Source/WebCore/ChangeLog

   1 2014-04-25  Javier Fernandez  <jfernandez@igalia.com>
   2
   3         [CSS Grid Layout] Implementation of the grid-template shorthand.
   4         https://bugs.webkit.org/show_bug.cgi?id=128980
   5
   6         Reviewed by Darin Adler.
   7
   8         This shorthand sets the values for the grid-template-columns,
   9         grid-template-rows and grid-template-areas, so the implementation
  10         tries to reuse as much available parsing functions as possible.
  11
  12         The "parsingGridTrackList" was refactored to return a CSSValue and
  13         let the "parseValue" function to assign the property value. The
  14         "forwardSlash" operator is now valid when the track-list clause is
  15         part of a shorthand. The "parseValue" function checkouts that only
  16         additional clauses are allowed when processing shorthands; the
  17         grid-columns-rows-get-set.html tests was modified to verify this.
  18
  19         The "parseGridTemplateAreas" was refactored too, in order to
  20         process single areas's rows. This is very useful for the
  21         gris-template secondary syntax, which mixes areas and rows values.
  22
  23         Finally, the "parseGirdLineNames" function was modified as well by
  24         defining an new argument to concatenate head/tail custom-ident
  25         elements and ensure the identList is at the heading index, since
  26         it's now possible the parseList was rewound.
  27
  28         The implementation of the grid-template shorthand tries first to
  29         match the <grid-template-columns> / <grid-template-rows> syntax,
  30         failing back to the secondary syntax if needed.  This approach
  31         requires to rewind the parseList but it produces a clearer code.
  32
  33         Test: fast/css-grid-layout/grid-template-shorthand-get-set.html
  34
  35         * css/CSSComputedStyleDeclaration.cpp:
  36         (WebCore::ComputedStyleExtractor::propertyValue):
  37         * css/CSSParser.cpp:
  38         (WebCore::CSSParser::parseValue):
  39         (WebCore::CSSParser::parseGridTemplateRowsAndAreas):
  40         (WebCore::CSSParser::parseGridTemplateShorthand):
  41         (WebCore::CSSParser::parseGridLineNames):
  42         (WebCore::CSSParser::parseGridTrackList):
  43         (WebCore::CSSParser::parseGridTemplateAreasRow):
  44         (WebCore::CSSParser::parseGridTemplateAreas):
  45         * css/CSSParser.h:
  46         * css/CSSParserValues.h:
  47         (WebCore::CSSParserValueList::setCurrentIndex):
  48         * css/CSSPropertyNames.in:
  49         * css/StylePropertyShorthand.cpp:
  50         (WebCore::webkitGridTemplateShorthand):
  51         * css/StylePropertyShorthand.h:
  52
  53 2014-04-25  Andreas Kling  <akling@apple.com>
  54
  55         Remove two unused SVGDocument functions.
  56         <https://webkit.org/b/132178>
  57
  58         Reviewed by Antti Koivisto.
  59
  60         * svg/SVGDocument.cpp:
  61         (WebCore::SVGDocument::dispatchZoomEvent): Deleted.
  62         (WebCore::SVGDocument::dispatchScrollEvent): Deleted.
  63         * svg/SVGDocument.h:
  64
  65 2014-04-25  Ion Rosca  <rosca@adobe.com>
  66
  67         Incomplete body painting when using blend modes
  68         https://bugs.webkit.org/show_bug.cgi?id=131889
  69
  70         The incomplete painting was caused by the transparency layer created for
  71         the root renderer. We can safely skip creating this transparency layer at
  72         the root level, as there is nothing else being painted behind this layer that
  73         could be used erroneously as a backdrop.
  74
  75         Reviewed by Simon Fraser.
  76
  77         Test: css3/compositing/blend-mode-with-body.html
  78
  79         * rendering/RenderLayer.h:
  80         Changing RenderLayer::paintsWithTransparency so that it will not
  81         return true when the root renderer needs to isolate blending.
  82
  83 2014-04-25  Darin Adler  <darin@apple.com>
  84
  85         ASSERTION FAILED: "!m_isolatedWorld->isNormal() || m_wrapper || !m_jsFunction" in svg/custom/use-instanceRoot-event-listeners.xhtml
  86         https://bugs.webkit.org/show_bug.cgi?id=132148
  87
  88         Reviewed by Andreas Kling.
  89
  90         Changed how JSCustomMarkFunction generation works. Instead of leaving out
  91         the generated visitChildren function, just generate a call to visitAdditionalChildren.
  92         This eliminates the need to repeat boilerplate.
  93
  94         The fix for the above bug was to correct mistaken logic where JSSVGElementInstance
  95         had a visitChildren that did not properly mark event listeners because it explicitly
  96         did not call through to the base class visitChildren. The new arrangement makes that
  97         mistake impossible.
  98
  99         * bindings/js/JSAttrCustom.cpp:
 100         (WebCore::JSAttr::visitAdditionalChildren): Use this instead of visitChildren.
 101         * bindings/js/JSAudioTrackCustom.cpp:
 102         (WebCore::JSAudioTrack::visitAdditionalChildren): Ditto.
 103         * bindings/js/JSAudioTrackListCustom.cpp:
 104         (WebCore::JSAudioTrackList::visitAdditionalChildren): Ditto.
 105         * bindings/js/JSCSSRuleCustom.cpp:
 106         (WebCore::JSCSSRule::visitAdditionalChildren): Ditto.
 107         * bindings/js/JSCSSStyleDeclarationCustom.cpp:
 108         (WebCore::JSCSSStyleDeclaration::visitAdditionalChildren): Ditto.
 109         * bindings/js/JSCanvasRenderingContextCustom.cpp:
 110         (WebCore::JSCanvasRenderingContext::visitAdditionalChildren): Ditto.
 111         * bindings/js/JSCryptoKeyPairCustom.cpp:
 112         (WebCore::JSCryptoKeyPair::visitAdditionalChildren): Ditto.
 113         * bindings/js/JSDOMWindowCustom.cpp:
 114         (WebCore::JSDOMWindow::visitAdditionalChildren): Ditto.
 115         * bindings/js/JSMessageChannelCustom.cpp:
 116         (WebCore::JSMessageChannel::visitAdditionalChildren): Ditto.
 117         * bindings/js/JSMessagePortCustom.cpp:
 118         (WebCore::JSMessagePort::visitAdditionalChildren): Ditto.
 119         * bindings/js/JSNodeCustom.cpp:
 120         (WebCore::JSNode::visitAdditionalChildren): Ditto.
 121         * bindings/js/JSNodeFilterCustom.cpp:
 122         (WebCore::JSNodeFilter::visitAdditionalChildren): Ditto.
 123         * bindings/js/JSNodeIteratorCustom.cpp:
 124         (WebCore::JSNodeIterator::visitAdditionalChildren): Ditto.
 125         * bindings/js/JSSVGElementInstanceCustom.cpp:
 126         (WebCore::JSSVGElementInstance::visitAdditionalChildren): Ditto.
 127         * bindings/js/JSSharedWorkerCustom.cpp:
 128         (WebCore::JSSharedWorker::visitAdditionalChildren): Ditto.
 129         * bindings/js/JSStyleSheetCustom.cpp:
 130         (WebCore::JSStyleSheet::visitAdditionalChildren): Ditto.
 131         * bindings/js/JSTextTrackCueCustom.cpp:
 132         (WebCore::JSTextTrackCue::visitAdditionalChildren): Ditto.
 133         * bindings/js/JSTextTrackCustom.cpp:
 134         (WebCore::JSTextTrack::visitAdditionalChildren): Ditto.
 135         * bindings/js/JSTextTrackListCustom.cpp:
 136         (WebCore::JSTextTrackList::visitAdditionalChildren): Ditto.
 137         * bindings/js/JSTreeWalkerCustom.cpp:
 138         (WebCore::JSTreeWalker::visitAdditionalChildren): Ditto.
 139         * bindings/js/JSVideoTrackCustom.cpp:
 140         (WebCore::JSVideoTrack::visitAdditionalChildren): Ditto.
 141         * bindings/js/JSVideoTrackListCustom.cpp:
 142         (WebCore::JSVideoTrackList::visitAdditionalChildren): Ditto.
 143         * bindings/js/JSWebGLRenderingContextCustom.cpp:
 144         (WebCore::JSWebGLRenderingContext::visitAdditionalChildren): Ditto.
 145         * bindings/js/JSWorkerGlobalScopeCustom.cpp:
 146         (WebCore::JSWorkerGlobalScope::visitAdditionalChildren): Ditto.
 147         * bindings/js/JSXMLHttpRequestCustom.cpp:
 148         (WebCore::JSXMLHttpRequest::visitAdditionalChildren): Ditto.
 149         * bindings/js/JSXPathResultCustom.cpp:
 150         (WebCore::JSXPathResult::visitAdditionalChildren): Ditto.
 151
 152         * bindings/js/JSDOMGlobalObject.cpp:
 153         (WebCore::JSDOMGlobalObject::visitChildren): Rewrote to use modern for loops.
 154
 155         * bindings/scripts/CodeGeneratorJS.pm:
 156         (GenerateHeader): Generate declaration of visitAdditionalChildren.
 157         (GenerateImplementation): Generate call to visitAdditionalChildren.
 158
 159 2014-04-24  Andreas Kling  <akling@apple.com>
 160
 161         [iOS WebKit2] Enable optimization to mmap downloaded resources once they become file-backed.
 162         <https://webkit.org/b/132171>
 163         <rdar://problem/16720733>
 164
 165         Add a missing export for the USE(CFNETWORK) + WebKit2 combo.
 166
 167         Reviewed by Antti Koivisto.
 168
 169         * WebCore.exp.in:
 170
 171 2014-04-24  Darin Adler  <darin@apple.com>
 172
 173         FrameLoader::checkCompleted can hit the "ref'ing while destroyed" assertion
 174         https://bugs.webkit.org/show_bug.cgi?id=132163
 175         rdar://problem/16720640
 176
 177         Reviewed by Brady Eidson.
 178
 179         Couldn't find a way to test this yet. Would be nice to have a test.
 180
 181         * loader/FrameLoader.cpp:
 182         (WebCore::FrameLoader::checkCompleted): Move protector until after we check
 183         if the frame is already complete. That can happen in practice when this is
 184         called from within the frame's destructor. All the code that runs before the
 185         protector simply checks state and does not require protection.
 186
 187 2014-04-24  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
 188
 189         Mark Supplement instead of RefCountedSupplement in NavigatorContentUtils
 190         https://bugs.webkit.org/show_bug.cgi?id=132151
 191
 192         Reviewed by Darin Adler.
 193
 194         Though Original goal was to make it sharable across navigator instances, the NavigatorContentUtils
 195         has used RefCountedSupplement<Page> instead of RefCountedSupplement<Navigator>. This patch makes it
 196         use Supplement<Page> because there is no scenario which needs to be shared across navigator instances.
 197
 198         Blink merge from https://src.chromium.org/viewvc/blink?view=rev&revision=171403.
 199
 200         No new tests, no behavior changes.
 201
 202         * Modules/navigatorcontentutils/NavigatorContentUtils.cpp:
 203         (WebCore::NavigatorContentUtils::from):
 204         (WebCore::NavigatorContentUtils::create):
 205         (WebCore::provideNavigatorContentUtilsTo):
 206         * Modules/navigatorcontentutils/NavigatorContentUtils.h:
 207
 208 2014-04-24  Commit Queue  <commit-queue@webkit.org>
 209
 210         Unreviewed, rolling out r167700.
 211         https://bugs.webkit.org/show_bug.cgi?id=132142
 212
 213         Incorrectly reverted the change in r167547 for
 214         webkit.org/b/131898 (Requested by rniwa on #webkit).
 215
 216         Reverted changeset:
 217
 218         "Cursor doesn't change back to pointer when leaving the Safari
 219         window"
 220         https://bugs.webkit.org/show_bug.cgi?id=132038
 221         http://trac.webkit.org/changeset/167700
 222
 223 2014-04-24  Brady Eidson  <beidson@apple.com>
 224
 225         Rename "IMAGE_CONTROLS" feature to "SERVICE_CONTROLS"
 226         https://bugs.webkit.org/show_bug.cgi?id=132155
 227
 228         Reviewed by Tim Horton.
 229
 230         No new tests (No change in behavior).
 231
 232         * Configurations/FeatureDefines.xcconfig:
 233         * DerivedSources.make:
 234         * WebCore.exp.in:
 235         * css/CSSDefaultStyleSheets.cpp:
 236         (WebCore::CSSDefaultStyleSheets::ensureDefaultStyleSheetsForElement):
 237         * css/CSSPrimitiveValueMappings.h:
 238         (WebCore::CSSPrimitiveValue::CSSPrimitiveValue):
 239         * css/CSSValueKeywords.in:
 240         * dom/Node.h:
 241         * html/HTMLImageElement.cpp:
 242         (WebCore::HTMLImageElement::HTMLImageElement):
 243         (WebCore::HTMLImageElement::parseAttribute):
 244         (WebCore::HTMLImageElement::didAttachRenderers):
 245         * html/HTMLImageElement.h:
 246         * html/shadow/ImageControlsRootElement.cpp:
 247         * html/shadow/ImageControlsRootElement.h:
 248         * html/shadow/mac/ImageControlsButtonElementMac.cpp:
 249         * html/shadow/mac/ImageControlsButtonElementMac.h:
 250         * html/shadow/mac/ImageControlsRootElementMac.cpp:
 251         * html/shadow/mac/ImageControlsRootElementMac.h:
 252         * page/ContextMenuContext.cpp:
 253         (WebCore::ContextMenuContext::ContextMenuContext):
 254         * page/ContextMenuContext.h:
 255         * page/ContextMenuController.cpp:
 256         (WebCore::ContextMenuController::maybeCreateContextMenu):
 257         (WebCore::ContextMenuController::populate):
 258         * page/ContextMenuController.h:
 259         * page/Settings.in:
 260         * platform/ThemeTypes.h:
 261         * rendering/RenderImage.cpp:
 262         (WebCore::RenderImage::canHaveChildren):
 263         * rendering/RenderTheme.cpp:
 264         (WebCore::RenderTheme::adjustStyle):
 265         (WebCore::RenderTheme::paint):
 266         (WebCore::RenderTheme::paintBorderOnly):
 267         (WebCore::RenderTheme::paintDecorations):
 268         * rendering/RenderTheme.h:
 269         * rendering/RenderThemeMac.h:
 270         * rendering/RenderThemeMac.mm:
 271         (WebCore::RenderThemeMac::servicesRolloverButtonCell):
 272         (WebCore::RenderThemeMac::paintImageControlsButton):
 273         (WebCore::RenderThemeMac::imageControlsButtonSize):
 274
 275 2014-04-24  Timothy Hatcher  <timothy@apple.com>
 276
 277         Web Inspector: Restore PageDebuggerAgent::enable / disable
 278         https://bugs.webkit.org/show_bug.cgi?id=132156
 279
 280         Restore functions that were eroniously removed in r167530.
 281
 282         Reviewed by Joseph Pecoraro.
 283
 284         * inspector/PageDebuggerAgent.cpp:
 285         (WebCore::PageDebuggerAgent::enable): Added.
 286         (WebCore::PageDebuggerAgent::disable): Added.
 287         * inspector/PageDebuggerAgent.h:
 288
 289 2014-04-24  Alexey Proskuryakov  <ap@apple.com>
 290
 291         Dropzone effects don't work in non-file documents
 292         https://bugs.webkit.org/show_bug.cgi?id=131770
 293
 294         Reviewed by Darin Adler.
 295
 296         File documents have two quirks that were making dropzone work in these before:
 297         1. An ancient hack for Dashboard allows pasteboard access from JS.
 298         2. On Mac, sandbox doesn't prevent File object creation, as we already have the access.
 299
 300         * dom/DataTransfer.cpp:
 301         (WebCore::DataTransfer::hasFileOfType):
 302         (WebCore::DataTransfer::hasStringOfType):
 303         * dom/DataTransfer.h:
 304         Moved these functions from EventHandler to DataTransfer. We can't create a DataTransfer
 305         with Files while dragging, security doesn't permit us to. But we can get the file name.
 306
 307         * fileapi/File.cpp:
 308         (WebCore::createBlobDataForFile):
 309         (WebCore::createBlobDataForFileWithName):
 310         (WebCore::File::contentTypeFromFilePath):
 311         (WebCore::getContentTypeFromFileName): Deleted.
 312         * fileapi/File.h:
 313         Exposed a function to get file type from path without creating a File first.
 314         This is much cheaper than creating a File, and works even when sandbox disallows
 315         read access to content, such as when dragging over a target.
 316
 317         * page/EventHandler.cpp:
 318         (WebCore::hasDropZoneType):
 319         (WebCore::hasFileOfType): Deleted.
 320         (WebCore::hasStringOfType): Deleted.
 321
 322 2014-04-24  Commit Queue  <commit-queue@webkit.org>
 323
 324         Unreviewed, rolling out r167441.
 325         https://bugs.webkit.org/show_bug.cgi?id=132152
 326
 327         Caused full screen regressions on vimeo, youtube, and others.
 328         (Requested by jernoble on #webkit).
 329
 330         Reverted changeset:
 331
 332         "Fullscreen media controls are unusable in pagination mode"
 333         https://bugs.webkit.org/show_bug.cgi?id=131705
 334         http://trac.webkit.org/changeset/167441
 335
 336 2014-04-24  Adenilson Cavalcanti  <cavalcantii@gmail.com>
 337
 338         Unused class forward declarations in Page
 339         https://bugs.webkit.org/show_bug.cgi?id=132141
 340
 341         Reviewed by Benjamin Poulain.
 342
 343         No new tests, no change on behavior.
 344
 345         * page/Page.h:
 346
 347 2014-04-24  Eric Carlson  <eric.carlson@apple.com>
 348
 349         [Mac] don't ask for AVAssetTrack properties before they are available
 350         https://bugs.webkit.org/show_bug.cgi?id=131902
 351         <rdar://problem/16505076>
 352
 353         Reviewed by Brent Fulgham.
 354
 355         No new tests, the behavior this changes can not be tested with a layout test.
 356
 357         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
 358         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
 359         (WebCore::MediaPlayerPrivateAVFoundationObjC::MediaPlayerPrivateAVFoundationObjC): Initialize
 360             m_cachedTotalBytes.
 361         (WebCore::MediaPlayerPrivateAVFoundationObjC::beginLoadingMetadata): Don't report that
 362             metadata has been loaded until the track properties we need have been loaded too.
 363         (WebCore::MediaPlayerPrivateAVFoundationObjC::totalBytes): Cache totalBytes instead
 364             of recalculating it every time.
 365         (WebCore::MediaPlayerPrivateAVFoundationObjC::tracksDidChange): Invalidate cached
 366             total bytes.
 367         (WebCore::assetTrackMetadataKeyNames): Array of AVAssetTrack properties we use.
 368
 369 2014-04-24  Myles C. Maxfield  <mmaxfield@apple.com>
 370
 371         Unify platformWidthForGlyph across OS X and iOS
 372         https://bugs.webkit.org/show_bug.cgi?id=132036
 373
 374         Reviewed by Darin Adler.
 375
 376         This patch creates on shared SimpleFontData::platformWidthForGlyph() function for both OS X and iOS.
 377
 378         No new tests are necessary because there should be no behavior changes.
 379
 380         * platform/graphics/SimpleFontData.h: Signatures for two helper functions
 381         * platform/graphics/ios/SimpleFontDataIOS.mm: Replace iOS implementation of platformWidthForGlyph() with
 382         implementations of only the two helper functions
 383         (WebCore::SimpleFontData::getRenderingStyle): Compute style argument to CGFontGetGlyphAdvancesForStyle()
 384         (WebCore::SimpleFontData::advanceForColorBitmapFont): iOS doesn't have color bitmap fonts
 385         (WebCore::SimpleFontData::platformWidthForGlyph): Deleted.
 386         * platform/graphics/mac/SimpleFontDataMac.mm:
 387         (WebCore::SimpleFontData::getRenderingStyle): Compute style argument to CGFontGetGlyphAdvancesForStyle()
 388         (WebCore::SimpleFontData::advanceForColorBitmapFont): Use [NSFont advancementForGlyph] to compute the advance
 389         (WebCore::hasCustomTracking): Removed #if
 390         (WebCore::isEmoji): Only relevant on iOS
 391         (WebCore::SimpleFontData::platformWidthForGlyph): Shared implementation. Calls helper functions.
 392
 393 2014-04-24  Zalan Bujtas  <zalan@apple.com>
 394
 395         Subpixel rendering: Clipping on text areas when shifted by one device pixel.
 396         https://bugs.webkit.org/show_bug.cgi?id=132008
 397
 398         Reviewed by Darin Adler.
 399
 400         Make RenderTheme paint* functions LayoutRect aware. Textarea is device pixel snapped, while
 401         other theme controls are still on integral size/positions.
 402
 403         Test: fast/forms/hidpi-textarea-on-subpixel-position.html
 404
 405         * rendering/RenderBox.cpp:
 406         (WebCore::RenderBox::paintBoxDecorations):
 407         * rendering/RenderTheme.cpp:
 408         (WebCore::RenderTheme::paint):
 409         (WebCore::RenderTheme::paintBorderOnly):
 410         (WebCore::RenderTheme::paintDecorations):
 411         * rendering/RenderTheme.h:
 412         (WebCore::RenderTheme::paintTextField):
 413         (WebCore::RenderTheme::paintTextFieldDecorations):
 414         (WebCore::RenderTheme::paintTextArea):
 415         (WebCore::RenderTheme::paintTextAreaDecorations):
 416         * rendering/RenderThemeIOS.h:
 417         * rendering/RenderThemeIOS.mm:
 418         (WebCore::RenderThemeIOS::paintTextFieldDecorations):
 419         (WebCore::RenderThemeIOS::paintTextAreaDecorations):
 420         * rendering/RenderThemeMac.h:
 421         * rendering/RenderThemeMac.mm:
 422         (WebCore::RenderThemeMac::paintTextField):
 423         (WebCore::RenderThemeMac::paintTextArea):
 424
 425 2014-04-24  Myles C. Maxfield  <mmaxfield@apple.com>
 426
 427         FontCache::fontCache() never returns nullptr so it can be made to return a reference instead
 428         https://bugs.webkit.org/show_bug.cgi?id=132110
 429
 430         Reviewed by Tim Horton.
 431
 432         Updates callers to use '.' instead of '->'.
 433
 434         No new tests are necessary because there should be no behavior change.
 435
 436         * css/CSSFontFaceSource.cpp:
 437         (WebCore::CSSFontFaceSource::getFontData):
 438         * css/CSSFontSelector.cpp:
 439         (WebCore::CSSFontSelector::CSSFontSelector):
 440         (WebCore::CSSFontSelector::~CSSFontSelector):
 441         (WebCore::CSSFontSelector::addFontFaceRule):
 442         (WebCore::fontDataForGenericFamily):
 443         (WebCore::CSSFontSelector::getFallbackFontData):
 444         * platform/MemoryPressureHandler.cpp:
 445         (WebCore::MemoryPressureHandler::releaseMemory):
 446         * platform/graphics/FontCache.cpp:
 447         (WebCore::fontCache): Return a reference
 448         * platform/graphics/FontCache.h:
 449         (WebCore::FontCachePurgePreventer::FontCachePurgePreventer):
 450         (WebCore::FontCachePurgePreventer::~FontCachePurgePreventer):
 451         * platform/graphics/FontGlyphs.cpp:
 452         (WebCore::FontGlyphs::FontGlyphs):
 453         (WebCore::FontGlyphs::releaseFontData):
 454         (WebCore::FontGlyphs::realizeFontDataAt):
 455         (WebCore::FontGlyphs::glyphDataAndPageForCharacter):
 456         * platform/graphics/freetype/FontPlatformDataFreeType.cpp:
 457         (WebCore::FontPlatformData::verticalData):
 458         * platform/graphics/ios/SimpleFontDataIOS.mm:
 459         (WebCore::SimpleFontData::platformCreateScaledFontData):
 460         * platform/graphics/mac/ComplexTextControllerCoreText.mm:
 461         (WebCore::ComplexTextController::collectComplexTextRunsForCharacters):
 462         * platform/graphics/mac/FontCacheMac.mm:
 463         (WebCore::invalidateFontCache):
 464         (WebCore::fontCacheRegisteredFontsChangedNotificationCallback):
 465         * platform/graphics/mac/SimpleFontDataMac.mm:
 466         (WebCore::SimpleFontData::platformDestroy):
 467         (WebCore::SimpleFontData::platformCreateScaledFontData):
 468         * platform/graphics/win/FontCacheWin.cpp:
 469         (WebCore::getCJKCodePageMasks):
 470         * platform/graphics/win/SimpleFontDataWin.cpp:
 471         (WebCore::SimpleFontData::containsCharacters):
 472         * platform/graphics/wince/FontCacheWinCE.cpp:
 473         (WebCore::getCJKCodePageMasks):
 474         * platform/graphics/wince/FontPlatformData.cpp:
 475         (WebCore::FontFamilyCodePageInfo::codePages):
 476         (WebCore::FixedSizeFontData::create):
 477         * platform/graphics/wince/GlyphPageTreeNodeWinCE.cpp:
 478         (WebCore::GlyphPage::fill):
 479         * platform/graphics/wince/SimpleFontDataWinCE.cpp:
 480         (WebCore::SimpleFontData::platformCreateScaledFontData):
 481         (WebCore::SimpleFontData::containsCharacters):
 482
 483 2014-04-24  Eric Carlson  <eric.carlson@apple.com>
 484
 485         [iOS] Manage AudioSession category according to media type
 486         https://bugs.webkit.org/show_bug.cgi?id=132096
 487
 488         Reviewed by Jer Noble.
 489
 490         * WebCore.exp.in: Export setting.
 491
 492         * html/HTMLMediaSession.cpp:
 493         (WebCore::HTMLMediaSession::HTMLMediaSession):
 494         (WebCore::initializeAudioSession): Deleted.
 495
 496         * page/Settings.cpp:
 497         * page/Settings.h:
 498         (WebCore::Settings::setShouldManageAudioSession): New.
 499         (WebCore::Settings::shouldManageAudioSession): Ditto.
 500
 501         * platform/audio/ios/AudioDestinationIOS.cpp:
 502         (WebCore::AudioDestinationIOS::AudioDestinationIOS): Use a MediaSession instead of inheriting
 503             from AudioListener and calling the AudioSession directly.
 504         (WebCore::AudioDestinationIOS::~AudioDestinationIOS): Ditto.
 505         (WebCore::AudioDestinationIOS::start): Notify session.
 506         (WebCore::AudioDestinationIOS::stop): Ditto.
 507         (WebCore::AudioDestinationIOS::beganAudioInterruption): Deleted.
 508         (WebCore::AudioDestinationIOS::endedAudioInterruption): Deleted.
 509         * platform/audio/ios/AudioDestinationIOS.h:
 510         (WebCore::AudioDestinationIOS::mediaType):
 511         (WebCore::AudioDestinationIOS::canReceiveRemoteControlCommands):
 512         (WebCore::AudioDestinationIOS::didReceiveRemoteControlCommand):
 513         (WebCore::AudioDestinationIOS::isPlaying): Deleted.
 514
 515         * platform/audio/ios/AudioSessionIOS.mm:
 516         (WebCore::categoryName): Debug-only logging function.
 517         (WebCore::AudioSession::setCategory): Don't stick with "media" once it is set.
 518
 519         * platform/audio/ios/MediaSessionManagerIOS.mm:
 520         (WebCore::MediaSessionManageriOS::resetRestrictions): Set up restrictions for WebAudio.
 521         (WebCore::MediaSessionManageriOS::updateNowPlayingInfo): Don't set invalid start time.
 522
 523         * platform/audio/mac/MediaSessionManagerMac.cpp:
 524         (MediaSessionManager::updateSessionState): Manage AudioSession.active when WebAudio clients
 525             come and go. Manage AudioSession.category according to the number of WebAudio and
 526             HTMLMediaElement clients.
 527
 528 2014-04-24  David Hyatt  <hyatt@apple.com>
 529
 530         [New Multicolumn] Client rects don't work with column spans.
 531         https://bugs.webkit.org/show_bug.cgi?id=132131
 532
 533         Reviewed by Dean Jackson.
 534
 535         Don't factor in the offset of the multicolumn set from the top
 536         of the multicolumn block. This was added already, and it doesn't
 537         need to be a part of columnTranslationForOffset.
 538
 539         Added fast/multicol/client-rects-spanners.html
 540
 541         * rendering/RenderMultiColumnSet.cpp:
 542         (WebCore::RenderMultiColumnSet::columnTranslationForOffset):
 543
 544 2014-04-24  Praveen R Jadhav  <praveen.j@samsung.com>
 545
 546         [EFL] WebKit build fails when MEDIA_SOURCE is enabled
 547         https://bugs.webkit.org/show_bug.cgi?id=132118
 548
 549         Reviewed by Brent Fulgham.
 550
 551         Files MediaSourceGStreamer.cpp, SourceBufferPrivateGStreamer.cpp and
 552         WebKitMediaSourceGStreamer.cpp are included for EFL port build.
 553
 554         No new tests. No change in behaviour.
 555
 556         * PlatformEfl.cmake: MediaSourceGStreamer.cpp, SourceBufferPrivateGStreamer.cpp
 557         and WebKitMediaSourceGStreamer.cpp are included for compilation.
 558
 559 2014-04-24  Tamas Gergely  <tgergely.u-szeged@partner.samsung.com>
 560
 561         ASSERTION FAILED: !begin.isIndefinite() in WebCore::SVGSMILElement::resolveFirstInterval.
 562         https://bugs.webkit.org/show_bug.cgi?id=131097
 563
 564         Reviewed by Darin Adler.
 565
 566         According to smil animation reference, max attribute cannot be 0.
 567
 568         Test: svg/animations/smil-animation-max-attribute-zero-crash.svg
 569
 570         * svg/animation/SVGSMILElement.cpp:
 571         (WebCore::SVGSMILElement::maxValue):
 572           changed (result < 0) to (result <= 0)
 573
 574 2014-04-24  Ryuan Choi  <ryuan.choi@samsung.com>
 575
 576         Remove screenColorProfile()
 577         https://bugs.webkit.org/show_bug.cgi?id=132035
 578
 579         Reviewed by Darin Adler.
 580
 581         Only chromium used screenColorProfile() since r120789.
 582
 583         * platform/PlatformScreen.h:
 584         * platform/efl/PlatformScreenEfl.cpp:
 585         (WebCore::screenColorProfile): Deleted.
 586         * platform/gtk/PlatformScreenGtk.cpp:
 587         (WebCore::screenColorProfile): Deleted.
 588         * platform/image-decoders/ImageDecoder.h:
 589         (WebCore::ImageDecoder::qcmsOutputDeviceProfile):
 590         * platform/ios/PlatformScreenIOS.mm:
 591         (WebCore::screenColorProfile): Deleted.
 592         * platform/mac/PlatformScreenMac.mm:
 593         (WebCore::screenColorProfile): Deleted.
 594         * platform/win/PlatformScreenWin.cpp:
 595         (WebCore::screenColorProfile): Deleted.
 596
 597 2014-04-24  Zalan Bujtas  <zalan@apple.com>
 598
 599         One more unreviewed build fix after r167755.
 600
 601         * html/shadow/mac/ImageControlsButtonElementMac.cpp:
 602         (WebCore::RenderImageControlsButton::updateLogicalWidth):
 603         (WebCore::RenderImageControlsButton::computeLogicalHeight):
 604
 605 2014-04-24  Zalan Bujtas  <zalan@apple.com>
 606
 607         Unreviewed build fix after r167755.
 608
 609         * rendering/RenderThemeMac.h:
 610
 611 2014-04-24  Zalan Bujtas  <zalan@apple.com>
 612
 613         Transition RenderTheme API from RenderObject* to const RenderObject&
 614         https://bugs.webkit.org/show_bug.cgi?id=132037
 615
 616         Reviewed by Andreas Kling.
 617
 618         Using const references provides better encapsulation and improve security.
 619
 620         No change in behavior.
 621
 622         * accessibility/AccessibilityObject.cpp:
 623         (WebCore::AccessibilityObject::boundingBoxForQuads):
 624         * dom/Element.cpp:
 625         (WebCore::Element::setActive):
 626         (WebCore::Element::setHovered):
 627         * editing/FrameSelection.cpp:
 628         (WebCore::FrameSelection::focusedOrActiveStateChanged):
 629         * html/HTMLFormControlElement.cpp:
 630         (WebCore::HTMLFormControlElement::disabledStateChanged):
 631         (WebCore::HTMLFormControlElement::readOnlyAttributeChanged):
 632         * html/HTMLInputElement.cpp:
 633         (WebCore::HTMLInputElement::setChecked):
 634         (WebCore::HTMLInputElement::setIndeterminate):
 635         * html/HTMLOptionElement.cpp:
 636         (WebCore::HTMLOptionElement::parseAttribute):
 637         * rendering/RenderBlock.cpp:
 638         (WebCore::RenderBlock::addVisualOverflowFromTheme):
 639         (WebCore::RenderBlock::baselinePosition):
 640         * rendering/RenderBox.cpp:
 641         (WebCore::RenderBox::paintBoxDecorations):
 642         * rendering/RenderButton.cpp:
 643         (WebCore::RenderButton::styleDidChange):
 644         * rendering/RenderFileUploadControl.cpp:
 645         (WebCore::RenderFileUploadControl::paintObject):
 646         * rendering/RenderFlowThread.cpp:
 647         (WebCore::RenderFlowThread::addRegionsVisualOverflowFromTheme):
 648         * rendering/RenderObject.cpp:
 649         (WebCore::RenderObject::drawLineForBoxSide):
 650         * rendering/RenderObject.h:
 651         * rendering/RenderProgress.cpp:
 652         (WebCore::RenderProgress::computeLogicalHeight):
 653         * rendering/RenderTextControlSingleLine.cpp:
 654         (WebCore::RenderTextControlSingleLine::paint):
 655         * rendering/RenderTheme.cpp:
 656         (WebCore::RenderTheme::paint):
 657         (WebCore::RenderTheme::paintBorderOnly):
 658         (WebCore::RenderTheme::paintDecorations):
 659         (WebCore::RenderTheme::baselinePosition):
 660         (WebCore::RenderTheme::adjustRepaintRect):
 661         (WebCore::RenderTheme::stateChanged):
 662         (WebCore::RenderTheme::updateControlStatesForRenderer):
 663         (WebCore::RenderTheme::extractControlStatesForRenderer):
 664         (WebCore::RenderTheme::isActive):
 665         (WebCore::RenderTheme::isChecked):
 666         (WebCore::RenderTheme::isIndeterminate):
 667         (WebCore::RenderTheme::isEnabled):
 668         (WebCore::RenderTheme::isFocused):
 669         (WebCore::RenderTheme::isPressed):
 670         (WebCore::RenderTheme::isSpinUpButtonPartPressed):
 671         (WebCore::RenderTheme::isReadOnlyControl):
 672         (WebCore::RenderTheme::isHovered):
 673         (WebCore::RenderTheme::isSpinUpButtonPartHovered):
 674         (WebCore::RenderTheme::isDefault):
 675         (WebCore::RenderTheme::paintInputFieldSpeechButton):
 676         (WebCore::RenderTheme::paintMeter):
 677         (WebCore::RenderTheme::paintSliderTicks):
 678         (WebCore::RenderTheme::progressBarRectForBounds):
 679         * rendering/RenderTheme.h:
 680         (WebCore::RenderTheme::controlSupportsTints):
 681         (WebCore::RenderTheme::paintCapsLockIndicator):
 682         (WebCore::RenderTheme::paintFileUploadIconDecorations):
 683         (WebCore::RenderTheme::imageControlsButtonSize):
 684         (WebCore::RenderTheme::paintCheckbox):
 685         (WebCore::RenderTheme::paintRadio):
 686         (WebCore::RenderTheme::paintButton):
 687         (WebCore::RenderTheme::paintInnerSpinButton):
 688         (WebCore::RenderTheme::paintCheckboxDecorations):
 689         (WebCore::RenderTheme::paintRadioDecorations):
 690         (WebCore::RenderTheme::paintButtonDecorations):
 691         (WebCore::RenderTheme::paintTextField):
 692         (WebCore::RenderTheme::paintTextFieldDecorations):
 693         (WebCore::RenderTheme::paintTextArea):
 694         (WebCore::RenderTheme::paintTextAreaDecorations):
 695         (WebCore::RenderTheme::paintMenuList):
 696         (WebCore::RenderTheme::paintMenuListDecorations):
 697         (WebCore::RenderTheme::paintMenuListButtonDecorations):
 698         (WebCore::RenderTheme::paintPushButtonDecorations):
 699         (WebCore::RenderTheme::paintSquareButtonDecorations):
 700         (WebCore::RenderTheme::paintProgressBar):
 701         (WebCore::RenderTheme::paintSliderTrack):
 702         (WebCore::RenderTheme::paintSliderThumb):
 703         (WebCore::RenderTheme::paintSliderThumbDecorations):
 704         (WebCore::RenderTheme::paintSearchField):
 705         (WebCore::RenderTheme::paintSearchFieldDecorations):
 706         (WebCore::RenderTheme::paintSearchFieldCancelButton):
 707         (WebCore::RenderTheme::paintSearchFieldDecorationPart):
 708         (WebCore::RenderTheme::paintSearchFieldResultsDecorationPart):
 709         (WebCore::RenderTheme::paintSearchFieldResultsButton):
 710         (WebCore::RenderTheme::paintMediaFullscreenButton):
 711         (WebCore::RenderTheme::paintMediaPlayButton):
 712         (WebCore::RenderTheme::paintMediaOverlayPlayButton):
 713         (WebCore::RenderTheme::paintMediaMuteButton):
 714         (WebCore::RenderTheme::paintMediaSeekBackButton):
 715         (WebCore::RenderTheme::paintMediaSeekForwardButton):
 716         (WebCore::RenderTheme::paintMediaSliderTrack):
 717         (WebCore::RenderTheme::paintMediaSliderThumb):
 718         (WebCore::RenderTheme::paintMediaVolumeSliderContainer):
 719         (WebCore::RenderTheme::paintMediaVolumeSliderTrack):
 720         (WebCore::RenderTheme::paintMediaVolumeSliderThumb):
 721         (WebCore::RenderTheme::paintMediaRewindButton):
 722         (WebCore::RenderTheme::paintMediaReturnToRealtimeButton):
 723         (WebCore::RenderTheme::paintMediaToggleClosedCaptionsButton):
 724         (WebCore::RenderTheme::paintMediaControlsBackground):
 725         (WebCore::RenderTheme::paintMediaCurrentTime):
 726         (WebCore::RenderTheme::paintMediaTimeRemaining):
 727         (WebCore::RenderTheme::paintMediaFullScreenVolumeSliderTrack):
 728         (WebCore::RenderTheme::paintMediaFullScreenVolumeSliderThumb):
 729         (WebCore::RenderTheme::paintSnapshottedPluginOverlay):
 730         (WebCore::RenderTheme::paintImageControlsButton):
 731         * rendering/RenderThemeIOS.h:
 732         * rendering/RenderThemeIOS.mm:
 733         (WebCore::RenderThemeIOS::addRoundedBorderClip):
 734         (WebCore::RenderThemeIOS::paintCheckboxDecorations):
 735         (WebCore::RenderThemeIOS::baselinePosition):
 736         (WebCore::RenderThemeIOS::paintRadioDecorations):
 737         (WebCore::RenderThemeIOS::paintTextFieldDecorations):
 738         (WebCore::RenderThemeIOS::paintTextAreaDecorations):
 739         (WebCore::RenderThemeIOS::paintMenuListButtonDecorations):
 740         (WebCore::RenderThemeIOS::paintSliderTrack):
 741         (WebCore::RenderThemeIOS::paintSliderThumbDecorations):
 742         (WebCore::RenderThemeIOS::paintProgressBar):
 743         (WebCore::RenderThemeIOS::paintSearchFieldDecorations):
 744         (WebCore::RenderThemeIOS::paintButtonDecorations):
 745         (WebCore::RenderThemeIOS::paintPushButtonDecorations):
 746         (WebCore::RenderThemeIOS::paintFileUploadIconDecorations):
 747         * rendering/RenderThemeMac.h:
 748         (WebCore::RenderThemeMac::updateActiveState):
 749         * rendering/RenderThemeMac.mm:
 750         (WebCore::RenderThemeMac::documentViewFor):
 751         (WebCore::RenderThemeMac::adjustRepaintRect):
 752         (WebCore::RenderThemeMac::convertToPaintingRect):
 753         (WebCore::RenderThemeMac::updateCheckedState):
 754         (WebCore::RenderThemeMac::updateEnabledState):
 755         (WebCore::RenderThemeMac::updateFocusedState):
 756         (WebCore::RenderThemeMac::updatePressedState):
 757         (WebCore::RenderThemeMac::controlSupportsTints):
 758         (WebCore::RenderThemeMac::paintTextField):
 759         (WebCore::RenderThemeMac::paintCapsLockIndicator):
 760         (WebCore::RenderThemeMac::paintTextArea):
 761         (WebCore::RenderThemeMac::paintMenuList):
 762         (WebCore::RenderThemeMac::paintMeter):
 763         (WebCore::RenderThemeMac::progressBarRectForBounds):
 764         (WebCore::RenderThemeMac::paintProgressBar):
 765         (WebCore::RenderThemeMac::paintMenuListButtonGradients):
 766         (WebCore::RenderThemeMac::paintMenuListButtonDecorations):
 767         (WebCore::RenderThemeMac::setPopupButtonCellState):
 768         (WebCore::RenderThemeMac::paintSliderTrack):
 769         (WebCore::RenderThemeMac::paintSliderThumb):
 770         (WebCore::RenderThemeMac::paintSearchField):
 771         (WebCore::RenderThemeMac::setSearchCellState):
 772         (WebCore::RenderThemeMac::paintSearchFieldCancelButton):
 773         (WebCore::RenderThemeMac::paintSearchFieldDecorationPart):
 774         (WebCore::RenderThemeMac::paintSearchFieldResultsDecorationPart):
 775         (WebCore::RenderThemeMac::paintSearchFieldResultsButton):
 776         (WebCore::RenderThemeMac::paintSnapshottedPluginOverlay):
 777         (WebCore::RenderThemeMac::paintImageControlsButton):
 778         (WebCore::RenderThemeMac::imageControlsButtonSize):
 779
 780 2014-04-23  Carlos Garcia Campos  <cgarcia@igalia.com>
 781
 782         [GTK] HTML Media capture attribute is a boolean since r163958
 783         https://bugs.webkit.org/show_bug.cgi?id=132061
 784
 785         Reviewed by Gustavo Noronha Silva.
 786
 787         Add new methods webkit_dom_html_input_element_get_capture_enabled
 788         and webkit_dom_html_input_element_set_capture_enabled using a
 789         boolean and deprecate the old methods.
 790
 791         * bindings/gobject/WebKitDOMDeprecated.cpp:
 792         (webkit_dom_html_input_element_get_capture):
 793         (webkit_dom_html_input_element_set_capture):
 794         * bindings/gobject/WebKitDOMDeprecated.h:
 795         * bindings/gobject/WebKitDOMDeprecated.symbols:
 796         * bindings/gobject/webkitdom.symbols:
 797         * bindings/scripts/CodeGeneratorGObject.pm:
 798         (GetEffectiveFunctionName): Helper function to rename API methods
 799         for special cases.
 800         (GenerateFunction): Use GetEffectiveFunctionName().
 801
 802 2014-04-23  Praveen R Jadhav  <praveen.j@samsung.com>
 803
 804         [MediaStream] Implement MediaStream active attribute
 805         https://bugs.webkit.org/show_bug.cgi?id=131973
 806
 807         Reviewed by Eric Carlson.
 808
 809         MediaStream .active attribute are introduced which will replace
 810         .ended attribute. This patch implements the newly introduced attributes.
 811
 812         MediaStream-add-remove-tracks.html is updated to handle this scenario.
 813
 814         * Modules/mediastream/MediaStream.cpp:
 815         (WebCore::MediaStream::active): Added.
 816         (WebCore::MediaStream::setActive): Added.
 817         (WebCore::MediaStream::addTrack): Propagates 'onactive' event when required.
 818         (WebCore::MediaStream::removeTrack): Propagates 'oninactive' event when required.
 819         (WebCore::MediaStream::trackDidEnd): Propagates 'oninactive' event when required.
 820         (WebCore::MediaStream::streamDidEnd):
 821         (WebCore::MediaStream::setStreamIsActive): Added.
 822         * Modules/mediastream/MediaStream.h:
 823         * Modules/mediastream/MediaStream.idl:
 824         * dom/EventNames.h:
 825         * platform/mediastream/MediaStreamPrivate.cpp:
 826         (WebCore::MediaStreamPrivate::MediaStreamPrivate): Initialize .active attribute
 827         (WebCore::MediaStreamPrivate::setEnded):
 828         (WebCore::MediaStreamPrivate::setActive): Added.
 829         * platform/mediastream/MediaStreamPrivate.h:
 830         (WebCore::MediaStreamPrivate::active): Added.
 831
 832 2014-04-23  Darin Adler  <darin@apple.com>
 833
 834         [Cocoa] fix CF leaks found by code inspection
 835         https://bugs.webkit.org/show_bug.cgi?id=132106
 836
 837         Reviewed by Andreas Kling.
 838
 839         * page/CaptionUserPreferencesMediaAF.cpp:
 840         (WebCore::trackDisplayName): Added a missing adoptCF.
 841
 842         * platform/Language.cpp:
 843         (WebCore::displayNameForLanguageLocale): Added a missing adoptCF.
 844
 845         * platform/graphics/FontPlatformData.cpp:
 846         (WebCore::FontPlatformData::openTypeTable): Added a missing adoptCF.
 847
 848         * platform/graphics/avfoundation/cf/InbandTextTrackPrivateAVCF.cpp:
 849         (WebCore::InbandTextTrackPrivateAVCF::label): Added two missing adoptCF.
 850
 851         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
 852         (WebCore::AVFWrapper::createImageForTimeInRect): Added two missing adoptCF.
 853
 854         * platform/graphics/cg/PDFDocumentImage.cpp:
 855         (WebCore::PDFDocumentImage::createPDFDocument): Added missing adoptCF.
 856
 857         * platform/graphics/cocoa/FontPlatformDataCocoa.mm:
 858         (WebCore::cascadeToLastResortFontDescriptor): Added two missing adoptCF.
 859
 860         * platform/graphics/mac/FontMac.mm:
 861         (WebCore::Font::primaryFontDataIsSystemFont): Added missing adoptCF.
 862
 863         * platform/graphics/mac/SimpleFontDataMac.mm:
 864         (WebCore::hasCustomTracking): Added missing adoptCF.
 865
 866         * platform/image-decoders/ImageDecoder.h:
 867         (WebCore::ImageDecoder::qcmsOutputDeviceProfile): Added CFRelease.
 868
 869         * plugins/mac/PluginPackageMac.cpp:
 870         (WebCore::readPListFile): Added two missing adoptCF.
 871
 872 2014-04-23  Ryosuke Niwa  <rniwa@webkit.org>
 873
 874         REGRESSION (r157328): popover to check into flight ba.com dismisses instantly when focusing form
 875         https://bugs.webkit.org/show_bug.cgi?id=131949
 876
 877         Address the review comment.
 878
 879         * dom/EventDispatcher.cpp:
 880         (WebCore::EventRelatedNodeResolver::findHostOfTreeScopeInTargetTreeScope):
 881
 882 2014-04-23  Andreas Kling  <akling@apple.com>
 883
 884         CachedResourceLoader hoards URLs indefinitely for no good reason.
 885         <https://webkit.org/b/132102>
 886         <rdar://problem/16708265>
 887
 888         Since we don't care about CachedResourceLoader::m_validatedURL's after
 889         the Document has finished dispatching its initial load event, clear the
 890         set at that point, and don't add any new URLs to it.
 891
 892         Reviewed by Anders Carlsson.
 893
 894         * dom/Document.cpp:
 895         (WebCore::Document::dispatchWindowLoadEvent):
 896         * loader/cache/CachedResourceLoader.cpp:
 897         (WebCore::CachedResourceLoader::requestResource):
 898         (WebCore::CachedResourceLoader::documentDidFinishLoadEvent):
 899         * loader/cache/CachedResourceLoader.h:
 900
 901 2014-04-23  Andreas Kling  <akling@apple.com>
 902
 903         Canvas cache of clean URLs can grow without bounds.
 904         <https://webkit.org/b/132091>
 905         <rdar://problem/16695665>
 906
 907         Remove a silly "optimization" that kept a cache of clean URLs
 908         that can be drawn into a canvas without tainting it, all to avoid
 909         the "expensive" checks to determine whether it would taint.
 910
 911         Reviewed by Benjamin Poulain.
 912
 913         * html/canvas/CanvasRenderingContext.cpp:
 914         (WebCore::CanvasRenderingContext::wouldTaintOrigin):
 915         * html/canvas/CanvasRenderingContext.h:
 916
 917 2014-04-23  Benjamin Poulain  <bpoulain@apple.com>
 918
 919         [iOS][WK2] Fix a few mistakes affecting the initial layout and the initial unobscured rect
 920         https://bugs.webkit.org/show_bug.cgi?id=132093
 921
 922         Reviewed by Tim Horton.
 923
 924         Change the minimum layout size to float point values to account for size defined on retina displays.
 925         The minimum layout size supports half-pixels, the value is rounded later when computing the layout size
 926         in document coordinates.
 927
 928         * WebCore.exp.in:
 929         * page/ViewportConfiguration.cpp:
 930         (WebCore::ViewportConfiguration::ViewportConfiguration):
 931         Setting the initial content size is incorrect. The layout size computation already take into account
 932         empty size for the first layout.
 933
 934         Setting the content size upfront make the first computation incorrect when the viewport arguments specify
 935         the initial scale.
 936
 937         (WebCore::ViewportConfiguration::setMinimumLayoutSize):
 938         * page/ViewportConfiguration.h:
 939         (WebCore::ViewportConfiguration::minimumLayoutSize):
 940
 941 2014-04-23  Brent Fulgham  <bfulgham@apple.com>
 942
 943         [Mac, iOS] Stop buffering media when on an inactive tab.
 944         https://bugs.webkit.org/show_bug.cgi?id=132077
 945
 946         Reviewed by Eric Carlson.
 947
 948         * html/HTMLMediaElement.cpp: Rename 'm_isDisplaySleepDisablingSuspended'
 949         to 'm_elementIsHidden'.
 950         (WebCore::HTMLMediaElement::HTMLMediaElement):
 951         (WebCore::HTMLMediaElement::visibilityStatusChanged): Notify the
 952         media session that the element is (or is not) hidden.
 953         (WebCore::HTMLMediaElement::setShouldBufferData): Added.
 954         * html/HTMLMediaElement.h:
 955         * platform/audio/MediaSession.cpp:
 956         (WebCore::MediaSession::clientWillBeginPlayback): Tell media to
 957         buffer if not hidden or playing.
 958         (WebCore::MediaSession::clientWillPausePlayback): Ditto.
 959         (WebCore::MediaSession::visibilityChanged): Added. Client API, just relays call
 960         to updateClientDataBuffering.
 961         (WebCore::MediaSession::updateClientDataBuffering): Tell client it should only
 962         buffer data if it's currently playing, or not hidden.
 963         * platform/audio/MediaSession.h:
 964         * platform/graphics/MediaPlayer.cpp:
 965         (WebCore::MediaPlayer::setShouldBufferData): Added: Just relays to
 966         MediaPlayerPrivate object.
 967         * platform/graphics/MediaPlayer.h:
 968         * platform/graphics/MediaPlayerPrivate.h:
 969         (WebCore::MediaPlayerPrivateInterface::setShouldBufferData):
 970         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
 971         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
 972         (WebCore::MediaPlayerPrivateAVFoundationObjC::MediaPlayerPrivateAVFoundationObjC):
 973         (WebCore::MediaPlayerPrivateAVFoundationObjC::setShouldBufferData): Added.
 974         Detach the player item from the player if we don't want to continue
 975         buffering or other background tasks.
 976
 977 2014-04-23  Alexey Proskuryakov  <ap@apple.com>
 978
 979         Eliminate internals.setMockScrollbarsEnabled()
 980         https://bugs.webkit.org/show_bug.cgi?id=132085
 981
 982         Reviewed by Tim Horton.
 983
 984         This was essentially unused, and also didnt work.
 985
 986         * testing/InternalSettings.cpp:
 987         (WebCore::InternalSettings::Backup::restoreTo):
 988         (WebCore::InternalSettings::setMockScrollbarsEnabled): Deleted.
 989         * testing/InternalSettings.h:
 990         * testing/InternalSettings.idl:
 991
 992 2014-04-23  Anders Carlsson  <andersca@apple.com>
 993
 994         Don't migrate the WKView.h header from WebCore to WebKit
 995         https://bugs.webkit.org/show_bug.cgi?id=132086
 996
 997         Reviewed by Dan Bernstein.
 998
 999         * WebCore.xcodeproj/project.pbxproj:
1000         Add WAKViewInternal.h.
1001
1002         * platform/WAKViewInternal.h: Added.
1003
1004         * platform/ios/wak/WAKClipView.m:
1005         Import WAKViewInternal.h instead of WAKViewPrivate.h.
1006
1007         * platform/ios/wak/WAKScrollView.mm:
1008         Import WAKViewInternal.h instead of WAKViewPrivate.h.
1009
1010         * platform/ios/wak/WAKView.h:
1011         Move ivars to a class extension in WAKViewInternal.h and remove WKView.h import.
1012
1013         * platform/ios/wak/WAKView.mm:
1014         Import WAKViewInternal.h instead of WAKViewPrivate.h.
1015
1016         * platform/ios/wak/WAKViewPrivate.h:
1017         Import WKViewPrivate.h.
1018
1019 2014-04-23  David Hyatt  <hyatt@apple.com>
1020
1021         [New Multicolumn] fast/multicol/fixed-column-percent-logical-height-orthogonal-writing-mode.html fails
1022         https://bugs.webkit.org/show_bug.cgi?id=132078
1023
1024         Reviewed by Anders Carlsson.
1025
1026         * rendering/RenderBox.cpp:
1027         (WebCore::RenderBox::skipContainingBlockForPercentHeightCalculation):
1028         Add a parameter to test for orthogonal writing modes. If we're perpendicular,
1029         then we should not skip the flow thread, since we resolve relative to the column width,
1030         and that is always set.
1031
1032         (WebCore::RenderBox::computePercentageLogicalHeight):
1033         Patched to pass in whether or not the box and the ancestor block are perpendicular.
1034
1035         * rendering/RenderBox.h:
1036         Patched the signature of skipContainingBlockForPercentHeightCalculation
1037
1038 2014-04-23  Myles C. Maxfield  <mmaxfield@apple.com>
1039
1040         [OS X] Make checking if a font is the system font more robust
1041         https://bugs.webkit.org/show_bug.cgi?id=132030
1042
1043         Reviewed by Dean Jackson.
1044
1045         Instead of inspecting a font's name to determine if it is a system font,
1046         on OS X we can ask the system directly.
1047
1048         This patch also moves a platform-specific check into platform-specific
1049         code, so that other platforms don't check for OS X-specific behavior.
1050
1051         Covered by existing tests.
1052
1053         * platform/graphics/Font.cpp:
1054         (WebCore::Font::hasValidAverageCharWidth):
1055         * platform/graphics/Font.h:
1056         * platform/graphics/mac/FontMac.mm:
1057         (WebCore::Font::primaryFontDataIsSystemFont):
1058
1059 2014-04-23  David Hyatt  <hyatt@apple.com>
1060
1061         [New Multicolumn] Assertion failure in huge-column-count.html
1062         https://bugs.webkit.org/show_bug.cgi?id=132071
1063
1064         Reviewed by Dean Jackson.
1065
1066         * rendering/RenderBlock.cpp:
1067         (WebCore::RenderBlock::regionAtBlockOffset):
1068         Back out this change, since it wasn't general enough.
1069
1070         * rendering/RenderFlowThread.cpp:
1071         (WebCore::RenderFlowThread::getRegionRangeForBox):
1072         The real issue was that this loop needed to consider the actual box
1073         rather than starting from the parent. This was a non-issue for normal
1074         regions (which cannot have nested flow threads), but for columns, you
1075         have to consider the fact that the box could itself be a flow thread.
1076
1077 2014-04-23  Andreas Kling  <akling@apple.com>
1078
1079         [iOS] Memory pressure notification should fire on main thread.
1080         <https://webkit.org/b/132074>
1081
1082         Rejig the memory pressure notification to fire on the main queue
1083         directly instead of rerouting it manually.
1084
1085         Reviewed by Mark Rowe.
1086
1087         * platform/cocoa/MemoryPressureHandlerCocoa.mm:
1088         (WebCore::MemoryPressureHandler::install):
1089
1090 2014-04-23  Commit Queue  <commit-queue@webkit.org>
1091
1092         Unreviewed, rolling out r167720.
1093         https://bugs.webkit.org/show_bug.cgi?id=132075
1094
1095         broke eight newmulticol tests (Requested by thorton on
1096         #webkit).
1097
1098         Reverted changeset:
1099
1100         "[New Multicolumn] Assertion failure in huge-column-
1101         count.html"
1102         https://bugs.webkit.org/show_bug.cgi?id=132071
1103         http://trac.webkit.org/changeset/167720
1104
1105 2014-04-23  David Hyatt  <hyatt@apple.com>
1106
1107         [New Multicolumn] Assertion failure in huge-column-count.html
1108         https://bugs.webkit.org/show_bug.cgi?id=132071
1109
1110         Reviewed by Dean Jackson.
1111
1112         * rendering/RenderBlock.cpp:
1113         (WebCore::RenderBlock::regionAtBlockOffset):
1114         Remove the code that returned 0 here, since we're going to patch a lower-level
1115         function to catch all cases.
1116
1117         * rendering/RenderFlowThread.cpp:
1118         (WebCore::RenderFlowThread::getRegionRangeForBox):
1119         Don't allow in-flow RenderFlowThreads to ever have a region range. The sets
1120         are what should have ranges... the flow thread needs to just be ignored.
1121
1122 2014-04-23  David Hyatt  <hyatt@apple.com>
1123
1124         [New Multicolumn] Crasher when clearing out a flow thread in multicolumn layout.
1125         https://bugs.webkit.org/show_bug.cgi?id=132069
1126
1127         Reviewed by Dean Jackson.
1128
1129         This is imported from a patch Morten did for Blink, but I had to change it a fair
1130         bit. deleteLines() is used to handle simple line box layout instead of just calling
1131         deleteLineBoxTree.
1132
1133         I also had to disable the layout state to stop asserts on repaint when the children
1134         get moved. Not sure why Blink didn't hit this, but it's simple enough to add a
1135         LayoutStateDisabler to stop the assert.
1136
1137         Added fast/multicol/inline-children-crash.html
1138
1139         * rendering/RenderMultiColumnFlowThread.cpp:
1140         (WebCore::RenderMultiColumnFlowThread::evacuateAndDestroy):
1141
1142 2014-04-23  Andreas Kling  <akling@apple.com>
1143
1144         [iOS WebKit2] IOSurfacePool should force CA to actually garbage collect surfaces.
1145         <https://webkit.org/b/132065>
1146         <rdar://problem/16110687>
1147
1148         Add a platformGarbageCollectNow() helper function to IOSurfacePool that
1149         triggers a sweep of the IOSurfaces. Call this from collectionTimerFired()
1150         and discardAllSurfaces().
1151
1152         This lets us drop all otherwise-unused 420f surfaces on memory pressure.
1153
1154         Reviewed by Tim Horton.
1155
1156         * WebCore.xcodeproj/project.pbxproj:
1157         * platform/graphics/cg/IOSurfacePool.cpp:
1158         (WebCore::IOSurfacePool::collectionTimerFired):
1159         (WebCore::IOSurfacePool::discardAllSurfaces):
1160         * platform/graphics/cg/IOSurfacePool.h:
1161         * platform/graphics/cocoa/IOSurfacePoolCocoa.mm: Added.
1162         (WebCore::IOSurfacePool::platformGarbageCollectNow):
1163
1164 2014-04-23  Morten Stenshorne  <mstensho@opera.com>
1165
1166         REGRESSION (Safari 6 - ToT): Incorrectly assumes that RenderStyle data can be shared
1167         https://bugs.webkit.org/show_bug.cgi?id=113058
1168
1169         Reviewed by David Hyatt.
1170
1171         Before sharing CSS properties with an element in the cache, we need to
1172         check that the new element is suitable for this, just like we check
1173         elements before inserting them into the cache.
1174
1175         Test: fast/css/identical-logical-height-decl.html
1176
1177         * css/StyleResolver.cpp:
1178         (WebCore::StyleResolver::applyMatchedProperties):
1179
1180 2014-04-23  Commit Queue  <commit-queue@webkit.org>
1181
1182         Unreviewed, rolling out r167713.
1183         https://bugs.webkit.org/show_bug.cgi?id=132070
1184
1185         broke hundreds of tests (Requested by thorton on #webkit).
1186
1187         Reverted changeset:
1188
1189         "[OS X] Make checking if a font is the system font more
1190         robust"
1191         https://bugs.webkit.org/show_bug.cgi?id=132030
1192         http://trac.webkit.org/changeset/167713
1193
1194 2014-04-22  David Hyatt  <hyatt@apple.com>
1195
1196         [New Multicolumn] Nested columns not working at all.
1197         https://bugs.webkit.org/show_bug.cgi?id=131805
1198
1199         Reviewed by Dean Jackson.
1200
1201         Add support for nested pagination contexts, allowing for an arbitrary level
1202         of nesting of multicolumn layouts. There were a number of things that had to
1203         be patched in order for this to work.
1204
1205         * rendering/RenderBlock.cpp:
1206         (WebCore::RenderBlock::regionAtBlockOffset):
1207         Make sure RenderMultiColumnFlowThreads just return null for regions at any
1208         block offset. Individual region sets will be created as you cross ancestor
1209         regions eventually, so this is just getting in the way.
1210
1211         * rendering/RenderLayer.cpp:
1212         (WebCore::RenderLayer::enclosingPaginationLayerInSubtree):
1213         Add a new helper method for obtaining an enclosingPaginationLayer when
1214         constrained by some root. This function ensures you don't accidentally
1215         cross your subtree root when looking for enclosing pagination layers.
1216
1217         (WebCore::RenderLayer::collectFragments):
1218         Patch collectFragments to know how to recur to collect ancestor fragments
1219         in order to apply nested splitting as you cross pagination boundaries.
1220
1221         (WebCore::RenderLayer::updatePaintingInfoForFragments):
1222         (WebCore::RenderLayer::calculateClipRects):
1223         * rendering/RenderLayer.h:
1224         (WebCore::LayerFragment::LayerFragment):
1225         (WebCore::LayerFragment::setRects):
1226         (WebCore::LayerFragment::moveBy):
1227         (WebCore::LayerFragment::intersect):
1228         Improve the LayerFragment so that it caches transformed bounding boxes as
1229         well. This is needed to fix intersectsDamageRect so that it doesn't grab
1230         the wrong bounding box when checking inline layers that are paginated.
1231
1232         * rendering/RenderMultiColumnFlowThread.cpp:
1233         (WebCore::RenderMultiColumnFlowThread::flowThreadDescendantInserted):
1234         Ignore inserted flow threads inside an ancestor flow thread, since we only
1235         care about what the sets do.
1236
1237         * rendering/RenderObject.cpp:
1238         (WebCore::RenderObject::insertedIntoTree):
1239         Make sure that nested flow thread layers return themselves when a child
1240         is inserted directly under them.
1241
1242 2014-04-22  Myles C. Maxfield  <mmaxfield@apple.com>
1243
1244         [OS X] Make checking if a font is the system font more robust
1245         https://bugs.webkit.org/show_bug.cgi?id=132030
1246
1247         Reviewed by Dean Jackson.
1248
1249         Instead of inspecting a font's name to determine if it is a system font,
1250         on OS X we can ask the system directly.
1251
1252         This patch also moves a platform-specific check into platform-specific
1253         code, so that other platforms don't check for OS X-specific behavior.
1254
1255         Covered by existing tests.
1256
1257         * platform/graphics/Font.cpp:
1258         (WebCore::Font::hasValidAverageCharWidth):
1259         * platform/graphics/Font.h:
1260         * platform/graphics/mac/FontMac.mm:
1261         (WebCore::Font::primaryFontDataIsSystemFont):
1262
1263 2014-04-23  Philippe Normand  <pnormand@igalia.com>
1264
1265         [GTK] Focus management API was moved from HTMLDocument to Document in r166668
1266         https://bugs.webkit.org/show_bug.cgi?id=132060
1267
1268         Reviewed by Carlos Garcia Campos.
1269
1270         Deprecate the removed methods in WebKitDOMHTMLDocument to keep API compatibility.
1271
1272         * bindings/gobject/WebKitDOMDeprecated.cpp:
1273         (webkit_dom_html_document_get_active_element):
1274         (webkit_dom_html_document_has_focus):
1275         * bindings/gobject/WebKitDOMDeprecated.h:
1276         * bindings/gobject/WebKitDOMDeprecated.symbols:
1277         * bindings/gobject/webkitdom.symbols:
1278
1279 2014-04-22  Andreas Kling  <akling@apple.com>
1280
1281         [iOS WebKit1] MemoryPressureHandler::respondToMemoryPressure called on wrong thread.
1282         <https://webkit.org/b/132041>
1283         <rdar://problem/16687238>
1284
1285         Always dispatch the memory pressure relief code on the main queue.
1286         This is really only needed for iOS/WK1, but we might as well share the code.
1287
1288         Reviewed by Michael Saboff.
1289
1290         * platform/cocoa/MemoryPressureHandlerCocoa.mm:
1291         (WebCore::MemoryPressureHandler::install):
1292
1293 2014-04-23  Philippe Normand  <pnormand@igalia.com>
1294
1295         [GTK] ShadowRoot API was removed in r164131
1296         https://bugs.webkit.org/show_bug.cgi?id=132059
1297
1298         Reviewed by Gustavo Noronha Silva.
1299
1300         Add ShadowRoot removed API to WebKitDOMDeprecated to keep API/ABI compatibility.
1301
1302         * bindings/gobject/WebKitDOMDeprecated.cpp:
1303         (webkit_dom_shadow_root_class_init):
1304         (webkit_dom_shadow_root_init):
1305         (webkit_dom_shadow_root_element_from_point):
1306         (webkit_dom_shadow_root_get_active_element):
1307         (webkit_dom_shadow_root_get_apply_author_styles):
1308         (webkit_dom_shadow_root_get_element_by_id):
1309         (webkit_dom_shadow_root_get_elements_by_class_name):
1310         (webkit_dom_shadow_root_get_elements_by_tag_name):
1311         (webkit_dom_shadow_root_get_elements_by_tag_name_ns):
1312         (webkit_dom_shadow_root_get_inner_html):
1313         (webkit_dom_shadow_root_get_reset_style_inheritance):
1314         (webkit_dom_shadow_root_get_selection):
1315         (webkit_dom_shadow_root_set_apply_author_styles):
1316         (webkit_dom_shadow_root_set_inner_html):
1317         (webkit_dom_shadow_root_set_reset_style_inheritance):
1318         * bindings/gobject/WebKitDOMDeprecated.h:
1319         * bindings/gobject/WebKitDOMDeprecated.symbols:
1320
1321 2014-04-23  Morten Stenshorne  <mstensho@opera.com>
1322
1323         [New Multicolumn] fast/multicol/overflow-content.html displays red
1324         https://bugs.webkit.org/show_bug.cgi?id=131809
1325
1326         Reviewed by David Hyatt.
1327
1328         Insert a break at end of content on our own in the multicol code, to make sure
1329         that overflow is accounted for, and also to make sure that we account for all
1330         content in non-final sets (i.e. those preceding spanners).
1331
1332         In other words, this will additionally fix balancing issues in sets preceding
1333         a spanner. Added a test for that.
1334
1335         Tests: fast/multicol/break-in-columns-before-spanner.html
1336                fast/multicol/newmulticol/compare-with-old-impl/overflow-content.html
1337
1338         * rendering/RenderFlowThread.cpp:
1339         (WebCore::RenderFlowThread::regionInRange): Deleted.
1340         * rendering/RenderFlowThread.h:
1341         * rendering/RenderMultiColumnSet.cpp:
1342         (WebCore::RenderMultiColumnSet::distributeImplicitBreaks):
1343         * rendering/RenderNamedFlowThread.cpp:
1344         (WebCore::RenderNamedFlowThread::applyBreakAfterContent):
1345         * rendering/RenderNamedFlowThread.h:
1346
1347 2014-04-23  Morten Stenshorne  <mstensho@opera.com>
1348
1349         Overflow propagation broken in BTT and RTL writing-modes
1350         https://bugs.webkit.org/show_bug.cgi?id=113781
1351
1352         Reviewed by David Hyatt.
1353
1354         Overflow rectangles are not quite physical, not quite logical. This
1355         means that we cannot use clientBoxRect() directly to represent a
1356         rectangle that expresses exactly no overflow. This rectangle is the
1357         padding box (relative to the border box) in vertical-lr and
1358         horizontal-tb, but the block-direction borders need to be flipped in
1359         vertical-rl and horizontal-bt.
1360
1361         Tests: fast/css/overflow-btt-border-after.html
1362                fast/css/overflow-rtl-border-after.html
1363
1364         * WebCore.exp.in:
1365         * rendering/RenderBlock.cpp:
1366         (WebCore::RenderBlock::computeOverflow):
1367         * rendering/RenderBox.cpp:
1368         (WebCore::RenderBox::addLayoutOverflow):
1369         (WebCore::RenderBox::addVisualOverflow):
1370         (WebCore::RenderBox::layoutOverflowRectForPropagation):
1371         * rendering/RenderBox.h:
1372         (WebCore::RenderBox::layoutOverflowRect):
1373
1374 2014-04-23  Andrei Bucur  <abucur@adobe.com>
1375
1376         [CSS Regions] Improve the debugging infrastructure
1377         https://bugs.webkit.org/show_bug.cgi?id=132042
1378
1379         Reviewed by Mihnea Ovidenie.
1380
1381         This patch improves the debugging code for CSS Regions.
1382
1383         Tests: No function change. No new tests.
1384
1385         * rendering/RenderObject.cpp:
1386         (WebCore::RenderObject::showRegionsInformation): Print brackets around the region range when dumping
1387         the render tree to the console.
1388         * rendering/RootInlineBox.cpp:
1389         (WebCore::RootInlineBox::containingRegion): Convert the ASSERT to an ASSERT_WITH_SECURITY_IMPLICATION.
1390
1391 2014-04-22  Ryosuke Niwa  <rniwa@webkit.org>
1392
1393         Cursor doesn't change back to pointer when leaving the Safari window
1394         https://bugs.webkit.org/show_bug.cgi?id=132038
1395
1396         Reviewed by Alexey Proskuryakov.
1397
1398         r147739 incorrectly added an early exit in EventHandler::selectCursor when hit test result didn't have
1399         any node associated with it. Since we will hit this code when the cursor is outside of the WebView,
1400         we still need to take the CURSOR_AUTO path as did the code before r147739.
1401
1402         No new test is added since this behavior can't be tested in DRT or WTR.
1403
1404         * page/EventHandler.cpp:
1405         (WebCore::EventHandler::selectCursor):
1406
1407 2014-04-22  Zalan Bujtas  <zalan@apple.com>
1408
1409         Do not paint border image when the border rect is empty.
1410         https://bugs.webkit.org/show_bug.cgi?id=131988
1411
1412         Reviewed by Darin Adler.
1413
1414         http://trac.webkit.org/changeset/167351 introduced an early return when border
1415         rect is empty. This patch ensures that border image is not painted either in that case.
1416
1417         Modified padding-margin-negative-border.html to cover border-image case.
1418
1419         * rendering/RenderBoxModelObject.cpp:
1420         (WebCore::RenderBoxModelObject::paintBorder):
1421
1422 2014-04-22  Tim Horton  <timothy_horton@apple.com>
1423
1424         ASSERTION FAILED: scrollerImp == scrollbarPainterForScrollbar(_scrollbar) on two API tests
1425         https://bugs.webkit.org/show_bug.cgi?id=132034
1426         <rdar://problem/16624332>
1427
1428         Reviewed by Simon Fraser.
1429
1430         * platform/mac/ScrollAnimatorMac.mm:
1431         (-[WebScrollbarPainterDelegate shouldUseLayerPerPartForScrollerImp:]):
1432         scrollerImpWithStyle:controlSize:horizontal:replacingScrollerImp: moves the delegate
1433         from the old scrollerImp to the new one, and also happens to call shouldUseLayerPerPartForScrollerImp.
1434         Since scrollerImpWithStyle: has not returned yet, the scrollbarPainterForScrollbar still returns the old scrollerImp,
1435         so this assertion fires. It's safe to remove this because supportsUpdateOnSecondaryThread doesn't make use of the imp.
1436
1437 2014-04-22  Ryosuke Niwa  <rniwa@webkit.org>
1438
1439         REGRESSION (r157328): popover to check into flight ba.com dismisses instantly when focusing form
1440         https://bugs.webkit.org/show_bug.cgi?id=131949
1441
1442         Reviewed by Darin Adler.
1443
1444         The regression was caused by two bugs:
1445         1. The event didn't stop propagating itself even when it should.
1446            If the related target is same as the event origin, the event propagation should stop when the event reaches
1447            the root of the related target's tree scope. Otherwise, it should stop when it reaches the related target.
1448
1449         2. Mouse event's related target exposed nodes inside a user-agent shadow DOM when the related target appeared
1450            inside the origin.
1451
1452         Fixed the bugs by re-introducing path shrinkage algorithm removed in r157328 into EventPath::setRelatedTarget
1453         and adding an algorithm to determine the least common ancestor of the related target and the current target
1454         in moveToParentOrShadowHost. The latter algorithm doesn't match the shadow DOM specification:
1455         http://www.w3.org/TR/2013/WD-shadow-dom-20130514/
1456         but it's good enough in terms of the Web exposed behavior as we don't support author defined insertion points.
1457
1458         Test: fast/events/shadow-event-path.html
1459
1460         * dom/EventDispatcher.cpp:
1461         (WebCore::EventRelatedNodeResolver::moveToParentOrShadowHost):
1462         (WebCore::EventRelatedNodeResolver::findHostOfTreeScopeInTargetTreeScope): Added.
1463         (WebCore::EventDispatcher::dispatchEvent):
1464         (WebCore::EventPath::setRelatedTarget):
1465
1466 2014-04-22  Ryosuke Niwa  <rniwa@webkit.org>
1467
1468         Rollout r156635 since the old behavior was intentional.
1469
1470         * page/EventHandler.cpp:
1471         (WebCore::EventHandler::selectCursor):
1472
1473 2014-04-22  Commit Queue  <commit-queue@webkit.org>
1474
1475         Unreviewed, rolling out r167674.
1476         https://bugs.webkit.org/show_bug.cgi?id=132025
1477
1478         Going a different way with this (Requested by bradee-oh on
1479         #webkit).
1480
1481         Reverted changeset:
1482
1483         "Change Image Controls replacement to use selection and paste"
1484         https://bugs.webkit.org/show_bug.cgi?id=131992
1485         http://trac.webkit.org/changeset/167674
1486
1487 2014-04-22  Brent Fulgham  <bfulgham@apple.com>
1488
1489         [Win] Support Python 2.7 in Cygwin
1490         https://bugs.webkit.org/show_bug.cgi?id=132023
1491
1492         Reviewed by Michael Saboff.
1493
1494         * DerivedSources.make: Use proper path to Cygwin on
1495         all platforms.
1496
1497 2014-04-22  Andreas Kling  <akling@apple.com>
1498
1499         REGRESSION (r151839): Subframe keeps getting mousemove events with the same coordinates after hiding a hovered element.
1500         <https://webkit.org/b/131974>
1501         <rdar://problem/15907469>
1502
1503         When the currently hovered element disappears as a result of style recalc,
1504         we send a fake mousemove event to the page, to see if anything newly added
1505         should become hovered.
1506
1507         The faking mechanism lives in EventHandler and simply synthesizes a new
1508         mousemove event using the last seen mouse location.
1509
1510         The problem here is that we were sending this fake mousemove event to the
1511         subframe where the hovered element lived. Since subframes aren't kept up
1512         to date on recent mouse locations, this could cause some strange behavior
1513         where a subframe would dispatch mousemove events with stale coordinates.
1514
1515         The solution is to always dispatch fake mousemove events from the main
1516         frame's event handler. This is how real event delivery happens, and hit
1517         testing will then find the appropriate subframe, if any.
1518
1519         Reviewed by Benjamin Poulain.
1520
1521         Test: fast/events/ghostly-mousemoves-in-subframe.html
1522
1523         * dom/Document.cpp:
1524         (WebCore::Document::recalcStyle):
1525
1526 2014-04-22  Myles C. Maxfield  <mmaxfield@apple.com>
1527
1528         [OS X] Glyph spacing for system fonts may be incorrect
1529         https://bugs.webkit.org/show_bug.cgi?id=131967
1530
1531         Unreviewed iOS build fix after r167679.
1532
1533         * platform/graphics/mac/SimpleFontDataMac.mm:
1534
1535 2014-04-22  Myles C. Maxfield  <mmaxfield@apple.com>
1536
1537         [OS X] Glyph spacing for system fonts may be incorrect
1538         https://bugs.webkit.org/show_bug.cgi?id=131967
1539
1540         Unreviewed build fix after r167679.
1541
1542         * platform/graphics/mac/SimpleFontDataMac.mm:
1543         (WebCore::hasCustomTracking):
1544
1545 2014-04-21  Myles C. Maxfield  <mmaxfield@apple.com>
1546
1547         [OS X] Glyph spacing for system fonts may be incorrect
1548         https://bugs.webkit.org/show_bug.cgi?id=131967
1549
1550         Reviewed by Simon Fraser.
1551
1552         Covered by existing tests.
1553
1554         * platform/graphics/mac/SimpleFontDataMac.mm:
1555         (WebCore::SimpleFontData::platformWidthForGlyph): Update to use CTFontGetAdvancesForGlyphs() for system fonts
1556         (WebCore::hasCustomTracking):
1557
1558 2014-04-22  David Hyatt  <hyatt@apple.com>
1559
1560         [New Multicolumn] widows/orphans cause assertion failures.
1561         https://bugs.webkit.org/show_bug.cgi?id=131233
1562
1563         Reviewed by Dean Jackson.
1564
1565         * rendering/RenderMultiColumnSet.cpp:
1566         (WebCore::RenderMultiColumnSet::recordSpaceShortage):
1567         Fix an obvious bug where the space shortage is recorded twice (and the
1568         if statement that kept it from being negative is accidentally ignored).
1569
1570 2014-04-22  David Hyatt  <hyatt@apple.com>
1571
1572         [New Multicolumn] Make sure columnTranslationForOffset has the same column-span-aware
1573         translation that fragment collection does.
1574         https://bugs.webkit.org/show_bug.cgi?id=131738
1575
1576         Reviewed by Anders Carlsson.
1577
1578         * rendering/RenderMultiColumnSet.cpp:
1579         (WebCore::RenderMultiColumnSet::collectLayerFragments):
1580         Remove a FIXME that is no longer needed, since the translation offset of the
1581         column set did get patched.
1582
1583         (WebCore::RenderMultiColumnSet::columnTranslationForOffset):
1584         Add the exact same fix to columnTranslationForOffset that was applied to collectLayerFragments.
1585
1586 2014-04-22  Mark Lam  <mark.lam@apple.com>
1587
1588         WebCore::HTMLMediaElement::ensureMediaControlsInjectedScript() needs to acquire the JSLock before calling into JS.
1589         <https://webkit.org/b/132021>
1590
1591         Reviewed by Mark Hahnenberg.
1592
1593         Covered by existing layout test.
1594
1595         * html/HTMLMediaElement.cpp:
1596         (WebCore::HTMLMediaElement::parseAttribute):
1597
1598 2014-04-22  Manuel Rego Casasnovas  <rego@igalia.com>
1599
1600         REGRESSION (r167652): Broke fast/regions/cssom/region-range-for-box-crash.html in debug mode
1601         https://bugs.webkit.org/show_bug.cgi?id=131982
1602
1603         Reviewed by David Hyatt.
1604
1605         The problem was creating the Range from the arguments received at RenderView::setSelection(). Specifically
1606         in this test endPos is 1 when the element has not children which creates an invalid Range.
1607
1608         * rendering/RenderView.cpp:
1609         (WebCore::RenderView::splitSelectionBetweenSubtrees): Pass 0 as startPos and endPos to Range::create() as we are
1610         not interested in the positions at this point.
1611
1612 2014-04-22  Brady Eidson  <beidson@apple.com>
1613
1614         Change Image Controls replacement to use selection and paste
1615         <rdar://problem/16302722> and https://bugs.webkit.org/show_bug.cgi?id=131992
1616
1617         Reviewed by Tim Horton.
1618
1619         * WebCore.exp.in: Remove deleted symbol.
1620
1621         * html/shadow/mac/ImageControlsButtonElementMac.cpp:
1622         (WebCore::ImageControlsButtonElementMac::defaultEventHandler):
1623
1624         * page/ContextMenuController.cpp:
1625         (WebCore::ContextMenuController::showImageControlsMenu): Select the image element
1626           before showing the menu.
1627         (WebCore::ContextMenuController::replaceControlledImage): Deleted.
1628         * page/ContextMenuController.h:
1629
1630 2014-04-22  David Hyatt  <hyatt@apple.com>
1631
1632         REGRESSION: Hitting asserts in new flow thread selection code.
1633         https://bugs.webkit.org/show_bug.cgi?id=132019
1634
1635         Reviewed by Anders Carlsson.
1636
1637         Null checking flowThreadController is not enough. The list of threads
1638         can be null even when the flowThreadController is not. That's why you
1639         use hasRenderNamedFlowThreads. I patched the two bad places in the code
1640         to use hasRenderNamedFlowThreads to stop the crashes.
1641
1642         * rendering/RenderView.cpp:
1643         (WebCore::RenderView::selectionBounds):
1644         (WebCore::RenderView::repaintSelection):
1645
1646 2014-04-22  Brent Fulgham  <bfulgham@apple.com>
1647
1648         Check (rather than assume) element is a RenderTableSection before using it
1649         https://bugs.webkit.org/show_bug.cgi?id=121858
1650
1651         Reviewed by David Kilzer.
1652
1653         Test: fast/table/table-insert-object-before-td-crash.html
1654
1655         * rendering/RenderTable.cpp:
1656         (WebCore::RenderTable::addChild): Check that lastBox is a table section before
1657         using the node as if it was one.
1658
1659 2014-04-22  Alex Christensen  <achristensen@webkit.org>
1660
1661         Begin implementation of video using Media Foundation.
1662         https://bugs.webkit.org/show_bug.cgi?id=131830
1663
1664         Reviewed by Brent Fulgham.
1665
1666         * WebCore.vcxproj/WebCore.vcxproj:
1667         * WebCore.vcxproj/WebCore.vcxproj.filters:
1668         Added MediaPlayerPrivateMediaFoundation files.
1669         * platform/graphics/MediaPlayer.cpp:
1670         Include MediaPlayerPrivateMediaFoundation inside new USE(MEDIA_FOUNDATION) flag.
1671         * platform/graphics/win/MediaPlayerPrivateMediaFoundation.cpp: Added.
1672         * platform/graphics/win/MediaPlayerPrivateMediaFoundation.h: Added.
1673
1674 2014-04-22  Commit Queue  <commit-queue@webkit.org>
1675
1676         Unreviewed, rolling out r167658.
1677         https://bugs.webkit.org/show_bug.cgi?id=131993
1678
1679         Caused many crashes on media tests (Requested by ap on
1680         #webkit).
1681
1682         Reverted changeset:
1683
1684         "[Mac] don't ask for AVAssetTrack properties before they are
1685         available"
1686         https://bugs.webkit.org/show_bug.cgi?id=131902
1687         http://trac.webkit.org/changeset/167658
1688
1689 2014-04-22  David Hyatt  <hyatt@apple.com>
1690
1691         [New Multicolumn] fast/multicol/multicol-with-child-renderLayer-for-input.html puts the textfield in the wrong place
1692         https://bugs.webkit.org/show_bug.cgi?id=131808
1693
1694         Reviewed by Brady Eidson.
1695
1696         Make sure FrameView is patched to pass in the update pagination flag when subtree layout
1697         is done on textfields. It had code for the old multicolumn system but did not have a check
1698         for the new multicolumn system.
1699
1700         Added fast/multicol/newmulticol/compare-with-old-impl/multicol-with-child-renderLayer-for-input.html
1701
1702         * page/FrameView.cpp:
1703         (WebCore::updateLayerPositionFlags):
1704
1705 2014-04-22  Zalan Bujtas  <zalan@apple.com>
1706
1707         border-style: double rendered as solid when combined with border-radius
1708         https://bugs.webkit.org/show_bug.cgi?id=131927
1709
1710         Reviewed by Simon Fraser.
1711
1712         BorderEdge::getDoubleBorderStripeWidths() should return the calculated inner, outer border widths.
1713
1714         Test: fast/borders/hidpi-double-border-with-border-radius-always-produce-solid-line.html
1715
1716         * rendering/BorderEdge.cpp:
1717         (WebCore::BorderEdge::getDoubleBorderStripeWidths):
1718         * rendering/BorderEdge.h:
1719
1720 2014-04-22  Eric Carlson  <eric.carlson@apple.com>
1721
1722         [Mac] don't ask for AVAssetTrack properties before they are available
1723         https://bugs.webkit.org/show_bug.cgi?id=131902
1724         <rdar://problem/16505076>
1725
1726         Reviewed by Jer Noble.
1727
1728         No new tests, the behavior this changes can not be tested with a layout test.
1729
1730         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
1731         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
1732         (WebCore::MediaPlayerPrivateAVFoundationObjC::MediaPlayerPrivateAVFoundationObjC): Initialize
1733             m_cachedTotalBytes.
1734         (WebCore::MediaPlayerPrivateAVFoundationObjC::beginLoadingMetadata): Don't report that
1735             metadata has been loaded until the track properties we need have been loaded too.
1736         (WebCore::MediaPlayerPrivateAVFoundationObjC::totalBytes): Cache totalBytes instead
1737             of recalculating it every time.
1738         (WebCore::MediaPlayerPrivateAVFoundationObjC::tracksDidChange): Invalidate cached
1739             total bytes.
1740         (WebCore::assetTrackMetadataKeyNames): Array of AVAssetTrack properties we use.
1741
1742 2014-04-22  Peter Molnar  <pmolnar.u-szeged@partner.samsung.com>
1743
1744         Buildfix after r167650 broke non-Cocoa builds
1745         https://bugs.webkit.org/show_bug.cgi?id=131980
1746
1747         Reviewed by Mario Sanchez Prada.
1748
1749         * page/EventHandler.cpp:
1750         (WebCore::EventHandler::handleWheelEvent):
1751
1752 2014-04-22  Manuel Rego Casasnovas  <rego@igalia.com>
1753
1754         [CSS Regions] Selection highlight doesn't match DOM selection
1755         https://bugs.webkit.org/show_bug.cgi?id=131511
1756
1757         Reviewed by David Hyatt.
1758
1759         This patch implements a subtrees approach to fix selection interaction in CSS Regions.
1760
1761         The idea is to divide the Render tree in subtrees. One subtree will be the RenderView (which is always present),
1762         then for each RenderNamedFlowThread (which are children of RenderView) we will have an extra subtree.
1763
1764         For each subtree it determines the start and end positions according to the DOM tree. So, every subtree will
1765         always have a start and end position, and the selection algorithm will be able to reach the end from the start
1766         without problems (this allows us to remove the changes introduced in r155058).
1767
1768         Then selection related methods in RenderView are adapted to ensure that they are called for each subtree. Making
1769         the selection highlight to match DOM selection as expected.
1770
1771         Test: fast/regions/selection/selecting-text-including-region.html
1772
1773         * CMakeLists.txt: Add new class SelectionSubtreeRoot to build file.
1774         * WebCore.vcxproj/WebCore.vcxproj: Ditto.
1775         * WebCore.vcxproj/WebCore.vcxproj.filters: Ditto.
1776         * WebCore.xcodeproj/project.pbxproj: Ditto.
1777         * rendering/RenderNamedFlowThread.h: Inherit from SelectionSubtreeRoot.
1778         * rendering/RenderObject.cpp:
1779         (WebCore::RenderObject::selectionRoot): New method to return the RenderNamedFlowThread if any or the RenderView
1780         otherwise.
1781         (WebCore::RenderObject::selectionStartEnd): If it's inside a RenderNamedFlowThread use specific method instead
1782         of the RenderView's one.
1783         * rendering/RenderObject.h: Add new method signature.
1784         * rendering/RenderView.cpp:
1785         (WebCore::RenderView::selectionBounds): Adapt method to subtrees approach.
1786         (WebCore::RenderView::subtreeSelectionBounds): Modify previous method in order to use SelectionSubtreeRoot
1787         objects.
1788         (WebCore::RenderView::repaintSelection): Adapt method to subtrees approach.
1789         (WebCore::RenderView::repaintSubtreeSelection): Modify previous method in order to use SelectionSubtreeRoot
1790         objects.
1791         (getNextOrPrevRenderObjectBasedOnDirection): Deleted method from r155058 as it is not needed anymore. Each
1792         subtree will have start and end positions, so the selection algorithm will always find the end.
1793         (WebCore::RenderView::setSelection): Adapt method to subtrees approach.
1794         (WebCore::RenderView::splitSelectionBetweenSubtrees): Determine start and end positions for each subtree and
1795         call setSubtreeSelection().
1796         (WebCore::RenderView::setSubtreeSelection): Modify previous method in order to use SelectionSubtreeRoot objects.
1797         * rendering/RenderView.h: Inherit from SelectionSubtreeRoot. Add signature for new helper methods receiving
1798         SelectionSubtreeRoot objects.
1799         * rendering/SelectionSubtreeRoot.cpp: Added.
1800         (WebCore::SelectionSubtreeRoot::SelectionSubtreeRoot):
1801         (WebCore::SelectionSubtreeRoot::selectionStartEndPositions):
1802         * rendering/SelectionSubtreeRoot.h: Added. New class to identify root elements for the subtrees. With some basic
1803         attributes and methods to store the start and end positions and offsets.
1804         (WebCore::SelectionSubtreeRoot::selectionStart):
1805         (WebCore::SelectionSubtreeRoot::selectionStartPos):
1806         (WebCore::SelectionSubtreeRoot::selectionEnd):
1807         (WebCore::SelectionSubtreeRoot::selectionEndPos):
1808         (WebCore::SelectionSubtreeRoot::setSelectionStart):
1809         (WebCore::SelectionSubtreeRoot::setSelectionStartPos):
1810         (WebCore::SelectionSubtreeRoot::setSelectionEnd):
1811         (WebCore::SelectionSubtreeRoot::setSelectionEndPos):
1812
1813 2014-04-21  Brent Fulgham  <bfulgham@apple.com>
1814
1815         [Mac] Follow-up: After a horizontal overflow scroll with a mechanical wheel mouse, vertical scrolling no longer works
1816         https://bugs.webkit.org/show_bug.cgi?id=131960
1817         <rdar://problem/16142100>
1818
1819         Unreviewed build fix.
1820
1821         * page/EventHandler.cpp:
1822         (WebCore::EventHandler::handleWheelEvent): The test to clear the
1823         last scrolled node was too broad. It should only follow this path
1824         if both the phase and momentum phase are 'None'.
1825
1826 2014-04-21  Brent Fulgham  <bfulgham@apple.com>
1827
1828         [Mac] After a horizontal overflow scroll with a mechanical wheel mouse, vertical scrolling no longer works
1829         https://bugs.webkit.org/show_bug.cgi?id=131960
1830         <rdar://problem/16142100>
1831
1832         Reviewed by Darin Adler.
1833
1834         Test: platform/mac/fast/scrolling/scroll-div-no-latching.html
1835
1836         * page/EventHandler.cpp:
1837         (WebCore::EventHandler::handleWheelEvent): When not latching, always clear m_latchedWheelEventElement and m_previousWheelScrolledElement.
1838
1839 2014-04-21  Joseph Pecoraro  <pecoraro@apple.com>
1840
1841         WebProfilerAgent::disable is calling the wrong superclass method
1842         https://bugs.webkit.org/show_bug.cgi?id=131968
1843
1844         Reviewed by Simon Fraser.
1845
1846         * inspector/WebProfilerAgent.cpp:
1847         (WebCore::WebProfilerAgent::disable):
1848
1849 2014-04-21  Enrica Casucci  <enrica@apple.com>
1850
1851         Followup from http://trac.webkit.org/changeset/167624
1852
1853         Reviewed by Darin Adler.
1854
1855         Fixing incorrect use of ? : syntax.
1856
1857         * editing/VisibleUnits.cpp:
1858         (WebCore::characterBeforePosition):
1859
1860 2014-04-21  Ryuan Choi  <ryuan.choi@samsung.com>
1861
1862         Move CoordinatedGraphics related codes out of CMakeLists.txt
1863         https://bugs.webkit.org/show_bug.cgi?id=131931
1864
1865         Reviewed by Gyuyoung Kim.
1866
1867         Move CoordinatedGraphics related files to PlatformEfl.cmake because they are not common files.
1868
1869         * CMakeLists.txt:
1870         * PlatformEfl.cmake:
1871
1872 2014-04-21  Jer Noble  <jer.noble@apple.com>
1873
1874         [MSE][Mac] AVSampleBufferDisplayLayer and AudioRenderer is overretained.
1875         https://bugs.webkit.org/show_bug.cgi?id=131963
1876
1877         Reviewed by Jon Honeycutt.
1878
1879         Adopt the results of -[[class alloc] init] rather than retaining them.
1880
1881         * platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:
1882         (WebCore::SourceBufferPrivateAVFObjC::hasVideo):
1883         (WebCore::SourceBufferPrivateAVFObjC::trackDidChangeEnabled):
1884
1885 2014-04-21  Zalan Bujtas  <zalan@apple.com>
1886
1887         REGRESSION (r166784): Gradient at background of iCloud login page doesn’t go all the way to the bottom
1888         https://bugs.webkit.org/show_bug.cgi?id=131924
1889
1890         Computing tile dimension for contain/cover requires higher precision than what LayoutUnit has. Switching to floats.
1891
1892         Reviewed by Simon Fraser.
1893
1894         Test: fast/backgrounds/hidpi-background-image-contain-cover-scale-needs-more-precision.html
1895
1896         * rendering/RenderBoxModelObject.cpp:
1897         (WebCore::RenderBoxModelObject::calculateFillTileSize):
1898
1899 2014-04-21  Dean Jackson  <dino@apple.com>
1900
1901         Build fix for Mountain Lion.
1902
1903         * platform/mac/SerializedPlatformRepresentationMac.mm:
1904         (WebCore::SerializedPlatformRepresentationMac::deserialize):
1905
1906 2014-04-21  Dan Bernstein  <mitz@apple.com>
1907
1908         WebCore part of <rdar://problem/16631323> Page::setDefersLoading doesn’t do anything when using the Network process
1909         https://bugs.webkit.org/show_bug.cgi?id=131939
1910
1911         Reviewed by David Kilzer.
1912
1913         * WebCore.exp.in: Exported WebCore::ResourceHandle::setDefersLoading.
1914
1915         * loader/ResourceLoadScheduler.cpp:
1916         (WebCore::ResourceLoadScheduler::setDefersLoading): Added a no-op implementation.
1917         * loader/ResourceLoadScheduler.h: Declared new virtual member function setDefersLoading.
1918
1919         * loader/ResourceLoader.cpp:
1920         (WebCore::ResourceLoader::setDefersLoading): Added a call to the above function.
1921
1922 2014-04-21  Brent Fulgham  <bfulgham@apple.com>
1923
1924         [Mac] Difficulty gesture scrolling vertically with trackpad after scrolling horizontally
1925         https://bugs.webkit.org/show_bug.cgi?id=131959
1926         <rdar://problem/16654523>
1927
1928         Reviewed by Simon Fraser.
1929
1930         * page/mac/EventHandlerMac.mm:
1931         (WebCore::deltaIsPredominantlyVertical): Added.
1932         (WebCore::scrolledToEdgeInDominantDirection): Only consider current mouse wheel event. We don't care about
1933         overall history when deciding if we are bumping against the edge of a scrollable region. Short-circuit if
1934         the element style indicates that overflow is hidden, since this means there is no scroll possible in that
1935         direction.
1936         (WebCore::EventHandler::platformPrepareForWheelEvents): Update for new signature.
1937
1938 2014-04-21  Eric Carlson  <eric.carlson@apple.com>
1939
1940         [Mac] implement WebKitDataCue
1941         https://bugs.webkit.org/show_bug.cgi?id=131799
1942
1943         Reviewed by Dean Jackson.
1944
1945         Tests: http/tests/media/track-in-band-hls-metadata.html
1946                media/track/track-datacue-value.html
1947
1948         * Configurations/FeatureDefines.xcconfig: Define ENABLE_DATACUE_VALUE.
1949         * DerivedSources.make: Add ENABLE_DATACUE_VALUE to HTML_FLAGS when appropriate.
1950
1951         * CMakeLists.txt: Add JSDataCueCustom.cpp.
1952         * bindings/js/JSBindingsAllInOne.cpp:
1953         * WebCore.xcodeproj/project.pbxproj: Add new files.
1954
1955         * bindings/js/JSDataCueCustom.cpp: Added.
1956         (WebCore::JSDataCue::value):
1957         (WebCore::JSDataCue::setValue):
1958         (WebCore::JSDataCueConstructor::constructJSDataCue): Custom constructor.
1959
1960         * html/HTMLMediaElement.cpp:
1961         (WebCore::HTMLMediaElement::updateActiveTextTrackCues): Drive-by fixes: don't schedule timeupdate
1962             events when paused, don't call sort on an Vector that can't be sorted.
1963
1964         * html/track/DataCue.cpp:
1965         (WebCore::DataCue::DataCue): Initialize m_type.
1966         (WebCore::DataCue::~DataCue): Unprotect the JSValue if necessary.
1967         (WebCore::DataCue::data): Ask the platform value for data if non-null.
1968         (WebCore::DataCue::setData): Clear m_platformValue and m_value.
1969         (WebCore::DataCue::isEqual): New.
1970         (WebCore::DataCue::value): Return a JSValue from the platform value, or the value passed
1971             to the constructor/set by script.
1972         (WebCore::DataCue::setValue): Set m_value.
1973         * html/track/DataCue.h:
1974         * html/track/DataCue.idl:
1975
1976         * html/track/InbandDataTextTrack.cpp:
1977         (WebCore::InbandDataTextTrack::addDataCue): Don't add the same cue more than once.
1978         (WebCore::InbandDataTextTrack::updateDataCue): Update a cue's duration.
1979         (WebCore::InbandDataTextTrack::removeDataCue): Remove an incomplete cue.
1980         (WebCore::InbandDataTextTrack::removeCue): Remove a cue from the incomplete cue map if necessary.
1981         * html/track/InbandDataTextTrack.h:
1982
1983         * html/track/InbandGenericTextTrack.cpp:
1984         (WebCore::InbandGenericTextTrack::addGenericCue): CueMatchRules is now in TextTrackCue instead
1985             of VTTCue.
1986         * html/track/InbandTextTrack.h:
1987
1988         * html/track/InbandWebVTTTextTrack.cpp:
1989         (WebCore::InbandWebVTTTextTrack::newCuesParsed): Ditto.
1990
1991         * html/track/TextTrack.cpp:
1992         (WebCore::TextTrack::hasCue): Ditto.
1993         * html/track/TextTrack.h:
1994
1995         * html/track/TextTrackCue.cpp:
1996         (WebCore::TextTrackCue::isEqual): New, test base class equality.
1997         * html/track/TextTrackCue.h:
1998
1999         * html/track/TextTrackCueGeneric.cpp:
2000         (WebCore::TextTrackCueGeneric::isEqual): Call TextTrackCue::isEqual first.
2001         * html/track/TextTrackCueGeneric.h:
2002
2003         * html/track/VTTCue.cpp:
2004         (WebCore::VTTCue::isEqual): Call TextTrackCue::isEqual first.
2005         * html/track/VTTCue.h:
2006
2007         * platform/SerializedPlatformRepresentation.h: Added.
2008         (WebCore::SerializedPlatformRepresentation::~SerializedPlatformRepresentation):
2009         (WebCore::SerializedPlatformRepresentation::SerializedPlatformRepresentation):
2010
2011         * platform/graphics/InbandTextTrackPrivateClient.h: Add methods for DataCue with SerializedPlatformRepresentation.
2012
2013         * platform/graphics/avfoundation/InbandMetadataTextTrackPrivateAVF.cpp: Added.
2014         (WebCore::InbandMetadataTextTrackPrivateAVF::create):
2015         (WebCore::InbandMetadataTextTrackPrivateAVF::InbandMetadataTextTrackPrivateAVF):
2016         (WebCore::InbandMetadataTextTrackPrivateAVF::~InbandMetadataTextTrackPrivateAVF):
2017         (WebCore::InbandMetadataTextTrackPrivateAVF::addDataCue):
2018         (WebCore::InbandMetadataTextTrackPrivateAVF::updatePendingCueEndTimes):
2019         (WebCore::InbandMetadataTextTrackPrivateAVF::flushPartialCues):
2020         * platform/graphics/avfoundation/InbandMetadataTextTrackPrivateAVF.h: Added.
2021
2022         * platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.cpp:
2023         (WebCore::MediaPlayerPrivateAVFoundation::seekWithTolerance): currentTrack -> currentTextTrack.
2024         (WebCore::MediaPlayerPrivateAVFoundation::seekCompleted): Ditto.
2025         (WebCore::MediaPlayerPrivateAVFoundation::configureInbandTracks): Ditto.
2026         * platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.h:
2027         (WebCore::MediaPlayerPrivateAVFoundation::setCurrentTextTrack):
2028         (WebCore::MediaPlayerPrivateAVFoundation::setCurrentTrack): Deleted.
2029
2030         Renamed currentTrack and setCurrentTrack to currentTextTrack and setCurrentTextTrack.
2031         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
2032         (WebCore::AVFWrapper::currentTextTrack):
2033         (WebCore::MediaPlayerPrivateAVFoundationCF::setCurrentTextTrack):
2034         (WebCore::MediaPlayerPrivateAVFoundationCF::currentTextTrack):
2035         (WebCore::AVFWrapper::setCurrentTextTrack):
2036         (WebCore::AVFWrapper::AVFWrapper):
2037         (WebCore::AVFWrapper::processCue):
2038         (WebCore::AVFWrapper::currentTrack): Deleted.
2039         (WebCore::MediaPlayerPrivateAVFoundationCF::setCurrentTrack): Deleted.
2040         (WebCore::MediaPlayerPrivateAVFoundationCF::currentTrack): Deleted.
2041         (WebCore::AVFWrapper::setCurrentTrack): Deleted.
2042         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.h:
2043
2044         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
2045         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
2046         (WebCore::MediaPlayerPrivateAVFoundationObjC::MediaPlayerPrivateAVFoundationObjC):
2047         (WebCore::MediaPlayerPrivateAVFoundationObjC::seekToTime): currentTrack -> currentTextTrack
2048         (WebCore::MediaPlayerPrivateAVFoundationObjC::tracksChanged): currentTrack -> currentTextTrack.
2049         (WebCore::MediaPlayerPrivateAVFoundationObjC::processMetadataTrack): New.
2050         (WebCore::MediaPlayerPrivateAVFoundationObjC::processCue): m_currentTrack -> m_currentTextTrack.
2051         (WebCore::MediaPlayerPrivateAVFoundationObjC::flushCues): Ditto.
2052         (WebCore::MediaPlayerPrivateAVFoundationObjC::setCurrentTextTrack): Renamed from setCurrentTextTrack.
2053         (WebCore::metadataType): Map an AVFoundation metadata key space to a metadata cue type.
2054         (WebCore::MediaPlayerPrivateAVFoundationObjC::metadataDidArrive): Process new metadata.
2055         (-[WebCoreAVFMovieObserver observeValueForKeyPath:ofObject:change:context:]):
2056         (WebCore::MediaPlayerPrivateAVFoundationObjC::setCurrentTrack): Deleted.
2057
2058         Create a JSValue representation from an AVMetadataItem.
2059         * platform/mac/SerializedPlatformRepresentationMac.h: Added.
2060         (WebCore::SerializedPlatformRepresentationMac::platformType):
2061         (WebCore::SerializedPlatformRepresentationMac::nativeValue):
2062         * platform/mac/SerializedPlatformRepresentationMac.mm: Added.
2063         (WebCore::SerializedPlatformRepresentationMac::SerializedPlatformRepresentationMac):
2064         (WebCore::SerializedPlatformRepresentationMac::~SerializedPlatformRepresentationMac):
2065         (WebCore::SerializedPlatformRepresentationMac::create):
2066         (WebCore::SerializedPlatformRepresentationMac::data):
2067         (WebCore::SerializedPlatformRepresentationMac::deserialize):
2068         (WebCore::SerializedPlatformRepresentationMac::isEqual):
2069         (WebCore::toSerializedPlatformRepresentationMac):
2070         (WebCore::jsValueWithValueInContext):
2071         (WebCore::jsValueWithDataInContext):
2072         (WebCore::jsValueWithArrayInContext):
2073         (WebCore::jsValueWithDictionaryInContext):
2074         (WebCore::jsValueWithAVMetadataItemInContext):
2075
2076 2014-04-21  Benjamin Poulain  <benjamin@webkit.org>
2077
2078         Add Element.matches, the standard name for webkitMatchesSelector
2079         https://bugs.webkit.org/show_bug.cgi?id=131922
2080
2081         Reviewed by Andreas Kling.
2082
2083         This patch just renames webkitMatchesSelector() to matches().
2084
2085         * bindings/scripts/CodeGeneratorObjC.pm:
2086         (GenerateImplementation):
2087         The Objective-C binding generator only supported ImplementedAs for getters and setters.
2088         This adds support for the option so that Element.matches can be generated for Obj-C.
2089
2090         * dom/Element.cpp:
2091         (WebCore::Element::matches):
2092         (WebCore::Element::webkitMatchesSelector): Deleted.
2093         * dom/Element.h:
2094         * dom/Element.idl:
2095         * inspector/InspectorCSSAgent.cpp:
2096         (WebCore::InspectorCSSAgent::buildArrayForMatchedRuleList):
2097
2098 2014-04-21  Beth Dakin  <bdakin@apple.com>
2099
2100         topContentInset does not play well with fullscreen elements
2101         https://bugs.webkit.org/show_bug.cgi?id=131955
2102         -and corresponding-
2103         <rdar://problem/16651925>
2104
2105         Reviewed by Sam Weinig.
2106
2107         Calling setNeedsLayout() is not sufficient when the topContentInset has changed
2108         dynamically. We need to perform the layout right away and update the scrollbars.
2109         This works completely when the inset changes due to entering/exiting fullscreen,
2110         but I left a FIXME behind because it is possible to change the inset dynamically
2111         just through the WK2 API and end up in a situation where everything looks right
2112         except for the scrollbars.
2113         * page/FrameView.cpp:
2114         (WebCore::FrameView::topContentInsetDidChange):
2115         * page/FrameView.h:
2116         * page/Page.cpp:
2117         (WebCore::Page::setTopContentInset):
2118
2119 2014-04-18  Jon Honeycutt  <jhoneycutt@apple.com>
2120
2121         Empty RenderInline objects should not be line break objects.
2122
2123         https://bugs.webkit.org/show_bug.cgi?id=131861
2124         <rdar://problem/15663617>
2125
2126         Reviewed by David Hyatt.
2127
2128         * rendering/line/BreakingContextInlineHeaders.h:
2129         (WebCore::BreakingContext::canBreakAtThisPosition):
2130         Don't break at this position if this is an empty RenderInline.
2131
2132 2014-04-21  Enrica Casucci  <enrica@apple.com>
2133
2134         [iOS WebKit2] support replacements for misspelled words.
2135         https://bugs.webkit.org/show_bug.cgi?id=131827
2136         <rdar://problem/16319657>
2137
2138         Reviewed by Darin Adler.
2139
2140         Adds utility function to return the character before the current selection
2141         and converts the nbsp to the space character.
2142
2143         * WebCore.exp.in:
2144         * editing/VisibleUnits.cpp:
2145         (WebCore::characterBeforePosition):
2146         * editing/VisibleUnits.h:
2147
2148 2014-04-21  Eric Carlson  <eric.carlson@apple.com>
2149
2150         [iOS] enable plug-in replacement
2151         https://bugs.webkit.org/show_bug.cgi?id=131947
2152
2153         Reviewed by Jer Noble.
2154
2155         * bindings/generic/RuntimeEnabledFeatures.cpp:
2156         (WebCore::RuntimeEnabledFeatures::RuntimeEnabledFeatures): initialize m_isPluginReplacementEnabled
2157             to true on iOS.
2158
2159 2014-04-21  Rik Cabanier  <cabanier@adobe.com>
2160
2161         [CSS Blending] Remove support for non-separable blend modes from mix-blend-mode
2162         https://bugs.webkit.org/show_bug.cgi?id=131824
2163
2164         Reviewed by Dean Jackson.
2165
2166         Removed support for non-separable blend modes from the CSS parser.
2167
2168         Covered by existing tests
2169
2170         * css/CSSParser.cpp:
2171         (WebCore::isValidKeywordPropertyAndValue):
2172         * platform/graphics/ca/mac/PlatformCAFiltersMac.mm:
2173         (PlatformCAFilters::setBlendingFiltersOnLayer):
2174
2175 2014-04-21  David Hyatt  <hyatt@apple.com>
2176
2177         [New Multicolumn] Column set drawing under horizontal scrollbar.
2178         https://bugs.webkit.org/show_bug.cgi?id=131812.
2179
2180         Reviewed by Sam Weinig.
2181
2182         Added fast/multicol/newmulticol/compare-with-old-impl/LeftToRight-tb.html.
2183
2184         * rendering/RenderBlockFlow.cpp:
2185         (WebCore::RenderBlockFlow::layoutBlock):
2186         (WebCore::RenderBlockFlow::checkForPaginationLogicalHeightChange):
2187         * rendering/RenderBlockFlow.h:
2188         Pass in relayoutChildren as a reference boolean to checkForPaginationLogicalHeightChange.
2189         The new columns need to be able to set it to true if the amount of available column
2190         height ever changes, so that the RenderFlowThread and the sets get a layout call.
2191
2192         * rendering/style/RenderStyle.cpp:
2193         (WebCore::RenderStyle::setColumnStylesFromPaginationMode):
2194         Make sure that the pagination mode is set up with an auto column-fill specified. You
2195         don't ever want to balance when you set a pagination mode on the RenderView.
2196
2197 2014-04-21  Tim Horton  <timothy_horton@apple.com>
2198
2199         [iOS WebKit2] Implement CSS viewport units
2200         https://bugs.webkit.org/show_bug.cgi?id=131863
2201         <rdar://problem/16279088>
2202
2203         Reviewed by Darin Adler.
2204
2205         * WebCore.exp.in:
2206         * page/FrameView.cpp:
2207         (WebCore::FrameView::setViewportSize):
2208         * page/FrameView.h:
2209         (WebCore::FrameView::viewportSize):
2210         Add setViewportSize() and viewportSize().
2211         If the viewport size changes, we have to do style recalc.
2212         If we have an overridden viewport size, use it.
2213
2214         * rendering/RenderView.cpp:
2215         (WebCore::RenderView::viewportSize):
2216         * rendering/RenderView.h:
2217         Get the size from FrameView.
2218
2219 2014-04-21  Joseph Pecoraro  <pecoraro@apple.com>
2220
2221         Remove unused WebKitSystemInterface functions
2222         https://bugs.webkit.org/show_bug.cgi?id=131938
2223
2224         Reviewed by Alexey Proskuryakov.
2225
2226         * WebCore.exp.in:
2227         * platform/ios/WebCoreSystemInterfaceIOS.mm:
2228         * platform/mac/WebCoreSystemInterface.h:
2229         * platform/mac/WebCoreSystemInterface.mm:
2230
2231 2014-04-21  Andreas Kling  <akling@apple.com>
2232
2233         Move the JSString cache from DOMWrapperWorld to VM.
2234         <https://webkit.org/b/131940>
2235
2236         Since there's no need for JSStrings to be world-specific, this patch
2237         moves the string cache to JSC::VM. This makes jsStringWithCache()
2238         a lot faster since it no longer has to jump through twenty-eleven
2239         hoops to find the DOMWrapperWorld.
2240
2241         Reviewed by Geoff Garen.
2242
2243         * bindings/js/DOMWrapperWorld.cpp:
2244         (WebCore::DOMWrapperWorld::clearWrappers):
2245         * bindings/js/DOMWrapperWorld.h:
2246         * bindings/js/JSDOMBinding.cpp:
2247         (WebCore::jsStringWithCache):
2248
2249 2014-04-21  David Hyatt  <hyatt@apple.com>
2250
2251         [New Multicolumn] Column balancing is slow on float-multicol.html
2252         https://bugs.webkit.org/show_bug.cgi?id=131801
2253
2254         Reviewed by Enrica Casucci.
2255
2256         Added fast/multicol/tall-float.html
2257
2258         * rendering/RenderBlock.cpp:
2259         (WebCore::RenderBlock::expandsToEncloseOverhangingFloats):
2260         Make RenderFlowThreads expand to encompass floats. Also from Morten's patch for
2261         Blink (#361551).
2262
2263         * rendering/RenderBlockFlow.cpp:
2264         (WebCore::RenderBlockFlow::computeOverflow):
2265         Fix a bug where the new multi-column code was adding in overflow for floats
2266         when it should not. This was resulting in the creation of scrollbars when none
2267         should exist. This is my own addition.
2268
2269         (WebCore::RenderBlockFlow::addOverhangingFloats):
2270         Make RenderFlowThreads and RenderRegions ignore overhanging float additions. This
2271         comes from Morten's patch for Blink (#361551).
2272
2273 2014-04-21  Darin Adler  <darin@apple.com>
2274
2275         REGRESSION (r160908): Safari doesn't draw rotated images properly first time
2276         https://bugs.webkit.org/show_bug.cgi?id=131930
2277
2278         Reviewed by Anders Carlsson.
2279
2280         With render tree creation sufficiently lazy, we had to put some explicit
2281         calls to updateStyleIfNeeded into the ImageDocument class.
2282
2283         * dom/DOMImplementation.cpp:
2284         (WebCore::DOMImplementation::createDocument): Pass a Frame&.
2285
2286         * html/ImageDocument.cpp:
2287         (WebCore::ImageDocument::imageSize): Added this helper function. This updates
2288         style so we can use the renderer to compute an image size taking things like
2289         orientation into account, and also passes in the zoom factor.
2290         (WebCore::ImageDocument::updateDuringParsing): Moved the body of the
2291         ImageDocumentParser::appendBytes function into this function. Since it's a
2292         member of ImageDocument, it has better access to appropriate private members,
2293         and we don't have to expose as many functions to the public. Also changed to
2294         not use the frame so much. Also merged in the code that creates the document
2295         instead of having it be a side effect in the cachedImage helper function; all
2296         the other call sites were checking for null before calling cachedImage.
2297         (WebCore::ImageDocument::finishedParsing): Moved the body of the
2298         ImageDocumentParser::finish function here. Added a call to updateStyleIfNeeded
2299         before using the renderer to get the image size.
2300         (WebCore::ImageDocumentParser::document): Made this return a reference. It's
2301         just a convenience function used in the two functions below.
2302         (WebCore::ImageDocumentParser::appendBytes): Just call updateDuringParsing.
2303         (WebCore::ImageDocumentParser::finish): Just call finishedParsing.
2304         (WebCore::ImageDocument::ImageDocument): Take a reference for the argument,
2305         use nullptr, and also initialize m_shouldShrinkImage directly instead of using
2306         a shouldShrinkToFit() helper function to do it.
2307         (WebCore::ImageDocument::createDocumentStructure): Removed unneeded check of
2308         null for the frame, since the caller does it. Removed some unneeded
2309         IGNORE_EXCEPTON and a stray semicolon. Use m_shouldShrinkImage instead of
2310         calling shouldShrinkToFit since at this early point they will be identical.
2311         (WebCore::ImageDocument::scale): Make non-const so we can call the new
2312         imageSize helper. Use Document::view instead of Frame::view. Improved the
2313         conversion to floating point so it doesn't involve a static_cast and an extra
2314         conversion from int to LayoutUnit.
2315         (WebCore::ImageDocument::resizeImageToFit): Ditto.
2316         (WebCore::ImageDocument::imageClicked): Ditto.
2317         (WebCore::ImageDocument::imageUpdated): Ditto. Also use m_shouldShrinkImage
2318         ibnstead of shouldShrinkToFit.
2319         (WebCore::ImageDocument::restoreImageSize): Ditto.
2320         (WebCore::ImageDocument::imageFitsInWindow): Ditto.
2321         (WebCore::ImageDocument::windowSizeChanged): Ditto.
2322         (WebCore::ImageDocument::cachedImage): Deleted. Moved logic into updateDuringParsing.
2323         (WebCore::ImageDocument::shouldShrinkToFit): Deleted. Moved logic into constructor.
2324         (WebCore::ImageEventListener::handleEvent): Updated for changes above.
2325         (WebCore::ImageEventListener::operator==): Simplified. No reason this has to check
2326         whether both listeners are in the same document.
2327         (WebCore::ImageDocumentElement::didMoveToNewDocument): Use nullptr.
2328
2329         * html/ImageDocument.h: Deleted some unused things, updated for changes above,
2330         and made some memer functions private.
2331
2332         * loader/FrameLoaderClient.h: Deleted the allowScriptFromSource, allowPlugins,
2333         allowImage, and allowDisplayingInsecureContent functions. These were only used in
2334         Chromium and are now unneeded complexity. We could bring them back if anyone wants them.
2335
2336         * loader/MixedContentChecker.cpp:
2337         (WebCore::MixedContentChecker::canDisplayInsecureContent): Removed call to
2338         allowDisplayingInsecureContent, just use the setting directly.
2339         (WebCore::MixedContentChecker::canRunInsecureContent): Removed call to
2340         allowRunningInsecureContent, just use the setting directly.
2341
2342         * loader/SubframeLoader.cpp:
2343         (WebCore::SubframeLoader::allowPlugins): Removed call to allowPlugins, just
2344         use the arePluginsEnabled setting directly.
2345
2346         * loader/cache/CachedResourceLoader.cpp:
2347         (WebCore::CachedResourceLoader::canRequest): Removed call to allowScriptFromSource, just
2348         use the isScriptEnabled setting directly.
2349         (WebCore::CachedResourceLoader::clientDefersImage): Removed call to allowImage.
2350
2351 2014-04-21  David Hyatt  <hyatt@apple.com>
2352
2353         [New Multicolumn] Pagination mode messed up with non-inline axis and reversed direction.
2354         https://bugs.webkit.org/show_bug.cgi?id=131811
2355
2356         Reviewed by Dean Jackson.
2357
2358         Added fast/multicol/newmulticol/compare-with-old-impl/BottomToTop-tb.html
2359
2360         With block axis pagination mode, it is possible to set a column height that is not the same
2361         as the available fill height for a block. The new multi-column code had the assumption that
2362         the column height was the same as the amount of fill room you had available. This is not
2363         the case.
2364
2365         To correct the issue, I added a member variable to RenderMultiColumnSet that stores the
2366         available column height as a separate variable from the computed column height. This allows
2367         the pagination API to specify a different column height that is not the same as the view's
2368         content height.
2369
2370         Even though it isn't involved in the solution, I also patched pageOrViewLogicalHeight on
2371         RenderView to work with the new column code as well.
2372
2373         To address the layout test failures (that caused the previous rollout), I made sure to
2374         initialize m_availableHeight to 0 when m_computedColumnHeight also gets reset to 0.
2375
2376         The assertion is not something I could reproduce on any machine, but I can see the problem.
2377         I patched Page's pageCount method to not have column code directly in Page.cpp,
2378         and to make a new pageCount() method on RenderView that Page calls
2379         into. This method is now patched to handle the new column code as well as the old. I have
2380         no real way of testing this method though, since I can't reproduce the assertion that the
2381         bots were experiencing.
2382
2383         * page/Page.cpp:
2384         (WebCore::Page::pageCount):
2385         * rendering/RenderMultiColumnSet.cpp:
2386         (WebCore::RenderMultiColumnSet::RenderMultiColumnSet):
2387         (WebCore::RenderMultiColumnSet::setAndConstrainColumnHeight):
2388         (WebCore::RenderMultiColumnSet::prepareForLayout):
2389         (WebCore::RenderMultiColumnSet::computeLogicalHeight):
2390         * rendering/RenderMultiColumnSet.h:
2391         * rendering/RenderView.cpp:
2392         (WebCore::RenderView::pageOrViewLogicalHeight):
2393         (WebCore::RenderView::pageCount):
2394         * rendering/RenderView.h:
2395
2396 2014-04-18  Dean Jackson  <dino@apple.com>
2397
2398         [Media] Clean up localized strings in controls
2399         https://bugs.webkit.org/show_bug.cgi?id=131857
2400         <rdar://problem/16663005>
2401
2402         Reviewed by Eric Carlson.
2403
2404         There is no need for separate localization files
2405         on OS X and iOS. Move the small number of iOS-specific
2406         localizations into a shared file.
2407
2408         We also don't need a function to return a static table.
2409         Ultimately I think we want a top-level WebKit object that
2410         hold the table, but for now use a global variable
2411         called UIStringTable. Since we're in an isolated world, we
2412         won't clash with anything.
2413
2414         I noticed a few cases were we grab a localized string and
2415         then instantly run a replacement on it. This should be
2416         supported directly in the UIString() accessor.
2417
2418         * English.lproj/mediaControlsLocalizedStrings.js: Add iOS strings. Just
2419         use a static table.
2420         (mediaControlsLocalizedStrings): Deleted.
2421         * English.lproj/mediaControlsLocalizedStringsiOS.js: Removed.
2422         * Modules/mediacontrols/mediaControlsApple.js:
2423         (Controller.prototype.extend): Fix braces.
2424         (Controller.prototype.UIString): Use the string table, and allow replacement.
2425         (Controller.prototype.createControls): Use new UIString replacement.
2426         * Modules/mediacontrols/mediaControlsiOS.js:
2427         (ControllerIOS.prototype.updateWirelessPlaybackStatus): Use new UIString.
2428         (ControllerIOS.prototype.UIString): Deleted. Use the parent class version.
2429         * WebCore.xcodeproj/project.pbxproj: Remove file from Resources.
2430         * rendering/RenderThemeIOS.mm:
2431         (WebCore::RenderThemeIOS::mediaControlsScript): No need to include the iOS
2432         specific localization file.
2433
2434 2014-04-21  Brent Fulgham  <bfulgham@apple.com>
2435
2436         [Win] Unreviewed build fix.
2437
2438         * inspector/TimelineRecordFactory.cpp: Add missing include.
2439
2440 2014-04-21  Zan Dobersek  <zdobersek@igalia.com>
2441
2442         Move cross-port Source/WebCore/page/ code to std::unique_ptr
2443         https://bugs.webkit.org/show_bug.cgi?id=130534
2444
2445         Reviewed by Darin Adler.
2446
2447         Replace uses of OwnPtr and PassOwnPtr in cross-port code under Source/WebCore/page/ with
2448         std::unique_ptr. Unnecessary OwnPtr.h and PassOwnPtr.h header inclusions are removed as well.
2449
2450         * dom/Document.cpp:
2451         (WebCore::Document::initSecurityContext):
2452         * dom/SecurityContext.cpp:
2453         (WebCore::SecurityContext::setContentSecurityPolicy):
2454         * dom/SecurityContext.h:
2455         * page/CaptionUserPreferences.h:
2456         * page/ContentSecurityPolicy.cpp:
2457         (WebCore::CSPDirectiveList::create):
2458         (WebCore::CSPDirectiveList::checkEvalAndReportViolation):
2459         (WebCore::CSPDirectiveList::checkInlineAndReportViolation):
2460         (WebCore::CSPDirectiveList::checkSourceAndReportViolation):
2461         (WebCore::CSPDirectiveList::setCSPDirective):
2462         (WebCore::ContentSecurityPolicy::didReceiveHeader):
2463         * page/ContentSecurityPolicy.h:
2464         * page/DOMSecurityPolicy.h:
2465         * page/DOMWindow.cpp:
2466         (WebCore::PostMessageTimer::fired):
2467         (WebCore::DOMWindow::postMessageTimerFired):
2468         * page/DOMWindow.h:
2469         * page/EventHandler.cpp:
2470         (WebCore::EventHandler::EventHandler):
2471         * page/EventHandler.h:
2472         * page/Frame.cpp:
2473         (WebCore::Frame::Frame):
2474         (WebCore::createRegExpForLabels):
2475         (WebCore::Frame::searchForLabelsAboveCell):
2476         (WebCore::Frame::searchForLabelsBeforeElement):
2477         (WebCore::matchLabelsAgainstString):
2478         (WebCore::Frame::setTiledBackingStoreEnabled):
2479         * page/Frame.h:
2480         * page/FrameView.cpp:
2481         (WebCore::FrameView::addEmbeddedObjectToUpdate):
2482         (WebCore::FrameView::addSlowRepaintObject):
2483         (WebCore::FrameView::addViewportConstrainedObject):
2484         (WebCore::FrameView::addScrollableArea):
2485         * page/FrameView.h:
2486         * page/Page.cpp:
2487         (WebCore::Page::Page):
2488         * page/Page.h:
2489         * page/PageConsole.h:
2490         * page/PageThrottler.h:
2491         * page/SecurityPolicy.cpp:
2492         (WebCore::SecurityPolicy::addOriginAccessWhitelistEntry):
2493         * page/animation/AnimationController.h:
2494         * page/animation/CSSPropertyAnimation.cpp:
2495         (WebCore::PropertyWrapperVisitedAffectedColor::PropertyWrapperVisitedAffectedColor):
2496         (WebCore::CSSPropertyAnimationWrapperMap::instance):
2497         (WebCore::CSSPropertyAnimationWrapperMap::CSSPropertyAnimationWrapperMap):
2498         * page/mac/PageMac.cpp:
2499         (WebCore::Page::addSchedulePair):
2500         * workers/WorkerGlobalScope.cpp:
2501         (WebCore::WorkerGlobalScope::applyContentSecurityPolicyFromString):
2502
2503 2014-04-21  Darin Adler  <darin@apple.com>
2504
2505         Add HashSet::takeAny
2506         https://bugs.webkit.org/show_bug.cgi?id=131928
2507
2508         Reviewed by Benjamin Poulain.
2509
2510         * dom/Document.cpp:
2511         (WebCore::Document::takeAnyMediaCanStartListener): Use HashSet::takeAny.
2512         * dom/ScriptExecutionContext.cpp:
2513         (WebCore::takeAny): Deleted.
2514         (WebCore::ScriptExecutionContext::~ScriptExecutionContext): Use HashSet::takeAny.
2515
2516 2014-04-21  Zan Dobersek  <zdobersek@igalia.com>
2517
2518         iOS and non-iOS RemoteCommandListener::create() functions should use std::make_unique<>()
2519         https://bugs.webkit.org/show_bug.cgi?id=131916
2520
2521         Reviewed by Darin Adler.
2522
2523         A follow-up to r167197. Make RemoteCommandListener ctor public and use std::make_unique<>()
2524         in both versions of RemoteCommandListener::create() functions (iOS and non-iOS).
2525
2526         * platform/RemoteCommandListener.cpp:
2527         (WebCore::RemoteCommandListener::create):
2528         * platform/RemoteCommandListener.h:
2529         (WebCore::RemoteCommandListener::RemoteCommandListener):
2530         * platform/ios/RemoteCommandListenerIOS.mm:
2531         (WebCore::RemoteCommandListener::create):
2532
2533 2014-04-21  Andreas Kling  <akling@apple.com>
2534
2535         Micro-optimize the way we hand NodeLists to JSC.
2536         <https://webkit.org/b/131932>
2537
2538         Use HashMap::fastAdd() when returning cached node lists and collections.
2539         10.9% progression on Bindings/get-elements-by-tag-name.html
2540
2541         Reviewed by Antti Koivisto.
2542
2543         * dom/NodeRareData.h:
2544         (WebCore::NodeListsNodeData::addCacheWithAtomicName):
2545         (WebCore::NodeListsNodeData::addCacheWithName):
2546         (WebCore::NodeListsNodeData::addCacheWithQualifiedName):
2547         (WebCore::NodeListsNodeData::addCachedCollection):
2548
2549 2014-04-21  Commit Queue  <commit-queue@webkit.org>
2550
2551         Unreviewed, rolling out r167584.
2552         https://bugs.webkit.org/show_bug.cgi?id=131929
2553
2554         Broke Objective-C bindings test (Requested by ap on #webkit).
2555
2556         Reverted changeset:
2557
2558         "Add Element.matches, the standard name for
2559         webkitMatchesSelector"
2560         https://bugs.webkit.org/show_bug.cgi?id=131922
2561         http://trac.webkit.org/changeset/167584
2562
2563 2014-04-20  Benjamin Poulain  <benjamin@webkit.org>
2564
2565         Compile the :root pseudo class and fix a related issue with :nth-child()
2566         https://bugs.webkit.org/show_bug.cgi?id=131926
2567
2568         Reviewed by Andreas Kling.
2569
2570         Add the :root pseudo class. This is another trivial selector, we just need to compare
2571         the element pointer with the documentElement.
2572
2573         I discovered some issues with :nth-child(n) through the layout tests for ":root".
2574         When the pseudo class nth-child could match anything, no code was generated. That decision
2575         was taken when generating the fragments.
2576
2577         The specification of :nth-child() has two tests: the parent test and the counter test.
2578         Since some fragments would not generate any code for :nth-child(n), they would succeed on the root,
2579         which is incorrect since the root should fail the parent test.
2580
2581         This was fixed by moving the filtering of non-counting :nth-child() after we generate the parent
2582         check.
2583         We still don't generate any counter test unless required.
2584
2585         Test: fast/selectors/nth-child-on-root.html
2586
2587         * cssjit/SelectorCompiler.cpp:
2588         (WebCore::SelectorCompiler::addPseudoClassType):
2589         (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementMatching):
2590         (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementIsNthChild):
2591         (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementIsRoot):
2592         * dom/Document.h:
2593         (WebCore::Document::documentElementMemoryOffset):
2594
2595 2014-04-20  Benjamin Poulain  <benjamin@webkit.org>
2596
2597         Add Element.matches, the standard name for webkitMatchesSelector
2598         https://bugs.webkit.org/show_bug.cgi?id=131922
2599
2600         Reviewed by Andreas Kling.
2601
2602         This patch just renames webkitMatchesSelector() to matches().
2603
2604         * bindings/scripts/CodeGeneratorObjC.pm:
2605         (GenerateImplementation):
2606         The Objective-C binding generator only supported ImplementedAs for getters and setters.
2607         This adds support for the option so that Element.matches can be generated for Obj-C.
2608
2609         * dom/Element.cpp:
2610         (WebCore::Element::matches):
2611         (WebCore::Element::webkitMatchesSelector): Deleted.
2612         * dom/Element.h:
2613         * dom/Element.idl:
2614         * inspector/InspectorCSSAgent.cpp:
2615         (WebCore::InspectorCSSAgent::buildArrayForMatchedRuleList):
2616
2617 2014-04-20  Zalan Bujtas  <zalan@apple.com>
2618
2619         Subpixel rendering: RenderLayer's size is set using enclosingRect() which can result in cruft.
2620         https://bugs.webkit.org/show_bug.cgi?id=131468
2621
2622         Reviewed by Darin Adler.
2623
2624         Layers marked opaque may not paint edge pixels, when RenderLayer bounds are non-integral and
2625         get device pixel snapped.
2626         For details see this: https://trac.webkit.org/r159463
2627
2628         Test: compositing/hidpi-non-simple-compositing-layer-with-fractional-size-and-background.html
2629
2630         * rendering/RenderLayerBacking.cpp:
2631         (WebCore::RenderLayerBacking::updateGraphicsLayerGeometry):
2632
2633 2014-04-19  Darin Adler  <darin@apple.com>
2634
2635         Simple layout can get confused by coordinate overflow
2636         https://bugs.webkit.org/show_bug.cgi?id=131890
2637         rdar://problem/15558510
2638
2639         Reviewed by Andreas Kling.
2640
2641         Test: fast/css/simple-layout-overflow.html
2642
2643         * rendering/SimpleLineLayoutResolver.h:
2644         (WebCore::SimpleLineLayout::RunResolver::rangeForRect):
2645         If lastLine is smaller than firstLine, just ignore it.
2646
2647 2014-04-20  Darin Adler  <darin@apple.com>
2648
2649         ScriptExecutionContext::stopActiveDOMObjects iterates a hash map that can change during iteration (for multiple reasons, including GC)
2650         https://bugs.webkit.org/show_bug.cgi?id=52719
2651
2652         Reviewed by Alexey Proskuryakov.
2653
2654         At least two specific ways this can happen:
2655
2656         1) XMLHttpRequest::stop can trigger a JavaScript garbage collection.
2657         2) NotificationCenter::stop can delete the last references to notifications;
2658            those notifications are also active DOM objects.
2659
2660         Besides fixing the iteration in that function, did some other fixes for the
2661         ScriptExecutionContext class, including some coding style changes. Many uses
2662         of nullptr instead of 0, without listing each function separately below.
2663
2664         * Modules/webdatabase/DatabaseContext.cpp:
2665         (WebCore::DatabaseContext::contextDestroyed): Call through to the base class
2666         version of contextDestroyed rather than repeating what it does (with a large
2667         comment that doesn't acknowledge the base class alread does it).
2668         * Modules/webdatabase/DatabaseContext.h: Removed some unneeded includes.
2669         Wrote out "private" explicitly for deriving from ActiveDOMObject. Made the
2670         ActiveDOMObject function overrides private, and marked them override and final.
2671
2672         * dom/ActiveDOMObject.h: Updated comments. Replaced suspendIfNeededCalled with
2673         assertSuspendIfNeededWasCalled, which has an empty inline version in the header.
2674         Renamed m_suspendIfNeededCalled to m_suspendIfNeededWasCalled.
2675
2676         * dom/ActiveDOMObject.cpp:
2677         (WebCore::ActiveDOMObject::ActiveDOMObject): Pass a reference instead of a pointer.
2678         (WebCore::ActiveDOMObject::~ActiveDOMObject): Ditto.
2679         (WebCore::ActiveDOMObject::suspendIfNeeded): Ditto.
2680
2681         * dom/ContextDestructionObserver.cpp:
2682         (WebCore::ContextDestructionObserver::observeContext): Pass a reference instead of a pointer.
2683
2684         * dom/MessagePort.cpp:
2685         (WebCore::MessagePort::MessagePort): Pass a reference instead of a pointer.
2686         (WebCore::MessagePort::~MessagePort): Ditto.
2687         (WebCore::MessagePort::disentangle): Ditto.
2688
2689         * dom/ScriptExecutionContext.cpp:
2690         (WebCore::ScriptExecutionContext::ScriptExecutionContext): Updated flags used
2691         for assertions so they are conditional and updated their names.
2692         (WebCore::takeAny): Added. Helper function that we can consider for HashSet in
2693         the future; makes loop below easier to read.
2694         (WebCore::checkConsistency): Added. Assertions that were done multiple places below,
2695         and should not be written over and over again.
2696         (WebCore::ScriptExecutionContext::~ScriptExecutionContext): Changed to use C++11
2697         for loops and the takeAny function above.
2698         (WebCore::ScriptExecutionContext::dispatchMessagePortEvents): Ditto.
2699         (WebCore::ScriptExecutionContext::createdMessagePort): Changed to take a reference
2700         for clarity and so it doesn't have to do an assert the pointer is non-null.
2701         (WebCore::ScriptExecutionContext::destroyedMessagePort): Ditto.
2702         (WebCore::ScriptExecutionContext::canSuspendActiveDOMObjects): Changed to use
2703         C++11 for loop and reworded comment and redid assertions.
2704         (WebCore::ScriptExecutionContext::suspendActiveDOMObjects): Ditto.
2705         (WebCore::ScriptExecutionContext::resumeActiveDOMObjects): Ditto.
2706         (WebCore::ScriptExecutionContext::stopActiveDOMObjects): Changed to support
2707         removal of an active DOM object during the stop function. Included new comments
2708         to clarify what the rules are.
2709         (WebCore::ScriptExecutionContext::suspendActiveDOMObjectIfNeeded): Changed to take
2710         a reference for clarity and so it doesn't have to assert a pointer is non-null.
2711         (WebCore::ScriptExecutionContext::didCreateActiveDOMObject): Ditto. Also changed to
2712         use RELEASE_ASSERT instead of CRASH.
2713         (WebCore::ScriptExecutionContext::willDestroyActiveDOMObject): Ditto.
2714         (WebCore::ScriptExecutionContext::didCreateDestructionObserver): Ditto.
2715         (WebCore::ScriptExecutionContext::willDestroyDestructionObserver): Ditto.
2716         (WebCore::ScriptExecutionContext::closeMessagePorts): Moved the body of this
2717         function into its one call site, ScriptExecutionContext::stopActiveDOMObjects,
2718         since it's simple enough when written as a C++11 for loop.
2719         (WebCore::ScriptExecutionContext::hasPendingActivity): Added. This function was
2720         already exported for workers, and implementing it outside this class required
2721         exposing the private HashSet members; more sensible to implement it here and
2722         simply make it public in WorkerGlobalScope.
2723
2724         * dom/ScriptExecutionContext.h: Removed unnecessary includes and forward declarations.
2725         Removed a long-ago-fixed FIXME. Changed various functions to take references instead of
2726         pointers. Added a protected hasPendingActivity function, deleted the closeMessagePorts
2727         function, deleted the ActiveDOMObjectsSet typedef, made the assertion flags be
2728         !ASSERT_DISABLED only, and deleted the messagePorts and activeDOMObjects functions.
2729
2730         * workers/WorkerGlobalScope.cpp:
2731         (WebCore::WorkerGlobalScope::hasPendingActivity): Deleted. This is now implemented
2732         in the base class.
2733
2734         * workers/WorkerGlobalScope.h: Make hasPendingActivity function from the base class
2735         public instead of declaring it in this class.
2736
2737 2014-04-20  Brent Fulgham  <bfulgham@apple.com>
2738
2739         [Mac] Unable to select 'Off' or 'Auto' from track menu when tracks consist of unsupported track types
2740         https://bugs.webkit.org/show_bug.cgi?id=131908
2741         <rdar://problem/15999882>
2742
2743         Reviewed by Eric Carlson.
2744
2745         * html/HTMLMediaElement.cpp:
2746         (WebCore::HTMLMediaElement::setSelectedTextTrack): When the user selects 'Off', make sure the currently 'showing'
2747         track is switched to 'disabled'.
2748
2749 2014-04-20  Andreas Kling  <akling@apple.com>
2750
2751         Speed up jsStringWithCache() through WeakGCMap inlining.
2752         <https://webkit.org/b/131923>
2753
2754         Inline the common path of WeakGCMap::add() in jsStringWithCache().
2755         26% progression on Bindings/id-getter.html
2756
2757         Reviewed by Darin Adler.
2758
2759         * WebCore.exp.in:
2760         * bindings/js/JSDOMBinding.h:
2761         * bindings/js/JSDOMBinding.cpp:
2762         (WebCore::jsStringWithCache):
2763
2764             Move jsStringWithCache() out of line since we're now blowing up
2765             its size quite a bit.
2766
2767 2014-04-20  Benjamin Poulain  <benjamin@webkit.org>
2768
2769         Fix the build after r167574
2770
2771         * cssjit/SelectorCompiler.cpp:
2772         (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementMatching):
2773
2774 2014-04-20  Benjamin Poulain  <benjamin@webkit.org>
2775
2776         Compile the pseudo class :target
2777         https://bugs.webkit.org/show_bug.cgi?id=131904
2778
2779         Reviewed by Andreas Kling.
2780
2781         The pseudo class :target is trivial, it just compare the current element with
2782         Document::cssTarget().
2783
2784         * cssjit/SelectorCompiler.cpp:
2785         (WebCore::SelectorCompiler::addPseudoType):
2786         (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementMatching):
2787         (WebCore::SelectorCompiler::getDocument):
2788         (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementAttributeValueExactMatching):
2789         (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementIsTarget):
2790         * dom/Document.h:
2791         (WebCore::Document::cssTargetMemoryOffset):
2792
2793 2014-04-20  Benjamin Poulain  <benjamin@webkit.org>
2794
2795         Rename the CSSSelector PseudoType to PseudoClassType
2796         https://bugs.webkit.org/show_bug.cgi?id=131907
2797
2798         Reviewed by Andreas Kling.
2799
2800         Pseudo Elements and Page Pseudo Classes have been moved out of PseudoType in previous patches.
2801         The remaining values in the PseudoType enumeration are the pseudo classes.
2802
2803         This patch is the final clean up, PseudoType is renamed to PseudoClassType.
2804
2805         * css/CSSGrammar.y.in:
2806         * css/CSSParserValues.cpp:
2807         (WebCore::CSSParserSelector::parsePseudoClassAndCompatibilityElementSelector):
2808         * css/CSSParserValues.h:
2809         (WebCore::CSSParserSelector::pseudoClassType):
2810         (WebCore::CSSParserSelector::pseudoType): Deleted.
2811         * css/CSSSelector.cpp:
2812         (WebCore::CSSSelector::specificityForOneSelector):
2813         (WebCore::appendPseudoClassFunctionTail):
2814         (WebCore::CSSSelector::selectorText):
2815         * css/CSSSelector.h:
2816         (WebCore::CSSSelector::pseudoClassType):
2817         (WebCore::CSSSelector::pseudoElementType):
2818         (WebCore::CSSSelector::pagePseudoClassType):
2819         (WebCore::pseudoClassIsRelativeToSiblings):
2820         (WebCore::CSSSelector::isSiblingSelector):
2821         (WebCore::CSSSelector::CSSSelector):
2822         (WebCore::CSSSelector::pseudoType): Deleted.
2823         * css/RuleSet.cpp:
2824         (WebCore::RuleSet::findBestRuleSetAndAdd):
2825         * css/SelectorChecker.cpp:
2826         (WebCore::SelectorChecker::checkOne):
2827         (WebCore::SelectorChecker::checkScrollbarPseudoClass):
2828         (WebCore::SelectorChecker::determineLinkMatchType):
2829         (WebCore::SelectorChecker::matchesFocusPseudoClass):
2830         * css/SelectorChecker.h:
2831         (WebCore::SelectorChecker::isCommonPseudoClassSelector):
2832         * css/SelectorCheckerFastPath.cpp:
2833         (WebCore::SelectorCheckerFastPath::commonPseudoClassSelectorMatches):
2834         * css/SelectorPseudoClassAndCompatibilityElementMap.in:
2835         * css/SelectorPseudoTypeMap.h:
2836         * css/StyleResolver.cpp:
2837         (WebCore::StyleResolver::styleForElement):
2838         * css/makeSelectorPseudoClassAndCompatibilityElementMap.py:
2839         (enumerablePseudoType):
2840         * cssjit/SelectorCompiler.cpp:
2841         (WebCore::SelectorCompiler::addPseudoClassType):
2842         (WebCore::SelectorCompiler::SelectorCodeGenerator::SelectorCodeGenerator):
2843         (WebCore::SelectorCompiler::SelectorCodeGenerator::generateElementMatching):
2844         (WebCore::SelectorCompiler::addPseudoType): Deleted.
2845         * inspector/InspectorCSSAgent.cpp:
2846         (WebCore::computePseudoClassMask):
2847         (WebCore::InspectorCSSAgent::forcePseudoState):
2848         * inspector/InspectorCSSAgent.h:
2849         * inspector/InspectorInstrumentation.cpp:
2850         (WebCore::InspectorInstrumentation::forcePseudoStateImpl):
2851         * inspector/InspectorInstrumentation.h:
2852         (WebCore::InspectorInstrumentation::forcePseudoState):
2853
2854 2014-04-19  Alexey Proskuryakov  <ap@apple.com>
2855
2856         Crashes in HTMLFormElement::submit.
2857         https://bugs.webkit.org/show_bug.cgi?id=131910
2858         <rdar://problem/15661790>
2859
2860         Based on a patch by Kent Tamura.
2861
2862         Reviewed by Anders Carlsson.
2863
2864         Tests: fast/forms/form-submission-crash-2.html
2865                fast/forms/form-submission-crash.html
2866
2867         Code that executes arbitrary JS needs to protect objects that it uses afterwards.
2868
2869         * html/HTMLFormElement.cpp:
2870         (WebCore::HTMLFormElement::prepareForSubmission):
2871         (WebCore::HTMLFormElement::submit):
2872         * loader/FrameLoader.cpp:
2873         (WebCore::FrameLoader::submitForm):
2874
2875 2014-04-20  Antti Koivisto  <antti@apple.com>
2876
2877         Text bounding box computation for simple line layout is wrong
2878         https://bugs.webkit.org/show_bug.cgi?id=131912
2879
2880         Reviewed by Andreas Kling.
2881
2882         Top-left is currently the first line top-left which is not always correct.
2883
2884         * WebCore.exp.in:
2885         * rendering/RenderText.cpp:
2886         (WebCore::RenderText::firstRunLocation):
2887         (WebCore::RenderText::firstRunOrigin): Deleted.
2888         (WebCore::RenderText::firstRunX): Deleted.
2889         (WebCore::RenderText::firstRunY): Deleted.
2890
2891             Keep just one accessor and rename it.
2892             Encapsulate the line box and simple line versions.
2893
2894         * rendering/RenderText.h:
2895         * rendering/RenderTextLineBoxes.cpp:
2896         (WebCore::RenderTextLineBoxes::firstRunLocation):
2897
2898             Line box version.
2899
2900         * rendering/RenderTextLineBoxes.h:
2901         * rendering/RenderTreeAsText.cpp:
2902
2903             Simplify RenderText dumping.
2904
2905         (WebCore::RenderTreeAsText::writeRenderObject):
2906         * rendering/SimpleLineLayoutFunctions.cpp:
2907         (WebCore::SimpleLineLayout::computeTextBoundingBox):
2908
2909             Return the correct x position.
2910
2911         (WebCore::SimpleLineLayout::computeTextFirstRunLocation):
2912
2913             Simple line version.
2914
2915         * rendering/SimpleLineLayoutFunctions.h:
2916         * rendering/svg/SVGRenderTreeAsText.cpp:
2917         (WebCore::writeSVGInlineText):
2918
2919 2014-04-19  Zalan Bujtas  <zalan@apple.com>
2920
2921         Subpixel rendering: RenderLayer's clipping should snap to device pixel boundaries.
2922         https://bugs.webkit.org/show_bug.cgi?id=131466
2923
2924         Reviewed by Simon Fraser.
2925
2926         Fractional pixel clipping can produce cruft on RenderLayers. Since RenderLayer
2927         sizing and painting are snapped, painting clip rect needs to be snapped the same way.
2928
2929         Test: fast/clip/clip-when-rect-has-fractional-pixel-value.html
2930
2931         * rendering/RenderLayer.cpp:
2932         (WebCore::RenderLayer::clipToRect):
2933
2934 2014-04-19  Brent Fulgham  <bfulgham@apple.com>
2935
2936         Latched scrolling may interact badly with custom programmatic scrolling
2937         https://bugs.webkit.org/show_bug.cgi?id=131869
2938         <rdar://problem/16249557>
2939
2940         Reviewed by Darin Adler.
2941
2942         * dom/Element.cpp:
2943         (WebCore::Element::setScrollLeft): Mark scrollable area as having
2944         been scrolled programmatically.
2945         (WebCore::Element::setScrollTop): Ditto.
2946         * page/EventHandler.cpp:
2947         (WebCore::EventHandler::handleWheelEvent): Check for programmatic scroll, and
2948         clear latched state if the handler manually scrolled. Clear programmatic
2949         scroll state at the end of event handling.
2950         (WebCore::EventHandler::clearLatchedState): Refactored code.
2951         * page/EventHandler.h:
2952         * page/mac/EventHandlerMac.mm:
2953         (WebCore::EventHandler::platformPrepareForWheelEvents): Check
2954         if scrollable area was scrolled programmatically. If it was, do
2955         not honor latching behavior.
2956         * platform/ScrollableArea.cpp:
2957         (WebCore::ScrollableArea::ScrollableArea): Initialize new member.
2958         * platform/ScrollableArea.h:
2959         (WebCore::ScrollableArea::isScrolledProgrammatically): Added.
2960         (WebCore::ScrollableArea::setScrolledProgrammatically): Added.
2961
2962 2014-04-19  Chris Fleizach  <cfleizach@apple.com>
2963
2964         AX: grid rows are not recognized do to lack of explicit role="row", role="gridcell"
2965         https://bugs.webkit.org/show_bug.cgi?id=131819
2966
2967         Reviewed by Darin Adler.
2968
2969         Test: accessibility/table-with-missing-aria-role-rows.html
2970
2971         If a table is using role="grid", and it has <tr> elements that do not have
2972         role="row", we should still treat them as row elements.
2973
2974         * accessibility/AccessibilityARIAGrid.cpp:
2975         (WebCore::AccessibilityARIAGrid::addTableCellChild):
2976
2977 2014-04-19  Chris Fleizach  <cfleizach@apple.com>
2978
2979         AX: Malformed tables exposing incorrect col and colSpans
2980         https://bugs.webkit.org/show_bug.cgi?id=131796
2981
2982         Reviewed by Darin Adler.
2983
2984         Test: accessibility/table-incorrect-colspan-cell.html
2985
2986         When a developer has specified malformed colspans, accessibility is reporting those values instead of the effective column information.
2987
2988         * accessibility/AccessibilityTableCell.cpp:
2989         (WebCore::AccessibilityTableCell::columnIndexRange):
2990
2991 2014-04-19  Benjamin Poulain  <bpoulain@apple.com>
2992
2993         Make the CSS JIT compile for ARM64
2994         https://bugs.webkit.org/show_bug.cgi?id=131834
2995
2996         Reviewed by Gavin Barraclough.
2997
2998         Add the CPU specific code required to compile SelectorCompiler on ARM64.
2999         The code is absolutely non-functional, it would crash immediately, but it compiles.
3000
3001         * cssjit/FunctionCall.h:
3002         (WebCore::FunctionCall::FunctionCall):
3003         (WebCore::FunctionCall::callAndBranchOnCondition):
3004         * cssjit/RegisterAllocator.h:
3005         What am I supposed to do with all those registers? There are so many of them :)
3006         The array calleeSavedRegisters is defined for compatibility but it cannot be reached at the moment.
3007
3008         (WebCore::RegisterAllocator::isValidRegister):
3009         * cssjit/SelectorCompiler.cpp:
3010         (WebCore::SelectorCompiler::SelectorCodeGenerator::modulo):
3011         Implement modulo by doing dividend - (divisor * int(dividend/divisor)).
3012
3013         * cssjit/StackAllocator.h:
3014         StackAllocator will need a small redesign to handle the 16 bytes alligned stack of
3015         ARM64.
3016         The code has been modified to build but it is pretty useless.
3017
3018         (WebCore::StackAllocator::allocateUninitialized):
3019         (WebCore::StackAllocator::push):
3020         (WebCore::StackAllocator::pop):
3021         (WebCore::StackAllocator::popAndDiscard):
3022         (WebCore::StackAllocator::popAndDiscardUpTo):
3023         (WebCore::StackAllocator::alignStackPreFunctionCall):
3024         (WebCore::StackAllocator::unalignStackPostFunctionCall):
3025
3026 2014-04-19  Simon Fraser  <simon.fraser@apple.com>
3027
3028         [UI-side compositing] Implement blend modes
3029         https://bugs.webkit.org/show_bug.cgi?id=131891
3030         <rdar://problem/16490085>
3031
3032         Reviewed by Tim Horton.
3033
3034         Implement blend modes with UI-side compositing.
3035
3036         PlatformCALayer::setBlendMode() becomes pure virtual, and overridden in
3037         PlatformCALayerMac and PlatformCALayerRemote.
3038         PlatformCAFilters::setBlendingFiltersOnLayer() now takes a raw platform
3039         layer, so it can be called from the UI process on a CALayer*.
3040
3041         Fix an issue with GraphicsLayerCA::LayerChange flags. 1 << 31 shifts
3042         a signed integer, so it ended up setting all the high bits in m_uncommittedChanges,
3043         causing us to push blend modes to the UI process for every layer. Fix by shifting
3044         an unsigned long long (1LLU << 31). This should also fix a build warning on Windows.
3045
3046         * WebCore.exp.in:
3047         * platform/graphics/ca/GraphicsLayerCA.h:
3048         * platform/graphics/ca/PlatformCAFilters.h:
3049         * platform/graphics/ca/PlatformCALayer.h:
3050         * platform/graphics/ca/mac/PlatformCAFiltersMac.mm:
3051         (PlatformCAFilters::setBlendingFiltersOnLayer):
3052         * platform/graphics/ca/mac/PlatformCALayerMac.h:
3053         * platform/graphics/ca/mac/PlatformCALayerMac.mm:
3054         (PlatformCALayerMac::setFilters):
3055         (PlatformCALayerMac::setBlendMode):
3056         (PlatformCALayer::setBlendMode): Deleted.
3057
3058 2014-04-19  Joseph Pecoraro  <pecoraro@apple.com>
3059
3060         Fix AVPlayerController leak on pages with video
3061         https://bugs.webkit.org/show_bug.cgi?id=131878
3062
3063         Reviewed by Eric Carlson.
3064
3065         * platform/ios/WebVideoFullscreenInterfaceAVKit.mm:
3066         (-[WebAVPlayerController init]):
3067         Assigning to the retain property will increment the count,
3068         so balance the alloc with an autorelease.
3069
3070 2014-04-19  Andreas Kling  <akling@apple.com>
3071
3072         Slap ALWAYS_INLINE on Element attribute lookup things.
3073         <https://webkit.org/b/131892>
3074
3075         Most of the "inline" helpers for setting/getting attributes were not
3076         actually getting inlined by the compiler, so let's make that happen.
3077
3078         Reviewed by Geoffrey Garen.
3079
3080         * dom/Element.cpp:
3081         (WebCore::Element::synchronizeAttribute):
3082         * dom/ElementData.h:
3083         (WebCore::ElementData::findAttributeByName):
3084         (WebCore::ElementData::findAttributeIndexByName):
3085
3086 2014-04-19  Darin Adler  <darin@apple.com>
3087
3088         Telephone number detection should respect its setting consistently
3089         https://bugs.webkit.org/show_bug.cgi?id=131893
3090         rdar://problem/16597639
3091
3092         Reviewed by Tim Horton.
3093
3094         * editing/Editor.cpp:
3095         (WebCore::Editor::respondToChangedSelection): Only start the timer
3096         if shouldDetectTelephoneNumbers returns true.
3097         (WebCore::Editor::shouldDetectTelephoneNumbers): Added. Calls both
3098         isTelephoneNumberParsingEnabled and TelephoneNumberDetector::isSupported.
3099         (WebCore::Editor::scanSelectionForTelephoneNumbers): Use
3100         shouldDetectTelephoneNumbers.
3101         (WebCore::Editor::clearDataDetectedTelephoneNumbers): Use document()
3102         instead of m_frame.document().
3103
3104         * editing/Editor.h: Added declaration of shouldDetectTelephoneNumbers.
3105
3106 2014-04-19  Andrei Bucur  <abucur@adobe.com>
3107
3108         [CSS Regions] Harden the layout in case there are no regions
3109         https://bugs.webkit.org/show_bug.cgi?id=131517
3110
3111         Reviewed by Mihnea Ovidenie.
3112
3113         The patch fixes the cases when the content of a flow thread is not
3114         properly invalidated when all the regions of its chain are removed.
3115
3116         Test: fast/regions/simplified-layout-no-regions.html
3117
3118         * rendering/RenderFlowThread.cpp:
3119         (WebCore::RenderFlowThread::logicalWidthChangedInRegionsForBlock):
3120         * rendering/RenderRegion.cpp:
3121         (WebCore::RenderRegion::ensureOverflowForBox):
3122
3123 2014-04-19  Zalan Bujtas  <zalan@apple.com>
3124
3125         https://bugs.webkit.org/show_bug.cgi?id=131594
3126         Subpixel rendering: WK1: Wrong repaint rect is calculated when layer has non-compositing transform.
3127
3128         Reviewed by Simon Fraser.
3129
3130         We need to snap the repaint rect for transformed object the same way when we paint them.
3131
3132         Test: fast/repaint/hidpi-wrong-repaint-rect-when-parent-has-noncompositing-transform.html
3133
3134         * rendering/RenderBox.cpp:
3135         (WebCore::RenderBox::computeRectForRepaint):
3136
3137 2014-04-18  Simon Fraser  <simon.fraser@apple.com>
3138
3139         Use 'override' in GraphicsLayerCA
3140         https://bugs.webkit.org/show_bug.cgi?id=131882
3141
3142         Reviewed by Tim Horton.
3143
3144         Add 'override' to all overridden functions. Found a few that need
3145         not have been virtual. Removed "allowTiledLayer" stuff that was
3146         never called.
3147
3148         * WebCore.exp.in:
3149         * platform/graphics/ca/GraphicsLayerCA.cpp:
3150         (WebCore::GraphicsLayerCA::GraphicsLayerCA):
3151         (WebCore::GraphicsLayerCA::requiresTiledLayer):
3152         (WebCore::GraphicsLayerCA::setAllowTiledLayer): Deleted.
3153         * platform/graphics/ca/GraphicsLayerCA.h:
3154         (WebCore::GraphicsLayerCA::platformCALayer):
3155         (WebCore::GraphicsLayerCA::hasContentsLayer): Deleted.
3156         (WebCore::GraphicsLayerCA::allowTiledLayer): Deleted.
3157         (WebCore::GraphicsLayerCA::isGraphicsLayerCA): Deleted.
3158         (WebCore::GraphicsLayerCA::platformCALayerLayoutSublayersOfLayer): Deleted.
3159         (WebCore::GraphicsLayerCA::platformCALayerRespondsToLayoutChanges): Deleted.
3160         (WebCore::GraphicsLayerCA::platformCALayerContentsOrientation): Deleted.
3161         (WebCore::GraphicsLayerCA::platformCALayerShowDebugBorders): Deleted.
3162         (WebCore::GraphicsLayerCA::platformCALayerIncrementRepaintCount): Deleted.
3163         (WebCore::GraphicsLayerCA::platformCALayerContentsOpaque): Deleted.
3164         (WebCore::GraphicsLayerCA::platformCALayerDrawsContent): Deleted.
3165         (WebCore::GraphicsLayerCA::platformCALayerLayerDidDisplay): Deleted.
3166
3167 2014-04-18  Timothy Hatcher  <timothy@apple.com>
3168
3169         Web Inspector: Move InspectorProfilerAgent to JavaScriptCore
3170         https://bugs.webkit.org/show_bug.cgi?id=131673
3171
3172         Passes existing profiler and inspector tests.
3173
3174         Reviewed by Joseph Pecoraro.
3175
3176         * CMakeLists.txt:
3177         * DerivedSources.make:
3178         * ForwardingHeaders/inspector/agents/InspectorProfilerAgent.h: Added.
3179         * WebCore.exp.in:
3180         * WebCore.vcxproj/WebCore.vcxproj:
3181         * WebCore.vcxproj/WebCore.vcxproj.filters:
3182         * WebCore.xcodeproj/project.pbxproj:
3183         * bindings/js/ScriptProfile.cpp: Removed.
3184         * bindings/js/ScriptProfile.h:
3185         * bindings/js/ScriptProfiler.cpp: Removed.
3186         * bindings/js/ScriptProfiler.h: Removed.
3187         * inspector/InspectorAllInOne.cpp:
3188         * inspector/InspectorConsoleInstrumentation.h:
3189         (WebCore::InspectorInstrumentation::profilerEnabled):
3190         (WebCore::InspectorInstrumentation::startProfiling):
3191         (WebCore::InspectorInstrumentation::stopProfiling):
3192         * inspector/InspectorController.cpp:
3193         (WebCore::InspectorController::InspectorController):
3194         * inspector/InspectorController.h:
3195         * inspector/InspectorInstrumentation.cpp:
3196         (WebCore::InspectorInstrumentation::didCommitLoadImpl):
3197         (WebCore::InspectorInstrumentation::startProfilingImpl):
3198         (WebCore::InspectorInstrumentation::stopProfilingImpl):
3199         (WebCore::InspectorInstrumentation::consoleTimeStampImpl): Deleted.
3200         * inspector/InspectorInstrumentation.h:
3201         * inspector/InspectorProfilerAgent.cpp: Removed.
3202         * inspector/InspectorProfilerAgent.h: Removed.
3203         * inspector/InspectorTimelineAgent.cpp:
3204         (WebCore::startProfiling):
3205         (WebCore::stopProfiling):
3206         (WebCore::InspectorTimelineAgent::willCallFunction):
3207         (WebCore::InspectorTimelineAgent::didCallFunction):
3208         (WebCore::InspectorTimelineAgent::willEvaluateScript):
3209         (WebCore::InspectorTimelineAgent::didEvaluateScript):
3210         (WebCore::InspectorTimelineAgent::didDispatchXHRLoadEvent): Deleted.
3211         * inspector/InstrumentingAgents.h:
3212         (WebCore::InstrumentingAgents::inspectorProfilerAgent):
3213         (WebCore::InstrumentingAgents::setInspectorProfilerAgent):
3214         * inspector/PageDebuggerAgent.cpp:
3215         (WebCore::PageDebuggerAgent::sourceMapURLForScript): Deleted.
3216         * inspector/PageDebuggerAgent.h:
3217         * inspector/PageProfilerAgent.cpp: Added.
3218         (WebCore::PageProfilerAgent::PageProfilerAgent):
3219         (WebCore::PageProfilerAgent::profilingGlobalExecState):
3220         * inspector/PageProfilerAgent.h: Added.
3221         (WebCore::PageProfilerAgent::~PageProfilerAgent):
3222         * inspector/ScriptProfile.idl:
3223         * inspector/TimelineRecordFactory.cpp:
3224         (WebCore::TimelineRecordFactory::appendProfile):
3225         * inspector/TimelineRecordFactory.h:
3226         * inspector/WebConsoleAgent.cpp:
3227         * inspector/WebProfilerAgent.cpp: Added.
3228         (WebCore::WebProfilerAgent::WebProfilerAgent):
3229         (WebCore::WebProfilerAgent::enable):
3230         (WebCore::WebProfilerAgent::disable):
3231         * inspector/WebProfilerAgent.h: Added.
3232         (WebCore::WebProfilerAgent::~WebProfilerAgent):
3233         * inspector/WorkerInspectorController.cpp:
3234         (WebCore::WorkerInspectorController::WorkerInspectorController):
3235         * inspector/WorkerProfilerAgent.cpp: Added.
3236         (WebCore::WorkerProfilerAgent::WorkerProfilerAgent):
3237         (WebCore::WorkerProfilerAgent::profilingGlobalExecState):
3238         * inspector/WorkerProfilerAgent.h: Added.
3239         (WebCore::WorkerProfilerAgent::~WorkerProfilerAgent):
3240         * loader/appcache/ApplicationCacheGroup.cpp:
3241         * page/PageConsole.cpp:
3242         (WebCore::PageConsole::profile):
3243         (WebCore::PageConsole::profileEnd):
3244         (WebCore::PageConsole::time): Deleted.
3245         (WebCore::PageConsole::timeEnd): Deleted.
3246         (WebCore::PageConsole::timeStamp): Deleted.
3247         * page/PageConsole.h:
3248         * testing/Internals.cpp:
3249         * testing/Internals.h:
3250         * xml/XMLHttpRequest.cpp:
3251
3252 2014-04-18  Simon Fraser  <simon.fraser@apple.com>
3253
3254         Contents of directly composited image layers are sometimes missing
3255         https://bugs.webkit.org/show_bug.cgi?id=131880
3256
3257         Reviewed by Tim Horton.
3258
3259         RenderLayerBacking::updateGraphicsLayerGeometry() tested whether a
3260         GraphicsLayer had a contents layer to determine whether to call code
3261         that updates the contents rect. However, on Mac, we don't make a contents
3262         layer until later at layer flush time. That would result in an empty
3263         contentsRect being set.
3264
3265         Fix by renaming GraphicsLayer::hasContentsLayer() to usesContentsLayer(),
3266         and on Mac consulting the m_contentsLayerPurpose flag, which is set
3267         at the appropriate time.
3268
3269         Test: compositing/images/update-content-rect.html
3270
3271         * platform/ScrollView.cpp:
3272         (WebCore::positionScrollbarLayer):
3273         * platform/graphics/GraphicsLayer.cpp:
3274         (WebCore::GraphicsLayer::getDebugBorderInfo):
3275         * platform/graphics/GraphicsLayer.h:
3276         (WebCore::GraphicsLayer::usesContentsLayer):
3277         (WebCore::GraphicsLayer::hasContentsLayer): Deleted.
3278         * platform/graphics/ca/GraphicsLayerCA.h:
3279         (WebCore::GraphicsLayerCA::hasContentsLayer): Deleted.
3280         * platform/graphics/texmap/GraphicsLayerTextureMapper.h:
3281         (WebCore::GraphicsLayerTextureMapper::usesContentsLayer):
3282         (WebCore::GraphicsLayerTextureMapper::hasContentsLayer): Deleted.
3283         * platform/graphics/texmap/coordinated/CoordinatedGraphicsLayer.h:
3284         * rendering/RenderLayerBacking.cpp:
3285         (WebCore::RenderLayerBacking::updateGraphicsLayerGeometry):
3286         (WebCore::RenderLayerBacking::positionOverflowControlsLayers):
3287         (WebCore::RenderLayerBacking::compositingLayerType):
3288
3289 2014-04-18  Commit Queue  <commit-queue@webkit.org>
3290
3291         Unreviewed, rolling out r167527.
3292         https://bugs.webkit.org/show_bug.cgi?id=131883
3293
3294         Broke 32-bit build (Requested by ap on #webkit).
3295
3296         Reverted changeset:
3297
3298         "[Mac] implement WebKitDataCue"
3299         https://bugs.webkit.org/show_bug.cgi?id=131799
3300         http://trac.webkit.org/changeset/167527
3301
3302 2014-04-18  Eric Carlson  <eric.carlson@apple.com>
3303
3304         [Mac] implement WebKitDataCue
3305         https://bugs.webkit.org/show_bug.cgi?id=131799
3306
3307         Reviewed by Dean Jackson.
3308
3309         Tests: http/tests/media/track-in-band-hls-metadata.html
3310                media/track/track-datacue-value.html
3311
3312         * Configurations/FeatureDefines.xcconfig: Define ENABLE_DATACUE_VALUE.
3313         * DerivedSources.make: Add ENABLE_DATACUE_VALUE to HTML_FLAGS when appropriate.
3314
3315         * CMakeLists.txt: Add JSDataCueCustom.cpp.
3316         * bindings/js/JSBindingsAllInOne.cpp:
3317         * WebCore.xcodeproj/project.pbxproj: Add new files.
3318
3319         * bindings/js/JSDataCueCustom.cpp: Added.
3320         (WebCore::JSDataCue::value):
3321         (WebCore::JSDataCue::setValue):
3322         (WebCore::JSDataCueConstructor::constructJSDataCue): Custom constructor.
3323
3324         * html/HTMLMediaElement.cpp:
3325         (WebCore::HTMLMediaElement::updateActiveTextTrackCues): Drive-by fixes: don't schedule timeupdate
3326             events when paused, don't call sort on an Vector that can't be sorted.
3327
3328         * html/track/DataCue.cpp:
3329         (WebCore::DataCue::DataCue): Initialize m_type.
3330         (WebCore::DataCue::~DataCue): Unprotect the JSValue if necessary.
3331         (WebCore::DataCue::data): Ask the platform value for data if non-null.
3332         (WebCore::DataCue::setData): Clear m_platformValue and m_value.
3333         (WebCore::DataCue::isEqual): New.
3334         (WebCore::DataCue::value): Return a JSValue from the platform value, or the value passed
3335             to the constructor/set by script.
3336         (WebCore::DataCue::setValue): Set m_value.
3337         * html/track/DataCue.h:
3338         * html/track/DataCue.idl:
3339
3340         * html/track/InbandDataTextTrack.cpp:
3341         (WebCore::InbandDataTextTrack::addDataCue): Don't add the same cue more than once.
3342         (WebCore::InbandDataTextTrack::updateDataCue): Update a cue's duration.
3343         (WebCore::InbandDataTextTrack::removeDataCue): Remove an incomplete cue.
3344         (WebCore::InbandDataTextTrack::removeCue): Remove a cue from the incomplete cue map if necessary.
3345         * html/track/InbandDataTextTrack.h:
3346
3347         * html/track/InbandGenericTextTrack.cpp:
3348         (WebCore::InbandGenericTextTrack::addGenericCue): CueMatchRules is now in TextTrackCue instead
3349             of VTTCue.
3350         * html/track/InbandTextTrack.h:
3351
3352         * html/track/InbandWebVTTTextTrack.cpp:
3353         (WebCore::InbandWebVTTTextTrack::newCuesParsed): Ditto.
3354
3355         * html/track/TextTrack.cpp:
3356         (WebCore::TextTrack::hasCue): Ditto.
3357         * html/track/TextTrack.h:
3358
3359         * html/track/TextTrackCue.cpp:
3360         (WebCore::TextTrackCue::isEqual): New, test base class equality.
3361         * html/track/TextTrackCue.h:
3362
3363         * html/track/TextTrackCueGeneric.cpp:
3364         (WebCore::TextTrackCueGeneric::isEqual): Call TextTrackCue::isEqual first.
3365         * html/track/TextTrackCueGeneric.h:
3366
3367         * html/track/VTTCue.cpp:
3368         (WebCore::VTTCue::isEqual): Call TextTrackCue::isEqual first.
3369         * html/track/VTTCue.h:
3370
3371         * platform/SerializedPlatformRepresentation.h: Added.
3372         (WebCore::SerializedPlatformRepresentation::~SerializedPlatformRepresentation):
3373         (WebCore::SerializedPlatformRepresentation::SerializedPlatformRepresentation):
3374
3375         * platform/graphics/InbandTextTrackPrivateClient.h: Add methods for DataCue with SerializedPlatformRepresentation.
3376
3377         * platform/graphics/avfoundation/InbandMetadataTextTrackPrivateAVF.cpp: Added.
3378         (WebCore::InbandMetadataTextTrackPrivateAVF::create):
3379         (WebCore::InbandMetadataTextTrackPrivateAVF::InbandMetadataTextTrackPrivateAVF):
3380         (WebCore::InbandMetadataTextTrackPrivateAVF::~InbandMetadataTextTrackPrivateAVF):
3381         (WebCore::InbandMetadataTextTrackPrivateAVF::addDataCue):
3382         (WebCore::InbandMetadataTextTrackPrivateAVF::updatePendingCueEndTimes):
3383         (WebCore::InbandMetadataTextTrackPrivateAVF::flushPartialCues):
3384         * platform/graphics/avfoundation/InbandMetadataTextTrackPrivateAVF.h: Added.
3385
3386         * platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.cpp:
3387         (WebCore::MediaPlayerPrivateAVFoundation::seekWithTolerance): currentTrack -> currentTextTrack.
3388         (WebCore::MediaPlayerPrivateAVFoundation::seekCompleted): Ditto.
3389         (WebCore::MediaPlayerPrivateAVFoundation::configureInbandTracks): Ditto.
3390         * platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.h:
3391         (WebCore::MediaPlayerPrivateAVFoundation::setCurrentTextTrack):
3392         (WebCore::MediaPlayerPrivateAVFoundation::setCurrentTrack): Deleted.
3393
3394         Renamed currentTrack and setCurrentTrack to currentTextTrack and setCurrentTextTrack.
3395         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
3396         (WebCore::AVFWrapper::currentTextTrack):
3397         (WebCore::MediaPlayerPrivateAVFoundationCF::setCurrentTextTrack):
3398         (WebCore::MediaPlayerPrivateAVFoundationCF::currentTextTrack):
3399         (WebCore::AVFWrapper::setCurrentTextTrack):
3400         (WebCore::AVFWrapper::AVFWrapper):
3401         (WebCore::AVFWrapper::processCue):
3402         (WebCore::AVFWrapper::currentTrack): Deleted.
3403         (WebCore::MediaPlayerPrivateAVFoundationCF::setCurrentTrack): Deleted.
3404         (WebCore::MediaPlayerPrivateAVFoundationCF::currentTrack): Deleted.
3405         (WebCore::AVFWrapper::setCurrentTrack): Deleted.
3406         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.h:
3407
3408         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
3409         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
3410         (WebCore::MediaPlayerPrivateAVFoundationObjC::MediaPlayerPrivateAVFoundationObjC):
3411         (WebCore::MediaPlayerPrivateAVFoundationObjC::seekToTime): currentTrack -> currentTextTrack
3412         (WebCore::MediaPlayerPrivateAVFoundationObjC::tracksChanged): currentTrack -> currentTextTrack.
3413         (WebCore::MediaPlayerPrivateAVFoundationObjC::processMetadataTrack): New.
3414         (WebCore::MediaPlayerPrivateAVFoundationObjC::processCue): m_currentTrack -> m_currentTextTrack.
3415         (WebCore::MediaPlayerPrivateAVFoundationObjC::flushCues): Ditto.
3416         (WebCore::MediaPlayerPrivateAVFoundationObjC::setCurrentTextTrack): Renamed from setCurrentTextTrack.
3417         (WebCore::metadataType): Map an AVFoundation metadata key space to a metadata cue type.
3418         (WebCore::MediaPlayerPrivateAVFoundationObjC::metadataDidArrive): Process new metadata.
3419         (-[WebCoreAVFMovieObserver observeValueForKeyPath:ofObject:change:context:]):
3420         (WebCore::MediaPlayerPrivateAVFoundationObjC::setCurrentTrack): Deleted.
3421
3422         Create a JSValue representation from an AVMetadataItem.
3423         * platform/mac/SerializedPlatformRepresentationMac.h: Added.
3424         (WebCore::SerializedPlatformRepresentationMac::platformType):
3425         (WebCore::SerializedPlatformRepresentationMac::nativeValue):
3426         * platform/mac/SerializedPlatformRepresentationMac.mm: Added.
3427         (WebCore::SerializedPlatformRepresentationMac::SerializedPlatformRepresentationMac):
3428         (WebCore::SerializedPlatformRepresentationMac::~SerializedPlatformRepresentationMac):
3429         (WebCore::SerializedPlatformRepresentationMac::create):
3430         (WebCore::SerializedPlatformRepresentationMac::data):
3431         (WebCore::SerializedPlatformRepresentationMac::deserialize):
3432         (WebCore::SerializedPlatformRepresentationMac::isEqual):
3433         (WebCore::toSerializedPlatformRepresentationMac):
3434         (WebCore::jsValueWithValueInContext):
3435         (WebCore::jsValueWithDataInContext):
3436         (WebCore::jsValueWithArrayInContext):
3437         (WebCore::jsValueWithDictionaryInContext):
3438         (WebCore::jsValueWithAVMetadataItemInContext):
3439
3440 2014-04-18  Jon Honeycutt  <jhoneycutt@apple.com>
3441
3442         Harden RenderInline::inlineElementContinuation()
3443
3444         <https://bugs.webkit.org/show_bug.cgi?id=131858>
3445
3446         Reviewed by Sam Weinig.
3447
3448         No new tests, as there are no known cases of this happening.
3449
3450         * rendering/RenderInline.cpp:
3451         (WebCore::RenderInline::inlineElementContinuation):
3452         Return nullptr if the continuation is neither a RenderInline nor a
3453         RenderBlock.
3454
3455 2014-04-18  Stephanie Lewis  <slewis@apple.com>
3456