2010-08-13 Gustavo Noronha Silva <gustavo.noronha@collabora.co.uk>
[WebKit-https.git] / WebCore / ChangeLog
1 2010-08-13  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>
2
3         Reviewed by Kenneth Rohde Christiansen and Martin Robinson.
4
5         [GTK] GTK+ drawing idle starved
6         https://bugs.webkit.org/show_bug.cgi?id=43994
7
8         Tune down the priority of the shared timers to guarantee GTK+ has
9         an opportunity to redraw.
10
11         * platform/gtk/SharedTimerGtk.cpp:
12         (WebCore::setSharedTimerFireTime):
13
14 2010-08-16  Leandro Pereira  <leandro@profusion.mobi>
15
16         [EFL] Build fix after r65366.
17
18         EFL port does not support automated tests yet.
19
20         * CMakeLists.txt: Use if (VAR) instead of if (${VAR}) to check if
21         they're empty.
22
23 2010-08-16  Pavel Feldman  <pfeldman@chromium.org>
24
25         Reviewed by Yury Semikhatsky.
26
27         Web Inspector: Make InjectedScript proto-based.
28         https://bugs.webkit.org/show_bug.cgi?id=44028
29
30         * inspector/front-end/InjectedScript.js:
31         (injectedScriptConstructor):
32         (injectedScriptConstructor.):
33
34 2010-08-16  Csaba Osztrogonác  <ossy@webkit.org>
35
36         Reviewed by Adam Roben.
37
38         Fix warning in WebCore/plugins/win/PluginPackageWin.cpp
39         https://bugs.webkit.org/show_bug.cgi?id=43920
40
41         * plugins/win/PluginPackageWin.cpp:
42         (WebCore::PluginPackage::fetchInfo):
43
44 2010-08-16  Pavel Feldman  <pfeldman@chromium.org>
45
46         Not reviewed. Touch inspector file to force frontend deployment.
47
48         * inspector/InspectorController.cpp:
49
50 2010-08-16  Kevin Ollivier  <kevino@theolliviers.com>
51
52         [wx] Build fix for non-Win ports on Windows.
53
54         * platform/graphics/WOFFFileFormat.cpp:
55
56 2010-08-16  Adam Langley  <agl@chromium.org>
57
58         Reviewed by Tony Chang.
59
60         [chromium] Improve spacing support for complex text on Linux
61
62         https://bugs.webkit.org/show_bug.cgi?id=39014
63
64         Previously, our complex text support ignored word-spacing,
65         justification and letter-spacing. This fixes the first two issues and
66         allows us to render Scribd's HTML5 documents much better.
67
68         Test: fast/text/atsui-spacing-features.html
69
70         * platform/graphics/chromium/FontLinux.cpp:
71         (WebCore::TextRunWalker::TextRunWalker):
72         (WebCore::TextRunWalker::setWordSpacingAdjustment):
73         (WebCore::TextRunWalker::setLetterSpacingAdjustment):
74         (WebCore::TextRunWalker::setPadding):
75         (WebCore::TextRunWalker::setGlyphXPositions):
76         (WebCore::TextRunWalker::isCodepointSpace):
77         (WebCore::Font::drawComplexText):
78         (WebCore::Font::floatWidthForComplexText):
79         (WebCore::Font::offsetForPositionForComplexText):
80         (WebCore::Font::selectionRectForComplexText):
81
82 2010-08-16  Csaba Osztrogonác  <ossy@webkit.org>
83
84         Reviewed by Kenneth Rohde Christiansen.
85
86         Fix warnings: unknown conversion type character 'l' in format
87         https://bugs.webkit.org/show_bug.cgi?id=43359
88
89         * loader/icon/IconDatabase.cpp:
90         (WebCore::IconDatabase::performURLImport): Replace %zu with %lu, because Windows doesn't understand z modifier.
91         (WebCore::IconDatabase::pruneUnretainedIcons): Use %I64i intsead of %lli on Windows.
92         * platform/graphics/qt/MediaPlayerPrivatePhonon.cpp:
93         (WebCore::MediaPlayerPrivate::totalTimeChanged): Use %I64d intsead of %lld on Windows.
94         * platform/sql/SQLiteDatabase.cpp:
95         (WebCore::SQLiteDatabase::setMaximumSize): Use %I64i intsead of %lli on Windows.
96         * storage/DatabaseTracker.cpp:
97         (WebCore::DatabaseTracker::setQuota): Use %I64u intsead of %llu on Windows.
98
99 2010-08-16  Csaba Osztrogonác  <ossy@webkit.org>
100
101         Reviewed by Kenneth Rohde Christiansen.
102
103         Fix warnings: unknown conversion type character 'l' in format
104         https://bugs.webkit.org/show_bug.cgi?id=43359
105
106         * loader/FTPDirectoryParser.cpp: Use %I64u format specifier instead of %llu on Windows.
107         (WebCore::parseOneFTPLine):
108
109 2010-08-16  Ariya Hidayat  <ariya@sencha.com>
110
111         Reviewed by Kenneth Rohde Christiansen.
112
113         [Qt] Shadow blur for rectangle fill
114         https://bugs.webkit.org/show_bug.cgi?id=44025
115
116         The blur implementation follows the approximation of Gaussian blur
117         with three successive box blurs, working on the alpha channel only.
118         (see http://www.w3.org/TR/SVG/filters.html#feGaussianBlur).
119
120         * WebCore.pro: Refer to ContexShadow.* files.
121         * platform/graphics/qt/ContextShadow.cpp: Added.
122         (WebCore::ContextShadow::ContextShadow):
123         (WebCore::ContextShadow::clear):
124         (WebCore::blurHorizontal): Box blurs in horizontal direction, working
125         on QImage (in-place).
126         (WebCore::shadowBlur): Main blur function which just calls the
127         above blurHorizontal function twice, once with transposed image
128         (equivalent as it would have done vertical box blurs).
129         (WebCore::ContextShadow::drawShadowRect):
130         * platform/graphics/qt/ContextShadow.h: Added.
131         (WebCore::ContextShadow::):
132         * platform/graphics/qt/GraphicsContextQt.cpp:
133         (WebCore::GraphicsContextPlatformPrivate::hasShadow):
134         (WebCore::GraphicsContext::restorePlatformState):
135         (WebCore::GraphicsContext::fillRect): Use ContextShadow for the shadow.
136         (WebCore::GraphicsContext::setPlatformShadow):
137
138 2010-08-16  Andras Becsi  <abecsi@webkit.org>
139
140         Rubber-stamped by Gustavo Noronha Silva.
141
142         [GTK] Build fix for MathML code generation.
143
144         No new tests needed.
145
146         * GNUmakefile.am:
147
148 2010-08-16  Sheriff Bot  <webkit.review.bot@gmail.com>
149
150         Unreviewed, rolling out r65414.
151         http://trac.webkit.org/changeset/65414
152         https://bugs.webkit.org/show_bug.cgi?id=44055
153
154         GTL & QT tests failing. (Requested by pfeldman on #webkit).
155
156         * inspector/front-end/InjectedScript.js:
157         (injectedScriptConstructor):
158         (injectedScriptConstructor.):
159
160 2010-08-16  Ariya Hidayat  <ariya@sencha.com>
161
162         Reviewed by Antonio Gomes.
163
164         [Qt] Save and restore shadow state in GraphicsContextQt
165         https://bugs.webkit.org/show_bug.cgi?id=44031
166
167         This is needed after r65362.
168         Use a new class ContextShadowParameter to track and keep the shadow
169         state. We use this rather than using GraphicsContextState to allow
170         possible optimizations (right now only to determine the shadow type,
171         but in future it might cover things like cached scratch image,
172         persistent shader, etc).
173
174         * platform/graphics/qt/GraphicsContextQt.cpp:
175         (WebCore::ContextShadowParameter::):
176         (WebCore::ContextShadowParameter::ContextShadowParameter):
177         (WebCore::ContextShadowParameter::clear):
178         (WebCore::GraphicsContextPlatformPrivate::hasShadow):
179         (WebCore::GraphicsContextPlatformPrivate::GraphicsContextPlatformPrivate):
180         (WebCore::GraphicsContext::savePlatformState):
181         (WebCore::GraphicsContext::restorePlatformState):
182         (WebCore::GraphicsContext::drawRect):
183         (WebCore::GraphicsContext::drawLine):
184         (WebCore::GraphicsContext::strokeArc):
185         (WebCore::GraphicsContext::drawConvexPolygon):
186         (WebCore::GraphicsContext::fillPath):
187         (WebCore::GraphicsContext::strokePath):
188         (WebCore::GraphicsContext::fillRect):
189         (WebCore::GraphicsContext::fillRoundedRect):
190         (WebCore::GraphicsContext::setPlatformShadow):
191         (WebCore::GraphicsContext::clearPlatformShadow):
192
193 2010-07-28  Marcus Bulach  <bulach@chromium.org>
194
195         Reviewed by Adam Barth.
196
197         Implement IDBKeyPath parser.
198         https://bugs.webkit.org/show_bug.cgi?id=42976
199
200         IDBKeyPath is an internal class, exposed only for unit tests for now.
201         It'll allow indexed database implementation to traverse a given key path and obtain the corresponding properties
202         from a SerializedScriptValue.
203
204         * WebCore.gypi:
205         * storage/IDBKeyPath.cpp: Added.
206         (WebCore::IDBKeyPath::create):
207         (WebCore::IDBKeyPath::IDBKeyPath):
208         (WebCore::IDBKeyPath::~IDBKeyPath):
209         (WebCore::IDBKeyPath::hasNext):
210         (WebCore::IDBKeyPath::next):
211         (WebCore::IDBKeyPath::identifier):
212         (WebCore::IDBKeyPath::hasIndex):
213         (WebCore::IDBKeyPath::indexValue):
214         (WebCore::IDBKeyPath::parserError):
215         (WebCore::IDBKeyPath::parse):
216         (WebCore::IDBKeyPath::Lexer::lex):
217         (WebCore::isSafeIdentifierStartCharacter):
218         (WebCore::isSafeIdentifierCharacter):
219         (WebCore::IDBKeyPath::Lexer::lexIdentifier):
220         (WebCore::IDBKeyPath::Lexer::lexNumber):
221         * storage/IDBKeyPath.h: Added.
222         (WebCore::IDBKeyPath::Token::Token):
223         (WebCore::IDBKeyPath::):
224         (WebCore::IDBKeyPath::Lexer::Lexer):
225         (WebCore::IDBKeyPath::Lexer::next):
226         (WebCore::IDBKeyPath::Lexer::currentToken):
227
228 2010-08-13  Steve Block  <steveblock@google.com>
229
230         Reviewed by Alexey Proskuryakov.
231
232         Geolocation requests in progress when the frame is disconnected should invoke the error callback
233         https://bugs.webkit.org/show_bug.cgi?id=43974
234
235         If requests are ongoing when the Frame is disconnected, we abort them with a fatal error.
236         To do this, when Geolocation::disconnectFrame() is called we call cancelAllRequests() to
237         set a fatal error on all ongoing requests. Once the requests have made their error
238         callbacks, they are deleted. Note that we check at callback time that the script
239         execution context for the callback is still valid, so it's safe to attempt these
240         callbacks even after the Geolocation object's Frame has gone.
241
242         This requires a change to allow multiple calls to GeoNotifier::setFatalError().
243         For repeated calls, we do not replace the existing error. This ensures that when
244         permission has been denied and the frame is then disconnected, the error callback
245         reports the permission error, as required by the spec.
246
247         Tests: fast/dom/Geolocation/disconnected-frame.html
248                fast/dom/Geolocation/disconnected-frame-permission-denied.html
249
250         * page/Geolocation.cpp:
251         (WebCore::Geolocation::GeoNotifier::setFatalError):
252         (WebCore::Geolocation::disconnectFrame):
253         (WebCore::Geolocation::startRequest):
254         (WebCore::Geolocation::cancelAllRequests):
255         * page/Geolocation.h:
256
257 2010-08-16  Pavel Feldman  <pfeldman@chromium.org>
258
259         Reviewed by Joseph Pecoraro.
260
261         Web Inspector: upstream frontend-side WebSocket transport.
262         https://bugs.webkit.org/show_bug.cgi?id=43970
263
264         Chromium already has an alternate WebSocket-based communication channel with
265         the backend. Upstreaming it in this change. We will agree on the URI
266         of the remote service as the protocol matures.
267
268         * inspector/front-end/inspector.js:
269         (.WebInspector.socket.onmessage):
270         (.WebInspector.socket.onerror):
271         (.WebInspector.socket.onopen):
272         (WebInspector.loaded):
273         (WebInspector.doLoadedDone):
274         (WebInspector_syncDispatch):
275
276 2010-08-16  Pavel Feldman  <pfeldman@chromium.org>
277
278         Reviewed by Yury Semikhatsky.
279
280         Web Inspector: Make InjectedScript proto-based.
281         https://bugs.webkit.org/show_bug.cgi?id=44028
282
283         * inspector/front-end/InjectedScript.js:
284         (injectedScriptConstructor):
285         (injectedScriptConstructor.):
286         * inspector/front-end/inspector.js:
287         (WebInspector.loaded):
288
289 2010-08-16  Pavel Feldman  <pfeldman@chromium.org>
290
291         Reviewed by Joseph Pecoraro.
292
293         Web Inspector: [REGRESSION] console's clear command does not work.
294         https://bugs.webkit.org/show_bug.cgi?id=44027
295
296         Test: inspector/console-command-clear.html
297
298         * inspector/Inspector.idl:
299         * inspector/InspectorBackend.cpp:
300         * inspector/InspectorBackend.h:
301         * inspector/InspectorController.cpp:
302         (WebCore::InspectorController::clearConsoleMessages):
303         * inspector/front-end/ConsoleView.js:
304         (WebInspector.ConsoleView.prototype.requestClearMessages):
305         * inspector/front-end/inspector.js:
306         (WebInspector.consoleMessagesCleared):
307
308 2010-08-16  Jeremy Orlow  <jorlow@chromium.org>
309
310         Ugh, the #if ENABLE() was supposed to be added to Settings not GroupSettings.
311
312         * page/GroupSettings.cpp:
313         (WebCore::GroupSettings::GroupSettings):
314         (WebCore::GroupSettings::setLocalStorageQuotaBytes):
315         * page/GroupSettings.h:
316         * page/Settings.cpp:
317         (WebCore::Settings::Settings):
318         * page/Settings.h:
319
320 2010-08-16  Jeremy Orlow  <jorlow@chromium.org>
321
322         Last build fix?
323
324         * page/PageGroup.cpp:
325         (WebCore::PageGroup::PageGroup):
326
327 2010-08-16  Jeremy Orlow  <jorlow@chromium.org>
328
329         Ugh...another build fix.
330
331         * page/GroupSettings.cpp:
332         (WebCore::GroupSettings::GroupSettings):
333         * page/GroupSettings.h:
334
335 2010-08-16  Jeremy Orlow  <jorlow@chromium.org>
336
337         Build fix.
338
339         * GNUmakefile.am:
340
341 2010-08-16  Jeremy Orlow  <jorlow@chromium.org>
342
343         Build fix.
344
345         * page/PageGroup.cpp:
346         (WebCore::PageGroup::localStorage):
347
348 2010-08-13  Jeremy Orlow  <jorlow@chromium.org>
349
350         Some settings are linked to the PageGroup not the Page.  Create a new class for those.
351         https://bugs.webkit.org/show_bug.cgi?id=43794
352
353         Since persistient storage is shared between pages, there's no way to modify
354         some settings related to it on a per page basis.  As such, it's not technically
355         correct for these settings to be on the Page's settings.  Create a new class
356         called GroupSettings, move  group-wide local storage setting there (the other we
357         can't), and add a new setting for IndexedDB's path (which is prompting this change).
358
359         No behavior has changed, so no tests.
360
361         * Android.mk:
362         * CMakeLists.txt:
363         * GNUmakefile.am:
364         * WebCore.gypi:
365         * WebCore.pro:
366         * WebCore.vcproj/WebCore.vcproj:
367         * WebCore.xcodeproj/project.pbxproj:
368         * page/GroupSettings.cpp: Added.
369         (WebCore::GroupSettings::GroupSettings):
370         (WebCore::GroupSettings::setLocalStorageQuotaBytes):
371         (WebCore::GroupSettings::setLocalStorageDatabasePath):
372         (WebCore::GroupSettings::setIndexedDBDatabasePath):
373         * page/GroupSettings.h: Added.
374         (WebCore::GroupSettings::create):
375         (WebCore::GroupSettings::localStorageQuotaBytes):
376         (WebCore::GroupSettings::localStorageDatabasePath):
377         (WebCore::GroupSettings::indexedDBDatabasePath):
378         * page/PageGroup.cpp:
379         (WebCore::PageGroup::PageGroup):
380         (WebCore::PageGroup::localStorage):
381         * page/PageGroup.h:
382         (WebCore::PageGroup::groupSettings):
383         * page/Settings.cpp:
384         (WebCore::Settings::Settings):
385         (WebCore::Settings::setSessionStorageQuota):
386         * page/Settings.h:
387
388 2010-08-16  Adam Barth  <abarth@webkit.org>
389
390         Reviewed by Alexey Proskuryakov.
391
392         HTML5 fragment parsing doesn't load iframes
393         https://bugs.webkit.org/show_bug.cgi?id=44038
394
395         Previously, this code assumed that every source iframe was actually
396         attached to a render tree.  It tried to "keep the iframe alive" as it
397         moved documents.  However, if the iframe wasn't attached to begin with,
398         it was never "alive," so trying to keep a dead iframe alive lead to a
399         ...  zombie iframe.
400
401         Test: fast/frames/adopt-from-created-document.html
402
403         * dom/Document.cpp:
404         (WebCore::Document::adoptNode):
405
406 2010-08-13  MORITA Hajime  <morrita@google.com>
407
408         Reviewed by Tony Chang.
409
410         https://bugs.webkit.org/show_bug.cgi?id=43778
411         Dropping should fire textInput event
412
413         This change:
414         - added TextEvent::m_isDrop to distinguish drop-initiated events.
415         - added to dispatch TextEvent from DragController before making a side effect
416           and cancel the side effect if the event prevents default
417
418         Tests: editing/pasteboard/drop-text-events-sideeffect.html
419                editing/pasteboard/drop-text-events.html
420
421         * dom/TextEvent.cpp:
422         (WebCore::TextEvent::createForDrop): Added.
423         (WebCore::TextEvent::TextEvent):
424         * dom/TextEvent.h:
425         (WebCore::TextEvent::isDrop): Added.
426         * editing/Editor.cpp:
427         (WebCore::Editor::handleTextEvent):
428         (WebCore::Editor::findEventTargetFrom): Extracted from findEventTargetFromSelection().
429         (WebCore::Editor::findEventTargetFromSelection): 
430         * editing/Editor.h:
431         * page/DragController.cpp:
432         (WebCore::DragController::dispatchTextInputEventFor): Added.
433         (WebCore::DragController::concludeEditDrag): Added an event dispatch path.
434         * page/DragController.h:
435
436 2010-08-15  Ariya Hidayat  <ariya@sencha.com>
437
438         Unreviewed, rolling out r65393.
439         http://trac.webkit.org/changeset/65393
440         https://bugs.webkit.org/show_bug.cgi?id=44031
441
442         Breaks some canvas tests.
443
444         * platform/graphics/qt/GraphicsContextQt.cpp:
445         (WebCore::GraphicsContextPlatformPrivate::):
446         (WebCore::GraphicsContextPlatformPrivate::hasShadow):
447         (WebCore::GraphicsContextPlatformPrivate::GraphicsContextPlatformPrivate):
448         (WebCore::GraphicsContext::savePlatformState):
449         (WebCore::GraphicsContext::restorePlatformState):
450         (WebCore::GraphicsContext::strokeArc):
451         (WebCore::GraphicsContext::drawConvexPolygon):
452         (WebCore::GraphicsContext::fillPath):
453         (WebCore::GraphicsContext::strokePath):
454         (WebCore::GraphicsContext::fillRect):
455         (WebCore::GraphicsContext::fillRoundedRect):
456         (WebCore::GraphicsContext::setPlatformShadow):
457         (WebCore::GraphicsContext::clearPlatformShadow):
458
459 2010-08-15  Ariya Hidayat  <ariya@sencha.com>
460
461         Reviewed by Antonio Gomes.
462
463         [Qt] Save and restore shadow state in GraphicsContextQt
464         https://bugs.webkit.org/show_bug.cgi?id=44031
465
466         This is needed after r65362.
467         Use a new class ContextShadowParameter to track and keep the shadow
468         state. We use this rather than using GraphicsContextState to allow
469         possible optimizations (right now only to determine the shadow type,
470         but in future it might cover things like cached scratch image,
471         persistent shader, etc).
472
473         * platform/graphics/qt/GraphicsContextQt.cpp:
474         (WebCore::ContextShadowParameter::):
475         (WebCore::ContextShadowParameter::ContextShadowParameter):
476         (WebCore::ContextShadowParameter::clear):
477         (WebCore::GraphicsContextPlatformPrivate::hasShadow):
478         (WebCore::GraphicsContextPlatformPrivate::GraphicsContextPlatformPrivate):
479         (WebCore::GraphicsContext::savePlatformState):
480         (WebCore::GraphicsContext::restorePlatformState):
481         (WebCore::GraphicsContext::drawRect):
482         (WebCore::GraphicsContext::drawLine):
483         (WebCore::GraphicsContext::strokeArc):
484         (WebCore::GraphicsContext::drawConvexPolygon):
485         (WebCore::GraphicsContext::fillPath):
486         (WebCore::GraphicsContext::strokePath):
487         (WebCore::GraphicsContext::fillRect):
488         (WebCore::GraphicsContext::fillRoundedRect):
489         (WebCore::GraphicsContext::setPlatformShadow):
490         (WebCore::GraphicsContext::clearPlatformShadow):
491
492 2010-08-15  Ariya Hidayat  <ariya@sencha.com>
493
494         Reviewed by Antonio Gomes.
495
496         [Qt] Border should not cast shadows
497         https://bugs.webkit.org/show_bug.cgi?id=44015
498
499         Shadows should be not casted except for brush fill (which is not what
500         drawLine and drawRect are supposed to do).
501
502         * platform/graphics/qt/GraphicsContextQt.cpp:
503         (WebCore::GraphicsContext::drawRect): Remove shadow painting,
504         (WebCore::GraphicsContext::drawLine): ditto.
505
506 2010-08-15  Ryuan Choi  <ryuan.choi@samsung.com>
507
508         Reviewed by Antonio Gomes.
509
510         [EFL] Build error on r65378
511         https://bugs.webkit.org/show_bug.cgi?id=44019
512
513         Change build script for HTMLEntityNames.json instead of HTMLEntityNames.gperf
514
515         * CMakeLists.txt:
516
517 2010-08-15  Kevin Ollivier  <kevino@theolliviers.com>
518
519         [wx] Build fix, add missing header.
520
521         * platform/graphics/WOFFFileFormat.cpp:
522
523 2010-08-15  Kevin Ollivier  <kevino@theolliviers.com>
524
525         [wx] Build fix, don't build PluginDataNone.cpp even on GTK as PluginDataWx.cpp compiles.
526
527         * wscript:
528
529 2010-08-15  Adam Barth  <abarth@webkit.org>
530
531         Reviewed by Eric Seidel.
532
533         document.write() doesn't flush plain text
534         https://bugs.webkit.org/show_bug.cgi?id=8961
535
536         Originally I thought we should buffer the character tokens until we've
537         reached the end of the input stream, but that's not how the spec
538         handles things (it emits the character tokens one-by-one).
539
540         Tests: fast/tokenizer/flush-characters-in-document-write-evil.html
541                fast/tokenizer/flush-characters-in-document-write.html
542
543         * html/HTMLTokenizer.cpp:
544         (WebCore::HTMLTokenizer::emitEndOfFile):
545         (WebCore::HTMLTokenizer::nextToken):
546         (WebCore::HTMLTokenizer::haveBufferedCharacterToken):
547             - Renamed this function now that it's simplier.
548         * html/HTMLTokenizer.h:
549
550 2010-08-15  Adam Barth  <abarth@webkit.org>
551
552         Reviewed by Eric Seidel.
553
554         Don't try to replace a non-existent document after executing JavaScript URLs
555         https://bugs.webkit.org/show_bug.cgi?id=44024
556
557         Synchronous JavaScript execution is evil.  Previously, the frame was
558         deleted after executing the JavaScript URL, so we'd get confused when
559         we tried to replace its document.
560
561         Test: fast/frames/javascript-url-for-deleted-frame.html
562
563         * bindings/ScriptControllerBase.cpp:
564         (WebCore::ScriptController::executeIfJavaScriptURL):
565
566 2010-08-14  Sheriff Bot  <webkit.review.bot@gmail.com>
567
568         Unreviewed, rolling out r65374.
569         http://trac.webkit.org/changeset/65374
570         https://bugs.webkit.org/show_bug.cgi?id=44018
571
572         broke a myriad of svg tests (Requested by kling on #webkit).
573
574         * platform/graphics/qt/PathQt.cpp:
575         (WebCore::Path::strokeBoundingRect):
576
577 2010-08-14  Eric Seidel  <eric@webkit.org>
578
579         Reviewed by Adam Barth.
580
581         XSLTProcessor.transformToFragment should not pass contextElement to html/xml document parsers
582         https://bugs.webkit.org/show_bug.cgi?id=44017
583
584         This fixes two tests which otherwise would fail when the
585         HTML5 TreeBuilder is enabled for fragment parsing.
586
587         This revealed a bug in HTMLTreeBuilder() which I fixed as well.
588
589         * html/HTMLTreeBuilder.cpp:
590         (WebCore::HTMLTreeBuilder::HTMLTreeBuilder):
591         * xml/XSLTProcessor.cpp:
592         (WebCore::createFragmentFromSource):
593
594 2010-08-14  Andreas Kling  <andreas.kling@nokia.com>
595
596         Reviewed by Ariya Hidayat.
597
598         [Qt] Path: Fast approximation of stroke bounding rects
599         https://bugs.webkit.org/show_bug.cgi?id=44014
600
601         * platform/graphics/qt/PathQt.cpp:
602         (WebCore::Path::strokeBoundingRect): Use QPainterPath::controlPointRect()
603         to calculate a slightly larger bounding rect in a fraction of the time.
604
605 2010-08-14  Eric Seidel  <eric@webkit.org>
606
607         Reviewed by Adam Barth.
608
609         Move innerHTML off of Range::createContextualFragment
610         https://bugs.webkit.org/show_bug.cgi?id=44011
611
612         This makes all the editing tests stop failing under the
613         HTML5 TreeBuilder.
614
615         In a previous patch Adam disabled some of the (wrong) code
616         in createContextualFragment when in html5 treebuilder mode.
617         The problem is that the editing code depends on
618         createContextualFragment (as well as the Range DOM API),
619         so we have to leave this code in until we can figure
620         out if it's right or wrong behavior for the Range API.
621
622         This patch moves innerHTML/outerHTML off of using
623         createContextualFragment and instead calls
624         DocumentFragment::parseHTML directly (which is much simpler).
625
626         I expect Adam and I will end up turning on the HTML5 TreeBuilder
627         in the next couple days, and most of the if branches I added
628         here will go away for good.
629
630         I renamed Element::createContextualFragment to
631         deprecatedCreateContextualFragment, to indicate that it's a
632         dead API.  The editing code should move off of it and consider
633         using DocumentFragment::parseHTML instead.
634
635         * dom/Element.cpp:
636         (WebCore::Element::deprecatedCreateContextualFragment):
637         * dom/Element.h:
638         * dom/Range.cpp:
639         (WebCore::Range::createContextualFragment):
640         * editing/markup.cpp:
641         (WebCore::createFragmentFromMarkup):
642         * html/HTMLElement.cpp:
643         (WebCore::useLegacyTreeBuilder):
644         (WebCore::HTMLElement::deprecatedCreateContextualFragment):
645         (WebCore::HTMLElement::setInnerHTML):
646         (WebCore::HTMLElement::setOuterHTML):
647         * html/HTMLElement.h:
648
649 2010-08-14  Tasuku Suzuki  <tasuku.suzuki@nokia.com>
650
651         Reviewed by Antonio Gomes.
652
653         [Qt] Fix compilation with QT_NO_LINEEDIT
654         https://bugs.webkit.org/show_bug.cgi?id=38324
655
656         * bridge/qt/qt_runtime.cpp:
657         (JSC::Bindings::convertValueToQVariant):
658         (JSC::Bindings::convertQVariantToValue):
659
660 2010-08-14  Rob Buis  <rwlbuis@gmail.com>
661
662         Reviewed by Dirk Schulze.
663
664         IE SVG empty path test fails
665         https://bugs.webkit.org/show_bug.cgi?id=44008
666
667         Change createLine to allow creating zero-length lines. This makes it possible
668         to have markers and line-caps applied to zero-length lines.
669
670         Test: svg/custom/marker-empty-path.svg
671
672         * platform/graphics/Path.cpp:
673         (WebCore::Path::createLine):
674
675 2010-08-14  Patrick Gansterer  <paroga@paroga.com>
676
677         Reviewed by Kenneth Rohde Christiansen.
678
679         [CMake] Move SQL files into a IF (ENABLE_DATABASE) section
680         https://bugs.webkit.org/show_bug.cgi?id=43979
681
682         * CMakeLists.txt:
683
684 2010-08-14  Matthew Delaney  <mdelaney@apple.com>
685
686         Reviewed by Kenneth Rohde Christiansen.
687
688         Canvas is not reset when setting canvas.width
689         https://bugs.webkit.org/show_bug.cgi?id=43341
690
691         * html/HTMLCanvasElement.cpp:
692         (WebCore::HTMLCanvasElement::reset): Moved the call to
693         hasCreatedImageBuffer up *before* the call to setSurface since setSurface
694         sets hasCreatedImageBuffer to false every time.
695
696 2010-08-14  Patrick Gansterer  <paroga@paroga.com>
697
698         Reviewed by Kenneth Rohde Christiansen.
699
700         [CMake] Set target properties only if available
701         https://bugs.webkit.org/show_bug.cgi?id=43978
702
703         * CMakeLists.txt:
704
705 2010-08-14  Dirk Schulze  <krit@webkit.org>
706
707         Unreviewed sort of Xcode project file.
708
709         * WebCore.xcodeproj/project.pbxproj:
710
711 2010-08-14  Ariya Hidayat  <ariya@sencha.com>
712
713         Reviewed by Simon Hausmann and Kenneth Rohde Christiansen.
714
715         [Qt] Refactor shadow state handling in GraphicsContextQt
716         https://bugs.webkit.org/show_bug.cgi?id=44006
717
718         * platform/graphics/qt/GraphicsContextQt.cpp:
719         (WebCore::GraphicsContextPlatformPrivate::): Added shadow states as
720         member variables.
721         (WebCore::GraphicsContextPlatformPrivate::hasShadow): Convenient
722         function to check whether there is shadow or not.
723         (WebCore::GraphicsContextPlatformPrivate::GraphicsContextPlatformPrivate):
724         (WebCore::GraphicsContext::drawRect): Use shadow states instead of
725         calling getShadow.
726         (WebCore::GraphicsContext::drawLine): ditto.
727         (WebCore::GraphicsContext::strokeArc): ditto.
728         (WebCore::GraphicsContext::drawConvexPolygon): ditto.
729         (WebCore::GraphicsContext::fillPath): ditto.
730         (WebCore::GraphicsContext::strokePath): ditto.
731         (WebCore::GraphicsContext::fillRect): Removes the use of helper function
732         drawBorderlessRectShadow as the code already becomes a lot simpler.
733         (WebCore::GraphicsContext::fillRoundedRect): Removes the use of helper
734         function drawFilledShadowPath as the code already becomes a lot simpler.
735         (WebCore::GraphicsContext::setPlatformShadow): Store shadow states
736         and find out the shadow type (complexity) for future use.
737         (WebCore::GraphicsContext::clearPlatformShadow): Reset shadow states.
738
739 2010-08-13  Adam Barth  <abarth@webkit.org>
740
741         Reviewed by Eric Seidel.
742
743         Rename some concepts in HTML entity search to be more self-documenting
744         https://bugs.webkit.org/show_bug.cgi?id=44004
745
746         Start and end are now first and last, which lets us use some better
747         names for the statics in HTMLEntityTable.
748
749         * html/HTMLEntityParser.cpp:
750         (WebCore::consumeHTMLEntity):
751         * html/HTMLEntitySearch.cpp:
752         (WebCore::HTMLEntitySearch::HTMLEntitySearch):
753         (WebCore::HTMLEntitySearch::findFirst):
754         (WebCore::HTMLEntitySearch::findLast):
755         (WebCore::HTMLEntitySearch::advance):
756         * html/HTMLEntitySearch.h:
757         (WebCore::HTMLEntitySearch::isEntityPrefix):
758         (WebCore::HTMLEntitySearch::mostRecentMatch):
759         (WebCore::HTMLEntitySearch::fail):
760         * html/HTMLEntityTable.h:
761
762 2010-08-13  Adam Barth  <abarth@webkit.org>
763
764         Fix ASSERT in XML entity parsing.  You can't advance the entity search
765         if you've already discovered that your string is not an entity.
766
767         * html/HTMLEntityParser.cpp:
768         (WebCore::decodeNamedEntity):
769
770 2010-08-13  Eric Seidel  <eric@webkit.org>
771
772         Unreviewed.  Attempt at fixing Chromium build.
773
774         Add support for MathML entities
775         https://bugs.webkit.org/show_bug.cgi?id=43949
776
777         * WebCore.gyp/WebCore.gyp:
778          - I have no clue if this is right or not.
779
780 2010-08-13  Eric Seidel  <eric@webkit.org>
781
782         Unreviewed.  Attempted build fix for Windows.
783
784         Add support for MathML entities
785         https://bugs.webkit.org/show_bug.cgi?id=43949
786
787         * DerivedSources.cpp:
788          - Add HTMLEntityTable.cpp
789
790 2010-08-09  Adam Barth  <abarth@webkit.org>
791
792         Reviewed by Eric Seidel.
793
794         Add support for MathML entities
795         https://bugs.webkit.org/show_bug.cgi?id=43949
796
797         Implementing the HTML5 entity parsing algorithm require refactoring how
798         we search for entity names.  Instead of using a perfect hash, we now
799         use a sorted list.  As we advance through the input, we walk down a
800         binary search of the table looking for an entity.
801
802         Using this data structure lets us keep track of whether the current
803         string is a prefix of an existing entity, which we need for the
804         algorithm.  In a future patch, I plan to add some indices to the
805         table, which should let us narrow down the range of interesting entries
806         more quickly.
807
808         The one nasty piece of the algorithm is if we walk too far down the
809         input and we need to back up to a previous match.  In this patch, we
810         accomplish this by rewinding the input and consuming a known number of
811         characters to resync the source.
812
813         * WebCore.xcodeproj/project.pbxproj:
814         * html/HTMLEntityParser.cpp:
815         (WebCore::consumeHTMLEntity):
816         * html/HTMLEntitySearch.cpp: Added.
817         (WebCore::):
818         (WebCore::HTMLEntitySearch::HTMLEntitySearch):
819         (WebCore::HTMLEntitySearch::compare):
820         (WebCore::HTMLEntitySearch::findStart):
821         (WebCore::HTMLEntitySearch::findEnd):
822         (WebCore::HTMLEntitySearch::advance):
823         * html/HTMLEntitySearch.h: Added.
824         (WebCore::HTMLEntitySearch::isEntityPrefix):
825         (WebCore::HTMLEntitySearch::currentValue):
826         (WebCore::HTMLEntitySearch::lastMatch):
827         (WebCore::HTMLEntitySearch::):
828         (WebCore::HTMLEntitySearch::fail):
829         * html/HTMLEntityTable.h: Added.
830         (WebCore::HTMLEntityTableEntry::lastCharacter):
831
832 2010-08-13  Tony Gentilcore  <tonyg@chromium.org>
833
834         Reviewed by Eric Seidel.
835
836         Refactor HTMLScriptRunner to allow deferred scripts to share code
837         https://bugs.webkit.org/show_bug.cgi?id=43736
838
839         No new tests because no functional change
840
841         * html/HTMLScriptRunner.cpp:
842         (WebCore::HTMLScriptRunner::sourceFromPendingScript): Make const to enforce the idea that it doesn't change m_parsingBlockingScript.
843         (WebCore::HTMLScriptRunner::executeParsingBlockingScript): Rename to make it clear this deals with the parsing blocking script and refactor to use executePendingScriptAndDispatchEvent.
844         (WebCore::HTMLScriptRunner::executePendingScriptAndDispatchEvent): Factored out of executePendingScript, this can be used for executing any external script.
845         (WebCore::HTMLScriptRunner::executeScript): This can be used for executing any inline or external script. ASSERTs moved to the two calling points.
846         (WebCore::HTMLScriptRunner::executeParsingBlockingScripts): Moved method.
847         (WebCore::HTMLScriptRunner::requestParsingBlockingScript): Factored out requestPendingScript.
848         (WebCore::HTMLScriptRunner::requestPendingScript): Factored out from reqeustParsingBlockingScript.
849         (WebCore::HTMLScriptRunner::runScript):
850         * html/HTMLScriptRunner.h:
851
852 2010-08-13  Gavin Barraclough  <barraclough@apple.com>
853
854         Windows build fix.
855
856         * WebCore.xcodeproj/project.pbxproj:
857         * platform/win/SharedBufferWin.cpp:
858
859 2010-08-13  Gavin Barraclough  <barraclough@apple.com>
860
861         Rubber stamped by Sam Weinig.
862         Switch String::/UString::ascii() to return a CString.
863
864         * WebCore.xcodeproj/project.pbxproj:
865         * bridge/c/c_class.cpp:
866         (JSC::Bindings::CClass::methodsNamed):
867         (JSC::Bindings::CClass::fieldNamed):
868         * bridge/jni/jsc/JNIBridgeJSC.cpp:
869         (JavaField::valueFromInstance):
870         (JavaField::setValueToInstance):
871         * bridge/jni/jsc/JavaInstanceJSC.cpp:
872         (JavaInstance::invokeMethod):
873         * bridge/objc/objc_class.mm:
874         (JSC::Bindings::ObjcClass::methodsNamed):
875         (JSC::Bindings::ObjcClass::fieldNamed):
876         * bridge/objc/objc_instance.mm:
877         (ObjcInstance::setValueOfUndefinedField):
878         (ObjcInstance::getValueOfUndefinedField):
879         * loader/icon/IconRecord.cpp:
880         * platform/sql/SQLiteDatabase.cpp:
881         * platform/sql/SQLiteStatement.cpp:
882         * storage/SQLStatement.cpp:
883
884 2010-08-13  Mihai Parparita  <mihaip@chromium.org>
885
886         Reviewed by Dimitri Glazkov.
887
888         Session history should skip over JS redirects
889         https://bugs.webkit.org/show_bug.cgi?id=42861
890         
891         Lock the back/forward list for location changes and form submits that
892         happen before the onload event fires that are not the result of user
893         gestures.
894         
895         Made form submission (at the ScheduledFormSubmission level) more similar
896         to ScheduledURLNavigation by having it call clientRedirected too, fixing
897         a long-standing FIXME.
898
899         Test: fast/history/gesture-before-onload-location-href.html,
900         fast/history/gesture-before-onload-form-submit.html and updated
901         expectations for http/tests/history tests that used to fail.
902
903         * loader/FormSubmission.cpp:
904         (WebCore::FormSubmission::requestURL): 
905         (WebCore::FormSubmission::populateFrameLoadRequest):
906         * loader/FormSubmission.h:
907         * loader/RedirectScheduler.cpp:
908         (WebCore::ScheduledFormSubmission::ScheduledFormSubmission):
909         (WebCore::ScheduledFormSubmission::fire):
910         (WebCore::ScheduledFormSubmission::didStartTimer):
911         (WebCore::ScheduledFormSubmission::didStopTimer):
912         (WebCore::RedirectScheduler::scheduleRedirect):
913         (WebCore::RedirectScheduler::mustLockBackForwardList):
914         (WebCore::RedirectScheduler::scheduleLocationChange):
915         (WebCore::RedirectScheduler::scheduleFormSubmission):
916         * loader/RedirectScheduler.h:
917
918 2010-08-13  Leandro Pereira  <leandro@profusion.mobi>
919
920         [EFL] Unreviewed build fix.
921         These changes were supposed to be landed on r65319, but due to some
922         svn-apply/commit-queue bug it wasn't fully applied. A bug was filed
923         by Patrick Gansterer (author of the changes in this commit):
924         https://bugs.webkit.org/show_bug.cgi?id=43981
925
926         * CMakeLists.txt: Update calls to GENERATE_GPERF and
927         GENERATE_GRAMMAR macros.
928
929 2010-08-13  Chris Fleizach  <cfleizach@apple.com>
930
931         AX: isNativeCheckbox does not work as advertised
932         https://bugs.webkit.org/show_bug.cgi?id=43872
933
934         Re-adding this patch after it was erroneously rolled out.
935
936         Test: platform/mac/accessibility/native-vs-nonnative-checkboxes.html
937
938         * accessibility/AccessibilityRenderObject.cpp:
939         (WebCore::AccessibilityRenderObject::isInputImage):
940         (WebCore::AccessibilityRenderObject::isNativeCheckboxOrRadio):
941         * dom/InputElement.h:
942         * html/HTMLInputElement.h:
943         (WebCore::HTMLInputElement::isCheckbox):
944
945 2010-08-13  Sheriff Bot  <webkit.review.bot@gmail.com>
946
947         Unreviewed, rolling out r65331.
948         http://trac.webkit.org/changeset/65331
949         https://bugs.webkit.org/show_bug.cgi?id=43980
950
951         This patch breaks some shadow tests on Qt Linux release
952         (Requested by ariya on #webkit).
953
954         * platform/graphics/qt/GraphicsContextQt.cpp:
955         (WebCore::GraphicsContextPlatformPrivate::GraphicsContextPlatformPrivate):
956         (WebCore::GraphicsContext::drawRect):
957         (WebCore::GraphicsContext::drawLine):
958         (WebCore::GraphicsContext::strokeArc):
959         (WebCore::GraphicsContext::drawConvexPolygon):
960         (WebCore::drawFilledShadowPath):
961         (WebCore::GraphicsContext::fillPath):
962         (WebCore::GraphicsContext::strokePath):
963         (WebCore::drawBorderlessRectShadow):
964         (WebCore::GraphicsContext::fillRect):
965         (WebCore::GraphicsContext::fillRoundedRect):
966         (WebCore::GraphicsContext::setPlatformShadow):
967         (WebCore::GraphicsContext::clearPlatformShadow):
968
969 2010-08-13  Kinuko Yasuda  <kinuko@chromium.org>
970
971         Reviewed by David Levin.
972
973         Update run-bindings-tests expectations
974         https://bugs.webkit.org/show_bug.cgi?id=43944
975
976         The results have become out of sync after changeset 64552 and 65269.
977
978         Test: WebKitTools/Scripts/run-bindings-tests
979
980         * bindings/scripts/test/GObject/WebKitDOMTestCallback.cpp:
981         (webkit_dom_test_callback_callback_with_class1param):
982         (webkit_dom_test_callback_callback_with_class2param):
983         (webkit_dom_test_callback_callback_with_non_bool_return_type):
984         * bindings/scripts/test/GObject/WebKitDOMTestCallback.h:
985         * bindings/scripts/test/GObject/WebKitDOMTestObj.cpp:
986         (webkit_dom_test_obj_void_method_with_args):
987         (webkit_dom_test_obj_int_method_with_args):
988         (webkit_dom_test_obj_obj_method):
989         (webkit_dom_test_obj_obj_method_with_args):
990         (webkit_dom_test_obj_method_that_requires_all_args):
991         (webkit_dom_test_obj_method_that_requires_all_args_and_throws):
992         (webkit_dom_test_obj_serialized_value):
993         (webkit_dom_test_obj_idb_key):
994         (webkit_dom_test_obj_with_dynamic_frame_and_user_gesture):
995         (webkit_dom_test_obj_with_dynamic_frame_and_user_gesture_asad):
996         (webkit_dom_test_obj_with_script_state_obj):
997         (webkit_dom_test_obj_with_script_state_obj_exception):
998         (webkit_dom_test_obj_get_read_only_string_attr):
999         (webkit_dom_test_obj_get_read_only_test_obj_attr):
1000         (webkit_dom_test_obj_get_string_attr):
1001         (webkit_dom_test_obj_set_string_attr):
1002         (webkit_dom_test_obj_get_test_obj_attr):
1003         (webkit_dom_test_obj_set_test_obj_attr):
1004         (webkit_dom_test_obj_get_xml_obj_attr):
1005         (webkit_dom_test_obj_set_xml_obj_attr):
1006         (webkit_dom_test_obj_get_reflected_string_attr):
1007         (webkit_dom_test_obj_set_reflected_string_attr):
1008         (webkit_dom_test_obj_get_reflected_url_attr):
1009         (webkit_dom_test_obj_set_reflected_url_attr):
1010         (webkit_dom_test_obj_get_reflected_non_empty_url_attr):
1011         (webkit_dom_test_obj_set_reflected_non_empty_url_attr):
1012         (webkit_dom_test_obj_get_reflected_custom_url_attr):
1013         (webkit_dom_test_obj_set_reflected_custom_url_attr):
1014         (webkit_dom_test_obj_get_reflected_custom_non_empty_url_attr):
1015         (webkit_dom_test_obj_set_reflected_custom_non_empty_url_attr):
1016         (webkit_dom_test_obj_get_string_attr_with_getter_exception):
1017         (webkit_dom_test_obj_set_string_attr_with_getter_exception):
1018         (webkit_dom_test_obj_get_string_attr_with_setter_exception):
1019         (webkit_dom_test_obj_set_string_attr_with_setter_exception):
1020         (webkit_dom_test_obj_get_script_string_attr):
1021         (webkit_dom_test_obj_get_hash):
1022         (webkit_dom_test_obj_class_init):
1023         * bindings/scripts/test/GObject/WebKitDOMTestObj.h:
1024         * bindings/scripts/test/JS/JSTestObj.cpp:
1025         (WebCore::jsTestObjPrototypeFunctionOverloadedMethod):
1026
1027 2010-08-13  Ariya Hidayat  <ariya@sencha.com>
1028
1029         Reviewed by Simon Hausmann.
1030
1031         [Qt] Canvas and CSS: blur option in shadow not working
1032         https://bugs.webkit.org/show_bug.cgi?id=34479
1033
1034         Patch 1: Refactor shadow states handling.
1035
1036         * platform/graphics/qt/GraphicsContextQt.cpp:
1037         (WebCore::GraphicsContextPlatformPrivate::): Added shadow states as
1038         member variables.
1039         (WebCore::GraphicsContextPlatformPrivate::hasShadow): Convenient
1040         function to check whether there is shadow or not.
1041         (WebCore::GraphicsContextPlatformPrivate::GraphicsContextPlatformPrivate):
1042         (WebCore::GraphicsContext::drawRect): Use shadow states instead of
1043         calling getShadow.
1044         (WebCore::GraphicsContext::drawLine): ditto.
1045         (WebCore::GraphicsContext::strokeArc): ditto.
1046         (WebCore::GraphicsContext::drawConvexPolygon): ditto.
1047         (WebCore::GraphicsContext::fillPath): ditto.
1048         (WebCore::GraphicsContext::strokePath): ditto.
1049         (WebCore::GraphicsContext::fillRect): Removes the use of helper function
1050         drawBorderlessRectShadow as the code already becomes a lot simpler.
1051         (WebCore::GraphicsContext::fillRoundedRect): Removes the use of helper
1052         function drawFilledShadowPath as the code already becomes a lot simpler.
1053         (WebCore::GraphicsContext::setPlatformShadow): Store shadow states
1054         and find out the shadow type (complexity) for future use.
1055         (WebCore::GraphicsContext::clearPlatformShadow): Reset shadow states.
1056
1057 2010-08-11  Zhenyao Mo  <zmo@google.com>
1058
1059         Reviewed by Dimitri Glazkov.
1060
1061         Regression in linking of programs
1062         https://bugs.webkit.org/show_bug.cgi?id=43820
1063
1064         Shaders and programs shouldn't be deleted until their attachments are removed, therefore, we need to track the attach/detach/useProgram.
1065
1066         * html/canvas/WebGLObject.cpp: Track the object attachment count and whether it should really be deleted.
1067         (WebCore::WebGLObject::WebGLObject):
1068         (WebCore::WebGLObject::setObject):
1069         (WebCore::WebGLObject::deleteObject):
1070         * html/canvas/WebGLObject.h: Track the object attachment count and whether it should really be deleted.
1071         (WebCore::WebGLObject::onAttached):
1072         (WebCore::WebGLObject::onDetached):
1073         (WebCore::WebGLObject::getAttachmentCount):
1074         * html/canvas/WebGLProgram.cpp: Track the attached shaders.
1075         (WebCore::WebGLProgram::WebGLProgram):
1076         (WebCore::WebGLProgram::deleteObjectImpl):
1077         (WebCore::WebGLProgram::getAttachedShader):
1078         (WebCore::WebGLProgram::attachShader):
1079         (WebCore::WebGLProgram::detachShader):
1080         * html/canvas/WebGLProgram.h: Track the attached shaders.
1081         * html/canvas/WebGLRenderingContext.cpp:
1082         (WebCore::WebGLRenderingContext::attachShader): Track the attachment of a shader to a program.
1083         (WebCore::WebGLRenderingContext::deleteProgram): Detach shaders.
1084         (WebCore::WebGLRenderingContext::detachShader): Track the attachment of a shader to a program.
1085         (WebCore::WebGLRenderingContext::validateWebGLObject): Also check if object == 0.
1086         (WebCore::WebGLRenderingContext::linkProgram): Using the cached attached shaders instead of query from driver.
1087         (WebCore::WebGLRenderingContext::useProgram): Track the attachment of a program to the current rendering pipeline.
1088
1089 2010-08-13  Steve Block  <steveblock@google.com>
1090
1091         Reviewed by Alexey Proskuryakov.
1092
1093         Geolocation activity started after frame has been disconnected can cause crash
1094         https://bugs.webkit.org/show_bug.cgi?id=39879
1095
1096         New requests started after the Frame has been disconnected are ignored. We do
1097         not invoke the error callback as this would allow buggy or malicious pages to
1098         hose the CPU. Such a page could hold a reference to a Geolocation object from
1099         a since closed Page and register new requests from the error callback to
1100         create an infinite loop.
1101
1102         Tests: fast/dom/Geolocation/disconnected-frame-already.html
1103
1104         * page/Geolocation.cpp:
1105
1106 2010-08-13  Satish Sampath  <satish@chromium.org>
1107
1108         Reviewed by Jeremy Orlow.
1109
1110         Support for multiple speech enabled elements in same page.
1111         https://bugs.webkit.org/show_bug.cgi?id=43922
1112
1113         * page/SpeechInput.cpp: Generates request ids as necessary when each speech enabled input element gets
1114         created/destroyed and multiplexes the listener callbacks to the appropriate input element based on the request id.
1115         (WebCore::SpeechInput::SpeechInput):
1116         (WebCore::SpeechInput::~SpeechInput):
1117         (WebCore::SpeechInput::registerListener):
1118         (WebCore::SpeechInput::unregisterListener):
1119         (WebCore::SpeechInput::didCompleteRecording):
1120         (WebCore::SpeechInput::didCompleteRecognition):
1121         (WebCore::SpeechInput::setRecognitionResult):
1122         (WebCore::SpeechInput::startRecognition):
1123         (WebCore::SpeechInput::stopRecording):
1124         (WebCore::SpeechInput::cancelRecognition):
1125         * page/SpeechInput.h:
1126         * page/SpeechInputClient.h: Now requires a one time set for the WebCore::SpeechInputListener and
1127         takes in a 'requestId' for all calls.
1128         * page/SpeechInputListener.h: Now returns the above mentioned requestId in all the listener callbacks for
1129         identifying which input element the event goes to.
1130         * platform/mock/SpeechInputClientMock.cpp:
1131         (WebCore::SpeechInputClientMock::SpeechInputClientMock):
1132         (WebCore::SpeechInputClientMock::setListener):
1133         (WebCore::SpeechInputClientMock::startRecognition):
1134         (WebCore::SpeechInputClientMock::stopRecording):
1135         (WebCore::SpeechInputClientMock::cancelRecognition):
1136         (WebCore::SpeechInputClientMock::timerFired):
1137         * platform/mock/SpeechInputClientMock.h:
1138         * rendering/TextControlInnerElements.cpp:
1139         (WebCore::InputFieldSpeechButtonElement::InputFieldSpeechButtonElement):
1140         (WebCore::InputFieldSpeechButtonElement::~InputFieldSpeechButtonElement):
1141         (WebCore::InputFieldSpeechButtonElement::defaultEventHandler):
1142         (WebCore::InputFieldSpeechButtonElement::speechInput):
1143         (WebCore::InputFieldSpeechButtonElement::didCompleteRecording):
1144         (WebCore::InputFieldSpeechButtonElement::didCompleteRecognition):
1145         (WebCore::InputFieldSpeechButtonElement::setRecognitionResult):
1146         * rendering/TextControlInnerElements.h:
1147
1148 2010-08-13  Eric Carlson  <eric.carlson@apple.com>
1149
1150         Reviewed by David Kilzer.
1151
1152         HTMLMediaElement::prepareForLoad should cancel current plug-in proxy media player
1153         https://bugs.webkit.org/show_bug.cgi?id=43923
1154
1155         * html/HTMLMediaElement.cpp:
1156         (WebCore::HTMLMediaElement::prepareForLoad): Tell mediaPlayer to cancel the current load
1157         if it already exists.
1158
1159 2010-08-12  Stephen White  <senorblanco@chromium.org>
1160
1161         Reviewed by David Levin.
1162
1163         [CHROMIUM] Use the BGRA format for canvas 2D accel upload and readbacks.
1164         https://bugs.webkit.org/show_bug.cgi?id=43804
1165
1166         Use the BGRA format from GraphicsContext3D, if supported.  Also keep a
1167         texture around for uploads, rather than re-creating it each time.
1168
1169         Covered by many layout tests (once we're running them).
1170
1171         * platform/graphics/chromium/GLES2Texture.cpp:
1172         (WebCore::convertFormat):
1173         Move convertFormat() ahead of texture creation, so we can check for
1174         BGRA support.
1175         (WebCore::GLES2Texture::create):
1176         Use convertFormat() to determine the correct format and types to use;
1177         use the returned format also for internalFormat, since GLES2 insists
1178         they match.
1179         * platform/graphics/skia/PlatformContextSkia.cpp:
1180         (WebCore::PlatformContextSkia::setGraphicsContext3D):
1181         Clear the upload texture when a new context is set.
1182         (WebCore::PlatformContextSkia::uploadSoftwareToHardware):
1183         Use m_uploadTexture instead of creating a new one each time.
1184         (WebCore::PlatformContextSkia::readbackHardwareToSoftware):
1185         When the context supports it, use the BGRA format for readbacks
1186         instead of swizzling.
1187         * platform/graphics/skia/PlatformContextSkia.h:
1188         Add m_uploadTexture for consecutive uploads.
1189
1190 2010-08-13  Simon Hausmann  <simon.hausmann@nokia.com>
1191
1192         Reviewed by Ariya Hidayat.
1193
1194         [Qt] Clean up mobile feature useage
1195         https://bugs.webkit.org/show_bug.cgi?id=43968
1196
1197         * WebCore.pro: Replaced use of Nokia specific platform scopes with concrete
1198         feature names.
1199
1200 2010-08-13  Pavel Feldman  <pfeldman@chromium.org>
1201
1202         Reviewed by Yury Semikhatsky.
1203
1204         Web Inspector: rename InspectorValue::push and set to
1205         push{Value|Object|Array} and set{...}.
1206         https://bugs.webkit.org/show_bug.cgi?id=43966
1207
1208         * bindings/js/ScriptProfile.cpp:
1209         (WebCore::buildInspectorObjectFor):
1210         * bindings/js/ScriptValue.cpp:
1211         (WebCore::jsToInspectorValue):
1212         * bindings/v8/ScriptCallStack.cpp:
1213         (WebCore::ScriptCallStack::stackTrace):
1214         * bindings/v8/ScriptProfile.cpp:
1215         (WebCore::buildInspectorObjectFor):
1216         * bindings/v8/ScriptValue.cpp:
1217         (WebCore::v8ToInspectorValue):
1218         * inspector/CodeGeneratorInspector.pm:
1219         * inspector/ConsoleMessage.cpp:
1220         (WebCore::ConsoleMessage::addToFrontend):
1221         * inspector/Inspector.idl:
1222         * inspector/InspectorApplicationCacheAgent.cpp:
1223         (WebCore::InspectorApplicationCacheAgent::buildObjectForApplicationCache):
1224         (WebCore::InspectorApplicationCacheAgent::buildArrayForApplicationCacheResources):
1225         * inspector/InspectorController.cpp:
1226         (WebCore::InspectorController::buildArrayForCookies):
1227         (WebCore::InspectorController::getDOMStorageEntries):
1228         (WebCore::InspectorController::getProfileHeaders):
1229         (WebCore::InspectorController::getProfile):
1230         * inspector/InspectorDOMAgent.cpp:
1231         (WebCore::InspectorDOMAgent::getEventListenersForNode):
1232         (WebCore::InspectorDOMAgent::buildObjectForNode):
1233         (WebCore::InspectorDOMAgent::buildArrayForContainerChildren):
1234         (WebCore::InspectorDOMAgent::getStyles):
1235         (WebCore::InspectorDOMAgent::getAllStyles):
1236         (WebCore::InspectorDOMAgent::getRuleRanges):
1237         (WebCore::InspectorDOMAgent::buildObjectForAttributeStyles):
1238         (WebCore::InspectorDOMAgent::buildArrayForCSSRules):
1239         (WebCore::InspectorDOMAgent::buildArrayForPseudoElements):
1240         (WebCore::InspectorDOMAgent::buildObjectForStyle):
1241         (WebCore::InspectorDOMAgent::populateObjectWithStyleProperties):
1242         (WebCore::InspectorDOMAgent::buildArrayForDisabledStyleProperties):
1243         (WebCore::InspectorDOMAgent::buildObjectForStyleSheet):
1244         (WebCore::InspectorDOMAgent::buildObjectForRule):
1245         * inspector/InspectorDebuggerAgent.cpp:
1246         (WebCore::InspectorDebuggerAgent::saveBreakpoints):
1247         * inspector/InspectorResource.cpp:
1248         (WebCore::InspectorResource::updateScriptObject):
1249         * inspector/InspectorStorageAgent.cpp:
1250         (WebCore::):
1251         (WebCore::InspectorStorageAgent::executeSQL):
1252         * inspector/InspectorTimelineAgent.cpp:
1253         (WebCore::InspectorTimelineAgent::pushGCEventRecords):
1254         (WebCore::InspectorTimelineAgent::didInstallTimer):
1255         (WebCore::InspectorTimelineAgent::didRemoveTimer):
1256         (WebCore::InspectorTimelineAgent::didScheduleResourceRequest):
1257         (WebCore::InspectorTimelineAgent::willSendResourceRequest):
1258         (WebCore::InspectorTimelineAgent::didFinishLoadingResource):
1259         (WebCore::InspectorTimelineAgent::didMarkTimeline):
1260         (WebCore::InspectorTimelineAgent::addRecordToTimeline):
1261         (WebCore::InspectorTimelineAgent::didCompleteCurrentRecord):
1262         * inspector/InspectorValues.cpp:
1263         (WebCore::):
1264         (WebCore::InspectorValue::asValue):
1265         (WebCore::InspectorValue::asObject):
1266         (WebCore::InspectorValue::asArray):
1267         (WebCore::InspectorObject::asObject):
1268         (WebCore::InspectorArray::asArray):
1269         * inspector/InspectorValues.h:
1270         (WebCore::InspectorObject::setBool):
1271         (WebCore::InspectorObject::setNumber):
1272         (WebCore::InspectorObject::setString):
1273         (WebCore::InspectorObject::setValue):
1274         (WebCore::InspectorObject::setObject):
1275         (WebCore::InspectorObject::setArray):
1276         (WebCore::InspectorArray::pushValue):
1277         (WebCore::InspectorArray::pushObject):
1278         (WebCore::InspectorArray::pushArray):
1279         * inspector/ScriptBreakpoint.cpp:
1280         (WebCore::ScriptBreakpoint::inspectorObjectFromSourceBreakpoints):
1281         * inspector/TimelineRecordFactory.cpp:
1282         (WebCore::TimelineRecordFactory::createGenericRecord):
1283
1284 2010-08-13  Sheriff Bot  <webkit.review.bot@gmail.com>
1285
1286         Unreviewed, rolling out r65204.
1287         http://trac.webkit.org/changeset/65204
1288         https://bugs.webkit.org/show_bug.cgi?id=43965
1289
1290         Caused a lot of test crashes in Chromium (Requested by yutak
1291         on #webkit).
1292
1293         * accessibility/AccessibilityRenderObject.cpp:
1294         (WebCore::AccessibilityRenderObject::isInputImage):
1295         (WebCore::AccessibilityRenderObject::isNativeCheckboxOrRadio):
1296         * dom/InputElement.h:
1297         * html/HTMLInputElement.h:
1298
1299 2010-08-13  Anton Muhin  <antonm@chromium.org>
1300
1301         Reviewed by Nate Chapin.
1302
1303         Reapplying http://trac.webkit.org/changeset/65284 after fixed include.
1304
1305         [v8] Create callbacks for named access of elements from document object on fly.
1306         https://bugs.webkit.org/show_bug.cgi?id=43725
1307
1308         That should speed up access to methods on document as there is no need to
1309         consult with interceptor before invoking the method.
1310
1311         This reapplies http://trac.webkit.org/changeset/60670 reverted with
1312         http://trac.webkit.org/changeset/63845.  The main difference is in treatment
1313         of removed named elements: we don't remove an accessor any more, but leave it.
1314         Therefore any accessor should be able to treat the case when there is no
1315         expando property.
1316
1317         * bindings/scripts/CodeGeneratorV8.pm:
1318         * bindings/v8/ScriptController.cpp:
1319         (WebCore::ScriptController::namedItemAdded):
1320         (WebCore::ScriptController::namedItemRemoved):
1321         * bindings/v8/V8DOMWindowShell.cpp:
1322         (WebCore::checkDocumentWrapper):
1323         (WebCore::V8DOMWindowShell::updateDocumentWrapperCache):
1324         (WebCore::getter):
1325         (WebCore::V8DOMWindowShell::namedItemAdded):
1326         (WebCore::V8DOMWindowShell::namedItemRemoved):
1327         * bindings/v8/V8DOMWindowShell.h:
1328         * bindings/v8/V8DOMWrapper.cpp:
1329         (WebCore::V8DOMWrapper::instantiateV8Object):
1330         * bindings/v8/custom/V8HTMLDocumentCustom.cpp:
1331         (WebCore::V8HTMLDocument::WrapInShadowObject):
1332         (WebCore::V8HTMLDocument::GetNamedProperty):
1333         (WebCore::V8HTMLDocument::allAccessorSetter):
1334         (WebCore::toV8):
1335         * html/HTMLDocument.cpp:
1336         (WebCore::HTMLDocument::removeItemFromMap):
1337
1338 2010-08-13  Pavel Feldman  <pfeldman@chromium.org>
1339
1340         Reviewed by Yury Semikhatsky.
1341
1342         Web Inspector: remove openInInspectorWindow and evaluateAndStringify from InjectedScript.
1343         https://bugs.webkit.org/show_bug.cgi?id=43854
1344
1345         * inspector/Inspector.idl:
1346         * inspector/InspectorController.cpp:
1347         (WebCore::InspectorController::openInInspectedWindow):
1348         * inspector/InspectorController.h:
1349         * inspector/front-end/ExtensionServer.js:
1350         (WebInspector.ExtensionServer.prototype._onEvaluateOnInspectedPage.callback):
1351         (WebInspector.ExtensionServer.prototype._onEvaluateOnInspectedPage):
1352         * inspector/front-end/InjectedScript.js:
1353         (injectedScriptConstructor):
1354         * inspector/front-end/InjectedScriptAccess.js:
1355         * inspector/front-end/InspectorBackendStub.js:
1356         (WebInspector.InspectorBackendStub):
1357         * inspector/front-end/ResourcesPanel.js:
1358         (WebInspector.ResourceSidebarTreeElement.prototype.ondblclick):
1359
1360 2010-08-13  Fumitoshi Ukai  <ukai@chromium.org>
1361
1362         Reviewed by Alexey Proskuryakov.
1363
1364         flaky websocket/tests/frame-length-overflow.html
1365         https://bugs.webkit.org/show_bug.cgi?id=43902
1366
1367         Add m_shouldDiscardReceivedData flag to indicate it will no longer
1368         receive data from network.
1369
1370         * websockets/WebSocketChannel.cpp:
1371         (WebCore::WebSocketChannel::WebSocketChannel):
1372         (WebCore::WebSocketChannel::didReceiveData):
1373         (WebCore::WebSocketChannel::didFail):
1374         (WebCore::WebSocketChannel::processBuffer):
1375         - when frame length overflows, we couldn't process data any more.
1376           clear buffer and mark m_shouldDiscardReceivedData true to make sure
1377           it doesn't process the same buffer again.
1378         * websockets/WebSocketChannel.h:
1379
1380 2010-08-13  Dirk Schulze  <krit@webkit.org>
1381
1382         Reviewed by Nikolas Zimmermann.
1383
1384         Use SVGPathByteStream to animate SVGPath
1385         https://bugs.webkit.org/show_bug.cgi?id=43929
1386
1387         Use SVGPathByteStream to perform animations of Paths. SVGPathBlender blends
1388         the starting point to the end point according to the current progress value.
1389         Cleanup SVGPathSegList, by removing the unnecessary legacy blending code.
1390
1391         No new tests because no functional changes.
1392
1393         * Android.mk:
1394         * CMakeLists.txt:
1395         * GNUmakefile.am:
1396         * WebCore.gypi:
1397         * WebCore.pro:
1398         * WebCore.vcproj/WebCore.vcproj:
1399         * WebCore.xcodeproj/project.pbxproj:
1400         * svg/SVGAllInOne.cpp:
1401         * svg/SVGAnimateElement.cpp:
1402         (WebCore::SVGAnimateElement::SVGAnimateElement):
1403         (WebCore::SVGAnimateElement::calculateAnimatedValue):
1404         (WebCore::SVGAnimateElement::calculateFromAndToValues):
1405         (WebCore::SVGAnimateElement::resetToBaseValue):
1406         (WebCore::SVGAnimateElement::applyResultsToTarget):
1407         * svg/SVGAnimateElement.h:
1408         * svg/SVGPathBlender.cpp: Added.
1409         (WebCore::SVGPathBlender::SVGPathBlender):
1410         (WebCore::SVGPathBlender::~SVGPathBlender):
1411         (WebCore::SVGPathBlender::blendAnimatedFloat):
1412         (WebCore::SVGPathBlender::blendAnimatedFloatPoint):
1413         (WebCore::SVGPathBlender::blendMoveToSegment):
1414         (WebCore::SVGPathBlender::blendLineToSegment):
1415         (WebCore::SVGPathBlender::blendLineToHorizontalSegment):
1416         (WebCore::SVGPathBlender::blendLineToVerticalSegment):
1417         (WebCore::SVGPathBlender::blendCurveToCubicSegment):
1418         (WebCore::SVGPathBlender::blendCurveToCubicSmoothSegment):
1419         (WebCore::SVGPathBlender::blendCurveToQuadraticSegment):
1420         (WebCore::SVGPathBlender::blendCurveToQuadraticSmoothSegment):
1421         (WebCore::SVGPathBlender::blendArcToSegment):
1422         (WebCore::SVGPathBlender::blendAnimatedPath):
1423         (WebCore::SVGPathBlender::cleanup):
1424         * svg/SVGPathBlender.h: Added.
1425         * svg/SVGPathParserFactory.cpp:
1426         (WebCore::globalSVGPathBlender):
1427         (WebCore::SVGPathParserFactory::buildAnimatedSVGPathByteStream):
1428         * svg/SVGPathParserFactory.h:
1429         * svg/SVGPathSegList.cpp:
1430         (WebCore::SVGPathSegList::getPathSegAtLength):
1431         * svg/SVGPathSegList.h:
1432         (WebCore::SVGPathSegList::create):
1433         * svg/SVGPointList.cpp:
1434         (WebCore::adjustAnimatedValue):
1435
1436 2010-08-12  Nikolas Zimmermann  <nzimmermann@rim.com>
1437
1438         Reviewed by Dirk Schulze.
1439
1440         Window size changes + resources on absolute sized content leads to pixelation
1441         https://bugs.webkit.org/show_bug.cgi?id=43921
1442
1443         Avoid pixelation when using resources on absolute sized objects. Be sure to invalidate the used resources if window size changes.
1444         Otherwhise masks etc. will be scaled up and pixelate. As nice side-effect, this fixes run-webkit-tests --singly svg/W3C-SVG-1.1, which
1445         used to produce wrong results for the gradient tests, as the window size change from 800x600 for 480x360 (svg/W3C-SVG-1.1 specific DRT quirk)
1446         didn't invalidate the resources.
1447
1448         Renamed invalidateClient -> removeClientFromCache and invalidateClients -> removeAllClientsFromCache.
1449         Add a boolean "markForInvalidation" parameter, that controls wheter the clients should be marked for repaint/boundaries invalidation/layout
1450         or wheter we only want to notify the parents that the resource changed, to avoid doing unncessary work in SVGRenderSupport::layoutChildren.
1451
1452         Test: svg/custom/absolute-sized-content-with-resources.xhtml
1453
1454         * rendering/RenderSVGGradientStop.cpp:
1455         (WebCore::RenderSVGGradientStop::styleDidChange): Don't do anything for StyleDifferenceEqual.
1456         * rendering/RenderSVGImage.cpp:
1457         (WebCore::RenderSVGImage::imageChanged):
1458         * rendering/RenderSVGResource.cpp:
1459         (WebCore::RenderSVGResource::markForLayoutAndParentResourceInvalidation):
1460         * rendering/RenderSVGResource.h:
1461         * rendering/RenderSVGResourceClipper.cpp:
1462         (WebCore::RenderSVGResourceClipper::removeAllClientsFromCache):
1463         (WebCore::RenderSVGResourceClipper::removeClientFromCache):
1464         * rendering/RenderSVGResourceClipper.h:
1465         * rendering/RenderSVGResourceContainer.cpp:
1466         (WebCore::RenderSVGResourceContainer::layout):
1467         (WebCore::RenderSVGResourceContainer::idChanged):
1468         (WebCore::RenderSVGResourceContainer::markAllClientsForInvalidation):
1469         (WebCore::RenderSVGResourceContainer::markClientForInvalidation):
1470         * rendering/RenderSVGResourceContainer.h:
1471         (WebCore::RenderSVGResourceContainer::):
1472         * rendering/RenderSVGResourceFilter.cpp:
1473         (WebCore::RenderSVGResourceFilter::removeAllClientsFromCache):
1474         (WebCore::RenderSVGResourceFilter::removeClientFromCache):
1475         (WebCore::RenderSVGResourceFilter::postApplyResource):
1476         * rendering/RenderSVGResourceFilter.h:
1477         * rendering/RenderSVGResourceGradient.cpp:
1478         (WebCore::RenderSVGResourceGradient::removeAllClientsFromCache):
1479         (WebCore::RenderSVGResourceGradient::removeClientFromCache):
1480         (WebCore::RenderSVGResourceGradient::applyResource):
1481         * rendering/RenderSVGResourceGradient.h:
1482         * rendering/RenderSVGResourceMarker.cpp:
1483         (WebCore::RenderSVGResourceMarker::layout):
1484         (WebCore::RenderSVGResourceMarker::removeAllClientsFromCache):
1485         (WebCore::RenderSVGResourceMarker::removeClientFromCache):
1486         * rendering/RenderSVGResourceMarker.h:
1487         * rendering/RenderSVGResourceMasker.cpp:
1488         (WebCore::RenderSVGResourceMasker::removeAllClientsFromCache):
1489         (WebCore::RenderSVGResourceMasker::removeClientFromCache):
1490         * rendering/RenderSVGResourceMasker.h:
1491         * rendering/RenderSVGResourcePattern.cpp:
1492         (WebCore::RenderSVGResourcePattern::removeAllClientsFromCache):
1493         (WebCore::RenderSVGResourcePattern::removeClientFromCache):
1494         (WebCore::RenderSVGResourcePattern::applyResource):
1495         * rendering/RenderSVGResourcePattern.h:
1496         * rendering/RenderSVGResourceSolidColor.h:
1497         (WebCore::RenderSVGResourceSolidColor::removeAllClientsFromCache):
1498         (WebCore::RenderSVGResourceSolidColor::removeClientFromCache):
1499         * rendering/SVGRenderSupport.cpp:
1500         (WebCore::invalidateResourcesOfChildren): Add new helper funtion recursively invalidating resources of a sub render tree.
1501         (WebCore::SVGRenderSupport::layoutChildren): Keep track of all objects that didn't need layout, as we still have to invalidate their resources, if window size changes.
1502         * rendering/SVGResources.cpp:
1503         (WebCore::SVGResources::removeClientFromCache):
1504         (WebCore::SVGResources::resourceDestroyed):
1505         * rendering/SVGResources.h:
1506         * rendering/SVGResourcesCache.cpp:
1507         (WebCore::SVGResourcesCache::clientLayoutChanged):
1508
1509 2010-08-13  Pavel Feldman  <pfeldman@chromium.org>
1510
1511         Reviewed by Yury Semikhatsky.
1512
1513         Chromium DevTools: [CRASH] Autocomplete on arrays causes crashes.
1514         https://bugs.webkit.org/show_bug.cgi?id=43955
1515
1516         * bindings/v8/ScriptValue.cpp:
1517         (WebCore::v8ToInspectorValue):
1518
1519 2010-08-13  Gavin Barraclough  <barraclough@apple.com>
1520
1521         Reviewed by Sam Weinig
1522
1523         Unify UString::UTF8String() & String::utf8() methods,
1524         remove UString::cost() & make atArrayIndex a free function.
1525
1526         * WebCore.xcodeproj/project.pbxproj:
1527         * bindings/js/JSDOMWindowCustom.cpp:
1528         (WebCore::JSDOMWindow::getOwnPropertySlot):
1529         (WebCore::JSDOMWindow::getOwnPropertyDescriptor):
1530         * bridge/NP_jsobject.cpp:
1531         (_NPN_Enumerate):
1532         * bridge/c/c_utility.cpp:
1533         (JSC::Bindings::convertValueToNPVariant):
1534         * bridge/jni/JNIBridge.cpp:
1535         (JavaParameter::JavaParameter):
1536         (JavaMethod::JavaMethod):
1537         (JavaMethod::signature):
1538         (JavaMethod::methodID):
1539         * bridge/jni/JNIBridge.h:
1540         (JSC::Bindings::JavaString::utf8):
1541         (JSC::Bindings::JavaParameter::type):
1542         (JSC::Bindings::JavaMethod::returnType):
1543         * bridge/jni/jni_jsobject.mm:
1544         (JavaJSObject::call):
1545         (JavaJSObject::eval):
1546         (JavaJSObject::getMember):
1547         (JavaJSObject::setMember):
1548         (JavaJSObject::removeMember):
1549         (JavaJSObject::convertJObjectToValue):
1550         * bridge/jni/jsc/JNIBridgeJSC.cpp:
1551         (JavaField::JavaField):
1552         (JavaField::valueFromInstance):
1553         (JavaField::setValueToInstance):
1554         * bridge/jni/jsc/JNIBridgeJSC.h:
1555         (JSC::Bindings::JavaField::type):
1556         * bridge/jni/jsc/JavaInstanceJSC.cpp:
1557         (JavaInstance::invokeMethod):
1558         * bridge/jni/jsc/JavaStringJSC.h:
1559         (JSC::Bindings::JavaStringImpl::utf8):
1560         * bridge/runtime_array.cpp:
1561         (JSC::RuntimeArray::getOwnPropertySlot):
1562         (JSC::RuntimeArray::getOwnPropertyDescriptor):
1563         (JSC::RuntimeArray::put):
1564
1565 2010-08-12  Yury Semikhatsky  <yurys@chromium.org>
1566
1567         Reviewed by Pavel Feldman.
1568
1569         Web Inspector: allow interrupting long running JS to execute inspector command
1570         https://bugs.webkit.org/show_bug.cgi?id=43900
1571
1572         * bindings/v8/ScriptDebugServer.cpp:
1573         (WebCore::):
1574         (WebCore::ScriptDebugServer::pause):
1575         (WebCore::ScriptDebugServer::interruptAndRun):
1576         (WebCore::ScriptDebugServer::runPendingTasks):
1577         (WebCore::ScriptDebugServer::handleV8DebugEvent):
1578         * bindings/v8/ScriptDebugServer.h:
1579         (WebCore::ScriptDebugServer::Task::~Task):
1580
1581 2010-08-12  Sheriff Bot  <webkit.review.bot@gmail.com>
1582
1583         Unreviewed, rolling out r65295.
1584         http://trac.webkit.org/changeset/65295
1585         https://bugs.webkit.org/show_bug.cgi?id=43950
1586
1587         It broke 4 sputnik tests (Requested by Ossy on #webkit).
1588
1589         * WebCore.xcodeproj/project.pbxproj:
1590         * bindings/js/JSDOMWindowCustom.cpp:
1591         (WebCore::JSDOMWindow::getOwnPropertySlot):
1592         (WebCore::JSDOMWindow::getOwnPropertyDescriptor):
1593         * bridge/NP_jsobject.cpp:
1594         (_NPN_Enumerate):
1595         * bridge/c/c_utility.cpp:
1596         (JSC::Bindings::convertValueToNPVariant):
1597         * bridge/jni/JNIBridge.cpp:
1598         (JavaParameter::JavaParameter):
1599         (JavaMethod::JavaMethod):
1600         (JavaMethod::signature):
1601         (JavaMethod::methodID):
1602         * bridge/jni/JNIBridge.h:
1603         (JSC::Bindings::JavaString::UTF8String):
1604         (JSC::Bindings::JavaParameter::type):
1605         (JSC::Bindings::JavaMethod::returnType):
1606         * bridge/jni/jni_jsobject.mm:
1607         (JavaJSObject::call):
1608         (JavaJSObject::eval):
1609         (JavaJSObject::getMember):
1610         (JavaJSObject::setMember):
1611         (JavaJSObject::removeMember):
1612         (JavaJSObject::convertJObjectToValue):
1613         * bridge/jni/jsc/JNIBridgeJSC.cpp:
1614         (JavaField::JavaField):
1615         (JavaField::valueFromInstance):
1616         (JavaField::setValueToInstance):
1617         * bridge/jni/jsc/JNIBridgeJSC.h:
1618         (JSC::Bindings::JavaField::type):
1619         * bridge/jni/jsc/JavaInstanceJSC.cpp:
1620         (JavaInstance::invokeMethod):
1621         * bridge/jni/jsc/JavaStringJSC.h:
1622         (JSC::Bindings::JavaStringImpl::UTF8String):
1623         * bridge/runtime_array.cpp:
1624         (JSC::RuntimeArray::getOwnPropertySlot):
1625         (JSC::RuntimeArray::getOwnPropertyDescriptor):
1626         (JSC::RuntimeArray::put):
1627
1628 2010-08-12  Pavel Feldman  <pfeldman@chromium.org>
1629
1630         Not reviewed. Fixed typo in inspector js.
1631         Follow up to r65300.
1632
1633         * inspector/front-end/inspector.js:
1634         (WebInspector.log.isLogAvailable):
1635
1636 2010-08-12  Pavel Feldman  <pfeldman@chromium.org>
1637
1638         Reviewed by Yury Semikhatsky.
1639
1640         Web Inspector: Make getPrototypes return objects, not names.
1641         https://bugs.webkit.org/show_bug.cgi?id=43934
1642
1643         * inspector/front-end/InjectedScript.js:
1644         (injectedScriptConstructor):
1645         (injectedScriptConstructor.):
1646         * inspector/front-end/PropertiesSidebarPane.js:
1647         (WebInspector.PropertiesSidebarPane.prototype.update.callback):
1648         * inspector/front-end/inspector.js:
1649         (WebInspector.log):
1650
1651 2010-08-12  James Robinson  <jamesr@chromium.org>
1652
1653         Unreviewed, rolling out r65284 and r65288.  Broke chromium compile.
1654         https://bugs.webkit.org/show_bug.cgi?id=43725
1655         http://trac.webkit.org/changeset/65284
1656         http://trac.webkit.org/changeset/65288
1657
1658         * bindings/scripts/CodeGeneratorV8.pm:
1659         * bindings/v8/ScriptController.cpp:
1660         (WebCore::ScriptController::namedItemAdded):
1661         (WebCore::ScriptController::namedItemRemoved):
1662         * bindings/v8/V8DOMWindowShell.cpp:
1663         (WebCore::V8DOMWindowShell::updateDocumentWrapperCache):
1664         * bindings/v8/V8DOMWindowShell.h:
1665         * bindings/v8/V8DOMWrapper.cpp:
1666         (WebCore::V8DOMWrapper::instantiateV8Object):
1667         * bindings/v8/custom/V8HTMLDocumentCustom.cpp:
1668         (WebCore::V8HTMLDocument::namedPropertyDeleter):
1669         (WebCore::V8HTMLDocument::namedPropertyGetter):
1670         (WebCore::V8HTMLDocument::indexedPropertyGetter):
1671         (WebCore::V8HTMLDocument::allAccessorSetter):
1672         (WebCore::toV8):
1673         * html/HTMLDocument.cpp:
1674         (WebCore::HTMLDocument::removeItemFromMap):
1675
1676 2010-08-12  Gavin Barraclough  <barraclough@apple.com>
1677
1678         Reviewed by Sam Weinig
1679
1680         Unify UString::UTF8String() & String::utf8() methods,
1681         remove UString::cost() & make atArrayIndex a free function.
1682
1683         * WebCore.xcodeproj/project.pbxproj:
1684         * bindings/js/JSDOMWindowCustom.cpp:
1685         (WebCore::JSDOMWindow::getOwnPropertySlot):
1686         (WebCore::JSDOMWindow::getOwnPropertyDescriptor):
1687         * bridge/NP_jsobject.cpp:
1688         (_NPN_Enumerate):
1689         * bridge/c/c_utility.cpp:
1690         (JSC::Bindings::convertValueToNPVariant):
1691         * bridge/jni/JNIBridge.cpp:
1692         (JavaParameter::JavaParameter):
1693         (JavaMethod::JavaMethod):
1694         (JavaMethod::signature):
1695         (JavaMethod::methodID):
1696         * bridge/jni/JNIBridge.h:
1697         (JSC::Bindings::JavaString::utf8):
1698         (JSC::Bindings::JavaParameter::type):
1699         (JSC::Bindings::JavaMethod::returnType):
1700         * bridge/jni/jni_jsobject.mm:
1701         (JavaJSObject::call):
1702         (JavaJSObject::eval):
1703         (JavaJSObject::getMember):
1704         (JavaJSObject::setMember):
1705         (JavaJSObject::removeMember):
1706         (JavaJSObject::convertJObjectToValue):
1707         * bridge/jni/jsc/JNIBridgeJSC.cpp:
1708         (JavaField::JavaField):
1709         (JavaField::valueFromInstance):
1710         (JavaField::setValueToInstance):
1711         * bridge/jni/jsc/JNIBridgeJSC.h:
1712         (JSC::Bindings::JavaField::type):
1713         * bridge/jni/jsc/JavaInstanceJSC.cpp:
1714         (JavaInstance::invokeMethod):
1715         * bridge/jni/jsc/JavaStringJSC.h:
1716         (JSC::Bindings::JavaStringImpl::utf8):
1717         * bridge/runtime_array.cpp:
1718         (JSC::RuntimeArray::getOwnPropertySlot):
1719         (JSC::RuntimeArray::getOwnPropertyDescriptor):
1720         (JSC::RuntimeArray::put):
1721
1722 2010-08-12  Gavin Barraclough  <barraclough@apple.com>
1723
1724         Speculative build fix for chromium bindings (someone added the old include style!).
1725
1726         * bindings/v8/V8DOMWindowShell.h:
1727
1728 2010-08-12  MORITA Hajime  <morrita@google.com>
1729
1730         Reviewed by Tony Chang.
1731
1732         [Refactoring] TextEvent class has to many flags
1733         https://bugs.webkit.org/show_bug.cgi?id=43893
1734
1735         Turned TextEvent::m_isLineBreak, TextEvent::m_isBacktab, TextEvent::m_isPaste 
1736         into single TextEvent::m_inputType enumeration.
1737         
1738         No functional change. No new tests.
1739
1740         * dom/TextEvent.cpp:
1741         (WebCore::TextEvent::selectInputType):
1742         (WebCore::TextEvent::create):
1743         (WebCore::TextEvent::createForPlainTextPaste):
1744         (WebCore::TextEvent::createForFragmentPaste):
1745         (WebCore::TextEvent::TextEvent):
1746         * dom/TextEvent.h:
1747         (WebCore::TextEvent::):
1748         (WebCore::TextEvent::isLineBreak):
1749         (WebCore::TextEvent::isBackTab):
1750         (WebCore::TextEvent::isPaste):
1751         * page/EventHandler.cpp:
1752         (WebCore::EventHandler::handleTextInputEvent):
1753
1754 2010-08-12  Kenneth Russell  <kbr@google.com>
1755
1756         Reviewed by David Levin.
1757
1758         Delete obsolete CanvasNumberArray
1759         https://bugs.webkit.org/show_bug.cgi?id=38619
1760
1761         Deleted unreferenced code. Built WebKit and Chromium to test.
1762
1763         * Android.derived.jscbindings.mk:
1764         * Android.derived.v8bindings.mk:
1765         * Android.jscbindings.mk:
1766         * Android.mk:
1767         * bindings/js/JSCanvasNumberArrayCustom.cpp: Removed.
1768         * html/canvas/CanvasNumberArray.cpp: Removed.
1769         * html/canvas/CanvasNumberArray.h: Removed.
1770         * html/canvas/CanvasNumberArray.idl: Removed.
1771
1772 2010-08-12  Anton Muhin  <antonm@chromium.org>
1773
1774         Reviewed by Nate Chapin.
1775
1776         [v8] Create callbacks for named access of elements from document object on fly.
1777         https://bugs.webkit.org/show_bug.cgi?id=43725
1778
1779         That should speed up access to methods on document as there is no need to
1780         consult with interceptor before invoking the method.
1781
1782         This reapplies http://trac.webkit.org/changeset/60670 reverted with
1783         http://trac.webkit.org/changeset/63845.  The main difference is in treatment
1784         of removed named elements: we don't remove an accessor any more, but leave it.
1785         Therefore any accessor should be able to treat the case when there is no
1786         expando property.
1787
1788         * bindings/scripts/CodeGeneratorV8.pm:
1789         * bindings/v8/ScriptController.cpp:
1790         (WebCore::ScriptController::namedItemAdded):
1791         (WebCore::ScriptController::namedItemRemoved):
1792         * bindings/v8/V8DOMWindowShell.cpp:
1793         (WebCore::checkDocumentWrapper):
1794         (WebCore::V8DOMWindowShell::updateDocumentWrapperCache):
1795         (WebCore::getter):
1796         (WebCore::V8DOMWindowShell::namedItemAdded):
1797         (WebCore::V8DOMWindowShell::namedItemRemoved):
1798         * bindings/v8/V8DOMWindowShell.h:
1799         * bindings/v8/V8DOMWrapper.cpp:
1800         (WebCore::V8DOMWrapper::instantiateV8Object):
1801         * bindings/v8/custom/V8HTMLDocumentCustom.cpp:
1802         (WebCore::V8HTMLDocument::WrapInShadowObject):
1803         (WebCore::V8HTMLDocument::GetNamedProperty):
1804         (WebCore::V8HTMLDocument::allAccessorSetter):
1805         (WebCore::toV8):
1806         * html/HTMLDocument.cpp:
1807         (WebCore::HTMLDocument::removeItemFromMap):
1808
1809 2010-08-12  Dimitri Glazkov  <dglazkov@chromium.org>
1810
1811         Reviewed by Adam Barth.
1812
1813         Ensure that parser doesn't attach children that have been removed by JavaScript event handlers.
1814         https://bugs.webkit.org/show_bug.cgi?id=43813
1815
1816         This patch re-fixes bug 40742 in a way that keeps allowing HTMLLinkElement
1817         to lazy-attach.
1818
1819         * html/HTMLConstructionSite.cpp:
1820         (WebCore::HTMLConstructionSite::attach): Added parent check.
1821         * html/HTMLLinkElement.cpp: Basically undoes changes introduced by r61424.
1822         * html/HTMLLinkElement.h: Ditto.
1823
1824 2010-08-12  Justin Schuh  <jschuh@chromium.org>
1825
1826         Reviewed by Dumitru Daniliuc.
1827
1828         Clear PluginData's page pointer on page refresh
1829         https://bugs.webkit.org/show_bug.cgi?id=43888
1830
1831         Test: plugins/access-after-page-destroyed.html
1832
1833         * page/Page.cpp:
1834         (WebCore::Page::refreshPlugins):
1835
1836 2010-08-12  Zhenyao Mo  <zmo@google.com>
1837
1838         Reviewed by Dimitri Glazkov.
1839
1840         getParameter(STENCIL_VALUE_MASK) (and others) returns the wrong value
1841         https://bugs.webkit.org/show_bug.cgi?id=31842
1842
1843         * html/canvas/WebGLRenderingContext.cpp:
1844         (WebCore::WebGLRenderingContext::getUnsignedLongParameter): Casting from int to unsigned long correctly.
1845
1846 2010-08-12  Pavel Feldman  <pfeldman@chromium.org>
1847
1848         Reviewed by Joseph Pecoraro.
1849
1850         Web Inspector: mimic FireBug's array detection.
1851         https://bugs.webkit.org/show_bug.cgi?id=30974
1852
1853         * inspector/front-end/InjectedScript.js:
1854         (injectedScriptConstructor):
1855
1856 2010-08-12  Zhenyao Mo  <zmo@google.com>
1857
1858         Reviewed by Dimitri Glazkov.
1859
1860         regression in generateMipmap
1861         https://bugs.webkit.org/show_bug.cgi?id=43873
1862
1863         * html/canvas/WebGLTexture.cpp:
1864         (WebCore::WebGLTexture::generateMipmapLevelInfo): should set m_needToUseBlackTexture to false if generateMipmapLevelInfo runs through successfully.
1865
1866 2010-08-12  Ariya Hidayat  <ariya@sencha.com>
1867
1868         Reviewed by David Hyatt.
1869
1870         [Qt] Split TransparencyLayer into its own header file.
1871         https://bugs.webkit.org/show_bug.cgi?id=43935
1872
1873         * WebCore.pro: Include TransparencyLayer.h.
1874         * platform/graphics/qt/GraphicsContextQt.cpp:
1875         * platform/graphics/qt/TransparencyLayer.h: Added.
1876         (WebCore::TransparencyLayer::TransparencyLayer):
1877         (WebCore::TransparencyLayer::operator=):
1878
1879 2010-08-12  François Sausset  <sausset@gmail.com>
1880
1881         Reviewed by Beth Dakin.
1882
1883         Horizontal padding should be 0 on mrow elements.
1884         https://bugs.webkit.org/show_bug.cgi?id=43908
1885
1886         Updated tests in /mathml
1887
1888         * css/mathml.css:
1889         (mrow, mfenced):
1890         (mfenced):
1891
1892 2010-08-12  Aaron Boodman  <aa@chromium.org>
1893
1894         Reviewed by David Hyatt.
1895
1896         Make the cascade level of "user" styles configurable
1897         https://bugs.webkit.org/show_bug.cgi?id=43457
1898
1899         * WebCore.exp.in:
1900         * css/CSSStyleSelector.cpp:
1901         (WebCore::CSSStyleSelector::CSSStyleSelector):
1902         * dom/Document.cpp:
1903         (WebCore::Document::pageGroupUserSheets):
1904         * page/PageGroup.cpp:
1905         (WebCore::PageGroup::addUserStyleSheetToWorld):
1906         * page/PageGroup.h:
1907         * page/UserStyleSheet.h:
1908         (WebCore::UserStyleSheet::):
1909         (WebCore::UserStyleSheet::UserStyleSheet):
1910         (WebCore::UserStyleSheet::level):
1911
1912 2010-08-12  Jian Li  <jianli@chromium.org>
1913
1914         Fix qt build break.
1915
1916         * html/ThreadableBlobRegistry.cpp:
1917         (WebCore::postTaskToMainThread):
1918
1919 2010-08-12  Jian Li  <jianli@chromium.org>
1920
1921         Reviewed by David Levin.
1922
1923         Add ThreadableBlobRegistry to support calling BlobRegistry methods in
1924         main thread.
1925         https://bugs.webkit.org/show_bug.cgi?id=43874
1926
1927         * Android.mk:
1928         * CMakeLists.txt:
1929         * GNUmakefile.am:
1930         * WebCore.gypi:
1931         * WebCore.pro:
1932         * WebCore.vcproj/WebCore.vcproj:
1933         * WebCore.xcodeproj/project.pbxproj:
1934         * dom/ScriptExecutionContext.cpp:
1935         (WebCore::ScriptExecutionContext::~ScriptExecutionContext):
1936         (WebCore::ScriptExecutionContext::addBlob):
1937         (WebCore::ScriptExecutionContext::removeBlob):
1938         (WebCore::ScriptExecutionContext::createPublicBlobURL):
1939         (WebCore::ScriptExecutionContext::revokePublicBlobURL):
1940         * dom/ScriptExecutionContext.h:
1941         * html/Blob.cpp:
1942         (WebCore::Blob::Blob):
1943         (WebCore::Blob::~Blob):
1944         (WebCore::Blob::contextDestroyed):
1945         * html/Blob.h:
1946         * html/BlobRegistryImpl.cpp:
1947         (WebCore::BlobRegistryImpl::registerBlobURL):
1948         (WebCore::BlobRegistryImpl::unregisterBlobURL):
1949         (WebCore::BlobRegistryImpl::getBlobDataFromURL):
1950         * html/ThreadableBlobRegistry.cpp: Added.
1951         * html/ThreadableBlobRegistry.h: Added.
1952
1953 2010-08-12  Csaba Osztrogonác  <ossy@webkit.org>
1954
1955         Reviewed by Darin Adler.
1956
1957         generate-bindings.pl should generate warning free code
1958         https://bugs.webkit.org/show_bug.cgi?id=43353
1959
1960         * bindings/scripts/CodeGeneratorJS.pm: Suggested parentheses added around && within ||
1961
1962 2010-08-11  Ryosuke Niwa  <rniwa@webkit.org>
1963
1964         Reviewed by Kent Tamura.
1965
1966         merge MarkupAccumulator and MarkupAccumulatorWrapper
1967         https://bugs.webkit.org/show_bug.cgi?id=43834
1968
1969         Removed MarkupAccumulator and added serializeNodesWithNamespaces to use MarkupAccumulatorWrapper
1970         in both versions of createMarkup.  Accumulation of nodes done manually in serializeNodes is
1971         now done by MarkupAccumulatorWrapper as done in the original MarkupAccumulator.
1972
1973         No new tests added since this is a cleanup.
1974
1975         * editing/markup.cpp:
1976         (WebCore::MarkupAccumulatorWrapper::MarkupAccumulatorWrapper): Takes vector of nodes and set it to m_nodes.
1977         (WebCore::MarkupAccumulatorWrapper::insertOpenTag): Adds node to m_nodes.
1978         (WebCore::MarkupAccumulatorWrapper::wrapWithNode): Adds node to m_nodes.
1979         (WebCore::serializeNodes): Adding node to nodes is moved into MarkupAccumulatorWrapper.
1980         (WebCore::createMarkup): Instantiates MarkupAccumulatorWrapper.
1981         (WebCore::serializeNodesWithNamespaces): Renamed from MarkupAccumulator::appendMarkup.
1982
1983 2010-08-12  Dirk Schulze  <krit@webkit.org>
1984
1985         Unreviewed sort of Xcode project file.
1986
1987         * WebCore.xcodeproj/project.pbxproj:
1988
1989 2010-08-12  Ben Murdoch  <benm@google.com>
1990
1991         Reviewed by Pavel Feldman.
1992
1993         InspectorController.h missing InspectorObject forward declaration
1994         https://bugs.webkit.org/show_bug.cgi?id=43913
1995
1996         Build fix so no new tests.
1997
1998         * inspector/InspectorController.h: Adds the necessary forward declaration.
1999
2000 2010-08-12  Hui Huang  <hui.2.huang@nokia.com>
2001
2002         Reviewed by Eric Seidel.
2003
2004         m_windowRect is not updated because parent of PluginView is not
2005         set when updatePluginWidget is called. plugin gets wrong window
2006         size on setwindow.
2007         https://bugs.webkit.org/show_bug.cgi?id=43635
2008
2009         * plugins/symbian/PluginViewSymbian.cpp:
2010         (WebCore::PluginView::setParent):
2011
2012 2010-08-12  Pavel Feldman  <pfeldman@chromium.org>
2013
2014         Not reviewed, re-apply r65241 and 65243.
2015
2016         Web Inspector: brush up object proxies, introduce remote object model.
2017         https://bugs.webkit.org/show_bug.cgi?id=43847
2018
2019         * WebCore.gypi:
2020         * WebCore.vcproj/WebCore.vcproj:
2021         * inspector/front-end/AuditRules.js:
2022         (WebInspector.AuditRules.CssInHeadRule.prototype.doRun):
2023         (WebInspector.AuditRules.StylesScriptsOrderRule.prototype.doRun):
2024         * inspector/front-end/ConsoleView.js:
2025         (WebInspector.ConsoleView.prototype.doEvalInWindow):
2026         (WebInspector.ConsoleView.prototype._enterKeyPressed.printResult):
2027         (WebInspector.ConsoleView.prototype._enterKeyPressed):
2028         (WebInspector.ConsoleView.prototype._format):
2029         (WebInspector.ConsoleView.prototype._formatarray):
2030         (WebInspector.ConsoleView.prototype._formatAsArrayEntry):
2031         (WebInspector.ConsoleMessage.prototype._format):
2032         (WebInspector.ConsoleCommandResult):
2033         * inspector/front-end/DOMAgent.js:
2034         (WebInspector.DOMNode):
2035         * inspector/front-end/ElementsPanel.js:
2036         (WebInspector.ElementsPanel.prototype.performSearch):
2037         * inspector/front-end/ElementsTreeOutline.js:
2038         (WebInspector.ElementsTreeElement.prototype._createTooltipForNode):
2039         * inspector/front-end/EventListenersSidebarPane.js:
2040         ():
2041         * inspector/front-end/InjectedScript.js:
2042         (injectedScriptConstructor):
2043         (injectedScriptConstructor.):
2044         * inspector/front-end/ObjectPropertiesSection.js:
2045         (WebInspector.ObjectPropertiesSection.prototype.update):
2046         (WebInspector.ObjectPropertiesSection.prototype.updateProperties):
2047         (WebInspector.ObjectPropertyTreeElement.prototype.onpopulate):
2048         (WebInspector.ObjectPropertyTreeElement.prototype.update):
2049         (WebInspector.ObjectPropertyTreeElement.prototype.applyExpression):
2050         * inspector/front-end/ObjectProxy.js: Removed.
2051         * inspector/front-end/PropertiesSidebarPane.js:
2052         (WebInspector.PropertiesSidebarPane.prototype.update.callback):
2053         * inspector/front-end/RemoteObject.js: Added.
2054         (WebInspector.RemoteObjectId):
2055         (WebInspector.RemoteObject):
2056         (WebInspector.RemoteObject.fromPrimitiveValue):
2057         (WebInspector.RemoteObject.fromNode):
2058         (WebInspector.RemoteObject.fromPayload):
2059         (WebInspector.RemoteObject.type):
2060         (WebInspector.RemoteObject.prototype.get objectId):
2061         (WebInspector.RemoteObject.prototype.get type):
2062         (WebInspector.RemoteObject.prototype.get description):
2063         (WebInspector.RemoteObject.prototype.get hasChildren):
2064         (WebInspector.RemoteObject.prototype.isError):
2065         (WebInspector.RemoteObject.prototype.getPropertyValueDescriptions):
2066         (WebInspector.RemoteObject.prototype.getOwnProperties):
2067         (WebInspector.RemoteObject.prototype.getProperties.remoteObjectBinder):
2068         (WebInspector.RemoteObjectProperty):
2069         * inspector/front-end/ScopeChainSidebarPane.js:
2070         (WebInspector.ScopeChainSidebarPane.prototype.update):
2071         * inspector/front-end/ScriptsPanel.js:
2072         (WebInspector.ScriptsPanel.prototype.evaluateInSelectedCallFrame.updatingCallbackWrapper):
2073         (WebInspector.ScriptsPanel.prototype.evaluateInSelectedCallFrame):
2074         (WebInspector.ScriptsPanel.prototype.doEvalInCallFrame):
2075         * inspector/front-end/SourceFrame.js:
2076         (WebInspector.SourceFrame.prototype._showPopup.evaluateCallback):
2077         (WebInspector.SourceFrame.prototype._showPopup):
2078         (WebInspector.SourceFrame.prototype._evalSelectionInCallFrame):
2079         * inspector/front-end/WatchExpressionsSidebarPane.js:
2080         (WebInspector.WatchExpressionsSection.prototype.update):
2081         (WebInspector.WatchExpressionTreeElement.prototype.update):
2082         * inspector/front-end/WebKit.qrc:
2083         * inspector/front-end/inspector.html:
2084         * inspector/front-end/inspector.js:
2085         (WebInspector.log.logMessage):
2086         (WebInspector.log):
2087         * inspector/front-end/utilities.js:
2088         ():
2089
2090 2010-08-11  Stephen White  <senorblanco@chromium.org>
2091
2092         Reviewed by David Levin.
2093
2094         Add support for BGRA pixel format to GraphicsContext3D.
2095         https://bugs.webkit.org/show_bug.cgi?id=43858
2096
2097         Although this has been standard in desktop GL since 1.2, it's an 
2098         extension in GL ES 2.0, so a query function is provided to check
2099         for support.  Since they differ on semantics (GL ES 2.0 requires 
2100         TexImage2D()'s format and internalFormat to match, while desktop GL
2101         will not accept BGRA as an internalFormat), the stub implementation
2102         returns false until these quirks have been implemented and tested on
2103         each port.
2104
2105         * platform/graphics/GraphicsContext3D.cpp:
2106         (WebCore::GraphicsContext3D::supportsBGRA):
2107         Add implementation for non-chromium platforms which returns false.
2108         * platform/graphics/GraphicsContext3D.h:
2109         (WebCore::GraphicsContext3D::):
2110         Add query function (supportsBGRA()) and the BGRA_EXT token.
2111
2112 2010-08-12  Pavel Feldman  <pfeldman@chromium.org>
2113
2114         Not reviewed. Rolling out r65241 and 65243 for breaking inspector tests.
2115
2116 2010-08-12  Pavel Feldman  <pfeldman@chromium.org>
2117
2118         Not reviewed. Adding the file missing in r65241.
2119
2120         * inspector/front-end/RemoteObject.js: Added.
2121         (WebInspector.RemoteObjectId):
2122         (WebInspector.RemoteObject):
2123         (WebInspector.RemoteObject.fromPrimitiveValue):
2124         (WebInspector.RemoteObject.fromNode):
2125         (WebInspector.RemoteObject.fromPayload):
2126         (WebInspector.RemoteObject.type):
2127         (WebInspector.RemoteObject.prototype.get objectId):
2128         (WebInspector.RemoteObject.prototype.get type):
2129         (WebInspector.RemoteObject.prototype.get description):
2130         (WebInspector.RemoteObject.prototype.get hasChildren):
2131         (WebInspector.RemoteObject.prototype.isError):
2132         (WebInspector.RemoteObject.prototype.getPropertyValueDescriptions):
2133         (WebInspector.RemoteObject.prototype.getOwnProperties):
2134         (WebInspector.RemoteObject.prototype.getProperties.remoteObjectBinder):
2135         (WebInspector.RemoteObjectProperty):
2136
2137 2010-08-12  Pavel Feldman  <pfeldman@chromium.org>
2138
2139         Reviewed by Yury Semikhatsky.
2140
2141         Web Inspector: brush up object proxies, introduce remote object model.
2142         https://bugs.webkit.org/show_bug.cgi?id=43847
2143
2144         * WebCore.gypi:
2145         * WebCore.vcproj/WebCore.vcproj:
2146         * inspector/front-end/AuditRules.js:
2147         (WebInspector.AuditRules.CssInHeadRule.prototype.doRun):
2148         (WebInspector.AuditRules.StylesScriptsOrderRule.prototype.doRun):
2149         * inspector/front-end/ConsoleView.js:
2150         (WebInspector.ConsoleView.prototype.doEvalInWindow):
2151         (WebInspector.ConsoleView.prototype._enterKeyPressed.printResult):
2152         (WebInspector.ConsoleView.prototype._enterKeyPressed):
2153         (WebInspector.ConsoleView.prototype._format):
2154         (WebInspector.ConsoleView.prototype._formatnode):
2155         (WebInspector.ConsoleView.prototype._formatarray):
2156         (WebInspector.ConsoleView.prototype._formatAsArrayEntry):
2157         (WebInspector.ConsoleMessage.prototype._format):
2158         (WebInspector.ConsoleCommandResult):
2159         * inspector/front-end/DOMAgent.js:
2160         (WebInspector.DOMNode):
2161         * inspector/front-end/ElementsTreeOutline.js:
2162         (WebInspector.ElementsTreeElement.prototype._createTooltipForNode):
2163         * inspector/front-end/EventListenersSidebarPane.js:
2164         * inspector/front-end/InjectedScript.js:
2165         (injectedScriptConstructor):
2166         (injectedScriptConstructor.):
2167         * inspector/front-end/ObjectPropertiesSection.js:
2168         (WebInspector.ObjectPropertiesSection.prototype.update):
2169         (WebInspector.ObjectPropertiesSection.prototype.updateProperties):
2170         (WebInspector.ObjectPropertyTreeElement.prototype.onpopulate):
2171         (WebInspector.ObjectPropertyTreeElement.prototype.update):
2172         (WebInspector.ObjectPropertyTreeElement.prototype.applyExpression):
2173         * inspector/front-end/ObjectProxy.js: Removed.
2174         * inspector/front-end/PropertiesSidebarPane.js:
2175         (WebInspector.PropertiesSidebarPane.prototype.update.callback):
2176         * inspector/front-end/RemoteObject.js: Added.
2177         (WebInspector.RemoteObjectId):
2178         (WebInspector.RemoteObject):
2179         (WebInspector.RemoteObject.fromPrimitiveValue):
2180         (WebInspector.RemoteObject.fromErrorMessage):
2181         (WebInspector.RemoteObject.fromNode):
2182         (WebInspector.RemoteObject.fromPayload):
2183         (WebInspector.RemoteObject.type):
2184         (WebInspector.RemoteObject.prototype.get objectId):
2185         (WebInspector.RemoteObject.prototype.get type):
2186         (WebInspector.RemoteObject.prototype.get description):
2187         (WebInspector.RemoteObject.prototype.get hasChildren):
2188         (WebInspector.RemoteObject.prototype.isError):
2189         (WebInspector.RemoteObject.prototype.getPropertyValueDescriptions):
2190         (WebInspector.RemoteObject.prototype.getOwnProperties):
2191         (WebInspector.RemoteObject.prototype.getProperties.remoteObjectBinder):
2192         (WebInspector.RemoteObjectProperty):
2193         * inspector/front-end/ScopeChainSidebarPane.js:
2194         (WebInspector.ScopeChainSidebarPane.prototype.update):
2195         * inspector/front-end/ScriptsPanel.js:
2196         (WebInspector.ScriptsPanel.prototype.evaluateInSelectedCallFrame.updatingCallbackWrapper):
2197         (WebInspector.ScriptsPanel.prototype.evaluateInSelectedCallFrame):
2198         (WebInspector.ScriptsPanel.prototype.doEvalInCallFrame):
2199         * inspector/front-end/SourceFrame.js:
2200         (WebInspector.SourceFrame.prototype._showPopup.evaluateCallback):
2201         (WebInspector.SourceFrame.prototype._showPopup):
2202         (WebInspector.SourceFrame.prototype._evalSelectionInCallFrame):
2203         * inspector/front-end/WatchExpressionsSidebarPane.js:
2204         (WebInspector.WatchExpressionsSection.prototype.update):
2205         (WebInspector.WatchExpressionTreeElement.prototype.update):
2206         * inspector/front-end/WebKit.qrc:
2207         * inspector/front-end/inspector.html:
2208         * inspector/front-end/inspector.js:
2209         (WebInspector.log.logMessage):
2210         (WebInspector.log):
2211         * inspector/front-end/utilities.js:
2212
2213 2010-08-12  Ben Murdoch  <benm@google.com>
2214
2215         Reviewed by Steve Block.
2216
2217         V8 code generator does not add V8BindingMacros.h include in all places it is needed
2218         https://bugs.webkit.org/show_bug.cgi?id=43907
2219
2220         Adds the V8BindingMacros.h header before using the STRING_TO_V8PARAMETER_EXCEPTION_BLOCK
2221         macro.
2222
2223         Build fix so no new tests.
2224
2225         * bindings/scripts/CodeGeneratorV8.pm: include the header at the right time.
2226
2227 2010-08-12  Jeremy Orlow  <jorlow@chromium.org>
2228
2229         Revert for now
2230         https://bugs.webkit.org/show_bug.cgi?id=43794 
2231
2232         * Android.mk:
2233         * CMakeLists.txt:
2234         * GNUmakefile.am:
2235         * WebCore.exp.in:
2236         * WebCore.gypi:
2237         * WebCore.pro:
2238         * WebCore.vcproj/WebCore.vcproj:
2239         * WebCore.xcodeproj/project.pbxproj:
2240         * page/GroupSettings.cpp: Removed.
2241         * page/GroupSettings.h: Removed.
2242         * page/PageGroup.cpp:
2243         (WebCore::PageGroup::PageGroup):
2244         (WebCore::PageGroup::localStorage):
2245         * page/PageGroup.h:
2246         * page/Settings.cpp:
2247         (WebCore::Settings::Settings):
2248         (WebCore::Settings::setLocalStorageQuota):
2249         (WebCore::Settings::setLocalStorageDatabasePath):
2250         * page/Settings.h:
2251         (WebCore::Settings::localStorageQuota):
2252         (WebCore::Settings::localStorageDatabasePath):
2253
2254 2010-08-12  Nikolas Zimmermann  <nzimmermann@rim.com>
2255
2256         Not reviewed. Follow up to Jeremys fix, just remove the context parameter, it's not needed anymore.
2257
2258         * rendering/RenderSVGResourceGradient.cpp:
2259         (WebCore::createMaskAndSwapContextForTextGradient):
2260         * rendering/RenderSVGResourceMasker.cpp:
2261         (WebCore::RenderSVGResourceMasker::applyResource):
2262         * rendering/SVGImageBufferTools.cpp:
2263         (WebCore::SVGImageBufferTools::createImageBuffer):
2264         * rendering/SVGImageBufferTools.h:
2265
2266 2010-08-12  Jeremy Orlow  <jorlow@chromium.org>
2267
2268         Build fix.
2269
2270         * rendering/SVGImageBufferTools.cpp:
2271         (WebCore::SVGImageBufferTools::createImageBuffer):
2272
2273 2010-08-12  Nikolas Zimmermann  <nzimmermann@rim.com>
2274
2275         Reviewed by Dirk Schulze.
2276
2277         SVG masking performance very slow
2278         https://bugs.webkit.org/show_bug.cgi?id=43622
2279
2280         Create ImageBuffers only as large as the final rendered size on screen. Only gradient on text on CG handled this correctly so far.
2281         Refactored the code into a new SVGImageBufferTools class, and use the code from RenderSVGResourceMasker -> makes the IE9 demo SVG dice usable.
2282         Clippers, Patterns and Filters remain to be converted.
2283
2284         Test: svg/zoom/page/zoom-mask-with-percentages.svg
2285
2286         * Android.mk: Add SVGImageBufferTools.* to build.
2287         * CMakeLists.txt: Ditto.
2288         * GNUmakefile.am: Ditto.
2289         * WebCore.gypi: Ditto.
2290         * WebCore.pro: Ditto.
2291         * WebCore.xcodeproj/project.pbxproj: Ditto.
2292         * rendering/RenderSVGAllInOne.cpp: Ditto.
2293         * rendering/RenderSVGResourceGradient.cpp: Refactored CG gradient specific "create image buffer in absolute coordinates" code into SVGImageBufferTools class.
2294         (WebCore::createMaskAndSwapContextForTextGradient):
2295         (WebCore::clipToTextMask):
2296         (WebCore::RenderSVGResourceGradient::applyResource):
2297         * rendering/RenderSVGResourceMasker.cpp: Use new SVGImageBufferTools class, to avoid pixelation when zooming and to create image buffers as big as the final rendered size on screen, not more.
2298         (WebCore::RenderSVGResourceMasker::invalidateClients):
2299         (WebCore::RenderSVGResourceMasker::applyResource):
2300         (WebCore::RenderSVGResourceMasker::drawContentIntoMaskImage):
2301         (WebCore::RenderSVGResourceMasker::calculateMaskContentRepaintRect):
2302         (WebCore::RenderSVGResourceMasker::resourceBoundingBox):
2303         * rendering/RenderSVGResourceMasker.h:
2304         * rendering/SVGImageBufferTools.cpp: Added.
2305         (WebCore::SVGImageBufferTools::absoluteTransformFromContext):
2306         (WebCore::SVGImageBufferTools::createImageBuffer):
2307         (WebCore::SVGImageBufferTools::clipToImageBuffer):
2308         * rendering/SVGImageBufferTools.h: Added.
2309         (WebCore::SVGImageBufferTools::SVGImageBufferTools):
2310         (WebCore::SVGImageBufferTools::~SVGImageBufferTools):
2311
2312 2010-08-10  Jeremy Orlow  <jorlow@chromium.org>
2313
2314         Reviewed by Adam Barth.
2315
2316         Some settings are linked to the PageGroup not the Page.  Create a new class for those.
2317         https://bugs.webkit.org/show_bug.cgi?id=43794
2318
2319         Since persistient storage is shared between pages, there's no way to modify
2320         some settings related to it on a per page basis.  As such, it's not technically
2321         correct for these settings to be on the Page's settings.  Create a new class
2322         called GroupSettings, move the 2 group-wide local storage settings there, and
2323         add a new setting for IndexedDB's path (which is prompting this change).
2324
2325         No behavior has changed, so no tests.
2326
2327         * Android.mk:
2328         * CMakeLists.txt:
2329         * GNUmakefile.am:
2330         * WebCore.exp.in:
2331         * WebCore.gypi:
2332         * WebCore.pro:
2333         * WebCore.vcproj/WebCore.vcproj:
2334         * WebCore.xcodeproj/project.pbxproj:
2335         * page/PageGroup.cpp:
2336         (WebCore::PageGroup::PageGroup):
2337         (WebCore::PageGroup::localStorage):
2338         * page/PageGroup.h:
2339         (WebCore::PageGroup::groupSettings):
2340         * page/Settings.cpp:
2341         (WebCore::Settings::Settings):
2342         (WebCore::Settings::setSessionStorageQuota):
2343         * page/Settings.h:
2344
2345 2010-08-12  Benjamin Poulain  <benjamin.poulain@nokia.com>
2346
2347         Reviewed by Simon Hausmann.
2348
2349         [Qt] Decode images directly to QPixmap
2350         https://bugs.webkit.org/show_bug.cgi?id=40797
2351
2352         Use the new API of Qt 4.7 to decode data from the image
2353         reader directly to QPixmap.
2354
2355         This allow us to use JDCT_IFAST when decoding jpeg images
2356         to pixmap, and to decode animated GIF images, while still
2357         using in-place conversion of color space.
2358
2359         * platform/graphics/qt/ImageDecoderQt.cpp:
2360         (WebCore::ImageDecoderQt::setData):
2361         (WebCore::ImageDecoderQt::internalHandleCurrentImage):
2362
2363 2010-07-14  Marcus Bulach  <bulach@chromium.org>
2364
2365         Reviewed by darin@apple.com.
2366
2367         Removes DontCheckEnums from some IDLs and fixes the corresponding enums.
2368         https://bugs.webkit.org/show_bug.cgi?id=42278
2369
2370         This is a follow-up to https://bugs.webkit.org/show_bug.cgi?id=42250: it fixes some IDL/.h and enables generating compile-time checks for enums.
2371
2372         Tests: idls should generate valid code.
2373
2374         * dom/Node.h:
2375         (WebCore::Node::):
2376         * dom/Node.idl:
2377         * dom/OverflowEvent.h:
2378         (WebCore::OverflowEvent::):
2379         * dom/OverflowEvent.idl:
2380         * html/FileReader.cpp:
2381         (WebCore::FileReader::readyState):
2382         * html/FileReader.h:
2383         (WebCore::FileReader::):
2384         * html/FileReader.idl:
2385
2386 2010-08-12  Alejandro G. Castro  <alex@igalia.com>
2387
2388         Reviewed by David Levin.
2389
2390         The #if clause enabling a feature should avoid the all includes of
2391         the feature
2392         https://bugs.webkit.org/show_bug.cgi?id=43866
2393
2394         Modified the position of the guard to avoid including all the
2395         headers if the feature controlling the guard is not enabled.
2396
2397         * dom/make_names.pl:
2398
2399 2010-08-12  Sheriff Bot  <webkit.review.bot@gmail.com>
2400
2401         Unreviewed, rolling out r65221.
2402         http://trac.webkit.org/changeset/65221
2403         https://bugs.webkit.org/show_bug.cgi?id=43896
2404
2405         Broke snow leopard build. (Requested by dave_levin on
2406         #webkit).
2407
2408         * platform/graphics/GraphicsContext3D.cpp:
2409         * platform/graphics/GraphicsContext3D.h:
2410         (WebCore::GraphicsContext3D::):
2411
2412 2010-08-11  Stephen White  <senorblanco@chromium.org>
2413
2414         Reviewed by David Levin.
2415
2416         Add support for BGRA pixel format to GraphicsContext3D.
2417         https://bugs.webkit.org/show_bug.cgi?id=43858
2418
2419         Although this has been standard in desktop GL since 1.2, it's an 
2420         extension in GL ES 2.0, so a query function is provided to check
2421         for support.  Since they differ on semantics (GL ES 2.0 requires 
2422         TexImage2D()'s format and internalFormat to match, while desktop GL
2423         will not accept BGRA as an internalFormat), the stub implementation
2424         returns false until these quirks have been implemented and tested on
2425         each port.
2426
2427         * platform/graphics/GraphicsContext3D.cpp:
2428         (WebCore::GraphicsContext3D::supportsBGRA):
2429         Add implementation for non-chromium platforms which returns false.
2430         * platform/graphics/GraphicsContext3D.h:
2431         (WebCore::GraphicsContext3D::):
2432         Add query function (supportsBGRA()) and the BGRA_EXT token.
2433
2434 2010-08-11  Satish Sampath  <satish@chromium.org>
2435
2436         Reviewed by Kent Tamura.
2437
2438         Handle clicks and recognition events for the speech input button
2439         https://bugs.webkit.org/show_bug.cgi?id=43857
2440
2441         * rendering/RenderInputSpeech.cpp:
2442         (WebCore::RenderInputSpeech::paintInputFieldSpeechButton): Selects the image based on current state.
2443         * rendering/TextControlInnerElements.cpp:
2444         (WebCore::InputFieldSpeechButtonElement::InputFieldSpeechButtonElement):
2445         (WebCore::InputFieldSpeechButtonElement::defaultEventHandler):
2446         (WebCore::InputFieldSpeechButtonElement::set_state): Switch to a new state and update the UI.
2447         (WebCore::InputFieldSpeechButtonElement::didCompleteRecording): 
2448         (WebCore::InputFieldSpeechButtonElement::didCompleteRecognition):
2449         * rendering/TextControlInnerElements.h:
2450         (WebCore::InputFieldSpeechButtonElement::):
2451         (WebCore::InputFieldSpeechButtonElement::state):
2452
2453 2010-08-11  Fumitoshi Ukai  <ukai@chromium.org>
2454
2455         Reviewed by Alexey Proskuryakov.
2456
2457         REGRESSION(65135): format specifier warnings
2458         https://bugs.webkit.org/show_bug.cgi?id=43876
2459
2460         * websockets/WebSocketChannel.cpp:
2461         (WebCore::WebSocketChannel::appendToBuffer): cast size_t to unsigned long for %lu.
2462         (WebCore::WebSocketChannel::processBuffer): cast size_t to unsigned long for %lu.
2463
2464 2010-08-11  Kevin Ollivier  <kevino@theolliviers.com>
2465
2466         [wx] Build fix, don't generate WebDOMEventTarget.h/.cpp as we use the ones in bindings/cpp.
2467         
2468         * DerivedSources.make:
2469
2470 2010-08-11  Tony Gentilcore  <tonyg@chromium.org>
2471
2472         Reviewed by Adam Barth.
2473
2474         Make PendingScript copyable so that it can be stored in containers
2475         https://bugs.webkit.org/show_bug.cgi?id=43875
2476
2477         This will be useful for storing PendingScripts in a Vector or Deque for
2478         async and defer scripts.
2479
2480         No new tests because no functional changes.
2481
2482         * dom/PendingScript.h:
2483         (WebCore::PendingScript::PendingScript):
2484         (WebCore::PendingScript::operator=):
2485
2486 2010-08-11  Victoria Kirst  <vrk@google.com>
2487
2488         Reviewed by David Levin.
2489
2490         Fixing a regression in the UI for Mac default audio/video controls.
2491         https://bugs.webkit.org/show_bug.cgi?id=43750
2492
2493         * rendering/RenderThemeChromiumMac.h:
2494         * rendering/RenderThemeChromiumMac.mm:
2495         (WebCore::RenderThemeChromiumMac::paintMediaVolumeSliderContainer):
2496         (WebCore::RenderThemeChromiumMac::paintMediaSliderThumb):
2497         (WebCore::RenderThemeChromiumMac::volumeSliderOffsetFromMuteButton):
2498
2499 2010-08-11  Adam Barth  <abarth@webkit.org>
2500
2501         Reviewed by Eric Seidel.
2502
2503         HTML TreeBuilder hits ASSERT in fragment case with insertAdjacentHTML and colgroup
2504         https://bugs.webkit.org/show_bug.cgi?id=43758
2505
2506         This patch conditionalizes some LegacyHTMLTreeBuilder-specific code in
2507         HTMLElement::createContextualFragment that interferes with the new
2508         HTMLTreeBuilder.  Doing that exposes the above ASSERT, which I've fixed
2509         in this patch too.  Fixing that ASSERT involved a small refactoring in
2510         ExternalCharacterTokenBuffer.
2511
2512         * html/HTMLElement.cpp:
2513         (WebCore::HTMLElement::createContextualFragment):
2514         * html/HTMLTreeBuilder.cpp:
2515         (WebCore::HTMLTreeBuilder::ExternalCharacterTokenBuffer::skipLeadingWhitespace):
2516         (WebCore::HTMLTreeBuilder::ExternalCharacterTokenBuffer::takeLeadingWhitespace):
2517         (WebCore::HTMLTreeBuilder::ExternalCharacterTokenBuffer::takeLeadingNonWhitespace):
2518         (WebCore::HTMLTreeBuilder::ExternalCharacterTokenBuffer::skipLeading):
2519         (WebCore::HTMLTreeBuilder::ExternalCharacterTokenBuffer::takeLeading):
2520         (WebCore::HTMLTreeBuilder::processCharacterBuffer):
2521
2522 2010-08-11  Adam Barth  <abarth@webkit.org>
2523
2524         Reviewed by Eric Seidel.
2525
2526         Conditionalize wrong fragment parsing code to pass more HTML5lib tests
2527         https://bugs.webkit.org/show_bug.cgi?id=43877
2528
2529         This code exists to support the LegacyHTMLTreeBuilder.  Unfortunately,
2530         it causes problems for the new HTMLTreeBuilder (which is more
2531         self-contained).
2532
2533         * dom/Element.cpp:
2534         (WebCore::Element::createContextualFragment):
2535
2536 2010-08-11  Ryosuke Niwa  <rniwa@webkit.org>
2537
2538         Reviewed by Tony Chang.
2539
2540         Can't unbold text in div in font-weight span
2541         https://bugs.webkit.org/show_bug.cgi?id=26871
2542
2543         The bug was caused by removeInlineStyle not being able to include styled inline nodes around the start.
2544         Solved this problem by pushing down all inline styles instead of just text-decorations.
2545         This approach allows removeInlineStyle to remove styled ancestors properly and generates compact markups.
2546
2547         Test: editing/style/push-down-inline-styles.html
2548
2549         * editing/ApplyStyleCommand.cpp:
2550         (WebCore::ApplyStyleCommand::removeCSSStyle): No longer removes attributes or node when mode == RemoveNone.
2551         (WebCore::ApplyStyleCommand::highestAncestorWithConflictingInlineStyle): Calls shouldRemoveInlineStyleFromElement
2552         to determine the highest ancestor whose style needs to be pushed down.
2553         (WebCore::ApplyStyleCommand::extractInlineStyleToPushDown): Renamed from extractTextDecorationStyle.
2554         Extracts all inline CSS properties specified instead of just text decorations.
2555         (WebCore::ApplyStyleCommand::applyInlineStyleToPushDown): Renamed from applyTextDecorationStyle.
2556         Applies inline styles using addInlineStyleIfNeeded or adds inline CSS values.
2557         (WebCore::ApplyStyleCommand::pushDownInlineStyleAroundNode): Renamed from pushDownTextDecorationStyleAroundNode.
2558         (WebCore::ApplyStyleCommand::removeInlineStyle): Calls pushDownTextDecorationStyleAroundNode.
2559         * editing/ApplyStyleCommand.h:
2560         * editing/DeleteSelectionCommand.cpp:
2561         (WebCore::DeleteSelectionCommand::mergeParagraphs): Prevents moveParagraph from preserving
2562         the style of an empty paragraph when merged with the previous paragraph because we don't use that style anyways.
2563
2564 2010-08-11  Julien Chaffraix  <jchaffraix@codeaurora.org>
2565
2566         Reviewed by Ariya Hidayat.
2567
2568         Remove dead code in CSSStyleSelector.
2569         https://bugs.webkit.org/show_bug.cgi?id=43805
2570
2571         * css/CSSStyleSelector.cpp: Remove some debugging code that
2572         was compiled out.
2573
2574 2010-08-11  Chris Fleizach  <cfleizach@apple.com>
2575
2576         Reviewed by Beth Dakin.
2577
2578         AX: isNativeCheckbox does not work as advertised
2579         https://bugs.webkit.org/show_bug.cgi?id=43872
2580
2581         Test: platform/mac/accessibility/native-vs-nonnative-checkboxes.html
2582
2583         * accessibility/AccessibilityRenderObject.cpp:
2584         (WebCore::AccessibilityRenderObject::isInputImage):
2585             Only say it's an input image if its role has not been overriden
2586         (WebCore::AccessibilityRenderObject::isNativeCheckboxOrRadio):
2587         * dom/InputElement.h:
2588             Add methods to query if it's a checkbox or radio button type element.
2589         * html/HTMLInputElement.h:
2590         (WebCore::HTMLInputElement::isCheckbox):
2591
2592 2010-08-11  Kenneth Russell  <kbr@google.com>
2593
2594         Reviewed by David Levin.
2595
2596         Remove obsolete WebGLArray type names
2597         https://bugs.webkit.org/show_bug.cgi?id=43885
2598
2599         Deleted aliases to obsolete WebGLArray type names. Updated
2600         affected layout tests, synchronizing them with the versions in the
2601         Khronos repository where appropriate. Ran all layout tests.
2602
2603         * bindings/js/JSDOMWindowCustom.cpp:
2604         * page/DOMWindow.idl:
2605
2606 2010-08-11  Steve Block  <steveblock@google.com>
2607
2608         Reviewed by David Levin.
2609
2610         [chromium] Geolocation breaks the compile if ENABLE_GEOLOCATION is set to 0
2611         https://bugs.webkit.org/show_bug.cgi?id=43753
2612
2613         This change guards the entire contents of the V8 custom bindings files
2614         with ENABLE(GEOLOCATION) to match JSC.
2615
2616         No new tests, build fix only.
2617
2618         * bindings/v8/custom/V8CustomPositionCallback.cpp:
2619         * bindings/v8/custom/V8CustomPositionErrorCallback.cpp:
2620         * bindings/v8/custom/V8GeolocationCustom.cpp:
2621
2622 2010-08-11  Kwang Yul Seo  <skyul@company100.net>
2623
2624         Reviewed by David Levin.
2625
2626         Include <wtf/UnusedParam.h> in GraphicsContextSkia.cpp
2627         https://bugs.webkit.org/show_bug.cgi?id=43843
2628
2629         wtf/UnusedParam.h must be included because UNUSED_PARAM macro is used.
2630
2631         * platform/graphics/skia/GraphicsContextSkia.cpp:
2632
2633 2010-08-11  Nico Weber  <thakis@chromium.org>
2634
2635         Reviewed by David Levin.
2636
2637         [chromium/mac]: Fix colors in compositing mode
2638         https://bugs.webkit.org/show_bug.cgi?id=43823
2639
2640         The textures are drawn by opengl, which doesn't do any color space
2641         correction -- it assumes everything is in the device color space
2642         already. Hence, when drawing into the memory that will be uploaded
2643         as opengl texture, use the device color space.
2644
2645         * platform/graphics/chromium/ImageLayerChromium.cpp:
2646         (WebCore::ImageLayerChromium::updateTextureContents):
2647         * platform/graphics/chromium/LayerChromium.cpp:
2648         (WebCore::LayerChromium::updateTextureContents):
2649         * platform/graphics/chromium/LayerRendererChromium.cpp:
2650         (WebCore::LayerRendererChromium::setRootLayerCanvasSize):
2651
2652 2010-08-11  Kenneth Russell  <kbr@google.com>
2653
2654         Reviewed by Dimitri Glazkov.
2655
2656         Remove obsolete texImage2D and texSubImage2D entry points
2657         https://bugs.webkit.org/show_bug.cgi?id=40320
2658
2659         No new tests; covered by existing WebGL tests.
2660
2661         * html/canvas/WebGLRenderingContext.cpp:
2662         * html/canvas/WebGLRenderingContext.h:
2663         * html/canvas/WebGLRenderingContext.idl:
2664
2665 2010-08-11  Gavin Barraclough  <barraclough@apple.com>
2666
2667         Windows build fix.
2668
2669         * platform/win/BString.cpp:
2670         (WebCore::BString::BString):
2671
2672 2010-08-11  Gavin Barraclough  <barraclough@apple.com>
2673
2674         Qt build fix II.
2675
2676         * WebCore.xcodeproj/project.pbxproj:
2677         * bridge/qt/qt_runtime.cpp:
2678         (JSC::Bindings::convertValueToQVariant):
2679
2680 2010-08-11  Gavin Barraclough  <barraclough@apple.com>
2681
2682         Qt build fix.
2683
2684         * WebCore.xcodeproj/project.pbxproj:
2685         * bridge/qt/qt_runtime.cpp:
2686         (JSC::Bindings::convertValueToQVariant):
2687
2688 2010-08-11  Steve Block  <steveblock@google.com>
2689
2690         Reviewed by Jeremy Orlow.
2691
2692         Need EmptyDeviceOrientationClient and EmptyDeviceMotionClient for use with SVGImage
2693         https://bugs.webkit.org/show_bug.cgi?id=43848
2694
2695         Tested by existing DeviceOrientation tests.
2696
2697         * loader/EmptyClients.h:
2698         (WebCore::EmptyDeviceMotionClient::setController):
2699         (WebCore::EmptyDeviceMotionClient::startUpdating):
2700         (WebCore::EmptyDeviceMotionClient::stopUpdating):
2701         (WebCore::EmptyDeviceMotionClient::currentDeviceMotion):
2702         (WebCore::EmptyDeviceOrientationClient::setController):
2703         (WebCore::EmptyDeviceOrientationClient::startUpdating):
2704         (WebCore::EmptyDeviceOrientationClient::stopUpdating):
2705         (WebCore::EmptyDeviceOrientationClient::lastOrientation):
2706         * svg/graphics/SVGImage.cpp:
2707         (WebCore::SVGImage::dataChanged):
2708
2709 2010-08-11  Rob Buis  <rwlbuis@gmail.com>
2710
2711         Reviewed by Dan Bernstein.
2712
2713         https://bugs.webkit.org/show_bug.cgi?id=42042
2714         An empty value for xml:lang isn't considered
2715
2716         Revert last line change of my patch ; it happens to
2717         behave correctly and same as before, but makes the
2718         use case of empty values unneededly slower.
2719
2720         No new tests. (No change in behavior)
2721
2722         * css/CSSStyleSelector.cpp:
2723         (WebCore::CSSStyleSelector::SelectorChecker::checkOneSelector):
2724
2725 2010-08-11  Kevin Ollivier  <kevino@theolliviers.com>
2726
2727         [wx, WebDOM] Build fix. Fix callback params and WebCore::String -> WTF::String fixes.
2728
2729         * bindings/cpp/WebDOMCustomVoidCallback.cpp: Added.
2730         (WebDOMCustomVoidCallback::WebDOMCustomVoidCallback):
2731         (WebDOMCustomVoidCallback::~WebDOMCustomVoidCallback):
2732         (WebDOMCustomVoidCallback::handleEvent):
2733         (toWebCore):
2734         * bindings/cpp/WebDOMCustomVoidCallback.h: Added.
2735         (WebDOMCustomVoidCallback::create):
2736         * bindings/cpp/WebDOMHTMLDocumentCustom.cpp:
2737         (documentWrite):
2738         * bindings/cpp/WebDOMString.cpp:
2739         (WebDOMString::equals):
2740         * bindings/scripts/CodeGeneratorCPP.pm:
2741         * bindings/scripts/test/CPP/WebDOMTestCallback.cpp:
2742         (WebDOMTestCallback::callbackWithClass1Param):
2743         (WebDOMTestCallback::callbackWithClass2Param):
2744         (WebDOMTestCallback::callbackWithNonBoolReturnType):
2745         * bindings/scripts/test/CPP/WebDOMTestObj.cpp:
2746         * bindings/scripts/test/V8/V8TestInterface.h:
2747         * bindings/scripts/test/V8/V8TestObj.h:
2748         * wscript:
2749
2750 2010-08-11  Gavin Barraclough  <barraclough@apple.com>
2751
2752         Rubber stamps by Darin Adler & Sam Weinig.
2753
2754         Bug 43867 - Some UString cleanup
2755
2756         Change JSC::UString data(), size(), and from(), to characters(), length(), and number() to match WTF::String.
2757         Move string concatenation methods to a new header to simplify down UString.h.  Remove is8Bit().
2758
2759         * WebCore.xcodeproj/project.pbxproj:
2760         * bindings/js/JSCSSStyleDeclarationCustom.cpp:
2761         (WebCore::hasCSSPropertyNamePrefix):
2762         (WebCore::cssPropertyName):
2763         * bindings/js/JSLocationCustom.cpp:
2764         (WebCore::JSLocation::setPort):
2765         * bindings/js/ScriptDebugServer.cpp:
2766         (WebCore::ScriptDebugServer::dispatchDidParseSource):
2767         * bindings/js/ScriptString.h:
2768         (WebCore::ScriptString::size):
2769         * bindings/js/ScriptValue.cpp:
2770         (WebCore::jsToInspectorValue):
2771         * bindings/objc/WebScriptObject.mm:
2772         (+[WebScriptObject _convertValueToObjcValue:originRootObject:rootObject:]):
2773         * bridge/jni/jni_jsobject.mm:
2774         (JavaJSObject::convertValueToJObject):
2775         * bridge/jni/jsc/JNIUtilityPrivate.cpp:
2776         (JSC::Bindings::convertArrayInstanceToJavaArray):
2777         (JSC::Bindings::convertValueToJValue):
2778         * bridge/objc/objc_runtime.mm:
2779         (JSC::Bindings::callObjCFallbackObject):
2780
2781 2010-08-10  David Hyatt  <hyatt@apple.com>
2782
2783         Reviewed by Dan Bernstein.
2784
2785         https://bugs.webkit.org/show_bug.cgi?id=43806, add ability to paginate screen content.
2786         
2787         Add a new boolean to Document to indicate that content should be paginated.  This is checked when doing page breaks
2788         and in a couple of other cases.
2789
2790         * dom/Document.cpp:
2791         (WebCore::Document::Document):
2792         * dom/Document.h:
2793         (WebCore::Document::paginatedForScreen):
2794         (WebCore::Document::setPaginatedForScreen):
2795         (WebCore::Document::paginated):
2796         * rendering/RenderBlock.cpp:
2797         (WebCore::RenderBlock::paintChildren):
2798         (WebCore::RenderBlock::calcColumnWidth):
2799         * rendering/RenderBox.cpp:
2800         (WebCore::RenderBox::calcHeight):
2801         * rendering/RenderView.cpp:
2802         (WebCore::RenderView::paint):
2803
2804 2010-08-11  Adam Barth  <abarth@webkit.org>
2805
2806         Windows build fix.  Turns out we need to unconditionally include
2807         HTMLEntityNames.cpp now that LegacyHTMLDocumentParser doesn't exist.
2808
2809         * html/HTMLEntityParser.cpp:
2810
2811 2010-08-11  Eric Carlson  <eric.carlson@apple.com>
2812
2813         Reviewed by Darin Adler.
2814
2815         Inserting a <source> element should immediately set networkState to NETWORK_NO_SOURCE.
2816         https://bugs.webkit.org/show_bug.cgi?id=42897
2817
2818         A <source> element inserted as a child of a media element with a networkState of NETWORK_EMPTY
2819         should cause the networkState to immediately change to NETWORK_NO_SOURCE.
2820
2821         * WebCore.xcodeproj/project.pbxproj:
2822         * html/HTMLMediaElement.cpp:
2823         (WebCore::HTMLMediaElement::HTMLMediaElement): Remove unused variable m_processingLoad.
2824         (WebCore::HTMLMediaElement::prepareForLoad): Set m_networkState to NETWORK_NO_SOURCE.
2825         (WebCore::HTMLMediaElement::loadInternal): Remove unused variable m_processingLoad.
2826         (WebCore::HTMLMediaElement::selectMediaResource): No need to change m_networkState, it was done in prepareForLoad.
2827         * html/HTMLMediaElement.h:
2828
2829         * html/HTMLSourceElement.cpp:
2830         (WebCore::HTMLSourceElement::insertedIntoTree): Renamed from insertedIntoDocument because we want
2831         to do this setup whether or not the parent media element is in the document.
2832         * html/HTMLSourceElement.h:
2833
2834 2010-08-11  Adam Barth  <abarth@webkit.org>
2835
2836         Rubber-stamped by Eric Seidel.
2837
2838         Delete LegacyHTMLDocumentParser and LegacyPreloadScanner
2839         https://bugs.webkit.org/show_bug.cgi?id=43836
2840
2841         These classes are bit-rotting disturbingly quickly.  We removed the
2842         last dependencies on these classes and so we can remove them.
2843
2844         We might need to preserve some of their particular quirks to handle
2845         some legacy applications, but our current thinking is to do that with
2846         individual settings that affect the new parser.
2847
2848         * Android.mk:
2849         * CMakeLists.txt:
2850         * GNUmakefile.am:
2851         * WebCore.gypi:
2852         * WebCore.pro:
2853         * WebCore.vcproj/WebCore.vcproj:
2854         * WebCore.xcodeproj/project.pbxproj:
2855         * dom/Document.cpp:
2856         * dom/DocumentFragment.cpp:
2857         (WebCore::DocumentFragment::parseHTML):
2858         * dom/DocumentParser.h:
2859         * dom/XMLDocumentParserLibxml2.cpp:
2860         * dom/XMLDocumentParserQt.cpp:
2861         * html/HTMLConstructionSite.cpp:
2862         * html/HTMLDocument.cpp:
2863         (WebCore::HTMLDocument::createParser):
2864         * html/HTMLEntityParser.cpp:
2865         (WebCore::decodeNamedEntity):
2866         * html/HTMLEntityParser.h:
2867         * html/HTMLFormControlElement.cpp:
2868         * html/HTMLTreeBuilder.cpp:
2869         * html/HTMLViewSourceDocument.cpp:
2870         (WebCore::HTMLViewSourceDocument::createParser):
2871         * html/LegacyHTMLDocumentParser.cpp: Removed.
2872         * html/LegacyHTMLDocumentParser.h: Removed.
2873         * html/LegacyHTMLTreeBuilder.cpp:
2874         * html/LegacyPreloadScanner.cpp: Removed.
2875         * html/LegacyPreloadScanner.h: Removed.
2876         * loader/FTPDirectoryDocument.cpp:
2877         (WebCore::FTPDirectoryDocumentParser::FTPDirectoryDocumentParser):
2878         (WebCore::FTPDirectoryDocumentParser::appendEntry):
2879         (WebCore::FTPDirectoryDocumentParser::createTDForFilename):
2880         (WebCore::processFilesizeString):
2881         (WebCore::wasLastDayOfMonth):
2882         (WebCore::processFileDateString):
2883         (WebCore::FTPDirectoryDocumentParser::parseAndAppendOneLine):
2884         (WebCore::FTPDirectoryDocumentParser::loadDocumentTemplate):
2885         (WebCore::FTPDirectoryDocumentParser::createBasicDocument):
2886         (WebCore::FTPDirectoryDocumentParser::append):
2887         (WebCore::FTPDirectoryDocumentParser::finish):
2888         * page/XSSAuditor.cpp:
2889         (WebCore::XSSAuditor::decodeHTMLEntities):
2890         * page/XSSAuditor.h:
2891         * platform/text/SegmentedString.cpp:
2892         (WebCore::SegmentedString::prepend):
2893
2894 2010-08-11  Chris Jerdonek  <cjerdonek@webkit.org>
2895
2896         Reviewed by Adam Barth.
2897
2898         Refactored FrameLoader::createWindow() out of the FrameLoader class
2899         since FrameLoader should not be responsible for creating windows.
2900
2901         https://bugs.webkit.org/show_bug.cgi?id=39156
2902
2903         No new functionality, so no new tests.
2904
2905         * bindings/generic/BindingDOMWindow.h:
2906         (WebCore::::createWindow):
2907           - Updated the call to FrameLoader's createWindow().
2908         * bindings/js/JSDOMWindowCustom.cpp:
2909         (WebCore::createWindow):
2910           - Updated the call to FrameLoader's createWindow().
2911         * loader/FrameLoader.cpp:
2912         (WebCore::createWindowFromFrameLoader):
2913           - Made FrameLoader::createWindow() non-member, non-friend and changed it
2914             to accept Frame* parameters instead of FrameLoader*.
2915         * loader/FrameLoader.h:
2916           - Updated the header file to reflect the above change to FrameLoader.cpp.
2917
2918 2010-08-11  Ilya Tikhonovsky  <loislo@chromium.org>
2919
2920         Reviewed by Pavel Feldman.
2921
2922         WebInspector: small refactoring for CodeGeneratorInspector.
2923         all "notify" methods are generating in InspectorFrontend
2924         the other methods are generating in InspectorBackend.
2925         https://bugs.webkit.org/show_bug.cgi?id=43849
2926
2927         * inspector/CodeGeneratorInspector.pm:
2928
2929 2010-08-11  Gustavo Noronha Silva  <gustavo.noronha@collabora.co.uk>
2930
2931         Reviewed by Xan Lopez.
2932
2933         [GTK] Misc improvements to the scrolling code
2934         https://bugs.webkit.org/show_bug.cgi?id=41926
2935
2936         Refactor code that uses deprecated ways of dealing with the
2937         adjustments to use the GtkVersioning infrastructure.
2938
2939         * GNUmakefile.am:
2940         * platform/gtk/GtkVersioning.cpp: Added.
2941         * platform/gtk/GtkVersioning.h:
2942         * platform/gtk/ScrollViewGtk.cpp:
2943         * platform/gtk/ScrollbarGtk.cpp:
2944         (ScrollbarGtk::detachAdjustment):
2945         (ScrollbarGtk::updateThumbPosition):
2946         (ScrollbarGtk::updateThumbProportion):
2947
2948 2010-08-11  Nate Chapin  <japhet@chromium.org>
2949
2950         Chromium mac build fix (variables weren't being initialized).
2951
2952         * platform/graphics/filters/FEGaussianBlur.cpp: 
2953         (WebCore::FEGaussianBlur::apply): Initalize dxLeft, dxRight, dyLeft, dyRight.
2954
2955 2010-08-11  Pavel Feldman  <pfeldman@chromium.org>
2956
2957         Reviewed by Yury Semikhatsky.
2958
2959         Web Inspector: apply correct fix to storage panel status bar buttons.
2960
2961         https://bugs.webkit.org/show_bug.cgi?id=43846
2962
2963         * inspector/front-end/DatabaseTableView.js:
2964         (WebInspector.DatabaseTableView.prototype.get statusBarItems):
2965         * inspector/front-end/StoragePanel.js:
2966         (WebInspector.StoragePanel.prototype._genericViewSetup):
2967
2968 2010-08-11  Kevin Ollivier  <kevino@theolliviers.com>
2969
2970         Reviewed by Adam Barth.
2971
2972         Override operator= for C++ DOM binding classes with private structs
2973         and update C++ DOM bindings test results.
2974         https://bugs.webkit.org/show_bug.cgi?id=43735
2975
2976         * bindings/cpp/WebDOMEventTarget.cpp:
2977         (WebDOMEventTarget::operator=):
2978         * bindings/cpp/WebDOMEventTarget.h:
2979         * bindings/scripts/CodeGeneratorCPP.pm:
2980         * bindings/scripts/test/CPP/WebDOMTestCallback.cpp:
2981         (WebDOMTestCallback::operator=):
2982         * bindings/scripts/test/CPP/WebDOMTestCallback.h:
2983         * bindings/scripts/test/CPP/WebDOMTestInterface.cpp:
2984         (WebDOMTestInterface::operator=):
2985         * bindings/scripts/test/CPP/WebDOMTestInterface.h:
2986         * bindings/scripts/test/CPP/WebDOMTestObj.cpp:
2987         (WebDOMTestObj::operator=):
2988         * bindings/scripts/test/CPP/WebDOMTestObj.h:
2989
2990 2010-08-11  Julie Parent  <jparent@chromium.org>
2991
2992         Reviewed by Justin Garcia.
2993
2994         Crash in replaceSelectionCommand with RTL text.
2995         https://bugs.webkit.org/show_bug.cgi?id=41485
2996         
2997         For text with mixed directionality, sort the text boxes before
2998         computing gaps, since that code assumes the boxes are in order.
2999
3000         Test: editing/execCommand/insert-image-on-top-of-directional-text.html
3001
3002         * editing/CompositeEditCommand.cpp:
3003         (WebCore::CompositeEditCommand::deleteInsignificantText): Sort boxes
3004         like we do in TextIterator before computing gaps.
3005         * editing/TextIterator.cpp:
3006         (WebCore::TextIterator::handleTextNode): Use new compareByStart
3007         rather than compareBoxStart.  No functional change.
3008         * rendering/InlineTextBox.h:
3009         (WebCore::InlineTextBox::compareByStart): Moved compareBoxStart
3010         from TextIterator here so it can be used in multiple places.
3011
3012 2010-08-11  Fumitoshi Ukai  <ukai@chromium.org>
3013
3014         Unreviewed build fix of Leopard Intel Debug (Build)
3015
3016         * websockets/WebSocketChannel.cpp:
3017         (WebCore::WebSocketChannel::processBuffer): use %lu, instead of %ul
3018
3019 2010-08-11  Alejandro G. Castro  <alex@igalia.com>
3020
3021         Reviewed by Dirk Schulze.
3022
3023         Fixes to the gaussian blur algorithm
3024         https://bugs.webkit.org/show_bug.cgi?id=41605
3025
3026         Added the recomendations of SVG regarding the kernel size
3027         calculation for the approximation of the gaussian
3028         blur. http://www.w3.org/TR/SVG/filters.html#feGaussianBlurElement.
3029         Modified also some variable names to clarify algorithm, reviewed
3030         the situation where both standard deviations are 0.
3031
3032         * platform/graphics/filters/FEGaussianBlur.cpp:
3033         (WebCore::FEGaussianBlur::FEGaussianBlur):
3034         (WebCore::FEGaussianBlur::stdDeviationX):
3035         (WebCore::FEGaussianBlur::setStdDeviationX):
3036         (WebCore::FEGaussianBlur::stdDeviationY):
3037         (WebCore::FEGaussianBlur::setStdDeviationY):
3038         (WebCore::boxBlur):
3039         (WebCore::FEGaussianBlur::kernelPosition): Added the SVG kernel
3040         position recomendation.
3041         (WebCore::FEGaussianBlur::apply):
3042         (WebCore::FEGaussianBlur::externalRepresentation):
3043         * platform/graphics/filters/FEGaussianBlur.h:
3044
3045 2010-08-11  Kristian Monsen  <kristianm@google.com>
3046
3047         Reviewed by Steve Block.
3048
3049         Compile fix for Android, added include for <wtf/text/Cstring.h, this
3050         did not get included in Android.
3051         https://bugs.webkit.org/show_bug.cgi?id=41517
3052         
3053         Build fix only, no new tests.
3054
3055         * page/PrintContext.cpp:
3056
3057 2010-08-11  Fumitoshi Ukai  <ukai@chromium.org>
3058
3059         Reviewed by Alexey Proskuryakov.
3060
3061         Fix length calculation to be more robust.
3062         https://bugs.webkit.org/show_bug.cgi?id=43777
3063
3064         Test: websocket/tests/frame-length-overflow.html
3065
3066         * websockets/WebSocketChannel.cpp:
3067         (WebCore::WebSocketChannel::appendToBuffer): len is size_t.
3068          - add sanity check for integer wraps.
3069         (WebCore::WebSocketChannel::skipBuffer): len is size_t.
3070         (WebCore::WebSocketChannel::processBuffer): length is size_t.
3071          - add sanity check for integer wraps.
3072         * websockets/WebSocketChannel.h: change m_bufferSize and len to size_t.
3073
3074 2010-08-10  Eric Seidel  <eric@webkit.org>
3075
3076         Reviewed by Adam Barth.
3077
3078         Fix three ASSERTs hit with the HTML5 TreeBuilder in fragment mode
3079         https://bugs.webkit.org/show_bug.cgi?id=43762
3080
3081         In fixing the insertAdjacentHTML I had to make small changes to
3082         the error logic, which ended up bringing us closer to HTML5.
3083
3084         Test: fast/dom/HTMLElement/insertAdjacentHTML-errors.html
3085
3086         * html/HTMLConstructionSite.cpp:
3087         (WebCore::HTMLConstructionSite::attach):
3088          - This ASSERT was just wrong, the child should not be attached
3089            in the case the parent is not attached.
3090         * html/HTMLElement.cpp:
3091         (WebCore::contextElementForInsertion):
3092          - This is part of HTML5, needed to know what to call the
3093            HTML or XML parsers with.  The previous code always passed
3094            "this" which is wrong in the case of beforeBegin or afterEnd insertion.
3095         (WebCore::HTMLElement::insertAdjacentHTML):
3096          - Use the right contextElement now that we know how to compute it.
3097         * html/HTMLTreeBuilder.cpp:
3098         (WebCore::HTMLTreeBuilder::processEndTagForInCell):
3099          - This ASSERT is direct from the spec, but it's wrong, so I've filed a spec bug.
3100
3101 2010-08-11  Adam Barth  <abarth@webkit.org>
3102
3103         Attempt to fix Windows build.
3104
3105         * platform/win/COMPtr.h:
3106         (WTF::):
3107
3108 2010-08-11  Adam Barth  <abarth@webkit.org>
3109
3110         Reviewed by Eric Seidel.
3111
3112         Port view-source to new parser
3113         https://bugs.webkit.org/show_bug.cgi?id=43746
3114
3115         This patch switches the view-source mode for frames over to using the
3116         new HTML parsing infrastructure.  This patch is an architectural change
3117         to how we parser view source documents.
3118
3119         Previously, the LegacyHTMLDocumentParser would output a "guide string"
3120         that consided of the inter-attribute whitespace and various "control"
3121         characters.  The HTMLViewSourceDocument would then interpret this guide
3122         string to approximately reconstruct the source of the original document
3123         and colorize various syntatic constructs.
3124
3125         Unfortunately, that approach is inherently low-fidelity.  It's not
3126         really feasible to reconstruct the input document from the token
3127         stream.  The old view source mode also had a number of hacks in the old
3128         parser (e.g., to turn of decoding of HTML entities).
3129
3130         Instead of trying to reconstruct the original document from the token
3131         stream, we use the segmentation information given by the tokens to
3132         colorize the input document itself.  Each token now caries information
3133         about where in the input stream it came from and where various
3134         subcomponents (e.g., attribute names and values) are located.  This
3135         approach is higher fidelity because we use this segmentation
3136         information to colorize the original input instead of attempting to
3137         reconstruct the original input.
3138
3139         * Android.mk:
3140         * CMakeLists.txt:
3141         * GNUmakefile.am:
3142         * WebCore.gypi:
3143         * WebCore.pro:
3144         * WebCore.vcproj/WebCore.vcproj:
3145         * WebCore.xcodeproj/project.pbxproj:
3146         * html/HTMLDocumentParser.cpp:
3147         * html/HTMLToken.h:
3148         (WebCore::HTMLToken::clear):
3149         (WebCore::HTMLToken::startIndex):
3150         (WebCore::HTMLToken::length):
3151         (WebCore::HTMLToken::end):
3152         * html/HTMLViewSourceDocument.cpp:
3153         (WebCore::HTMLViewSourceDocument::createParser):
3154         (WebCore::HTMLViewSourceDocument::addSource):
3155         (WebCore::HTMLViewSourceDocument::processDoctypeToken):
3156         (WebCore::HTMLViewSourceDocument::processTagToken):
3157         (WebCore::HTMLViewSourceDocument::processCommentToken):
3158         (WebCore::HTMLViewSourceDocument::processCharacterToken):
3159         (WebCore::HTMLViewSourceDocument::addRange):
3160         * html/HTMLViewSourceDocument.h:
3161         * html/HTMLViewSourceParser.cpp: Added.
3162         (WebCore::HTMLViewSourceParser::~HTMLViewSourceParser):
3163         (WebCore::HTMLViewSourceParser::insert):
3164         (WebCore::HTMLViewSourceParser::pumpTokenizer):
3165         (WebCore::HTMLViewSourceParser::append):
3166         (WebCore::HTMLViewSourceParser::sourceForToken):
3167         (WebCore::HTMLViewSourceParser::updateTokenizerState):
3168         (WebCore::HTMLViewSourceParser::finish):
3169         (WebCore::HTMLViewSourceParser::finishWasCalled):
3170         * html/HTMLViewSourceParser.h: Added.
3171         (WebCore::HTMLViewSourceParser::HTMLViewSourceParser):
3172         (WebCore::HTMLViewSourceParser::document):
3173         * html/LegacyHTMLDocumentParser.cpp:
3174         (WebCore::LegacyHTMLDocumentParser::processToken):
3175         (WebCore::LegacyHTMLDocumentParser::processDoctypeToken):
3176
3177 2010-08-11  Yoshiki Hayashi  <yhayashi@google.com>
3178
3179         Reviewed by Shinichiro Hamaji.
3180
3181         Make form controls, textarea, legend and datagrid treat width:
3182         auto as specified in stretching vertical flexbox.  Otherwise it's
3183         treated as intrinsic as before.
3184         https://bugs.webkit.org/show_bug.cgi?id=17918
3185
3186         Test: fast/flexbox/vertical-box-form-controls.html
3187
3188         * css/CSSStyleSelector.cpp:
3189         (WebCore::CSSStyleSelector::adjustRenderStyle):
3190         * rendering/RenderBox.cpp:
3191         (WebCore::RenderBox::sizesToIntrinsicWidth):
3192
3193 2010-08-10  Darin Adler  <darin@apple.com>
3194
3195         Reviewed by Sam Weinig.
3196
3197         Add leakRef and clear to all RefPtr variants
3198         https://bugs.webkit.org/show_bug.cgi?id=42389
3199
3200         * platform/win/COMPtr.h: Changed all uses of "template <...>" to instead do
3201         "template<...>". Defined many of the inlined functions outside the class
3202         definition, to avoid style checker warnings about multiple statements on
3203         a single line and for slightly better clarity of the class definition itself.
3204         Renamed releaseRef to leakRef. Added a releaseRef that calls leakRef so we
3205         don't have to rename all callers at once. Added a clear function. Changed
3206         the hash table code so it doesn't need to call releaseRef, and so it uses
3207         the hash table deleted value hooks already present within the class.
3208
3209 2010-08-10  Ariya Hidayat  <ariya@sencha.com>
3210
3211         [Qt] Fix build warning: remove reference to html/BlobRegistryImp.h
3212
3213         * WebCore.pro:
3214
3215 2010-08-10  Alexey Proskuryakov  <ap@apple.com>
3216
3217         Reviewed by Darin Adler.
3218
3219         https://bugs.webkit.org/show_bug.cgi?id=43803
3220         <rdar://problem/6988023> Only the last Korean character can be shown if font style is enabled while typing (43803)
3221
3222         Test: platform/mac/editing/input/bold-node.html
3223
3224         * editing/Editor.cpp: (WebCore::Editor::setComposition): It is not clear to me what the
3225         "baseNode == extentNode" check is about in this function , but let's try to make it more
3226         likely that it passes.
3227
3228 2010-08-10  Hayato Ito  <hayato@chromium.org>
3229
3230         Reviewed by Kent Tamura.
3231
3232         Make CSSSelector::specificity() non-recursive.
3233         https://bugs.webkit.org/show_bug.cgi?id=43784
3234
3235         This change is one of the required changes to fix the following master bug:
3236         https://bugs.webkit.org/show_bug.cgi?id=42806
3237
3238         No functional change, thus no tests.
3239
3240         * css/CSSSelector.cpp:
3241         (WebCore::CSSSelector::specificity):
3242         (WebCore::CSSSelector::specificityForOneSelector):
3243         (WebCore::CSSSelector::specificityForPage):
3244         * css/CSSSelector.h:
3245
3246 2010-08-10  Matthew Delaney  <mdelaney@apple.com>
3247
3248         Reviewed by Darin Adler.
3249
3250         2d.path.clip.empty.html test is failing
3251         https://bugs.webkit.org/show_bug.cgi?id=43161
3252
3253         * platform/graphics/cg/GraphicsContextCG.cpp:
3254         (WebCore::GraphicsContext::clip):
3255         Catching the empty path case from being sent directly to 
3256         CGContextClip - which would ignore it. Instead, using 
3257         CGContextClip with a CGRectZero to achieve the desired
3258         behavior of reducing the clipping region to nothing.
3259
3260 2010-08-10  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
3261
3262         Reviewed by Antonio Gomes.
3263
3264         [EFL] Add video files to enable video in WebKit EFL
3265         https://bugs.webkit.org/show_bug.cgi?id=43790
3266
3267         Add files regarding html5 video to CMakeList.txt for WebKit Efl.
3268         But, this patch is not run html5 video on WebKit EFL yet.
3269
3270         * CMakeLists.txt: Add files related to html5video.
3271         * platform/efl/LocalizedStringsEfl.cpp: Add functions for html5 video.
3272         (WebCore::localizedMediaControlElementString):
3273         (WebCore::localizedMediaControlElementHelpText):
3274         (WebCore::localizedMediaTimeDescription):
3275         * platform/graphics/MediaPlayer.cpp:
3276         (WebCore::installedMediaEngines): Disable MediaPlayerPrivate::r
3277       egisterMediaEngine for WebKit EFL.
3278
3279 2010-08-10  Patrick Gansterer  <paroga@paroga.com>
3280
3281         Reviewed by Nikolas Zimmermann.
3282
3283         Add a toRenderSVGImage function
3284         https://bugs.webkit.org/show_bug.cgi?id=43789
3285
3286         Add a function for casting to RenderSVGImage and use it.
3287
3288         * rendering/RenderSVGImage.h:
3289         (WebCore::toRenderSVGImage):
3290         * rendering/RenderTreeAsText.cpp:
3291         (WebCore::write):
3292         * rendering/SVGRenderTreeAsText.cpp:
3293         (WebCore::writeSVGImage):
3294         * rendering/SVGRenderTreeAsText.h:
3295         * svg/SVGImageElement.cpp:
3296         (WebCore::SVGImageElement::attach):
3297
3298 2010-08-10  Adam Barth  <abarth@webkit.org>
3299
3300         Reviewed by Eric Seidel.
3301
3302         Clients of HTMLTokenizer should be able to see where characters went in the token
3303         https://bugs.webkit.org/show_bug.cgi?id=43766
3304
3305         When viewing the source of a document, we want to colorize different
3306         parts of the input depending on how they were tokenized.  In this
3307         patch, we expose the internal segmentation of a token by recording the
3308         start and end offsets for each attribute name and each attribute value.
3309
3310         * html/HTMLToken.h:
3311         (WebCore::HTMLToken::addNewAttribute):
3312         (WebCore::HTMLToken::beginAttributeName):
3313         (WebCore::HTMLToken::endAttributeName):
3314         (WebCore::HTMLToken::beginAttributeValue):
3315         (WebCore::HTMLToken::endAttributeValue):
3316         (WebCore::HTMLToken::appendToAttributeName):
3317         (WebCore::HTMLToken::appendToAttributeValue):
3318         (WebCore::AtomicHTMLToken::AtomicHTMLToken):
3319         * html/HTMLTokenizer.cpp:
3320         (WebCore::HTMLTokenizer::nextToken):
3321
3322 2010-08-10  Dumitru Daniliuc  <dumi@chromium.org>
3323
3324         Reviewed by David Levin.
3325
3326         Do not post a sync task to the DB thread if it's terminating.
3327         https://bugs.webkit.org/show_bug.cgi?id=43676
3328
3329         * bindings/generic/ActiveDOMCallback.cpp: The context can
3330         sometimes be NULL. Check if it's NULL instead of asserting that
3331         it's not.
3332         (WebCore::ActiveDOMCallback::ActiveDOMCallback):
3333         * dom/ActiveDOMObject.cpp:
3334         (WebCore::ActiveDOMObject::ActiveDOMObject):
3335
3336         * storage/Database.cpp: Do not post a sync task to the DB thread
3337         if the DB thread is terminating. Also, when a task is destroyed,
3338         ASSERT that it was completed, or that it had no synchronizer.
3339         (WebCore::Database::openAndVerifyVersion):
3340         (WebCore::Database::markAsDeletedAndClose):
3341         (WebCore::Database::tableNames):
3342         * storage/DatabaseTask.cpp:
3343         (WebCore::DatabaseTaskSynchronizer::DatabaseTaskSynchronizer):
3344         (WebCore::DatabaseTask::~DatabaseTask):
3345         (WebCore::DatabaseTask::performTask):
3346         * storage/DatabaseTask.h:
3347         (WebCore::DatabaseTaskSynchronizer::hasCheckedForTermination):
3348         (WebCore::DatabaseTaskSynchronizer::setHasCheckedForTermination):
3349         (WebCore::DatabaseTask::hasSynchronizer):
3350         (WebCore::DatabaseTask::hasCheckedForTermination):
3351         * storage/DatabaseThread.cpp:
3352         (WebCore::DatabaseThread::terminationRequested):
3353         (WebCore::DatabaseThread::scheduleTask):
3354         (WebCore::DatabaseThread::scheduleImmediateTask):
3355         * storage/DatabaseThread.h:
3356
3357 2010-08-10  Gavin Barraclough  <barraclough@apple.com>
3358
3359         Qt build fix.
3360
3361         * bridge/qt/qt_runtime.cpp:
3362         (JSC::Bindings::convertValueToQVariant):
3363
3364 2010-08-10  Jian Li  <jianli@chromium.org>
3365
3366         Fix more build break.
3367
3368         * bindings/v8/SerializedScriptValue.cpp:
3369         (WebCore::ZigZag::Reader::readBlob):
3370         (WebCore::ZigZag::Reader::readFile):
3371         * dom/ScriptExecutionContext.cpp:
3372         (WebCore::ScriptExecutionContext::~ScriptExecutionContext):
3373         (WebCore::ScriptExecutionContext::revokeBlobURL):
3374         * dom/ScriptExecutionContext.h:
3375         * html/Blob.cpp:
3376         (WebCore::Blob::slice):
3377
3378 2010-08-10  Gavin Barraclough  <barraclough@apple.com>
3379
3380         Rubber stamped by Sam Weinig
3381
3382         Bug 43817 - Remove UString::Rep
3383         UString::Rep has for a long time been replaced by UStringImpl (Rep
3384         remaining as a typedef).  UStringImpl has since been removed too
3385         (unified with StringImpl). Remove Rep, rename rep() to impl() and
3386         m_rep to m_impl.  Also add impl() method to Identifier, and rename
3387         its UString member from _ustring to m_string.
3388
3389         * bindings/js/JSDOMBinding.cpp:
3390         (WebCore::findAtomicString):
3391         * bindings/js/JSDOMBinding.h:
3392         (WebCore::ustringToString):
3393         (WebCore::identifierToString):
3394         (WebCore::ustringToAtomicString):
3395         (WebCore::identifierToAtomicString):
3396         * bindings/js/SerializedScriptValue.h:
3397         (WebCore::SerializedScriptValueData::SerializedScriptValueData):
3398         * bridge/IdentifierRep.cpp:
3399         (WebCore::IdentifierRep::get):
3400         * bridge/c/c_class.cpp:
3401         (JSC::Bindings::CClass::methodsNamed):
3402         (JSC::Bindings::CClass::fieldNamed):
3403         * bridge/jni/jsc/JavaClassJSC.cpp:
3404         (JavaClass::JavaClass):
3405         (JavaClass::methodsNamed):
3406         (JavaClass::fieldNamed):
3407         * bridge/jni/jsc/JavaStringJSC.h:
3408         (JSC::Bindings::JavaStringImpl::~JavaStringImpl):
3409         (JSC::Bindings::JavaStringImpl::init):
3410         (JSC::Bindings::JavaStringImpl::UTF8String):
3411         (JSC::Bindings::JavaStringImpl::uchars):
3412         (JSC::Bindings::JavaStringImpl::length):
3413         (JSC::Bindings::JavaStringImpl::uString):
3414         * bridge/jsc/BridgeJSC.h:
3415
3416 2010-08-10  Jian Li  <jianli@chromium.org>
3417
3418         Fix qt build break.
3419
3420         * html/BlobRegistryImpl.cpp:
3421         * html/BlobRegistryImpl.h:
3422         * html/BlobURL.cpp:
3423         * html/BlobURL.h:
3424         * platform/BlobRegistry.h:
3425
3426 2010-08-10  Jian Li  <jianli@chromium.org>
3427
3428         Reviewed by David Levin.
3429
3430         Add the support to register the blob data.
3431         https://bugs.webkit.org/show_bug.cgi?id=43600
3432
3433         This is the first step towards adding Blob URL support. The blob has
3434         an internal URL (used by FileReader) and public URLs (used by Blob.url).
3435         The URL and its associated data are managed by BlobRegistry interface.
3436
3437         This patch does not remove the usage of BlobItem. It will be removed
3438         when all the blob data support is hooked up and the new tests will be
3439         added then.
3440
3441         * CMakeLists.txt:
3442         * GNUmakefile.am:
3443         * WebCore.gypi:
3444         * WebCore.pro:
3445         * WebCore.vcproj/WebCore.vcproj:
3446         * WebCore.xcodeproj/project.pbxproj:
3447         * bindings/js/SerializedScriptValue.cpp:
3448         (WebCore::SerializedBlob::create):
3449         (WebCore::SerializedBlob::url):
3450         (WebCore::SerializedBlob::type):
3451         (WebCore::SerializedBlob::size):
3452         (WebCore::SerializedBlob::SerializedBlob):
3453         (WebCore::SerializedFile::create):
3454         (WebCore::SerializedFile::path):
3455         (WebCore::SerializedFile::url):
3456         (WebCore::SerializedFile::type):
3457         (WebCore::SerializedFile::SerializedFile):
3458         (WebCore::SerializedFileList::item):
3459         (WebCore::SerializedFileList::SerializedFileList):
3460         (WebCore::SerializedScriptValueData::SerializedScriptValueData):
3461         (WebCore::SharedSerializedData::asBlob):
3462         (WebCore::SharedSerializedData::asFile):
3463         (WebCore::SerializingTreeWalker::convertIfTerminal):
3464         (WebCore::DeserializingTreeWalker::convertIfTerminal):
3465         (WebCore::TeardownTreeWalker::convertIfTerminal):
3466         * bindings/js/SerializedScriptValue.h:
3467         (WebCore::SerializedScriptValueData::):
3468         (WebCore::SerializedScriptValueData::asString):
3469         (WebCore::SerializedScriptValueData::asBlob):
3470         (WebCore::SerializedScriptValueData::asFile):
3471         * bindings/v8/SerializedScriptValue.cpp:
3472         (WebCore::ZigZag::Writer::writeBlob):
3473         (WebCore::ZigZag::Writer::writeFile):
3474         (WebCore::ZigZag::Writer::writeFileList):
3475         (WebCore::ZigZag::Writer::doWriteUintHelper):
3476         (WebCore::ZigZag::Writer::doWriteUint32):
3477         (WebCore::ZigZag::Writer::doWriteUint64):
3478         (WebCore::ZigZag::Serializer::writeBlob):
3479         (WebCore::ZigZag::Serializer::writeFile):
3480         (WebCore::ZigZag::Reader::readBlob):
3481         (WebCore::ZigZag::Reader::readFile):
3482         (WebCore::ZigZag::Reader::readFileList):
3483         (WebCore::ZigZag::Reader::doReadUintHelper):
3484         (WebCore::ZigZag::Reader::doReadUint32):
3485         (WebCore::ZigZag::Reader::doReadUint64):
3486         * dom/ScriptExecutionContext.cpp:
3487         (WebCore::ScriptExecutionContext::~ScriptExecutionContext):
3488         (WebCore::ScriptExecutionContext::trackBlobURL):
3489         (WebCore::ScriptExecutionContext::revokeBlobURL):
3490         * dom/ScriptExecutionContext.h:
3491         * html/Blob.cpp:
3492         (WebCore::Blob::Blob):
3493         (WebCore::Blob::~Blob):
3494         (WebCore::Blob::createPublicURL):
3495         * html/Blob.h:
3496         (WebCore::Blob::create):
3497         (WebCore::Blob::url):
3498         * html/BlobRegistryImpl.cpp: Added.
3499         * html/BlobRegistryImpl.h: Added.
3500         * html/BlobStorageData.h: Added.
3501         * html/BlobURL.cpp: Added.
3502         * html/BlobURL.h: Added.
3503         * html/File.cpp:
3504         (WebCore::File::File):
3505         * html/File.h:
3506         (WebCore::File::create):
3507         * platform/BlobData.cpp: Added.
3508         * platform/BlobData.h: Added.
3509         * platform/BlobRegistry.h: Added.
3510
3511 2010-08-10  Adam Barth  <abarth@webkit.org>
3512
3513         Reviewed by Eric Seidel.
3514
3515         SegmentedString should keep track of how many characters it consumes
3516         https://bugs.webkit.org/show_bug.cgi?id=43765
3517
3518         The easiest way to keep track of how many characters we've consumed
3519         would be to increment a counter every time we advance, but that's too
3520         slow.  Instead, we keep track lazily and update our summary durning the
3521         slow case of advance (which is sufficiently rare).
3522
3523         There's some subtly to how this works w.r.t. "unconsuming" characters
3524         by pushing them back on the front of the segmented string.  This isn't
3525         really a "right answer" here because the notion of unconsuming input
3526         isn't present in the API.  This patch makes some assumptions about how
3527         clients of this class use its API.  In a future patch, we might want to
3528         rename some of the method names to make this more explicit.
3529
3530         * platform/text/SegmentedString.cpp:
3531         (WebCore::SegmentedString::operator=):
3532         (WebCore::SegmentedString::append):
3533         (WebCore::SegmentedString::prepend):
3534         (WebCore::SegmentedString::advanceSubstring):
3535         * platform/text/SegmentedString.h:
3536         (WebCore::SegmentedSubstring::numberOfCharactersConsumed):
3537         (WebCore::SegmentedString::SegmentedString):
3538         (WebCore::SegmentedString::numberOfCharactersConsumed):
3539
3540 2010-08-10  Gavin Barraclough  <barraclough@apple.com>
3541
3542         Rubber stamped by Sam Weinig.
3543         
3544         Bug 43816 - Remove UStringImpl
3545         The class was actually removed a long time ago, replaced by StringImpl.
3546         UStringImpl is just a typedef onto StringImpl.  Remove this.
3547
3548         * WebCore.xcodeproj/project.pbxproj:
3549         * bindings/js/JSDOMBinding.cpp:
3550         (WebCore::findAtomicString):
3551
3552 2010-08-09  Stephen White  <senorblanco@chromium.org>
3553
3554         Reviewed by David Levin.
3555
3556         Prefer the medium-quality (bilinear) image resampling for 2D canvas on
3557         Chromium.
3558         https://bugs.webkit.org/show_bug.cgi?id=43646
3559         
3560         This is what the accelerated path is going to do,
3561         so do it in the software path as well.
3562
3563         * html/HTMLCanvasElement.cpp:
3564         (WebCore::HTMLCanvasElement::createImageBuffer):
3565         * platform/graphics/GraphicsContext.h:
3566
3567 2010-08-10  Chris Fleizach  <cfleizach@apple.com>
3568
3569         Reviewed by Dave Hyatt.
3570
3571         Removing an element from an anonymous block causes crash
3572         https://bugs.webkit.org/show_bug.cgi?id=42309
3573
3574         There was a case where a continuation was added as a child, but if you asked
3575         that child who is your parent, it would return the wrong answer.
3576    
3577         The specific scenario was when a sibling of an element who was the start of a 
3578         continuation was present. Retrieving the parent object had then follow the sibling
3579         chain and then follow the originating continuation chain.
3580
3581         Test: accessibility/removed-anonymous-block-child-causes-crash.html
3582
3583         * accessibility/AccessibilityRenderObject.cpp:
3584         (WebCore::AccessibilityRenderObject::nextSibling):
3585             Fix erroneous comment
3586         (WebCore::nextContinuation):
3587         (WebCore::AccessibilityRenderObject::renderParentObject):
3588             Handle unhandled continuation case.
3589         (WebCore::AccessibilityRenderObject::addChildren):
3590             ASSERT that the parentObject() is the same when adding a new child.
3591
3592 2010-08-10  Abhishek Arya  <inferno@chromium.org>
3593
3594         Reviewed by David Hyatt.
3595
3596         Take checks for ruby base existence out of&nb