Source/WebCore/ChangeLog

   1 2012-09-24  Tony Chang  <tony@chromium.org>
   2
   3         Replace 2 uses of updateLogicalHeight with computeLogicalHeight
   4         https://bugs.webkit.org/show_bug.cgi?id=97486
   5
   6         Reviewed by Ojan Vafai.
   7
   8         More work for bug 96804.  This is just a refactoring.
   9
  10         No new tests, behavior should be the same as before.
  11
  12         * rendering/RenderFlowThread.cpp:
  13         (WebCore::RenderFlowThread::computeLogicalHeight):
  14         * rendering/RenderFlowThread.h:
  15         * rendering/RenderMultiColumnFlowThread.cpp:
  16         (WebCore::RenderMultiColumnFlowThread::computeLogicalHeight):
  17         (WebCore):
  18         * rendering/RenderMultiColumnFlowThread.h:
  19         (RenderMultiColumnFlowThread):
  20
  21 2012-09-21  Kenneth Russell  <kbr@google.com>
  22
  23         [V8] ArrayBuffer code should not pass a negative length to SetIndexedPropertiesToExternalArrayData()
  24         https://bugs.webkit.org/show_bug.cgi?id=96703
  25
  26         Reviewed by Adam Barth.
  27
  28         Check length arguments that may be passed to SetIndexedPropertiesToExternalArrayData.
  29
  30         No tests because it is not guaranteed that buffers this large can actually be allocated.
  31
  32         * bindings/v8/custom/V8ArrayBufferViewCustom.h:
  33         (WebCore::wrapArrayBufferView):
  34         (WebCore::constructWebGLArrayWithArrayBufferArgument):
  35         (WebCore::constructWebGLArray):
  36
  37 2012-09-24  Antti Koivisto  <antti@apple.com>
  38
  39         Split stylesheet related code out from Document
  40         https://bugs.webkit.org/show_bug.cgi?id=97353
  41
  42         Reviewed by Andreas Kling.
  43
  44         Document is big and unwieldy. The code related to tracking active stylesheets can be factored out.
  45
  46         The patch moves stylesheet upkeep, collection and invalidation code from Document to a separate
  47         DocumentStyleSheetCollection class. There are no functional changes.
  48
  49         The usesLinkRules stylesheet feature bit and the related code is removed as no one was using it.
  50
  51         * WebCore.xcodeproj/project.pbxproj:
  52         * css/SelectorChecker.cpp:
  53         (WebCore::SelectorChecker::checkOneSelector):
  54         * css/StyleResolver.cpp:
  55         (WebCore::StyleResolver::StyleResolver):
  56         (WebCore::StyleResolver::Features::Features):
  57         (WebCore::StyleResolver::Features::add):
  58         (WebCore::StyleResolver::Features::clear):
  59         (WebCore::StyleResolver::collectMatchingRulesForList):
  60         * css/StyleResolver.h:
  61         (WebCore::StyleResolver::usesBeforeAfterRules):
  62         (Features):
  63         * dom/Document.cpp:
  64         (WebCore::Document::Document):
  65         (WebCore::Document::~Document):
  66         (WebCore::Document::setCompatibilityMode):
  67         (WebCore::Document::recalcStyle):
  68         (WebCore):
  69         (WebCore::Document::createStyleResolver):
  70         (WebCore::Document::seamlessParentUpdatedStylesheets):
  71         (WebCore::Document::didRemoveAllPendingStylesheet):
  72         (WebCore::Document::processHttpEquiv):
  73         (WebCore::Document::styleSheets):
  74         (WebCore::Document::preferredStylesheetSet):
  75         (WebCore::Document::selectedStylesheetSet):
  76         (WebCore::Document::setSelectedStylesheetSet):
  77         (WebCore::Document::styleResolverChanged):
  78         (WebCore::Document::reportMemoryUsage):
  79         (WebCore::Document::haveStylesheetsLoaded):
  80         * dom/Document.h:
  81         (WebCore):
  82         (Document):
  83         (WebCore::Document::styleSheetCollection):
  84         (WebCore::Document::hasNodesWithPlaceholderStyle):
  85         (WebCore::Document::setNeedsNotifyRemoveAllPendingStylesheet):
  86         (WebCore::Document::inStyleRecalc):
  87         * dom/DocumentStyleSheetCollection.cpp: Added.
  88         (WebCore):
  89         (WebCore::DocumentStyleSheetCollection::DocumentStyleSheetCollection):
  90         (WebCore::DocumentStyleSheetCollection::~DocumentStyleSheetCollection):
  91         (WebCore::DocumentStyleSheetCollection::pageUserSheet):
  92         (WebCore::DocumentStyleSheetCollection::clearPageUserSheet):
  93         (WebCore::DocumentStyleSheetCollection::updatePageUserSheet):
  94         (WebCore::DocumentStyleSheetCollection::pageGroupUserSheets):
  95         (WebCore::DocumentStyleSheetCollection::clearPageGroupUserSheets):
  96         (WebCore::DocumentStyleSheetCollection::updatePageGroupUserSheets):
  97         (WebCore::DocumentStyleSheetCollection::addUserSheet):
  98         (WebCore::DocumentStyleSheetCollection::removePendingSheet):
  99         (WebCore::DocumentStyleSheetCollection::addStyleSheetCandidateNode):
 100         (WebCore::DocumentStyleSheetCollection::removeStyleSheetCandidateNode):
 101         (WebCore::DocumentStyleSheetCollection::collectActiveStyleSheets):
 102         (WebCore::DocumentStyleSheetCollection::testAddedStyleSheetRequiresStyleRecalc):
 103         (WebCore::DocumentStyleSheetCollection::analyzeStyleSheetChange):
 104         (WebCore::styleSheetsUseRemUnits):
 105         (WebCore::DocumentStyleSheetCollection::updateActiveStyleSheets):
 106         (WebCore::DocumentStyleSheetCollection::combineCSSFeatureFlags):
 107         (WebCore::DocumentStyleSheetCollection::resetCSSFeatureFlags):
 108         (WebCore::DocumentStyleSheetCollection::reportMemoryUsage):
 109         * dom/DocumentStyleSheetCollection.h: Added.
 110         (WebCore):
 111         (DocumentStyleSheetCollection):
 112         (WebCore::DocumentStyleSheetCollection::authorStyleSheets):
 113         (WebCore::DocumentStyleSheetCollection::documentUserSheets):
 114         (WebCore::DocumentStyleSheetCollection::needsUpdateActiveStylesheetsOnStyleRecalc):
 115         (WebCore::DocumentStyleSheetCollection::preferredStylesheetSetName):
 116         (WebCore::DocumentStyleSheetCollection::selectedStylesheetSetName):
 117         (WebCore::DocumentStyleSheetCollection::setPreferredStylesheetSetName):
 118         (WebCore::DocumentStyleSheetCollection::setSelectedStylesheetSetName):
 119         (WebCore::DocumentStyleSheetCollection::addPendingSheet):
 120         (WebCore::DocumentStyleSheetCollection::hasPendingSheets):
 121         (WebCore::DocumentStyleSheetCollection::usesSiblingRules):
 122         (WebCore::DocumentStyleSheetCollection::setUsesSiblingRulesOverride):
 123         (WebCore::DocumentStyleSheetCollection::usesFirstLineRules):
 124         (WebCore::DocumentStyleSheetCollection::usesFirstLetterRules):
 125         (WebCore::DocumentStyleSheetCollection::setUsesFirstLetterRules):
 126         (WebCore::DocumentStyleSheetCollection::usesBeforeAfterRules):
 127         (WebCore::DocumentStyleSheetCollection::setUsesBeforeAfterRulesOverride):
 128         (WebCore::DocumentStyleSheetCollection::usesRemUnits):
 129         (WebCore::DocumentStyleSheetCollection::setUsesRemUnit):
 130         * dom/Element.cpp:
 131         (WebCore::Element::recalcStyle):
 132         * dom/ProcessingInstruction.cpp:
 133         (WebCore::ProcessingInstruction::~ProcessingInstruction):
 134         (WebCore::ProcessingInstruction::checkStyleSheet):
 135         (WebCore::ProcessingInstruction::sheetLoaded):
 136         (WebCore::ProcessingInstruction::insertedInto):
 137         (WebCore::ProcessingInstruction::removedFrom):
 138         * dom/StyleElement.cpp:
 139         (WebCore::StyleElement::insertedIntoDocument):
 140         (WebCore::StyleElement::removedFromDocument):
 141         (WebCore::StyleElement::clearDocumentData):
 142         (WebCore::StyleElement::createSheet):
 143         (WebCore::StyleElement::sheetLoaded):
 144         (WebCore::StyleElement::startLoadingDynamicSheet):
 145         * html/HTMLLinkElement.cpp:
 146         (WebCore::HTMLLinkElement::~HTMLLinkElement):
 147         (WebCore::HTMLLinkElement::insertedInto):
 148         (WebCore::HTMLLinkElement::removedFrom):
 149         (WebCore::HTMLLinkElement::addPendingSheet):
 150         (WebCore::HTMLLinkElement::removePendingSheet):
 151         * html/HTMLQuoteElement.cpp:
 152         (WebCore::HTMLQuoteElement::insertedInto):
 153         * html/HTMLViewSourceDocument.cpp:
 154         (WebCore::HTMLViewSourceDocument::HTMLViewSourceDocument):
 155         * mathml/MathMLMathElement.cpp:
 156         (WebCore::MathMLMathElement::insertedInto):
 157         * page/Page.cpp:
 158         (WebCore::Page::userStyleSheetLocationChanged):
 159         * page/PageGroup.cpp:
 160         (WebCore::PageGroup::resetUserStyleCacheInAllFrames):
 161         * rendering/RenderBR.cpp:
 162         (WebCore::RenderBR::lineHeight):
 163         * rendering/RenderBlock.cpp:
 164         (WebCore::RenderBlock::styleDidChange):
 165         (WebCore::RenderBlock::splitBlocks):
 166         (WebCore::RenderBlock::addChildIgnoringAnonymousColumnBlocks):
 167         (WebCore::RenderBlock::lineHeight):
 168         (WebCore::RenderBlock::updateFirstLetter):
 169         * rendering/RenderInline.cpp:
 170         (WebCore::RenderInline::styleDidChange):
 171         (WebCore::RenderInline::updateAlwaysCreateLineBoxes):
 172         (WebCore::RenderInline::addChildIgnoringContinuation):
 173         (WebCore::RenderInline::splitInlines):
 174         (WebCore::RenderInline::lineHeight):
 175         * rendering/RenderObject.cpp:
 176         (WebCore::RenderObject::uncachedFirstLineStyle):
 177         (WebCore::RenderObject::firstLineStyleSlowCase):
 178         * rendering/RenderObject.h:
 179         (WebCore::RenderObject::firstLineStyle):
 180         * rendering/RenderObjectChildList.cpp:
 181         (WebCore::RenderObjectChildList::updateBeforeAfterContent):
 182         * rendering/RenderTableRow.cpp:
 183         (WebCore::RenderTableRow::updateBeforeAndAfterContent):
 184         * rendering/RootInlineBox.cpp:
 185         (WebCore::RootInlineBox::verticalPositionForBox):
 186
 187 2012-09-24  Otto Derek Cheung  <otcheung@rim.com>
 188
 189         [BlackBerry] Reverting implementation for 407 error pages
 190         https://bugs.webkit.org/show_bug.cgi?id=97455
 191
 192         Reviewed by Rob Buis.
 193
 194         Revert "[BlackBerry] Reverting implementation for 407 error pages"
 195         This reverts commit fda0a1b6ac40c06c03bb6293b4a7d7353c3ca238.
 196
 197         This revert also reverts commit 0cffe01961fb80204138505bcec29a83818efb73
 198         due to dependency issues.
 199
 200         * PlatformBlackBerry.cmake:
 201         * platform/blackberry/AuthenticationChallengeManager.cpp: Removed.
 202         * platform/blackberry/AuthenticationChallengeManager.h:
 203         * platform/blackberry/PageClientBlackBerry.h:
 204         * platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:
 205         (WebCore::MediaPlayerPrivate::MediaPlayerPrivate):
 206         (WebCore::MediaPlayerPrivate::~MediaPlayerPrivate):
 207         (WebCore::MediaPlayerPrivate::onAuthenticationNeeded):
 208         (WebCore::MediaPlayerPrivate::notifyChallengeResult):
 209         * platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.h:
 210         (MediaPlayerPrivate):
 211         * platform/network/blackberry/NetworkJob.cpp:
 212         (WebCore::NetworkJob::NetworkJob):
 213         (WebCore::NetworkJob::handleNotifyStatusReceived):
 214         (WebCore::NetworkJob::notifyAuthReceived):
 215         (WebCore::NetworkJob::handleNotifyClose):
 216         (WebCore::NetworkJob::sendRequestWithCredentials):
 217         (WebCore::NetworkJob::notifyChallengeResult):
 218         * platform/network/blackberry/NetworkJob.h:
 219         (NetworkJob):
 220
 221 2012-09-24  Chris Rogers  <crogers@google.com>
 222
 223         [REGRESSION] Layout Test webaudio/biquad-getFrequencyResponse.html is failing
 224         https://bugs.webkit.org/show_bug.cgi?id=97439
 225
 226         Reviewed by Kenneth Russell.
 227
 228         Fixes uninitialized member variable.
 229
 230         This should fix flaky failing test: webaudio/biquad-getFrequencyResponse.html
 231
 232         * Modules/webaudio/BiquadProcessor.cpp:
 233         (WebCore::BiquadProcessor::BiquadProcessor):
 234
 235 2012-09-24  Tony Chang  <tony@chromium.org>
 236
 237         flex-grow should be 1 when omitted from flex shorthand
 238         https://bugs.webkit.org/show_bug.cgi?id=97480
 239
 240         Reviewed by Ojan Vafai.
 241
 242         We were using 0, based on an outdated version of the spec.
 243
 244         Tests: css3/flexbox/flex-property-parsing.html
 245                css3/flexbox/flex-algorithm.html: New test case.
 246
 247         * css/CSSParser.cpp:
 248         (WebCore::CSSParser::parseFlex):
 249
 250 2012-09-24  Benjamin Poulain  <benjamin@webkit.org>
 251
 252         Add support for query encoding to WTFURL
 253         https://bugs.webkit.org/show_bug.cgi?id=97422
 254
 255         Reviewed by Adam Barth.
 256
 257         Add the Charset conversion on WebCore side.
 258
 259         * platform/KURLWTFURL.cpp:
 260         (WebCore::KURL::KURL):
 261         (CharsetConverter):
 262         (WebCore::CharsetConverter::CharsetConverter):
 263         * platform/mac/KURLMac.mm:
 264         (WebCore::KURL::KURL):
 265
 266 2012-09-24  Benjamin Poulain  <benjamin@webkit.org>
 267
 268         Integrate most of GoogleURL in WTFURL
 269         https://bugs.webkit.org/show_bug.cgi?id=97405
 270
 271         Reviewed by Adam Barth.
 272
 273         Loosen KURLWTFURL to be able to run most tests in Debug.
 274
 275         * platform/KURLWTFURL.cpp:
 276         (WebCore::KURL::KURL):
 277         (WebCore::KURL::hasPort):
 278         (WebCore::KURL::user):
 279         (WebCore::KURL::pass):
 280         (WebCore::KURL::hasPath):
 281         (WebCore::KURL::path):
 282         (WebCore::KURL::query):
 283         (WebCore::KURL::fragmentIdentifier):
 284         (WebCore::KURL::fileSystemPath):
 285
 286 2012-09-24  Hans Muller  <hmuller@adobe.com>
 287
 288         [CSS Exclusions] ExclusionShape API should use logical coordinates for input/output
 289         https://bugs.webkit.org/show_bug.cgi?id=96156
 290
 291         Reviewed by Dirk Schulze.
 292
 293         Correct how ExclusionShapes deal with logical coordinates and enable
 294         shape-inside exclusion layout for vertical writing-modes.
 295
 296         BasicShape's are defined in physical cooordinates, incoming line and box
 297         dimensions are logical coordinates, and the ExclusionShape internals assume
 298         that lines are aligned with the Y axis. The createExclusionShape() method
 299         is responsible for converting the BasicShape to the internal coordinate
 300         system when the writing-mode is vertical. Similarly, the getInclude,ExcludedIntervals()
 301         methods are responsible for converting their logical line parameters to Y
 302         values in the internal coordinate system. The min,maxYForLogicalLine()
 303         methods do the conversion, based on the WritingMode the ExclusionShape
 304         was created with. The getInclude,ExcludedIntervals() methods return the
 305         logical left and right edges of line segments. No transformation is needed for this.
 306
 307         The ExclusionShape's internal coordinate system is essentially the
 308         "logical" one, except that top/bottom always map to Y, no matter what
 309         the writing-mode is.  This is just to simplify writing geometrical shape
 310         algorithms, notably the complex ones for polygons. The bug report includes a
 311         pair of diagrams that clarify how internal coordinates are related to logical
 312         and physical coordinates.
 313
 314         Test: fast/exclusions/shape-inside/shape-inside-vertical-text.html
 315
 316         * rendering/ExclusionRectangle.cpp:
 317         (WebCore::ExclusionRectangle::getExcludedIntervals): rename more accurately reflects return value
 318         (WebCore::ExclusionRectangle::getIncludedIntervals): rename for consistency with "excluded" version
 319         * rendering/ExclusionRectangle.h:
 320         * rendering/ExclusionShape.cpp:
 321         (WebCore):
 322         (WebCore::ExclusionShape::createExclusionShape):
 323         * rendering/ExclusionShape.h:
 324         (LineSegment):
 325         (WebCore::LineSegment::LineSegment): relocated from WrapShapeInfo.h
 326         (WebCore):
 327         (ExclusionShape):
 328         (WebCore::ExclusionShape::minYForLogicalLine):
 329         (WebCore::ExclusionShape::maxYForLogicalLine):
 330         (WebCore::ExclusionShape::internalToLogicalBoundingBox):
 331         * rendering/WrapShapeInfo.cpp:
 332         (WebCore::WrapShapeInfo::isWrapShapeInfoEnabledForRenderBlock):
 333         (WebCore::WrapShapeInfo::computeShapeSize): pass writingMode to createExclusionShape()
 334         (WebCore::WrapShapeInfo::computeSegmentsForLine): removed short-circuit for vertical writing-modes
 335         * rendering/WrapShapeInfo.h:
 336         (WebCore):
 337
 338 2012-09-24  Tony Chang  <tony@chromium.org>
 339
 340         Replace RenderMeter::updateLogicalHeight to RenderMeter::computeLogicalHeight
 341         https://bugs.webkit.org/show_bug.cgi?id=97475
 342
 343         Reviewed by Ojan Vafai.
 344
 345         Using RenderMeter::computeLogicalHeight is part of bug 96804.
 346         Also fix some code to be vertical writing mode aware. This doesn't actually cause
 347         a behavioral difference because we use percentage heights/widths which don't depend on
 348         updateLogicalWidth or computeLogicalHeight.  You can still see bugs if you try to set
 349         the min-width on a <meter> node in a vertical writing mode.
 350
 351         No new tests, no behavioral changes.
 352
 353         * rendering/RenderMeter.cpp:
 354         (WebCore::RenderMeter::updateLogicalWidth): Make vertical writing mode aware.
 355         (WebCore::RenderMeter::computeLogicalHeight): Switch from updateLogicalHeight and make vertical writing mode aware.
 356         * rendering/RenderMeter.h:
 357         (RenderMeter):
 358
 359 2012-09-24  Dimitri Glazkov  <dglazkov@chromium.org>
 360
 361         Remove unbaked support for :scope pseudo-class.
 362         https://bugs.webkit.org/show_bug.cgi?id=97467
 363
 364         Reviewed by Antti Koivisto.
 365
 366         It turns out, the Selectors 4 require ":scope" to match contextual reference element set, which would be the root node in querySelector:
 367         http://dev.w3.org/csswg/selectors4/#the-scope-pseudo
 368
 369         Right now, we simply make ":scope" equivalent to ":root", which is not correct. Let's remove the partial implementation until we have
 370         time/energy to fully implement it.
 371
 372         No new tests, removing half-baked feature.
 373
 374         * css/CSSSelector.cpp:
 375         (WebCore::CSSSelector::pseudoId): Removed all mentions of PseudoScope
 376         (WebCore::nameToPseudoTypeMap): Ditto.
 377         (WebCore::CSSSelector::extractPseudoType): Ditto.
 378         * css/CSSSelector.h: Ditto.
 379         * css/SelectorChecker.cpp:
 380         (WebCore::SelectorChecker::checkOneSelector): Ditto.
 381
 382 2012-09-24  Ryosuke Niwa  <rniwa@webkit.org>
 383
 384         suspend/resumeWidgetHierarchyUpdates should be a RAII object
 385         https://bugs.webkit.org/show_bug.cgi?id=96706
 386
 387         Reviewed by Simon Fraser.
 388
 389         Replaced suspendWidgetHierarchyUpdates and resumeWidgetHierarchyUpdates by WidgetHierarchyUpdatesSuspensionScope.
 390
 391         * WebCore.exp.in: Export new symbols.
 392         * dom/ContainerNode.cpp:
 393         (WebCore::ContainerNode::removeChild):
 394         (WebCore::ContainerNode::removeChildren):
 395         * dom/Document.cpp:
 396         (WebCore::Document::recalcStyle):
 397         * dom/Element.cpp:
 398         (WebCore::Element::attach):
 399         (WebCore::Element::detach):
 400         * page/mac/EventHandlerMac.mm:
 401         (WebCore::EventHandler::passMouseDownEventToWidget):
 402         * rendering/RenderWidget.cpp:
 403         (WebCore):
 404         (WebCore::WidgetHierarchyUpdatesSuspensionScope::widgetNewParentMap):
 405         (WebCore::WidgetHierarchyUpdatesSuspensionScope::moveWidgets):
 406         (WebCore::moveWidgetToParentSoon):
 407         * rendering/RenderWidget.h:
 408         (WidgetHierarchyUpdatesSuspensionScope):
 409         (WebCore::WidgetHierarchyUpdatesSuspensionScope::WidgetHierarchyUpdatesSuspensionScope):
 410         (WebCore::WidgetHierarchyUpdatesSuspensionScope::~WidgetHierarchyUpdatesSuspensionScope):
 411         (WebCore::WidgetHierarchyUpdatesSuspensionScope::isSuspended):
 412         (WebCore::WidgetHierarchyUpdatesSuspensionScope::scheduleWidgetToMove):
 413         (WebCore):
 414         (RenderWidget):
 415
 416 2012-09-24  Otto Derek Cheung  <otcheung@rim.com>
 417
 418         [BlackBerry] Reverting implementation for 407 error pages
 419         https://bugs.webkit.org/show_bug.cgi?id=97455
 420
 421         Reviewed by Rob Buis.
 422
 423         Revert "[BlackBerry] Show custom error page when 407 is received"
 424         This reverts commit e6d14529428fe47916fcb997528095b8acad2f2b.
 425
 426         * platform/network/blackberry/NetworkJob.cpp:
 427         (WebCore::NetworkJob::handleNotifyStatusReceived):
 428         (WebCore::NetworkJob::notifyAuthReceived):
 429         (WebCore::NetworkJob::sendRequestWithCredentials):
 430
 431 2012-09-24  Sheriff Bot  <webkit.review.bot@gmail.com>
 432
 433         Unreviewed, rolling out r129388.
 434         http://trac.webkit.org/changeset/129388
 435         https://bugs.webkit.org/show_bug.cgi?id=97477
 436
 437         Caused an assertion in a WebKit2 unit test (Requested by
 438         abarth on #webkit).
 439
 440         * WebCore.exp.in:
 441         * loader/FrameLoader.cpp:
 442         (WebCore::FrameLoader::didBeginDocument):
 443         * loader/SubresourceLoader.cpp:
 444         (WebCore::SubresourceLoader::willSendRequest):
 445         * loader/cache/CachedImage.cpp:
 446         * loader/cache/CachedImage.h:
 447         (WebCore::CachedImage::stillNeedsLoad):
 448         (CachedImage):
 449         * loader/cache/CachedResource.cpp:
 450         (WebCore::CachedResource::didAddClient):
 451         * loader/cache/CachedResource.h:
 452         (WebCore):
 453         * loader/cache/CachedResourceLoader.cpp:
 454         (WebCore::CachedResourceLoader::CachedResourceLoader):
 455         (WebCore::CachedResourceLoader::requestImage):
 456         (WebCore::CachedResourceLoader::canRequest):
 457         (WebCore::CachedResourceLoader::requestResource):
 458         (WebCore::CachedResourceLoader::revalidateResource):
 459         (WebCore::CachedResourceLoader::loadResource):
 460         (WebCore::CachedResourceLoader::determineRevalidationPolicy):
 461         (WebCore::CachedResourceLoader::setAutoLoadImages):
 462         * loader/cache/CachedResourceLoader.h:
 463         (CachedResourceLoader):
 464         * page/Settings.cpp:
 465         (WebCore::setLoadsImagesAutomaticallyInAllFrames):
 466         (WebCore::Settings::Settings):
 467         (WebCore::Settings::setLoadsImagesAutomatically):
 468         (WebCore::Settings::loadsImagesAutomaticallyTimerFired):
 469         (WebCore::Settings::setImagesEnabled):
 470         * page/Settings.h:
 471         (Settings):
 472         * testing/InternalSettings.cpp:
 473         (WebCore::InternalSettings::Backup::Backup):
 474         (WebCore::InternalSettings::Backup::restoreTo):
 475         * testing/InternalSettings.h:
 476         (Backup):
 477         (InternalSettings):
 478         * testing/InternalSettings.idl:
 479
 480 2012-09-24  Erik Arvidsson  <arv@chromium.org>
 481
 482         DOM4 remove method
 483         https://bugs.webkit.org/show_bug.cgi?id=73885
 484
 485         Reviewed by Ojan Vafai.
 486
 487         This adds the DOM 4 remove method.
 488
 489         http://dvcs.w3.org/hg/domcore/raw-file/tip/Overview.html#dom-childnode-remove
 490
 491         The remove method removes the node from its parent if it has a parent.
 492
 493         Tests: fast/dom/Comment/remove.html
 494                fast/dom/DocumentType/remove.html
 495                fast/dom/Element/remove.html
 496                fast/dom/Text/remove.html
 497
 498         * dom/CharacterData.idl:
 499         * dom/DocumentType.idl:
 500         * dom/Element.idl:
 501         * dom/Node.cpp:
 502         (WebCore::Node::remove): The spec says to do nothing if the node has no parent.
 503         * html/parser/HTMLTreeBuilder.cpp:
 504         (WebCore::HTMLTreeBuilder::processStartTagForInBody): Cleanup.
 505         * html/track/TextTrackCue.cpp:
 506         (WebCore::TextTrackCue::removeDisplayTree): No need to check parenNode twice.
 507
 508 2012-09-24  Tommy Widenflycht  <tommyw@google.com>
 509
 510         MediaStream API: Fix the incorrectly spelled RTCPeerConnection::onnegotiationneeded callback
 511         https://bugs.webkit.org/show_bug.cgi?id=97456
 512
 513         Reviewed by Adam Barth.
 514
 515         Renamed onnegotationneeded to onnegotiationneeded.
 516
 517         Existing test updated.
 518
 519         * Modules/mediastream/RTCPeerConnection.cpp:
 520         (WebCore::RTCPeerConnection::negotiationNeeded):
 521         * Modules/mediastream/RTCPeerConnection.h:
 522         (RTCPeerConnection):
 523         * Modules/mediastream/RTCPeerConnection.idl:
 524         * dom/EventNames.h:
 525         (WebCore):
 526
 527 2012-09-24  Peter Beverloo  <peter@chromium.org>
 528
 529         Android's mock scrollbars shows up as a difference in layout test results
 530         https://bugs.webkit.org/show_bug.cgi?id=96382
 531
 532         Reviewed by Adam Barth.
 533
 534         Remove the exceptions made for layout tests in Android's scrollbar theme.
 535         This will make our actual scrollbars show up in layout test pixel results,
 536         bringing the tests closer to what we actually ship.
 537
 538         An important difference with other platforms is that scrollbars do not
 539         take any width on Android, they're rendered on top of the content. Therefore
 540         each test that has a visible scrollbar does not just need a new pixel
 541         result, but will also need a new text result. This will be handled as part
 542         of a larger rebaselining process.
 543
 544         Will be exercised by every layout test that has a scrollbar.
 545
 546         * platform/chromium/ScrollbarThemeChromiumAndroid.cpp:
 547         (WebCore::ScrollbarThemeChromiumAndroid::scrollbarThickness):
 548         (WebCore::ScrollbarThemeChromiumAndroid::usesOverlayScrollbars):
 549         (WebCore::ScrollbarThemeChromiumAndroid::hasThumb):
 550         * platform/chromium/ScrollbarThemeChromiumAndroid.h:
 551         (ScrollbarThemeChromiumAndroid):
 552
 553 2012-09-24  Sean Wang  <Xuewen.Wang@torchmobile.com.cn>
 554
 555         [BlackBerry] Basic authentication challenge credentials for stored credentials again after restarting browser
 556         https://bugs.webkit.org/show_bug.cgi?id=96362
 557
 558         Reviewed by Rob Buis.
 559
 560         Fix a mistake of the commit 11fdc73c7c74bbd736ed4160248ff59636a01864
 561         Trunk has been changed during reviewing that patch.
 562
 563         No new tests, this is to correct a build error.
 564
 565         * platform/network/blackberry/CredentialBackingStore.cpp:
 566         (WebCore::CredentialBackingStore::getProtectionSpace):
 567
 568 2012-09-24  Dan Bernstein  <mitz@apple.com>
 569
 570         Reverted r129176, the fix for <http://webkit.org/b/97269>, because it introduced a
 571         discrepancy between line breaking and max width computation.
 572
 573         Rubber-stamped by Anders Carlsson.
 574
 575         * platform/graphics/mac/ComplexTextController.cpp:
 576         (WebCore::ComplexTextController::adjustGlyphsAndAdvances):
 577
 578 2012-09-24  Bo Liu  <boliu@chromium.org>
 579
 580         Reland "Add in-place reload behavior to ImagesEnabled setting" with optimizations
 581         https://bugs.webkit.org/show_bug.cgi?id=97055
 582
 583         Reviewed by Adam Barth.
 584
 585         Relanding 128780, 128676, 128645. Was reverted in 128914 due to
 586         performance regression in Chromium.
 587
 588         New changes in addition to previously reverted patches:
 589
 590         Refactored CachedResource::requestResource, loadResource, and
 591         revalidateResource. Moved CachedResource::load method to end of
 592         requestResource so there is one place where load is called for all
 593         resources.
 594
 595         Added a enum parameter for requestResource and
 596         determineRevalidationPolicy so that FrameLoaderClient::allowImage call
 597         do not need to be called multiple times.
 598
 599         Removed CachedImage::load call in requestImage so it is not called
 600         twice.
 601
 602         Removed unnecessary Frame.h includes in CachedResource and
 603         CachedImage.
 604
 605         Removed dead load() method declaration in CachedImage.
 606
 607         Updated text expectation for two image-permissions tests to reflect
 608         the removed calls to allowImage.
 609
 610         Tests: fast/loader/display-image-unset-allows-cached-image-load.html
 611                fast/loader/display-image-unset-can-block-image-and-can-reload-in-place.html
 612                fast/loader/images-enabled-unset-can-block-image-and-can-reload-in-place.html
 613
 614         * WebCore.exp.in:
 615         * loader/FrameLoader.cpp:
 616         (WebCore::FrameLoader::didBeginDocument):
 617         * loader/SubresourceLoader.cpp:
 618         (WebCore::SubresourceLoader::willSendRequest):
 619         * loader/cache/CachedImage.cpp:
 620         * loader/cache/CachedImage.h:
 621         * loader/cache/CachedResource.cpp:
 622         (WebCore::CachedResource::didAddClient):
 623         * loader/cache/CachedResource.h:
 624         (WebCore):
 625         (WebCore::CachedResource::stillNeedsLoad):
 626         * loader/cache/CachedResourceLoader.cpp:
 627         (WebCore::CachedResourceLoader::CachedResourceLoader):
 628         (WebCore::CachedResourceLoader::requestImage):
 629         (WebCore::CachedResourceLoader::canRequest):
 630         (WebCore::CachedResourceLoader::requestResource):
 631         (WebCore::CachedResourceLoader::revalidateResource):
 632         (WebCore::CachedResourceLoader::loadResource):
 633         (WebCore::CachedResourceLoader::determineRevalidationPolicy):
 634         (WebCore::CachedResourceLoader::setAutoLoadImages):
 635         (WebCore):
 636         (WebCore::CachedResourceLoader::setImagesEnabled):
 637         (WebCore::CachedResourceLoader::clientDefersImage):
 638         (WebCore::CachedResourceLoader::shouldDeferImageLoad):
 639         (WebCore::CachedResourceLoader::reloadImagesIfNotDeferred):
 640         * loader/cache/CachedResourceLoader.h:
 641         (CachedResourceLoader):
 642         * page/Settings.cpp:
 643         (WebCore::setImageLoadingSettings):
 644         (WebCore::Settings::Settings):
 645         (WebCore::Settings::setLoadsImagesAutomatically):
 646         (WebCore::Settings::imageLoadingSettingsTimerFired):
 647         (WebCore::Settings::setImagesEnabled):
 648         * page/Settings.h:
 649         (Settings):
 650         * testing/InternalSettings.cpp:
 651         (WebCore::InternalSettings::Backup::Backup):
 652         (WebCore::InternalSettings::Backup::restoreTo):
 653         (WebCore::InternalSettings::setImagesEnabled):
 654         (WebCore):
 655         * testing/InternalSettings.h:
 656         (Backup):
 657         (InternalSettings):
 658         * testing/InternalSettings.idl:
 659
 660 2012-09-24  Joone Hur  <joone.hur@intel.com>, Gustavo Noronha Silva  <gustavo.noronha@collabora.com>
 661
 662         [GTK] Implement GraphicsLayer using Clutter
 663         https://bugs.webkit.org/show_bug.cgi?id=73767
 664
 665         Reviewed by Martin Robinson.
 666
 667         This patch is needed for enabling Accelerated Compositing(Clutter backend)
 668         with the patches submitted in bug 92045 and 91940.
 669
 670         No new tests. This will be covered by pixel tests for Qt/GTK+ accelerated
 671         compositing and 3D transforms.
 672
 673         * GNUmakefile.list.am:
 674         * platform/clutter/GRefPtrClutter.cpp: Removed.
 675         * platform/clutter/GRefPtrClutter.h: Removed.
 676         * platform/graphics/clutter/PlatformClutterLayerClient.h: Added.
 677         (WebCore):
 678         (PlatformClutterLayerClient):
 679         (WebCore::PlatformClutterLayerClient::~PlatformClutterLayerClient):
 680         * platform/graphics/clutter/TransformationMatrixClutter.cpp: Added to convert CoglMatrix to
 681         TransformationMatrix.
 682         (WebCore):
 683         (WebCore::TransformationMatrix::operator CoglMatrix):
 684         * platform/graphics/transforms/TransformationMatrix.h:
 685         (TransformationMatrix):
 686
 687 2012-09-24  David Grogan  <dgrogan@chromium.org>
 688
 689         Unprefix IndexedDB
 690         https://bugs.webkit.org/show_bug.cgi?id=96548
 691
 692         Reviewed by Adam Barth.
 693
 694         We are largely compatible with the FF implementation and the w3c test
 695         suite submitted by MS. The w3c test suite doesn't yet check
 696         for lack of prefix; this is mostly to signal to devs that we think our
 697         implementation is stable.
 698
 699         This patch uses the new FeatureObserver to get data about prefixed
 700         vs unprefixed usage.
 701
 702         Tests: storage/indexeddb/unprefix-workers.html
 703                storage/indexeddb/unprefix.html
 704
 705         * Modules/indexeddb/DOMWindowIndexedDatabase.cpp:
 706         (WebCore::DOMWindowIndexedDatabase::indexedDB):
 707         * Modules/indexeddb/DOMWindowIndexedDatabase.h:
 708         (DOMWindowIndexedDatabase):
 709         * Modules/indexeddb/DOMWindowIndexedDatabase.idl:
 710         * Modules/indexeddb/WorkerContextIndexedDatabase.cpp:
 711         (WebCore::WorkerContextIndexedDatabase::indexedDB):
 712         * Modules/indexeddb/WorkerContextIndexedDatabase.h:
 713         (WorkerContextIndexedDatabase):
 714         * Modules/indexeddb/WorkerContextIndexedDatabase.idl:
 715         * bindings/generic/RuntimeEnabledFeatures.h:
 716         Making the auxiliary objects RuntimeEnabled didn't make much sense as
 717         they are useless without the factory, so always enable them.
 718
 719         (WebCore::RuntimeEnabledFeatures::indexedDBEnabled):
 720         * bindings/scripts/CodeGeneratorGObject.pm:
 721         (SkipAttribute):
 722         * page/FeatureObserver.h:
 723
 724 2012-09-24  Andrey Adaikin  <aandrey@chromium.org>
 725
 726         Web Inspector: [WebGL] First step towards 2D canvas instrumentation in injected script
 727         https://bugs.webkit.org/show_bug.cgi?id=96746
 728
 729         Reviewed by Pavel Feldman.
 730
 731         First step towards the 2D canvas instrumentation from the WebGL injected script module side.
 732         We trace and save all calls that affect the 2D canvas context state and it's resources (Gradient and Pattern)
 733         so that we could replay the context and resources states later.
 734         The 2D canvas context state consists of:
 735         - current transformation matrix
 736         - current default path
 737         - current clipping region (affected by the "clip" command)
 738         - current values of the context attributes (like strokeStyle, fillStyle, etc.)
 739         - a stack of saved drawing states (affected by the "save" and "restore" commands)
 740
 741         * inspector/InjectedScriptCanvasModuleSource.js:
 742         (.):
 743
 744 2012-09-24  Andrey Adaikin  <aandrey@chromium.org>
 745
 746         Web Inspector: [TextEditor] conditional breakpoint popup not showing up the first time
 747         https://bugs.webkit.org/show_bug.cgi?id=97442
 748
 749         Reviewed by Pavel Feldman.
 750
 751         1) The conditional breakpoint popup may not show up the first time we click on the gutter.
 752         This was the case because the popup decoration element would be deleted from the line while
 753         it was being highlighted and appended afterwards. Now we do not remove decorations from the
 754         DOM while highlighting.
 755
 756         2) Also the popup will close itself on any mouse click event, even if it's targeted to the
 757         input box itself. This was due to pointer-events: none; CSS style for the parent element.
 758         Now we just override this style for the input box.
 759
 760         * inspector/front-end/DefaultTextEditor.js:
 761         (WebInspector.TextEditorMainPanel.prototype._paintLine):
 762         (WebInspector.TextEditorMainPanel.prototype._insertSpanBefore):
 763         (WebInspector.TextEditorMainPanel.prototype._insertTextNodeBefore):
 764         * inspector/front-end/inspector.css:
 765         (.source-frame-breakpoint-condition):
 766
 767 2012-09-24  Andrey Adaikin  <aandrey@chromium.org>
 768
 769         Web Inspector: [Canvas] set CanvasAgent in InstrumentingAgents upon calling enable command
 770         https://bugs.webkit.org/show_bug.cgi?id=97331
 771
 772         Reviewed by Yury Semikhatsky.
 773
 774         We should set CanvasAgent in InstrumentingAgents from enable() command and remove from disable(), instead of doing this in constructor and destructor.
 775
 776         * inspector/InspectorCanvasAgent.cpp:
 777         (WebCore::InspectorCanvasAgent::InspectorCanvasAgent):
 778         (WebCore::InspectorCanvasAgent::~InspectorCanvasAgent):
 779         (WebCore::InspectorCanvasAgent::restore):
 780         (WebCore::InspectorCanvasAgent::enable):
 781         (WebCore::InspectorCanvasAgent::disable):
 782         * inspector/InspectorCanvasAgent.h:
 783         (InspectorCanvasAgent):
 784         * inspector/InspectorCanvasInstrumentation.h:
 785         (WebCore::InspectorInstrumentation::wrapCanvas2DRenderingContextForInstrumentation):
 786         (WebCore::InspectorInstrumentation::wrapWebGLRenderingContextForInstrumentation):
 787
 788 2012-09-24  Patrick Gansterer  <paroga@webkit.org>
 789
 790         Remove remaining WTF_DEPRECATED_STRING_OPERATORS from cpp files
 791         https://bugs.webkit.org/show_bug.cgi?id=97412
 792
 793         Reviewed by Adam Barth.
 794
 795         * css/StylePropertySet.cpp: Removed useless define.
 796
 797 2012-09-20  Emil A Eklund  <eae@chromium.org>
 798
 799         snapToSize rounds the incorrectly for negative locations
 800         https://bugs.webkit.org/show_bug.cgi?id=97265
 801
 802         Reviewed by Eric Seidel.
 803
 804         Change snapSizeToPixel to preserve sign for location which
 805         affects rounding.
 806
 807         Test: fast/sub-pixel/snap-negative-location.html
 808
 809         * platform/FractionalLayoutUnit.h:
 810         (WebCore::snapSizeToPixel):
 811
 812 2012-09-24  Jonathan Liu  <net147@gmail.com>
 813
 814         Use unsigned char for bitfield instead of unsigned.
 815         https://bugs.webkit.org/show_bug.cgi?id=97447
 816
 817         Reviewed by Andreas Kling.
 818
 819         Not all compilers will pad an unsigned bitfield to the smallest size.
 820         Use unsigned char instead of unsigned to reduce padding for compilers
 821         that pad to the underlying type.
 822
 823         No new tests. There is already a compile-time assertion.
 824
 825         * css/CSSRule.cpp:
 826         (SameSizeAsCSSRule):
 827         * css/CSSRule.h:
 828         (CSSRule):
 829
 830 2012-09-24  Peter Rybin  <peter.rybin@gmail.com>
 831
 832         Web Inspector: don't allow exception in front-end when expanding function scope
 833         https://bugs.webkit.org/show_bug.cgi?id=97346
 834
 835         Reviewed by Yury Semikhatsky.
 836
 837         A missing guard condition is added.
 838
 839         * inspector/front-end/ObjectPropertiesSection.js:
 840         (WebInspector.FunctionScopeMainTreeElement.prototype.onpopulate.didGetDetails):
 841         (WebInspector.FunctionScopeMainTreeElement.prototype.onpopulate):
 842
 843 2012-09-24  Arvid Nilsson  <anilsson@rim.com>
 844
 845         [BlackBerry] Add cookie database API
 846         https://bugs.webkit.org/show_bug.cgi?id=97102
 847
 848         Reviewed by Antonio Gomes.
 849
 850         Add a CookieManager method that takes a list of strings to parse
 851         instead of just one string. Expose CookieParser::parseOneCookie.
 852
 853         Reviewed internally by Otto D. Cheung.
 854
 855         No change in behavior, no new tests.
 856
 857         * platform/blackberry/CookieManager.cpp:
 858         (WebCore::CookieManager::setCookies):
 859         (WebCore):
 860         * platform/blackberry/CookieManager.h:
 861         * platform/blackberry/CookieParser.cpp:
 862         (WebCore::CookieParser::parseOneCookie):
 863         (WebCore):
 864         * platform/blackberry/CookieParser.h:
 865         (CookieParser):
 866
 867 2012-09-24  Vsevolod Vlasov  <vsevik@chromium.org>
 868
 869         Web Inspector: Open resource dialog has too many false positive matches.
 870         https://bugs.webkit.org/show_bug.cgi?id=97332
 871
 872         Reviewed by Alexander Pavlov.
 873
 874         Improved open resource dialog filtering to make the amount of false
 875         positive search results less.
 876
 877         * inspector/front-end/FilteredItemSelectionDialog.js:
 878         (WebInspector.FilteredItemSelectionDialog.prototype._innerCreateSearchRegExp):
 879         (WebInspector.FilteredItemSelectionDialog.prototype._highlightItem.get var):
 880         (WebInspector.FilteredItemSelectionDialog.prototype._highlightItem):
 881
 882 2012-09-24  Andrey Adaikin  <aandrey@chromium.org>
 883
 884         Web Inspector: [Canvas] support 2D canvas instrumentation from the inspector C++ code
 885         https://bugs.webkit.org/show_bug.cgi?id=97203
 886
 887         Reviewed by Yury Semikhatsky.
 888
 889         Implements wrapping a 2D canvas context through the injected canvas module script facility.
 890
 891         * bindings/js/JSHTMLCanvasElementCustom.cpp:
 892         (WebCore::JSHTMLCanvasElement::getContext):
 893         * bindings/v8/custom/V8HTMLCanvasElementCustom.cpp:
 894         (WebCore::V8HTMLCanvasElement::getContextCallback):
 895         * inspector/InjectedScriptCanvasModule.cpp:
 896         (WebCore::InjectedScriptCanvasModule::wrapCanvas2DContext):
 897         (WebCore):
 898         (WebCore::InjectedScriptCanvasModule::wrapWebGLContext):
 899         (WebCore::InjectedScriptCanvasModule::callWrapContextFunction):
 900         * inspector/InjectedScriptCanvasModule.h:
 901         (InjectedScriptCanvasModule):
 902         * inspector/InspectorCanvasAgent.cpp:
 903         (WebCore::InspectorCanvasAgent::wrapCanvas2DRenderingContextForInstrumentation):
 904         (WebCore):
 905         * inspector/InspectorCanvasAgent.h:
 906         (InspectorCanvasAgent):
 907         * inspector/InspectorCanvasInstrumentation.h:
 908         (WebCore::InspectorInstrumentation::wrapCanvas2DRenderingContextForInstrumentation):
 909         (WebCore):
 910         * inspector/InspectorInstrumentation.h:
 911         (InspectorInstrumentation):
 912
 913 2012-09-24  Pavel Feldman  <pfeldman@chromium.org>
 914
 915         Web Inspector:EXC_BAD_ACCESS upon closing page with node highlighted
 916         https://bugs.webkit.org/show_bug.cgi?id=97446
 917
 918         Reviewed by Alexander Pavlov.
 919
 920         Added 0 check on frame view.
 921
 922         * inspector/InspectorOverlay.cpp:
 923         (WebCore::InspectorOverlay::update):
 924
 925 2012-09-21  Alexander Pavlov  <apavlov@chromium.org>
 926
 927         Web Inspector: [REGRESSION] Cmd-Shift-C doesn't enable element inspection mode when inspector hidden
 928         https://bugs.webkit.org/show_bug.cgi?id=97249
 929
 930         Reviewed by Yury Semikhatsky.
 931
 932         Fix regression from r125871.
 933
 934         * inspector/front-end/InspectorFrontendAPI.js:
 935         (InspectorFrontendAPI.enterInspectElementMode):
 936         * inspector/front-end/inspector.js:
 937         (WebInspector._createGlobalStatusBarItems):
 938         (WebInspector.documentKeyDown):
 939
 940 2012-09-24  Alexander Pavlov  <apavlov@chromium.org>
 941
 942         Web Inspector: Disable persistence of master switches in the "Overrides" settings tab
 943         https://bugs.webkit.org/show_bug.cgi?id=97440
 944
 945         Reviewed by Yury Semikhatsky.
 946
 947         Disables enforcement of device metrics, geolocation, and device orientation overrides on Web Inspector start.
 948
 949         * inspector/InspectorPageAgent.cpp:
 950         (WebCore::InspectorPageAgent::restore):
 951         * inspector/front-end/SettingsScreen.js:
 952         (WebInspector.UserAgentSettingsTab.prototype._onMetricsCheckboxClicked):
 953         (WebInspector.UserAgentSettingsTab.prototype._onGeolocationOverrideCheckboxClicked):
 954         (WebInspector.UserAgentSettingsTab.prototype._onDeviceOrientationOverrideCheckboxClicked):
 955         * inspector/front-end/UserAgentSupport.js:
 956         (WebInspector.UserAgentSupport):
 957         (WebInspector.UserAgentSupport.prototype.toggleDeviceMetricsOverride):
 958         (WebInspector.UserAgentSupport.prototype.toggleGeolocationPositionOverride):
 959         (WebInspector.UserAgentSupport.prototype.toggleDeviceOrientationOverride):
 960         * inspector/front-end/externs.js:
 961
 962 2012-09-24  Vivek Galatage  <vivekgalatage@gmail.com>
 963
 964         Web Inspector: implement testing harness for pure protocol tests.
 965         https://bugs.webkit.org/show_bug.cgi?id=90675
 966
 967         Reviewed by Yury Semikhatsky.
 968
 969         Implementing the testing harness as the APIs of Internals object not
 970         only reduced platform specific patching of DRTs but also minimized the
 971         the effort required to open up the new dummy inspector Frontend.
 972
 973         The openDummyInspectorFrontend method will return the handle to newly
 974         created DOMWindow. This DOMWindow object can be utilized inside the test
 975         case to communicate using postMessage WebAPI.
 976
 977         The newly created DOMWindow will host the protocol-test.html which
 978         will seed the necessary JS libraries to communicate with the
 979         InspectorBackend.
 980
 981         Test: inspector-protocol/css-getSupportedCSSProperties.html
 982
 983         * WebCore.exp.in:
 984         * WebCore.gypi:
 985         * inspector/InspectorClient.h:
 986         (InspectorClient):
 987         * testing/Internals.cpp:
 988         (InspectorFrontendClientDummy):
 989         (WebCore::InspectorFrontendClientDummy::~InspectorFrontendClientDummy):
 990         (WebCore):
 991         (WebCore::InspectorFrontendClientDummy::InspectorFrontendClientDummy):
 992         (InspectorFrontendChannelDummy):
 993         (WebCore::InspectorFrontendChannelDummy::~InspectorFrontendChannelDummy):
 994         (WebCore::InspectorFrontendChannelDummy::InspectorFrontendChannelDummy):
 995         (WebCore::InspectorFrontendChannelDummy::sendMessageToFrontend):
 996         (WebCore::Internals::consoleMessageArgumentCounts):
 997         (WebCore::Internals::openDummyInspectorFrontend):
 998         (WebCore::Internals::closeDummyInspectorFrontend):
 999         * testing/Internals.h:
1000         (WebCore):
1001         (Internals):
1002         * testing/Internals.idl:
1003
1004 2012-09-24  Keishi Hattori  <keishi@webkit.org>
1005
1006         REGRESSION(r127727): Calendar picker focus doesn't loop in calendar-picker.html
1007         https://bugs.webkit.org/show_bug.cgi?id=97183
1008
1009         Reviewed by Kent Tamura.
1010
1011         Fixing bug calendar-picker.html. The page popup behaves fine because
1012         this code is only necessary in calendar-picker.html where the focus can
1013         move outside of the picker because we are just using an iframe.
1014
1015         No new tests. Added tests to calendar-picker-key-operations.html.
1016
1017         * Resources/pagepopups/calendarPicker.js:
1018         (YearMonthController.prototype.attachTo):
1019         (CalendarPicker.prototype._handleBodyKeyDown):
1020
1021 2012-09-24  Andrey Kosyakov  <caseq@chromium.org>
1022
1023         Unreviewed follow-up to r129336 -- fixed closure compiler warnings.
1024
1025         * inspector/front-end/HAREntry.js:
1026         * inspector/front-end/TimelineModel.js:
1027
1028 2012-09-24  Andrey Adaikin  <aandrey@chromium.org>
1029
1030         Web Inspector: [TextEditor] Line numbers out of sync after closing breakpoint condition editor
1031         https://bugs.webkit.org/show_bug.cgi?id=96510
1032
1033         Reviewed by Vsevolod Vlasov.
1034
1035         Sync line heights on adding/removing decorations.
1036
1037         * inspector/front-end/DefaultTextEditor.js:
1038         (WebInspector.TextEditorGutterChunk):
1039         (WebInspector.TextEditorGutterChunk.prototype.addDecoration):
1040         (WebInspector.TextEditorGutterChunk.prototype.removeDecoration):
1041         (WebInspector.TextEditorGutterChunk.prototype.set expanded):
1042         (WebInspector.TextEditorGutterChunk.prototype.get height):
1043         (WebInspector.TextEditorGutterChunk.prototype._createRow):
1044         (WebInspector.TextEditorMainChunk.prototype.addDecoration):
1045         (WebInspector.TextEditorMainChunk.prototype.removeDecoration):
1046         (WebInspector.TextEditorMainChunk.prototype.removeAllDecorations):
1047
1048 2012-09-21  Andrey Kosyakov  <caseq@chromium.org>
1049
1050         Web Inspector: support saving HAR with resources content
1051         https://bugs.webkit.org/show_bug.cgi?id=97341
1052
1053         Reviewed by Vsevolod Vlasov.
1054
1055         - add WebInspector.HARWriter that writes HAR with content to a stream;
1056         - replace Save all as HAR with Save all with Content as HAR in network's panel context menu;
1057         - remove entries for copying/saving of individual request items to reduce clutter in context menu;
1058         - a drive-by fix of an exception when invoking context menu not on a network item;
1059
1060         Test: http/tests/inspector/network/har-content.html
1061
1062         * English.lproj/localizedStrings.js:
1063         * inspector/front-end/FileUtils.js:
1064         * inspector/front-end/HAREntry.js:
1065         (WebInspector.HAREntry.prototype._buildPostData):
1066         (WebInspector.HARLog.prototype.build):
1067         (WebInspector.HARLog.prototype._creator):
1068         (WebInspector.HARWriter):
1069         (WebInspector.HARWriter.prototype.write):
1070         (WebInspector.HARWriter.prototype._onContentAvailable):
1071         (WebInspector.HARWriter.prototype._beginWrite):
1072         (WebInspector.HARWriter.prototype._writeNextChunk):
1073         * inspector/front-end/NetworkPanel.js:
1074         (WebInspector.NetworkLogView.prototype.get statusBarItems):
1075         (WebInspector.NetworkLogView.prototype._createStatusBarItems):
1076         (WebInspector.NetworkLogView.prototype._contextMenu):
1077         (WebInspector.NetworkLogView.prototype._exportAll.openCallback):
1078         (WebInspector.NetworkLogView.prototype._exportAll):
1079
1080 2012-09-24  Nico Weber  <thakis@chromium.org>
1081
1082         [chromium] Remove BitLockerSkia. It's apparently not used.
1083         https://bugs.webkit.org/show_bug.cgi?id=97434
1084
1085         Reviewed by Kent Tamura.
1086
1087         * platform/graphics/skia/BitLockerSkia.cpp: Removed.
1088         * platform/graphics/skia/BitLockerSkia.h: Removed.
1089
1090 2012-09-24  Matt Falkenhagen  <falken@chromium.org>
1091
1092         Skeleton implementation of dialog.showModal()
1093         https://bugs.webkit.org/show_bug.cgi?id=97425
1094
1095         Reviewed by Kent Tamura.
1096
1097         This adds a basic implementation of showModal(), so it later can be
1098         used to test the top layer, once it is implemented. The main features
1099         of showModal(), modality and the top layer, are not yet implemented.
1100
1101         Test: fast/dom/HTMLDialogElement/dialog-show-modal.html
1102
1103         * html/HTMLDialogElement.cpp:
1104         (WebCore::HTMLDialogElement::showModal): The same as show(), but throws an error in the cases specified in the spec.
1105         (WebCore):
1106         * html/HTMLDialogElement.h:
1107         (HTMLDialogElement):
1108         * html/HTMLDialogElement.idl:
1109
1110 2012-09-24  Ryuan Choi  <ryuan.choi@samsung.com>
1111
1112         [EFL] Regression after r129328
1113         https://bugs.webkit.org/show_bug.cgi?id=97429
1114
1115         Unreviewed, rolling out r129328.
1116
1117         * platform/efl/RenderThemeEfl.cpp:
1118         (WebCore::RenderTheme::themeForPage):
1119
1120 2012-09-23  Ryuan Choi  <ryuan.choi@samsung.com>
1121
1122         [EFL] Duplicated RenderThemeEfl creation for same page
1123         https://bugs.webkit.org/show_bug.cgi?id=88391
1124
1125         Reviewed by Gyuyoung Kim.
1126
1127         Other ports return one or two RenderTheme regardless of page.
1128         But, WebKit/Efl can create different theme for each page and
1129         themeForPage creates new RenderTheme whenever it is called with page.
1130         So more than one RenderTheme can be created for a page.
1131
1132         This patches checks whether page already has RenderTheme not to create new
1133         RenderTheme for same page.
1134
1135         * platform/efl/RenderThemeEfl.cpp:
1136         (WebCore::RenderTheme::themeForPage):
1137
1138 2012-09-23  Keishi Hattori  <keishi@webkit.org>
1139
1140         Add suggestionPicker to CalendarPicker
1141         https://bugs.webkit.org/show_bug.cgi?id=97201
1142
1143         Reviewed by Kent Tamura.
1144
1145         This adds the suggestionPicker to CalendarPicker. It is not available yet as a datalist UI but it can be used inside ManualTests/forms/calendar-picker.html.
1146
1147         No new tests. Suggestion Picker can be tested in ManualTests/forms/calendar-picker.html.
1148
1149         * Resources/pagepopups/calendarPicker.js:
1150         (initialize):
1151         (openSuggestionPicker):
1152         * Resources/pagepopups/pickerCommon.js:
1153         (enclosingNodeOrSelfWithClass):
1154         * Resources/pagepopups/suggestionPicker.css: Added.
1155         (.suggestion-list):
1156         (.suggestion-list-entry):
1157         (.suggestion-list-entry:focus):
1158         (.suggestion-list-entry:focus .label):
1159         (.suggestion-list-entry .content):
1160         (.suggestion-list-entry .title):
1161         (.suggestion-list-entry .label):
1162         (.measuring-width .suggestion-list-entry .label):
1163         (.suggestion-list .separator):
1164         * Resources/pagepopups/suggestionPicker.js: Added.
1165         (SuggestionPicker):
1166         (SuggestionPicker.validateConfig):
1167         (SuggestionPicker.prototype.cleanup):
1168         (SuggestionPicker.prototype._setColors): Creates css rules that sets highlight colors.
1169         (SuggestionPicker.prototype._createSuggestionEntryElement): Creates an entry that when selected submits the value.
1170         (SuggestionPicker.prototype._createActionEntryElement): Creates an entry that causes an action. (e.x. "Other...")
1171         (SuggestionPicker.prototype._measureMaxContentWidth): Temporarily left align everything to measure the width.
1172         (SuggestionPicker.prototype._fixWindowSize):
1173         (SuggestionPicker.prototype._layout):
1174         (SuggestionPicker.prototype.selectEntry):
1175         (SuggestionPicker.prototype._handleEntryClick):
1176         (SuggestionPicker.prototype._findFirstVisibleEntry):
1177         (SuggestionPicker.prototype._findLastVisibleEntry):
1178         (SuggestionPicker.prototype._handleBodyKeyDown):
1179         (SuggestionPicker.prototype._handleEntryMouseOver):
1180         (SuggestionPicker.prototype._handleMouseOut):
1181         * WebCore.gyp/WebCore.gyp:
1182
1183 2012-09-23  Andreas Kling  <kling@webkit.org>
1184
1185         REGRESSION(r128239): Mutable ElementAttributeData leak their Attribute vectors.
1186         <http://webkit.org/b/97423>
1187
1188         Reviewed by Anders Carlsson.
1189
1190         r128239 added a WTF::deleteOwnedPtr() override for ElementAttributeData*, but ElementAttributeData is
1191         a ref-counted object. Fixed this by overriding deref() instead to call the appropriate subclass destructor.
1192
1193         * dom/ElementAttributeData.h:
1194         (WebCore::ElementAttributeData::deref):
1195
1196 2012-09-23  Byungwoo Lee  <bw80.lee@gmail.com>
1197
1198         Fix build warnings : -Wunused-parameter, -Wparentheses, -Wuninitialized.
1199         https://bugs.webkit.org/show_bug.cgi?id=97306
1200
1201         Reviewed by Benjamin Poulain.
1202
1203         Fix build warning about -Wunused-parameter on ImageBufferCairo.cpp,
1204         ImageDecoder.h by using ASSERT_UNUSED() macro.
1205
1206         * platform/graphics/cairo/ImageBufferCairo.cpp:
1207         (WebCore::encodeImage):
1208         * platform/image-decoders/ImageDecoder.h:
1209         (WebCore::ImageDecoder::rgbColorProfile):
1210         (WebCore::ImageDecoder::inputDeviceColorProfile):
1211
1212 2012-09-23  Andreas Kling  <kling@webkit.org>
1213
1214         Enable ElementAttributeData sharing for non-HTML elements.
1215         <http://webkit.org/b/97413>
1216
1217         Reviewed by Anders Carlsson.
1218
1219         Use the whole qualified tag name when creating the ElementAttributeData cache key
1220         instead of just the localName(). This allows sharing and caching of attribute data
1221         for non-HTML elements and makes the code a little cleaner.
1222
1223         * dom/Element.cpp:
1224         (WebCore::Element::parserSetAttributes):
1225
1226             Enable the ElementAttributeData sharing path for non-HTML elements.
1227
1228         * dom/ElementAttributeData.h:
1229
1230             Make immutableAttributeArray() public so Document can call it instead of getting
1231             the raw data address from a const_cast'ed attributeItem(0).
1232
1233         * dom/Document.cpp:
1234         (WebCore::ImmutableAttributeDataCacheKey::ImmutableAttributeDataCacheKey):
1235         (WebCore::ImmutableAttributeDataCacheKey::operator!=):
1236         (WebCore::ImmutableAttributeDataCacheKey::hash):
1237         (ImmutableAttributeDataCacheKey):
1238         (WebCore::ImmutableAttributeDataCacheEntry::ImmutableAttributeDataCacheEntry):
1239         (ImmutableAttributeDataCacheEntry):
1240         (WebCore::Document::cachedImmutableAttributeData):
1241
1242             Let the immutable attribute data cache use a QualifiedName/attributes key
1243             instead of the AtomicString/attributes we were using before. We still grab the
1244             existingHash() from the QualifiedName::localName() and use that for the actual
1245             HashMap key.
1246
1247 2012-09-23  Mike West  <mkwst@chromium.org>
1248
1249         Measure the usage of the "X-WebKit-CSP" header in the hopes of dropping the prefix completely.
1250         https://bugs.webkit.org/show_bug.cgi?id=97408
1251
1252         Reviewed by Adam Barth.
1253
1254         We plan on landing the unprefixed header in webkit.org/b/96765 once the
1255         spec moves to CR. Though we plan to use it for experimentation in the
1256         near future, we'll want to drop the prefixed header completely at some
1257         point in the future. Starting to measure its usage now will give us a
1258         good baseline when we start contemplating that decision.
1259
1260         No new functionality, so no new tests.
1261
1262         * page/ContentSecurityPolicy.cpp:
1263         (WebCore::ContentSecurityPolicy::didReceiveHeader):
1264         * page/FeatureObserver.h:
1265
1266 2012-09-22  Dominic Mazzoni  <dmazzoni@google.com>
1267
1268         AX: Layout tests would be easier to write if AccessibilityController could find an element by id
1269         https://bugs.webkit.org/show_bug.cgi?id=97301
1270
1271         Reviewed by Chris Fleizach.
1272
1273         Add a method to get the DOM "id" attribute from an accessible object, only for layout tests.
1274
1275         Makes existing test more cross-platform:
1276         Tests: accessibility/aria-hidden-with-elements.html
1277
1278         * accessibility/mac/WebAccessibilityObjectWrapper.mm:
1279         (-[WebAccessibilityObjectWrapper accessibilityAttributeValue:]):
1280
1281 2012-09-22  Thierry Reding  <thierry.reding@avionic-design.de>
1282
1283         [GTK] Build is broken with GLib Unicode backend
1284         https://bugs.webkit.org/show_bug.cgi?id=97308.
1285
1286         Reviewed by Carlos Garcia Campos.
1287
1288         The TextBreakIteratorGtk implementation uses ubrk_close() to delete
1289         the iterator. However that function is ICU specific and not declared
1290         when using the GLib backend. Instead the iterator can simply be freed
1291         using the delete operator.
1292
1293         This commit is based on work by Martin Schlemmer who posted a fix for
1294         this here:
1295
1296             http://permalink.gmane.org/gmane.os.opendarwin.webkit.gtk/1210
1297
1298         No new tests.
1299
1300         * platform/text/gtk/TextBreakIteratorGtk.cpp:
1301         (WebCore::NonSharedCharacterBreakIterator::~NonSharedCharacterBreakIterator):
1302         Free iterator using delete instead of ubrk_close().
1303
1304 2012-09-21  Ami Fischman  <fischman@chromium.org>
1305
1306         HTMLMediaElement isn't garbage collected between document reloads
1307         https://bugs.webkit.org/show_bug.cgi?id=97020
1308
1309         Reviewed by Eric Carlson.
1310
1311         JS-created (as opposed to DOM-created) Audio nodes never got collected, because they
1312         appear to always hasPendingActivity(), because m_playing is never set to false.
1313
1314         Manual test added: ManualTests/audio-freed-during-reload.html
1315
1316         * html/HTMLMediaElement.cpp:
1317         (WebCore::HTMLMediaElement::stop): set m_playing = false; explicitly.
1318
1319 2012-09-21  Simon Fraser  <simon.fraser@apple.com>
1320
1321         RenderMarquee causes ASSERTION FAILED: enclosingIntRect(rendererMappedResult) == enclosingIntRect(FloatQuad(result).boundingBox()) : WebCore::FloatRect WebCore::RenderGeometryMap::absoluteRect(const WebCore::FloatRect &) const
1322         https://bugs.webkit.org/show_bug.cgi?id=92464
1323
1324         Reviewed by Sam Weinig.
1325
1326         Marquees could cause an updateCompositingLayersAfterScroll() to be called when
1327         we're in the middle of updating layer positions. updateCompositingLayersAfterScroll()
1328         does a full RenderLayer tree walk, but its use of RenderGeomeryMap reveals that
1329         it's using layers whose positions haven't been updated yet.
1330
1331         Fix by avoiding the updateCompositingLayersAfterScroll() if we're in the process
1332         of updating a marquee when updating layer positions. We'll do a compositing update
1333         soon anyway.
1334
1335         Tested by fast/events/tabindex-focus-blur-all.html
1336
1337         * rendering/RenderLayer.cpp:
1338         (WebCore::RenderLayer::RenderLayer):
1339         (WebCore::RenderLayer::updateLayerPositions):
1340         (WebCore::RenderLayer::updateLayerPositionsAfterScroll):
1341         (WebCore::RenderLayer::scrollTo):
1342         * rendering/RenderLayer.h:
1343         (RenderLayer):
1344
1345 2012-09-21  Adam Klein  <adamk@chromium.org>
1346
1347         Remove bogus assertions from ChildListMutationScope
1348         https://bugs.webkit.org/show_bug.cgi?id=97372
1349
1350         Reviewed by Ryosuke Niwa.
1351
1352         Some asserts (and their accompanying comment) were trying to enforce
1353         proper usage of ChildListMutationScope from WebCore, but in the
1354         presence of MutationEvents they could fail due to arbitrary script
1355         execution.
1356
1357         This change gets rid of those asserts and adds tests exercising
1358         the (pre-existing) codepaths for handling these out-of-order cases.
1359         Without this patch, these tests ASSERT in debug builds.
1360
1361         Tests: fast/mutation/added-out-of-order.html
1362                fast/mutation/removed-out-of-order.html
1363
1364         * dom/ChildListMutationScope.cpp:
1365         (WebCore::ChildListMutationAccumulator::childAdded):
1366         (WebCore::ChildListMutationAccumulator::willRemoveChild):
1367         * dom/ChildListMutationScope.h:
1368         (WebCore):
1369
1370 2012-09-21  Dan Bernstein  <mitz@apple.com>
1371
1372         REGRESSION (r129176): Incorrect line breaking when kerning occurs between a space and the following character
1373         https://bugs.webkit.org/show_bug.cgi?id=97377
1374
1375         Reviewed by Enrica Casucci.
1376
1377         Test: fast/text/kerning-with-TextLayout.html
1378
1379         When kerning is enabled, the last character in a word may have its advance shortened because
1380         of its trailing space. To account for that, words are measured along with the trailing space,
1381         then the width of a space is subtracted from the result. This doesn’t work when the trailing
1382         space itself has its advance shortened due to the character following it, which can happen
1383         when using the TextLayout optimization. However, when the optimization is used, the advance
1384         of the last character of the word is already adjusted for the trailing space, so there is no
1385         need to measure with that space and subtract its advance.
1386
1387         * rendering/RenderBlockLineLayout.cpp:
1388         (WebCore::RenderBlock::LineBreaker::nextLineBreak): Changed to not use the trailingSpaceWidth
1389         mechanism when TextLayout is being used.
1390
1391 2012-09-21  Adam Klein  <adamk@chromium.org>
1392
1393         Simplify and optimize ChildListMutationScope
1394         https://bugs.webkit.org/show_bug.cgi?id=97352
1395
1396         Reviewed by Ryosuke Niwa.
1397
1398         ChildListMutationScope is one of the most complicated bits of
1399         MutationObserver implementation. This patch aims to simplify it for
1400         clarity and improve its performance (mostly by just doing less).
1401
1402         The big change is to remove the MutationAccumulatorRouter class,
1403         replacing it with lifetime-management logic in ChildListMutationAccumulator
1404         ChildListMutationScope is expected to call getOrCreate() in
1405         its constructor, and each scope holds a RefPtr to the accumulator.
1406         When the last scope holding such a RefPtr is destroyed,
1407         ChildListMutationAccumulator's destructor enqueues the accumulated record.
1408
1409         This greatly reduces the number of lines of code, and condenses
1410         two HashMaps into one. It also reduces hash lookups, which now
1411         occur only on scope creation and when the refcount for a given
1412         accumulator reaches 0 (previously, each childAdded and willRemoveChild
1413         call could result in two hash lookups each).
1414
1415         There are some minor changes as well: the ChildListMutationAccumulator::clear()
1416         method is gone, as it was doing more work than necessary;
1417         DEFINE_STATIC_LOCAL is now used instead of hand-rolled static-management
1418         code; ChildListMutationAccumulator::m_lastAdded is no longer a RefPtr, since it
1419         always points at a Node that's already being ref'd by the accumulator.
1420         Also various minor syntactic cleanups.
1421
1422         No new tests, no change in behavior.
1423
1424         * dom/ChildListMutationScope.cpp:
1425         (WebCore::accumulatorMap): Reduced two maps to one, and manage its lifetime with DEFINE_STATIC_LOCAL.
1426         (WebCore::ChildListMutationAccumulator::ChildListMutationAccumulator): Remove unnecessary call to clear() (which itself has been removed).
1427         (WebCore::ChildListMutationAccumulator::~ChildListMutationAccumulator): Enqueue record if not empty at destruction, and have the accumulator
1428         remove itself from the map.
1429         (WebCore::ChildListMutationAccumulator::getOrCreate): Replaces half of MutationAccumulatorRouter's job.
1430         (WebCore::ChildListMutationAccumulator::childAdded): Minor RefPtr usage improvements.
1431         (WebCore::ChildListMutationAccumulator::isRemovedNodeInOrder): Simplify RefPtr syntax.
1432         (WebCore::ChildListMutationAccumulator::willRemoveChild): Minor RefPtr usage improvements.
1433         (WebCore::ChildListMutationAccumulator::enqueueMutationRecord): Replace call to clear() with clearing m_lastAdded,
1434         since it's the only bit not cleared by the MutationRecord creation call. Also remove
1435         isEmpty check and replace with asserts now that it's a private method.
1436         (WebCore::ChildListMutationAccumulator::isEmpty): Added more assertions about emptiness.
1437         * dom/ChildListMutationScope.h:
1438         (WebCore):
1439         (ChildListMutationAccumulator): Extract the inner class to make everything easier to read.
1440         (WebCore::ChildListMutationScope::ChildListMutationScope): Store m_accumulator rather than m_target.
1441         (WebCore::ChildListMutationScope::~ChildListMutationScope): ditto
1442         (WebCore::ChildListMutationScope::childAdded): ditto
1443         (WebCore::ChildListMutationScope::willRemoveChild): ditto
1444         (ChildListMutationScope):
1445         * html/HTMLElement.cpp: Remove unused ChildListMutationScope.h #include.
1446
1447 2012-09-21  Chris Rogers  <crogers@google.com>
1448
1449         BiquadFilterNode must take audio-rate parameter changes into account
1450         https://bugs.webkit.org/show_bug.cgi?id=97369
1451
1452         Reviewed by Kenneth Russell.
1453
1454         BiquadFilterNode is currently ignoring any timeline or audio-rate changes to its parameters.
1455         We now check if any of its parameters have timeline or audio-rate changes and, if so, take
1456         them into account.  Otherwise, we use ordinary parameter smoothing/de-zippering which is
1457         the case when the parameters are adjusted, for example, from a knob or slider in the UI.
1458
1459         * Modules/webaudio/BiquadDSPKernel.cpp:
1460         (WebCore::BiquadDSPKernel::updateCoefficientsIfNecessary):
1461         * Modules/webaudio/BiquadProcessor.cpp:
1462         (WebCore::BiquadProcessor::checkForDirtyCoefficients):
1463         * Modules/webaudio/BiquadProcessor.h:
1464         (WebCore::BiquadProcessor::hasSampleAccurateValues):
1465         (BiquadProcessor):
1466
1467 2012-09-21  Brandon Jones  <bajones@google.com>
1468
1469         Add support for OES_vertex_array_object in chromium
1470         https://bugs.webkit.org/show_bug.cgi?id=96578
1471
1472         Reviewed by Kenneth Russell.
1473
1474         Adding basic reference counting to WebGLBuffer objects to satisfy spec requirements
1475         for the OES_vertex_array_object extension. Added code to allow calls to the
1476         OES_vertex_array_object extension to interface properly with the chromium WebGL
1477         implementation.
1478
1479         Test: fast/canvas/webgl/oes-vertex-array-object.html
1480
1481         * html/canvas/WebGLRenderingContext.cpp:
1482         (WebCore):
1483         (WebCore::WebGLRenderingContext::deleteBuffer):
1484         (WebCore::WebGLRenderingContext::vertexAttribPointer):
1485         * html/canvas/WebGLVertexArrayObjectOES.cpp:
1486         (WebCore::WebGLVertexArrayObjectOES::setElementArrayBuffer):
1487         (WebCore):
1488         * html/canvas/WebGLVertexArrayObjectOES.h:
1489         (WebGLVertexArrayObjectOES):
1490         (WebCore::WebGLVertexArrayObjectOES::getVertexAttribStateSize):
1491         * platform/chromium/support/Extensions3DChromium.cpp:
1492         (WebCore::Extensions3DChromium::createVertexArrayOES):
1493         (WebCore::Extensions3DChromium::deleteVertexArrayOES):
1494         (WebCore::Extensions3DChromium::isVertexArrayOES):
1495         (WebCore::Extensions3DChromium::bindVertexArrayOES):
1496
1497 2012-09-21  Kenichi Ishibashi  <bashi@chromium.org>
1498
1499         [Chromium] Use OpenTypeVerticalData on Linux
1500         https://bugs.webkit.org/show_bug.cgi?id=97277
1501
1502         Reviewed by Tony Chang.
1503
1504         Remove HarfBuzz dependency from GlyphPageTreeNodeSkia. Use OpenTypeVerticalData instead.
1505
1506         No new tests. Rebaselined existing tests.
1507
1508         * WebCore.gyp/WebCore.gyp: Added OpenTypeTypes.h and OpenTypeVerticalData.(cpp|h) for linux and android.
1509         * platform/graphics/FontCache.cpp:
1510         Inserted a space between > and > in typedef of FontVerticalDataCache so that making some compilers happy.
1511         (WebCore):
1512         * platform/graphics/SimpleFontData.h:
1513         (SimpleFontData): Moved declaration of m_verticalData to avoid compile warnings.
1514         * platform/graphics/harfbuzz/FontPlatformDataHarfBuzz.cpp:
1515         (WebCore):
1516         (WebCore::FontPlatformData::verticalData): Added.
1517         (WebCore::FontPlatformData::openTypeTable): Added.
1518         * platform/graphics/harfbuzz/FontPlatformDataHarfBuzz.h:
1519         (WebCore):
1520         (FontPlatformData):
1521         * platform/graphics/skia/GlyphPageTreeNodeSkia.cpp: Removed substituteWithVerticalGlyphs().
1522         (WebCore::GlyphPage::fill):
1523
1524 2012-09-21  Jeremy Apthorp  <jeremya@chromium.org>
1525
1526         Crash in WebCore::Document::fullScreenChangeDelayTimerFired
1527         https://bugs.webkit.org/show_bug.cgi?id=97367
1528
1529         Reviewed by Abhishek Arya.
1530
1531         The document could be destroyed during the processing of the
1532         fullscreenchange event, if the document was destroyed as a result of
1533         one of the dispatchEvent calls.
1534
1535         This bug isn't reliably reproducible, so no new tests.
1536
1537         * dom/Document.cpp:
1538         (WebCore::Document::fullScreenChangeDelayTimerFired):
1539
1540 2012-09-21  Pratik Solanki  <psolanki@apple.com>
1541
1542         No need to pass order file for WebCoreTestSupport build
1543         https://bugs.webkit.org/show_bug.cgi?id=97363
1544
1545         Reviewed by David Kilzer.
1546
1547         No new tests because no functional change.
1548
1549         * Configurations/WebCoreTestSupport.xcconfig:
1550
1551 2012-09-21  Chris Rogers  <crogers@google.com>
1552
1553         Add Web Audio support for deprecated/legacy APIs
1554         https://bugs.webkit.org/show_bug.cgi?id=97050
1555
1556         Reviewed by Eric Carlson.
1557
1558         The Web Audio API specification has undergone much review and some small API changes
1559         have been made (mostly naming-related changes).  This patch adds an ENABLE_LEGACY_WEB_AUDIO
1560         build option to allow ports to support the old names.
1561
1562         Tests changed:
1563         audiobuffersource-playbackrate.html
1564         audiobuffersource.html
1565         note-grain-on-testing.js
1566         oscillator-testing.js
1567
1568         * Configurations/FeatureDefines.xcconfig:
1569         * GNUmakefile.features.am:
1570         * Modules/webaudio/AudioBufferSourceNode.cpp:
1571         (WebCore::AudioBufferSourceNode::startGrain):
1572         (WebCore):
1573         (WebCore::AudioBufferSourceNode::noteGrainOn):
1574         * Modules/webaudio/AudioBufferSourceNode.h:
1575         (AudioBufferSourceNode):
1576         * Modules/webaudio/AudioBufferSourceNode.idl:
1577         * Modules/webaudio/AudioScheduledSourceNode.cpp:
1578         (WebCore::AudioScheduledSourceNode::start):
1579         (WebCore::AudioScheduledSourceNode::stop):
1580         (WebCore):
1581         (WebCore::AudioScheduledSourceNode::noteOn):
1582         (WebCore::AudioScheduledSourceNode::noteOff):
1583         * Modules/webaudio/AudioScheduledSourceNode.h:
1584         * Modules/webaudio/Oscillator.idl:
1585         * page/FeatureObserver.h:
1586
1587 2012-09-21  Mark Pilgrim  <pilgrim@chromium.org>
1588
1589         [Chromium] remove getFontFamilyForCharacters from PlatformSupport
1590         https://bugs.webkit.org/show_bug.cgi?id=96282
1591
1592         Reviewed by Tony Chang.
1593
1594         Migrating away from PlatformSupport. getFontFamilyForCharacters is
1595         moved to FontCache.h and overridden by the two platforms that
1596         need it (Chromium Linux and Blackberry). New files for the overrides.
1597         Part of a larger refactoring series. See tracking bug 82948.
1598
1599         * PlatformBlackBerry.cmake:
1600         * WebCore.gypi:
1601         * platform/chromium/PlatformSupport.h:
1602         (PlatformSupport):
1603         * platform/graphics/FontCache.h:
1604         (SimpleFontFamily):
1605         (FontCache):
1606         * platform/graphics/blackberry/FontCacheBlackberry.cpp: Added.
1607         (WebCore):
1608         (WebCore::FontCache::getFontFamilyForCharacters):
1609         * platform/graphics/blackberry/skia/PlatformSupport.cpp:
1610         (WebCore):
1611         * platform/graphics/blackberry/skia/PlatformSupport.h:
1612         (PlatformSupport):
1613         * platform/graphics/chromium/FontCacheAndroid.cpp:
1614         (WebCore::FontCache::getFontDataForCharacters):
1615         * platform/graphics/chromium/FontCacheChromiumLinux.cpp: Added.
1616         (WebCore):
1617         (WebCore::FontCache::getFontFamilyForCharacters):
1618         * platform/graphics/skia/FontCacheSkia.cpp:
1619         (WebCore::FontCache::getFontDataForCharacters):
1620
1621 2012-09-21  Chris Fleizach  <cfleizach@apple.com>
1622
1623         AX: WebKit exposes incorrect bounds for embedded SVG in HTML
1624         https://bugs.webkit.org/show_bug.cgi?id=96168
1625
1626         Reviewed by Eric Seidel.
1627
1628         Override absoluteFocusRingQuads() for SVG objects because the default
1629         implementation relies on addFocusRingRects(). In addFocusRingRects(), SVG
1630         objects adds local positions for its rects instead of absolute positions.
1631
1632         Test: accessibility/svg-bounds.html
1633
1634         * rendering/RenderObject.h:
1635         (RenderObject):
1636         * rendering/svg/RenderSVGModelObject.cpp:
1637         (WebCore):
1638         (WebCore::RenderSVGModelObject::absoluteFocusRingQuads):
1639         * rendering/svg/RenderSVGModelObject.h:
1640         (RenderSVGModelObject):
1641
1642 2012-09-21  Lianghui Chen  <liachen@rim.com>
1643
1644         [BlackBerry] Really fix bug 95488 that user can get the authentication challenge dialog while the other tab has focus.
1645         https://bugs.webkit.org/show_bug.cgi?id=97348
1646         Internal PR: 186597.
1647
1648         Internally reviewed by Yong Li, Joe Mason.
1649         Reviewed by Yong Li.
1650
1651         Add a singleton AuthenticationChallengeManager to manage authentication
1652         challenge dialog. It does following things:
1653         Record page creation/deletion, so it knows what page is present or not.
1654         Record page visibility change so it knows when to display a dialog or not.
1655         Accept authentication challenge, and decide whether to postpone the
1656             challenge dialog based on whether there is active authentication challenge
1657             dialog already and whether its page is visible or not.
1658         When a challenge result comes back, notify the result to all clients
1659             authenticating for the same protection space, and then start the next
1660             authentication challenge from the same page, if there is one.
1661         When a page becomes visible, start the first authentication challenge
1662             dialog that has been blocked before.
1663
1664         And to support this new AuthenticationChallengeManager, and making the
1665             challenge really asynchronous, NetworkJob has been updated to support
1666             the concept of "freeze", which means buffering all network loading status
1667             change but don't send them to NetworkJob clients.
1668         This is necessary when authentication challenge is asynchronous, as the
1669             previous network loading status will likely come before user make any
1670             decision.
1671
1672         No new tests for platform specific internal change.
1673
1674         * PlatformBlackBerry.cmake:
1675         * platform/blackberry/AuthenticationChallengeManager.cpp: Added.
1676         (WebCore):
1677         (ChallengeInfo):
1678         (WebCore::ChallengeInfo::ChallengeInfo):
1679         (AuthenticationChallengeManagerPrivate):
1680         (WebCore::AuthenticationChallengeManagerPrivate::AuthenticationChallengeManagerPrivate):
1681         (WebCore::AuthenticationChallengeManagerPrivate::resumeAuthenticationChallenge):
1682         (WebCore::AuthenticationChallengeManagerPrivate::startAuthenticationChallenge):
1683         (WebCore::AuthenticationChallengeManagerPrivate::pageExists):
1684         (WebCore::AuthenticationChallengeManager::AuthenticationChallengeManager):
1685         (WebCore::AuthenticationChallengeManager::pageCreated):
1686         (WebCore::AuthenticationChallengeManager::pageDeleted):
1687         (WebCore::AuthenticationChallengeManager::pageVisibilityChanged):
1688         (WebCore::AuthenticationChallengeManager::authenticationChallenge):
1689         (WebCore::AuthenticationChallengeManager::cancelAuthenticationChallenge):
1690         (WebCore::AuthenticationChallengeManager::notifyChallengeResult):
1691         (WebCore::AuthenticationChallengeManager::instance):
1692         (WebCore::AuthenticationChallengeManager::init):
1693         * platform/blackberry/AuthenticationChallengeManager.h:
1694         (WebCore):
1695         (AuthenticationChallengeManager):
1696         * platform/blackberry/PageClientBlackBerry.h:
1697         * platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:
1698         (WebCore::MediaPlayerPrivate::MediaPlayerPrivate):
1699         (WebCore::MediaPlayerPrivate::~MediaPlayerPrivate):
1700         (WebCore::MediaPlayerPrivate::onAuthenticationNeeded):
1701         (WebCore::MediaPlayerPrivate::notifyChallengeResult):
1702         * platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.h:
1703         (MediaPlayerPrivate):
1704         * platform/network/blackberry/NetworkJob.cpp:
1705         (WebCore::NetworkJob::NetworkJob):
1706         (WebCore::NetworkJob::~NetworkJob):
1707         (WebCore):
1708         (WebCore::NetworkJob::handleNotifyStatusReceived):
1709         (WebCore::NetworkJob::handleNotifyClose):
1710         (WebCore::NetworkJob::sendRequestWithCredentials):
1711         (WebCore::NetworkJob::notifyChallengeResult):
1712         * platform/network/blackberry/NetworkJob.h:
1713         (NetworkJob):
1714
1715 2012-09-21  Simon Hausmann  <simon.hausmann@digia.com>
1716
1717         Unreviewed, rolling out r129248.
1718         http://trac.webkit.org/changeset/129248
1719         https://bugs.webkit.org/show_bug.cgi?id=96000
1720
1721         Broke win build
1722
1723         * Target.pri:
1724         * WebCore.vcproj/WebCore.vcproj:
1725         * rendering/RenderingAllInOne.cpp:
1726
1727 2012-09-21  Simon Hausmann  <simon.hausmann@nokia.com>
1728
1729         Make RenderingAllInOne.cpp usable for ports other than Apple/Win
1730         https://bugs.webkit.org/show_bug.cgi?id=96000
1731
1732         Reviewed by Ryosuke Niwa.
1733
1734         RenderingAllInOne.cpp unconditionally includes RenderThemeWin. This patch separates
1735         it out from the file.
1736
1737         * Target.pri: Add RenderingAllInOne.cpp to the list of supported all-in-one files.
1738         * WebCore.vcproj/WebCore.vcproj: Compile RenderThemeWin.cpp separately.
1739         * rendering/RenderingAllInOne.cpp: Don't include RenderThemeWin.cpp here.
1740
1741 2012-09-21  Joanmarie Diggs  <jdiggs@igalia.com>
1742
1743         REGRESSION (r127882): accessibility/spinbutton-value.html failing on GTK
1744         https://bugs.webkit.org/show_bug.cgi?id=96196
1745
1746         Reviewed by Martin Robinson.
1747
1748         The "regression" is that a new test was added but the support was missing
1749         in the Gtk port for spin buttons.
1750
1751         No new tests. Instead the new test which had been skipped was unskipped
1752         as part of this fix.
1753
1754         * accessibility/gtk/WebKitAccessibleWrapperAtk.cpp:
1755         (atkRole): Map SpinButtonRole to ATK_ROLE_SPIN_BUTTON
1756         (getInterfaceMaskFromObject): Add SpinButtonRole to the roles implementing
1757         the AtkValue interface.
1758
1759 2012-09-21  Joanmarie Diggs  <jdiggs@igalia.com>
1760
1761         [GTK] [Stable] Infinite recursion in WebCore::AXObjectCache::getOrCreate
1762         https://bugs.webkit.org/show_bug.cgi?id=96932
1763
1764         Reviewed by Martin Robinson.
1765
1766         Make the decision based on RenderObjects rather than AccessibilityObjects
1767         to avoid the infinite recursion which occurs when remapAriaRoleDueToParent
1768         gets called.
1769
1770         Test: platform/gtk/accessibility/remapped-aria-crash.html
1771
1772         * accessibility/gtk/AccessibilityObjectAtk.cpp:
1773         (WebCore::AccessibilityObject::accessibilityPlatformIncludesObject):
1774
1775 2012-09-21  Jonathan Dong  <jonathan.dong@torchmobile.com.cn>
1776
1777         [BlackBerry] HTML5 media does not handle SSL certificate failures
1778         https://bugs.webkit.org/show_bug.cgi?id=93324
1779
1780         Reviewed by Eric Carlson.
1781
1782         RIM PR: 116205
1783         Passed FrameLoaderClientBlackBerry's playerId to MMRPlayer::load()
1784         because MMRPlayer::load() added playerId as a new parameter, which
1785         is required to initiate a MediaSSLHandlerStream to deal with
1786         certificate failure when loading a "https" media url.
1787
1788         Internally reviewed by Joe Mason <jmason@rim.com>.
1789
1790         No new tests since there's no functional change.
1791
1792         * platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:
1793         (WebCore::MediaPlayerPrivate::load):
1794
1795 2012-09-21  Evan Wallace  <evan.exe@gmail.com>
1796
1797         [WebSocket] Receiving a large message is really slow
1798         https://bugs.webkit.org/show_bug.cgi?id=97237
1799
1800         Reviewed by Alexey Proskuryakov.
1801
1802         WebSocketChannel always reallocates its internal buffer when it receives
1803         and appends new data which causes dramatic slowdowns for messages over
1804         2 MB in size. This patch changes the internal buffer of WebSocketChannel
1805         from a raw char array to a Vector<char> and uses its amortized append()
1806         method. This brings the time to receive a 5 MB message from 5.2 seconds
1807         to 0.25 seconds.
1808
1809         This patch is only for optimization. No new tests are needed.
1810
1811         * Modules/websockets/WebSocketChannel.cpp:
1812         (WebCore::WebSocketChannel::WebSocketChannel):
1813         (WebCore::WebSocketChannel::~WebSocketChannel):
1814         (WebCore::WebSocketChannel::fail):
1815         (WebCore::WebSocketChannel::resume):
1816         (WebCore::WebSocketChannel::didReceiveSocketStreamData):
1817         (WebCore::WebSocketChannel::appendToBuffer):
1818         (WebCore::WebSocketChannel::skipBuffer):
1819         (WebCore::WebSocketChannel::processBuffer):
1820         (WebCore::WebSocketChannel::resumeTimerFired):
1821         (WebCore::WebSocketChannel::processFrame):
1822         * Modules/websockets/WebSocketChannel.h:
1823
1824 2012-09-21  Andrey Adaikin  <aandrey@chromium.org>
1825
1826         Fix build with ENABLE_WEBGL=false
1827         https://bugs.webkit.org/show_bug.cgi?id=97309
1828
1829         Reviewed by Eric Seidel.
1830
1831         WebKit no longer builds when WEBGL is not enabled.
1832
1833         * rendering/FilterEffectRenderer.h:
1834         (FilterEffectRenderer):
1835
1836 2012-09-21  Jonathan Dong  <jonathan.dong@torchmobile.com.cn>
1837
1838         [BlackBerry] Media player won't resize for the following source when first source fail to load
1839         https://bugs.webkit.org/show_bug.cgi?id=97342
1840
1841         Reviewed by Yong Li.
1842
1843         As platformPlayer will notify MediaPlayerPrivate for size change
1844         when loading metadata failed (in this case hasVideo() is false),
1845         we should prevent MediaPlayerPrivate to set width and height
1846         attribute of media element, otherwise we won't get the correct
1847         dimension for the following media sources.
1848
1849         Internally reviewed by Max Feil.
1850
1851         Test case: media/video-size.html
1852
1853         * platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:
1854         (WebCore::MediaPlayerPrivate::resizeSourceDimensions):
1855
1856 2012-09-21  Dan Bernstein  <mitz@apple.com>
1857
1858         REGRESSION (r126763): Incorrect line breaking when both kerning and word spacing are enabled
1859         https://bugs.webkit.org/show_bug.cgi?id=97280
1860
1861         Reviewed by Adele Peterson.
1862
1863         Font::width() never applies word spacing to the first character in the TextRun. The
1864         TextLayout optimization tried to achieve this behavior by not applying word spacing to
1865         any character, which led to this bug.
1866
1867         Test: fast/text/word-space-with-kerning-2.html
1868
1869         * platform/graphics/mac/ComplexTextController.cpp:
1870         (WebCore::TextLayout::TextLayout): Changed to use the given font rather than a version
1871         without word spacing.
1872         (WebCore::TextLayout::width): Added a check if the run starts with a space at a non-zero
1873         offset. If that is the case, then the ComplexTextController has added word spacing to that
1874         space, so subtract it here in order to maintain the behavior described above.
1875
1876 2012-09-21  Sheriff Bot  <webkit.review.bot@gmail.com>
1877
1878         Unreviewed, rolling out r129219.
1879         http://trac.webkit.org/changeset/129219
1880         https://bugs.webkit.org/show_bug.cgi?id=97338
1881
1882         Presumably broke Apple Mac compilation (Requested by yurys_ on
1883         #webkit).
1884
1885         * bindings/js/JSHTMLCanvasElementCustom.cpp:
1886         (WebCore::JSHTMLCanvasElement::getContext):
1887         * bindings/v8/custom/V8HTMLCanvasElementCustom.cpp:
1888         (WebCore::V8HTMLCanvasElement::getContextCallback):
1889         * inspector/InjectedScriptCanvasModule.cpp:
1890         (WebCore::InjectedScriptCanvasModule::wrapWebGLContext):
1891         (WebCore):
1892         * inspector/InjectedScriptCanvasModule.h:
1893         (InjectedScriptCanvasModule):
1894         * inspector/InspectorCanvasAgent.cpp:
1895         * inspector/InspectorCanvasAgent.h:
1896         * inspector/InspectorCanvasInstrumentation.h:
1897         * inspector/InspectorInstrumentation.h:
1898         (InspectorInstrumentation):
1899
1900 2012-09-21  Andrey Adaikin  <aandrey@chromium.org>
1901
1902         Web Inspector: [Canvas] support 2D canvas instrumentation from the inspector C++ code
1903         https://bugs.webkit.org/show_bug.cgi?id=97203
1904
1905         Reviewed by Yury Semikhatsky.
1906
1907         Implements wrapping a 2D canvas context through the injected canvas module script facility.
1908
1909         * bindings/js/JSHTMLCanvasElementCustom.cpp:
1910         (WebCore::JSHTMLCanvasElement::getContext):
1911         * bindings/v8/custom/V8HTMLCanvasElementCustom.cpp:
1912         (WebCore::V8HTMLCanvasElement::getContextCallback):
1913         * inspector/InjectedScriptCanvasModule.cpp:
1914         (WebCore::InjectedScriptCanvasModule::wrapCanvas2DContext):
1915         (WebCore):
1916         (WebCore::InjectedScriptCanvasModule::wrapWebGLContext):
1917         (WebCore::InjectedScriptCanvasModule::callWrapContextFunction):
1918         * inspector/InjectedScriptCanvasModule.h:
1919         (InjectedScriptCanvasModule):
1920         * inspector/InspectorCanvasAgent.cpp:
1921         (WebCore::InspectorCanvasAgent::wrapCanvas2DRenderingContextForInstrumentation):
1922         (WebCore):
1923         * inspector/InspectorCanvasAgent.h:
1924         (InspectorCanvasAgent):
1925         * inspector/InspectorCanvasInstrumentation.h:
1926         (WebCore::InspectorInstrumentation::wrapCanvas2DRenderingContextForInstrumentation):
1927         (WebCore):
1928         * inspector/InspectorInstrumentation.h:
1929         (InspectorInstrumentation):
1930
1931 2012-09-20  Andrey Kosyakov  <caseq@chromium.org>
1932
1933         Web Inspector: [refactoring] simplify interface to FileOutputStream
1934         https://bugs.webkit.org/show_bug.cgi?id=97226
1935
1936         Reviewed by Yury Semikhatsky.
1937
1938         - change OutputStream interface to match that of stream;
1939         - fix OutputStream implementations (FileOutputStream and those in heap profiler);
1940         - fix usages in Timeline and HeapProfiler.
1941
1942         * inspector/front-end/FileUtils.js:
1943         (WebInspector.OutputStream.prototype.write):
1944         (WebInspector.OutputStream.prototype.close):
1945         (WebInspector.ChunkedFileReader.prototype.start):
1946         (WebInspector.ChunkedFileReader.prototype._onChunkLoaded):
1947         (WebInspector.ChunkedXHRReader.prototype.start):
1948         (WebInspector.ChunkedXHRReader.prototype._onProgress):
1949         (WebInspector.ChunkedXHRReader.prototype._onLoad):
1950         (WebInspector.FileOutputStream):
1951         (WebInspector.FileOutputStream.prototype.open.callbackWrapper):
1952         (WebInspector.FileOutputStream.prototype.open):
1953         (WebInspector.FileOutputStream.prototype.write):
1954         (WebInspector.FileOutputStream.prototype.close):
1955         (WebInspector.FileOutputStream.prototype._onAppendDone):
1956         * inspector/front-end/HeapSnapshotLoader.js:
1957         (WebInspector.HeapSnapshotLoader.prototype.close):
1958         (WebInspector.HeapSnapshotLoader.prototype.write):
1959         * inspector/front-end/HeapSnapshotProxy.js:
1960         (WebInspector.HeapSnapshotLoaderProxy.prototype.startTransfer):
1961         (WebInspector.HeapSnapshotLoaderProxy.prototype.write):
1962         (WebInspector.HeapSnapshotLoaderProxy.prototype.close):
1963         * inspector/front-end/HeapSnapshotView.js:
1964         (WebInspector.HeapProfileHeader):
1965         (WebInspector.HeapProfileHeader.prototype.load):
1966         (WebInspector.HeapProfileHeader.prototype._setupWorker):
1967         (WebInspector.HeapProfileHeader.prototype.dispose):
1968         (WebInspector.HeapProfileHeader.prototype.transferChunk.callback):
1969         (WebInspector.HeapProfileHeader.prototype.transferChunk):
1970         (WebInspector.HeapProfileHeader.prototype.finishHeapSnapshot):
1971         (WebInspector.HeapProfileHeader.prototype.saveToFile.onOpen):
1972         (WebInspector.HeapProfileHeader.prototype.saveToFile):
1973         * inspector/front-end/TimelineModel.js:
1974         (WebInspector.TimelineModel.prototype._createFileWriter):
1975         (WebInspector.TimelineModel.prototype.saveToFile.callback):
1976         (WebInspector.TimelineModel.prototype.saveToFile):
1977         (WebInspector.TimelineModelLoader):
1978         (WebInspector.TimelineModelLoader.prototype.write):
1979         (WebInspector.TimelineModelLoader.prototype.close):
1980         (WebInspector.TimelineSaver):
1981         (WebInspector.TimelineSaver.prototype.save):
1982         (WebInspector.TimelineSaver.prototype._writeNextChunk):
1983
1984 2012-09-21  Alexandru Chiculita  <achicu@adobe.com>
1985
1986         -webkit-clip-path is applied on elements that are not descendant of the container
1987         https://bugs.webkit.org/show_bug.cgi?id=97217
1988
1989         Reviewed by Dirk Schulze.
1990
1991         The clip-path was set on the GraphicsContext, but was never restored, thus making all the layers
1992         rendered in the same "group" of save/restore state use the same clip-path.
1993
1994         Test: css3/masking/clip-path-restore.html
1995
1996         * css/StyleResolver.cpp:
1997         (WebCore::StyleResolver::collectMatchingRulesForList):
1998         clip-path property should create a stacking-context, otherwise the RenderLayers will not be nested,
1999         meaning that the clip-path of the parent is not going to apply correctly.
2000
2001         * rendering/RenderLayer.cpp:
2002         (WebCore::RenderLayer::paintLayerContents):
2003
2004 2012-09-21  Kent Tamura  <tkent@chromium.org>
2005
2006         Remove unused functions of LocalizedDate.h
2007         https://bugs.webkit.org/show_bug.cgi?id=97311
2008
2009         Reviewed by Kentaro Hara.
2010
2011         WebCore::localizedTimeFormatText, WebCore::localizedShortTimeFormatText,
2012         and WebCore::timeAMPMLabels are not used any more because we switched to
2013         the corresponding functions of Localizer.
2014
2015         No new tests because of no behavior changes.
2016
2017         * platform/text/Localizer.h:
2018         (Localizer): Moved comments from LocalizedDate.h.
2019         * platform/text/LocalizedDate.h:
2020         (WebCore): Removed localizedTimeFormatText,
2021         localizedShortTimeFormatText, and timeAMPMLabels.
2022         * platform/text/LocalizedDateICU.cpp:
2023         (WebCore): ditto.
2024         * platform/text/LocalizedDateWin.cpp:
2025         (WebCore): ditto.
2026         * platform/text/mac/LocalizedDateMac.cpp:
2027         (WebCore): ditto.
2028
2029 2012-09-21  Keishi Hattori  <keishi@webkit.org>
2030
2031         Add datalist suggestions into DateTimeChooserParameters
2032         https://bugs.webkit.org/show_bug.cgi?id=97292
2033
2034         Reviewed by Kent Tamura.
2035
2036         We read datalist suggestions, add them to DateTimeChooserParameters,
2037         and pass them to the page popup.
2038
2039         No new tests. No behavior change yet.
2040
2041         * html/shadow/CalendarPickerElement.cpp:
2042         (WebCore::CalendarPickerElement::openPopup): Read datalist suggestions and add them to DateTimeChooserParameters
2043         * platform/DateTimeChooser.h:
2044         (DateTimeChooserParameters): Added localizedSuggestionValues so we can show localized values inside the page popup.
2045
2046 2012-09-21  Yoshifumi Inoue  <yosin@chromium.org>
2047
2048         [Forms] DateTimeEditElement::layout() should take date time format as a parameter
2049         https://bugs.webkit.org/show_bug.cgi?id=97300
2050
2051         Reviewed by Kent Tamura.
2052
2053         This patch introduces DateTimeEditElement::LayoutParameters struct for
2054         passing four parameters to DateTimeEditElement::layout() for passing
2055         date time format from client of DateTimeEditElement instead of
2056         DateTimeEditElement::layout() takes time or short time format.
2057
2058         This patch is a part of preparation of introducing multiple fields
2059         date/datetime/month/week input UI.
2060
2061         This patch affects ports which enable both ENABLE_INPUT_TYPE_TIME and
2062         ENABLE_INPUT_TYPE_TIME_MULTIPLE_FIELDS.
2063
2064         No new tests. This patch doesn't change behavior.
2065
2066         * html/TimeInputType.cpp:
2067         (WebCore::TimeInputType::updateInnerTextValue): Changed to use LayoutParmeters.
2068         * html/shadow/DateTimeEditElement.cpp:  Removed unused include files, LocalizedDate.h and LocalizedNumber.h.
2069         (DateTimeEditBuilder):
2070         (WebCore::DateTimeEditBuilder::DateTimeEditBuilder): Changed parameters to LayoutParmeters.
2071         (WebCore::DateTimeEditElement::LayoutParameters::shouldHaveSecondField): Moved from DateTimeEditBuilder::needSecondField().
2072         (WebCore::DateTimeEditElement::layout): Changed to take LayoutParameters.
2073         (WebCore::DateTimeEditElement::setValueAsDate): ditto
2074         (WebCore::DateTimeEditElement::setEmptyValue): ditto
2075         * html/shadow/DateTimeEditElement.h: Removed unused classe declarations DateComponents and DateTimeEditLayouter.
2076         (LayoutParameters):  Added to bundle parameters for layout().
2077         (WebCore::DateTimeEditElement::LayoutParameters::LayoutParameters): Added.
2078
2079 2012-09-20  Pavel Feldman  <pfeldman@chromium.org>
2080
2081         Web Inspector: render grid scale to the right / at bottom in case box is close to 0 on that axis.
2082         https://bugs.webkit.org/show_bug.cgi?id=97219
2083
2084         Reviewed by Vsevolod Vlasov.
2085
2086         Otherwise, it is hard to inspect objects close to (0, 0)
2087
2088         * inspector/InspectorOverlayPage.html:
2089
2090 2012-09-21  Sheriff Bot  <webkit.review.bot@gmail.com>
2091
2092         Unreviewed, rolling out r129086.
2093         http://trac.webkit.org/changeset/129086
2094         https://bugs.webkit.org/show_bug.cgi?id=97312
2095
2096         Broke input rendering (Requested by shinyak on #webkit).
2097
2098         * html/BaseButtonInputType.cpp:
2099         (WebCore):
2100         * html/BaseButtonInputType.h:
2101         (WebCore::BaseButtonInputType::BaseButtonInputType):
2102         (BaseButtonInputType):
2103         * html/FileInputType.cpp:
2104         (WebCore::UploadButtonElement::create):
2105         (WebCore::UploadButtonElement::createForMultiple):
2106         * html/HTMLInputElement.cpp:
2107         (WebCore::HTMLInputElement::parseAttribute):
2108         * html/InputType.cpp:
2109         * html/InputType.h:
2110         (InputType):
2111         * rendering/RenderButton.cpp:
2112         (WebCore::RenderButton::RenderButton):
2113         (WebCore::RenderButton::styleDidChange):
2114         (WebCore::RenderButton::updateFromElement):
2115         (WebCore):
2116         (WebCore::RenderButton::setText):
2117         (WebCore::RenderButton::text):
2118         * rendering/RenderButton.h:
2119         (RenderButton):
2120
2121 2012-09-21  Kihong Kwon  <kihong.kwon@samsung.com>
2122
2123         Remove useless class prototypes from Vibration.h
2124         https://bugs.webkit.org/show_bug.cgi?id=97304
2125
2126         Reviewed by Kentaro Hara.
2127
2128         Remove two useless class prototype statements in the Vibration.h
2129
2130         * Modules/vibration/Vibration.h:
2131
2132 2012-09-21  Keishi Hattori  <keishi@webkit.org>
2133
2134         Prepare CalendarPicker so we can add another picker, SuggetionPicker
2135         https://bugs.webkit.org/show_bug.cgi?id=97193
2136
2137         Reviewed by Kent Tamura.
2138
2139         Preparation so we can add another picker to CalendarPicker and switch
2140         between them.
2141
2142         No new tests. No behavior change.
2143
2144         * Resources/pagepopups/calendarPicker.css:
2145         (.calendar-picker): Added so we can apply these styles just to calendar picker.
2146         * Resources/pagepopups/calendarPicker.js:
2147         (CalendarPicker.validateConfig): Renamed so each picker can validate the config object.
2148         (initialize):
2149         (closePicker): Call Picker.cleanup().
2150         (openCalendarPicker):
2151         (CalendarPicker):
2152         (CalendarPicker.prototype.cleanup): Cleanup event listener on document.body.
2153         * Resources/pagepopups/pickerCommon.js:
2154         (Picker.prototype.cleanup):
2155
2156 2012-09-20  John Mellor  <johnme@chromium.org>
2157
2158         Text Autosizing: Cluster text at flow roots, for consistency and to avoid autosizing headers/footers.
2159         https://bugs.webkit.org/show_bug.cgi?id=97025
2160
2161         Reviewed by Julien Chaffraix.
2162
2163         This patch has 3 main changes:
2164
2165         1. All text within a "cluster" (roughly equivalent to a CSS flow root /
2166            block formatting context) must have a uniform textAutosizingMultiplier,
2167            except for subtrees which are themselves clusters. This improves the
2168            consistency of the final output, since sibling blocks are now more
2169            likely to have the same multiplier (hence grow in proportion).
2170
2171         2. Clusters must contain a minimum amount of text in order to be
2172            autosized (4 lines of text, assuming each char is 1em wide, so about
2173            2 lines of text in practice). This is to reduce the likelihood of
2174            autosizing things like headers and footers, which can be quite
2175            visually distracting. The rationale is that if a cluster contains
2176            very few lines of text then it's ok to have to zoom in and pan from
2177            side to side to read each line, since if there are very few lines of
2178            text you'll only need to pan across once or twice.
2179
2180         3. To avoid adding a 3rd tree traversal, processSubtree/processBox were
2181            refactored such that all of Text Autosizing now happens as a single
2182            tree traversal (hence halving the number of tree traversals done).
2183
2184         Tests: fast/text-autosizing/cluster-narrow-in-wide.html
2185                fast/text-autosizing/cluster-wide-in-narrow.html
2186                fast/text-autosizing/clusters-insufficient-text.html
2187                fast/text-autosizing/clusters-insufficient-width.html
2188                fast/text-autosizing/clusters-sufficient-text-except-in-root.html
2189                fast/text-autosizing/clusters-sufficient-width.html
2190
2191         * rendering/TextAutosizer.cpp:
2192         (TextAutosizingWindowInfo):
2193
2194             Added this struct to bundle together the various sizes that
2195             processSubtree needs to pass to every recursive call to
2196             processCluster and processContainer.
2197
2198         (WebCore::TextAutosizer::processSubtree):
2199
2200             - Bundle windowSize and minLayoutSize together as a single struct,
2201               TextAutosizingWindowInfo, rather than passing them separately.
2202             - Walk up the tree to find the current cluster and container, rather
2203               than (incorrectly) assuming that the layoutRoot is always a
2204               container.
2205             - Call processCluster instead of traversing the tree.
2206
2207         (WebCore::TextAutosizer::processCluster):
2208
2209             Calculates the multiplier based on the width of the cluster (moved
2210             the calculation here from processBox, since now the multiplier is
2211             fixed per cluster), and delegates to processContainer for the actual
2212             tree traversal (since clusters are also containers).
2213
2214         (WebCore::contentHeightIsConstrained):
2215
2216             Changed parameter to RenderBlock.
2217
2218         (WebCore::TextAutosizer::processContainer):
2219
2220             This now takes care of the whole tree traversal, recursively calling
2221             processCluster/processContainer when it encounters such an object,
2222             and setMultiplier on RenderText objects (as processBox used to).
2223             Also added a check that the RenderText's multiplier is not already
2224             equal to the target multiplier (to save needlessly setting it).
2225
2226         (WebCore::TextAutosizer::isContainer):
2227
2228             - Changed to be a positive (is) instead of negative (isNot) check.
2229             - Require objects to be RenderBlocks so it's easier to find them
2230               using containingBlock, and there don't seem to be many interesting
2231               RenderBoxes that aren't RenderBlocks.
2232
2233         (WebCore::TextAutosizer::isAutosizingCluster):
2234
2235             A container that is also a flow root / block formatting context
2236             (approximately), hence demarcates an independent region of the page,
2237             within which we want consistent autosizing.
2238
2239         (WebCore::TextAutosizer::clusterShouldBeAutosized):
2240
2241             Uses measureDescendantTextWidth and to check whether the cluster
2242             contains enough text to be worth autosizing.
2243
2244         (WebCore::TextAutosizer::measureDescendantTextWidth):
2245
2246             Recursively traverse the cluster, skipping constrained height
2247             containers as processContainer does, to measure how much autosizable
2248             text it contains. Early out as soon as the minimum text width is
2249             reached.
2250
2251         (WebCore::TextAutosizer::nextInPreOrderSkippingDescendantsOfContainers):
2252
2253             Replaces nextInPreOrderMatchingFilter. The filter is now fixed to
2254             isContainer (we no longer need an isAutosizingCluster filter, since
2255             I consolidated the tree traversal), and filtered objects are
2256             actually returned (so they can in turn be recursively traversed),
2257             it's just their descendants that get skipped.
2258
2259         * rendering/TextAutosizer.h:
2260         (TextAutosizer):
2261
2262             - Deleted RenderObjectFilterFunctor, since the filter of
2263               nextInPreOrderSkippingDescendantsOfContainers is now fixed.
2264
2265 2012-09-20  Yoshifumi Inoue  <yosin@chromium.org>
2266
2267         [Platform] There are memory leak in LocaleICU
2268         https://bugs.webkit.org/show_bug.cgi?id=97289
2269
2270         Reviewed by Kent Tamura.
2271
2272         This patch adds udt_close() calls for medium time format and short
2273         time format data used in LocaleICU class to avoid memory leak.
2274
2275         This memory leak is found by external tool Valgrind and reported in
2276         Chromium bug repositry(http://crbug.com/151006) with stack trace of
2277         call path of leaked memory.
2278
2279         This patch affects ports which enable both ENABLE_INPUT_TYPE_TIME and
2280         ENABLE_INPUT_TYPE_TIME_MULTIPLE_FIELDS.
2281
2282         No new tests. External tool such as Valgrind will check this memory leak.
2283
2284         * platform/text/LocaleICU.cpp:
2285         (WebCore::LocaleICU::~LocaleICU): Added to call udt_close() for m_mediumTimeFormat
2286         and m_shortTimeFormat which have UDateFormat objects.
2287
2288 2012-09-20  Kenichi Ishibashi  <bashi@chromium.org>
2289
2290         Chromium mac cannot display AppleColorEmoji
2291         https://bugs.webkit.org/show_bug.cgi?id=97286
2292
2293         Reviewed by Kent Tamura.
2294
2295         Disable AppleColorEmoji for now. We will re-enable it after Skia supports CTFontDrawGlyphs().
2296
2297         No new tests. Fallback fonts should be used for emoji codepoints.
2298
2299         * platform/graphics/mac/FontCacheMac.mm:
2300         (WebCore::FontCache::getFontDataForCharacters):
2301
2302 2012-09-20  Luke Macpherson   <macpherson@chromium.org>
2303
2304         Fix use-after free when using a variable to specify a -webkit-filter.
2305         https://bugs.webkit.org/show_bug.cgi?id=97153
2306
2307         Reviewed by Abhishek Arya.
2308
2309         Make StyleResolver's m_pendingSVGDocuments a hashmap of RefPtr instead of raw pointers such that the document values cannot be freed prematurely.
2310         Present assumption is that storing raw pointers is ok because CSSValues will live as long as the StyleResolver instance, however that it no longer
2311         true when variables are used, so we must ensure we increment the reference counter to ensure the CSSValues are not freed prematurely.
2312
2313         Test: fast/css/variables/var-filter.html
2314
2315         * css/StyleResolver.cpp:
2316         (WebCore::StyleResolver::collectMatchingRulesForList):
2317         * css/StyleResolver.h:
2318         (StyleResolver):
2319
2320 2012-09-20  Levi Weintraub  <leviw@chromium.org>
2321
2322         Prevent reading stale data from InlineTextBoxes
2323         https://bugs.webkit.org/show_bug.cgi?id=94750
2324
2325         Reviewed by Abhishek Arya.
2326
2327         Text from dirty InlineTextBoxes should never be read or used. This change
2328         enforces this design goal by forcefully zero-ing out the start and length
2329         of InlineTextBoxes when they're being marked dirty. Ideally, we'd also
2330         add asserts to the accessors for this data, but there are still several
2331         places in editing that cause this. https://bugs.webkit.org/show_bug.cgi?id=97264
2332         tracks these cases.
2333
2334         This change involves making markDirty virtual. Running the line-layout
2335         performance test as well as profiling resizing the html5 spec showed
2336         negligable impact with this change.
2337
2338         No new tests as this doesn't change any proper behavior.
2339
2340         * dom/Position.cpp:
2341         (WebCore::Position::downstream): Adding a FIXME.
2342         * rendering/InlineBox.h:
2343         (WebCore::InlineBox::markDirty): Marking virtual to allow InlineTextBox to
2344         overload and zero out its start and length.
2345         * rendering/InlineTextBox.cpp:
2346         (WebCore::InlineTextBox::markDirty): Zeroing out the start and length when
2347         we mark the box dirty.
2348         * rendering/InlineTextBox.h:
2349         * rendering/RenderText.cpp:
2350         (WebCore::RenderText::setTextWithOffset): Adding a FIXME.
2351
2352 2012-09-20  Adam Barth  <abarth@webkit.org>
2353
2354         Measure how often web pages use Worker and SharedWorker
2355         https://bugs.webkit.org/show_bug.cgi?id=97273
2356
2357         Reviewed by Ojan Vafai.
2358
2359         We're not considering removing these features, but it will give us a
2360         baseline idea of how often they're used.
2361
2362         * page/FeatureObserver.h:
2363         * workers/SharedWorker.cpp:
2364         (WebCore::SharedWorker::create):
2365         * workers/Worker.cpp:
2366         (WebCore::Worker::create):
2367
2368 2012-09-20  Tony Chang  <tony@chromium.org>
2369
2370         [chromium] Enable more clang warnings
2371         https://bugs.webkit.org/show_bug.cgi?id=97257
2372
2373         Reviewed by James Robinson.
2374
2375         Add -Wunused-parameter to WebCore targets. Apple's Mac compile already
2376         warns on this and it's a common source of error for Chromium patches.
2377
2378         No new tests, should compile cleanly.
2379
2380         * WebCore.gyp/WebCore.gyp:
2381
2382 2012-09-20  Dan Bernstein  <mitz@apple.com>
2383
2384         Kerning never occurs between a space and the following glyph
2385         https://bugs.webkit.org/show_bug.cgi?id=97269
2386
2387         Reviewed by Tim Horton.
2388
2389         Covered by several existing tests.
2390
2391         * platform/graphics/mac/ComplexTextController.cpp:
2392         (WebCore::ComplexTextController::adjustGlyphsAndAdvances): Changed to not reset the advance
2393         of a space, and added a comment about how this also needs to be fixed for other characters
2394         that are treated as spaces.
2395
2396 2012-09-20  Kenichi Ishibashi  <bashi@chromium.org>
2397
2398         [Chromium] Improve glyph selection of HarfBuzzShaper
2399         https://bugs.webkit.org/show_bug.cgi?id=97164
2400
2401         Reviewed by Tony Chang.
2402
2403         Take into account clusters for selection.
2404
2405         Test: fast/text/international/hebrew-selection.html
2406
2407         * platform/graphics/harfbuzz/ng/HarfBuzzShaper.cpp:
2408         (WebCore::HarfBuzzShaper::HarfBuzzRun::applyShapeResult): Removed m_logClusters.
2409         m_logCluster is no longer used.
2410         (WebCore::HarfBuzzShaper::HarfBuzzRun::characterIndexForXPosition):
2411         - If targetX is in the left side of the first cluster, return the leftmost character index.
2412         - If targetX is in the right side of the last cluster, return the rightmost character index.
2413         - If targetX is between the right side of the cluster N and the left side of the cluster N+1, then:
2414           - return N+1 for LTR.
2415           - return N for RTL.
2416         (WebCore::HarfBuzzShaper::HarfBuzzRun::xPositionForOffset):
2417         Find the cluster of index in question, then:
2418         - return the left side boundary of the cluster for LTR.
2419         - return the right side boundary of the cluster for RTL.
2420         * platform/graphics/harfbuzz/ng/HarfBuzzShaper.h:
2421         (HarfBuzzRun):
2422
2423 2012-09-20  Tony Chang  <tony@chromium.org>
2424
2425         Replace RenderListBox::updateLogicalHeight with RenderListBox::computeLogicalHeight
2426         https://bugs.webkit.org/show_bug.cgi?id=97263
2427
2428         Reviewed by Ojan Vafai.
2429
2430         This is part of making computeLogicalHeight virtual so with any RenderBox pointer, one
2431         can compute the logical height without mutating the RenderBox.
2432
2433         No new tests, this is a refactor and existing list box tests should pass.
2434
2435         * rendering/RenderBox.h:
2436         (RenderBox):
2437         * rendering/RenderListBox.cpp:
2438         (WebCore::RenderListBox::layout): Move layout related logic here.
2439         (WebCore::RenderListBox::computeLogicalHeight): Use const version and remove layout related code.
2440         * rendering/RenderListBox.h:
2441         (RenderListBox): Override computeLogicalHeight.
2442
2443 2012-09-20  Mike West  <mkwst@chromium.org>
2444
2445         CSP reports should send an empty "blocked-uri" rather than nothing.
2446         https://bugs.webkit.org/show_bug.cgi?id=97256
2447
2448         Reviewed by Adam Barth.
2449
2450         In cases where a Content Security Policy violation report is generated
2451         without blocking a resource at a particular URI (inline scripts, for
2452         example), we currently leave the "blocked-uri" attribute out of the
2453         report entirely. For the same reason that we included the "referrer"
2454         attribute in webkit.org/b/97233, we should include an explicitly empty
2455         "blocked-uri" in these cases.
2456
2457         This new behavior is covered by updates to existing test expectations
2458         around the reporting functionality.
2459
2460         * page/ContentSecurityPolicy.cpp:
2461         (WebCore::ContentSecurityPolicy::reportViolation):
2462             If the 'blocked-uri' is invalid, add a "blocked-uri" attribute that
2463             is explicitly empty.
2464
2465 2012-09-20  Patrick Gansterer  <paroga@webkit.org>
2466
2467         Add String::numberToFixedPrecisionString()
2468         https://bugs.webkit.org/show_bug.cgi?id=96330
2469
2470         Reviewed by Benjamin Poulain.
2471
2472         * platform/text/TextStream.cpp:
2473         (WebCore::TextStream::operator<<): Use the new function instead of String::number() with flags.
2474
2475 2012-09-20  Adam Klein  <adamk@chromium.org>
2476
2477         Rename ContainerNode::parserAddChild "parserAppendChild" for consistency
2478         https://bugs.webkit.org/show_bug.cgi?id=97254
2479
2480         Reviewed by Adam Barth.
2481
2482         No functional change, all the below changes are simple renames.
2483
2484         * dom/ContainerNode.cpp:
2485         (WebCore::ContainerNode::takeAllChildrenFrom):
2486         (WebCore::ContainerNode::parserAppendChild):
2487         * dom/ContainerNode.h:
2488         (ContainerNode):
2489         * dom/DOMImplementation.cpp:
2490         (WebCore::DOMImplementation::createDocument):
2491         * editing/markup.cpp:
2492         (WebCore::createFragmentForTransformToFragment):
2493         * html/HTMLViewSourceDocument.cpp:
2494         (WebCore::HTMLViewSourceDocument::createContainingTable):
2495         (WebCore::HTMLViewSourceDocument::addSpanWithClassName):
2496         (WebCore::HTMLViewSourceDocument::addLine):
2497         (WebCore::HTMLViewSourceDocument::finishLine):
2498         (WebCore::HTMLViewSourceDocument::addText):
2499         (WebCore::HTMLViewSourceDocument::addBase):
2500         (WebCore::HTMLViewSourceDocument::addLink):
2501         * html/parser/HTMLConstructionSite.cpp:
2502         (WebCore::executeTask):
2503         * html/parser/HTMLTreeBuilder.cpp:
2504         (WebCore::HTMLTreeBuilder::callTheAdoptionAgency):
2505         * html/track/WebVTTParser.cpp:
2506         (WebCore::WebVTTParser::constructTreeFromToken):
2507         * xml/XMLErrors.cpp:
2508         (WebCore::createXHTMLParserErrorHeader):
2509         (WebCore::XMLErrors::insertErrorMessageBlock):
2510         * xml/parser/XMLDocumentParser.cpp:
2511         (WebCore::XMLDocumentParser::enterText):
2512         (WebCore::XMLDocumentParser::parseDocumentFragment):
2513         * xml/parser/XMLDocumentParserLibxml2.cpp:
2514         (WebCore::XMLDocumentParser::startElementNs):
2515         (WebCore::XMLDocumentParser::processingInstruction):
2516         (WebCore::XMLDocumentParser::cdataBlock):
2517         (WebCore::XMLDocumentParser::comment):
2518         (WebCore::XMLDocumentParser::internalSubset):
2519         * xml/parser/XMLDocumentParserQt.cpp:
2520         (WebCore::XMLDocumentParser::parseStartElement):
2521         (WebCore::XMLDocumentParser::parseProcessingInstruction):
2522         (WebCore::XMLDocumentParser::parseCdata):
2523         (WebCore::XMLDocumentParser::parseComment):
2524         (WebCore::XMLDocumentParser::parseDtd):
2525         * xml/parser/XMLTreeBuilder.cpp:
2526         (WebCore::XMLTreeBuilder::processDOCTYPE):
2527         (WebCore::XMLTreeBuilder::processStartTag):
2528         (WebCore::XMLTreeBuilder::add):
2529
2530 2012-09-20  James Robinson  <jamesr@chromium.org>
2531
2532         [chromium] Remove unused copy of chromium compositor implementation files
2533         https://bugs.webkit.org/show_bug.cgi?id=97255
2534
2535         Reviewed by Adam Barth.
2536
2537         Now that use_libcc_for_compositor is set to 1 these files aren't used anywhere.
2538
2539         * WebCore.gyp/WebCore.gyp:
2540         * WebCore.gypi:
2541         * platform/chromium/support/CCThreadImpl.cpp: Removed.
2542         * platform/chromium/support/CCThreadImpl.h: Removed.
2543         * platform/chromium/support/WebCompositorImpl.cpp: Removed.
2544         * platform/chromium/support/WebCompositorImpl.h: Removed.
2545         * platform/graphics/chromium/BitmapCanvasLayerTextureUpdater.cpp: Removed.
2546         * platform/graphics/chromium/BitmapCanvasLayerTextureUpdater.h: Removed.
2547         * platform/graphics/chromium/BitmapSkPictureCanvasLayerTextureUpdater.cpp: Removed.
2548         * platform/graphics/chromium/BitmapSkPictureCanvasLayerTextureUpdater.h: Removed.
2549         * platform/graphics/chromium/Canvas2DLayerBridge.cpp:
2550         (WebCore::Canvas2DLayerBridge::Canvas2DLayerBridge):
2551         (WebCore::Canvas2DLayerBridge::~Canvas2DLayerBridge):
2552         * platform/graphics/chromium/CanvasLayerTextureUpdater.cpp: Removed.
2553         * platform/graphics/chromium/CanvasLayerTextureUpdater.h: Removed.
2554         * platform/graphics/chromium/ContentLayerChromium.cpp: Removed.
2555         * platform/graphics/chromium/ContentLayerChromium.h: Removed.
2556         * platform/graphics/chromium/ContentLayerChromiumClient.h: Removed.
2557         * platform/graphics/chromium/FrameBufferSkPictureCanvasLayerTextureUpdater.cpp: Removed.
2558         * platform/graphics/chromium/FrameBufferSkPictureCanvasLayerTextureUpdater.h: Removed.
2559         * platform/graphics/chromium/GeometryBinding.cpp: Removed.
2560         * platform/graphics/chromium/GeometryBinding.h: Removed.
2561         * platform/graphics/chromium/GraphicsLayerChromium.cpp:
2562         * platform/graphics/chromium/HeadsUpDisplayLayerChromium.cpp: Removed.
2563         * platform/graphics/chromium/HeadsUpDisplayLayerChromium.h: Removed.
2564         * platform/graphics/chromium/IOSurfaceLayerChromium.cpp: Removed.
2565         * platform/graphics/chromium/IOSurfaceLayerChromium.h: Removed.
2566         * platform/graphics/chromium/ImageLayerChromium.cpp: Removed.
2567         * platform/graphics/chromium/ImageLayerChromium.h: Removed.
2568         * platform/graphics/chromium/LayerChromium.cpp: Removed.
2569         * platform/graphics/chromium/LayerChromium.h: Removed.
2570         * platform/graphics/chromium/LayerTextureSubImage.cpp: Removed.
2571         * platform/graphics/chromium/LayerTextureSubImage.h: Removed.
2572         * platform/graphics/chromium/LayerTextureUpdater.h: Removed.
2573         * platform/graphics/chromium/PlatformColor.h: Removed.
2574         * platform/graphics/chromium/ProgramBinding.cpp: Removed.
2575         * platform/graphics/chromium/ProgramBinding.h: Removed.
2576         * platform/graphics/chromium/RateLimiter.cpp: Removed.
2577         * platform/graphics/chromium/RateLimiter.h: Removed.
2578         * platform/graphics/chromium/RenderSurfaceChromium.cpp: Removed.
2579         * platform/graphics/chromium/RenderSurfaceChromium.h: Removed.
2580         * platform/graphics/chromium/ScrollbarLayerChromium.cpp: Removed.
2581         * platform/graphics/chromium/ScrollbarLayerChromium.h: Removed.
2582         * platform/graphics/chromium/ShaderChromium.cpp: Removed.
2583         * platform/graphics/chromium/ShaderChromium.h: Removed.
2584         * platform/graphics/chromium/SkPictureCanvasLayerTextureUpdater.cpp: Removed.
2585         * platform/graphics/chromium/SkPictureCanvasLayerTextureUpdater.h: Removed.
2586         * platform/graphics/chromium/SolidColorLayerChromium.cpp: Removed.
2587         * platform/graphics/chromium/SolidColorLayerChromium.h: Removed.
2588         * platform/graphics/chromium/TextureCopier.cpp: Removed.
2589         * platform/graphics/chromium/TextureCopier.h: Removed.
2590         * platform/graphics/chromium/TextureLayerChromium.cpp: Removed.
2591         * platform/graphics/chromium/TextureLayerChromium.h: Removed.
2592         * platform/graphics/chromium/TextureLayerChromiumClient.h: Removed.
2593         * platform/graphics/chromium/ThrottledTextureUploader.cpp: Removed.
2594         * platform/graphics/chromium/ThrottledTextureUploader.h: Removed.
2595         * platform/graphics/chromium/TiledLayerChromium.cpp: Removed.
2596         * platform/graphics/chromium/TiledLayerChromium.h: Removed.
2597         * platform/graphics/chromium/TreeSynchronizer.cpp: Removed.
2598         * platform/graphics/chromium/TreeSynchronizer.h: Removed.
2599         * platform/graphics/chromium/UnthrottledTextureUploader.h: Removed.
2600         * platform/graphics/chromium/VideoLayerChromium.cpp: Removed.
2601         * platform/graphics/chromium/VideoLayerChromium.h: Removed.
2602         * platform/graphics/chromium/cc/CCActiveAnimation.cpp: Removed.
2603         * platform/graphics/chromium/cc/CCActiveAnimation.h: Removed.
2604         * platform/graphics/chromium/cc/CCAnimationCurve.cpp: Removed.
2605         * platform/graphics/chromium/cc/CCAnimationCurve.h: Removed.
2606         * platform/graphics/chromium/cc/CCAnimationEvents.h: Removed.
2607         * platform/graphics/chromium/cc/CCAppendQuadsData.h: Removed.
2608         * platform/graphics/chromium/cc/CCCheckerboardDrawQuad.cpp: Removed.
2609         * platform/graphics/chromium/cc/CCCheckerboardDrawQuad.h: Removed.
2610         * platform/graphics/chromium/cc/CCCompletionEvent.h: Removed.
2611         * platform/graphics/chromium/cc/CCDamageTracker.cpp: Removed.
2612         * platform/graphics/chromium/cc/CCDamageTracker.h: Removed.
2613         * platform/graphics/chromium/cc/CCDebugBorderDrawQuad.cpp: Removed.
2614         * platform/graphics/chromium/cc/CCDebugBorderDrawQuad.h: Removed.
2615         * platform/graphics/chromium/cc/CCDebugRectHistory.cpp: Removed.
2616         * platform/graphics/chromium/cc/CCDebugRectHistory.h: Removed.
2617         * platform/graphics/chromium/cc/CCDelayBasedTimeSource.cpp: Removed.
2618         * platform/graphics/chromium/cc/CCDelayBasedTimeSource.h: Removed.
2619         * platform/graphics/chromium/cc/CCDirectRenderer.cpp: Removed.
2620         * platform/graphics/chromium/cc/CCDirectRenderer.h: Removed.
2621         * platform/graphics/chromium/cc/CCDrawQuad.cpp: Removed.
2622         * platform/graphics/chromium/cc/CCDrawQuad.h: Removed.
2623         * platform/graphics/chromium/cc/CCFontAtlas.cpp: Removed.
2624         * platform/graphics/chromium/cc/CCFontAtlas.h: Removed.
2625         * platform/graphics/chromium/cc/CCFrameRateController.cpp: Removed.
2626         * platform/graphics/chromium/cc/CCFrameRateController.h: Removed.
2627         * platform/graphics/chromium/cc/CCFrameRateCounter.cpp: Removed.
2628         * platform/graphics/chromium/cc/CCFrameRateCounter.h: Removed.
2629         * platform/graphics/chromium/cc/CCGraphicsContext.h: Removed.
2630         * platform/graphics/chromium/cc/CCHeadsUpDisplayLayerImpl.cpp: Removed.
2631         * platform/graphics/chromium/cc/CCHeadsUpDisplayLayerImpl.h: Removed.
2632         * platform/graphics/chromium/cc/CCIOSurfaceDrawQuad.cpp: Removed.
2633         * platform/graphics/chromium/cc/CCIOSurfaceDrawQuad.h: Removed.
2634         * platform/graphics/chromium/cc/CCIOSurfaceLayerImpl.cpp: Removed.
2635         * platform/graphics/chromium/cc/CCIOSurfaceLayerImpl.h: Removed.
2636         * platform/graphics/chromium/cc/CCInputHandler.h: Removed.
2637         * platform/graphics/chromium/cc/CCKeyframedAnimationCurve.cpp: Removed.
2638         * platform/graphics/chromium/cc/CCKeyframedAnimationCurve.h: Removed.
2639         * platform/graphics/chromium/cc/CCLayerAnimationController.cpp: Removed.
2640         * platform/graphics/chromium/cc/CCLayerAnimationController.h: Removed.
2641         * platform/graphics/chromium/cc/CCLayerImpl.cpp: Removed.
2642         * platform/graphics/chromium/cc/CCLayerImpl.h: Removed.
2643         * platform/graphics/chromium/cc/CCLayerIterator.cpp: Removed.
2644         * platform/graphics/chromium/cc/CCLayerIterator.h: Removed.
2645         * platform/graphics/chromium/cc/CCLayerQuad.cpp: Removed.
2646         * platform/graphics/chromium/cc/CCLayerQuad.h: Removed.
2647         * platform/graphics/chromium/cc/CCLayerSorter.cpp: Removed.
2648         * platform/graphics/chromium/cc/CCLayerSorter.h: Removed.
2649         * platform/graphics/chromium/cc/CCLayerTilingData.cpp: Removed.
2650         * platform/graphics/chromium/cc/CCLayerTilingData.h: Removed.
2651         * platform/graphics/chromium/cc/CCLayerTreeHost.cpp: Removed.
2652         * platform/graphics/chromium/cc/CCLayerTreeHost.h: Removed.
2653         * platform/graphics/chromium/cc/CCLayerTreeHostClient.h: Removed.
2654         * platform/graphics/chromium/cc/CCLayerTreeHostCommon.cpp: Removed.
2655         * platform/graphics/chromium/cc/CCLayerTreeHostCommon.h: Removed.
2656         * platform/graphics/chromium/cc/CCLayerTreeHostImpl.cpp: Removed.
2657         * platform/graphics/chromium/cc/CCLayerTreeHostImpl.h: Removed.
2658         * platform/graphics/chromium/cc/CCMathUtil.cpp: Removed.
2659         * platform/graphics/chromium/cc/CCMathUtil.h: Removed.
2660         * platform/graphics/chromium/cc/CCOcclusionTracker.cpp: Removed.
2661         * platform/graphics/chromium/cc/CCOcclusionTracker.h: Removed.
2662         * platform/graphics/chromium/cc/CCOverdrawMetrics.cpp: Removed.
2663         * platform/graphics/chromium/cc/CCOverdrawMetrics.h: Removed.
2664         * platform/graphics/chromium/cc/CCPageScaleAnimation.cpp: Removed.
2665         * platform/graphics/chromium/cc/CCPageScaleAnimation.h: Removed.
2666         * platform/graphics/chromium/cc/CCPrioritizedTexture.cpp: Removed.
2667         * platform/graphics/chromium/cc/CCPrioritizedTexture.h: Removed.
2668         * platform/graphics/chromium/cc/CCPrioritizedTextureManager.cpp: Removed.
2669         * platform/graphics/chromium/cc/CCPrioritizedTextureManager.h: Removed.
2670         * platform/graphics/chromium/cc/CCPriorityCalculator.cpp: Removed.
2671         * platform/graphics/chromium/cc/CCPriorityCalculator.h: Removed.
2672         * platform/graphics/chromium/cc/CCProxy.cpp: Removed.
2673         * platform/graphics/chromium/cc/CCProxy.h: Removed.
2674         * platform/graphics/chromium/cc/CCQuadCuller.cpp: Removed.
2675         * platform/graphics/chromium/cc/CCQuadCuller.h: Removed.
2676         * platform/graphics/chromium/cc/CCQuadSink.h: Removed.
2677         * platform/graphics/chromium/cc/CCRenderPass.cpp: Removed.
2678         * platform/graphics/chromium/cc/CCRenderPass.h: Removed.
2679         * platform/graphics/chromium/cc/CCRenderPassDrawQuad.cpp: Removed.
2680         * platform/graphics/chromium/cc/CCRenderPassDrawQuad.h: Removed.
2681         * platform/graphics/chromium/cc/CCRenderPassSink.h: Removed.
2682         * platform/graphics/chromium/cc/CCRenderSurface.cpp: Removed.
2683         * platform/graphics/chromium/cc/CCRenderSurface.h: Removed.
2684         * platform/graphics/chromium/cc/CCRenderSurfaceFilters.cpp: Removed.
2685         * platform/graphics/chromium/cc/CCRenderSurfaceFilters.h: Removed.
2686         * platform/graphics/chromium/cc/CCRenderer.h: Removed.
2687         * platform/graphics/chromium/cc/CCRendererGL.cpp: Removed.
2688         * platform/graphics/chromium/cc/CCRendererGL.h: Removed.
2689         * platform/graphics/chromium/cc/CCRenderingStats.h: Removed.
2690         * platform/graphics/chromium/cc/CCResourceProvider.cpp: Removed.
2691         * platform/graphics/chromium/cc/CCResourceProvider.h: Removed.
2692         * platform/graphics/chromium/cc/CCScheduler.cpp: Removed.
2693         * platform/graphics/chromium/cc/CCScheduler.h: Removed.
2694         * platform/graphics/chromium/cc/CCSchedulerStateMachine.cpp: Removed.
2695         * platform/graphics/chromium/cc/CCSchedulerStateMachine.h: Removed.
2696         * platform/graphics/chromium/cc/CCScopedTexture.cpp: Removed.
2697         * platform/graphics/chromium/cc/CCScopedTexture.h: Removed.
2698         * platform/graphics/chromium/cc/CCScopedThreadProxy.h: Removed.
2699         * platform/graphics/chromium/cc/CCScrollbarAnimationController.cpp: Removed.
2700         * platform/graphics/chromium/cc/CCScrollbarAnimationController.h: Removed.
2701         * platform/graphics/chromium/cc/CCScrollbarAnimationControllerLinearFade.cpp: Removed.
2702         * platform/graphics/chromium/cc/CCScrollbarAnimationControllerLinearFade.h: Removed.
2703         * platform/graphics/chromium/cc/CCScrollbarGeometryFixedThumb.cpp: Removed.
2704         * platform/graphics/chromium/cc/CCScrollbarGeometryFixedThumb.h: Removed.
2705         * platform/graphics/chromium/cc/CCScrollbarGeometryStub.cpp: Removed.
2706         * platform/graphics/chromium/cc/CCScrollbarGeometryStub.h: Removed.
2707         * platform/graphics/chromium/cc/CCScrollbarLayerImpl.cpp: Removed.
2708         * platform/graphics/chromium/cc/CCScrollbarLayerImpl.h: Removed.
2709         * platform/graphics/chromium/cc/CCSettings.cpp: Removed.
2710         * platform/graphics/chromium/cc/CCSettings.h: Removed.
2711         * platform/graphics/chromium/cc/CCSharedQuadState.cpp: Removed.
2712         * platform/graphics/chromium/cc/CCSharedQuadState.h: Removed.
2713         * platform/graphics/chromium/cc/CCSingleThreadProxy.cpp: Removed.
2714         * platform/graphics/chromium/cc/CCSingleThreadProxy.h: Removed.
2715         * platform/graphics/chromium/cc/CCSolidColorDrawQuad.cpp: Removed.
2716         * platform/graphics/chromium/cc/CCSolidColorDrawQuad.h: Removed.
2717         * platform/graphics/chromium/cc/CCSolidColorLayerImpl.cpp: Removed.
2718         * platform/graphics/chromium/cc/CCSolidColorLayerImpl.h: Removed.
2719         * platform/graphics/chromium/cc/CCStreamVideoDrawQuad.cpp: Removed.
2720         * platform/graphics/chromium/cc/CCStreamVideoDrawQuad.h: Removed.
2721         * platform/graphics/chromium/cc/CCTexture.cpp: Removed.
2722         * platform/graphics/chromium/cc/CCTexture.h: Removed.
2723         * platform/graphics/chromium/cc/CCTextureDrawQuad.cpp: Removed.
2724         * platform/graphics/chromium/cc/CCTextureDrawQuad.h: Removed.
2725         * platform/graphics/chromium/cc/CCTextureLayerImpl.cpp: Removed.
2726         * platform/graphics/chromium/cc/CCTextureLayerImpl.h: Removed.
2727         * platform/graphics/chromium/cc/CCTextureUpdateController.cpp: Removed.
2728         * platform/graphics/chromium/cc/CCTextureUpdateController.h: Removed.
2729         * platform/graphics/chromium/cc/CCTextureUpdateQueue.cpp: Removed.
2730         * platform/graphics/chromium/cc/CCTextureUpdateQueue.h: Removed.
2731         * platform/graphics/chromium/cc/CCThread.h: Removed.
2732         * platform/graphics/chromium/cc/CCThreadProxy.cpp: Removed.
2733         * platform/graphics/chromium/cc/CCThreadProxy.h: Removed.
2734         * platform/graphics/chromium/cc/CCThreadTask.h: Removed.
2735         * platform/graphics/chromium/cc/CCTileDrawQuad.cpp: Removed.
2736         * platform/graphics/chromium/cc/CCTileDrawQuad.h: Removed.
2737         * platform/graphics/chromium/cc/CCTiledLayerImpl.cpp: Removed.
2738         * platform/graphics/chromium/cc/CCTiledLayerImpl.h: Removed.
2739         * platform/graphics/chromium/cc/CCTimeSource.h: Removed.
2740         * platform/graphics/chromium/cc/CCTimer.cpp: Removed.
2741         * platform/graphics/chromium/cc/CCTimer.h: Removed.
2742         * platform/graphics/chromium/cc/CCTimingFunction.cpp: Removed.
2743         * platform/graphics/chromium/cc/CCTimingFunction.h: Removed.
2744         * platform/graphics/chromium/cc/CCVideoLayerImpl.cpp: Removed.
2745         * platform/graphics/chromium/cc/CCVideoLayerImpl.h: Removed.
2746         * platform/graphics/chromium/cc/CCYUVVideoDrawQuad.cpp: Removed.
2747         * platform/graphics/chromium/cc/CCYUVVideoDrawQuad.h: Removed.
2748
2749 2012-09-20  Tony Chang  <tony@chromium.org>
2750
2751         Implement absolutely positioned flex items
2752         https://bugs.webkit.org/show_bug.cgi?id=93798
2753
2754         Reviewed by Ojan Vafai.
2755
2756         Previously, we treated absolutely positioned flex items as a 0x0 placeholder element.
2757         Now we position the 0x0 placeholder where the next item would go. This causes the
2758         following changes:
2759         - justify-content: space-{around,between} no longer change due to the existence of absolutely positioned flex items.
2760         - alignment doesn't change the placement of absolutely positioned flex items.
2761         - auto margins in the alignment direction don't do anything on absolutely positioned flex items.
2762
2763         Test: css3/flexbox/position-absolute-children.html
2764
2765         * rendering/RenderFlexibleBox.cpp:
2766         (WebCore::RenderFlexibleBox::availableAlignmentSpaceForChild): Absolutely positioned flex items should not use this.
2767         (WebCore::RenderFlexibleBox::updateAutoMarginsInCrossAxis): Absolutely positioned flex items should not use this.
2768         (WebCore::initialJustifyContentOffset): If there are no flex items, space-around should center an absolutely positioned flex item.
2769         (WebCore::RenderFlexibleBox::numberOfInFlowPositionedChildren): Helper method for helping to compute space-between and space-around.
2770         (WebCore::RenderFlexibleBox::layoutAndPlaceChildren): Fix spacing when space-between or space-around.
2771         (WebCore::RenderFlexibleBox::layoutColumnReverse): Fix spacing when space-between or space-around.
2772         (WebCore::RenderFlexibleBox::alignChildren): Treat absolutely positioned children like flex-start.
2773         * rendering/RenderFlexibleBox.h: numberOfInFlowPositionedChildren method.
2774
2775 2012-09-20  Sheriff Bot  <webkit.review.bot@gmail.com>
2776
2777         Unreviewed, rolling out r129144.
2778         http://trac.webkit.org/changeset/129144
2779         https://bugs.webkit.org/show_bug.cgi?id=97244
2780
2781         causing lots of assertions in tests (Requested by smfr on
2782         #webkit).
2783
2784         * rendering/InlineBox.h:
2785         (WebCore::InlineBox::markDirty):
2786         * rendering/InlineTextBox.cpp:
2787         * rendering/InlineTextBox.h:
2788         (WebCore::InlineTextBox::start):
2789         (WebCore::InlineTextBox::end):
2790         (WebCore::InlineTextBox::len):
2791         (WebCore::InlineTextBox::offsetRun):
2792
2793 2012-09-20  Mike West  <mkwst@chromium.org>
2794
2795         CSP reports should send an empty 'referrer' rather than nothing.
2796         https://bugs.webkit.org/show_bug.cgi?id=97233
2797
2798         Reviewed by Adam Barth.
2799
2800         Currently, if a protected resource doesn't have a referrer, then any
2801         Content Security Policy violations send a report that doesn't contain
2802         a referrer attribute. It's arguably friendlier to developers to include
2803         an explicitly empty attribute.
2804
2805         This new behavior is covered by updates to existing test expectations
2806         around the reporting functionality.
2807
2808         * page/ContentSecurityPolicy.cpp:
2809         (WebCore::ContentSecurityPolicy::reportViolation):
2810             Drop the 'if', and always write out a referrer.
2811
2812 2012-09-20  David Barton  <dbarton@mathscribe.com>
2813
2814         [MathML] Increase visual space around fraction parts, italic variables, and operators
2815         https://bugs.webkit.org/show_bug.cgi?id=97228
2816
2817         Reviewed by Eric Seidel.
2818
2819         This makes MathML more readable, and more in agreement with Firefox and TeX.
2820
2821         Tested by existing tests.
2822
2823         * css/mathml.css:
2824         (mfrac > :first-child):
2825         (mfrac > :last-child):
2826         (mfrac):
2827         (mi):
2828         (msub > mi:first-child, msubsup > mi:first-child):
2829             - Subscripts don't need to move right after an italic <mi>.
2830         (msubsup > mi:first-child + * + *):
2831             - Superscripts do need to move right after an italic <mi>.
2832         (math > mo, mrow > mo, msqrt > mo, mtd > mo):
2833             - These are ok values for now. It will be better to use different values for different operators, as the FIXME says.
2834         (math > mo:last-child, mrow > mo:last-child, msqrt > mo:last-child, mtd > mo:last-child):
2835             - Prefix and postfix operators, including fences, generally get less spacing than infix operators.
2836
2837         * rendering/mathml/RenderMathMLFenced.cpp:
2838         (WebCore::RenderMathMLFenced::createMathMLOperator):
2839             - Usually the separator is a comma or semicolon, so we only put space after it for now.
2840         (WebCore::RenderMathMLFenced::makeFences):
2841         (WebCore::RenderMathMLFenced::addChild):
2842         (WebCore::RenderMathMLFenced::styleDidChange):
2843         * rendering/mathml/RenderMathMLFenced.h:
2844         (RenderMathMLFenced):
2845
2846         * rendering/mathml/RenderMathMLFraction.cpp:
2847         (WebCore):
2848         (WebCore::RenderMathMLFraction::updateFromElement):
2849             - gDenominatorPad is now handled by mathml.css.
2850
2851         * rendering/mathml/RenderMathMLSubSup.cpp:
2852         (WebCore):
2853         (WebCore::RenderMathMLSubSup::fixScriptsStyle):
2854             - gSubsupScriptMargin is now handled by mathml.css.
2855
2856 2012-09-20  Tommy Widenflycht  <tommyw@google.com>
2857
2858         MediaStream API: Extend UserMediaRequest with a ownerDocument method
2859         https://bugs.webkit.org/show_bug.cgi?id=97095
2860
2861         Reviewed by Adam Barth.
2862
2863         Chromium need to know exactly which frame called getUserMedia so that it can
2864         clean away the stream when the frame goes away.
2865         Since that information is available in webkit add an accessor method.
2866
2867         Chromium mock class extended to test the added method.
2868
2869         * Modules/mediastream/UserMediaRequest.cpp:
2870         (WebCore::UserMediaRequest::ownerDocument):
2871         (WebCore):
2872         * Modules/mediastream/UserMediaRequest.h:
2873         (WebCore):
2874         (UserMediaRequest):
2875
2876 2012-09-20  Levi Weintraub  <leviw@chromium.org>
2877
2878         Prevent reading stale data from InlineTextBoxes
2879         https://bugs.webkit.org/show_bug.cgi?id=94750
2880
2881         Reviewed by Eric Seidel.
2882
2883         Text from dirty InlineTextBoxes should never be read or used. This change enforces this
2884         design goal by forcefully zero-ing out the start and length of InlineTextBoxes when
2885         they're being marked dirty. It also adds asserts to accessors for those members.
2886
2887         This change involves making markDirty virtual. Running the line-layout performance test
2888         as well as profiling resizing the html5 spec showed negligable impact with this change.
2889
2890         No new tests as this doesn't change any proper behavior.
2891
2892         * rendering/InlineBox.h:
2893         (WebCore::InlineBox::markDirty): Making virtual to allow InlineTextBox to overload and
2894         zero out its start and length.
2895         * rendering/InlineTextBox.cpp:
2896         (WebCore::InlineTextBox::markDirty): Zeroing out start and length when we mark the box dirty.
2897         * rendering/InlineTextBox.h:
2898         (WebCore::InlineTextBox::start): Adding an assert when we hit this case.
2899         (WebCore::InlineTextBox::end): Ditto.
2900         (WebCore::InlineTextBox::len): Ditto.
2901         (WebCore::InlineTextBox::offsetRun): Ditto.
2902
2903 2012-09-20  Mike West  <mkwst@chromium.org>
2904
2905         Support paths in Content Security Policy directives.
2906         https://bugs.webkit.org/show_bug.cgi?id=89750
2907
2908         Reviewed by Adam Barth.
2909
2910         In CSP 1.0, paths are simply ignored: 'script-src
2911         http://example.com/path/to/a/file' would allow script to be loaded from
2912         http://example.com/path/to/a/file/javascript.js, but also from
2913         http://example.com/javascript.js.
2914
2915         This patch is an experimental implementation of more granular path
2916         support in CSP source lists as proposed in the current editor's draft of
2917         CSP 1.1. Paths are treated as specifying directories in which resources
2918         can be found, and are implicitly terminated with a '/': in other words,
2919         'script-src http://a.com/path' is the same as
2920         'script-src http://a.com/path/'. Moreover, paths cannot contain either
2921         '?' or '#' characters.
2922
2923         This is implemented outside the CSP_NEXT flag. All ports will be
2924         effected.
2925
2926         Spec: https://dvcs.w3.org/hg/content-security-policy/raw-file/tip/csp-specification.dev.html#matching
2927
2928         Tests: http/tests/security/contentSecurityPolicy/source-list-parsing-paths-01.html
2929                http/tests/security/contentSecurityPolicy/source-list-parsing-paths-02.html
2930
2931         * page/ContentSecurityPolicy.cpp:
2932         (WebCore::CSPSource::CSPSource):
2933             Store a path along with each CSP source.
2934         (WebCore::CSPSource::matches):
2935             Check the path when comparing a URL to the source.
2936         (WebCore::CSPSource::pathMatches):
2937             Compare the URL-decoded version of the resource to validate against
2938             the source's stored path. If the resource's path begins with the
2939             stored path, then it matches! If not, it doesn't.
2940         (CSPSource):
2941             Store a path along with each CSP source.
2942         (WebCore::CSPSourceList::parse):
2943             Pass a 'path' in when creating CSPSource objects.
2944         (WebCore::CSPSourceList::parsePath):
2945             Actually parse the path, flagging errors if '?' or '#' are present,
2946             URL-decoding the result, and ensuring that a terminal '/' is
2947             added if necessary.
2948         (WebCore::CSPSourceList::addSourceSelf):
2949             Ensure that 'self' sources have an empty path.
2950         * page/ContentSecurityPolicy.h:
2951             Dropping the "ignored path component" console warning.
2952
2953 2012-09-20  Joanmarie Diggs  <jdiggs@igalia.com>
2954
2955         [GTK] ControlsPanel string is not localized in LocalizedStringsGtk
2956         https://bugs.webkit.org/show_bug.cgi?id=96502
2957
2958         Reviewed by Chris Fleizach.
2959
2960         Adds the ControlsPanel string to the strings localized in LocalizedStringsGtk.
2961
2962         Test: platform/gtk/accessibility/media-controls-panel-title.html
2963
2964         * platform/gtk/LocalizedStringsGtk.cpp:
2965         (WebCore::localizedMediaControlElementString):
2966
2967 2012-09-20  Julien Chaffraix  <jchaffraix@webkit.org>
2968
2969         Remove isStartColumn in the border collapsing code
2970         https://bugs.webkit.org/show_bug.cgi?id=97024
2971
2972         Reviewed by Abhishek Arya.
2973
2974         isStartColumn is embedding the same information as prevCell. As we need to compute it
2975         in most of the cases, we may as well just reuse them.
2976
2977         While touching this code, I cleaned up the code by removing some unneeded checks and renaming
2978         some variables in preparation for bug 79272.
2979
2980         Refactoring covered by existing collapsing borders tests.
2981
2982         * rendering/RenderTableCell.cpp:
2983         (WebCore::RenderTableCell::computeCollapsedStartBorder):
2984         Removed |isStartColumn|.
2985
2986         (WebCore::RenderTableCell::computeCollapsedEndBorder):
2987         Added a comment about why |isEndColumn| is needed. Also cleaned up this code to be
2988         consistent with computeCollapsedStartBorder.
2989
2990 2012-09-20  Andrey Adaikin  <aandrey@chromium.org>
2991
2992         Web Inspector: setPropertyValue does not work for non-finite numbers
2993         https://bugs.webkit.org/show_bug.cgi?id=97016
2994
2995         Reviewed by Vsevolod Vlasov.
2996
2997         Fix: setting a property to NaN, Infinity or -Infinity numbers did not work.
2998
2999         * inspector/front-end/RemoteObject.js:
3000         (WebInspector.RemoteObject.prototype.setPropertyValue):
3001
3002 2012-09-20  Otto Derek Cheung  <otcheung@rim.com>
3003
3004         Web Inspector: Cookie info in Network Resources Cookies tab are incorrect
3005         https://bugs.webkit.org/show_bug.cgi?id=95491
3006
3007         Reviewed by Vsevolod Vlasov.
3008
3009         Web Inspector: Cookie info in Network Resources Cookies tab are incorrect
3010         https://bugs.webkit.org/show_bug.cgi?id=95491
3011
3012         Fixing a typo that causes cookies to appear as a "Session"
3013         cookie in the Networking panel because it can never find a properly named
3014         "expired" header.
3015
3016         This bug exposes another bug where the cookie GMT strings are inproperly
3017         parsed, causing an "invalid date" error to show up in the cookies tab in
3018         the Networking panel.
3019
3020         Also changed test expectations in LayoutTests/inspector/cookie-parser-expected.txt
3021
3022         * inspector/front-end/CookieParser.js:
3023         (WebInspector.Cookie.prototype.get session):
3024
3025 2012-09-20  Alexander Pavlov  <apavlov@chromium.org>
3026
3027         Web Inspector: [Elements] Word wrap disablement in the DOM tree broken
3028         https://bugs.webkit.org/show_bug.cgi?id=97185
3029
3030         Reviewed by Pavel Feldman.
3031
3032         Restored the effective "white-space: nowrap" for the tree elements.
3033
3034         * inspector/front-end/inspector.css:
3035         (.nowrap):
3036
3037 2012-09-20  Allan Sandfeld Jensen  <allan.jensen@nokia.com>
3038
3039         [TouchAdjustment] Simplify and improve hybrid distance function.
3040         https://bugs.webkit.org/show_bug.cgi?id=96519
3041
3042         Reviewed by Antonio Gomes.
3043
3044         The current distance function is a combination of two functions. One measuring the distance from
3045         the hot-spot in the touch-area to the centerline of the target, and one measuring how much of the
3046         target is covered.
3047
3048         The distance to the center-line was used instead of just the distance to the target, to make it
3049         easier to hit small targets near big targets. The very same feature is however also the reason
3050         measuring how much the target is covered is added. Using the distance to center-line is therefore
3051         redundant now, and can be replaced with the simpler 'distance the hot-spot needs to be adjusted'.
3052
3053         Tested by existing touchadjustment tests.
3054
3055         * page/TouchAdjustment.cpp:
3056         (TouchAdjustment):
3057         (WebCore::TouchAdjustment::hybridDistanceFunction):
3058         * platform/graphics/IntRect.cpp:
3059         * platform/graphics/IntRect.h:
3060         (IntRect):
3061
3062 2012-09-20  Yoshifumi Inoue  <yosin@chromium.org>
3063
3064         [Forms] HTMLSelectElement should call formStateDidChange on both menulist and lisbox mode
3065         https://bugs.webkit.org/show_bug.cgi?id=97177
3066
3067         Reviewed by Kent Tamura.
3068
3069         This patch makes listbox mode select element to call formStateDidChange()
3070         when selected options are changed.
3071
3072         For this change, this patch moves notifyFormStateChanged() to
3073         HTMLFormControlElementWithState class from HTMLTextFormControlElement
3074         for sharing code HTMLSelectElement class and HTMLInputElement/HTMLTextAreaElement
3075         derived from HTMLTextFormControlElement.
3076
3077         No new tests. We can't test this change in WebKit test tools. Test script
3078         will be implemented in Chromium side.
3079
3080         * html/HTMLFormControlElementWithState.cpp:
3081         (WebCore::HTMLFormControlElementWithState::notifyFormStateChanged):
3082         (WebCore):
3083         * html/HTMLFormControlElementWithState.h: Moved a declaration of notifyFormStateChanged() from HTMLTextFormControlElement.
3084         (HTMLFormControlElementWithState): Moved an implemented of notifyFormStateChanged() from HTMLTextFormControlElement.
3085         * html/HTMLSelectElement.cpp:
3086         (WebCore::HTMLSelectElement::selectOption): Changed to call notifyFormStateChanged() instead of formStateDidChange().
3087         (WebCore::HTMLSelectElement::updateListBoxSelection): Changed to call notifyFormStateChanged().
3088         * html/HTMLTextFormControlElement.cpp: Moved an implemented of notifyFormStateChanged() to HTMLFormControlElementWithState class.
3089         * html/HTMLTextFormControlElement.h: Moved a declaration of notifyFormStateChanged() to HTMLFormControlElementWithState class.
3090
3091 2012-09-20  Keishi Hattori  <keishi@webkit.org>
3092
3093         REGRESSION(r127727): Can't navigate between months with arrow keys in calendar picker
3094         https://bugs.webkit.org/show_bug.cgi?id=97166
3095
3096         Reviewed by Kent Tamura.
3097
3098         Fixing bug in r127727 so arrow keys work properly.
3099
3100         Test: fast/forms/date/calendar-picker-key-operations.html
3101
3102         * Resources/pagepopups/calendarPicker.js:
3103         (DaysTable.prototype._maybeSetPreviousMonth):
3104         (DaysTable.prototype._maybeSetNextMonth):
3105
3106 2012-09-18  Alexander Pavlov  <apavlov@chromium.org>
3107
3108         Web Inspector: Use and process the actual ScriptId in the protocol EventListener object
3109         https://bugs.webkit.org/show_bug.cgi?id=93271
3110
3111         Reviewed by Yury Semikhatsky.
3112
3113         - Use the actual script identifier in the "location" object's "scriptId" field
3114           for the DOM.EventListener protocol type, but send the script URL in the new "sourceName" field.
3115         - Use 0-based lines in the "location" object's "lineNumber" field for linkifyRawLocation() to work correctly.
3116         - Fixed formatting of links to listener locations to contain "(program)" rather than empty string.
3117
3118         * bindings/js/ScriptEventListener.cpp:
3119         (WebCore::eventListenerHandlerLocation):
3120         * bindings/js/ScriptEventListener.h:
3121         (WebCore):
3122         * bindings/v8/ScriptEventListener.cpp:
3123         (WebCore::eventListenerHandlerLocation):
3124         * bindings/v8/ScriptEventListener.h:
3125         (WebCore):
3126         * inspector/Inspector.json:
3127         * inspector/InspectorDOMAgent.cpp:
3128         (WebCore::InspectorDOMAgent::buildObjectForEventListener):
3129         * inspector/front-end/BreakpointsSidebarPane.js:
3130         * inspector/front-end/EventListenersSidebarPane.js:
3131         * inspector/front-end/Linkifier.js:
3132         (WebInspector.Linkifier.DefaultFormatter.prototype.formatLiveAnchor):
3133         * inspector/front-end/ResourceUtils.js:
3134         (WebInspector.formatLinkText): Use "(program)" if URL is empty.
3135
3136 2012-09-19  Dan Bernstein  <mitz@apple.com>
3137
3138         WebCore part of adding a setting and API for disabling screen font substitution
3139         https://bugs.webkit.org/show_bug.cgi?id=97168
3140
3141         Reviewed by Tim Horton.
3142
3143         * WebCore.exp.in: Added an entry for Settings::setScreenFontSubstitutionEnabled.
3144         * css/StyleResolver.cpp:
3145         (WebCore::StyleResolver::collectMatchingRulesForList): Changed to use printer fonts if
3146         screen font substitution is not enabled.
3147         * page/Settings.cpp:
3148         (WebCore::Settings::Settings): Added initializer for new m_screenFontSubstitutionEnabled
3149         member variable. The initial value is true, matching existing behavior.
3150         (WebCore::Settings::setScreenFontSubstitutionEnabled): Added this setter, which updated the
3151         member variable and forces a style recalc in all pages using this Settings.
3152         * page/Settings.h:
3153         (Settings): Added m_screenFontSubstitutionEnabled boolean member variable.
3154         (WebCore::Settings::screenFontSubstitutionEnabled): Added this getter.
3155
3156 2012-09-19  Yoshifumi Inoue  <yosin@chromium.org>
3157
3158         [Forms] multiple fields time input UI should call notifyFormStateChanged() when value of field is changed
3159         https://bugs.webkit.org/show_bug.cgi?id=97169
3160
3161         Reviewed by Kent Tamura.
3162
3163         This patch makes multiple fields time input UI calls notifyFormStateChanged()
3164         when field value is changed as other input types do.
3165
3166         This patch affects ports which enable both ENABLE_INPUT_TYPE_TIME and
3167         ENABLE_INPUT_TYPE_TIME_MULTIPLE_FIELDS.
3168
3169         No new tests. We can't test this change in WebKit test tools. Test script
3170         will be implemented in Chromium side.
3171
3172         * html/TimeInputType.cpp:
3173         (WebCore::TimeInputType::DateTimeEditControlOwnerImpl::editControlValueChanged):
3174
3175 2012-09-19  Keishi Hattori  <keishi@webkit.org>
3176
3177         REGRESSION(r127727): Calendar picker focus ring should be hidden until key event
3178         https://bugs.webkit.org/show_bug.cgi?id=97165
3179
3180         Reviewed by Kent Tamura.
3181
3182         The regression was caused because NoFocusRing class was being removed
3183         from then main element inside resetMain().
3184
3185         No new tests. Covered by calendar-picker-appearance.html.
3186
3187         * Resources/pagepopups/calendarPicker.js:
3188         (initialize):
3189         (CalendarPicker):
3190         (CalendarPicker.prototype._layout):
3191         (DaysTable.prototype._handleKey):
3192         (CalendarPicker.prototype._handleBodyKeyDown):
3193         (CalendarPicker.prototype.maybeUpdateFocusStyle): Make this a method of CalendarPicker because we don't use it for other pickers.
3194
3195 2012-09-19  David Grogan  <dgrogan@chromium.org>
3196
3197         IndexedDB: Print console warning about setVersion
3198         https://bugs.webkit.org/show_bug.cgi?id=96575
3199
3200         Reviewed by Tony Chang.
3201
3202         setVersion has been out of the spec for almost a year but there are
3203         still a lot of users.
3204
3205         We show the warning once per database object as an approximation for
3206         once per page.
3207
3208         No new tests, but 150-something rebaselines.
3209
3210         * Modules/indexeddb/IDBDatabase.cpp:
3211         (WebCore::IDBDatabase::IDBDatabase):
3212         (WebCore::IDBDatabase::setVersion):
3213         * Modules/indexeddb/IDBDatabase.h:
3214         (IDBDatabase):
3215
3216 2012-09-19  Sean Wang  <Xuewen.Wang@torchmobile.com.cn>
3217
3218         [BlackBerry] Basic authentication challenge credentials for stored credentials again after restarting browser
3219         https://bugs.webkit.org/show_bug.cgi?id=96362
3220
3221         Reviewed by Rob Buis.
3222
3223         This patch enable reading credentials from the persistent credential storage
3224         when it is not private browsing mode and there is not a credential in the RAM
3225         for the requested resource.
3226
3227         Since we don't load persistent stored credentials into RAM at the starting time,
3228         even we have saved the credentials at the last browsing, after restarting the browser,
3229         it will still challenge for credentials for the requesting resources.
3230
3231         No new tests, it uses the original authentication tests. There is no way to
3232         clear all credentials or restarting browsers to test this feature.
3233
3234         * platform/network/blackberry/CredentialBackingStore.cpp:
3235         (WebCore::CredentialBackingStore::getProtectionSpace):
3236         (WebCore):
3237         * platform/network/blackberry/CredentialBackingStore.h:
3238         (CredentialBackingStore):
3239         * platform/network/blackberry/NetworkManager.cpp:
3240         (WebCore::NetworkManager::startJob):
3241
3242 2012-09-19  Shinya Kawanaka  <shinyak@chromium.org>
3243
3244         [Refactoring] ButtonInputType of <input> element should have innerElement to make <input> AuthorShadowDOM-ready
3245         https://bugs.webkit.org/show_bug.cgi?id=95939
3246
3247         Reviewed by Dimitri Glazkov.
3248
3249         We had 2 ways to show text in RenderButton. One is to use RenderButton::setText() and the other is to add renderer
3250         as a child of RenderButton. <input type="button"> used the former one, and <button> used the latter one.
3251         The former makes RenderButton a bit messy, and also prevents from making <input> AuthorShadowDOM ready.
3252
3253         So we remove RenderButton::setText() and refactor <input type="button"> to use a Shadow DOM to show text.
3254
3255         Since the text in <input type="button"> should create RenderTextFragment, we introduce TextForButtonInputType
3256         class. RenderText will allow us to select the inner text of <input>, but it should not.
3257
3258         No new tests, existing tests should cover the change.
3259
3260         * html/BaseButtonInputType.cpp:
3261         (TextForButtonInputType): Special Text node which creates RenderTextFragment.
3262         (WebCore):
3263         (WebCore::TextForButtonInputType::create):
3264         (WebCore::TextForButtonInputType::TextForButtonInputType):
3265         (WebCore::TextForButtonInputType::createRenderer):
3266         (WebCore::BaseButtonInputType::BaseButtonInputType):
3267         (WebCore::BaseButtonInputType::createShadowSubtree):
3268         (WebCore::BaseButtonInputType::destroyShadowSubtree):
3269         (WebCore::BaseButtonInputType::valueAttributeChanged): When a button value is changed, we reflect the value to
3270         the text node in ShadowDOM.
3271         * html/BaseButtonInputType.h:
3272         (BaseButtonInputType):
3273         * html/FileInputType.cpp:
3274         (WebCore::UploadButtonElement::create): Creates Shadow DOM subtree.
3275         (WebCore::UploadButtonElement::createForMultiple): Creates Shadow DOM subtree.
3276         * html/HTMLInputElement.cpp:
3277         (WebCore::HTMLInputElement::parseAttribute):
3278         * html/InputType.cpp:
3279         (WebCore::InputType::valueAttributeChanged):
3280         (WebCore):
3281         * html/InputType.h:
3282         (InputType):
3283         * rendering/RenderButton.cpp:
3284         (WebCore::RenderButton::RenderButton):
3285         (WebCore::RenderButton::styleDidChange): Removed unnecessary text related code.
3286         * rendering/RenderButton.h:
3287         (RenderButton):
3288
3289 2012-09-19  Jason Liu  <jason.liu@torchmobile.com.cn>
3290
3291         [BlackBerry] Get infinite amount of requests after attempting re-authentication basic.
3292         https://bugs.webkit.org/show_bug.cgi?id=96994
3293
3294         Reviewed by Rob Buis.
3295
3296         We should remove the wrong credentials before calling sendRequestWithCredentials again.
3297         #PR 200226
3298         Reviewed internally by Jonathan Dong.
3299
3300         No new tests. It is covered by ManualTests/blackberry/http-auth-challenge.html.
3301
3302
3303         * platform/network/blackberry/NetworkJob.cpp:
3304         (WebCore::NetworkJob::notifyAuthReceived):
3305
3306 2012-09-19  Adam Barth  <abarth@webkit.org>
3307
3308         Remove Blob.webkitSlice
3309         https://bugs.webkit.org/show_bug.cgi?id=96715
3310
3311         Reviewed by Darin Fisher.
3312
3313         Based on these usage metrics, it appears that it is safe to remove
3314         Blob.webkitSlice. Folks that were previously calling webkitSlice should
3315         just call slice instead. They do the same thing.
3316
3317         Ratio of Blob.webkitSlice calls to Blob.slice: 14.87%
3318         Ratio of Blob.webkitSlice calls to Document creation: 0.0053%
3319
3320         * fileapi/Blob.cpp:
3321         (WebCore::Blob::slice):
3322         * fileapi/Blob.h:
3323         (Blob):
3324         * fileapi/Blob.idl:
3325         * fileapi/File.h:
3326         (File):
3327         * inspector/front-end/FileUtils.js:
3328         (WebInspector.ChunkedFileReader.prototype._loadChunk):
3329
3330 2012-09-19  Julien Chaffraix  <jchaffraix@webkit.org>
3331
3332         The collapsing border code needs direction-aware border getters
3333         https://bugs.webkit.org/show_bug.cgi?id=96710
3334
3335         Reviewed by Ojan Vafai.
3336
3337         This refactoring is needed to extend our collapsing border support for mixed directionality
3338         at the table cell level (we currently wrongly ignore any direction below the row-group). For
3339         now, the new helpers are dumb and return exactly the old result but they will be made
3340         direction-aware in a follow-up change.
3341
3342         Refactoring covered by existing tests.
3343
3344         * rendering/RenderBox.h:
3345         (WebCore::RenderBox::hasSameDirectionAs):
3346         Added this helper function. For now, it's only used to compare against
3347         the table direction but we will reuse it to compare the current cell
3348         direction against the other table parts.
3349
3350         * rendering/RenderTableCell.h:
3351         (WebCore::RenderTableCell::computeCollapsedStartBorder):
3352         (WebCore::RenderTableCell::computeCollapsedEndBorder):
3353         Transitioned those 2 functions to using the new direction-aware functions.
3354
3355         * rendering/RenderTable.cpp:
3356         (WebCore::RenderTable::tableStartBorderAdjoiningCell):
3357         (WebCore::RenderTable::tableEndBorderAdjoiningCell):
3358         * rendering/RenderTableSection.cpp:
3359         (WebCore::RenderTableSection::firstRowCellAdjoiningTableStart):
3360         (WebCore::RenderTableSection::firstRowCellAdjoiningTableEnd):
3361         * rendering/RenderTableCell.h:
3362         (WebCore::RenderTableCell::borderAdjoiningTableStart):
3363         (WebCore::RenderTableCell::borderAdjoiningTableEnd):
3364         * rendering/RenderTableSection.h:
3365         (WebCore::RenderTableSection::borderAdjoiningTableStart):
3366         (WebCore::RenderTableSection::borderAdjoiningTableEnd):
3367         Updated those call sites to use RenderBox::hasSameDirectionAs.
3368
3369         * rendering/RenderTableCell.h:
3370         (WebCore::RenderTableCell::borderAdjoiningNextCell):
3371         (WebCore::RenderTableCell::borderAdjoiningPreviousCell):
3372         * rendering/RenderTableCol.cpp:
3373         (WebCore::RenderTableCol::borderAdjoiningCellStartBorder):
3374         (WebCore::RenderTableCol::borderAdjoiningCellEndBorder):
3375         (WebCore::RenderTableCol::borderAdjoiningCellBefore):
3376         (WebCore::RenderTableCol::borderAdjoiningCellAfter):
3377         * rendering/RenderTableCol.h:
3378         * rendering/RenderTableRow.cpp:
3379         (WebCore::RenderTableRow::borderAdjoiningStartCell):
3380         (WebCore::RenderTableRow::borderAdjoiningEndCell):
3381         * rendering/RenderTableRow.h:
3382         (WebCore::RenderTableRow::borderAdjoiningTableStart):
3383         (WebCore::RenderTableRow::borderAdjoiningTableEnd):
3384         * rendering/RenderTableSection.cpp:
3385         (WebCore::RenderTableSection::borderAdjoiningStartCell):
3386         (WebCore::RenderTableSection::borderAdjoiningEndCell):
3387         New direction-aware functions. Added some ASSERT to ensure
3388         we don't call them with the wrong parameters.
3389
3390 2012-09-19  Kentaro Hara  <haraken@chromium.org>
3391
3392         [V8] ScriptController::compileAndRunScript() can crash
3393         https://bugs.webkit.org/show_bug.cgi?id=96567
3394
3395         Reviewed by Adam Barth.
3396
3397         See chromium bug: http://code.google.com/p/chromium/issues/detail?id=146776
3398
3399         The root cause is that v8::PreCompile() can return 0 when the stack of
3400         V8's parser overflows (c.f. http://code.google.com/codesearch#OAMlx_jo-ck/src/v8/src/parser.cc&exact_package=chromium&q=kPreParseStackOverflow&type=cs&l=6021).
3401
3402         This patch adds the 0 check to the caller side. Given that precompileScript()
3403         is just trying to speculatively precompile a script, it's OK to give up
3404         precompiling for such edge cases.
3405
3406         Manually tested with the html generated by the following shell script:
3407
3408           echo '<script language="JavaScript" type="text/javascript" src="asan-crash.js"></script>' > asan-crash.html
3409           echo 'if(wURLF.search("")>=0) {}' > asan-crash.js
3410           for i in `seq 14830`
3411           do
3412             echo 'else if(wURLF.search("")>=0) {}' >> asan-crash.js
3413           done
3414
3415         I didn't add the test because '14380' depends on an environment
3416         and because we don't want to add a huge html test.
3417
3418         * bindings/v8/ScriptSourceCode.cpp:
3419         (WebCore::ScriptSourceCode::precompileScript):
3420
3421 2012-09-19  Joshua Bell  <jsbell@chromium.org>
3422
3423         IndexedDB: Pending call cleanup
3424         https://bugs.webkit.org/show_bug.cgi?id=96952
3425
3426         Reviewed by Tony Chang.