (WebCore::BitmapInfo::create):
1329         (WebCore::BitmapInfo::createBottomUp):
1330         * platform/win/BitmapInfo.h: Added property svn:eol-style.
1331         (WebCore::BitmapInfo::is16bit):
1332         (WebCore::BitmapInfo::is32bit):
1333         (WebCore::BitmapInfo::width):
1334         (WebCore::BitmapInfo::height):
1335         (WebCore::BitmapInfo::size):
1336         (WebCore::BitmapInfo::paddedWidth):
1337         (WebCore::BitmapInfo::numPixels):
1338         (WebCore::BitmapInfo::paddedBytesPerLine):
1339         (WebCore::BitmapInfo::bytesPerLine):
1340
1341 2010-07-22  MORITA Hajime  <morrita@google.com>
1342
1343         Reviewed by Ojan Vafai.
1344
1345         Autoscroll on selection should take scrollbars into account as outside of the content.
1346         https://bugs.webkit.org/show_bug.cgi?id=40403
1347
1348         ScrollView's scroll-bar was considered as a part of content rect,
1349         that prevents autoscroll to happen when the pointer is over the
1350         scroll-bar.  This change excluded scroll-bar region from scrolling
1351         delta calculation.
1352         
1353         Test: WebCore/manual-tests/autoscroll-over-scrollbar.html
1354
1355         * rendering/RenderLayer.cpp:
1356         (WebCore::RenderLayer::scrollRectToVisible):
1357
1358 2010-07-22  Kent Tamura  <tkent@chromium.org>
1359
1360         Reviewed by Darin Adler.
1361
1362         maxLength should not be applied to non-text types
1363         https://bugs.webkit.org/show_bug.cgi?id=42814
1364
1365         According to the HTML5 specification, maxLength attribute should
1366         be applied to text, search, url, tel, email, and password types.
1367
1368         Test: fast/forms/input-maxlength-unsupported.html
1369
1370         * dom/InputElement.cpp:
1371         (WebCore::InputElement::handleBeforeTextInsertedEvent):
1372          Apply maxLength() only if supportsMaxLength() is true.
1373         * dom/InputElement.h:
1374         * html/HTMLInputElement.cpp:
1375         (WebCore::HTMLInputElement::patternMismatch):
1376          Code cleanup by isTextType().
1377         (WebCore::HTMLInputElement::tooLong):
1378          ditto.
1379         (WebCore::HTMLInputElement::isTextType):
1380          Added. This returns true for types with maxLength/pattern/placeholder.
1381         * html/HTMLInputElement.h:
1382         (WebCore::HTMLInputElement::supportsMaxLength):
1383          Added. Just calls isTextType().
1384         (WebCore::HTMLInputElement::supportsPlaceholder):
1385          Code cleanup by isTextType().
1386         * wml/WMLInputElement.h:
1387         (WebCore::WMLInputElement::supportsMaxLength):
1388          Added. Always returns true.
1389
1390 2010-07-22  Ryosuke Niwa  <rniwa@webkit.org>
1391
1392         Reviewed by Kent Tamura.
1393
1394         [cleanup] logic to extract adjacent lists and list children in listifyParagraph and unlistifyParagraph should be isolated
1395         https://bugs.webkit.org/show_bug.cgi?id=42841
1396
1397         Added enclosingListChild that takes the list node containing the node for the second argument
1398         to ensure that it only returns the enclosing list child of the specified list node.
1399
1400         Added adjacentEnclosingList that finds adjacent lists of the specified position.
1401         It takes the current position, the adjacent position that may or may not have a list, and the list type,
1402         and avoids a list that belongs to a different table cell or an outer list that contains the current position.
1403
1404         No new tests needed.
1405
1406         * editing/InsertListCommand.cpp:
1407         (WebCore::enclosingListChild): Added. See above.
1408         (WebCore::InsertListCommand::unlistifyParagraph): Uses enclosingListChild.
1409         (WebCore::adjacentEnclosingList): Added. See above.
1410         (WebCore::InsertListCommand::listifyParagraph): Uses adjacentEnclosingList.
1411
1412 2010-07-22  Patrick Gansterer  <paroga@paroga.com>
1413
1414         Reviewed by Adam Roben.
1415
1416         [WIN] Correct DeleteObject usage in PasteboardWin
1417         https://bugs.webkit.org/show_bug.cgi?id=41956
1418
1419         Call DeleteObject() if OpenClipboard() fails and don't
1420         call DeleteObject() on objects returned from SelectObject().
1421
1422         * platform/win/PasteboardWin.cpp:
1423         (WebCore::Pasteboard::writeImage):
1424
1425 2010-07-22  Lucas De Marchi  <lucas.demarchi@profusion.mobi>
1426
1427         Reviewed by Antonio Gomes.
1428
1429         [EFL] Log function is incorrectly used.
1430         r63804 used LOG function, but it should receive one more parameter.
1431         Used LOG_ERROR instead, since this is an error message.
1432         https://bugs.webkit.org/show_bug.cgi?id=42842
1433
1434         * platform/efl/WidgetEfl.cpp:
1435         (WebCore::Widget::applyFallbackCursor): use LOG_ERROR in place of LOG
1436
1437 2010-07-22  Kent Tamura  <tkent@chromium.org>
1438
1439         Reviewed by Ojan Vafai.
1440
1441         Cancel button of search input field should fire input/change events
1442         https://bugs.webkit.org/show_bug.cgi?id=34393
1443
1444         Test: fast/forms/search-cancel-button-events.html
1445
1446         * rendering/TextControlInnerElements.cpp:
1447         (WebCore::SearchFieldCancelButtonElement::defaultEventHandler):
1448
1449 2010-07-22  Sam Weinig  <sam@webkit.org>
1450
1451         Reviewed by Maciej Stachowiak.
1452
1453         Fix for <rdar://problem/8222626>
1454         Send textDidChangeInTextField delegate callback only in response to typing or other forms of user text input.
1455
1456         * GNUmakefile.am:
1457         * WebCore.gypi:
1458         * WebCore.pro:
1459         * WebCore.vcproj/WebCore.vcproj:
1460         * WebCore.xcodeproj/project.pbxproj:
1461         Update project files.
1462         
1463         * dom/UserTypingGestureIndicator.cpp: Copied from dom/UserGestureIndicator.cpp.
1464         (WebCore::UserTypingGestureIndicator::processingUserTypingGesture):
1465         (WebCore::focusedNode):
1466         (WebCore::UserTypingGestureIndicator::focusedElementAtGestureStart):
1467         (WebCore::UserTypingGestureIndicator::UserTypingGestureIndicator):
1468         (WebCore::UserTypingGestureIndicator::~UserTypingGestureIndicator):
1469         * dom/UserTypingGestureIndicator.h: Copied from dom/UserGestureIndicator.h.
1470         Version of UserGestureIndicator that just tracks keyboard gestures and the focused
1471         node they were targeting.
1472         
1473         * editing/Editor.cpp:
1474         (WebCore::Editor::confirmComposition):
1475         (WebCore::Editor::setComposition):
1476         * page/EventHandler.cpp:
1477         (WebCore::EventHandler::keyEvent):
1478         Put UserTypingGestureIndicator on the stack for user generated keyboard events.
1479
1480 2010-07-22  Kent Tamura  <tkent@chromium.org>
1481
1482         Reviewed by Ojan Vafai.
1483
1484         <input type=number> stepper buttons should dispatch input/change events
1485         https://bugs.webkit.org/show_bug.cgi?id=42440
1486
1487         Test: fast/forms/input-number-events.html
1488
1489         * html/HTMLInputElement.cpp:
1490         (WebCore::HTMLInputElement::stepUpFromRenderer):
1491          Sets a flag to dispatch 'change' event and dispatches 'input' event
1492          if the value is changed.
1493
1494 2010-07-22  Justin Schuh  <jschuh@chromium.org>
1495
1496         Reviewed by Adam Barth.
1497
1498         Make history state methods validate against document URL
1499         https://bugs.webkit.org/show_bug.cgi?id=42858
1500
1501         Test: fast/history/history-replace-illegal.html
1502
1503         * page/History.cpp:
1504         (WebCore::History::urlForState):
1505
1506 2010-07-22  Andy Estes  <aestes@apple.com>
1507
1508         Reviewed by Maciej Stachowiak.
1509
1510         When there are no beforeload listeners on a document,
1511         HTMLLinkElement::process() should be called immediately when the node is
1512         inserted into the document, rather than waiting until after attach is
1513         performed.
1514         <https://bugs.webkit.org/show_bug.cgi?id=42859>
1515         <rdar://problem/8194528>
1516
1517         No new tests.
1518
1519         * html/HTMLLinkElement.cpp:
1520         (WebCore::HTMLLinkElement::insertedIntoDocument): If there are no
1521         beforeload listeners, call process() immediately.  Otherwise, add it as
1522         a post-attach callback.
1523
1524 2010-07-22  James Robinson  <jamesr@chromium.org>
1525
1526         Reviewed by Simon Fraser.
1527
1528         REGRESSION: box shadows on WebGL canvas elements repainted every frame
1529         https://bugs.webkit.org/show_bug.cgi?id=42852
1530
1531         Avoid calling HTMLCanvasElement::willDraw() every time a WebGL canvas'
1532         contents update.  This call is only needed in the non-compositing
1533         rendering path.
1534
1535         Covered by fast/canvas/webgl/canvas-test.html
1536
1537         * html/canvas/WebGLRenderingContext.cpp:
1538         (WebCore::WebGLRenderingContext::markContextChanged):
1539         (WebCore::WebGLRenderingContext::paintRenderingResultsToCanvas):
1540
1541 2010-07-22  Patrick Gansterer  <paroga@paroga.com>
1542
1543         Reviewed by Ojan Vafai.
1544
1545         Remove useless PlatformScrollBar
1546         https://bugs.webkit.org/show_bug.cgi?id=41911
1547
1548         * WebCore.gypi:
1549         * platform/win/PlatformScrollBar.h: Removed.
1550         * platform/win/PlatformScrollBarWin.cpp: Removed.
1551
1552 2010-07-22  Ryosuke Niwa  <rniwa@webkit.org>
1553
1554         Reviewed by Ojan Vafai.
1555
1556         [Windows] Home hey doesn't work in first DIV inside a TABLE
1557         https://bugs.webkit.org/show_bug.cgi?id=42767
1558
1559         The bug was caused by positionAvoidingFirstPositionInTable moving outside of the editing region.
1560         Added a check so that it avoids the first position in the table only if the table is editable.
1561
1562         Test: editing/selection/home-inside-noneditable-table.html
1563
1564         * editing/visible_units.cpp:
1565         (WebCore::positionAvoidingFirstPositionInTable):
1566
1567 2010-07-22  Antonio Gomes  <tonikitoo@webkit.org>
1568
1569         Reviewed by Laszlo Gombos.
1570
1571         Fix build warning for missing parentheses.
1572
1573         WebCore/html/HTMLInputElement.cpp:113: warning: suggest parentheses around '&&' within '||'
1574
1575         * html/HTMLInputElement.cpp:
1576         (WebCore::isNumberCharacter):
1577
1578 2010-07-22  John Gregg  <johnnyg@google.com>
1579
1580         Reviewed by Kent Tamura.
1581
1582         [chromium] fast/forms/input-file-directory-upload.html fails on win after r63454
1583         https://bugs.webkit.org/show_bug.cgi?id=42768
1584
1585         Normalize path separators for the script API by convering \'s to /'s.
1586
1587         Covered by input-file-directory-upload.html
1588
1589         * html/HTMLInputElement.cpp:
1590         (WebCore::HTMLInputElement::setFileListFromRenderer):
1591
1592 2010-07-22  Dmitry Titov  <dimich@chromium.org>
1593
1594         Rubberstamped by Simon Fraser.
1595
1596         * platform/PlatformStrategies.h:
1597         (WebCore::PlatformStrategies::PlatformStrategies): Add initialization of m_localizationStrategy.
1598
1599 2010-07-22  Dumitru Daniliuc  <dumi@chromium.org>
1600
1601         Reviewed by Ojan Vafai.
1602
1603         Make db.changeVersion()'s third parameter optional, as required by the spec.
1604         https://bugs.webkit.org/show_bug.cgi?id=42787
1605
1606         * bindings/js/JSDatabaseCustom.cpp:
1607         (WebCore::JSDatabase::changeVersion):
1608         * bindings/js/JSDatabaseSyncCustom.cpp:
1609         (WebCore::JSDatabaseSync::changeVersion):
1610         * bindings/v8/custom/V8DatabaseCustom.cpp:
1611         (WebCore::V8Database::changeVersionCallback):
1612         * bindings/v8/custom/V8DatabaseSyncCustom.cpp:
1613         (WebCore::V8DatabaseSync::changeVersionCallback):
1614         * storage/SQLTransaction.cpp:
1615         (WebCore::SQLTransaction::deliverTransactionCallback):
1616         * storage/SQLTransactionSync.cpp:
1617         (WebCore::SQLTransactionSync::execute):
1618
1619 2010-07-20  Jessie Berlin  <jberlin@apple.com>
1620
1621         Reviewed by Darin Adler.
1622
1623         Code Generator: Allow negative and string constants.
1624         Also add tests for hexadecimal numbers.
1625
1626         * bindings/scripts/CodeGeneratorJS.pm:
1627         If it is a string, do not try to cast it to an int.
1628
1629         * bindings/scripts/IDLStructure.pm:
1630         Add a regex just for constant values that allows strings, hexadecimal numbers, and integral numbers.
1631
1632         * bindings/scripts/test/CPP/WebDOMTestObj.h:
1633         Update the tests.
1634
1635         * bindings/scripts/test/JS/JSTestObj.cpp:
1636         Ditto.
1637
1638         * bindings/scripts/test/JS/JSTestObj.h:
1639         Ditto.
1640
1641         * bindings/scripts/test/ObjC/DOMTestObj.h:
1642         Ditto.
1643
1644         * bindings/scripts/test/TestObj.idl:
1645         Add tests for negative numbers, strings, and hexadecimal numbers.
1646
1647         * bindings/scripts/test/V8/V8TestObj.cpp:
1648         Update the tests.
1649
1650 2010-07-22  Simon Fraser  <simon.fraser@apple.com>
1651
1652         Reviewed by Darin Adler.
1653
1654         Composited layers don't scroll in WebKit2
1655         https://bugs.webkit.org/show_bug.cgi?id=42771
1656
1657         More prep work: make a single scrollPositionChanged() bottleneck method, and use it to 
1658         send the scroll event, repaint fixed elements, and update compositing layers.
1659
1660         * page/FrameView.cpp:
1661         (WebCore::FrameView::scrollPositionChangedViaPlatformWidget):
1662         (WebCore::FrameView::scrollPositionChanged):
1663         (WebCore::FrameView::repaintFixedElementsAfterScrolling):
1664         (WebCore::FrameView::valueChanged):
1665         * page/FrameView.h:
1666         * platform/ScrollView.cpp:
1667         (WebCore::ScrollView::valueChanged):
1668
1669 2010-07-22  Sam Weinig  <sam@webkit.org>
1670
1671         Reviewed by Anders Carlsson.
1672
1673         Patch for https://bugs.webkit.org/show_bug.cgi?id=42836
1674         Add localized strings stubs for WebKit2.
1675
1676         - Convert WebKit (mac) to use a LocalizationStrategy rather than the WebCoreViewFactory
1677           for localized strings.
1678
1679         Gets us below 200 tests failing in WebKitTestRunner.
1680
1681         * WebCore.exp.in:
1682         * WebCore.xcodeproj/project.pbxproj:
1683         * page/mac/WebCoreViewFactory.h:
1684         * platform/LocalizationStrategy.h: Added.
1685         (WebCore::LocalizationStrategy::~LocalizationStrategy):
1686         * platform/PlatformStrategies.h:
1687         (WebCore::PlatformStrategies::pluginStrategy):
1688         (WebCore::PlatformStrategies::localizationStrategy):
1689         (WebCore::PlatformStrategies::PlatformStrategies):
1690         (WebCore::PlatformStrategies::~PlatformStrategies):
1691         * platform/mac/LocalizedStringsMac.mm:
1692
1693 2010-07-22  Andras Becsi  <abecsi@webkit.org>
1694
1695         Reviewed by Antonio Gomes.
1696
1697         [Qt] Implement WebContext::applicationCacheDirectory() for the Qt port of WebKit2.
1698         https://bugs.webkit.org/show_bug.cgi?id=42830
1699
1700         No new tests needed.
1701
1702         * WebCore.pri: add ApplicationCacheStorage.h to wk2 headers
1703         * WebCore.pro: add WebContextQt.cpp to sources
1704
1705 2010-07-22  Joone Hur  <joone.hur@samsung.com>
1706
1707         Reviewed by Antonio Gomes.
1708
1709         [EFL] add support for logging
1710         https://bugs.webkit.org/show_bug.cgi?id=40807
1711
1712         Enable to turn log channels on for EFL port
1713
1714         * platform/efl/LoggingEfl.cpp:
1715         (WebCore::InitializeLoggingChannelsIfNecessary):
1716
1717 2010-07-22  Balazs Kelemen  <kb@inf.u-szeged.hu>
1718
1719         Ureviewed build fix.
1720
1721         [Qt] Fix WebKit2 build
1722
1723         https://bugs.webkit.org/show_bug.cgi?id=42638
1724
1725         * WebCore.pri:
1726         * WebCore.pro:
1727
1728 2010-07-22  Andrey Kosyakov  <caseq@chromium.org>
1729
1730         Reviewed by Yury Semikhatsky.
1731
1732         Web Inspector: Stop granting universal access to inspector front-end page.
1733         https://bugs.webkit.org/show_bug.cgi?id=42817
1734
1735         * inspector/InspectorFrontendClientLocal.cpp:
1736         (WebCore::InspectorFrontendClientLocal::windowObjectCleared):
1737
1738 2010-07-22  Yury Semikhatsky  <yurys@chromium.org>
1739
1740         Unreviewed. Fix inspector console tests crashes.
1741
1742         * bindings/v8/ScriptValue.cpp:
1743         (WebCore::v8ToInspectorValue):
1744         (WebCore::ScriptValue::toInspectorValue): enter v8 context before calling GetPropertyNames
1745
1746 2010-07-22  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1747
1748         Unreviewed build fix.
1749
1750         [EFL] Build break for SVN r63881
1751         https://bugs.webkit.org/show_bug.cgi?id=42813
1752
1753         * CMakeLists.txt: Add platform/SchemeRegistry.cpp
1754
1755 2010-07-22  Hayato Ito  <hayato@chromium.org>
1756
1757         Reviewed by Darin Adler.
1758
1759         Refactor CSSSelector's destructor and make it inline.
1760
1761         https://bugs.webkit.org/show_bug.cgi?id=42726
1762
1763         Refactor with no behavior change, thus no new tests.
1764
1765         * css/CSSSelector.cpp:
1766         (WebCore::CSSSelectorBag::~CSSSelectorBag): Make sure the bag is empty.
1767         (WebCore::CSSSelectorBag::add): Renamed from append().
1768         (WebCore::CSSSelector::releaseOwnedSelectorsToBag):
1769         (WebCore::CSSSelector::deleteReachableSelectors):
1770         * css/CSSSelector.h:
1771         (WebCore::CSSSelector::~CSSSelector):
1772
1773 2010-07-22  Yury Semikhatsky  <yurys@chromium.org>
1774
1775         Reviewed by Pavel Feldman.
1776
1777         Web Inspector: should be possible to convert console message arguments to InspectorValues
1778         https://bugs.webkit.org/show_bug.cgi?id=42457
1779
1780         Now it is possible to convert simple JavaScript objects to
1781         InspectorValues.
1782
1783         * bindings/js/ScriptValue.cpp:
1784         (WebCore::jsToInspectorValue):
1785         (WebCore::ScriptValue::toInspectorValue):
1786         * bindings/js/ScriptValue.h:
1787         * bindings/v8/ScriptValue.cpp:
1788         (WebCore::v8ToInspectorValue):
1789         (WebCore::ScriptValue::toInspectorValue):
1790         * bindings/v8/ScriptValue.h:
1791         * inspector/ConsoleMessage.cpp: console notifications are pushed to RemoteInspectorFrontend instead of InspectorFrontend.
1792         (WebCore::ConsoleMessage::CallFrame::buildInspectorObject):
1793         (WebCore::ConsoleMessage::addToFrontend):
1794         (WebCore::ConsoleMessage::updateRepeatCountInConsole):
1795         * inspector/ConsoleMessage.h:
1796         * inspector/InjectedScript.cpp:
1797         (WebCore::InjectedScript::wrapForConsole): return InspectorValue instead of SerializedScriptValue
1798         * inspector/InjectedScript.h:
1799         * inspector/Inspector.idl:
1800         * inspector/InspectorBackend.cpp:
1801         (WebCore::InspectorBackend::clearConsoleMessages): send response directly from the backend
1802         (WebCore::InspectorBackend::remoteFrontend):
1803         * inspector/InspectorBackend.h:
1804         * inspector/InspectorController.cpp:
1805         (WebCore::InspectorController::addConsoleMessage):
1806         (WebCore::InspectorController::clearConsoleMessages): don't send notification to the front end,
1807         it will be send either from InspectorBackend.cpp if it was a user request or reset message will
1808         be send to the front end if the messages are cleared due to page navigation.
1809         (WebCore::InspectorController::populateScriptObjects):
1810         * inspector/front-end/ConsoleView.js:
1811         * inspector/front-end/inspector.js:
1812
1813 2010-07-22  Alexander Pavlov  <apavlov@chromium.org>
1814
1815         Reviewed by Pavel Feldman.
1816
1817         Web Inspector: Modifying settings actually modifies defaultValues
1818         https://bugs.webkit.org/show_bug.cgi?id=42816
1819
1820         Avoid using live modifiable objects as property default values.
1821
1822         * inspector/front-end/Settings.js:
1823         (WebInspector.Settings):
1824         (WebInspector.Settings.prototype.reset):
1825         (WebInspector.Settings.prototype._load):
1826         (WebInspector.Settings.prototype.installSetting):
1827         (WebInspector.Settings.prototype._get):
1828
1829 2010-07-22  Ben Murdoch  <benm@google.com>
1830
1831         Reviewed by Simon Fraser.
1832
1833         Touch events do not affect the :active CSS state
1834         https://bugs.webkit.org/show_bug.cgi?id=39493
1835
1836         Clean the code up by using a typedef based on comments
1837         post landing of the original patch.
1838
1839         No change in functionality so no new tests.
1840
1841         * rendering/HitTestRequest.h: Typedef the HitTestRequestType to unsigned.
1842         * WebCore.exp.in: Update export signature.
1843         * WebCore.order: Ditto.
1844         * page/EventHandler.cpp: use the new typedef.
1845         * page/EventHandler.h: ditto.
1846
1847
1848 2010-07-22  Alexander Pavlov  <apavlov@chromium.org>
1849
1850         Reviewed by Pavel Feldman.
1851
1852         Web Inspector: Rename method for CSS rule source range retrieval and fix return object format
1853         https://bugs.webkit.org/show_bug.cgi?id=42379
1854
1855         * inspector/Inspector.idl:
1856         * inspector/InspectorBackend.cpp:
1857         (WebCore::InspectorBackend::getRuleRanges):
1858         * inspector/InspectorBackend.h:
1859         * inspector/InspectorBackend.idl:
1860         * inspector/InspectorCSSStore.cpp:
1861         (WebCore::InspectorCSSStore::getRuleRanges):
1862         * inspector/InspectorCSSStore.h:
1863         * inspector/InspectorDOMAgent.cpp:
1864         (WebCore::InspectorDOMAgent::getRuleRanges):
1865         * inspector/InspectorDOMAgent.h:
1866         * inspector/front-end/DOMAgent.js:
1867
1868 2010-07-21  Pavel Feldman  <pfeldman@chromium.org>
1869
1870         Reviewed by Yury Semikhatsky.
1871
1872         Chromium Dev Tools: Separators are missing in the context menu.
1873
1874         https://bugs.webkit.org/show_bug.cgi?id=42761
1875
1876         * bindings/js/JSInspectorFrontendHostCustom.cpp:
1877         (WebCore::JSInspectorFrontendHost::showContextMenu):
1878         * bindings/v8/custom/V8InspectorFrontendHostCustom.cpp:
1879         (WebCore::V8InspectorFrontendHost::showContextMenuCallback):
1880         * platform/ContextMenuItem.h:
1881         (WebCore::):
1882
1883 2010-07-22  Zoltan Herczeg  <zherczeg@webkit.org>
1884
1885         Reviewed by Nikolas Zimmermann.
1886
1887         SVGFilterElement & SVGFE*Element don't support dynamic invalidation, when attributes change
1888         https://bugs.webkit.org/show_bug.cgi?id=42244
1889
1890         Implementing svgAttributeChanged and childrenChanged for the base class of all filters,
1891         called SVGFilterPrimitiveStandardAttributes.
1892
1893         Tests: svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-height-attr.html
1894                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-result-attr.html
1895                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-width-attr.html
1896                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-x-attr.html
1897                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-dom-y-attr.html
1898                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-height-prop.html
1899                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-result-prop.html
1900                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-width-prop.html
1901                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-x-prop.html
1902                svg/dynamic-updates/SVGFilterPrimitiveStandardAttributes-svgdom-y-prop.html
1903
1904         * svg/SVGFilterPrimitiveStandardAttributes.cpp:
1905         (WebCore::SVGFilterPrimitiveStandardAttributes::svgAttributeChanged):
1906         (WebCore::SVGFilterPrimitiveStandardAttributes::childrenChanged):
1907         * svg/SVGFilterPrimitiveStandardAttributes.h:
1908         (WebCore::SVGFilterPrimitiveStandardAttributes::invalidateFilter):
1909
1910 2010-07-22  Kent Tamura  <tkent@chromium.org>
1911
1912         Reviewed by Darin Adler.
1913
1914         Small refactoring for input value sanitization
1915         https://bugs.webkit.org/show_bug.cgi?id=42807
1916
1917         Rename some functions to clarify their roles.
1918
1919         * dom/InputElement.cpp:
1920         (WebCore::replaceEOLAndLimitLength):
1921           Renamed from sanitizeUserInputValue().
1922         (WebCore::InputElement::sanitizeValueForTextField):
1923           Renamed from sanitizeValue(), and call replaceEOLAndLimitLength()
1924           instead of sanitizeUserInputValue().
1925         (WebCore::InputElement::sanitizeUserInputValue):
1926           Just call replaceEOLAndLimitLength().
1927         (WebCore::InputElement::updateValueIfNeeded):
1928           Use non-static sanitizeValue() for consistency.
1929         * dom/InputElement.h:
1930         * html/HTMLInputElement.cpp:
1931         (WebCore::HTMLInputElement::setInputType):
1932           Revert the change of r63876. It is not needed because of the
1933           updateValueIfNeeded() change.
1934         (WebCore::HTMLInputElement::sanitizeValue):
1935           Apply the sanitizeValue() renaming.
1936
1937 2010-07-22  Tony Gentilcore  <tonyg@chromium.org>
1938
1939         Reviewed by Darin Fisher.
1940
1941         webkitPerformance.timing.responseEnd should not include document parse time
1942         https://bugs.webkit.org/show_bug.cgi?id=42797
1943
1944         No new tests because timing based test would be flaky.
1945
1946         * loader/FrameLoader.cpp:
1947         (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.
1948         * loader/MainResourceLoader.cpp:
1949         (WebCore::MainResourceLoader::didReceiveData): Update the time each time didReceiveData() is called.
1950         (WebCore::MainResourceLoader::didFinishLoading): When finished, and after parsing, set responseEnd appropriately.
1951         * loader/MainResourceLoader.h:
1952         * page/Timing.cpp:
1953         (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.
1954
1955 2010-07-21  Kent Tamura  <tkent@chromium.org>
1956
1957         Reviewed by Darin Adler.
1958
1959         Assertion failure by changing the type of an input element with a
1960         non-number value to 'range'.
1961         https://bugs.webkit.org/show_bug.cgi?id=42643
1962
1963         Test: fast/forms/input-value-sanitization.html
1964
1965         * html/HTMLInputElement.cpp:
1966         (WebCore::HTMLInputElement::setInputType):
1967         Update the value by HTMLInputElement::sanitizeValue() in a case that
1968         storesValueSeparateFromAttribute() state is not changed.
1969
1970 2010-07-21  Adam Barth  <abarth@webkit.org>
1971
1972         Reviewed by Eric Seidel.
1973
1974         Implement spec changes for basefont and bgsound
1975         https://bugs.webkit.org/show_bug.cgi?id=42792
1976
1977         * html/HTMLTreeBuilder.cpp:
1978         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
1979         (WebCore::HTMLTreeBuilder::processStartTag):
1980         (WebCore::HTMLTreeBuilder::processStartTagForInHead):
1981
1982 2010-07-21  Adam Barth  <abarth@webkit.org>
1983
1984         Reviewed by Eric Seidel.
1985
1986         Update a list of tags to match changes in the HTML5 spec
1987         https://bugs.webkit.org/show_bug.cgi?id=42791
1988
1989         We need to atomize these strings at some point.
1990
1991         * html/HTMLTreeBuilder.cpp:
1992         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
1993
1994 2010-07-21  Adam Barth  <abarth@webkit.org>
1995
1996         Reviewed by Eric Seidel.
1997
1998         HTMLTreeBuilder needs to update to match new spec behavior for <button>
1999         https://bugs.webkit.org/show_bug.cgi?id=42233
2000
2001         The spec has some typos in this area, but this is my best guess for
2002         what Ian means.  I've filed bugs against the spec for each typo.
2003
2004         * html/HTMLElementStack.cpp:
2005         (WebCore::HTMLNames::isScopeMarker):
2006         (WebCore::HTMLNames::isButtonScopeMarker):
2007         (WebCore::HTMLElementStack::inButtonScope):
2008         * html/HTMLElementStack.h:
2009         * html/HTMLTreeBuilder.cpp:
2010         (WebCore::HTMLTreeBuilder::processFakePEndTagIfPInButtonScope):
2011         (WebCore::HTMLTreeBuilder::processCloseWhenNestedTag):
2012         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
2013         (WebCore::HTMLTreeBuilder::processEndTagForInBody):
2014         * html/HTMLTreeBuilder.h:
2015
2016 2010-07-21  Justin Schuh  <jschuh@chromium.org>
2017
2018         Reviewed by Oliver Hunt.
2019
2020         Prevent DeleteButtonController enable state from changing when not editing
2021         https://bugs.webkit.org/show_bug.cgi?id=42659
2022
2023         Test: svg/custom/use-invalid-html.xhtml
2024
2025         * dom/ContainerNode.cpp:
2026         (WebCore::ContainerNode::cloneChildNodes):
2027
2028 2010-07-21  Beth Dakin  <bdakin@apple.com>
2029
2030         Reviewed by Dan Bernstein.
2031
2032         Fix for https://bugs.webkit.org/show_bug.cgi?id=42605 New border-
2033         radius path-based drawing code has some issues with corner-joins
2034
2035         GraphicsContext::clipConvexPolygon() now takes a boolean parameter 
2036         indicating whether or not the clip should be antialiased.
2037         * platform/graphics/GraphicsContext.h:
2038         * platform/graphics/cairo/GraphicsContextCairo.cpp:
2039         (WebCore::GraphicsContext::clipConvexPolygon):
2040         * platform/graphics/cg/GraphicsContextCG.cpp:
2041         (WebCore::GraphicsContext::clipConvexPolygon):
2042         * platform/graphics/haiku/GraphicsContextHaiku.cpp:
2043         (WebCore::GraphicsContext::clipConvexPolygon):
2044         * platform/graphics/qt/GraphicsContextQt.cpp:
2045         (WebCore::GraphicsContext::clipConvexPolygon):
2046         * platform/graphics/skia/GraphicsContextSkia.cpp:
2047         (WebCore::GraphicsContext::clipConvexPolygon):
2048         * platform/graphics/wince/GraphicsContextWince.cpp:
2049         (WebCore::GraphicsContext::clipConvexPolygon):
2050         * platform/graphics/wx/GraphicsContextWx.cpp:
2051         (WebCore::GraphicsContext::clipConvexPolygon):
2052
2053         Compute 4 booleans determining if each edge is a "matching" edge of 
2054         not. Send the two appropriate values to clipBorderSidePolygon() so 
2055         we can decide there whether or not to antialias the clip based on 
2056         matchy-ness.
2057         * rendering/RenderBoxModelObject.cpp:
2058         (WebCore::RenderBoxModelObject::paintBorder):
2059
2060         If both edges are matching, don't antialias. If neither edge 
2061         matches, never antialias. If one edge matches and one does not, 
2062         apply two clips to get the appropriate antialiasing in each corner.
2063         (WebCore::RenderBoxModelObject::clipBorderSidePolygon):
2064         * rendering/RenderBoxModelObject.h:
2065
2066 2010-07-21  Brady Eidson  <beidson@apple.com>
2067
2068         Reviewed by Geoffrey Garen.
2069
2070         Break out "scheme registration" functionality from SecurityOrigin to a SchemeRegistry
2071         https://bugs.webkit.org/show_bug.cgi?id=42783
2072
2073         No new tests. (No change in behavior)
2074
2075         * GNUmakefile.am:
2076         * WebCore.exp.in:
2077         * WebCore.gypi:
2078         * WebCore.pro:
2079         * WebCore.vcproj/WebCore.vcproj:
2080         * WebCore.xcodeproj/project.pbxproj:
2081         
2082         * loader/FrameLoader.cpp:
2083         (WebCore::FrameLoader::isMixedContent):
2084         (WebCore::FrameLoader::loadFrameRequest):
2085         
2086         * page/SecurityOrigin.cpp:
2087         (WebCore::SecurityOrigin::SecurityOrigin):
2088         (WebCore::SecurityOrigin::canLoad):
2089         (WebCore::SecurityOrigin::isLocal):
2090         * page/SecurityOrigin.h:
2091         
2092         * platform/SchemeRegistry.cpp: Added.
2093         (WebCore::localSchemes):
2094         (WebCore::secureSchemes):
2095         (WebCore::schemesWithUniqueOrigins):
2096         (WebCore::SchemeRegistry::registerURLSchemeAsLocal):
2097         (WebCore::SchemeRegistry::removeURLSchemeRegisteredAsLocal):
2098         (WebCore::SchemeRegistry::localURLSchemes):
2099         (WebCore::SchemeRegistry::shouldTreatURLAsLocal):
2100         (WebCore::SchemeRegistry::shouldTreatURLSchemeAsLocal):
2101         (WebCore::SchemeRegistry::registerURLSchemeAsNoAccess):
2102         (WebCore::SchemeRegistry::shouldTreatURLSchemeAsNoAccess):
2103         (WebCore::SchemeRegistry::registerURLSchemeAsSecure):
2104         (WebCore::SchemeRegistry::shouldTreatURLSchemeAsSecure):
2105         * platform/SchemeRegistry.h: Added.
2106
2107 2010-07-21  Chris Marrin  <cmarrin@apple.com>
2108
2109         Reviewed by Simon Fraser.
2110
2111         Assertion failure in AnimationBase::updateStateMachine() coming out of paused state
2112         https://bugs.webkit.org/show_bug.cgi?id=37993
2113         
2114         Added logic to properly handle pausing and resuming when in the 
2115         AnimationStateStartWaitStyleAvailable state. This was causing an
2116         assert when going out of the pause state because the paused flag
2117         was not set.
2118         
2119         The fix is a straightforward implementation, going into a new
2120         AnimationStatePausedWaitStyleAvailable state and setting the paused
2121         flag (actually setting the m_pauseTime variable to something other
2122         than -1). Also added handling of the new state, both when the 
2123         "style available" callback comes in while in this state and when
2124         unpausing while in this state. 
2125         
2126         For now a LayoutTest is not possible since there's no way to go in
2127         and out of the pause state. I've opened https://bugs.webkit.org/show_bug.cgi?id=42790
2128         to track this.
2129
2130         * page/animation/AnimationBase.cpp:
2131         (WebCore::AnimationBase::updateStateMachine):
2132         * page/animation/AnimationBase.h:
2133         (WebCore::AnimationBase::):
2134
2135 2010-07-21  Bo Liu  <boliu@chromium.org>
2136
2137         Reviewed by Darin Fisher.
2138
2139         [Chromium] Add chromium WebMediaPlayer to PlatformMedia
2140         https://bugs.webkit.org/show_bug.cgi?id=41295
2141
2142         * platform/graphics/MediaPlayer.h:
2143         (WebCore::PlatformMedia::):
2144
2145 2010-07-21  Simon Fraser  <simon.fraser@apple.com>
2146
2147         Reviewed by Anders Carlsson.
2148
2149         Composited layers don't scroll in WebKit2
2150         https://bugs.webkit.org/show_bug.cgi?id=42771
2151
2152         Prep work: FrameView::scrollPositionChanged() sounds like a generic "did scroll" bottleneck,
2153         but this is deceiving. It's only every called on one platform (Mac) when the NSScrollView gets
2154         scrolled, so rename it to FrameView::scrollPositionChangedViaPlatformWidget().
2155
2156         * WebCore.exp.in:
2157         * page/FrameView.cpp:
2158         (WebCore::FrameView::scrollPositionChangedViaPlatformWidget):
2159         * page/FrameView.h:
2160
2161 2010-07-21  Peter Beverloo  <peter@lvp-media.com>
2162
2163         Reviewed by Eric Carlson.
2164
2165         Removed support for the -khtml CSS vendor prefix and limit the
2166         -apple prefix to two properties (dashboard-region and line-clamp).
2167
2168         Test: fast/css/limited-vendor-prefix-behavior.html
2169
2170         * css/CSSParser.cpp:
2171         (WebCore::cssPropertyID):
2172
2173 2010-07-21  Adam Barth  <abarth@webkit.org>
2174
2175         Reviewed by Eric Seidel.
2176
2177         Fix the last tree HTML5 tree builder crashes
2178         https://bugs.webkit.org/show_bug.cgi?id=42773
2179
2180         This patch changes the internal representation of a bookmark to handle
2181         the case where one of the adjecent entries in the list of active
2182         formatting elements is actually a marker.
2183
2184         After this patch, the bookmarking mechanism isn't as general, but it
2185         works for the cases we need in the adoption agency.
2186
2187         Also, after this patch, there aren't any more known crashers in the
2188         HTML5 tree builder.  :)
2189
2190         * html/HTMLFormattingElementList.cpp:
2191         (WebCore::HTMLFormattingElementList::bookmarkFor):
2192         (WebCore::HTMLFormattingElementList::swapTo):
2193         * html/HTMLFormattingElementList.h:
2194         (WebCore::HTMLFormattingElementList::Bookmark::Bookmark):
2195         (WebCore::HTMLFormattingElementList::Bookmark::moveToAfter):
2196         (WebCore::HTMLFormattingElementList::Bookmark::hasBeenMoved):
2197         (WebCore::HTMLFormattingElementList::Bookmark::mark):
2198         (WebCore::HTMLFormattingElementList::first):
2199         * html/HTMLTreeBuilder.cpp:
2200         (WebCore::HTMLTreeBuilder::callTheAdoptionAgency):
2201
2202 2010-07-21  Tony Gentilcore  <tonyg@chromium.org>
2203
2204         Unreviewed build fix.
2205
2206         Disable overzealous ASSERT
2207         https://bugs.webkit.org/show_bug.cgi?id=42775
2208
2209         No new tests because no changed functionality.
2210
2211         * loader/FrameLoader.cpp:
2212         (WebCore::FrameLoader::stopLoading):
2213
2214 2010-07-21  Yael Aharon  <yael.aharon@nokia.com>
2215
2216         Reviewed by Darin Adler.
2217
2218         Crash in Notification::disconnectFrame() triggered by Frame::lifeSupportTimerFired()
2219         https://bugs.webkit.org/show_bug.cgi?id=42534
2220
2221         Call NotificationsCenter::disconnectFrame() when the frame is disconnected from the page.
2222         Calling it from the destructor of Frame is too late and sometimes causes access violation.
2223         I was not able to reproduce this crash, so did not add new tests.
2224         This patch is based on the error reported in 
2225         http://code.google.com/p/chromium/issues/detail?id=49323.
2226
2227         * page/DOMWindow.cpp:
2228         (WebCore::DOMWindow::pageDestroyed):
2229         * page/DOMWindow.h:
2230         * page/Frame.cpp:
2231         (WebCore::Frame::pageDestroyed):
2232
2233 2010-07-21  Anders Carlsson  <andersca@apple.com>
2234
2235         Reviewed by Sam Weinig.
2236
2237         Don't assert when clicking on a plug-in in WebKit2
2238         https://bugs.webkit.org/show_bug.cgi?id=42762
2239
2240         Remove a now invalid assert and return false instead.
2241
2242         * page/mac/EventHandlerMac.mm:
2243         (WebCore::EventHandler::passMouseDownEventToWidget):
2244
2245 2010-07-21  Anton Muhin  <antonm@chromium.org>
2246
2247         Reviewed by Adam Barth.
2248
2249         [v8] Revert r60670 as it introduced a regression: in some cases named children couldn't be retrieved.
2250         https://bugs.webkit.org/show_bug.cgi?id=42766
2251
2252         See http://code.google.com/p/chromium/issues/detail?id=48804 for more details.
2253
2254         * bindings/scripts/CodeGeneratorV8.pm:
2255         * bindings/v8/ScriptController.cpp:
2256         (WebCore::ScriptController::namedItemAdded):
2257         (WebCore::ScriptController::namedItemRemoved):
2258         * bindings/v8/V8DOMWindowShell.cpp:
2259         (WebCore::V8DOMWindowShell::updateDocumentWrapperCache):
2260         * bindings/v8/V8DOMWindowShell.h:
2261         * bindings/v8/V8DOMWrapper.cpp:
2262         (WebCore::V8DOMWrapper::instantiateV8Object):
2263         * bindings/v8/custom/V8HTMLDocumentCustom.cpp:
2264         (WebCore::V8HTMLDocument::namedPropertyDeleter):
2265         (WebCore::V8HTMLDocument::namedPropertyGetter):
2266         (WebCore::V8HTMLDocument::indexedPropertyGetter):
2267         (WebCore::V8HTMLDocument::allAccessorSetter):
2268         (WebCore::toV8):
2269
2270 2010-07-21  Eric Carlson  <eric.carlson@apple.com>
2271
2272         Reviewed by Simon Fraser.
2273
2274         Update plug-in proxy backed <video> and <audio> elements.
2275         https://bugs.webkit.org/show_bug.cgi?id=42770
2276         <rdar://problem/7963467>
2277
2278         * html/HTMLMediaElement.cpp:
2279         (WebCore::HTMLMediaElement::attributeChanged): Only call m_player->setControls when
2280         the 'controls' attribute changes.
2281         (WebCore::HTMLMediaElement::createRenderer): 
2282         (WebCore::HTMLMediaElement::attach): If we have a proxy widget, call the WebFrameLoaderClient's 
2283         'hide' and 'show' methods for the media proxy plug-in to make sure we remove and re-add the 
2284         plug-in's view to the global list of plugin views.
2285         (WebCore::HTMLMediaElement::createMediaPlayerProxy): Bail if m_proxyWidget is non-NULL
2286
2287         * loader/EmptyClients.h:
2288         (WebCore::EmptyFrameLoaderClient::hideMediaPlayerProxyPlugin): Empty stub method.
2289         (WebCore::EmptyFrameLoaderClient::showMediaPlayerProxyPlugin): Ditto.
2290
2291         * loader/FrameLoader.cpp:
2292         * loader/FrameLoaderClient.h:
2293         * loader/SubframeLoader.cpp:
2294         (WebCore::FrameLoader::loadMediaPlayerProxyPlugin): Always set m_containsPlugIns in loadMediaPlayerProxyPlugin(),
2295          even when we don't have a renderer.
2296         (WebCore::FrameLoader::hideMediaPlayerProxyPlugin): New, call hideMediaPlayerProxyPlugin.
2297         (WebCore::FrameLoader::showMediaPlayerProxyPlugin): New, call showMediaPlayerProxyPlugin.
2298
2299         * platform/graphics/MediaPlayer.cpp:
2300         (WebCore::NullMediaPlayerPrivate::setControls):
2301         (WebCore::MediaPlayer::setControls): New, pass through to media engine.
2302         (WebCore::MediaPlayer::enterFullscreen): Ditto.
2303         (WebCore::MediaPlayer::exitFullscreen): Ditto.
2304         * platform/graphics/MediaPlayer.h:
2305         * platform/graphics/MediaPlayerPrivate.h:
2306         (WebCore::MediaPlayerPrivateInterface::setControls):
2307         (WebCore::MediaPlayerPrivateInterface::enterFullscreen):
2308         (WebCore::MediaPlayerPrivateInterface::exitFullscreen):
2309
2310 2010-07-21  Dimitri Glazkov  <dglazkov@chromium.org>
2311
2312         Reviewed by Darin Adler.
2313
2314         Make more members of CSSStyleSelector private.
2315         https://bugs.webkit.org/show_bug.cgi?id=42757
2316
2317         No change in behavior, so no new tests.
2318
2319         * css/CSSStyleSelector.h: Made a publicly-unused methods private.
2320
2321 2010-07-21  Pavel Feldman  <pfeldman@chromium.org>
2322
2323         Reviewed by Timothy Hatcher.
2324
2325         Web Inspector: consider jQuery object to be of array nature.
2326
2327         https://bugs.webkit.org/show_bug.cgi?id=42758
2328
2329         * inspector/front-end/InjectedScript.js:
2330         (injectedScriptConstructor):
2331
2332 2010-07-21  Alexey Proskuryakov  <ap@apple.com>
2333
2334         Unreviewed Windows build fix.
2335
2336         https://bugs.webkit.org/show_bug.cgi?id=42717
2337         <rdar://problem/7062824> A wrong password entered for site or proxy auth remains in WebCore
2338         credential storage, and is sent with subsequent requests
2339
2340         * platform/network/cf/ResourceHandleCFNet.cpp:
2341         (WebCore::ResourceHandle::didReceiveAuthenticationChallenge): Don't use that direct a
2342         copy/paste, oops!
2343
2344 2010-07-21  Simon Fraser  <simon.fraser@apple.com>
2345
2346         Reviewed by Darin Adler.
2347
2348         When GraphicsLayer::contentsOrientation() is CompositingCoordinatesBottomUp, need to flip repaint rects
2349         https://bugs.webkit.org/show_bug.cgi?id=42662
2350         
2351         If the contentsOrientation() on a GraphicsLayer is CompositingCoordinatesBottomUp, then we
2352         need to flip the rects passed to setNeedsDisplayInRect:.
2353         
2354         Avoid writing this code twice by making a bare function to share code
2355         between WebLayer and WebTiledLayer. Convert the existing +drawContents:ofLayer:intoContext:
2356         class method into a bare function, because the Obj-C calling overhead doesn't buy us anything.
2357         
2358         Take out an assertion in GraphicsLayerCA::updateContentsTransform() that is not
2359         correct.
2360         
2361         No new tests because no layers use bottom-up contents at present.
2362
2363         * platform/graphics/mac/GraphicsLayerCA.mm:
2364         (WebCore::GraphicsLayerCA::updateContentsTransform):
2365         * platform/graphics/mac/WebLayer.h:
2366         * platform/graphics/mac/WebLayer.mm:
2367         (drawLayerContents):
2368         (setLayerNeedsDisplayInRect):
2369         (-[WebLayer setNeedsDisplayInRect:]):
2370         (-[WebLayer drawInContext:]):
2371         * platform/graphics/mac/WebTiledLayer.mm:
2372         (-[WebTiledLayer setNeedsDisplayInRect:]):
2373         (-[WebTiledLayer drawInContext:]):
2374
2375 2010-07-20  Alexey Proskuryakov  <ap@apple.com>
2376
2377         Reviewed by Darin Adler.
2378
2379         https://bugs.webkit.org/show_bug.cgi?id=42717
2380         <rdar://problem/7062824> A wrong password entered for site or proxy auth remains in WebCore
2381         credential storage, and is sent with subsequent requests
2382
2383         Tests: http/tests/security/401-logout/401-logout.php
2384                http/tests/xmlhttprequest/remember-bad-password.html
2385
2386         * platform/network/CredentialStorage.cpp: (WebCore::CredentialStorage::remove):
2387         * platform/network/CredentialStorage.h:
2388         Added a way to remove stored credentials for a given protection space.
2389
2390         * platform/network/cf/ResourceHandleCFNet.cpp: (WebCore::ResourceHandle::didReceiveAuthenticationChallenge):
2391         * platform/network/mac/ResourceHandleMac.mm: (WebCore::ResourceHandle::didReceiveAuthenticationChallenge):
2392         Remove stored credentials if they didn't work the first time.
2393
2394 2010-07-20  Steve Falkenburg  <sfalken@apple.com>
2395
2396         Reviewed by Adam Roben.
2397
2398         WebKit on Windows should build optionally with an unversioned ICU DLL
2399         https://bugs.webkit.org/show_bug.cgi?id=42722
2400         <rdar://problem/8211767> WebKit needs to link against unversioned ICU
2401
2402         To get the proper value for U_DISABLE_RENAMING into all source files, we force
2403         the include of ICUVersion.h (our generated header) via the compiler options.
2404         
2405         * WebCore.vcproj/WebCore.vcproj:
2406         Add forced include of ICUVersion.h.
2407         * WebCore.vcproj/WebCoreCommon.vsprops: Add forced include of ICUVersion.h.
2408
2409 2010-07-21  Ilya Tikhonovsky  <loislo@chromium.org>
2410
2411         Reviewed by Pavel Feldman.
2412
2413         WebInspector: Serialization to JSON in InspectorValue works incorrect
2414         if comma is assigned as decimal separator. Windows with Russian locale.
2415         https://bugs.webkit.org/show_bug.cgi?id=42755
2416
2417         * inspector/InspectorValues.cpp:
2418         (WebCore::InspectorBasicValue::writeJSON):
2419
2420 2010-07-20  Simon Fraser  <simon.fraser@apple.com>
2421
2422         Reviewed by Dan Bernstein.
2423
2424         CSS rotation transform can cause elements with certain styles to vanish during rotation.
2425         https://bugs.webkit.org/show_bug.cgi?id=42579
2426         
2427         Remove some untested, broken code that, as far as I can tell, could never get hit.
2428         Use the correct rect for the PaintInfo so that zero-sized elements with overflow:hidden
2429         but a border paint correctly.
2430
2431         Test: compositing/overflow/zero-size-overflow.html
2432
2433         * rendering/RenderLayerBacking.cpp:
2434         (WebCore::RenderLayerBacking::paintIntoLayer):
2435
2436 2010-07-21  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
2437
2438         Reviewed by Laszlo Gombos.
2439
2440         [QT] Leak and few more fixes to qt port of geolocation
2441         https://bugs.webkit.org/show_bug.cgi?id=42753
2442
2443         Fixes memory leak and default values set to attributes. 
2444         
2445         Memory leak fix no new test case added. 
2446
2447         * platform/qt/GeolocationServiceQt.cpp: Added.
2448         (WebCore::GeolocationServiceQt::~GeolocationServiceQt):
2449         (WebCore::GeolocationServiceQt::positionUpdated):
2450
2451 2010-07-21  Kristian Amlie <kristian.amlie@nokia.com>
2452
2453         Reviewed by Kenneth Rohde Christiansen.
2454
2455         Added automatic sqlite extraction for Symbian to QtWebKit.
2456
2457         Also added sqlite detection in case sqlite is not present in the SDK.
2458         This is possible if WebKit is compiled standalone.
2459
2460         The inclusion part is a consequence of commit c578c6c1d6d in the Qt
2461         repository. It will not work on Qt versions < 4.7.1, but that is ok,
2462         since the only build system it will affect is marked as experimental
2463         in the whole 4.7 series.
2464
2465         * WebCore.pro:
2466
2467 2010-07-21  Adam Barth  <abarth@webkit.org>
2468
2469         Reviewed by Eric Seidel.
2470
2471         Associate elements with the active form
2472         https://bugs.webkit.org/show_bug.cgi?id=42728
2473
2474         This patch fixes fast/forms/formmove3.html.  The test still doesn't
2475         pass due to some render tree differences, but it works as intended now.
2476
2477         To fix this test, I needed to deviate from the spec slight.  Minefield
2478         seems to have the same deviation:
2479           - http://www.w3.org/Bugs/Public/show_bug.cgi?id=10216
2480
2481         * html/HTMLConstructionSite.cpp:
2482         (WebCore::HTMLConstructionSite::takeForm):
2483         (WebCore::HTMLConstructionSite::setForm):
2484         (WebCore::HTMLConstructionSite::createHTMLElement):
2485         * html/HTMLConstructionSite.h:
2486         (WebCore::HTMLConstructionSite::form):
2487         * html/HTMLTreeBuilder.cpp:
2488         (WebCore::HTMLTreeBuilder::processStartTagForInBody):
2489
2490 2010-07-21  Adam Barth  <abarth@webkit.org>
2491
2492         Reviewed by Eric Seidel.
2493
2494         The adoption agency doesn't properly attach()
2495         https://bugs.webkit.org/show_bug.cgi?id=42727
2496
2497         The adoption agency is transliterated rather directly from the spec,
2498         but it misses some of the WebKit-specific machinations, such as
2499         attaching to the render tree.
2500
2501         The algorithm, as written, is a minor layer violation.  I've added to
2502         the problem by calling attach() from HTMLTreeBuilder (even though
2503         that's the job of the HTMLConstructionSite).  We'll need to clean up
2504         the layering a bit at some point.
2505
2506         This patch half fixes fast/forms/formmove3.html.  Hopefully I'll finish
2507         fixing it in the next patch.
2508
2509         * html/HTMLConstructionSite.cpp:
2510         (WebCore::HTMLConstructionSite::createHTMLElementFromSavedElement):
2511         * html/HTMLTreeBuilder.cpp:
2512         (WebCore::HTMLTreeBuilder::callTheAdoptionAgency):
2513
2514 2010-07-21  Adam Barth  <abarth@webkit.org>
2515
2516         Reviewed by Eric Seidel.
2517
2518         <input> elements with no type attribute should be foster parented
2519         https://bugs.webkit.org/show_bug.cgi?id=42725
2520
2521         Fix spec transcription error.
2522
2523         * html/HTMLTreeBuilder.cpp:
2524         (WebCore::HTMLTreeBuilder::processStartTagForInTable):
2525
2526 2010-07-21  Adam Barth  <abarth@webkit.org>
2527
2528         Reviewed by Eric Seidel.
2529
2530         Fix fast/css/last-child-style-sharing.html
2531         https://bugs.webkit.org/show_bug.cgi?id=42731
2532
2533         Prior to this patch, we weren't calling finishParsingChildren on the
2534         body element.  We need a more systematic way of catching these bugs.
2535
2536         * html/HTMLElementStack.cpp:
2537         (WebCore::HTMLElementStack::popAll):
2538         * html/HTMLElementStack.h:
2539         * html/HTMLTreeBuilder.cpp:
2540         (WebCore::HTMLTreeBuilder::processEndOfFile):
2541
2542 2010-07-21  Hans Wennborg  <hans@chromium.org>
2543
2544         Reviewed by Steve Block.
2545
2546         Runtime feature switch for device orientation
2547         https://bugs.webkit.org/show_bug.cgi?id=42265
2548
2549         Add a runtime feature switch that decides whether device orientation
2550         events are available or not. Defaults to true.
2551
2552         * WebCore.vcproj/WebCoreCommon.vsprops: Add bindings/generic to include path.
2553         * bindings/generic/RuntimeEnabledFeatures.cpp:
2554         * bindings/generic/RuntimeEnabledFeatures.h:
2555         (WebCore::RuntimeEnabledFeatures::setDeviceOrientationEnabled):
2556         (WebCore::RuntimeEnabledFeatures::deviceOrientationEnabled):
2557         (WebCore::RuntimeEnabledFeatures::deviceOrientationEventEnabled):
2558         (WebCore::RuntimeEnabledFeatures::ondeviceorientationEnabled):
2559         * page/DOMWindow.cpp:
2560         (WebCore::DOMWindow::addEventListener):
2561         (WebCore::DOMWindow::removeEventListener):
2562         (WebCore::DOMWindow::removeAllEventListeners):
2563         * page/DOMWindow.idl:
2564         * page/Page.cpp:
2565         (WebCore::Page::Page):
2566
2567 2010-07-21  Zoltan Herczeg  <zherczeg@webkit.org>
2568
2569         Reviewed Nikolas Zimmermann.
2570
2571         SVGFilterElement & SVGFE*Element don't support dynamic invalidation, when attributes change
2572         https://bugs.webkit.org/show_bug.cgi?id=42244
2573
2574         The implementation was done by Nikolas Zimmermann before, but
2575         there was no tests for it. The patch also implements the simple
2576         setFilterRes method.
2577
2578         Tests: svg/dynamic-updates/SVGFilterElement-dom-filterRes-attr.html
2579                svg/dynamic-updates/SVGFilterElement-dom-filterUnits-attr.html
2580                svg/dynamic-updates/SVGFilterElement-dom-height-attr.html
2581                svg/dynamic-updates/SVGFilterElement-dom-primitiveUnits-attr.html
2582                svg/dynamic-updates/SVGFilterElement-dom-width-attr.html
2583                svg/dynamic-updates/SVGFilterElement-dom-x-attr.html
2584                svg/dynamic-updates/SVGFilterElement-dom-y-attr.html
2585                svg/dynamic-updates/SVGFilterElement-svgdom-filterRes-call.html
2586                svg/dynamic-updates/SVGFilterElement-svgdom-filterResX-prop.html
2587                svg/dynamic-updates/SVGFilterElement-svgdom-filterResY-prop.html
2588                svg/dynamic-updates/SVGFilterElement-svgdom-filterUnits-prop.html
2589                svg/dynamic-updates/SVGFilterElement-svgdom-height-prop.html
2590                svg/dynamic-updates/SVGFilterElement-svgdom-primitiveUnits-prop.html
2591                svg/dynamic-updates/SVGFilterElement-svgdom-width-prop.html
2592                svg/dynamic-updates/SVGFilterElement-svgdom-x-prop.html
2593                svg/dynamic-updates/SVGFilterElement-svgdom-y-prop.html
2594
2595         * svg/SVGFilterElement.cpp:
2596         (WebCore::SVGFilterElement::setFilterRes):
2597         (WebCore::SVGFilterElement::svgAttributeChanged):
2598         (WebCore::SVGFilterElement::childrenChanged):
2599         * svg/SVGFilterElement.h:
2600
2601 2010-07-21  Ben Murdoch  <benm@google.com>
2602
2603         Unreviewed, build fix.
2604
2605         Forgot to update the role of HitTestRequest.h in
2606         the xcode project after landing patch from
2607         https://bugs.webkit.org/show_bug.cgi?id=39493
2608
2609         * WebCore.xcodeproj/project.pbxproj: Make HitTestRequest.h
2610             private so it can be included in WebKit by EventHandler.h
2611             which is already private..
2612
2613 2010-07-20  Ben Murdoch  <benm@google.com>
2614
2615         Reviewed by Steve Block.
2616
2617         Touch events do not affect the :active CSS state
2618         https://bugs.webkit.org/show_bug.cgi?id=39493
2619
2620         Test: fast/events/touch/touch-active-state.html
2621
2622         * WebCore.exp.in: Update exports for new signature of
2623             hitTestResultAtPoint.
2624         * WebCore.order: ditto.
2625         * page/EventHandler.cpp:
2626         (WebCore::EventHandler::EventHandler):
2627         (WebCore::EventHandler::hitTestResultAtPoint): Pass the type
2628             of the hit test to perform as a parameter with a default
2629             value rather than harcoding it in the function body.
2630         (WebCore::EventHandler::handleMouseMoveEvent): Do not modiify
2631             the active element during a mouse move if the user is
2632             touching the screen.
2633         (WebCore::EventHandler::handleTouchEvent): Set the correct
2634             type of hit test to perform depending on the type of the
2635             touch event we are handling.
2636         * page/EventHandler.h: Update the signature of hitTestResultAtPoint.
2637
2638 2010-07-20  Yury Semikhatsky  <yurys@chromium.org>
2639
2640         Reviewed by Pavel Feldman.
2641
2642         console.assert should include stack trace with line numbers.
2643         https://bugs.webkit.org/show_bug.cgi?id=22556
2644
2645         Test: inspector/console-assert.html
2646
2647         * bindings/v8/ScriptCallStack.cpp:
2648         (WebCore::getTopFrameLocation):
2649         (WebCore::toScriptCallFramesVector):
2650         (WebCore::ScriptCallStack::create):
2651         (WebCore::ScriptCallStack::ScriptCallStack):
2652         * bindings/v8/ScriptCallStack.h:
2653         * bindings/v8/ScriptController.cpp:
2654         (WebCore::ScriptController::setCaptureCallStackForUncaughtExceptions):
2655         * bindings/v8/custom/V8ConsoleCustom.cpp:
2656         (WebCore::V8Console::traceCallback):
2657         (WebCore::V8Console::assertCallback):
2658         * inspector/ConsoleMessage.cpp:
2659         (WebCore::ConsoleMessage::ConsoleMessage):
2660         * inspector/InspectorController.cpp:
2661         (WebCore::InspectorController::addMessageToConsole):
2662         * inspector/front-end/ConsoleView.js:
2663         (WebInspector.ConsoleMessage.prototype._formatMessage):
2664         (WebInspector.ConsoleMessage.prototype.toMessageElement):
2665         * page/Console.cpp:
2666         (WebCore::Console::lastWMLErrorMessage):
2667         * page/Console.idl:
2668
2669 2010-07-20  Rafael Antognolli  <antognolli@profusion.mobi>
2670
2671         Reviewed by Kent Tamura.
2672
2673         [EFL] Use log functions instead of fprintf
2674         https://bugs.webkit.org/show_bug.cgi?id=42576
2675
2676         Use LOG and LOG_ERROR instead of fprintf.
2677
2678         No new tests, no new functionality.
2679
2680         * platform/efl/WidgetEfl.cpp:
2681         (WebCore::Widget::applyFallbackCursor):
2682
2683 2010-07-20  Ilya Tikhonovsky  <loislo@chromium.org>
2684
2685         Reviewed by Yury Semikhatsky.
2686
2687         WebInspector: the bindings generation helper script can be
2688         simplified a bit. Chromium.
2689         https://bugs.webkit.org/show_bug.cgi?id=42523
2690
2691         * WebCore.gyp/WebCore.gyp:
2692         * WebCore.gyp/scripts/rule_binding.py:
2693         * bindings/scripts/CodeGenerator.pm:
2694         * bindings/scripts/CodeGeneratorCPP.pm:
2695         * bindings/scripts/CodeGeneratorJS.pm:
2696         * bindings/scripts/CodeGeneratorObjC.pm:
2697         * bindings/scripts/CodeGeneratorV8.pm:
2698         * bindings/scripts/generate-bindings.pl:
2699         * inspector/CodeGeneratorInspector.pm:
2700
2701 2010-07-20  Steve Falkenburg  <sfalken@apple.com>
2702
2703         Re-save vsprops file after no-op edit in Visual Studio
2704         to fix manual edit issues.
2705
2706         * WebCore.vcproj/WebCoreCommon.vsprops:
2707
2708 2010-07-20  Steve Falkenburg  <sfalken@apple.com>
2709
2710         Re-save vcproj file after no-op edit in Visual Studio
2711         to fix manual edit issues.
2712
2713         * WebCore.vcproj/WebCore.vcproj:
2714
2715 2010-07-20  Adam Barth  <abarth@webkit.org>
2716
2717         Reviewed by Eric Seidel.
2718
2719         Include attributes when reconstructing elements in HTML5 tree builder
2720         https://bugs.webkit.org/show_bug.cgi?id=42594
2721
2722         Remove FIXME now that we have a test.
2723
2724         * html/HTMLConstructionSite.cpp:
2725         (WebCore::HTMLConstructionSite::createHTMLElementFromElementRecord):
2726
2727 2010-07-20  Sam Weinig  <sam@webkit.org>
2728
2729         Reviewed by Brady Eidson.
2730
2731         Patch for https://bugs.webkit.org/show_bug.cgi?id=42719
2732         Make Acid2 pass in WebKit2
2733
2734         * WebCore.exp.in: Update exported functions.
2735
2736 2010-07-20  Gavin Peters  <gavinp@chromium.org>
2737
2738         Reviewed by Adam Barth.
2739
2740         prefetch categorization is exactly wrong
2741         https://bugs.webkit.org/show_bug.cgi?id=42651
2742
2743         While I was in the area, I saw that CachedResource::schedule() was dead, so I have removed it.
2744
2745         Test: fast/dom/HTMLLinkElement/onload-completion-test.html
2746
2747         * loader/CachedCSSStyleSheet.h:
2748         * loader/CachedFont.h:
2749         * loader/CachedImage.h:
2750         * loader/CachedResource.cpp:
2751         * loader/CachedResource.h:
2752         (WebCore::CachedResource::isPrefetch):
2753         * loader/CachedScript.h:
2754         * loader/CachedXBLDocument.h:
2755         * loader/CachedXSLStyleSheet.h:
2756
2757 2010-07-20  James Hawkins  <jhawkins@chromium.org>
2758
2759         Reviewed by Darin Fisher.
2760
2761         Expose the form submission trigger on the HTMLFormElement object. This
2762         is used to verify that the user submitted the form instead of JS when
2763         saving form data in Chrome AutoFill.
2764         https://bugs.webkit.org/show_bug.cgi?id=42479
2765
2766         No new tests as this is only used by the Chromium WebKit API.
2767
2768         * html/HTMLFormElement.cpp:
2769         (WebCore::HTMLFormElement::HTMLFormElement):
2770         (WebCore::HTMLFormElement::submit):
2771         (WebCore::HTMLFormElement::reset):
2772         (WebCore::HTMLFormElement::submissionTrigger):
2773         * html/HTMLFormElement.h:
2774
2775 2010-07-20  Adam Barth  <abarth@webkit.org>
2776
2777         Unreviewed.
2778
2779         Fix fast/dom/title-content-write-set.html for HTML5 tree builder
2780         https://bugs.webkit.org/show_bug.cgi?id=42668
2781
2782         Address some late-breaking review comments.
2783
2784         * dom/CharacterData.cpp:
2785         (WebCore::CharacterData::appendDataCommon):
2786         (WebCore::CharacterData::parserAppendData):
2787         (WebCore::CharacterData::appendData):
2788         (WebCore::CharacterData::insertData):
2789         (WebCore::CharacterData::replaceData):
2790         * dom/CharacterData.h:
2791
2792 2010-07-20  Daniel Erat  <derat@chromium.org>
2793
2794         Reviewed by Ojan Vafai.
2795
2796         Subpixel rendering always disabled for Chromium Linux
2797         https://bugs.webkit.org/show_bug.cgi?id=42220
2798
2799         Explicitly initialize global Chromium Linux isSkiaSubpixelGlyphs
2800         flag to false for clarity.
2801
2802         * platform/graphics/chromium/FontPlatformDataLinux.cpp:
2803
2804 2010-07-20  Adam Barth  <abarth@webkit.org>
2805
2806         Reviewed by Darin Adler.
2807
2808         Fix fast/dom/title-content-write-set.html for HTML5 tree builder
2809         https://bugs.webkit.org/show_bug.cgi?id=42668
2810
2811         When I split parserAppendData from appendData, it was tempting to cut
2812         the function at dispatchModifiedEvent, but that's not quite right.  We
2813         still need to notify the parent that it's children have changed.
2814
2815         * dom/CharacterData.cpp:
2816         (WebCore::CharacterData::appendDataCommon):
2817         (WebCore::CharacterData::parserAppendData):
2818         (WebCore::CharacterData::appendData):
2819         * dom/CharacterData.h:
2820
2821 2010-07-20  Chris Fleizach  <cfleizach@apple.com>
2822
2823         Reviewed by Beth Dakin.
2824
2825         CrashTracer: [USER] 300 crashes in Safari at com.apple.WebCore: WebCore::AccessibilityTable::isTableExposableThroughAccessibility + 573
2826         https://bugs.webkit.org/show_bug.cgi?id=42652
2827
2828         When a table cell accesses its parent table, we should not use getOrCreate, because creating an AXTable inspects its render tree state
2829         which may be out of date, leading to a crash.
2830         By using only get(), it implies that the AXTable must be created before AXTableCells. This should
2831         always be the case when AT clients access a table.
2832
2833         Test: accessibility/updating-attribute-in-table-causes-crash.html
2834
2835         * accessibility/AccessibilityTableCell.cpp:
2836         (WebCore::AccessibilityTableCell::parentTable):
2837
2838 2010-07-20  Abhishek Arya  <inferno@chromium.org>
2839
2840         Reviewed by David Hyatt.
2841
2842         Check the node is a text node before doing the static cast
2843         for editing commands.
2844         https://bugs.webkit.org/show_bug.cgi?id=42655
2845
2846         Test: editing/execCommand/editing-nontext-node-crash.xhtml
2847
2848         * editing/DeleteSelectionCommand.cpp:
2849         (WebCore::DeleteSelectionCommand::fixupWhitespace):
2850         * editing/InsertLineBreakCommand.cpp:
2851         (WebCore::InsertLineBreakCommand::doApply):
2852         * editing/InsertParagraphSeparatorCommand.cpp:
2853         (WebCore::InsertParagraphSeparatorCommand::doApply):
2854
2855 2010-07-20  Leo Yang  <leo.yang@torchmobile.com.cn>
2856
2857         Reviewed by David Hyatt.
2858
2859         Don't merge Anonymous block whose first child is inline run-in.
2860         Make run-in recalculate its style after its renderer is destroyed.
2861         https://bugs.webkit.org/show_bug.cgi?id=41375.
2862
2863         Test: fast/runin/crash-when-reparent-sibling.html
2864
2865         * rendering/RenderBlock.cpp:
2866         (WebCore::canMergeContiguousAnonymousBlocks):
2867         * rendering/RenderObjectChildList.cpp:
2868         (WebCore::RenderObjectChildList::destroyLeftoverChildren):
2869
2870 2010-07-20  Steve Block  <steveblock@google.com>
2871
2872         Unreviewed Qt test fix.
2873
2874         Qt should not use PREEMPT_GEOLOCATION_PERMISSION until Bug 42027 is fixed.
2875         https://bugs.webkit.org/show_bug.cgi?id=42068
2876
2877         No new tests, test fix only.
2878
2879         * WebCore.pro:
2880
2881 2010-07-20  Sheriff Bot  <webkit.review.bot@gmail.com>
2882
2883         Unreviewed, rolling out r63764.
2884         http://trac.webkit.org/changeset/63764
2885         https://bugs.webkit.org/show_bug.cgi?id=42658
2886
2887         have broken Chromium linux build (Requested by loislo on
2888         #webkit).
2889
2890         * WebCore.gyp/WebCore.gyp:
2891         * WebCore.gyp/scripts/rule_binding.py:
2892         * bindings/scripts/CodeGenerator.pm:
2893         * bindings/scripts/CodeGeneratorCPP.pm:
2894         * bindings/scripts/CodeGeneratorJS.pm:
2895         * bindings/scripts/CodeGeneratorObjC.pm:
2896         * bindings/scripts/CodeGeneratorV8.pm:
2897         * bindings/scripts/generate-bindings.pl:
2898         * inspector/CodeGeneratorInspector.pm:
2899
2900 2010-07-20  Alexey Proskuryakov  <ap@apple.com>
2901
2902         Reviewed by Brady Eidson.
2903
2904         https://bugs.webkit.org/show_bug.cgi?id=41531
2905         Asynchronous cross origin XMLHttpRequest doesn't expose 401 response when withCredentials is false
2906
2907         This doesn't match Firefox, but it matches our sync case, XHR2 spec and common sense.
2908
2909         Test: http/tests/xmlhttprequest/cross-origin-no-authorization.html (updated results).
2910
2911         * loader/DocumentThreadableLoader.cpp:
2912         (WebCore::DocumentThreadableLoader::didReceiveAuthenticationChallenge): Instead of canceling
2913         the request, continue without credentials - if the platform has a necessary method on
2914         ResourceHandle.
2915
2916         * loader/SubresourceLoader.cpp:
2917         (WebCore::SubresourceLoader::didReceiveAuthenticationChallenge): Don't ask resource loader
2918         client for credentials if subresource loader client already took care of those.
2919
2920         * platform/network/ResourceHandle.cpp: (WebCore::ResourceHandle::hasAuthenticationChallenge):
2921         * platform/network/ResourceHandle.h:
2922         Added an accessor to check whether ResourceHandle is currently waiting for credentials.
2923
2924 2010-07-20  Leandro Pereira  <leandro@profusion.mobi>
2925
2926         [EFL] Unreviewed build fix.
2927
2928         Build SVGPathBuilder, SVGPathParser and SVGPathSegListBuilder.
2929         Add websockets include directory.
2930
2931         The EFL port doesn't yet support automated tests.
2932
2933         * CMakeLists.txt:
2934
2935 2010-07-20  Ilya Tikhonovsky  <loislo@chromium.org>
2936
2937         Reviewed by Yury Semikhatsky.
2938
2939         WebInspector: the bindings generation helper script can be
2940         simplified a bit. Chromium.
2941         https://bugs.webkit.org/show_bug.cgi?id=42523
2942
2943         * WebCore.gyp/WebCore.gyp:
2944         * WebCore.gyp/scripts/rule_binding.py:
2945         * bindings/scripts/CodeGenerator.pm:
2946         * bindings/scripts/CodeGeneratorCPP.pm:
2947         * bindings/scripts/CodeGeneratorJS.pm:
2948         * bindings/scripts/CodeGeneratorObjC.pm:
2949         * bindings/scripts/CodeGeneratorV8.pm:
2950         * bindings/scripts/generate-bindings.pl:
2951         * inspector/CodeGeneratorInspector.pm:
2952
2953 2010-07-20  Joseph Pecoraro  <joepeck@webkit.org>
2954
2955         Reviewed by Geoffrey Garen.
2956
2957         WebScriptObject Should Allow Safely Checking For Key Existence
2958         https://bugs.webkit.org/show_bug.cgi?id=42613
2959
2960         Test: platform/mac/fast/objc/webScriptObject-hasWebScriptKey.html
2961
2962         Add private API "hasWebScriptKey" to check for key existence in
2963         a WebScriptObject. Like JavaScript's `in` syntax. This is intended
2964         to be made public eventually.
2965
2966         * bindings/objc/WebScriptObject.mm:
2967         (-[WebScriptObject hasWebScriptKey:]):
2968         * bindings/objc/WebScriptObjectPrivate.h:
2969
2970 2010-07-20  Adam Barth  <abarth@webkit.org>
2971
2972         Reviewed by Eric Seidel.
2973
2974         HTML5 tree builder needs to call dispatchDocumentElementAvailable
2975         https://bugs.webkit.org/show_bug.cgi?id=42654
2976
2977         This patch fixes the follout LayoutTests with --html5-treebuilder:
2978           - userscripts/script-not-run-for-fragments.html
2979           - userscripts/script-run-at-start.html
2980
2981         * html/HTMLConstructionSite.cpp:
2982         (WebCore::HTMLConstructionSite::HTMLConstructionSite):
2983         (WebCore::HTMLConstructionSite::dispatchDocumentElementAvailableIfNeeded):
2984         (WebCore::HTMLConstructionSite::insertHTMLHtmlStartTagBeforeHTML):
2985         (WebCore::HTMLConstructionSite::insertHTMLHtmlElement):
2986         * html/HTMLConstructionSite.h:
2987         * html/HTMLTreeBuilder.cpp:
2988         (WebCore::HTMLTreeBuilder::HTMLTreeBuilder):
2989
2990 2010-07-20  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
2991
2992         Reviewed by Steve Block.
2993
2994         Need to be able to configure Geolocation policy regarding user permissions
2995         https://bugs.webkit.org/show_bug.cgi?id=42068
2996
2997         Introducing new USE() flag PREEMPT_GEOLOCATION_PREMISSION using which
2998         acquires user permission first before starting location service
2999
3000         This change does not introduce any change in behavior for any platform. So there are no new tests
3001         added.
3002
3003         * WebCore.pro:
3004         * page/Geolocation.cpp:
3005         (WebCore::Geolocation::startRequest):
3006         (WebCore::Geolocation::setIsAllowed):
3007         (WebCore::Geolocation::startUpdating):
3008         * page/Geolocation.h:
3009
3010 2010-07-20  Adam Barth  <abarth@webkit.org>
3011
3012         Reviewed by Darin Adler.
3013
3014         HTML5 tree builder should restore form state
3015         https://bugs.webkit.org/show_bug.cgi?id=42644
3016
3017         We need to tell self-closing tags that we're done parsing their
3018         children.  This patch fixes the following LayoutTests when run with
3019         --html5-treebuilder:
3020
3021           fast/forms/button-state-restore.html
3022           fast/forms/state-restore-to-non-autocomplete-form.html
3023           fast/forms/state-restore-to-non-edited-controls.html
3024           fast/history/saves-state-after-fragment-nav.html
3025           http/tests/navigation/restore-form-state-https.html
3026
3027         * html/HTMLConstructionSite.cpp:
3028         (WebCore::HTMLConstructionSite::insertSelfClosingHTMLElement):
3029
3030 2010-07-20  Sheriff Bot  <webkit.review.bot@gmail.com>
3031
3032         Unreviewed, rolling out r63750.
3033         http://trac.webkit.org/changeset/63750
3034         https://bugs.webkit.org/show_bug.cgi?id=42648
3035
3036         This revision breaks the windows builds (Requested by
3037         hwennborg on #webkit).
3038
3039         * bindings/generic/RuntimeEnabledFeatures.cpp:
3040         * bindings/generic/RuntimeEnabledFeatures.h:
3041         * page/DOMWindow.cpp:
3042         (WebCore::DOMWindow::addEventListener):
3043         (WebCore::DOMWindow::removeEventListener):
3044         (WebCore::DOMWindow::removeAllEventListeners):
3045         * page/DOMWindow.idl:
3046         * page/Page.cpp:
3047         (WebCore::Page::Page):
3048
3049 2010-07-20  Ilya Tikhonovsky  <loislo@chromium.org>
3050
3051         Reviewed by Pavel Feldman.
3052
3053         WebInspector: It is possible to show full call stack instead of top frame for Caller
3054         and Call Site properties in Timeline panel.
3055         https://bugs.webkit.org/show_bug.cgi?id=42620
3056
3057         * English.lproj/localizedStrings.js:
3058         * inspector/front-end/TimelinePanel.js:
3059         (WebInspector.TimelinePanel.FormattedRecord):
3060         (WebInspector.TimelinePanel.FormattedRecord.prototype._generatePopupContent):
3061         (WebInspector.TimelinePanel.FormattedRecord.prototype._getRecordDetails):
3062         (WebInspector.TimelinePanel.PopupContentHelper.prototype._appendElementRow):
3063         (WebInspector.TimelinePanel.PopupContentHelper.prototype._appendLinkRow):
3064         (WebInspector.TimelinePanel.PopupContentHelper.prototype._appendStackTrace):
3065         * inspector/front-end/inspector.css:
3066         (.timeline-details):
3067         (.timeline-function-name):
3068         (.timeline-stacktrace-title):
3069
3070 2010-07-20  Leon Clarke  <leonclarke@google.com>
3071
3072         Reviewed by Pavel Feldman.
3073
3074         Make things compile again when the inspector is disabled, following
3075         recent inspector improvements.
3076         https://bugs.webkit.org/show_bug.cgi?id=42632
3077
3078         No new tests. Fixing a build break.
3079
3080
3081         * bindings/v8/ScriptCallStack.cpp:
3082         (WebCore::ScriptCallStack::stackTrace):
3083         * inspector/InspectorController.h:
3084         (WebCore::InspectorController::didInsertDOMNode):
3085         (WebCore::InspectorController::didRemoveDOMNode):
3086         (WebCore::InspectorController::didModifyDOMAttr):
3087         * inspector/InspectorDOMAgent.h:
3088         * loader/appcache/ApplicationCacheGroup.cpp:
3089
3090 2010-07-20  Anton Muhin  <antonm@chromium.org>
3091
3092         Reviewed by Adam Barth.
3093
3094         [v8] Allow handles to be disposed and WebKit objects to be dereferenced even if their map is already destroyed.
3095         https://bugs.webkit.org/show_bug.cgi?id=42634
3096
3097         Currently DOMDataStore could be destroyed even if it has some mappings (it gets destroyed
3098         when its isolated context gets GCed).  However in this case, handles allocated for
3099         such objects would never be disposed as we require presence of mapping from wrapped
3100         WebKit object to handle being collected in the map and now map is gone.  That leads to
3101         zombie objects in both WebKit (wrapped WebKit object doesn't get dereferenced) and V8
3102         (both handle and V8 wrapper object could not be destroyed).
3103
3104         See http://code.google.com/p/chromium/issues/detail?id=47125 for further discussion.
3105
3106         * bindings/v8/DOMData.h:
3107         (WebCore::DOMData::handleWeakObject):
3108         * bindings/v8/DOMDataStore.cpp:
3109         (WebCore::DOMDataStore::weakNodeCallback):
3110         * bindings/v8/V8DOMMap.h:
3111         (WebCore::WeakReferenceMap::~WeakReferenceMap):
3112
3113 2010-07-20  Hans Wennborg  <hans@chromium.org>
3114
3115         Reviewed by Steve Block.
3116
3117         Runtime feature switch for device orientation
3118         https://bugs.webkit.org/show_bug.cgi?id=42265
3119
3120         Add a runtime feature switch that decides whether device orientation
3121         events are available or not. Defaults to true.
3122
3123         * bindings/generic/RuntimeEnabledFeatures.cpp:
3124         * bindings/generic/RuntimeEnabledFeatures.h:
3125         (WebCore::RuntimeEnabledFeatures::setDeviceOrientationEnabled):
3126         (WebCore::RuntimeEnabledFeatures::deviceOrientationEnabled):
3127         (WebCore::RuntimeEnabledFeatures::deviceOrientationEventEnabled):
3128         (WebCore::RuntimeEnabledFeatures::ondeviceorientationEnabled):
3129         * page/DOMWindow.cpp:
3130         (WebCore::DOMWindow::addEventListener):
3131         (WebCore::DOMWindow::removeEventListener):
3132         (WebCore::DOMWindow::removeAllEventListeners):
3133         * page/DOMWindow.idl:
3134         * page/Page.cpp:
3135         (WebCore::Page::Page):
3136
3137 2010-07-20  Hayato Ito  <hayato@chromium.org>
3138
3139         Reviewed by Darin Adler.
3140
3141         Fixed a crash when deeply nested CSS selector is used.
3142         https://bugs.webkit.org/show_bug.cgi?id=41129
3143
3144         This patch deletes CSSSelectors iteratively so that it doesn't cause stack overflow.
3145
3146         Test: fast/css/css-selector-deeply-nested.html
3147
3148         * css/CSSSelector.cpp:
3149         (WebCore::CSSSelectorBag::~CSSSelectorBag):
3150         (WebCore::CSSSelectorBag::isEmpty):
3151         (WebCore::CSSSelectorBag::append):
3152         (WebCore::CSSSelectorBag::takeAny):
3153         (WebCore::CSSSelector::~CSSSelector):
3154         (WebCore::CSSSelector::specificity):
3155         * css/CSSSelector.h:
3156
3157 2010-07-20  Yury Semikhatsky  <yurys@chromium.org>
3158
3159         Reviewed by Pavel Feldman.
3160
3161         [v8] Web Inspector: inspected page renderer crashes when inspected page has JS syntax error
3162         https://bugs.webkit.org/show_bug.cgi?id=42642
3163
3164         * bindings/v8/V8ConsoleMessage.cpp:
3165         (WebCore::V8ConsoleMessage::handler):
3166
3167 2010-07-20  Sheriff Bot  <webkit.review.bot@gmail.com>
3168
3169         Unreviewed, rolling out r63742.
3170         http://trac.webkit.org/changeset/63742
3171         https://bugs.webkit.org/show_bug.cgi?id=42641
3172
3173         Broke Leopard Intel build. (Requested by bbandix on #webkit).
3174
3175         * WebCore.pro:
3176         * page/Geolocation.cpp:
3177         (WebCore::Geolocation::startRequest):
3178         (WebCore::Geolocation::setIsAllowed):
3179         (WebCore::Geolocation::startUpdating):
3180         * page/Geolocation.h:
3181
3182 2010-07-20  Adam Roben  <aroben@apple.com>
3183
3184         Windows build fix
3185
3186         * WebCore.vcproj/WebCore.vcproj: Fixed case of "Name" attribute.
3187
3188 2010-07-20  Mahesh Kulkarni  <mahesh.kulkarni@nokia.com>
3189
3190         Reviewed by Steve Block.
3191
3192         Need to be able to configure Geolocation policy regarding user permissions
3193         https://bugs.webkit.org/show_bug.cgi?id=42068
3194
3195         Introducing new USE() flag PREEMPT_GEOLOCATION_PREMISSION using which
3196         acquires user permission first before starting location service
3197
3198         This change does not introduce any change in behavior for any platform. So there are no new tests
3199         added.
3200
3201         * WebCore.pro:
3202         * page/Geolocation.cpp:
3203         (WebCore::Geolocation::startRequest):
3204         (WebCore::Geolocation::setIsAllowed):
3205         (WebCore::Geolocation::startUpdating):
3206         * page/Geolocation.h:
3207
3208 2010-07-20  Alexander Pavlov  <apavlov@chromium.org>
3209
3210         Reviewed by Pavel Feldman.
3211
3212         Inspector: Resources Search Should Search Only Filtered Resources
3213         https://bugs.webkit.org/show_bug.cgi?id=28290
3214
3215         * inspector/front-end/AbstractTimelinePanel.js:
3216         (WebInspector.AbstractTimelinePanel.prototype._updateFilter):
3217         * inspector/front-end/ResourcesPanel.js:
3218         (WebInspector.ResourcesPanel.prototype.get searchableViews):
3219         * inspector/front-end/inspector.js:
3220         (WebInspector.performSearch):
3221         (WebInspector.doPerformSearch):
3222
3223 2010-07-20  Alexander Pavlov  <apavlov@chromium.org>
3224
3225         Reviewed by Pavel Feldman.
3226
3227         Web Inspector: Incorrect absolute URLs in tooltips of links in the ElementsTreeOutline
3228         https://bugs.webkit.org/show_bug.cgi?id=42626
3229
3230         * inspector/front-end/ElementsTreeOutline.js:
3231         (WebInspector.ElementsTreeElement.prototype._attributeHTML):
3232
3233 2010-07-20  Shinichiro Hamaji  <hamaji@chromium.org>
3234
3235         Unreviewed comment fix. The comment in IDL should be //, not # .
3236
3237         Code generator: ensure generated constants match their corresponding enums.
3238         https://bugs.webkit.org/show_bug.cgi?id=42250
3239
3240         * dom/OverflowEvent.idl:
3241
3242 2010-07-20  Hans Wennborg  <hans@chromium.org>
3243
3244         Reviewed by Steve Block.
3245
3246         Add WebCore/bindings/generic/RuntimeEnabledFeatures.cpp to build files
3247         https://bugs.webkit.org/show_bug.cgi?id=42380
3248
3249         RuntimeEnabledFeatures.cpp and .h were moved from bindings/v8 to
3250         bindings/generic a while a go (in r54593), but need to
3251         be added to the build in order to be used.
3252
3253         No new functionality so no new tests.
3254
3255         * Android.mk:
3256         * CMakeLists.txt:
3257         * GNUmakefile.am:
3258         * WebCore.pro:
3259         * WebCore.vcproj/WebCore.vcproj:
3260         * WebCore.xcodeproj/project.pbxproj:
3261         * websockets/WebSocket.cpp: Remove #if USE(V8), as RuntimeEnabledFeatures.cpp is no longer V8-only.
3262         * websockets/WebSocket.h: Ditto.
3263
3264 2010-07-20  Tony Chang  <tony@chromium.org>
3265
3266         Reviewed by Kent Tamura.
3267
3268         clean up style in ClipboardWin and PasteboardWin
3269         https://bugs.webkit.org/show_bug.cgi?id=42609
3270
3271         No new tests since this is just a style cleanup.
3272
3273         * platform/win/ClipboardUtilitiesWin.cpp:
3274         (WebCore::createGlobalData):
3275         (WebCore::markupToCFHTML):
3276         (WebCore::getURL):
3277         (WebCore::getPlainText):
3278         (WebCore::getTextHTML):
3279         (WebCore::fragmentFromFilenames):
3280         (WebCore::containsFilenames):
3281         (WebCore::fragmentFromCFHTML):
3282         (WebCore::fragmentFromHTML):
3283         * platform/win/ClipboardUtilitiesWin.h:
3284         * platform/win/ClipboardWin.cpp:
3285         (WebCore::pathRemoveBadFSCharacters):
3286         (WebCore::createGlobalHDropContent):
3287         (WebCore::createGlobalUrlFileDescriptor):
3288         (WebCore::writeURL):
3289         (WebCore::ClipboardWin::clearData):
3290         (WebCore::ClipboardWin::clearAllData):
3291         (WebCore::ClipboardWin::getData):
3292         (WebCore::ClipboardWin::types):
3293         (WebCore::ClipboardWin::declareAndWriteDragImage):
3294         (WebCore::ClipboardWin::writeRange):
3295         * platform/win/ClipboardWin.h:
3296         (WebCore::ClipboardWin::create):
3297         (WebCore::ClipboardWin::dataObject):
3298         * platform/win/PasteboardWin.cpp:
3299         (WebCore::PasteboardOwnerWndProc):
3300         (WebCore::Pasteboard::writeSelection):
3301         (WebCore::Pasteboard::writeURL):
3302         (WebCore::Pasteboard::plainText):
3303         (WebCore::Pasteboard::documentFragment):
3304
3305 2010-07-20  Yury Semikhatsky  <yurys@chromium.org>
3306
3307         Reviewed by Pavel Feldman.
3308
3309         Web Inspector: jump to source is broken for call frames logged by console.trace
3310         https://bugs.webkit.org/show_bug.cgi?id=42619
3311
3312         * inspector/front-end/ConsoleView.js:
3313         (WebInspector.ConsoleMessage.prototype._populateStackTraceTreeElement):
3314         * inspector/front-end/treeoutline.js: allow passing HTML element as a title.
3315         (TreeElement.prototype.set title):
3316         (TreeElement.prototype.set shouldRefreshChildren):
3317         (TreeElement.prototype._setListItemNodeContent):
3318         (TreeElement.prototype._attach):
3319
3320 2010-07-20  Tony Chang  <tony@chromium.org>
3321
3322         Reviewed by Dan Bernstein.
3323
3324         [chromium] crash in Position::getInlineBoxAndOffset
3325         https://bugs.webkit.org/show_bug.cgi?id=42202
3326
3327         Test: editing/selection/firstRect-crash.html
3328         Manual test: win/contextmenu-key2.html
3329
3330         * page/EventHandler.cpp: Code no longer needed
3331         * page/Frame.cpp:
3332         (WebCore::Frame::firstRectForRange): Normalize Positions to VisiblePositions because
3333             Positions may be pointing to nodes that have no renderer.  If there's no renderer,
3334             getInlineBoxAndOffset will crash.
3335
3336 2010-07-20  Adam Barth  <abarth@webkit.org>
3337
3338         Reviewed by Darin Adler.
3339
3340         Foster parenting depends on the current element at insertion time
3341         https://bugs.webkit.org/show_bug.cgi?id=42599
3342
3343         We need to consider the current element at insertion-time when deciding
3344         whether to redirect insertion to the foster parent.  Previously, we
3345         were considering the current element *both* at insertion-time and at
3346         the time we created the guard.
3347
3348         * html/HTMLConstructionSite.h:
3349         (WebCore::HTMLConstructionSite::RedirectToFosterParentGuard::RedirectToFosterParentGuard):
3350         * html/HTMLTreeBuilder.cpp:
3351         (WebCore::HTMLTreeBuilder::processStartTagForInTable):
3352         (WebCore::HTMLTreeBuilder::processEndTagForInTable):
3353         (WebCore::HTMLTreeBuilder::defaultForInTableText):
3354
3355 2010-07-20  Matthew Delaney  <mdelaney@apple.com>
3356
3357         Reviewed by Kenneth Rohde Christiansen.
3358
3359         Failing 2d.path.stroke.prune.curve philip canvas test
3360         https://bugs.webkit.org/show_bug.cgi?id=42190
3361
3362         * html/canvas/CanvasRenderingContext2D.cpp:
3363         (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.
3364         (WebCore::CanvasRenderingContext2D::quadraticCurveTo): Added in simple bounds as per the spec.
3365         (WebCore::CanvasRenderingContext2D::bezierCurveTo): Added in simple bounds as per the spec.
3366         * platform/graphics/cg/PathCG.cpp:
3367         (WebCore::Path::closeSubpath): Moved the check for an empty path up on level to make it platform independent and remove redundancy.
3368
3369 2010-07-19  Victoria Kirst  <vrk@google.com>
3370
3371         Reviewed by David Levin.
3372
3373         Added a simple implementation of VideoLayerChromium. Uses the
3374         LayerChromium::updateTextureRect() to send video frames to the
3375         GPU.
3376         https://bugs.webkit.org/show_bug.cgi?id=42234
3377
3378         * WebCore.gypi: Added include for VideoLayerChromium.
3379         * platform/graphics/chromium/GraphicsLayerChromium.cpp:
3380         (WebCore::GraphicsLayerChromium::setContentsToMedia): Implemented
3381         setContentsToMedia, though it does not seem to trigger a repaint
3382         correctly.
3383         * platform/graphics/chromium/GraphicsLayerChromium.h:
3384         * platform/graphics/chromium/VideoLayerChromium.cpp: Added.
3385         (WebCore::VideoLayerChromium::create):
3386         (WebCore::VideoLayerChromium::VideoLayerChromium):
3387         (WebCore::VideoLayerChromium::updateTextureContents):
3388         * platform/graphics/chromium/VideoLayerChromium.h: Added.
3389         (WebCore::VideoLayerChromium::drawsContent):
3390
3391 2010-07-19  Dirk Schulze  <krit@webkit.org>
3392
3393         Reviewed by Nikolas Zimmermann.
3394
3395         SVG CleanUp of SVGPathData parsing
3396         https://bugs.webkit.org/show_bug.cgi?id=41410
3397
3398         Cleanup the parsing code for SVG Paths. Move classes out of the convoluted SVGParserUtilities.cpp
3399         in their own files. Make use of WebCore specific objects in parsing code like FloatPoint, AffineTransform,
3400         etc. instead of using home-brewn solutions.
3401         The SVGPathParser parses a path data string and delivers the parsed segments and values to the SVGPathConsumer.
3402         SVGPathConsumer is the base class for SVGPathBuilder and SVGPathSegListBuilder, that either build the platform
3403         Path object or a SVGPathSegList out of the segments.
3404         We're now directly parsing floats instead of truncating precision to float afterwards.
3405
3406         SVG Path with an arc with radius of 0 does not render
3407         https://bugs.webkit.org/show_bug.cgi?id=40448
3408
3409         If one of the radii on the elliptic arc are zero, we should draw a line from the starting point to
3410         the end point according to the spec. Fixed this bug with this patch, because an is zero check and 
3411         an DRT check was neccessary with the current clean-up.
3412         Extended svg/dom/path-parser.xhml to check the correct behavior.
3413
3414         * Android.mk:
3415         * GNUmakefile.am:
3416         * WebCore.gypi:
3417         * WebCore.pro:
3418         * WebCore.vcproj/WebCore.vcproj:
3419         * WebCore.xcodeproj/project.pbxproj:
3420         * platform/graphics/FloatPoint.h:
3421         (WebCore::FloatPoint::move): Follow WebKit style and use multiple lines for the function.
3422         (WebCore::FloatPoint::scale): Scale FloatPoint.
3423         (WebCore::operator+=):
3424         (WebCore::operator+): Add two FloatPoints and give back the sum as FloatPoint.
3425         * svg/SVGAllInOne.cpp: Added new created files.
3426         * svg/SVGAnimateElement.cpp:
3427         (WebCore::SVGAnimateElement::calculateFromAndToValues): Use new PathParser to create a PathSegList.
3428         * svg/SVGAnimateMotionElement.cpp:
3429         (WebCore::SVGAnimateMotionElement::parseMappedAttribute): Use new PathParser to create a Path.
3430         * svg/SVGGlyphElement.cpp:
3431         (WebCore::parsePathData): Use new PathParser to create a Path.
3432         * svg/SVGParserUtilities.cpp:
3433         (WebCore::parseArcFlag): Removed parseArcFlag(double&), we parse in float now.
3434         * svg/SVGParserUtilities.h: Removed path parsing code and a lot of unneccessary includes.
3435         * svg/SVGPathBuilder.cpp: Added.
3436         (WebCore::SVGPathBuilder::SVGPathBuilder):
3437         (WebCore::SVGPathBuilder::build): Builds a normalized Path.
3438         (WebCore::SVGPathBuilder::moveTo):
3439         (WebCore::SVGPathBuilder::lineTo):
3440         (WebCore::SVGPathBuilder::curveToCubic):
3441         (WebCore::SVGPathBuilder::closePath):
3442         * svg/SVGPathBuilder.h: Added. We just create normalized Paths, so this functions should never be reached.
3443         (WebCore::SVGPathBuilder::lineToHorizontal):
3444         (WebCore::SVGPathBuilder::lineToVertical):
3445         (WebCore::SVGPathBuilder::curveToCubicSmooth):
3446         (WebCore::SVGPathBuilder::curveToQuadratic):
3447         (WebCore::SVGPathBuilder::curveToQuadraticSmooth):
3448         (WebCore::SVGPathBuilder::arcTo):
3449         * svg/SVGPathConsumer.h: Added. Base class of SVGPathBuilder and SVGPathSegListBuilder.
3450         (WebCore::):
3451         (WebCore::SVGPathConsumer::SVGPathConsumer):
3452         (WebCore::SVGPathConsumer::~SVGPathConsumer):
3453         * svg/SVGPathElement.cpp:
3454         (WebCore::SVGPathElement::parseMappedAttribute): Use new PathParser to create a PathSegList.
3455         * svg/SVGPathParser.cpp: Added.
3456         (WebCore::SVGPathParser::SVGPathParser):
3457         (WebCore::SVGPathParser::~SVGPathParser):
3458         (WebCore::SVGPathParser::parseClosePathSegment):
3459         (WebCore::SVGPathParser::parseMoveToSegment):
3460         (WebCore::SVGPathParser::parseLineToSegment):
3461         (WebCore::SVGPathParser::parseLineToHorizontalSegment):
3462         (WebCore::SVGPathParser::parseLineToVerticalSegment):
3463         (WebCore::SVGPathParser::parseCurveToCubicSegment):
3464         (WebCore::SVGPathParser::parseCurveToCubicSmoothSegment):
3465         (WebCore::SVGPathParser::parseCurveToQuadraticSegment):
3466         (WebCore::SVGPathParser::parseCurveToQuadraticSmoothSegment):
3467         (WebCore::SVGPathParser::parseArcToSegment):
3468         (WebCore::SVGPathParser::parsePathDataString):
3469         (WebCore::SVGPathParser::decomposeArcToCubic): Normalizes an arc to multiple cubic curves.
3470         * svg/SVGPathParser.h: Added.
3471         * svg/SVGPathSegListBuilder.cpp: Added.
3472         (WebCore::SVGPathSegListBuilder::SVGPathSegListBuilder):
3473         (WebCore::SVGPathSegListBuilder::build):
3474         (WebCore::SVGPathSegListBuilder::moveTo):
3475         (WebCore::SVGPathSegListBuilder::lineTo):
3476         (WebCore::SVGPathSegListBuilder::lineToHorizontal):
3477         (WebCore::SVGPathSegListBuilder::lineToVertical):
3478         (WebCore::SVGPathSegListBuilder::curveToCubic):
3479         (WebCore::SVGPathSegListBuilder::curveToCubicSmooth):
3480         (WebCore::SVGPathSegListBuilder::curveToQuadratic):
3481         (WebCore::SVGPathSegListBuilder::curveToQuadraticSmooth):
3482         (WebCore::SVGPathSegListBuilder::arcTo):
3483         (WebCore::SVGPathSegListBuilder::closePath):
3484         * svg/SVGPathSegListBuilder.h: Added.
3485
3486 2010-07-15  Yuzo Fujishima  <yuzo@google.com>
3487
3488         Reviewed by Dan Bernstein.
3489
3490         Fix for Bug 42342 - Font download error for an @font-face rule invalidates other @font-face rules for the same font-family
3491         https://bugs.webkit.org/show_bug.cgi?id=42342
3492
3493         Test: fast/css/font-face-download-error.html
3494
3495         * css/CSSSegmentedFontFace.cpp:
3496         (WebCore::CSSSegmentedFontFace::isValid): Valid if at least one font
3497         face is valid.
3498         (WebCore::CSSSegmentedFontFace::getFontData): Check validity for each
3499         font face.
3500         * css/CSSSegmentedFontFace.h: Make isValid private.
3501
3502 2010-07-19  Kent Tamura  <tkent@chromium.org>
3503
3504         Reviewed by Shinichiro Hamaji.
3505
3506         [Chromium] Fix style errors of RenderThemeChromiumWin.cpp
3507         https://bugs.webkit.org/show_bug.cgi?id=42568
3508
3509         * rendering/RenderThemeChromiumWin.cpp:
3510         (WebCore::):
3511         (WebCore::getNonClientMetrics):
3512         (WebCore::systemFontSize):
3513         (WebCore::pointsToPixels):
3514         (WebCore::querySystemBlinkInterval):
3515         (WebCore::RenderThemeChromiumWin::platformActiveSelectionBackgroundColor):
3516         (WebCore::RenderThemeChromiumWin::platformInactiveSelectionBackgroundColor):
3517         (WebCore::RenderThemeChromiumWin::platformActiveSelectionForegroundColor):
3518         (WebCore::RenderThemeChromiumWin::platformActiveTextSearchHighlightColor):
3519         (WebCore::RenderThemeChromiumWin::paintButton):
3520         (WebCore::RenderThemeChromiumWin::paintSliderTrack):
3521         (WebCore::RenderThemeChromiumWin::paintMenuList):
3522         (WebCore::RenderThemeChromiumWin::paintTextFieldInternal):
3523         (WebCore::RenderThemeChromiumWin::paintInnerSpinButton):
3524
3525 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3526
3527         (Hopefully the last) Windows build fix.
3528         
3529         Version of CoreServices.h in WebKit Support Libraries uses
3530         pragma once, unlike the version I have locally. Switch based
3531         on __COLORSYNCDEPRECATED__ instead.
3532
3533         * WebCorePrefix.h:
3534
3535 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3536
3537         Windows build fix.
3538
3539         * WebCorePrefix.h:
3540
3541 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3542
3543         Build fix.
3544
3545         * WebCorePrefix.h:
3546
3547 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3548
3549         Build fix.
3550
3551         * WebCorePrefix.h:
3552
3553 2010-07-19  Steve Falkenburg  <sfalken@apple.com>
3554
3555         Windows Build fixes for new ColorSync API.
3556         We support both new and old APIs, since the newer headers aren't in the tree yet.
3557
3558         * WebCorePrefix.h: Removed include of CoreServices.h. Included via ColorSyncPriv.h instead, since header may not be present.
3559         * platform/graphics/cg/ColorCG.cpp:
3560         (WebCore::createCGColor): Conditionally use new ColorSync API.
3561         * platform/graphics/opentype/OpenTypeUtilities.cpp: Define Fixed if CoreServices.h doesn't.
3562         * platform/graphics/win/GraphicsLayerCACF.cpp:
3563         (WebCore::GraphicsLayerCACF::updateLayerDrawsContent): Use 0 instead of nil, since nil is no longer present via CoreServices.h.
3564         * platform/graphics/win/MediaPlayerPrivateQuickTimeVisualContext.cpp: Include AssertMacros.h.
3565         * platform/network/cf/ResourceErrorCF.cpp:
3566         (WebCore::ResourceError::operator CFErrorRef): Use 0 instead of nil, since nil is no longer present via CoreServices.h.
3567
3568 2010-07-19  Tony Gentilcore  <tonyg@chromium.org>
3569
3570         Reviewed by Dimitri Glazkov.
3571
3572         Only set unloadEventEnd when the unload event is actually fired
3573         https://bugs.webkit.org/show_bug.cgi?id=42607
3574
3575         r63689 introduced this ASSERT and it began failing on Qt and Mac debug.
3576         For some reason, it did not fail on Windows (which is where I was
3577         testing). The ASSERT was disabled in r63699.
3578
3579         Test: page-cache related layout tests don't crash in debug mode on Mac.
3580
3581         * loader/FrameLoader.cpp:
3582         (WebCore::FrameLoader::stopLoading):
3583
3584 2010-07-19  Andreas Kling  <andreas.kling@nokia.com>
3585
3586         Reviewed by Kenneth Rohde Christiansen.
3587
3588         [Qt] Don't unnecessarily copy QPainterPath in fillPath() and strokePath()
3589         https://bugs.webkit.org/show_bug.cgi?id=42513
3590
3591         Avoid making unnecessary deep-copies of QPainterPaths that will
3592         be discarded after use.
3593
3594         * platform/graphics/qt/GraphicsContextQt.cpp:
3595         (WebCore::GraphicsContext::fillPath):
3596         (WebCore::GraphicsContext::strokePath):
3597
3598 2010-07-19  Kenneth Russell  <kbr@google.com>
3599
3600         Reviewed by Nate Chapin.
3601
3602         WebGL rendering results must be made available to Canvas.toDataURL and 2D drawImage
3603         https://bugs.webkit.org/show_bug.cgi?id=34719
3604
3605         Fixed compiler warning introduced by original patch. No new tests;
3606         covered by existing tests.
3607
3608         * platform/graphics/mac/GraphicsContext3DMac.mm:
3609         (WebCore::GraphicsContext3D::paintRenderingResultsToCanvas):
3610
3611 2010-07-19  Anders Carlsson  <andersca@apple.com>
3612
3613         Reviewed by Darin Adler, Adam Roben, Dan Bernstein and Sam Weinig.
3614
3615         Handle NP_ASFILE and NP_ASFILEONLY transfer modes
3616         https://bugs.webkit.org/show_bug.cgi?id=42587
3617
3618         * WebCore.exp.in:
3619         Export functions from FileSystem.h
3620         
3621         * platform/mac/FileSystemMac.mm:
3622         (WebCore::openTemporaryFile):
3623         Try to create a temporary file using mkstemp.
3624
3625 2010-07-19  Anders Carlsson  <andersca@apple.com>
3626
3627         Reviewed by Sam Weinig.
3628
3629         Implement NPN_PostURLNotify
3630         https://bugs.webkit.org/show_bug.cgi?id=42602
3631
3632         Export symbols needed by WebKit2.
3633
3634         * WebCore.exp.in:
3635
3636 2010-07-19  Tony Gentilcore  <tonyg@chromium.org>
3637
3638         Unreviewed build fix.
3639
3640         Remove overzealous ASSERT from r63689
3641         https://bugs.webkit.org/show_bug.cgi?id=42606
3642
3643         No new tests because no new functionality.
3644