Source/WebCore/ChangeLog

   1 2014-11-18  Myles C. Maxfield  <mmaxfield@apple.com>
   2
   3         Subclass CachedFont for SVG fonts
   4         https://bugs.webkit.org/show_bug.cgi?id=138686
   5
   6         Reviewed by Simon Fraser.
   7
   8         Re-landing due to flakey bots.
   9
  10         CachedFont had many #if ENABLE(SVG_FONTS) scattered throughout it.
  11         Splitting out this SVG-specific code into a subclass of CachedFont
  12         cleans up the design.
  13
  14         No new tests because there is no behavior change.
  15
  16         * CMakeLists.txt:
  17         * WebCore.vcxproj/WebCore.vcxproj:
  18         * WebCore.vcxproj/WebCore.vcxproj.filters:
  19         * WebCore.xcodeproj/project.pbxproj:
  20         * css/CSSFontFaceSource.cpp:
  21         (WebCore::CSSFontFaceSource::getFontData): There are two sections
  22         in this function; one for SVG fonts and one for regular fonts.
  23         I've moved these two sections into CachedFont and SVGCachedFont,
  24         thereby simplifying this function.
  25         (WebCore::CSSFontFaceSource::ensureFontData): Ditto.
  26         (WebCore::CSSFontFaceSource::svgFontFaceElement): Moved to header.
  27         (WebCore::CSSFontFaceSource::setSVGFontFaceElement): Ditto.
  28         (WebCore::CSSFontFaceSource::isSVGFontFaceSource): Ditto.
  29         * css/CSSFontFaceSource.h: Put functions that are simple enough to
  30         be inlined here.
  31         * css/CSSFontFaceSrcValue.cpp:
  32         (WebCore::CSSFontFaceSrcValue::isSVGFontTarget): Convenience
  33         function.
  34         (WebCore::CSSFontFaceSrcValue::cachedFont): Pass through a boolean
  35         from CSSFontSelector to CachedResourceHandle regarding if we should
  36         be using SVGCachedFont instead of CachedFont.
  37         * css/CSSFontFaceSrcValue.h: New convenience function.
  38         * css/CSSFontSelector.cpp:
  39         (WebCore::CSSFontSelector::addFontFaceRule): Pass through a boolean
  40         * inspector/InspectorPageAgent.cpp:
  41         (WebCore::InspectorPageAgent::cachedResourceType): Update for new
  42         CachedResource type
  43         (WebCore::cachedResourcesForFrame): Ditto.
  44         * loader/cache/CachedFont.cpp:
  45         (WebCore::CachedFont::CachedFont):
  46         (WebCore::CachedFont::ensureCustomFontData): Taken from
  47         CSSFontFaceSource::getFontData()
  48         (WebCore::CachedFont::getFontData): Ditto.
  49         (WebCore::CachedFont::platformDataFromCustomData): Ditto.
  50         (WebCore::CachedFont::ensureSVGFontData): Moved to SVGCachedFont.
  51         (WebCore::CachedFont::getSVGFontById): Ditto.
  52         * loader/cache/CachedFont.h:
  53         * loader/cache/CachedResource.cpp:
  54         (WebCore::defaultPriorityForResourceType): Update for new enum
  55         type.
  56         * loader/cache/CachedResource.h:
  57         * loader/cache/CachedResourceLoader.cpp:
  58         (WebCore::createResource): Takes new boolean.
  59         (WebCore::CachedResourceLoader::requestFont): Ditto.
  60         (WebCore::CachedResourceLoader::checkInsecureContent): Update for
  61         new enum type
  62         (WebCore::CachedResourceLoader::canRequest): Ditto.
  63         * loader/cache/CachedResourceLoader.h:
  64         * loader/cache/CachedSVGFont.cpp: Added.
  65         (WebCore::CachedSVGFont::CachedSVGFont): Moved from
  66         CSSFontFaceSource::getFontData() and CachedFont
  67         (WebCore::CachedSVGFont::getFontData): Ditto.
  68         (WebCore::CachedSVGFont::platformDataFromCustomData): Ditto.
  69         (WebCore::CachedSVGFont::ensureCustomFontData): Ditto.
  70         (WebCore::CachedSVGFont::getSVGFontById): Ditto.
  71         (WebCore::CachedSVGFont::firstFontFace): Ditto.
  72         * loader/cache/CachedSVGFont.h: Subclass CachedFont.
  73         * loader/cache/MemoryCache.cpp:
  74         (WebCore::MemoryCache::getStatistics): Update for new enum.
  75         * svg/SVGFontFaceUriElement.cpp:
  76         (WebCore::isSVGFontTarget): Convenience function.
  77         (WebCore::SVGFontFaceUriElement::loadFont): Update for new
  78         boolean.
  79
  80 2014-11-18  Chris Dumez  <cdumez@apple.com>
  81
  82         Update the Vector API to deal with unsigned types instead of size_t
  83         https://bugs.webkit.org/show_bug.cgi?id=138824
  84
  85         Reviewed by Andreas Kling.
  86
  87         Update code base to fix build errors related to the typing changes
  88         in the Vector API (size_t -> unsigned).
  89
  90         No new tests, no behavior change.
  91
  92         * WebCore.exp.in:
  93         * bindings/js/JSDOMBinding.h:
  94         (WebCore::jsArray):
  95         * bindings/js/JSWebGLRenderingContextCustom.cpp:
  96         * cssjit/SelectorCompiler.cpp:
  97         * html/HTMLFormElement.cpp:
  98         (WebCore::removeFromVector):
  99         * html/parser/HTMLParserIdioms.h:
 100         * html/parser/XSSAuditor.cpp:
 101
 102 2014-11-18  Myles C. Maxfield  <mmaxfield@apple.com>
 103
 104         Unreviewed iOS build fix
 105
 106         * platform/graphics/ios/SimpleFontDataIOS.mm:
 107         (WebCore::SimpleFontData::platformInit):
 108
 109 2014-11-18  Dan Bernstein  <mitz@apple.com>
 110
 111         iOS build fix.
 112
 113         * platform/graphics/ios/SimpleFontDataIOS.mm:
 114         (WebCore::SimpleFontData::platformInit):
 115
 116 2014-11-18  Commit Queue  <commit-queue@webkit.org>
 117
 118         Unreviewed, rolling out r176264.
 119         https://bugs.webkit.org/show_bug.cgi?id=138837
 120
 121         caused many test failures (Requested by litherum on #webkit).
 122
 123         Reverted changeset:
 124
 125         "Subclass CachedFont for SVG fonts"
 126         https://bugs.webkit.org/show_bug.cgi?id=138686
 127         http://trac.webkit.org/changeset/176264
 128
 129 2014-11-18  Daniel Bates  <dabates@apple.com>
 130
 131         [iOS] Libdispatch, Dyld, IOKit, Mach, NSPointerFunctions, MobileGestalt
 132         SPI are not available in public SDK
 133         https://bugs.webkit.org/show_bug.cgi?id=138802
 134
 135         Reviewed by David Kilzer.
 136
 137         Add SPI wrapper headers {Dispatch, DynamicLinker, IOPMLib, MachVM, NSPointerFunctions, MobileGestalt}SPI.h
 138         that forward declare applicable libdispatch, dyld, IOKit, mach, Foundation, MobileGestalt SPI
 139         and use these headers instead of the including private headers corresponding to these libraries/frameworks.
 140
 141         * WebCore.xcodeproj/project.pbxproj: Add private headers MobileGestaltSPI.h and
 142         MachVMSPI.h, and project headers DynamicLinkerSPI.h, NSPointerFunctionsSPI.h,
 143         DispatchSPI.h and IOPMLibSPI.h. Also, sort files in Xcode project.
 144         * bindings/objc/DOMInternal.mm: Include header NSPointerFunctionsSPI.h.
 145         * bridge/objc/objc_instance.mm: Ditto; Also, sort #import directives.
 146         * platform/cocoa/DisplaySleepDisablerCocoa.cpp: Substitute header IOPMLibSPI.h
 147         for private header IOKit/pwr_mgt/IOPMLib.h.
 148         * platform/cocoa/MemoryPressureHandlerCocoa.mm: Substitute header DispatchSPI.h
 149         for private header dispatch/private.h.
 150         * platform/graphics/ios/FontServicesIOS.mm: Substitute header DynamicLinkerSPI.h
 151         for private header mach-o/dyld_priv.h.
 152         * platform/ios/PlatformScreenIOS.mm: Substitute header MobileGestaltSPI.h for
 153         private header MobileGestalt.h.
 154         * platform/spi/cocoa/DispatchSPI.h: Added.
 155         * platform/spi/cocoa/DynamicLinkerSPI.h: Added.
 156         * platform/spi/cocoa/IOPMLibSPI.h: Added.
 157         * platform/spi/cocoa/MachVMSPI.h: Added.
 158         * platform/spi/cocoa/NSPointerFunctionsSPI.h: Added.
 159         * platform/spi/ios/MobileGestaltSPI.h: Added.
 160
 161 2014-11-18  Myles C. Maxfield  <mmaxfield@apple.com>
 162
 163         Subclass CachedFont for SVG fonts
 164         https://bugs.webkit.org/show_bug.cgi?id=138686
 165
 166         Reviewed by Simon Fraser.
 167
 168         CachedFont had many #if ENABLE(SVG_FONTS) scattered throughout it.
 169         Splitting out this SVG-specific code into a subclass of CachedFont
 170         cleans up the design.
 171
 172         No new tests because there is no behavior change.
 173
 174         * CMakeLists.txt:
 175         * WebCore.vcxproj/WebCore.vcxproj:
 176         * WebCore.vcxproj/WebCore.vcxproj.filters:
 177         * WebCore.xcodeproj/project.pbxproj:
 178         * css/CSSFontFaceSource.cpp:
 179         (WebCore::CSSFontFaceSource::getFontData): There are two sections
 180         in this function; one for SVG fonts and one for regular fonts.
 181         I've moved these two sections into CachedFont and SVGCachedFont,
 182         thereby simplifying this function.
 183         (WebCore::CSSFontFaceSource::ensureFontData): Ditto.
 184         (WebCore::CSSFontFaceSource::svgFontFaceElement): Moved to header.
 185         (WebCore::CSSFontFaceSource::setSVGFontFaceElement): Ditto.
 186         (WebCore::CSSFontFaceSource::isSVGFontFaceSource): Ditto.
 187         * css/CSSFontFaceSource.h: Put functions that are simple enough to
 188         be inlined here.
 189         * css/CSSFontFaceSrcValue.cpp:
 190         (WebCore::CSSFontFaceSrcValue::isSVGFontTarget): Convenience
 191         function.
 192         (WebCore::CSSFontFaceSrcValue::cachedFont): Pass through a boolean
 193         from CSSFontSelector to CachedResourceHandle regarding if we should
 194         be using SVGCachedFont instead of CachedFont.
 195         * css/CSSFontFaceSrcValue.h: New convenience function.
 196         * css/CSSFontSelector.cpp:
 197         (WebCore::CSSFontSelector::addFontFaceRule): Pass through a boolean
 198         * inspector/InspectorPageAgent.cpp:
 199         (WebCore::InspectorPageAgent::cachedResourceType): Update for new
 200         CachedResource type
 201         (WebCore::cachedResourcesForFrame): Ditto.
 202         * loader/cache/CachedFont.cpp:
 203         (WebCore::CachedFont::CachedFont):
 204         (WebCore::CachedFont::ensureCustomFontData): Taken from
 205         CSSFontFaceSource::getFontData()
 206         (WebCore::CachedFont::getFontData): Ditto.
 207         (WebCore::CachedFont::platformDataFromCustomData): Ditto.
 208         (WebCore::CachedFont::ensureSVGFontData): Moved to SVGCachedFont.
 209         (WebCore::CachedFont::getSVGFontById): Ditto.
 210         * loader/cache/CachedFont.h:
 211         * loader/cache/CachedResource.cpp:
 212         (WebCore::defaultPriorityForResourceType): Update for new enum
 213         type.
 214         * loader/cache/CachedResource.h:
 215         * loader/cache/CachedResourceLoader.cpp:
 216         (WebCore::createResource): Takes new boolean.
 217         (WebCore::CachedResourceLoader::requestFont): Ditto.
 218         (WebCore::CachedResourceLoader::checkInsecureContent): Update for
 219         new enum type
 220         (WebCore::CachedResourceLoader::canRequest): Ditto.
 221         * loader/cache/CachedResourceLoader.h:
 222         * loader/cache/CachedSVGFont.cpp: Added.
 223         (WebCore::CachedSVGFont::CachedSVGFont): Moved from
 224         CSSFontFaceSource::getFontData() and CachedFont
 225         (WebCore::CachedSVGFont::getFontData): Ditto.
 226         (WebCore::CachedSVGFont::platformDataFromCustomData): Ditto.
 227         (WebCore::CachedSVGFont::ensureCustomFontData): Ditto.
 228         (WebCore::CachedSVGFont::getSVGFontById): Ditto.
 229         (WebCore::CachedSVGFont::firstFontFace): Ditto.
 230         * loader/cache/CachedSVGFont.h: Subclass CachedFont.
 231         * loader/cache/MemoryCache.cpp:
 232         (WebCore::MemoryCache::getStatistics): Update for new enum.
 233         * svg/SVGFontFaceUriElement.cpp:
 234         (WebCore::isSVGFontTarget): Convenience function.
 235         (WebCore::SVGFontFaceUriElement::loadFont): Update for new
 236         boolean.
 237
 238 2014-11-18  Myles C. Maxfield  <litherum@gmail.com>
 239
 240         Use underlining metrics from the font file
 241         https://bugs.webkit.org/show_bug.cgi?id=138762
 242
 243         Reviewed by Dave Hyatt.
 244
 245         Test: platform/mac/fast/css3-text/css3-text-decoration/text-decoration-thickness.html
 246
 247         * platform/graphics/FontMetrics.h: Adding variables for text decoration thickness and
 248         underline position.
 249         (WebCore::FontMetrics::FontMetrics):
 250         (WebCore::FontMetrics::decorationThickness):
 251         (WebCore::FontMetrics::setDecorationThickness):
 252         (WebCore::FontMetrics::underlinePosition):
 253         (WebCore::FontMetrics::setUnderlinePosition):
 254         * platform/graphics/freetype/SimpleFontDataFreeType.cpp:
 255         (WebCore::SimpleFontData::platformInit): Initialize new FontMetrics members.
 256         * platform/graphics/ios/SimpleFontDataIOS.mm:
 257         (WebCore::SimpleFontData::platformInit): Ditto.
 258         * platform/graphics/mac/SimpleFontDataMac.mm:
 259         (WebCore::SimpleFontData::platformInit): Ditto.
 260         * platform/graphics/win/SimpleFontDataCGWin.cpp:
 261         (WebCore::SimpleFontData::platformInit): Ditto.
 262         * platform/graphics/win/SimpleFontDataCairoWin.cpp:
 263         (WebCore::SimpleFontData::platformInit): Ditto.
 264         * platform/graphics/win/SimpleFontDataWin.cpp:
 265         (WebCore::SimpleFontData::initGDIFont): Ditto.
 266         * rendering/InlineTextBox.cpp:
 267         (WebCore::InlineTextBox::paintDecoration): Use FontMetrics data.
 268         * style/InlineTextBoxStyle.cpp:
 269         (WebCore::computeUnderlineOffset): Ditto.
 270         (WebCore::visualOverflowForDecorations): Ditto.
 271         * style/InlineTextBoxStyle.h:
 272         (WebCore::textDecorationStrokeThickness): Deleted.
 273         * svg/SVGFontData.cpp:
 274         (WebCore::SVGFontData::initializeFontData): Initialize new FontMetrics members.
 275
 276 2014-11-17  David Hyatt  <hyatt@apple.com>
 277
 278         Improve Ruby selection (getting rid of overlap and improving gap filling)
 279         https://bugs.webkit.org/show_bug.cgi?id=138250
 280
 281         Reviewed by Dean Jackson.
 282
 283         * rendering/RenderBlock.cpp:
 284         (WebCore::RenderBlock::selectionGaps):
 285         (WebCore::RenderBlock::blockSelectionGaps):
 286         Add Ruby text in along with the Ruby base.
 287
 288         * rendering/RenderBlockFlow.cpp:
 289         (WebCore::RenderBlockFlow::inlineSelectionGaps):
 290         Don't let block gaps get filled in here.
 291
 292         * rendering/RenderRubyBase.h:
 293         Expose accessor to the ruby run.
 294
 295         * rendering/RenderRubyText.cpp:
 296         (WebCore::RenderRubyText::rubyRun):
 297         * rendering/RenderRubyText.h:
 298         Add accessor to the ruby run.
 299
 300         * rendering/RootInlineBox.cpp:
 301         (WebCore::RootInlineBox::selectionTop):
 302         (WebCore::RootInlineBox::selectionBottom):
 303         Improve selectionTop and selectionBottom of ruby bases to avoid the
 304         ruby text. Improve the selectionTop and selectionBottom of ruby texts
 305         to fill the gap up to the previous/next line as appropriate.
 306
 307 2014-11-18  Shivakumar JM  <shiva.jm@samsung.com>
 308
 309         crypto: Use sequence<> instead of KeyUsage[] in IDL.
 310         https://bugs.webkit.org/show_bug.cgi?id=138823
 311
 312         Reviewed by Darin Adler.
 313
 314         Use sequence<KeyUsage> instead of KeyUsage[] in crypto IDL as per spec: https://dvcs.w3.org/hg/webcrypto-api/raw-file/tip/spec/
 315         Overview.html#SubtleCrypto-method-generateKey
 316
 317         No new tests, no behavior change.
 318
 319         * crypto/SubtleCrypto.idl:
 320
 321 2014-11-18  Philippe Normand  <pnormand@igalia.com>
 322
 323         HRTFDatabaseLoader is not an absolute condition to run audioContext
 324         https://bugs.webkit.org/show_bug.cgi?id=138829
 325
 326         Reviewed by Jer Noble.
 327
 328         This patch is a port of the following Blink revision by
 329         <keonho07.kim@samsung.com>:
 330         <https://src.chromium.org/viewvc/blink?revision=167887&view=revision>
 331
 332         * Modules/webaudio/AudioContext.cpp:
 333         (WebCore::AudioContext::AudioContext):
 334         (WebCore::AudioContext::isRunnable): Deleted.
 335         * Modules/webaudio/AudioContext.h:
 336         (WebCore::AudioContext::hrtfDatabaseLoader): Deleted.
 337         * Modules/webaudio/AudioDestinationNode.cpp:
 338         (WebCore::AudioDestinationNode::render):
 339         * Modules/webaudio/OfflineAudioDestinationNode.cpp:
 340         (WebCore::OfflineAudioDestinationNode::offlineRender):
 341         * Modules/webaudio/PannerNode.cpp:
 342         (WebCore::PannerNode::PannerNode):
 343         (WebCore::PannerNode::process):
 344         (WebCore::PannerNode::initialize):
 345         (WebCore::PannerNode::setPanningModel):
 346         * Modules/webaudio/PannerNode.h:
 347         * Modules/webaudio/RealtimeAnalyser.cpp:
 348         * Modules/webaudio/RealtimeAnalyser.h:
 349         * platform/audio/HRTFDatabaseLoader.cpp:
 350         (WebCore::HRTFDatabaseLoader::createAndLoadAsynchronouslyIfNecessary):
 351
 352 2014-11-18  Commit Queue  <commit-queue@webkit.org>
 353
 354         Unreviewed, rolling out r176218.
 355         https://bugs.webkit.org/show_bug.cgi?id=138827
 356
 357         seems to have caused regressions on the perf bots (Requested
 358         by lajava on #webkit).
 359
 360         Reverted changeset:
 361
 362         "[CSS Grid Layout] Upgrade align-self and align-items parsing
 363         to CSS 3"
 364         https://bugs.webkit.org/show_bug.cgi?id=133359
 365         http://trac.webkit.org/changeset/176218
 366
 367 2014-11-18  Joanmarie Diggs  <jdiggs@igalia.com>
 368
 369         AX: [ATK] Crash getting the orientation of a MenuListOption after the MenuList was removed from the document
 370         https://bugs.webkit.org/show_bug.cgi?id=138727
 371
 372         Reviewed by Chris Fleizach.
 373
 374         AccessibilityMenuListOption::elementRect() returns the value of the
 375         grandparent MenuList, asserting that the grandparent exists with that
 376         role. But it is possible to have an existing MenuListOption and remove
 377         the element which had been backing that MenuList from the document.
 378         Adding null checks prior to the assertions prevents our crashing if the
 379         parent or grandparent was removed.
 380
 381         Test: platform/gtk/accessibility/combobox-descendants-orientation-crash.html
 382
 383         * accessibility/AccessibilityMenuListOption.cpp:
 384         (WebCore::AccessibilityMenuListOption::elementRect):
 385
 386 2014-11-17  Benjamin Poulain  <bpoulain@apple.com>
 387
 388         Fix a small indentation mistake in SelectorCodeGenerator::generateNthFilterTest
 389
 390         * cssjit/SelectorCompiler.cpp:
 391         (WebCore::SelectorCompiler::SelectorCodeGenerator::generateNthFilterTest):
 392
 393 2014-11-17  Benjamin Poulain  <benjamin@webkit.org>
 394
 395         Fix two bad function names of HTMLFormControlElement
 396         https://bugs.webkit.org/show_bug.cgi?id=138790
 397
 398         Reviewed by Andreas Kling.
 399
 400         Darin suggested some name improvements in https://bugs.webkit.org/show_bug.cgi?id=138769
 401
 402         * dom/CheckedRadioButtons.cpp:
 403         (WebCore::RadioButtonGroup::add):
 404         (WebCore::RadioButtonGroup::updateCheckedState):
 405         (WebCore::RadioButtonGroup::requiredAttributeChanged):
 406         (WebCore::RadioButtonGroup::remove):
 407         (WebCore::RadioButtonGroup::updateValidityForAllButtons):
 408         (WebCore::RadioButtonGroup::setNeedsValidityCheckForAllButtons): Deleted.
 409         * html/FileInputType.cpp:
 410         (WebCore::FileInputType::setFiles):
 411         * html/HTMLButtonElement.cpp:
 412         (WebCore::HTMLButtonElement::computeWillValidate):
 413         (WebCore::HTMLButtonElement::recalcWillValidate): Deleted.
 414         * html/HTMLButtonElement.h:
 415         * html/HTMLFieldSetElement.h:
 416         * html/HTMLFormControlElement.cpp:
 417         (WebCore::HTMLFormControlElement::requiredAttributeChanged):
 418         (WebCore::HTMLFormControlElement::computeWillValidate):
 419         (WebCore::HTMLFormControlElement::willValidate):
 420         (WebCore::HTMLFormControlElement::setNeedsWillValidateCheck):
 421         (WebCore::HTMLFormControlElement::isValidFormControlElement):
 422         (WebCore::HTMLFormControlElement::updateValidity):
 423         (WebCore::HTMLFormControlElement::setCustomValidity):
 424         (WebCore::HTMLFormControlElement::recalcWillValidate): Deleted.
 425         (WebCore::HTMLFormControlElement::setNeedsValidityCheck): Deleted.
 426         * html/HTMLFormControlElement.h:
 427         * html/HTMLInputElement.cpp:
 428         (WebCore::HTMLInputElement::runPostTypeUpdateTasks):
 429         (WebCore::HTMLInputElement::parseAttribute):
 430         (WebCore::HTMLInputElement::setChecked):
 431         (WebCore::HTMLInputElement::setValueInternal):
 432         (WebCore::HTMLInputElement::setValueFromRenderer):
 433         (WebCore::HTMLInputElement::computeWillValidate):
 434         (WebCore::HTMLInputElement::parseMaxLengthAttribute):
 435         (WebCore::HTMLInputElement::recalcWillValidate): Deleted.
 436         * html/HTMLInputElement.h:
 437         * html/HTMLKeygenElement.h:
 438         * html/HTMLOutputElement.h:
 439         * html/HTMLSelectElement.cpp:
 440         (WebCore::HTMLSelectElement::deselectItems):
 441         (WebCore::HTMLSelectElement::optionSelectedByUser):
 442         (WebCore::HTMLSelectElement::listBoxSelectItem):
 443         (WebCore::HTMLSelectElement::add):
 444         (WebCore::HTMLSelectElement::parseAttribute):
 445         (WebCore::HTMLSelectElement::childrenChanged):
 446         (WebCore::HTMLSelectElement::optionElementChildrenChanged):
 447         (WebCore::HTMLSelectElement::setLength):
 448         (WebCore::HTMLSelectElement::selectAll):
 449         (WebCore::HTMLSelectElement::updateListBoxSelection):
 450         (WebCore::HTMLSelectElement::selectOption):
 451         (WebCore::HTMLSelectElement::restoreFormControlState):
 452         (WebCore::HTMLSelectElement::parseMultipleAttribute):
 453         (WebCore::HTMLSelectElement::reset):
 454         * html/HTMLTextAreaElement.cpp:
 455         (WebCore::HTMLTextAreaElement::parseAttribute):
 456         (WebCore::HTMLTextAreaElement::subtreeHasChanged):
 457         (WebCore::HTMLTextAreaElement::setValue):
 458         (WebCore::HTMLTextAreaElement::setNonDirtyValue):
 459
 460 2014-11-17  Sukolsak Sakshuwong  <sukolsak@gmail.com>
 461
 462         Add parsing for :role()
 463         https://bugs.webkit.org/show_bug.cgi?id=138310
 464
 465         Reviewed by Benjamin Poulain.
 466
 467         Add support for parsing :role() pseudo class. The implementation of selector
 468         matching will be in a follow-up patch.
 469
 470         * css/CSSGrammar.y.in:
 471         * css/CSSParser.cpp:
 472         (WebCore::CSSParser::detectFunctionTypeToken):
 473         * css/CSSSelector.cpp:
 474         (WebCore::appendPseudoClassFunctionTail):
 475         (WebCore::CSSSelector::selectorText):
 476         * css/CSSSelector.h:
 477         * css/SelectorChecker.cpp:
 478         (WebCore::SelectorChecker::checkOne):
 479         * css/SelectorPseudoClassAndCompatibilityElementMap.in:
 480         * cssjit/SelectorCompiler.cpp:
 481         (WebCore::SelectorCompiler::addPseudoClassType):
 482
 483 2014-11-17  Simon Fraser  <simon.fraser@apple.com>
 484
 485         [iOS WK1] Sometimes, missing tiles in -webkit-overflow-scrolling: touch in UIWebViews
 486         https://bugs.webkit.org/show_bug.cgi?id=138815
 487         rdar://problem/18746203
 488
 489         Reviewed by Tim Horton.
 490
 491         The scroll-velocity-related data members on FrameView, added in r168233, were uninitialized in UIWebViews,
 492         so FrameView::computeCoverageRect() could return garbage results, leading to too much or too little tile coverage.
 493
 494         We still add additional coverage from the new inclusion of a margin area, but, for simplicity, leave that the same
 495         between WK1 and WK2 for now.
 496
 497         * page/FrameView.cpp:
 498         (WebCore::FrameView::FrameView):
 499
 500 2014-11-17  Chris Dumez  <cdumez@apple.com>
 501
 502         Add initial layout testing coverage for timer throttling
 503         https://bugs.webkit.org/show_bug.cgi?id=138809
 504
 505         Reviewed by Simon Fraser.
 506
 507         Add isTimerThrottled() API to Internals so that we can add layout test
 508         coverage for timer throttling.
 509
 510         Tests: fast/dom/nested-timer-invisible-element-throttling.html
 511                fast/dom/nested-timer-visible-element-throttling.html
 512                fast/dom/repeating-timer-invisible-element-throttling.html
 513                fast/dom/repeating-timer-visible-element-throttling.html
 514
 515         * page/DOMTimer.h:
 516         * testing/Internals.cpp:
 517         (WebCore::Internals::isTimerThrottled):
 518         * testing/Internals.h:
 519         * testing/Internals.idl:
 520
 521 2014-11-17  Zalan Bujtas  <zalan@apple.com>
 522
 523         Simple line layout: Rename FlowContentIterator and move implementation to SimpleLineLayoutFlowContents.cpp
 524         https://bugs.webkit.org/show_bug.cgi?id=138799
 525
 526         Reviewed by Antti Koivisto.
 527
 528         This is in preparation to support multiple renderers.
 529         The FlowContents class will be used not just when we create runs, but also when we resolve them at painting time.
 530
 531         No change in functionality.
 532
 533         * CMakeLists.txt:
 534         * WebCore.vcxproj/WebCore.vcxproj:
 535         * WebCore.vcxproj/WebCore.vcxproj.filters:
 536         * WebCore.xcodeproj/project.pbxproj:
 537         * rendering/SimpleLineLayout.cpp:
 538         (WebCore::SimpleLineLayout::removeTrailingWhitespace):
 539         (WebCore::SimpleLineLayout::initializeNewLine):
 540         (WebCore::SimpleLineLayout::splitFragmentToFitLine):
 541         (WebCore::SimpleLineLayout::nextFragment):
 542         (WebCore::SimpleLineLayout::createLineRuns):
 543         (WebCore::SimpleLineLayout::closeLineEndingAndAdjustRuns):
 544         (WebCore::SimpleLineLayout::createTextRuns):
 545         (WebCore::SimpleLineLayout::FlowContentIterator::FlowContentIterator): Deleted.
 546         (WebCore::SimpleLineLayout::FlowContentIterator::findNextBreakablePosition): Deleted.
 547         (WebCore::SimpleLineLayout::FlowContentIterator::findNextNonWhitespacePosition): Deleted.
 548         (WebCore::SimpleLineLayout::FlowContentIterator::textWidth): Deleted.
 549         (WebCore::SimpleLineLayout::FlowContentIterator::isNewlineCharacter): Deleted.
 550         (WebCore::SimpleLineLayout::FlowContentIterator::isEndOfContent): Deleted.
 551         (WebCore::SimpleLineLayout::FlowContentIterator::style): Deleted.
 552         (WebCore::SimpleLineLayout::computeLineLeft): Deleted.
 553         * rendering/SimpleLineLayoutFlowContents.cpp: Added.
 554         (WebCore::SimpleLineLayout::FlowContents::FlowContents):
 555         (WebCore::SimpleLineLayout::FlowContents::findNextBreakablePosition):
 556         (WebCore::SimpleLineLayout::FlowContents::findNextNonWhitespacePosition):
 557         (WebCore::SimpleLineLayout::FlowContents::textWidth):
 558         * rendering/SimpleLineLayoutFlowContents.h: Added.
 559         (WebCore::SimpleLineLayout::FlowContents::Style::Style):
 560         (WebCore::SimpleLineLayout::FlowContents::style):
 561         (WebCore::SimpleLineLayout::FlowContents::isNewlineCharacter):
 562         (WebCore::SimpleLineLayout::FlowContents::isEndOfContent):
 563
 564 2014-11-17  Daniel Bates  <dabates@apple.com>
 565
 566         Unreviewed, rolling out r176232.
 567
 568         Broke the Mac build. Will investigate offline.
 569
 570         Reverted changeset:
 571
 572         "[iOS] Libdispatch, Dyld, IOKit, Mach, NSPointerFunctions,
 573         MobileGestalt"
 574         https://bugs.webkit.org/show_bug.cgi?id=138802
 575         http://trac.webkit.org/changeset/176232
 576
 577 2014-11-17  Daniel Bates  <dabates@apple.com>
 578
 579         [iOS] Libdispatch, Dyld, IOKit, Mach, NSPointerFunctions, MobileGestalt
 580         SPI are not available in public SDK
 581         https://bugs.webkit.org/show_bug.cgi?id=138802
 582
 583         Reviewed by David Kilzer.
 584
 585         Add SPI wrapper headers {Dispatch, DynamicLinker, IOPMLib, MachVM, NSPointerFunctions, MobileGestalt}SPI.h
 586         that forward declare applicable libdispatch, dyld, IOKit, mach, Foundation, MobileGestalt SPI
 587         and use these headers instead of the including private headers corresponding to these libraries/frameworks.
 588
 589         * WebCore.xcodeproj/project.pbxproj: Add private headers MobileGestaltSPI.h and
 590         MachVMSPI.h, and project headers DynamicLinkerSPI.h, NSPointerFunctionsSPI.h,
 591         DispatchSPI.h and IOPMLibSPI.h. Also, sort files in Xcode project.
 592         * bindings/objc/DOMInternal.mm: Include header NSPointerFunctionsSPI.h.
 593         * bridge/objc/objc_instance.mm: Ditto; Also, sort #import directives.
 594         * platform/cocoa/DisplaySleepDisablerCocoa.cpp: Substitute header IOPMLibSPI.h
 595         for private header IOKit/pwr_mgt/IOPMLib.h.
 596         * platform/cocoa/MemoryPressureHandlerCocoa.mm: Substitute header DispatchSPI.h
 597         for private header dispatch/private.h.
 598         * platform/graphics/ios/FontServicesIOS.mm: Substitute header DynamicLinkerSPI.h
 599         for private header mach-o/dyld_priv.h.
 600         * platform/ios/PlatformScreenIOS.mm: Substitute header MobileGestaltSPI.h for
 601         private header MobileGestalt.h.
 602         * platform/spi/cocoa/DispatchSPI.h: Added.
 603         * platform/spi/cocoa/DynamicLinkerSPI.h: Added.
 604         * platform/spi/cocoa/IOPMLibSPI.h: Added.
 605         * platform/spi/cocoa/MachVMSPI.h: Added.
 606         * platform/spi/cocoa/NSPointerFunctionsSPI.h: Added.
 607         * platform/spi/ios/MobileGestaltSPI.h: Added.
 608
 609 2014-11-17  Daniel Bates  <dabates@apple.com>
 610
 611         [iOS] CoreText and CoreUI SPI are not available in public SDK
 612         https://bugs.webkit.org/show_bug.cgi?id=138760
 613
 614         Reviewed by David Kilzer.
 615
 616         Add a SPI wrapper header CoreUISPI.h that forward declares CoreUI SPI
 617         and use this header instead of including private headers of CoreUI directly.
 618
 619         Additionally, add more forward declarations for CoreText SPI.
 620
 621         * WebCore.xcodeproj/project.pbxproj: Add project header CoreUISPI.h.
 622         * platform/graphics/ios/FontCacheIOS.mm: Include header CoreTextSPI.h instead
 623         of including CoreText headers directly.
 624         * platform/graphics/mac/FontMac.mm: Include header CoreUISPI.h instead
 625         of including CoreUI headers directly.
 626         * platform/graphics/mac/GlyphPageTreeNodeMac.cpp: Remove unnecessary
 627         include of CoreText headers. These headers will be included by header
 628         CoreTextSPI.h (at the top of the file).
 629         * platform/graphics/mac/SimpleFontDataMac.mm: Remove forward declaration
 630         of CTFontDescriptorIsSystemUIFont() and include of header CTFontDescriptorPriv.h
 631         as this duplicates the logic in CoreTextSPI.h, which is included in this file.
 632         * platform/spi/cocoa/CoreTextSPI.h: Added more forward declarations. Simplified
 633         #include directives.
 634         * platform/spi/ios/CoreUISPI.h: Added.
 635         * rendering/RenderThemeIOS.mm: Move CoreUI externs from this file to file
 636         CoreUISPI.h and include header CoreUISPI.h.
 637
 638 2014-11-17  Zalan Bujtas  <zalan@apple.com>
 639
 640         Simple line layout: Drop 16bit support temporarily.
 641         https://bugs.webkit.org/show_bug.cgi?id=138796
 642
 643         Reviewed by Antti Koivisto.
 644
 645         This is in preparation to multiple renderers support.
 646         It requires code refactoring on top of the coming multiple renderers codebase.
 647         (Not expecting any significant coverage shrinking.)
 648
 649         * rendering/SimpleLineLayout.cpp:
 650         (WebCore::SimpleLineLayout::canUseFor):
 651         (WebCore::SimpleLineLayout::FlowContentIterator::findNextBreakablePosition):
 652         (WebCore::SimpleLineLayout::FlowContentIterator::findNextNonWhitespacePosition):
 653         (WebCore::SimpleLineLayout::FlowContentIterator::textWidth):
 654         (WebCore::SimpleLineLayout::removeTrailingWhitespace):
 655         (WebCore::SimpleLineLayout::initializeNewLine):
 656         (WebCore::SimpleLineLayout::splitFragmentToFitLine):
 657         (WebCore::SimpleLineLayout::nextFragment):
 658         (WebCore::SimpleLineLayout::createLineRuns):
 659         (WebCore::SimpleLineLayout::closeLineEndingAndAdjustRuns):
 660         (WebCore::SimpleLineLayout::createTextRuns):
 661         (WebCore::SimpleLineLayout::create):
 662
 663 2014-11-17  Tim Horton  <timothy_horton@apple.com>
 664
 665         DataDetectors PageOverlay callbacks can come in to the wrong overlay
 666         https://bugs.webkit.org/show_bug.cgi?id=138801
 667         <rdar://problem/18991369>
 668
 669         Reviewed by Anders Carlsson.
 670
 671         * page/PageOverlay.cpp:
 672         (WebCore::generatePageOverlayID):
 673         (WebCore::PageOverlay::PageOverlay):
 674         * page/PageOverlay.h:
 675         Give each PageOverlay a unique ID.
 676
 677 2014-11-17  Javier Fernandez  <jfernandez@igalia.com>
 678
 679         [CSS Grid Layout] Upgrade align-self and align-items parsing to CSS 3
 680         https://bugs.webkit.org/show_bug.cgi?id=133359
 681
 682         Reviewed by David Hyatt.
 683
 684         Broaden justify-self's parsing name as a preparation of migrating
 685         align-self, align-items and justify-items to the CSS 3 Alignment syntax.
 686
 687         The current naming was too tied to justify-self and needs to
 688         be broadened. This will reduce the follow-up implementations'
 689         patches.
 690
 691         Upgrade align-self and align-items parsing to CSS 3
 692
 693         This change migrates the 2 properties to the CSS 3 Alignment
 694         parsing. The new parsing is identical to how we parse
 695         'justify-self'. The 2 properties need to be migrated together
 696         as they are used in tandem in CSSComputedStyleDeclaration.
 697
 698         This change also removes EAlignItems as it is now unused.
 699
 700         Tests: css3/parse-align-items.html
 701                css3/parse-align-self.html
 702
 703         * css/CSSComputedStyleDeclaration.cpp:
 704         (WebCore::valueForItemPositionWithOverflowAlignment): Added. Builds the CSSValue for the for the alignment properties.
 705         (WebCore::ComputedStyleExtractor::propertyValue):
 706         * css/CSSParser.cpp:
 707         (WebCore::isKeywordPropertyID): Removed align-items and align-self from the list.
 708         (WebCore::isValidKeywordPropertyAndValue): Removed align-items and align-self from the list.
 709         (WebCore::CSSParser::parseValue): Added align-items and align-self to the list and call to the generic parsing fuction for the alignment properties.
 710         (WebCore::isBaselinePositionKeyword): Utility function added to identify Baseline keyword values.
 711         (WebCore::CSSParser::parseItemPositionOverflowPosition): Generic parsing fuction for the alignment properties.
 712         (WebCore::CSSParser::parseJustifySelf): Deleted.
 713         * css/CSSParser.h:
 714         * css/CSSPrimitiveValueMappings.h:
 715         (WebCore::CSSPrimitiveValue::CSSPrimitiveValue): Refactoring some mapping functions.
 716        (WebCore::CSSPrimitiveValue::operator ItemPosition): Added. Replace the old EAlignItems enumeration.
 717         (WebCore::CSSPrimitiveValue::operator OverflowAlignment): Added.
 718         (WebCore::CSSPrimitiveValue::operator EAlignItems): Deleted.
 719         (WebCore::CSSPrimitiveValue::operator EJustifySelf): Deleted.
 720         (WebCore::CSSPrimitiveValue::operator EJustifySelfOverflowAlignment): Deleted.
 721         * css/CSSPropertyNames.in:
 722         * css/CSSValueKeywords.in:
 723         * css/StyleResolver.cpp:
 724         (WebCore::StyleResolver::adjustRenderStyle): Added a new adjustment to resolve the "auto" vale on the alignment properties.
 725         (WebCore::StyleResolver::adjustStyleForAlignment): Added. Function to resolve the "auto" vale on the alignment properties.
 726         (WebCore::StyleResolver::applyProperty):
 727         (WebCore::isDisplayFlexibleBox): Deleted. Moved to the RenderStyle.
 728         (WebCore::isDisplayGridBox): Deleted. Moved to the RenderStyle.
 729         (WebCore::isDisplayFlexibleOrGridBox): Deleted. Moved to the RenderStyle.
 730         * css/StyleResolver.h:
 731         * rendering/RenderBox.cpp:
 732         (WebCore::flexItemHasStretchAlignment): Adapted to the new ItemPosition enum.
 733         * rendering/RenderFlexibleBox.cpp:
 734         (WebCore::RenderFlexibleBox::firstLineBaseline): Adapted to the new ItemPosition enum.
 735         (WebCore::resolveAlignment): Consider also that align-items could be "auto", since it's the default now.
 736         (WebCore::RenderFlexibleBox::styleDidChange): Adapted to the new ItemPosition enum.
 737         (WebCore::RenderFlexibleBox::alignmentForChild): Adapted to the new ItemPosition enum.
 738         (WebCore::RenderFlexibleBox::needToStretchChild): Adapted to the new ItemPosition enum.
 739         (WebCore::RenderFlexibleBox::layoutAndPlaceChildren): Adapted to the new ItemPosition enum.
 740         (WebCore::RenderFlexibleBox::alignChildren): Adapted to the new ItemPosition enum.
 741         * rendering/RenderFlexibleBox.h:
 742         * rendering/RenderFullScreen.cpp:
 743         (WebCore::createFullScreenStyle): Adapted to the new ItemPosition enum.
 744         * rendering/RenderMenuList.cpp:
 745         (WebCore::RenderMenuList::adjustInnerStyle): Adapted to the new ItemPosition enum.
 746         * rendering/mathml/RenderMathMLRoot.cpp:
 747         (WebCore::RenderMathMLRoot::updateStyle): Adapted to the new ItemPosition enum.
 748         * rendering/mathml/RenderMathMLScripts.cpp:
 749         (WebCore::RenderMathMLScripts::fixAnonymousStyleForSubSupPair): Adapted to the new ItemPosition enum.
 750         (WebCore::RenderMathMLScripts::fixAnonymousStyles): Adapted to the new ItemPosition enum.
 751         * rendering/style/RenderStyle.h: Adapted to the new ItemPosition enum.
 752         * rendering/style/RenderStyleConstants.h: Adapted to the new ItemPosition enum.
 753         * rendering/style/StyleRareNonInheritedData.cpp: Adapted to the new ItemPosition enum.
 754         (WebCore::StyleRareNonInheritedData::StyleRareNonInheritedData): Adapted to the new ItemPosition enum.
 755         (WebCore::StyleRareNonInheritedData::operator==): Adapted to the new ItemPosition enum.
 756         * rendering/style/StyleRareNonInheritedData.h: Adapted to the new ItemPosition enum.
 757         * style/StyleResolveTree.cpp:
 758         (WebCore::Style::determineChange): Changes in the alignItems property will cause a Detach.
 759
 760 2014-11-17  Shivakumar JM  <shiva.jm@samsung.com>
 761
 762         Attribute text in HTMLAnchorElement should behave as per specification.
 763         https://bugs.webkit.org/show_bug.cgi?id=138557
 764
 765         Reviewed by Darin Adler.
 766
 767         attribute text in HTMLAnchorElement should not be readonly as per specification
 768         http://www.w3.org/TR/html/text-level-semantics.html#dom-a-text. It should behave as textContent attribute.
 769         Also This matches the behavior of Chrome 38 and FireFox.
 770
 771         Test: fast/dom/HTMLAnchorElement/anchor-text-attribute.html
 772
 773         * html/HTMLAnchorElement.cpp:
 774         (WebCore::HTMLAnchorElement::text):
 775         (WebCore::HTMLAnchorElement::setText):
 776         * html/HTMLAnchorElement.h:
 777         * html/HTMLAnchorElement.idl:
 778
 779 2014-11-17  Chris Dumez  <cdumez@apple.com>
 780
 781         Throttle timers that change the style of elements outside the viewport
 782         https://bugs.webkit.org/show_bug.cgi?id=138292
 783
 784         Reviewed by Antti Koivisto.
 785
 786         Throttle timers that change the style of elements outside the viewport
 787         to 1 second, similarly to what was already done for timers interacting
 788         with non user observable plugins. To be conservative, we don't throttle
 789         timers that also cause DOM Tree modifications (e.g. adding/removing
 790         nodes, modify element attributes).
 791
 792         On huffingtonpost.com, the CPU usage is at ~17% when the top scrolling
 793         banner is inside the viewport on my machine. Without this patch, CPU
 794         usage would stay ~17% when the banner is outside the viewport. Thanks
 795         to timer throttling, CPU usage now goes down to ~1.5%, without user
 796         observable side effects. The timers get unthrottled when they are
 797         inside the viewport again (i.e. due to scrolling or layout).
 798
 799         On espn.com, the CPU usage goes down from ~7% at the top of the page
 800         to ~1% when scrolling to the bottom of the page. On ebay.com, CPU
 801         usage goes down from ~25% at the top of the page to less than 1% when
 802         scrolling to the bottom of the page.
 803
 804 2014-11-17  peavo@outlook.com  <peavo@outlook.com>
 805
 806         [WinCairo] Compile error, CoreGraphics header file included.
 807         https://bugs.webkit.org/show_bug.cgi?id=138795
 808
 809         Reviewed by Brent Fulgham.
 810
 811         CoreGraphics is not available on WinCairo.
 812
 813         * platform/graphics/win/FontCacheWin.cpp:
 814
 815 2014-11-17  Commit Queue  <commit-queue@webkit.org>
 816
 817         Unreviewed, rolling out r176169.
 818         https://bugs.webkit.org/show_bug.cgi?id=138797
 819
 820         Appears to have broken svg/custom/anchor-on-use.svg (Requested
 821         by ap on #webkit).
 822
 823         Reverted changeset:
 824
 825         "Attribute text in HTMLAnchorElement should behave as per
 826         specification."
 827         https://bugs.webkit.org/show_bug.cgi?id=138557
 828         http://trac.webkit.org/changeset/176169
 829
 830 2014-11-17  Daniel Bates  <dabates@apple.com>
 831
 832         [iOS] AVKit and MediaPlayer SPI are not available in public SDK
 833         https://bugs.webkit.org/show_bug.cgi?id=138756
 834
 835         Reviewed by Eric Carlson.
 836
 837         Add a SPI wrapper headers {AVKit, MPAVRoutingController}SPI.h that forward declares
 838         AVKit and MPAVRoutingController SPI, respectively, and use these header instead of
 839         including private headers of AVKit and MediaPlayer directly.
 840
 841         * WebCore.xcodeproj/project.pbxproj: Add project and private header AVKitSPI.h
 842         and MPAVRoutingControllerSPI.h, respectively.
 843         * platform/audio/ios/MediaSessionManagerIOS.mm: Include header MPAVRoutingControllerSPI.h
 844         instead of including the private header MPAVRoutingController.h.
 845         * platform/ios/WebVideoFullscreenInterfaceAVKit.mm: Include header AVKitSPI.h
 846         instead of including AVKit headers directly.
 847
 848 2014-11-17  Daniel Bates  <dabates@apple.com>
 849
 850         Attempt to fix the Mac build after <https://trac.webkit.org/changeset/176204>
 851         (https://bugs.webkit.org/show_bug.cgi?id=138751)
 852
 853         Move declaration of CAContext to be before the declaration of CALayer since
 854         the latter depends on the former.
 855
 856         * platform/spi/cocoa/QuartzCoreSPI.h:
 857
 858 2014-11-17  Daniel Bates  <dabates@apple.com>
 859
 860         [iOS] QuartzCore SPI not available in public SDK
 861         https://bugs.webkit.org/show_bug.cgi?id=138751
 862
 863         Reviewed by David Kilzer.
 864
 865         Add a SPI wrapper header called QuartzCoreSPI.h that forward declares
 866         QuartzCore SPI and use this header instead of including private headers
 867         of QuartzCore directly.
 868
 869         * WebCore.xcodeproj/project.pbxproj: Added private header QuartzCoreSPI.h.
 870         * platform/graphics/ca/mac/PlatformCAFiltersMac.h: Removed; incorporated its
 871         functionality into QuartzCoreSPI.h.
 872         * platform/graphics/ca/mac/PlatformCAFiltersMac.mm: Substitute header
 873         QuartzCoreSPI.h for PlatformCAFiltersMac.h as the content of the latter
 874         was incorporated into the former.
 875         * platform/graphics/ca/mac/PlatformCALayerMac.mm: Move CALayer and CATiledLayer
 876         forward declarations from this file to file QuartzCoreSPI.h and include
 877         header QuartzCoreSPI.h instead of including CATiledLayerPrivate.h directly. Also
 878         remove header PlatformCAFiltersMac.h as its content is incorporated into QuartzCoreSPI.h
 879         * platform/ios/LegacyTileCache.mm: Include headers CoreGraphicsSPI.h and QuartzCoreSPI.h
 880         instead of including QuartzCore headers directly.
 881         * platform/ios/LegacyTileGrid.mm: Include header QuartzCoreSPI.h instead
 882         of including QuartzCore headers directly. Also sort the list of #include directives.
 883         * platform/ios/LegacyTileGridTile.mm: Include header QuartzCoreSPI.h
 884         instead of including QuartzCore headers directly.
 885         * platform/spi/cocoa/QuartzCoreSPI.h: Added.
 886
 887 2014-11-17  Chris Dumez  <cdumez@apple.com>
 888
 889         Move more CSS properties to the new StyleBuilder
 890         https://bugs.webkit.org/show_bug.cgi?id=138766
 891
 892         Reviewed by Darin Adler.
 893
 894         Move more CSS properties from DeprecatedStyleBuilder to the new
 895         StyleBuilder:
 896         line-height
 897         word-spacing
 898         -webkit-marquee-repetition
 899         -webkit-text-underline-position
 900
 901         No new tests, no behavior change.
 902
 903         * css/CSSPropertyNames.in:
 904         * css/DeprecatedStyleBuilder.cpp:
 905         (WebCore::DeprecatedStyleBuilder::DeprecatedStyleBuilder):
 906         (WebCore::csstoLengthConversionDataWithTextZoomFactor): Deleted.
 907         (WebCore::ApplyPropertyMarqueeRepetition::applyValue): Deleted.
 908         (WebCore::ApplyPropertyMarqueeRepetition::createHandler): Deleted.
 909         (WebCore::ApplyPropertyTextUnderlinePosition::applyValue): Deleted.
 910         (WebCore::ApplyPropertyTextUnderlinePosition::createHandler): Deleted.
 911         (WebCore::ApplyPropertyLineHeight::applyValue): Deleted.
 912         (WebCore::ApplyPropertyLineHeight::createHandler): Deleted.
 913         (WebCore::ApplyPropertyLineHeightForIOSTextAutosizing::applyValue): Deleted.
 914         (WebCore::ApplyPropertyLineHeightForIOSTextAutosizing::applyInitialValue): Deleted.
 915         (WebCore::ApplyPropertyLineHeightForIOSTextAutosizing::applyInheritValue): Deleted.
 916         (WebCore::ApplyPropertyLineHeightForIOSTextAutosizing::createHandler): Deleted.
 917         (WebCore::ApplyPropertyWordSpacing::applyValue): Deleted.
 918         (WebCore::ApplyPropertyWordSpacing::createHandler): Deleted.
 919         * css/StyleBuilderConverter.h:
 920         (WebCore::StyleBuilderConverter::convertMarqueeRepetition):
 921         (WebCore::StyleBuilderConverter::convertTextUnderlinePosition):
 922         * css/StyleBuilderCustom.h:
 923         (WebCore::StyleBuilderFunctions::csstoLengthConversionDataWithTextZoomFactor):
 924         (WebCore::StyleBuilderFunctions::convertLineHeight):
 925         (WebCore::StyleBuilderFunctions::applyValueWordSpacing):
 926         (WebCore::StyleBuilderFunctions::applyInheritLineHeight):
 927         (WebCore::StyleBuilderFunctions::applyInitialLineHeight):
 928         (WebCore::StyleBuilderFunctions::applyValueLineHeight):
 929
 930 2014-11-16  Andreas Kling  <akling@apple.com>
 931
 932         Avoid synchronous style recalc when mutating a Node inside FrameSelection.
 933         <https://webkit.org/b/138791>
 934
 935         Reviewed by Antti Koivisto.
 936
 937         We can safely clear the RenderView's selection without doing a synchronous
 938         style recalc first.
 939
 940         This is a small progression on Speedometer locally.
 941
 942         * editing/FrameSelection.cpp:
 943         (WebCore::DragCaretController::nodeWillBeRemoved):
 944         (WebCore::FrameSelection::respondToNodeModification):
 945         (WebCore::clearRenderViewSelection): Deleted.
 946
 947             Merged clearRenderViewSelection() into the call sites and removed
 948             the now-unnecessary protection of the Document.
 949
 950 2014-11-17  Eric Carlson  <eric.carlson@apple.com>
 951
 952         [iOS] allow host application to opt-out of alternate fullscreen
 953         https://bugs.webkit.org/show_bug.cgi?id=138787
 954
 955         Reviewed by Darin Adler.
 956
 957         * Modules/mediacontrols/MediaControlsHost.cpp:
 958         (WebCore::MediaControlsHost::optimizedFullscreenSupported): Check mediaSession().allowsAlternateFullscreen.
 959
 960         * html/HTMLMediaElement.cpp:
 961         (WebCore::HTMLMediaElement::enterFullscreenOptimized): Do nothing if
 962             mediaSession().allowsAlternateFullscreen says it isn't allowed.
 963
 964         * html/HTMLMediaSession.cpp:
 965         (WebCore::HTMLMediaSession::allowsAlternateFullscreen): New, check settings.
 966         * html/HTMLMediaSession.h:
 967
 968         * page/Settings.cpp: Add allowsAlternateFullscreen.
 969         * page/Settings.in: Ditto.
 970
 971 2014-11-16  Zan Dobersek  <zdobersek@igalia.com>
 972
 973         Replace RenderSVGResource::cast<T>() with downcast<T>()
 974         https://bugs.webkit.org/show_bug.cgi?id=138290
 975
 976         Reviewed by Chris Dumez.
 977
 978         Add type traits for the RenderSVGResource hierarchy. This enables
 979         using the typecasting framework in getRenderSVGResourceById() that
 980         in return renders the RenderSVGResource::cast<>() template and all
 981         the static s_resourceType member variables in the hierarchy removable.
 982
 983         * WebCore.order:
 984         * rendering/svg/RenderSVGResource.h:
 985         (WebCore::RenderSVGResource::cast): Deleted.
 986         * rendering/svg/RenderSVGResourceClipper.h:
 987         * rendering/svg/RenderSVGResourceContainer.h:
 988         (WebCore::getRenderSVGResourceById):
 989         * rendering/svg/RenderSVGResourceFilter.h:
 990         (isType):
 991         * rendering/svg/RenderSVGResourceLinearGradient.h:
 992         * rendering/svg/RenderSVGResourceMarker.h:
 993         * rendering/svg/RenderSVGResourceMasker.h:
 994         * rendering/svg/RenderSVGResourcePattern.h:
 995         * rendering/svg/RenderSVGResourceRadialGradient.h:
 996         * rendering/svg/RenderSVGResourceSolidColor.h:
 997
 998 2014-11-16  Benjamin Poulain  <benjamin@webkit.org>
 999
1000         Implement :valid and :invalid matching for the fieldset element
1001         https://bugs.webkit.org/show_bug.cgi?id=138769
1002
1003         Reviewed by Darin Adler.
1004
1005         In the latest HTML spec, the pseudo classes :valid and :invalid match
1006         a fieldset element based on its descendants:
1007             https://html.spec.whatwg.org/#selector-valid
1008             https://html.spec.whatwg.org/#selector-invalid
1009
1010         This patch adds that behavior.
1011
1012         There are two key problems to solve with these pseudo classes on fieldset:
1013         -Efficient matching.
1014         -Style invalidation when any of the descendant changes.
1015
1016         To implement the style invalidation, I have modified HTMLFormControlElement
1017         to notify its ancestor when its state changes.
1018
1019         The first change is making the state fully internal to HTMLFormControlElement,
1020         we do not want subclass to be able to change the behavior and forget to update
1021         the ancestors.
1022
1023         To achieve that encapsulation, the interface was changed a bit:
1024         -Neither willValidate() nor isValidFormControlElement() inherit from Element.
1025          Instead, willValidate() is the implementation of FormAssociatedElement's interface
1026          and it is final. The method isValidFormControlElement() becomes completely internal
1027          to HTMLFormControlElement.
1028         -Since willValidate() should no longer be re-implemented by subclass, the elements
1029          that were depending on it have been migrated to recalcWillValidate() to set
1030          the initial state as needed.
1031
1032         With the validity state fully encapsulated in HTMLFormControlElement, all I need
1033         is a way to communicate that information to HTMLFieldSetElement ancestors.
1034         This is done in two cases:
1035         -The validity state changes.
1036         -The tree changes in a way that would make the input element not a descendant
1037          of a HTMLFieldSetElement.
1038
1039         The invalidation is simply done by walking up the ancestors and adding the current
1040         element to a "validity dependency list" on each HTMLFieldSetElement.
1041
1042         Tests: fast/css/pseudo-invalid-fieldset-invalidation-optimization.html
1043                fast/css/pseudo-invalid-fieldset-style-sharing.html
1044                fast/css/pseudo-invalid-fieldset.html
1045                fast/css/pseudo-valid-fieldset-invalidation-optimization.html
1046                fast/css/pseudo-valid-fieldset-style-sharing.html
1047                fast/css/pseudo-valid-fieldset.html
1048                fast/selectors/invalid-fieldset-style-update-1.html
1049                fast/selectors/invalid-fieldset-style-update-2.html
1050                fast/selectors/invalid-fieldset-style-update-3.html
1051                fast/selectors/invalid-fieldset-style-update-4.html
1052                fast/selectors/invalid-fieldset-style-update-5.html
1053                fast/selectors/valid-fieldset-style-update-1.html
1054                fast/selectors/valid-fieldset-style-update-2.html
1055                fast/selectors/valid-fieldset-style-update-3.html
1056                fast/selectors/valid-fieldset-style-update-4.html
1057                fast/selectors/valid-fieldset-style-update-5.html
1058
1059         * css/SelectorCheckerTestFunctions.h:
1060         (WebCore::isInRange):
1061         (WebCore::isOutOfRange):
1062         (WebCore::isInvalid):
1063         (WebCore::isValid):
1064         The hack "ContainsValidityStyleRules" is in the way of correct styling
1065         of FieldSet and Form.
1066         It is not the right way to get stylesheet properties anyway.
1067
1068         * css/StyleResolver.cpp:
1069         (WebCore::StyleResolver::canShareStyleWithControl):
1070         Make sure style sharing does not incorrectly share style for fieldset elements.
1071
1072         * dom/Document.cpp:
1073         (WebCore::Document::Document):
1074         * dom/Document.h:
1075         (WebCore::Document::containsValidityStyleRules): Deleted.
1076         (WebCore::Document::setContainsValidityStyleRules): Deleted.
1077         * dom/Element.h:
1078         (WebCore::Element::matchesValidPseudoClass):
1079         (WebCore::Element::matchesInvalidPseudoClass):
1080         (WebCore::Element::willValidate): Deleted.
1081         (WebCore::Element::isValidFormControlElement): Deleted.
1082         * html/FormAssociatedElement.cpp:
1083         (WebCore::FormAssociatedElement::customError):
1084         * html/FormAssociatedElement.h:
1085
1086         * html/HTMLFieldSetElement.cpp:
1087         (WebCore::HTMLFieldSetElement::matchesValidPseudoClass):
1088         (WebCore::HTMLFieldSetElement::matchesInvalidPseudoClass):
1089         (WebCore::HTMLFieldSetElement::addInvalidDescendant):
1090         (WebCore::HTMLFieldSetElement::removeInvalidDescendant):
1091         Each HTMLFormControlElement that has constraint validation adds or removes
1092         itself from its HTMLFieldSetElement ancestors.
1093
1094         It should be possible to just keep track of a count instead of a HashSet.
1095         I decided to got with the HashSet to make the code more robust and easier
1096         to debug. A few assertions ensure that the HashSet is actually used as a counter.
1097
1098         * html/HTMLFieldSetElement.h:
1099         * html/HTMLFormControlElement.cpp:
1100         (WebCore::addInvalidElementToAncestorFromInsertionPoint):
1101         (WebCore::removeInvalidElementToAncestorFromInsertionPoint):
1102
1103         (WebCore::HTMLFormControlElement::insertedInto):
1104         (WebCore::HTMLFormControlElement::removedFrom):
1105         One tricky part of those two functions is that we cannot use
1106         matchesValidPseudoClass() or matchesInvalidPseudoClass().
1107
1108         The reason is that HTMLFieldSetElement is a subclass of HTMLFormControlElement
1109         and it has its own definition of what Valid and Invalid mean when matching selectors.
1110
1111         In HTMLFormControlElement, we must use the internal state,
1112         willValidate() and isValidFormControlElement() must be used directly.
1113
1114         (WebCore::HTMLFormControlElement::matchesValidPseudoClass):
1115         (WebCore::HTMLFormControlElement::matchesInvalidPseudoClass):
1116         (WebCore::HTMLFormControlElement::willValidate):
1117         (WebCore::addInvalidElementToAncestors):
1118         (WebCore::removeInvalidElementFromAncestors):
1119         (WebCore::HTMLFormControlElement::setNeedsWillValidateCheck):
1120         (WebCore::HTMLFormControlElement::setNeedsValidityCheck):
1121         (WebCore::HTMLFormControlElement::isValidFormControlElement): Deleted.
1122         * html/HTMLFormControlElement.h:
1123         (WebCore::HTMLFormControlElement::isValidFormControlElement):
1124         * html/HTMLKeygenElement.h:
1125         * html/HTMLObjectElement.h:
1126         * html/HTMLOutputElement.h:
1127
1128 2014-11-16  Zan Dobersek  <zdobersek@igalia.com>
1129
1130         [TexMap] Add typecasting support for GraphicsLayerTextureMapper
1131         https://bugs.webkit.org/show_bug.cgi?id=138741
1132
1133         Reviewed by Martin Robinson.
1134
1135         Add type traits for the GraphicsLayerTextureMapper class.
1136         This enables using typecasting functions instead of
1137         toGraphicsLayerTextureMapper().
1138
1139         Remove toTextureMapperLayer() casting function which actually
1140         returned the TextureMapperLayer object that belonged to the
1141         passed-in GraphicsLayerTextureMapper. Instead, the GraphicsLayer
1142         object is now downcasted and the TextureMapperLayer is accessed
1143         by calling the GraphicsLayerTextureMapper::layer() method.
1144
1145         * platform/graphics/GraphicsLayer.h:
1146         (WebCore::GraphicsLayer::isGraphicsLayerTextureMapper):
1147         * platform/graphics/texmap/GraphicsLayerTextureMapper.cpp:
1148         (WebCore::toTextureMapperLayerVector):
1149         (WebCore::GraphicsLayerTextureMapper::commitLayerChanges):
1150         (WebCore::toTextureMapperLayer): Deleted.
1151         * platform/graphics/texmap/GraphicsLayerTextureMapper.h:
1152         (WebCore::toGraphicsLayerTextureMapper): Deleted.
1153
1154 2014-11-16  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
1155
1156         Move WebCore/loader to std::unique_ptr
1157         https://bugs.webkit.org/show_bug.cgi?id=138646
1158
1159         Reviewed by Chris Dumez.
1160
1161         Replace OwnPtr with std::unique_ptr<> and std::make_unique<>.
1162         However this patch leaves ResourceRequest::adopt() because this need
1163         to be handled by different patch.
1164
1165         No new tests, no behavior changes.
1166
1167         * loader/DocumentThreadableLoader.cpp:
1168         (WebCore::DocumentThreadableLoader::makeCrossOriginAccessRequest):
1169         (WebCore::DocumentThreadableLoader::preflightSuccess):
1170         * loader/DocumentThreadableLoader.h:
1171         * loader/SubresourceLoader.cpp:
1172         (WebCore::SubresourceLoader::SubresourceLoader):
1173         (WebCore::SubresourceLoader::didReceiveResponse):
1174         (WebCore::SubresourceLoader::notifyDone):
1175         * loader/SubresourceLoader.h:
1176         * loader/TextTrackLoader.h:
1177         * loader/appcache/ApplicationCacheHost.h:
1178         * loader/appcache/ApplicationCacheStorage.cpp:
1179         (WebCore::ApplicationCacheStorage::storeCopyOfCache):
1180         * loader/archive/cf/LegacyWebArchive.cpp:
1181         (WebCore::LegacyWebArchive::create):
1182         * loader/cache/CachedResourceLoader.cpp:
1183         (WebCore::CachedResourceLoader::requestPreload):
1184         (WebCore::CachedResourceLoader::clearPreloads):
1185         (WebCore::CachedResourceLoader::printPreloadStats):
1186         * loader/cache/CachedResourceLoader.h:
1187         * loader/icon/IconDatabase.cpp:
1188         (WebCore::IconDatabase::deleteAllPreparedStatements):
1189         (WebCore::readySQLiteStatement):
1190         * loader/icon/IconDatabase.h:
1191         * loader/icon/IconRecord.h:
1192
1193 2014-11-16  Chris Dumez  <cdumez@apple.com>
1194
1195         Crash when setting 'order' CSS property to a calculated value
1196         https://bugs.webkit.org/show_bug.cgi?id=138780
1197
1198         Reviewed by Darin Adler.
1199
1200         CSS Calculated values were not handled by the CSS parser for 'order'
1201         CSS property. As a result, using calculated values wouldn't work in
1202         release builds and would hit an assertion in debug builds.
1203
1204         This patch updates the CSS parser to directly convert the
1205         CSS Calculated value into a simple integer CSSPrimitiveValue for
1206         'order' property. We could have marked CSS Calculated values as
1207         valid in the CSS Parser instead but this would have brought issues:
1208         - The calculated value needs to be adjusted to INT_MIN + 2 if it is less
1209           than that. This would force us to calculate the expression anyway.
1210         - The StyleBuilder would need updating to properly handle CSS Calculated
1211           values for 'order'.
1212
1213         Test: fast/css/order-calculated-value.html
1214
1215         * css/CSSParser.cpp:
1216         (WebCore::CSSParser::parseValue):
1217
1218 2014-11-16  Chris Dumez  <cdumez@apple.com>
1219
1220         Assertion hit when setting a very large value to 'border-width' / 'font-size' CSS properties
1221         https://bugs.webkit.org/show_bug.cgi?id=138770
1222
1223         Reviewed by Darin Adler.
1224
1225         When setting a very large value to a CSS property, it is represented internally as
1226         infinity. r166114 already tried to deal with this by adding an std::isinf() check
1227         in CSSValuePool::createValue() and returning an identifier CSSPrimitiveValue with
1228         CSSValueInvalid value in such case. The issue is that doing leads to the
1229         StyleBuilder getting a CSSValueInvalid CSSPrimitive value as input, which is not
1230         handled and leads to assertions.
1231
1232         This patch updates the CSSParser to detect cases where the double value is
1233         infinity earlier (in CSSParser::validUnit() and parseSimpleLengthValue()), so
1234         that we mark the value as invalid and actually drop it. As a result,
1235         CSSPrimitiveValues with CSSValueInvalid value no longer make their way to the
1236         StyleBuilder.
1237
1238         Test: fast/css/style-builder-infinite-value.html
1239               fast/css/infinite-floating-value.html
1240
1241         * css/CSSParser.cpp:
1242         (WebCore::parseSimpleLengthValue):
1243         (WebCore::CSSParser::validUnit):
1244         * css/CSSValuePool.cpp:
1245         (WebCore::CSSValuePool::createValue):
1246
1247 2014-11-16  Shivakumar JM  <shiva.jm@samsung.com>
1248
1249         Attribute text in HTMLAnchorElement should behave as per specification.
1250         https://bugs.webkit.org/show_bug.cgi?id=138557
1251
1252         Reviewed by Darin Adler.
1253
1254         attribute text in HTMLAnchorElement should not be readonly as per specification
1255         http://www.w3.org/TR/html/text-level-semantics.html#dom-a-text. It should behave as textContent attribute.
1256         Also This matches the behavior of Chrome 38 and FireFox.
1257
1258         Test: fast/dom/HTMLAnchorElement/anchor-text-attribute.html
1259
1260         * html/HTMLAnchorElement.cpp:
1261         (WebCore::HTMLAnchorElement::text):
1262         (WebCore::HTMLAnchorElement::setText):
1263         * html/HTMLAnchorElement.h:
1264         * html/HTMLAnchorElement.idl:
1265
1266 2014-11-16  Chris Dumez  <cdumez@apple.com>
1267
1268         Move 'vertical-align' CSS property to the new StyleBuilder
1269         https://bugs.webkit.org/show_bug.cgi?id=138764
1270
1271         Reviewed by Darin Adler.
1272
1273         Move 'vertical-align' CSS property from DeprecatedStyleBuilder to the
1274         new StyleBuilder by using custom code.
1275
1276         No new tests, no behavior change.
1277
1278         * css/CSSPropertyNames.in:
1279         * css/DeprecatedStyleBuilder.cpp:
1280         (WebCore::DeprecatedStyleBuilder::DeprecatedStyleBuilder):
1281         (WebCore::ApplyPropertyVerticalAlign::applyValue): Deleted.
1282         (WebCore::ApplyPropertyVerticalAlign::createHandler): Deleted.
1283         * css/StyleBuilderCustom.h:
1284         (WebCore::StyleBuilderFunctions::applyValueVerticalAlign):
1285
1286 2014-11-16  Tim Horton  <timothy_horton@apple.com>
1287
1288         Use TextIndicator instead of the built in Lookup highlight
1289         https://bugs.webkit.org/show_bug.cgi?id=138775
1290         <rdar://problem/18953982>
1291
1292         Follow-up to r176164.
1293
1294         * platform/spi/mac/LookupSPI.h:
1295         (WebCore::canDisableLookupIndicator): Deleted.
1296         Remove canDisableLookupIndicator.
1297
1298 2014-11-16  Chris Dumez  <cdumez@apple.com>
1299
1300         Crash when setting '-webkit-line-clamp' CSS property to a calculated value
1301         https://bugs.webkit.org/show_bug.cgi?id=138777
1302
1303         Reviewed by Benjamin Poulain.
1304
1305         CSS Calculated values were not handled when converting a CSSPrimitiveValue
1306         to a LineClampValue. As a result, we would hit an ASSERT_NOT_REACHED()
1307         assertion. To address this issue, this patch updates the code to use
1308         primitiveType() / getValue<>() instead of m_primitiveUnitType /
1309         m_value.num, as those getter function properly handle CSS Calculated
1310         values.
1311
1312         Test: fast/css/webkit-line-clamp-calculated-value.html
1313
1314         * css/CSSPrimitiveValueMappings.h:
1315         (WebCore::CSSPrimitiveValue::operator LineClampValue):
1316
1317 2014-11-16  Tim Horton  <timothy_horton@apple.com>
1318
1319         Use TextIndicator instead of the built in Lookup highlight
1320         https://bugs.webkit.org/show_bug.cgi?id=138775
1321         <rdar://problem/18953982>
1322
1323         Reviewed by Anders Carlsson.
1324
1325         * WebCore.exp.in:
1326         * platform/mac/WebCoreSystemInterface.h:
1327         * platform/mac/WebCoreSystemInterface.mm:
1328         Remove a WKSI import/export that we don't need anymore.
1329
1330         * platform/spi/mac/LookupSPI.h: Added.
1331         (WebCore::canDisableLookupIndicator):
1332         Add Lookup SPI header.
1333         Add a 'canDisableLookupIndicator', which must be called in any translation
1334         unit that wants to use either of the soft-linked constants related to
1335         that functionality, before calling the get* function.
1336
1337         * WebCore.xcodeproj/project.pbxproj:
1338         * editing/mac/DictionaryLookup.mm:
1339         (WebCore::rangeForDictionaryLookupForSelection):
1340         (WebCore::rangeForDictionaryLookupAtHitTestResult):
1341         Adopt the Lookup SPI header.
1342
1343 2014-11-16  Joanmarie Diggs  <jdiggs@igalia.com>
1344
1345         AX: [ATK] REGRESSION(r166288): Accessible values are no longer accessible via AT-SPI2
1346         https://bugs.webkit.org/show_bug.cgi?id=138776
1347
1348         Reviewed by Chris Fleizach.
1349
1350         r166288 used #if...#else instead of #if...#endif, thus disabling the old
1351         AtkValue support. Unfortunately AT-SPI2 doesn't yet have support for the
1352         new AtkValue API. Thus r166288 broke access to value-holding elements for
1353         AT-SPI2 assistive technologies. Replacing #else with #endif fixes things.
1354
1355         No new tests because catching this regression would require AT-SPI2.
1356
1357         * accessibility/atk/WebKitAccessibleInterfaceValue.cpp:
1358         (webkitAccessibleValueGetMinimumIncrement):
1359         (webkitAccessibleValueInterfaceInit):
1360
1361 2014-11-16  Joanmarie Diggs  <jdiggs@igalia.com>
1362
1363         AX: [ATK] Expose the blockquote element using ATK_ROLE_BLOCK_QUOTE
1364         https://bugs.webkit.org/show_bug.cgi?id=138771
1365
1366         Reviewed by Chris Fleizach.
1367
1368         Add BlockquoteRole and assigns it to blockquote elements. Make
1369         corresponding changes to platform role mappings: ATK_ROLE_BLOCK_QUOTE
1370         for GTK and EFL; AXGroup (which was being used already) for the Mac.
1371
1372         No new tests are required. Instead, updated the existing expectations
1373         to reflect the new mapping.
1374
1375         * accessibility/AccessibilityNodeObject.cpp:
1376         (WebCore::AccessibilityNodeObject::determineAccessibilityRole):
1377         * accessibility/AccessibilityObject.cpp:
1378         (WebCore::AccessibilityObject::isBlockquote):
1379         * accessibility/AccessibilityObject.h:
1380         * accessibility/AccessibilityRenderObject.cpp:
1381         (WebCore::AccessibilityRenderObject::determineAccessibilityRole):
1382         * accessibility/atk/WebKitAccessibleWrapperAtk.cpp:
1383         (atkRole):
1384         * accessibility/mac/WebAccessibilityObjectWrapperMac.mm:
1385         (createAccessibilityRoleMap):
1386
1387 2014-11-16  Chris Dumez  <cdumez@apple.com>
1388
1389         Crash when setting 'alt' CSS property to inherit or initial
1390         https://bugs.webkit.org/show_bug.cgi?id=138774
1391         <rdar://problem/18995409>
1392
1393         Reviewed by Chris Fleizach.
1394
1395         The StyleResolver was not handling 'inherit' or 'initial' values for
1396         'alt' CSS property. As a result, the code was assuming the CSSValue
1397         was a CSSPrimitiveValue (although it could be a CSSInitialValue or
1398         a CSSInheritValue). In such case, the code would dereference a null
1399         pointer because primitiveValue is initialized using:
1400         CSSPrimitiveValue* primitiveValue = is<CSSPrimitiveValue>(*value)
1401             ? downcast<CSSPrimitiveValue>(value)
1402             : nullptr;
1403
1404         Test: fast/css/alt-inherit-initial.html
1405
1406         * css/StyleResolver.cpp:
1407         (WebCore::StyleResolver::applyProperty):
1408
1409 2014-11-15  Joanmarie Diggs  <jdiggs@igalia.com>
1410
1411         AX: [ATK] Lists with selectable children should have ATK_ROLE_LIST_BOX
1412         https://bugs.webkit.org/show_bug.cgi?id=127060
1413
1414         Reviewed by Chris Fleizach.
1415
1416         Corrected the mapping from ATK_ROLE_LIST to ATK_ROLE_LIST_BOX.
1417
1418         No new tests. Instead, updated the expectations of roles-exposed.html to
1419         reflect the corrected mapping.
1420
1421         * accessibility/atk/WebKitAccessibleWrapperAtk.cpp:
1422         (atkRole):
1423
1424 2014-11-15  Joanmarie Diggs  <jdiggs@igalia.com>
1425
1426         AX: [ATK] Video and audio elements should be exposed using ATK's video and audio roles.
1427         https://bugs.webkit.org/show_bug.cgi?id=138767
1428
1429         Reviewed by Chris Fleizach.
1430
1431         Map AudioRole and VideoRole to ATK_ROLE_AUDIO and ATK_ROLE_VIDEO
1432         respectively, falling back on the existing mapping to ATK_ROLE_EMBEDDED
1433         in environments which have an older version of ATK.
1434
1435         No new tests are required. Instead, existing expectations were updated
1436         to reflect the new mapping.
1437
1438         * accessibility/atk/WebKitAccessibleWrapperAtk.cpp:
1439         (atkRole):
1440
1441 2014-11-15  Joanmarie Diggs  <jdiggs@igalia.com>
1442
1443         AX: [ATK] Expose the footer element as ATK_ROLE_FOOTER
1444         https://bugs.webkit.org/show_bug.cgi?id=138773
1445
1446         Reviewed by Chris Fleizach.
1447
1448         Map FooterRole to ATK_ROLE_FOOTER.
1449
1450         No new tests. Unskipped a test that verifies this mapping.
1451
1452         * accessibility/atk/WebKitAccessibleWrapperAtk.cpp:
1453         (atkRole):
1454
1455 2014-11-14  Dean Jackson  <dino@apple.com>
1456
1457         Rename WebKitCSSKeyframe(s)Rule into CSSKeyframe(s)Rule
1458         https://bugs.webkit.org/show_bug.cgi?id=138763
1459         <rdar://problem/18991569>
1460
1461         Reviewed by Simon Fraser.
1462
1463         Globally rename WebKitCSSKeyframeRule and WebKitCSSKeyframesRule to
1464         CSSKeyframeRule and CSSKeyframesRule respectively.
1465
1466         This *may* break some existing content, but is compatible with
1467         what Blink and Firefox are doing. To minimize breakage, I kept an
1468         alias for the old constant values on CSSRule: WEBKIT_KEYFRAMES_RULE
1469         and WEBKIT_KEYFRAME_RULE.
1470
1471         Covered by existing tests.
1472
1473         * CMakeLists.txt:
1474         * DerivedSources.cpp:
1475         * DerivedSources.make:
1476         * WebCore.vcxproj/WebCore.vcxproj:
1477         * WebCore.vcxproj/WebCore.vcxproj.filters:
1478         * WebCore.xcodeproj/project.pbxproj:
1479         * bindings/js/JSCSSRuleCustom.cpp:
1480         (WebCore::toJS):
1481         * bindings/objc/DOMCSS.mm:
1482         (kitClass):
1483         * css/CSSGrammar.y.includes:
1484         * css/CSSKeyframeRule.cpp: Renamed from Source/WebCore/css/WebKitCSSKeyframeRule.cpp.
1485         (WebCore::StyleKeyframe::StyleKeyframe):
1486         (WebCore::StyleKeyframe::~StyleKeyframe):
1487         (WebCore::StyleKeyframe::mutableProperties):
1488         (WebCore::StyleKeyframe::parseKeyString):
1489         (WebCore::StyleKeyframe::cssText):
1490         (WebCore::CSSKeyframeRule::CSSKeyframeRule):
1491         (WebCore::CSSKeyframeRule::~CSSKeyframeRule):
1492         (WebCore::CSSKeyframeRule::style):
1493         (WebCore::CSSKeyframeRule::reattach):
1494         * css/CSSKeyframeRule.h: Renamed from Source/WebCore/css/WebKitCSSKeyframeRule.h.
1495         (WebCore::StyleKeyframe::create):
1496         (WebCore::StyleKeyframe::keyText):
1497         (WebCore::StyleKeyframe::setKeyText):
1498         (WebCore::StyleKeyframe::getKeys):
1499         (WebCore::StyleKeyframe::properties):
1500         * css/CSSKeyframeRule.idl: Renamed from Source/WebCore/css/WebKitCSSKeyframeRule.idl.
1501         * css/CSSKeyframesRule.cpp: Renamed from Source/WebCore/css/WebKitCSSKeyframesRule.cpp.
1502         (WebCore::StyleRuleKeyframes::StyleRuleKeyframes):
1503         (WebCore::StyleRuleKeyframes::~StyleRuleKeyframes):
1504         (WebCore::StyleRuleKeyframes::parserAppendKeyframe):
1505         (WebCore::StyleRuleKeyframes::wrapperAppendKeyframe):
1506         (WebCore::StyleRuleKeyframes::wrapperRemoveKeyframe):
1507         (WebCore::StyleRuleKeyframes::findKeyframeIndex):
1508         (WebCore::CSSKeyframesRule::CSSKeyframesRule):
1509         (WebCore::CSSKeyframesRule::~CSSKeyframesRule):
1510         (WebCore::CSSKeyframesRule::setName):
1511         (WebCore::CSSKeyframesRule::appendRule):
1512         (WebCore::CSSKeyframesRule::insertRule):
1513         (WebCore::CSSKeyframesRule::deleteRule):
1514         (WebCore::CSSKeyframesRule::findRule):
1515         (WebCore::CSSKeyframesRule::cssText):
1516         (WebCore::CSSKeyframesRule::length):
1517         (WebCore::CSSKeyframesRule::item):
1518         (WebCore::CSSKeyframesRule::cssRules):
1519         (WebCore::CSSKeyframesRule::reattach):
1520         * css/CSSKeyframesRule.h: Renamed from Source/WebCore/css/WebKitCSSKeyframesRule.h.
1521         (WebCore::StyleRuleKeyframes::create):
1522         (WebCore::StyleRuleKeyframes::keyframes):
1523         (WebCore::StyleRuleKeyframes::name):
1524         (WebCore::StyleRuleKeyframes::setName):
1525         (WebCore::StyleRuleKeyframes::copy):
1526         * css/CSSKeyframesRule.idl: Renamed from Source/WebCore/css/WebKitCSSKeyframesRule.idl.
1527         * css/CSSParser.cpp:
1528         * css/CSSRule.h:
1529         * css/CSSRule.idl:
1530         * css/CSSStyleSheet.cpp:
1531         * css/RuleSet.cpp:
1532         * css/StyleResolver.cpp:
1533         * css/StyleRule.cpp:
1534         (WebCore::StyleRuleBase::createCSSOMWrapper):
1535         * inspector/InspectorStyleSheet.cpp:
1536         (WebCore::asCSSRuleList):
1537
1538 2014-11-15  David Kilzer  <ddkilzer@apple.com>
1539
1540         REGRESSION (r176137): iOS build broke after moving DataDetectors scanning code to WebCore
1541
1542         This regressed with the following commit:
1543
1544             Move DataDetectors scanning code to WebCore
1545             https://bugs.webkit.org/show_bug.cgi?id=138731
1546             <rdar://problem/18877535>
1547
1548         This fixes the following build failures:
1549
1550             In file included from WebCore/editing/mac/DataDetection.mm:29:
1551             WebCore/platform/spi/mac/DataDetectorsSPI.h:66:11: error: unknown type name 'NSRect'
1552             @property NSRect highlightFrame;
1553                       ^
1554             WebCore/editing/mac/DataDetection.mm:42:28: error: use of undeclared identifier 'DataDetection'
1555             RetainPtr<DDActionContext> DataDetection::detectItemAroundHitTestResult(const HitTestResult& hitTestResult, FloatRect& detectedDataBoundingBox, RefPtr<Range>& detectedDataRange)
1556                                        ^
1557             2 errors generated.
1558
1559         * WebCore.exp.in:
1560         (DataDetection::detectItemAroundHitTestResult): Move symbol from
1561         shared section to Mac-only section.
1562         * editing/mac/DataDetection.mm: Add PLATFORM(MAC) check around
1563         the entire implementation since it it unused on iOS.
1564         * platform/spi/mac/DataDetectorsSPI.h: Expand PLATFORM(MAC)
1565         check around DDActionContext class declarations.
1566
1567 2014-11-14  Benjamin Poulain  <benjamin@webkit.org>
1568
1569         Compute the selector specificity as we match simple selectors
1570         https://bugs.webkit.org/show_bug.cgi?id=138718
1571
1572         Reviewed by Andreas Kling.
1573
1574         This is an other tiny step toward dynamic specificity. Instead of computing
1575         the entire specificity at the end, compute it dynamically as we are matching
1576         each individual simple selector.
1577
1578         * css/CSSSelector.cpp:
1579         (WebCore::CSSSelector::specificity):
1580         (WebCore::CSSSelector::simpleSelectorSpecificity):
1581         (WebCore::CSSSelector::addSpecificities):
1582         (WebCore::CSSSelector::specificityForPage):
1583         (WebCore::CSSSelector::specificityForOneSelector): Deleted.
1584         * css/CSSSelector.h:
1585         * css/SelectorChecker.cpp:
1586         (WebCore::SelectorChecker::match):
1587         (WebCore::SelectorChecker::matchRecursively):
1588         (WebCore::SelectorChecker::checkOne):
1589         (WebCore::SelectorChecker::matchSelectorList):
1590         * css/SelectorChecker.h:
1591
1592 2014-11-14  Andreas Kling  <akling@apple.com>
1593
1594         [mac] Only images that are actually purgeable should be advertised as such.
1595         <https://webkit.org/b/138759>
1596         <rdar://problem/13205438>
1597
1598         Reviewed by Tim Horton.
1599
1600         Instead of saying that all images on >=Yosemite are purgeable, have BitmapImage
1601         forward the question to CGImage for the correct answer.
1602
1603         The memory may be marked non-volatile by frameworks underneath WebKit at any
1604         given time, but that's not something we need to worry about.
1605
1606         * platform/graphics/BitmapImage.cpp:
1607         (WebCore::BitmapImage::decodedDataIsPurgeable):
1608         * platform/graphics/BitmapImage.h:
1609         * platform/graphics/cg/BitmapImageCG.cpp:
1610         (WebCore::BitmapImage::decodedDataIsPurgeable):
1611         * platform/graphics/cg/ImageSourceCG.cpp:
1612         (WebCore::ImageSource::createFrameAtIndex):
1613         * platform/spi/cg/CoreGraphicsSPI.h:
1614
1615 2014-11-14  Zalan Bujtas  <zalan@apple.com>
1616
1617         Remove computeTextPaintStyle's RenderText dependency.
1618         https://bugs.webkit.org/show_bug.cgi?id=138754
1619
1620         Reviewed by Simon Fraser.
1621
1622         So that we can compute text paint style across renderers in simple line layout.
1623
1624         No change in functionality.
1625
1626         * rendering/InlineTextBox.cpp:
1627         (WebCore::InlineTextBox::paint):
1628         * rendering/SimpleLineLayoutFunctions.cpp:
1629         (WebCore::SimpleLineLayout::paintFlow):
1630         * rendering/TextPaintStyle.cpp:
1631         (WebCore::computeTextPaintStyle):
1632         * rendering/TextPaintStyle.h:
1633
1634 2014-11-14  Myles C. Maxfield  <mmaxfield@apple.com>
1635
1636         Get rid of WebFontCache
1637         https://bugs.webkit.org/show_bug.cgi?id=134752
1638
1639         Reviewed by Dave Hyatt.
1640
1641         No new tests because there is no behavior change.
1642
1643         * Configurations/WebCore.xcconfig:
1644         * PlatformMac.cmake:
1645         * WebCore.exp.in:
1646         * WebCore.order:
1647         * WebCore.xcodeproj/project.pbxproj:
1648         * platform/graphics/mac/FontCacheMac.mm:
1649         (WebCore::acceptableChoice):
1650         (WebCore::betterChoice):
1651         (WebCore::toTraitsMask):
1652         (WebCore::desiredFamilyToAvailableFamilyDictionary):
1653         (WebCore::rememberDesiredFamilyToAvailableFamilyMapping):
1654         (WebCore::fontWithFamily):
1655         (WebCore::invalidateFontCache):
1656         (WebCore::FontCache::getTraitsInFamily):
1657         (WebCore::FontCache::createFontPlatformData):
1658         * platform/mac/WebFontCache.h: Removed.
1659         * platform/mac/WebFontCache.mm: Removed.
1660
1661 2014-11-14  Daniel Bates  <dabates@apple.com>
1662
1663         [iOS] NSFileManager and QuickLook SPI not available in public SDK
1664         https://bugs.webkit.org/show_bug.cgi?id=138742
1665
1666         Reviewed by Pratik Solanki.
1667
1668         Add a SPI wrapper headers {NSFileManager, QuickLook}SPI.h that forward declares
1669         NSFileManager and QuickLook SPI, respectively, and use these header instead of
1670         including private headers of Foundation and QuickLook directly.
1671
1672         * WebCore.xcodeproj/project.pbxproj: Add project and private header QuickLookSPI.h
1673         and NSFileManagerSPI.h, respectively.
1674         * platform/network/ios/QuickLook.mm: Include headers QuickLookSPI.h and NSFileManagerSPI.h
1675         instead of including headers of QuickLook and Foundation directly.
1676         * platform/spi/cocoa/NSFileManagerSPI.h: Added.
1677         * platform/spi/ios/QuickLookSPI.h: Added.
1678
1679 2014-11-14  Daniel Bates  <dabates@apple.com>
1680
1681         [iOS] CoreGraphics SPI not available in public SDK
1682         https://bugs.webkit.org/show_bug.cgi?id=138709
1683
1684         Reviewed by David Kilzer.
1685
1686         Add a SPI wrapper header called CoreGraphicsSPI.h that forward declares
1687         CoreGraphics SPI and use this header instead of including private headers
1688         of CoreGraphics directly.
1689
1690         * WebCore.xcodeproj/project.pbxproj: Add WebCore private header CoreGraphicsSPI.h. Also remove header
1691         CGFontUnicodeSupportSPI.h as its functionality has been incorporated into header CoreGraphicsSPI.h.
1692         * WebCore.vcxproj/WebCore.vcxproj: Add WebCore private header CoreGraphicsSPI.h.
1693         * WebCore.vcxproj/WebCore.vcxproj.filters: Add directories platform/spi and platform/spi/cg.
1694         * WebCore.vcxproj/WebCoreCG.props: Add include directory platform/spi/cg.
1695         * WebCore.vcxproj/copyForwardingHeaders.cmd: Copy WebCore headers from directory platform/spi/cg. This
1696         is necessary because WebCore/platform/graphics/SimpleFontData.h includes CoreGraphicsSPI.h. And
1697         WebCore/platform/graphics/SimpleFontData.h is compiled for the Apple Windows port.
1698         * platform/graphics/SimpleFontData.h: Move CGFontRenderingStyle declaration and associated enum
1699         from this file to file CoreGraphicsSPI.h and include header CoreGraphicsSPI.h.
1700         * platform/graphics/cg/BitmapImageCG.cpp: Remove unnecessary #include of header CGContextPrivate.h.
1701         * platform/graphics/cg/ColorCG.cpp: Include header CoreGraphicsSPI.h instead of including headers
1702         of CoreGraphics directly.
1703         * platform/graphics/cg/GraphicsContextCG.cpp: Move forward declarations of CoreGraphics functions
1704         from this file to file CoreGraphicsSPI.h and include header CoreGraphicsSPI.h. Also sort the list
1705         of #include directives.
1706         * platform/graphics/cg/ImageSourceCG.cpp: Include header CoreGraphicsSPI.h instead of including
1707         headers of CoreGraphics directly.
1708         * platform/graphics/cg/PDFDocumentImage.cpp: Ditto.
1709         * platform/graphics/cocoa/FontPlatformDataCocoa.mm: Ditto.
1710         * platform/graphics/ios/FontCacheIOS.mm: Include header CoreGraphicsSPI.h instead of including
1711         header <CoreGraphics/CGFontUnicodeSupport.h> directly.
1712         * platform/graphics/ios/FontServicesIOS.mm: Include header CoreGraphicsSPI.h instead of including
1713         headers of CoreGraphics directly.
1714         * platform/graphics/ios/SimpleFontDataIOS.mm: Ditto.
1715         * platform/graphics/mac/FontMac.mm: Move forward declarations of CoreGraphics functions from this
1716         file to file CoreGraphicsSPI.h and include header CoreGraphicsSPI.h.
1717         * platform/graphics/mac/GlyphPageTreeNodeMac.cpp: Substitute header CoreGraphicsSPI.h for CGFontUnicodeSupportSPI.h
1718         and remove #include of headers CGFontUnicodeSupport.h (it will be included by CoreGraphicsSPI.h).
1719         * platform/graphics/mac/SimpleFontDataMac.mm: Move forward declarations of CoreGraphics functions
1720         from this file to file CoreGraphicsSPI.h and include header CoreGraphicsSPI.h.
1721         * platform/ios/LegacyTileGrid.mm: Include header CoreGraphicsSPI.h instead of including headers of
1722         CoreGraphics directly.
1723         * platform/ios/wak/WKGraphics.h: Remove #include of header CoreGraphicsPrivate.h and forward declare
1724         CGFontAntialiasingStyle Additionally, declare WKCompositeOperation as an alias of int and use this
1725         data type instead of CGCompositeOperation (SPI). This approach lets us keep the definition of
1726         CGCompositeOperation in file CoreGraphicsSPI.h, where all the other definitions/forward declarations
1727         of CoreGraphics data types reside, as opposed to defining it in this file.
1728         * platform/ios/wak/WKGraphics.mm:
1729         (WKRectFillUsingOperation): Cast data type WKCompositeOperation to CGCompositeOperation and added
1730         compile-time assert to ensure that the size of WKCompositeOperation is identical to the size of CGCompositeOperation.
1731         * platform/graphics/cg/ImageSourceCG.cpp: Move CGImageCachingFlags declaration, associated enum, and
1732         forward declaration of CGImageSetCachingFlags() from this file to file CoreGraphicsSPI.h and include
1733         header CoreGraphicsSPI.h.
1734         (WebCore::ImageSource::createFrameAtIndex): Fix incorrect comment.
1735         * platform/spi/cocoa/CGFontUnicodeSupportSPI.h: Removed; moved its functionality into header CoreGraphicsSPI.h.
1736         * platform/spi/cg/CoreGraphicsSPI.h: Added.
1737         (CGFloatMin):
1738         * rendering/RenderThemeIOS.mm: Substitute header CoreGraphics.h for CGPathPrivate.h as we no longer
1739         make use of any functionality from CGPathPrivate.h in this file. The functionality we used from
1740         CGPathPrivate.h, including CGPathAddRoundedRect(), has been moved to the public API header, CGPath.h,
1741         which is included from header CoreGraphics.h.
1742
1743 2014-11-14  Dean Jackson  <dino@apple.com>
1744
1745         [Media] Timeline scrubber not updating as the video plays
1746         https://bugs.webkit.org/show_bug.cgi?id=138717
1747         <rdar://problem/18957509>
1748
1749         Reviewed by Jer Noble.
1750
1751         There is some confusing logic that determines
1752         whether or not the controls are hidden, and a
1753         slight error caused timelines to not advance unless
1754         they were explicitly showing.
1755
1756         The fix was to be a bit more liberal in deciding
1757         when the controls are showing, but still cautious
1758         enough to stop up causing repaints when things
1759         are hidden.
1760
1761         Test: media/media-controls-timeline-updates.html
1762
1763         * Modules/mediacontrols/mediaControlsApple.js:
1764         (Controller.prototype.controlsAreHidden): We only need
1765         to check if we're an audio element or explicitly
1766         hidden.
1767         (Controller.prototype.showControls): Explicitly call
1768         updateTime.
1769
1770 2014-11-14  Tim Horton  <timothy_horton@apple.com>
1771
1772         Move DataDetectors scanning code to WebCore
1773         https://bugs.webkit.org/show_bug.cgi?id=138731
1774         <rdar://problem/18877535>
1775
1776         Reviewed by Anders Carlsson.
1777
1778         * WebCore.exp.in:
1779         Export our new function.
1780
1781         * WebCore.xcodeproj/project.pbxproj:
1782         Add DataDetection.{mm, h}
1783         Adjust the project so that editing/mac actually points to editing/mac instead of editing/
1784
1785         * editing/mac/DictionaryLookup.h:
1786         * editing/mac/DictionaryLookup.mm:
1787         Swap to PLATFORM(MAC) instead of !PLATFORM(IOS).
1788         Move DictionaryLookup.{mm, h} to editing/mac, where they were in the project but not on disk.
1789
1790         * editing/mac/DataDetection.h: Added.
1791         * editing/mac/DataDetection.mm: Added.
1792         (WebCore::DataDetection::detectItemAroundHitTestResult):
1793         Moved from WebKit2.
1794
1795 2014-11-14  Dan Bernstein  <mitz@apple.com>
1796
1797         <rdar://problem/18978497> Wrong (off-by-1) navigation snapshots shown after a mix of gesture and button back/forward navigation
1798         https://bugs.webkit.org/show_bug.cgi?id=138753
1799
1800         Reviewed by Tim Horton.
1801
1802         Removed a FrameLoaderClient function that no one overrides anymore.
1803
1804         * loader/FrameLoaderClient.h:
1805         (WebCore::FrameLoaderClient::willChangeCurrentHistoryItem): Deleted.
1806         * loader/HistoryController.cpp:
1807         (WebCore::HistoryController::setCurrentItem): Removed call to willChangeCurrentHistoryItem.
1808         (WebCore::HistoryController::replaceCurrentItem): Ditto.
1809
1810 2014-11-14  Jeremy Jones  <jeremyj@apple.com>
1811
1812         Do proper teardown for optimized fullscreen
1813         https://bugs.webkit.org/show_bug.cgi?id=138648
1814
1815         Reviewed by Eric Carlson.
1816
1817         * platform/ios/WebVideoFullscreenInterfaceAVKit.mm:
1818         (WebVideoFullscreenInterfaceAVKit::invalidate): must do corresponding teardown based on mode.
1819
1820 2014-11-11  David Hyatt  <hyatt@apple.com>
1821
1822         Table rows repaint entire table when their background color changes.
1823         https://bugs.webkit.org/show_bug.cgi?id=26662
1824
1825         Reviewed by Simon Fraser.
1826
1827         Added fast/repaint/table-row-repaint.html
1828
1829         * rendering/RenderTableCell.h:
1830         * rendering/RenderTableRow.cpp:
1831         (WebCore::RenderTableRow::clippedOverflowRectForRepaint):
1832         Improve table row's function to unite with the cell rects that can possibly
1833         have the the row's background color painted behind them.
1834
1835 2014-11-14  Zalan Bujtas  <zalan@apple.com>
1836
1837         Simple line layout: Move simple line layout RunResolver and LineResolver implementation to SimpleLineLayoutResolver.cpp
1838         https://bugs.webkit.org/show_bug.cgi?id=138704
1839
1840         Reviewed by Antti Koivisto.
1841
1842         No change in functionality.
1843
1844         * CMakeLists.txt:
1845         * WebCore.vcxproj/WebCore.vcxproj:
1846         * WebCore.vcxproj/WebCore.vcxproj.filters:
1847         * WebCore.xcodeproj/project.pbxproj:
1848         * rendering/SimpleLineLayoutResolver.cpp: Added.
1849         (WebCore::SimpleLineLayout::RunResolver::Run::Run):
1850         (WebCore::SimpleLineLayout::RunResolver::Run::rect):
1851         (WebCore::SimpleLineLayout::RunResolver::Run::baseline):
1852         (WebCore::SimpleLineLayout::RunResolver::Run::text):
1853         (WebCore::SimpleLineLayout::RunResolver::Iterator::Iterator):
1854         (WebCore::SimpleLineLayout::RunResolver::Iterator::advance):
1855         (WebCore::SimpleLineLayout::RunResolver::Iterator::advanceLines):
1856         (WebCore::SimpleLineLayout::RunResolver::RunResolver):
1857         (WebCore::SimpleLineLayout::RunResolver::lineIndexForHeight):
1858         (WebCore::SimpleLineLayout::RunResolver::rangeForRect):
1859         (WebCore::SimpleLineLayout::LineResolver::Iterator::Iterator):
1860         (WebCore::SimpleLineLayout::LineResolver::Iterator::operator*):
1861         (WebCore::SimpleLineLayout::LineResolver::LineResolver):
1862         * rendering/SimpleLineLayoutResolver.h:
1863         (WebCore::SimpleLineLayout::RunResolver::Run::Run): Deleted.
1864         (WebCore::SimpleLineLayout::RunResolver::Run::rect): Deleted.
1865         (WebCore::SimpleLineLayout::RunResolver::Run::baseline): Deleted.
1866         (WebCore::SimpleLineLayout::RunResolver::Run::text): Deleted.
1867         (WebCore::SimpleLineLayout::RunResolver::Iterator::Iterator): Deleted.
1868         (WebCore::SimpleLineLayout::RunResolver::Iterator::advance): Deleted.
1869         (WebCore::SimpleLineLayout::RunResolver::Iterator::advanceLines): Deleted.
1870         (WebCore::SimpleLineLayout::RunResolver::RunResolver): Deleted.
1871         (WebCore::SimpleLineLayout::RunResolver::lineIndexForHeight): Deleted.
1872         (WebCore::SimpleLineLayout::RunResolver::rangeForRect): Deleted.
1873         (WebCore::SimpleLineLayout::LineResolver::Iterator::Iterator): Deleted.
1874         (WebCore::SimpleLineLayout::LineResolver::Iterator::operator*): Deleted.
1875         (WebCore::SimpleLineLayout::LineResolver::LineResolver): Deleted.
1876
1877 2014-11-14  Andrzej Badowski  <a.badowski@samsung.com>
1878
1879         AX: Improve AccessibilityTableCell isColumnHeaderCell function.
1880         https://bugs.webkit.org/show_bug.cgi?id=138423
1881
1882         Reviewed by Chris Fleizach.
1883
1884         <th> cell in the zero row should be column header.
1885
1886         New tests are not necessary.
1887
1888         * accessibility/AccessibilityTableCell.cpp:
1889         (WebCore::AccessibilityTableCell::isColumnHeaderCell):
1890
1891 2014-11-14  Carlos Garcia Campos  <cgarcia@igalia.com>
1892
1893         [SOUP] Use GMainLoopSource for request timeout in ResourceHandle
1894         https://bugs.webkit.org/show_bug.cgi?id=138695
1895
1896         Reviewed by Sergio Villar Senin.
1897
1898         We are currently using soup_timeout_add() that simply creates a
1899         GSource and attaches it to the given context. Using
1900         GMainLoopSource we simplify the code and fix any potential problem
1901         of converting the double value into milliseconds.
1902
1903         * platform/network/ResourceHandleInternal.h:
1904         * platform/network/soup/ResourceHandleSoup.cpp:
1905         (WebCore::cleanupSoupRequestOperation):
1906         (WebCore::ResourceHandle::sendPendingRequest):
1907         (WebCore::ResourceHandle::platformSetDefersLoading):
1908         (WebCore::requestTimeoutCallback): Deleted.
1909
1910 2014-11-13  Tim Horton  <timothy_horton@apple.com>
1911
1912         [mac] Keep around more decoded image data, since it's purgeable
1913         https://bugs.webkit.org/show_bug.cgi?id=125273
1914         <rdar://problem/13205438>
1915
1916         Reviewed by Simon Fraser.
1917
1918         No new tests, just an optimization.
1919
1920         Instead of throwing away decoded image data eagerly, allow the operating
1921         system to manage the memory via the standard purgeability mechanism,
1922         where it can.
1923
1924         This improves the performance on some pathological cases (extremely large
1925         animated GIFs) by up to 8x.
1926
1927         * loader/cache/MemoryCache.cpp:
1928         (WebCore::MemoryCache::pruneLiveResourcesToSize):
1929         Don't prune live resources' decoded data if it is purgeable.
1930
1931         * platform/graphics/BitmapImage.cpp:
1932         (WebCore::BitmapImage::destroyDecodedDataIfNecessary):
1933         Don't eagerly throw away decoded image data if it's purgeable.
1934
1935         * loader/cache/CachedImage.h:
1936         * loader/cache/CachedResource.h:
1937         (WebCore::CachedResource::decodedDataIsPurgeable):
1938         * platform/graphics/BitmapImage.h:
1939         * platform/graphics/Image.h:
1940         (WebCore::Image::decodedDataIsPurgeable):
1941
1942 2014-11-13  Myles C. Maxfield  <litherum@gmail.com>
1943
1944         Addressing post-review comment
1945         https://bugs.webkit.org/show_bug.cgi?id=138633
1946
1947         * platform/spi/cg/CoreGraphicsSPI.h:
1948
1949 2014-11-13  Myles C. Maxfield  <mmaxfield@apple.com>
1950
1951         Move FontMac and FontCacheMac off of WKSI
1952         https://bugs.webkit.org/show_bug.cgi?id=138633
1953
1954         Reviewed by Simon Fraser.
1955
1956         No new tests because there is no behavior change.
1957
1958         * WebCore.exp.in:
1959         * WebCore.order:
1960         * WebCore.vcxproj/WebCoreCG.props:
1961         * WebCore.xcodeproj/project.pbxproj:
1962         * platform/graphics/ios/FontCacheIOS.mm:
1963         (WebCore::FontCache::platformInit):
1964         * platform/graphics/mac/FontCacheMac.mm:
1965         (WebCore::FontCache::platformInit):
1966         (WebCore::FontCache::systemFallbackForCharacters):
1967         * platform/graphics/mac/FontMac.mm:
1968         (WebCore::isIntegral):
1969         (WebCore::setCGFontRenderingMode):
1970         (WebCore::Font::drawGlyphs):
1971         * platform/graphics/mac/GlyphPageTreeNodeMac.cpp:
1972         * platform/graphics/win/FontCacheWin.cpp:
1973         (WebCore::FontCache::platformInit):
1974         * platform/ios/WebCoreSystemInterfaceIOS.mm:
1975         * platform/mac/WebCoreSystemInterface.h:
1976         * platform/mac/WebCoreSystemInterface.mm:
1977         * platform/spi/cocoa/CoreGraphicsSPI.h: Copied from Source/WebCore/platform/spi/cocoa/CGFontUnicodeSupportSPI.h.
1978         * platform/spi/cocoa/CoreTextSPI.h:
1979         * platform/spi/mac/NSFontSPI.h: Renamed from Source/WebCore/platform/spi/cocoa/CGFontUnicodeSupportSPI.h.
1980
1981 2014-11-13  Chris Dumez  <cdumez@apple.com>
1982
1983         Move 'image-resolution' CSS property to the new StyleBuilder
1984         https://bugs.webkit.org/show_bug.cgi?id=138715
1985
1986         Reviewed by Andreas Kling.
1987
1988         Move 'image-resolution' CSS property from DeprecatedStyleBuilder to
1989         the new StyleBuilder by using custom code.
1990
1991         No new tests, no behavior change.
1992
1993         * css/CSSPropertyNames.in:
1994         * css/DeprecatedStyleBuilder.cpp:
1995         (WebCore::DeprecatedStyleBuilder::DeprecatedStyleBuilder):
1996         (WebCore::ApplyPropertyImageResolution::applyInheritValue): Deleted.
1997         (WebCore::ApplyPropertyImageResolution::applyInitialValue): Deleted.
1998         (WebCore::ApplyPropertyImageResolution::applyValue): Deleted.
1999         (WebCore::ApplyPropertyImageResolution::createHandler): Deleted.
2000         * css/StyleBuilderCustom.h:
2001         (WebCore::StyleBuilderFunctions::applyInheritImageResolution):
2002         (WebCore::StyleBuilderFunctions::applyInitialImageResolution):
2003         (WebCore::StyleBuilderFunctions::applyValueImageResolution):
2004
2005 2014-11-13  Commit Queue  <commit-queue@webkit.org>
2006
2007         Unreviewed, rolling out r176106.
2008         https://bugs.webkit.org/show_bug.cgi?id=138719
2009
2010         needs build love (Requested by kling on #webkit).
2011
2012         Reverted changeset:
2013
2014         "[mac] Keep around more decoded image data, since it's
2015         purgeable"
2016         https://bugs.webkit.org/show_bug.cgi?id=125273
2017         http://trac.webkit.org/changeset/176106
2018
2019 2014-11-13  Eric Carlson  <eric.carlson@apple.com>
2020
2021         Context menus should not offer the "Download video" option for videos that cannot
2022         be downloaded
2023         https://bugs.webkit.org/show_bug.cgi?id=138530
2024         -and corresponding-
2025         rdar://problem/18919130
2026
2027         Reviewed by Tim Horton.
2028
2029         * html/HTMLMediaElement.cpp:
2030         (WebCore::HTMLMediaElement::parseAttribute):
2031         * html/HTMLMediaElement.h:
2032         * page/ContextMenuController.cpp:
2033         (WebCore::ContextMenuController::populate):
2034         * platform/graphics/MediaPlayer.cpp:
2035         (WebCore::MediaPlayer::canSaveMediaData):
2036         (WebCore::MediaPlayer::supportsSave): Deleted.
2037         * platform/graphics/MediaPlayer.h:
2038         * platform/graphics/MediaPlayerPrivate.h:
2039         (WebCore::MediaPlayerPrivateInterface::supportsFullscreen):
2040         (WebCore::MediaPlayerPrivateInterface::canSaveMediaData):
2041         (WebCore::MediaPlayerPrivateInterface::supportsSave): Deleted.
2042         * platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.cpp:
2043         (WebCore::MediaPlayerPrivateAVFoundation::resolvedURL):
2044         (WebCore::MediaPlayerPrivateAVFoundation::canSaveMediaData):
2045         * platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.h:
2046         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
2047         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
2048         (WebCore::MediaPlayerPrivateAVFoundationObjC::resolvedURL):
2049         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
2050         (WebCore::MediaPlayerPrivateGStreamer::canSaveMediaData):
2051         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.h:
2052         * platform/graphics/mac/MediaPlayerPrivateQTKit.h:
2053         * platform/graphics/mac/MediaPlayerPrivateQTKit.mm:
2054         (WebCore::MediaPlayerPrivateQTKit::canSaveMediaData):
2055         * rendering/HitTestResult.cpp:
2056         (WebCore::HitTestResult::isDownloadableMedia):
2057
2058 2014-11-02  Tim Horton  <timothy_horton@apple.com>
2059
2060         [mac] Keep around more decoded image data, since it's purgeable
2061         <https://webkit.org/b/125273>
2062         <rdar://problem/13205438>
2063
2064         Reviewed by Simon Fraser and Andreas Kling.
2065
2066         No new tests, just an optimization.
2067
2068         Instead of throwing away decoded image data eagerly, allow the operating
2069         system to manage the memory via the standard purgeability mechanism,
2070         where it can.
2071
2072         This improves the performance on some pathological cases (extremely large
2073         animated GIFs) by up to 8x.
2074
2075         * loader/cache/MemoryCache.cpp:
2076         (WebCore::MemoryCache::pruneLiveResourcesToSize):
2077         Don't prune live resources' decoded data if it is purgeable.
2078
2079         * platform/graphics/BitmapImage.cpp:
2080         (WebCore::BitmapImage::destroyDecodedDataIfNecessary):
2081         Don't eagerly throw away decoded image data if it's purgeable.
2082
2083         * loader/cache/CachedImage.h:
2084         * loader/cache/CachedResource.h:
2085         (WebCore::CachedResource::decodedDataIsPurgeable):
2086         * platform/graphics/BitmapImage.h:
2087         * platform/graphics/Image.h:
2088         (WebCore::Image::decodedDataIsPurgeable):
2089
2090 2014-11-13  Daniel Bates  <dabates@apple.com>
2091
2092         [iOS] NSGeometry data types are not available in the public SDK
2093         https://bugs.webkit.org/show_bug.cgi?id=137536
2094
2095         Reviewed by David Kilzer.
2096
2097         Towards building iOS WebKit with the public iOS SDK, define NSGeometry
2098         data types and functions in terms of CGGeometry data types and functions
2099         because the former is SPI on iOS.
2100
2101         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm: Include
2102         header WAKAppKitStubs.h when building for iOS. Otherwise, include header NSGeometry.h.
2103         * platform/graphics/ca/mac/PlatformCALayerMac.mm: Remove import of private header NSGeometry.h.
2104         We will use the NSGeometry data types defined in WAKAppKitStubs.h, which is implicitly imported
2105         by header WAKWindow.h.
2106         * platform/graphics/mac/MediaPlayerProxy.h: Substitute WAKAppKitStubs.h for NSGeometry.h.
2107         * platform/ios/PlatformEventFactoryIOS.mm: Ditto. Also fix up style issues with #import directives.
2108         * platform/ios/WebEvent.mm: Ditto.
2109         * platform/ios/wak/WAKAppKitStubs.h: Define typedefs and macros that map NSGeometry data types
2110         and functions to the analogous CGGeometry data types and functions. Also, remove #ifdef __OBJC__-
2111         guard as the existing content in WAKAppKitStubs.h assumes that this file will only be included
2112         in an Objective-C/Objective-C++ file.
2113         * platform/ios/wak/WAKView.h: Import header WAKAppKitStubs.h instead of defining macros for some
2114         NSGeometry data types.
2115
2116 2014-11-13  Simon Fraser  <simon.fraser@apple.com>
2117
2118         [WK2] Fire a layout milestone on session restore based on render tree size
2119         https://bugs.webkit.org/show_bug.cgi?id=138711
2120         rdar://problem/16033854
2121
2122         Reviewed by Anders Carlsson.
2123
2124         New layout milestone for session restore based on render tree size. Only used
2125         in WK2 at present.
2126
2127         * page/LayoutMilestones.h:
2128
2129 2014-11-13  Dan Bernstein  <mitz@apple.com>
2130
2131         Policy client not called for navigations through the page cache
2132         https://bugs.webkit.org/show_bug.cgi?id=138703
2133
2134         Reviewed by Alexey Proskuryakov.
2135
2136         Test added to TestWebKitAPI/Tests/WebKit2Cocoa/Navigation.mm.
2137
2138         * loader/FrameLoader.cpp:
2139         (WebCore::FrameLoader::loadDifferentDocumentItem): When using a cached page, which already
2140         has a document loader, set the document loader’s triggering action (so that the policy
2141         client sees that this is a back/forward navigation) and clear its last checked request (so
2142         that the policy client gets called).
2143
2144 2014-11-13  Joanmarie Diggs  <jdiggs@igalia.com>
2145
2146         AX: [ATK] Do not return ATK_ROLE_UNKNOWN for null or otherwise invalid accessible objects
2147         https://bugs.webkit.org/show_bug.cgi?id=137867
2148
2149         Reviewed by Chris Fleizach.
2150
2151         Return ATK_ROLE_INVALID for null or otherwise invalid accessible objects.
2152
2153         Test: platform/gtk/accessibility/detached-object-role.html
2154
2155         * accessibility/atk/WebKitAccessibleWrapperAtk.cpp:
2156         (webkitAccessibleGetRole):
2157
2158 2014-11-13  Eric Carlson  <eric.carlson@apple.com>
2159
2160         [iOS] update optimized fullscreen media controls artwork
2161         https://bugs.webkit.org/show_bug.cgi?id=138705
2162
2163         Reviewed by Dean Jackson.
2164
2165         * Modules/mediacontrols/MediaControlsHost.cpp:
2166         (WebCore::MediaControlsHost::mediaUIImageData): Remove one button.
2167         * Modules/mediacontrols/MediaControlsHost.idl:
2168
2169         * Modules/mediacontrols/mediaControlsApple.js:
2170         (Controller.prototype.controlsAreHidden): Add parentheses to fix the logic.
2171
2172         * Modules/mediacontrols/mediaControlsiOS.js:
2173         (ControllerIOS.prototype.shouldHaveStartPlaybackButton): The big play button should not be
2174             visible during playback.
2175         (ControllerIOS.prototype.createControls): No longer necessary to insert style dynamically,
2176         (ControllerIOS.prototype.setPlaying): Call the base class first so "isPlaying" is set before
2177             calling updateControls, which calls shouldHaveStartPlaybackButton.
2178
2179         * platform/ios/WebCoreSystemInterfaceIOS.h: Update constants.
2180
2181         * rendering/RenderThemeIOS.mm:
2182         (WebCore::RenderThemeIOS::mediaControlsStyleSheet): Add button style to the media controls sheet.
2183
2184 2014-11-13  Benjamin Poulain  <bpoulain@apple.com>
2185
2186         Implement the matching for :nth-last-child(An+B of selector-list)
2187         https://bugs.webkit.org/show_bug.cgi?id=138650
2188
2189         Reviewed by Andreas Kling.
2190
2191         Matching is trivial based on the existing CSS4 infrastructure.
2192
2193         The tricky part is style invalidation. In the cases of the old :nth-last-child()
2194         and :nth-last-of-type(), we were using the flag ChildrenAffectedByBackwardPositionalRules.
2195
2196         The problem with ChildrenAffectedByBackwardPositionalRules is that it only invalidate
2197         subtrees when nodes are being added or removed. It is a valuable optimization by itself
2198         since tree updates are less common than property updates.
2199
2200         For the case of property updates, I have added the new flag "ChildrenAffectedByPropertyBasedBackwardPositionalRules".
2201         It pretty much work the same way but invalidates subtree on style recalc.
2202
2203         Tests: fast/css/nth-child-and-nth-last-child.html
2204                fast/css/nth-last-child-of-classname.html
2205                fast/css/nth-last-child-of-complex-selector.html
2206                fast/css/nth-last-child-of-compound-selector.html
2207                fast/css/nth-last-child-of-style-sharing-1.html
2208                fast/css/nth-last-child-of-style-sharing-2.html
2209                fast/css/nth-last-child-of-style-update-optimization.html
2210                fast/css/nth-last-child-of-tagname.html
2211                fast/selectors/nth-last-child-of-class-style-update.html
2212
2213         * css/SelectorChecker.cpp:
2214         (WebCore::SelectorChecker::checkOne):
2215         * dom/Element.cpp:
2216         (WebCore::Element::setChildrenAffectedByPropertyBasedBackwardPositionalRules):
2217         (WebCore::Element::hasFlagsSetDuringStylingOfChildren):
2218         (WebCore::Element::rareDataChildrenAffectedByPropertyBasedBackwardPositionalRules):
2219         * dom/Element.h:
2220         (WebCore::Element::childrenAffectedByPropertyBasedBackwardPositionalRules):
2221         * dom/ElementRareData.h:
2222         (WebCore::ElementRareData::childrenAffectedByPropertyBasedBackwardPositionalRules):
2223         (WebCore::ElementRareData::setChildrenAffectedByPropertyBasedBackwardPositionalRules):
2224         (WebCore::ElementRareData::ElementRareData):
2225         (WebCore::ElementRareData::resetDynamicRestyleObservations):
2226         * dom/Node.cpp:
2227         (WebCore::Node::updateAncestorsForStyleRecalc):
2228         (WebCore::Node::setNeedsStyleRecalc):
2229         (WebCore::markAncestorsWithChildNeedsStyleRecalc): Deleted.
2230         * dom/Node.h:
2231         Unlike the other marking steps, ChildrenAffectedByPropertyBasedBackwardPositionalRules can only really
2232         affect one level (because we don't have a parent-selectors). It is quite easy to mark large
2233         trees with ChildrenAffectedByPropertyBasedBackwardPositionalRules by using a selector list that is a bit
2234         too generic. When that happen, we must be careful not invalidating everything, just the affected subtree.
2235
2236 2014-11-12  Jon Honeycutt  <jhoneycutt@apple.com>
2237
2238         REGRESSION (r172826): Password field placeholder text is missing if
2239         placeholder attribute precedes type attribute
2240
2241         <https://bugs.webkit.org/show_bug.cgi?id=138682>
2242         <rdar://problem/18574213>
2243
2244         When the placeholder attribute is encountered, a shadow tree is created
2245         to include the placeholder element. When the type attribute is
2246         encountered, the previous shadow tree is destroyed, and a new shadow
2247         tree is created. TextFieldInputType::createShadowSubtree() did not
2248         handle creating the placeholder element.
2249
2250         This regressed in r172826 because prior to r172826, the placeholder
2251         element would be recreated when
2252         HTMLTextFormControlElement::updatePlaceholderVisibility() was called.
2253         In r172826, updatePlaceholderVisibility() was changed to only show or
2254         hide the placeholder element if it exists.
2255
2256         Reviewed by Benjamin Poulain.
2257
2258         * html/TextFieldInputType.cpp:
2259         (WebCore::TextFieldInputType::createShadowSubtree):
2260         Call updatePlaceholderVisibility() to insert the placeholder element
2261         (after the inner text element has been inserted, as it appears before
2262         it in the shadow DOM).
2263
2264 2014-11-13  Joanmarie Diggs  <jdiggs@igalia.com>
2265
2266         AX: [ATK] Accessible focus events are emitted when the selection is changed in non-focused combo boxes
2267         https://bugs.webkit.org/show_bug.cgi?id=137866
2268
2269         Reviewed by Mario Sanchez Prada.
2270
2271         Check that the AccessibilityMenuList is really focused before notifying
2272         the platform that it is focused. Likewise, only notify the platform of
2273         an AccessibilityMenuListOption focus gain if the AccessibilityMenuList
2274         containing that option is expanded.
2275
2276         Test: platform/gtk/accessibility/menu-list-unfocused-notifications.html
2277
2278         * accessibility/atk/AXObjectCacheAtk.cpp:
2279         (WebCore::notifyChildrenSelectionChange):
2280         (WebCore::AXObjectCache::postPlatformNotification):
2281
2282 2014-11-13  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
2283
2284         [EFL] Unexpected crashes in layout test after r175953
2285         https://bugs.webkit.org/show_bug.cgi?id=138652
2286
2287         Reviewed by Csaba Osztrogonác.
2288
2289         In r175953, it set reusableNodeIndex value with m_partCache's size(). It causes
2290         out of bound index crash because the reusableNodeIndex is Vector object.
2291
2292         No new test. This patch is to fix broken tests related to widgets on EFL port.
2293
2294         * platform/efl/RenderThemeEfl.cpp:
2295         (WebCore::RenderThemeEfl::getThemePartFromCache): Need to set reusableNodeIndex with "size() - 1".
2296
2297 2014-11-13  Chris Dumez  <cdumez@apple.com>
2298
2299         Lazily create HTMLInputElement's inputType and shadow subtree
2300         https://bugs.webkit.org/show_bug.cgi?id=138524
2301
2302         Reviewed by Ryosuke Niwa.
2303
2304         When an HTMLInputElement was created by the parser, we would first call
2305         HTMLInputElement::create(), then call Element::parserSetAttributes() on
2306         the constructed input. With the previous implementation, this was a bit
2307         inefficient because HTMLInputElement::create() would construct a
2308         TextInputType inputType (as this is the default) as well as its
2309         corresponding shadow subtree. Then, parserSetAttributes() would often
2310         set the |type| attribute and would need to destroy this input type as
2311         well as its subtree if the new |type| is not 'text', to create a new
2312         inputType / shadow subtree of the right type. The profiler showed that
2313         this was fairly expensive.
2314
2315         To improve this, this patch delays the inputType / shadow subtree
2316         creation when the HTMLInputElement is constructed by the parser, until
2317         the attributes are actually set by the parser. This way, we directly
2318         create an inputType / shadow subtree of the right type.
2319
2320         I see a 1.4% speed up on speedometer (73.95 -> 75.0).
2321
2322         Test: fast/dom/HTMLInputElement/border-attribute-crash.html
2323
2324         * dom/Element.cpp:
2325         (WebCore::Element::parserSetAttributes):
2326         (WebCore::Element::parserDidSetAttributes):
2327         * dom/Element.h:
2328         * html/HTMLInputElement.cpp:
2329         (WebCore::HTMLInputElement::HTMLInputElement):
2330         (WebCore::HTMLInputElement::create):
2331         (WebCore::HTMLInputElement::updateType):
2332         (WebCore::HTMLInputElement::runPostTypeUpdateTasks):
2333         (WebCore::HTMLInputElement::initializeInputType):
2334         (WebCore::HTMLInputElement::parseAttribute):
2335         (WebCore::HTMLInputElement::parserDidSetAttributes):
2336         (WebCore::HTMLInputElement::finishParsingChildren):
2337         * html/HTMLInputElement.h:
2338
2339 2014-11-12  Chris Dumez  <cdumez@apple.com>
2340
2341         Have DOMTimer deal with more ScriptExecutionContext references
2342         https://bugs.webkit.org/show_bug.cgi?id=138679
2343
2344         Reviewed by Andreas Kling.
2345
2346         Have DOMTimer deal with more ScriptExecutionContext references instead
2347         of pointers, to make it clear the script execution context cannot be
2348         null.
2349
2350         No new tests, no behavior change.
2351
2352         * bindings/js/ScheduledAction.cpp:
2353         (WebCore::ScheduledAction::execute):
2354         (WebCore::ScheduledAction::executeFunctionInContext):
2355         * bindings/js/ScheduledAction.h:
2356         * dom/DocumentEventQueue.cpp:
2357         * page/DOMTimer.cpp:
2358         (WebCore::DOMTimerFireState::DOMTimerFireState):
2359         (WebCore::NestedTimersMap::instanceForContext):
2360         (WebCore::DOMTimer::DOMTimer):
2361         (WebCore::DOMTimer::install):
2362         (WebCore::DOMTimer::removeById):
2363         (WebCore::DOMTimer::fired):
2364         * page/DOMTimer.h:
2365         * page/DOMWindow.cpp:
2366         (WebCore::DOMWindow::setTimeout):
2367         (WebCore::DOMWindow::clearTimeout):
2368         (WebCore::DOMWindow::setInterval):
2369         (WebCore::DOMWindow::clearInterval):
2370         * page/SuspendableTimer.cpp:
2371         (WebCore::SuspendableTimer::SuspendableTimer):
2372         * page/SuspendableTimer.h:
2373         * workers/WorkerGlobalScope.cpp:
2374         (WebCore::WorkerGlobalScope::setTimeout):
2375         (WebCore::WorkerGlobalScope::clearTimeout):
2376         (WebCore::WorkerGlobalScope::setInterval):
2377         (WebCore::WorkerGlobalScope::clearInterval):
2378
2379 2014-11-12  Dean Jackson  <dino@apple.com>
2380
2381         Support unprefixed animation property names
2382         https://bugs.webkit.org/show_bug.cgi?id=138678
2383         <rdar://problem/18943059>
2384
2385         Reviewed by Simon Fraser.
2386
2387         Step 1 of 3 in unprefixing CSS animations: handling the
2388         property names and values.
2389
2390         Unlike other unprefixing, where we use an alias that
2391         effectively removes the prefixed values from the code,
2392         for animations we need to keep the old values around so that
2393         existing content will not break (e.g. computed style).
2394
2395         During testing I noticed that we didn't handle the
2396         (relatively) new animation-direction values of
2397         "reverse" and "alternate-reverse" when querying
2398         computed style.
2399
2400         Tests: animations/unprefixed-properties.html
2401                animations/unprefixed-shorthand.html
2402
2403         * css/CSSComputedStyleDeclaration.cpp:
2404         (WebCore::ComputedStyleExtractor::propertyValue): Handle the unprefixed
2405         values, but also get animation direction to produce the correct
2406         results when "reverse" and "alternate-reverse" is specified.
2407
2408         * css/CSSParser.cpp: Handle the new values. If necessary, pass a flag
2409         around indicating whether it is prefixed or unprefixed.
2410         (WebCore::CSSParser::parseValue):
2411         (WebCore::CSSParser::parseAnimationShorthand): Use the flag to decide
2412         which of the forms we're going to check.
2413         (WebCore::CSSParser::parseAnimationProperty):
2414         * css/CSSParser.h:
2415
2416         * css/CSSProperty.h:
2417         (WebCore::prefixingVariantForPropertyId): Add prefixing/unprefixing variants for
2418         all the animation values, and clean up the code a bit.
2419
2420         * css/CSSPropertyNames.in: Add new properties.
2421
2422         * css/DeprecatedStyleBuilder.cpp:
2423         (WebCore::DeprecatedStyleBuilder::DeprecatedStyleBuilder): We need handlers for the
2424         unprefixed forms. This should move to the new StyleBuilder soon.
2425
2426         * css/StyleProperties.cpp: Handle new values.
2427         (WebCore::StyleProperties::getPropertyValue):
2428         (WebCore::StyleProperties::asText):
2429
2430         * css/StylePropertyShorthand.cpp:
2431         (WebCore::animationShorthand):
2432         (WebCore::animationShorthandForParsing): Decide which list of properties to use.
2433         (WebCore::shorthandForProperty):
2434         (WebCore::matchingShorthandsForLonghand):
2435         (WebCore::webkitAnimationShorthandForParsing): Deleted.
2436         * css/StylePropertyShorthand.h:
2437
2438         * css/StyleResolver.cpp: Since the new properties come before "background" in
2439         alphabetical order, they are earlier in the CSSPropertyNames.in file, and thus
2440         we need to update the call sites that think CSSPropertyBackground is the first
2441         property.
2442         (WebCore::StyleResolver::styleForKeyframe): Replace CSSPropertyBackground with CSSPropertyAnimation.
2443         (WebCore::StyleResolver::styleForPage):
2444         (WebCore::StyleResolver::applyMatchedProperties):
2445         (WebCore::StyleResolver::applyProperty):
2446
2447 2014-11-12  Beth Dakin  <bdakin@apple.com>
2448
2449         Speculative build fix.
2450
2451         * WebCore.exp.in:
2452
2453 2014-11-12  Timothy Horton  <timothy_horton@apple.com>
2454
2455         Fix the iOS build.
2456
2457         * platform/spi/mac/DataDetectorsSPI.h:
2458         It's likely that more of this file should be compiled out on iOS,
2459         but this is all we need to build.
2460
2461 2014-11-12  Beth Dakin  <bdakin@apple.com>
2462
2463         WK1: Support default actions for images
2464         https://bugs.webkit.org/show_bug.cgi?id=138668
2465         -and corresponding-
2466         rdar://problem/18785434
2467
2468         Reviewed by Tim Horton.
2469
2470         Symbol needed for copying image.
2471         * WebCore.exp.in:
2472
2473 2014-11-05  Ada Chan  <adachan@apple.com>
2474
2475         Implement new plug-in API for muting plug-ins
2476         https://bugs.webkit.org/show_bug.cgi?id=138105
2477
2478         Reviewed by Anders Carlsson.
2479
2480         Add the new NPNVmuteAudioBool NPNVariable.
2481
2482         Test: platform/mac-wk2/plugins/muted-state.html
2483
2484         * plugins/npapi.h:
2485
2486 2014-11-12  Tim Horton  <timothy_horton@apple.com>
2487
2488         Need to show the text indicator when Data Detectors shows a popover
2489         https://bugs.webkit.org/show_bug.cgi?id=138664
2490         <rdar://problem/18869900>
2491
2492         Reviewed by Beth Dakin.
2493
2494         * platform/spi/mac/DataDetectorsSPI.h:
2495         Add new DataDetectors SPI.
2496
2497 2014-11-12  Chris Dumez  <cdumez@apple.com>
2498
2499         Minor improvements to RenderListItem
2500         https://bugs.webkit.org/show_bug.cgi?id=138601
2501
2502         Reviewed by Darin Adler.
2503
2504         Make several minor improvements to RenderListItem and clean up the
2505         code a bit.
2506
2507         No new tests, no behavior change.
2508
2509         * rendering/RenderListItem.cpp:
2510         (WebCore::isHTMLListElement):
2511         - Rename isList to isHTMLListElement() for clarity as it checks
2512           for HTMLOListElement and HTMLUListElement types
2513         - Inline the function as it is short.
2514         - Use is<HTML*Element>() for type checking
2515         - Update the argument to take a Node& instead of an Element*. This
2516           is because the argument is expected to the non-null. Also using
2517           looser typing here allows us to use this utility function is one
2518           more place, without impacting performance as
2519           hasTagName(HTMLQualifiedName) is defined on Node.
2520
2521         (WebCore::enclosingList):
2522         - Pass the argument as a reference instead of a pointer as it is
2523           expected to be non-null.
2524         - Initialize firstNode before the loop to avoid the if (!firstNode)
2525           check for every iteration.
2526
2527         (WebCore::nextListItem):
2528         - Take an Element as second argument instead of a RenderListItem*
2529           and provide convenience overloads so that we don't need to do
2530           null checks just because some calls sites call this function
2531           with 2 arguments and others with 1 argument. This way, we avoid
2532           unnecessary null checks as most call sites already do such
2533           checks (or have references).
2534         - Transform the while loop into a for loop for clarity.
2535         - Don't traverse an Element's subtree if the Element does not have
2536           a renderer as it is impossible of any of its descendant to have
2537           a renderer (and we are looking for a specific type of renderer).
2538
2539         (WebCore::previousListItem):
2540         - Pass the item argument as a reference instead of a pointer as it
2541           is expected to be non-null.
2542         - Reduce the scope of the |current| so that it is now declared
2543           inside the loop.
2544
2545         (WebCore::RenderListItem::updateItemValuesForOrderedList):
2546         (WebCore::RenderListItem::itemCountForOrderedList):
2547         - Pass argument as a reference instead of a pointer as it was expected
2548           to be non-null (There was an assertion in place to make sure of it).
2549
2550         (WebCore::RenderListItem::calcValue):
2551         - Use is<HTMLOListElement>() instead of hasTagName().
2552
2553         (WebCore::getParentOfFirstLineBox):
2554         - Rename variables to stop using abbreviations.
2555         - Pass arguments as references instead of pointers as they are expected
2556           to be non-null.
2557         - Remove the firstChild null check before the loop as it does not
2558           change behavior. The loop will abort early if firstChild is null
2559           and we will end up returning nullptr as well.
2560         - Use is<>() more for type checking.
2561         - Reuse the isHTMLListElement() utility function instead of duplicating
2562           its code inside this function.
2563
2564         (WebCore::firstNonMarkerChild):
2565         - Pass argument as a reference as it is expected to be non-null.
2566         - Rename variable from result to child for clarity, as we are traversing
2567           the children.
2568
2569         (WebCore::RenderListItem::markerTextWithSuffix):
2570         - Use String appending instead of StringBuilder as it simplifies the
2571           code a lot and should not impact performance in this case.
2572
2573         (WebCore::RenderListItem::explicitValueChanged):
2574         - Restructure the code a bit to do the |listNode| null check before the
2575           loop, now that nextListItem() takes a reference in argument. This is
2576           the only call site where we didn't already know that listNode is
2577           non-null.
2578
2579         (WebCore::previousOrNextItem):
2580         - Mark this function as inline as it is short and called repeatedly.
2581
2582 2014-11-12  Chris Dumez  <cdumez@apple.com>
2583
2584         Speed up HTMLTextFormControlElement::setInnerTextValue() a bit
2585         https://bugs.webkit.org/show_bug.cgi?id=138619
2586
2587         Reviewed by Darin Adler.
2588
2589         Speed up HTMLTextFormControlElement::setInnerTextValue() a bit by:
2590         - Not doing a virtual isTextFormControl() call. Relying on innerTextElement()
2591           not returning null is sufficient.
2592         - Caching the result of innerTextElement() instead of repeatedly calling that
2593           virtual function.
2594         - De-virtualizing setFormControlValueMatchesRenderer() /
2595           formControlValueMatchesRenderer() as these are never overridden.
2596         - Moving the code constructing the innerTextValue from a
2597           TextControlInnerTextElement from innerTextValue() to a new
2598           innerTextValueFrom(TextControlInnerTextElement&) function and call this new
2599           function from setInnerTextValue() to avoid calling innerTextElement() again
2600           and making sure it is non-null
2601         - Do the tree traversal from innerTextElement's firstChild instead of from
2602           innerTextElement. The innerTextElement is a TextControlInnerTextElement.
2603           Therefore, it cannot be an HTMLBRElement or a Text node.
2604
2605         No new tests, no behavior change.
2606
2607         * html/HTMLFormControlElement.h:
2608         (WebCore::HTMLFormControlElement::formControlValueMatchesRenderer):
2609         (WebCore::HTMLFormControlElement::setFormControlValueMatchesRenderer):
2610         * html/HTMLTextFormControlElement.cpp:
2611         (WebCore::stripTrailingNewline):
2612         (WebCore::innerTextValueFrom):
2613         (WebCore::HTMLTextFormControlElement::setInnerTextValue):
2614         (WebCore::HTMLTextFormControlElement::innerTextValue):
2615         (WebCore::HTMLTextFormControlElement::valueWithHardLineBreaks):
2616         (WebCore::finishText): Deleted.
2617
2618 2014-11-12  Jer Noble  <jer.noble@apple.com>
2619
2620         [Mac] media/track/audio-track.html is flakey
2621         https://bugs.webkit.org/show_bug.cgi?id=138394
2622
2623         Reviewed by Alexey Proskuryakov.
2624
2625         The above test fails if the HTMLMediaElement's "canplaythrough" event fires before the AudioTrackList's
2626         "addtrack" event fires. This can happen because each object keeps its own GenericEventQueue, which
2627         empties when that individual queue's timer fires. So events can be enqueued in the following order:
2628         1, A, 2, B, 3, C; but fired in a different order: 1, 2, 3, A, B, C.
2629
2630         Make events enqueued in GenericEventQueues globally ordered, so that events are fired in the order
2631         which they are enqueued, regardless of which object owns the queue. Use a static queue of
2632         GenericEventQueues to manage which GenericEventQueue fires and in what order.
2633
2634         GenericEventQueues will use a WeakPtrFactory to cancel pending events. Revoking a GenericEventQueue's
2635         weak pointers means those queue's entries will be skipped the next time the meta-queue is processed.
2636
2637         * dom/GenericEventQueue.cpp:
2638         (WebCore::GenericEventQueue::GenericEventQueue): Create a WeakPtrFactory instead of a Timer.
2639         (WebCore::GenericEventQueue::enqueueEvent): Enqueue this queue with the MetaQueue.
2640         (WebCore::GenericEventQueue::sharedTimer): Lazily-initializing accessor.
2641         (WebCore::GenericEventQueue::sharedTimerFired):  Ask each queue to dispatch one event.
2642         (WebCore::GenericEventQueue::pendingQueues): Lazily-initializing accessor.
2643         (WebCore::GenericEventQueue::dispatchOneEvent): Renamed from timerFired.
2644         (WebCore::GenericEventQueue::close): Revoke all WeakPtrs.
2645         (WebCore::GenericEventQueue::cancelAllEvents): Ditto.
2646         (WebCore::GenericEventQueue::hasPendingEvents): Use !isEmpty().
2647         (WebCore::GenericEventQueue::timerFired): Deleted.
2648         * dom/GenericEventQueue.h:
2649
2650 2014-11-12  Carlos Garcia Campos  <cgarcia@igalia.com>
2651
2652         [GTK] Expose user script messages to GObject DOM bindings
2653         https://bugs.webkit.org/show_bug.cgi?id=138411
2654
2655         Reviewed by Philippe Normand.
2656
2657         * PlatformGTK.cmake: Add WebKitDOMCustomUnstable.h to the list of
2658         installed headers and make a symlink for it in the DerivedSources dir.
2659         * bindings/gobject/WebKitDOMCustomUnstable.h: Added.
2660         * bindings/gobject/WebKitDOMCustom.cpp:
2661         (webkit_dom_dom_window_get_webkit_namespace):
2662         (webkit_dom_user_message_handlers_namespace_get_handler):
2663         * bindings/scripts/CodeGeneratorGObject.pm:
2664         (HasUnstableCustomAPI): Helper function to check if the given
2665         class has custom unstable API.
2666         (WriteData): Include WebKitDOMCustomUnstable.h header if the class
2667         has API defined there.
2668
2669 2014-11-12  Carlos Garcia Campos  <cgarcia@igalia.com>
2670
2671         [GTK] Fix how SerializedScriptValue is exposed to GObject DOM bindings
2672         https://bugs.webkit.org/show_bug.cgi?id=138409
2673
2674         Reviewed by Philippe Normand.
2675
2676         It's currently exposed as a WebKitDOMSerializedScriptValue class
2677         that doesn't really exist. It should be handled as a string, that
2678         is used to create a SerializedScriptValue when coming from the
2679         API, and converted to a string again when returned to the API.
2680
2681         * bindings/scripts/CodeGeneratorGObject.pm:
2682         (GetGlibTypeName): Return gchar* as glib type of SerializedScriptValue.
2683         (IsGDOMClassType): Don't consider SerializedScriptValue as a GDOM class.
2684         (GenerateFunction): Handle SerializedScriptValue parameters and
2685         return values specially.
2686         * bindings/scripts/test/GObject/WebKitDOMTestCallback.cpp:
2687         (webkit_dom_test_callback_callback_with_serialized_script_value_param):
2688         * bindings/scripts/test/GObject/WebKitDOMTestCallback.h:
2689         * bindings/scripts/test/GObject/WebKitDOMTestObj.cpp:
2690         (webkit_dom_test_obj_serialized_value):
2691         * bindings/scripts/test/GObject/WebKitDOMTestObj.h:
2692         * bindings/scripts/test/GObject/WebKitDOMTestSerializedScriptValueInterface.cpp:
2693         (webkit_dom_test_serialized_script_value_interface_class_init):
2694         (webkit_dom_test_serialized_script_value_interface_get_value):
2695         (webkit_dom_test_serialized_script_value_interface_set_value):
2696         (webkit_dom_test_serialized_script_value_interface_get_readonly_value):
2697         (webkit_dom_test_serialized_script_value_interface_get_cached_value):
2698         (webkit_dom_test_serialized_script_value_interface_set_cached_value):
2699         (webkit_dom_test_serialized_script_value_interface_get_cached_readonly_value):
2700         * bindings/scripts/test/GObject/WebKitDOMTestSerializedScriptValueInterface.h:
2701         * bindings/scripts/test/GObject/WebKitDOMTestTypedefs.cpp:
2702         (webkit_dom_test_typedefs_class_init):
2703         (webkit_dom_test_typedefs_get_immutable_serialized_script_value):
2704         (webkit_dom_test_typedefs_set_immutable_serialized_script_value):
2705         * bindings/scripts/test/GObject/WebKitDOMTestTypedefs.h:
2706
2707 2014-11-12  Carlos Garcia Campos  <cgarcia@igalia.com>
2708
2709         [GTK] Move RedirectedXCompositeWindow from platform to WebKit2 layer
2710         https://bugs.webkit.org/show_bug.cgi?id=138093
2711
2712         Reviewed by Martin Robinson.
2713
2714         Remove RedirectedXCompositeWindow.
2715
2716         * PlatformGTK.cmake:
2717
2718 2014-09-01  Philippe Normand  <pnormand@igalia.com>
2719
2720         [WK2] UserMediaClient support
2721         https://bugs.webkit.org/show_bug.cgi?id=123158
2722
2723         Reviewed by Benjamin Poulain.
2724
2725         Added document and frame accessors, needed by the
2726         UserMediaRequestManager of the WebProcess. Also added two methods
2727         used by the permission request manager to know if access to audio
2728         or video devices is requested.
2729
2730         Tests: fast/mediastream/error.html
2731                fast/mediastream/success.html
2732                fast/mediastream/delayed-permission-allowed.html
2733                fast/mediastream/delayed-permission-denied.html
2734                http/tests/media/media-stream/disconnected-frame-already.html
2735                http/tests/media/media-stream/disconnected-frame-permission-denied.html
2736                http/tests/media/media-stream/disconnected-frame.html
2737
2738         * Modules/mediastream/UserMediaRequest.cpp:
2739         (WebCore::UserMediaRequest::document):
2740         (WebCore::UserMediaRequest::frame):
2741         * Modules/mediastream/UserMediaRequest.h:
2742         (WebCore::UserMediaRequest::requiresAudio):
2743         (WebCore::UserMediaRequest::requiresVideo):
2744
2745 2014-11-12  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
2746
2747         [EFL] Clean up OwnPtr in RenderThemeEfl
2748         https://bugs.webkit.org/show_bug.cgi?id=138642
2749
2750         Reviewed by Zoltan Herczeg.
2751
2752         No new tests, no behavior change.
2753
2754         * platform/efl/RenderThemeEfl.cpp:
2755         (WebCore::RenderThemeEfl::applyEdjeRTLState): Clean up all remaining OwnPtr.
2756
2757 2014-11-11  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
2758
2759         Move Modules/websockets to std::unique_ptr
2760         https://bugs.webkit.org/show_bug.cgi?id=138649
2761
2762         Reviewed by Chris Dumez.
2763
2764         Remove create() factory function and use std::make_unique<>, std::unique_ptr<>
2765         instead of OwnPtr, PassOwnPtr.
2766         However didReceiveBinaryData() is still using PassOwnPtr. It should be replaced
2767         with std::unique_ptr<>
2768
2769         No new tests, no behavior.
2770
2771         * Modules/websockets/ThreadableWebSocketChannelClientWrapper.h:
2772         * Modules/websockets/WebSocket.cpp:
2773         * Modules/websockets/WebSocketChannel.cpp:
2774         (WebCore::WebSocketChannel::connect):
2775         (WebCore::WebSocketChannel::didFail):
2776         (WebCore::WebSocketChannel::processFrame):
2777         (WebCore::WebSocketChannel::enqueueTextFrame):
2778         (WebCore::WebSocketChannel::enqueueRawFrame):
2779         (WebCore::WebSocketChannel::enqueueBlobFrame):
2780         (WebCore::WebSocketChannel::processOutgoingFrameQueue):
2781         (WebCore::WebSocketChannel::sendFrame):
2782         * Modules/websockets/WebSocketChannel.h:
2783         * Modules/websockets/WebSocketDeflateFramer.cpp:
2784         (WebCore::WebSocketDeflateFramer::createExtensionProcessor):
2785         (WebCore::WebSocketDeflateFramer::enableDeflate):
2786         (WebCore::WebSocketDeflateFramer::deflate):
2787         (WebCore::WebSocketDeflateFramer::inflate):
2788         (WebCore::WebSocketExtensionDeflateFrame::create): Deleted.
2789         * Modules/websockets/WebSocketDeflateFramer.h:
2790         (WebCore::DeflateResultHolder::create): Deleted.
2791         (WebCore::InflateResultHolder::create): Deleted.
2792         * Modules/websockets/WebSocketDeflater.cpp:
2793         (WebCore::WebSocketDeflater::WebSocketDeflater):
2794         (WebCore::WebSocketInflater::WebSocketInflater):
2795         (WebCore::WebSocketDeflater::create): Deleted.
2796         (WebCore::WebSocketInflater::create): Deleted.
2797         * Modules/websockets/WebSocketDeflater.h:
2798         * Modules/websockets/WebSocketExtensionDispatcher.cpp:
2799         (WebCore::WebSocketExtensionDispatcher::addProcessor):
2800         * Modules/websockets/WebSocketExtensionDispatcher.h:
2801         * Modules/websockets/WebSocketHandshake.cpp:
2802         (WebCore::WebSocketHandshake::addExtensionProcessor):
2803         * Modules/websockets/WebSocketHandshake.h:
2804
2805 2014-11-11  Chris Dumez  <cdumez@apple.com>
2806
2807         Unreviewed iOS build fix after r176001.
2808
2809         Export an additional symbol.
2810
2811         * WebCore.exp.in:
2812
2813 2014-11-11  Jer Noble  <jer.noble@apple.com>
2814
2815         [Mac] Add diagnostic logging for per-media-engine load failures
2816         https://bugs.webkit.org/show_bug.cgi?id=138647
2817
2818         Reviewed by Eric Carlson.
2819
2820         Add diagnostic logging fired whenever a media engine fails to load media,
2821         even if another engine subsequentially succeeds. Add a mechanism for retrieving
2822         the platform-specific error code from a given engine.
2823
2824         * html/HTMLMediaElement.cpp:
2825         (WebCore::HTMLMediaElement::parseAttribute):
2826         * html/HTMLMediaElement.h:
2827         * page/DiagnosticLoggingKeys.cpp:
2828         (WebCore::DiagnosticLoggingKeys::engineFailedToLoadKey):
2829         * page/DiagnosticLoggingKeys.h:
2830         * platform/graphics/MediaPlayer.cpp:
2831         (WebCore::MediaPlayer::networkStateChanged):
2832         (WebCore::MediaPlayer::platformErrorCode):
2833         * platform/graphics/MediaPlayer.h:
2834         (WebCore::MediaPlayerClient::mediaPlayerEngineFailedToLoad):
2835         * platform/graphics/MediaPlayerPrivate.h:
2836         (WebCore::MediaPlayerPrivateInterface::platformErrorCode):
2837         * platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.h:
2838         (WebCore::MediaPlayerPrivateAVFoundation::platformErrorCode):
2839         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.cpp:
2840         (WebCore::MediaPlayerPrivateAVFoundationCF::assetErrorCode):
2841         * platform/graphics/avfoundation/cf/MediaPlayerPrivateAVFoundationCF.h:
2842         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
2843         * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
2844         (WebCore::MediaPlayerPrivateAVFoundationObjC::assetErrorCode):
2845         * platform/graphics/mac/MediaPlayerPrivateQTKit.h:
2846         * platform/graphics/mac/MediaPlayerPrivateQTKit.mm:
2847         (WebCore::MediaPlayerPrivateQTKit::platformErrorCode):
2848
2849 2014-11-11  Chris Dumez  <cdumez@apple.com>
2850
2851         Mark more virtual functions in HTMLInputElement as final
2852         https://bugs.webkit.org/show_bug.cgi?id=138644
2853
2854         Reviewed by Andreas Kling.
2855
2856         Mark more virtual functions in HTMLInputElement as final to help the
2857         compiler optimize better.
2858
2859         No new tests, no behavior change.
2860
2861         * html/HTMLInputElement.h:
2862
2863 2014-11-11  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
2864
2865         Remove IncrementalSweeper::create()
2866         https://bugs.webkit.org/show_bug.cgi?id=138243
2867
2868         Reviewed by Filip Pizlo.
2869
2870         As a step to use std::unique_ptr<> and std::make_unique<>, this patch removes
2871         IncrementalSweeper::create(), then set constructor of IncrementalSweeper to public.
2872         Now we begins to use std::make_unique<> to create IncrementalSweeper instance.
2873
2874         No new tests, no behavior changes.
2875
2876         * bindings/js/JSDOMWindowBase.cpp:
2877         (WebCore::JSDOMWindowBase::commonVM):
2878         * platform/ios/WebSafeIncrementalSweeperIOS.h:
2879
2880 2014-11-11  Chris Dumez  <cdumez@apple.com>
2881
2882         Move 'border-image-*' / '-webkit-mask-box-image-*' CSS properties to the new StyleBuilder
2883         https://bugs.webkit.org/show_bug.cgi?id=138594
2884
2885         Reviewed by Antti Koivisto.
2886
2887         Move 'border-image-*' / '-webkit-mask-box-image-*' CSS properties from
2888         DeprecatedStyleBuilder to the new StyleBuilder using custom code.
2889
2890         No new tests, no behavior change.
2891
2892         * css/CSSPropertyNames.in:
2893         * css/DeprecatedStyleBuilder.cpp:
2894         (WebCore::DeprecatedStyleBuilder::DeprecatedStyleBuilder):
2895         (WebCore::ApplyPropertyBorderImageModifier::getValue): Deleted.
2896         (WebCore::ApplyPropertyBorderImageModifier::setValue): Deleted.
2897         (WebCore::ApplyPropertyBorderImageModifier::applyInheritValue): Deleted.
2898         (WebCore::ApplyPropertyBorderImageModifier::applyInitialValue): Deleted.
2899         (WebCore::ApplyPropertyBorderImageModifier::applyValue): Deleted.
2900         (WebCore::ApplyPropertyBorderImageModifier::createHandler): Deleted.
2901         * css/StyleBuilderCustom.h:
2902         (WebCore::StyleBuilderFunctions::ApplyPropertyBorderImageModifier::applyInheritValue):
2903         (WebCore::StyleBuilderFunctions::ApplyPropertyBorderImageModifier::applyInitialValue):
2904         (WebCore::StyleBuilderFunctions::ApplyPropertyBorderImageModifier::applyValue):
2905         (WebCore::StyleBuilderFunctions::ApplyPropertyBorderImageModifier::getValue):
2906         (WebCore::StyleBuilderFunctions::ApplyPropertyBorderImageModifier::setValue):
2907
2908 2014-11-11  David Kilzer  <ddkilzer@apple.com>
2909
2910         Modernize $buildingForIPhone in CodeGeneratorObjC.pm
2911         <http://webkit.org/b/138640>
2912
2913         Reviewed by Daniel Bates.
2914
2915         Mac OS X is the only platform that supports NSColor, so rename
2916         $buildingForIPhone to $shouldUseCGColor (which is actually what
2917         the variable represents) and define it in terms of the "macosx"
2918         platform name.
2919
2920         * bindings/scripts/CodeGeneratorObjC.pm:
2921         ($buildingForIPhone): Rename to $shouldUseCGColor.
2922         ($shouldUseCGColor): False if PLATFORM_NAME is not defined,
2923         which is the case on older Xcodes when building for Mac OS X, or
2924         if PLATFORM_NAME is equal to "macosx".  True otherwise.
2925         (GetClassName): Switch to use new variable name.
2926         (AddIncludesForType): Ditto.
2927         (GenerateHeader): Remove dead code.  The
2928         $buildingForTigerOrEarlier variable was accidentally
2929         re-introduced in r161638, but was originally removed in r127032!
2930         (GenerateImplementation): Switch to use new variable name.
2931
2932 2014-11-11  Tim Horton  <timothy_horton@apple.com>
2933
2934         Add support for mailto: link action menus
2935         https://bugs.webkit.org/show_bug.cgi?id=138641
2936         <rdar://problem/18741567>
2937
2938         Reviewed by Beth Dakin.
2939
2940         * WebCore.exp.in:
2941         Add an export.
2942
2943         * platform/spi/mac/DataDetectorsSPI.h:
2944         Add some DataDetectors SPI.
2945
2946 2014-11-11  Eric Carlson  <eric.carlson@apple.com>
2947
2948         [iOS] video is sometimes allowed to play from the background
2949         https://bugs.webkit.org/show_bug.cgi?id=138522
2950
2951         Rubber-stamped by Jer Noble.
2952
2953         * platform/audio/MediaSession.cpp:
2954         (WebCore::MediaSession::updateClientDataBuffering): Call MediaSessionManager::sessionCanLoadMedia.
2955
2956 2014-11-11  Eric Carlson  <eric.carlson@apple.com>
2957
2958         [iOS] a page with video in optimized fullscreen is visible
2959         https://bugs.webkit.org/show_bug.cgi?id=138632
2960
2961         Reviewed by Anders Carlsson.
2962
2963         * platform/ios/WebVideoFullscreenInterfaceAVKit.h:
2964         (WebCore::WebVideoFullscreenInterfaceAVKit::mode): Expose fullscreen mode.
2965
2966 2014-11-11  Myles C. Maxfield  <mmaxfield@apple.com>
2967
2968         Mountain Lion build fix
2969         https://bugs.webkit.org/show_bug.cgi?id=138631
2970
2971         Reviewed by Simon Fraser.
2972
2973         * platform/spi/cocoa/CoreTextSPI.h:
2974
2975 2014-11-11  Eric Carlson  <eric.carlson@apple.com>
2976
2977         [iOS] cleanup wireless route monitoring
2978         https://bugs.webkit.org/show_bug.cgi?id=138614
2979
2980         Reviewed by Jer Noble.
2981
2982         * html/HTMLMediaElement.cpp:
2983         (WebCore::HTMLMediaElement::clearMediaPlayer): Add logging.
2984         (WebCore::HTMLMediaElement::removeEventListener): Ditto.
2985         (WebCore::HTMLMediaElement::createMediaPlayer): Ditto.
2986
2987         * platform/audio/ios/MediaSessionManagerIOS.mm:
2988         (WebCore::MediaSessionManageriOS::resetRestrictions): Add logging.
2989         (WebCore::MediaSessionManageriOS::configureWireLessTargetMonitoring): Ditto.
2990         (WebCore::MediaSessionManageriOS::updateNowPlayingInfo): Ditto.
2991         (-[WebMediaSessionHelper initWithCallback:]): Ditto.
2992         (-[WebMediaSessionHelper dealloc]): Always deallocate volume view and routing controller
2993             on the main thread.
2994         (-[WebMediaSessionHelper clearCallback]): Add logging.
2995         (-[WebMediaSessionHelper hasWirelessTargetsAvailable]): Ditto.
2996         (-[WebMediaSessionHelper startMonitoringAirPlayRoutes]): Always call routing controller on
2997             the main thread.
2998         (-[WebMediaSessionHelper stopMonitoringAirPlayRoutes]): Ditto.
2999         (-[WebMediaSessionHelper interruption:]): Add logging.
3000         (-[WebMediaSessionHelper applicationWillEnterForeground:]): Ditto.
3001         (-[WebMediaSessionHelper applicationDidBecomeActive:]): Ditto.
3002         (-[WebMediaSessionHelper applicationWillResignActive:]): Ditto.
3003         (-[WebMediaSessionHelper wirelessRoutesAvailableDidChange:]): Ditto.
3004
3005 2014-11-11  Myles C. Maxfield  <mmaxfield@apple.com>
3006
3007         Unreviewd iOS build fix
3008
3009         * platform/spi/cocoa/CoreTextSPI.h:
3010
3011 2014-11-11  David Kilzer  <ddkilzer@apple.com>
3012
3013         Protect Document in ProcessingInstruction::setXSLStyleSheet()
3014         <http://webkit.org/b/138621>
3015
3016         Reviewed by Andreas Kling.
3017
3018         The patch is inspired by the following Blink revision by
3019         <tasak@google.com>:
3020         <https://src.chromium.org/viewvc/blink?view=rev&revision=182309>
3021
3022         * dom/ProcessingInstruction.cpp:
3023         (WebCore::ProcessingInstruction::setXSLStyleSheet):
3024
3025 2014-11-11  Myles C. Maxfield  <mmaxfield@apple.com>
3026
3027         Finish moving CTFontGetVerticalGlyphsForCharacters and CTLineCreateWithUniCharProvider out from WKSI
3028         https://bugs.webkit.org/show_bug.cgi?id=138623
3029
3030         Reviewed by Geoff Garen.
3031
3032         No new tests because there is no behavior change.
3033
3034         * WebCore.exp.in:
3035         * WebCore.order:
3036         * platform/graphics/mac/GlyphPageTreeNodeMac.cpp:
3037         (WebCore::GlyphPage::fill):
3038         * platform/graphics/mac/SimpleFontDataMac.mm:
3039         (WebCore::SimpleFontData::canRenderCombiningCharacterSequence):
3040         * platform/ios/WebCoreSystemInterfaceIOS.mm:
3041         * platform/mac/WebCoreSystemInterface.h:
3042         * platform/mac/WebCoreSystemInterface.mm:
3043         * platform/spi/cocoa/CoreTextSPI.h:
3044
3045 2014-11-11  Myles C. Maxfield  <mmaxfield@apple.com>
3046
3047         Unreviewed Mountain Lion build fix
3048
3049         * platform/spi/cocoa/CoreTextSPI.h:
3050
3051 2014-11-11  Commit Queue  <commit-queue@webkit.org>
3052
3053         Unreviewed, rolling out r175852.
3054         https://bugs.webkit.org/show_bug.cgi?id=138626
3055
3056         Broke PLT by introducing a crash. (Requested by rniwa on
3057         #webkit).
3058
3059         Reverted changeset:
3060
3061         "Lazily create HTMLInputElement's inputType and shadow
3062         subtree"
3063         https://bugs.webkit.org/show_bug.cgi?id=138524
3064         http://trac.webkit.org/changeset/175852
3065
3066 2014-11-11  Chris Dumez  <cdumez@apple.com>
3067
3068         Regression(r175947): Caused assertions in debug builds
3069         https://bugs.webkit.org/show_bug.cgi?id=138620
3070
3071         Reviewed by Benjamin Poulain.
3072
3073         In HTMLCollection::traverseForward(), traversedCount was incremented 1
3074         time too many when hitting the end of the collection (i.e. element
3075         becomes null). Doing a partial revert.
3076
3077         No new tests, already covered by existing tests.
3078
3079         * html/HTMLCollection.cpp:
3080         (WebCore::HTMLCollection::traverseForward):
3081
3082 2014-11-11  Tim Horton  <timothy_horton@apple.com>
3083
3084         DataDetectors' menu items aren't presented in the telephone number menu
3085         https://bugs.webkit.org/show_bug.cgi?id=138602
3086         <rdar://problem/18866308>
3087
3088         Reviewed by Beth Dakin.
3089
3090         * English.lproj/Localizable.strings:
3091         Remove a now-unused localized string.
3092
3093 2014-11-11  Daniel Bates  <dabates@apple.com>
3094
3095         [Frame Flattening] ASSERT(transaction->view == &view()) fails in RenderBlock::removeFromUpdateScrollInfoAfterLayoutTransaction()
3096         https://bugs.webkit.org/show_bug.cgi?id=138615
3097         <rdar://problem/18928487>
3098
3099         Reviewed by David Hyatt.
3100
3101         Fixes an issue where we always expected that the RenderView associated with the RenderBlock called in
3102         RenderBlock::removeFromUpdateScrollInfoAfterLayoutTransaction() is the same as the RenderView associated with
3103         the top transaction in the UpdateScrollInfoAfterLayoutTransaction stack regardless of frame flattening. This
3104         expectation is only true when frame flattening is disabled.
3105
3106         Test: fast/frames/flattening/iframe-flattening-inside-flexbox-with-delayed-scroll-update.html
3107
3108         * rendering/RenderBlock.cpp:
3109         (WebCore::RenderBlock::removeFromUpdateScrollInfoAfterLayoutTransaction):
3110
3111 2014-11-11  Tim Horton  <timothy_horton@apple.com>
3112
3113         Adopt DDActionContext menu type SPI
3114         https://bugs.webkit.org/show_bug.cgi?id=138603
3115         <rdar://problem/18867627>
3116
3117         Reviewed by Anders Carlsson.
3118
3119         * platform/spi/mac/DataDetectorsSPI.h:
3120         Add the new SPI.
3121
3122 2014-11-11  Tim Horton  <timothy_horton@apple.com>
3123
3124         Data Detectors popovers should be dismissed upon scrolling
3125         https://bugs.webkit.org/show_bug.cgi?id=138600
3126         <rdar://problem/18932770>
3127
3128         Reviewed by Beth Dakin.
3129
3130         * platform/spi/mac/DataDetectorsSPI.h:
3131         Adjust the SPI header.
3132
3133 2014-11-11  Tim Horton  <timothy_horton@apple.com>
3134
3135         Localize strings for action menus
3136         https://bugs.webkit.org/show_bug.cgi?id=138591
3137         <rdar://problem/18815343>
3138
3139         Reviewed by Dan Bates.
3140
3141         * English.lproj/Localizable.strings:
3142         Add a bunch of localizable strings.
3143
3144 2014-11-11  Myles C. Maxfield  <mmaxfield@apple.com>
3145
3146         Cleanup from r175379
3147         https://bugs.webkit.org/show_bug.cgi?id=138616
3148
3149         Reviewed by Simon Fraser.
3150
3151         No new tests because there is no behavior change.
3152
3153         * WebCore.exp.in:
3154         * platform/ios/WebCoreSystemInterfaceIOS.mm:
3155         * platform/mac/WebCoreSystemInterface.h:
3156         * platform/mac/WebCoreSystemInterface.mm:
3157
3158 2014-11-11  Myles C. Maxfield  <mmaxfield@apple.com>
3159
3160         Move CTFontTransformGlyphs out from WKSI
3161         https://bugs.webkit.org/show_bug.cgi?id=138599
3162
3163         Reviewed by Simon Fraser.
3164
3165         No new tests because there is no behavior change.
3166
3167         * WebCore.exp.in:
3168         * WebCore.order:
3169         * platform/graphics/SimpleFontData.cpp:
3170         (WebCore::SimpleFontData::applyTransforms):
3171         * platform/graphics/SimpleFontData.h:
3172         * platform/ios/WebCoreSystemInterfaceIOS.mm:
3173         * platform/mac/WebCoreSystemInterface.h:
3174         * platform/mac/WebCoreSystemInterface.mm:
3175         * platform/spi/cocoa/CoreTextSPI.h:
3176
3177 2014-11-11  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
3178
3179         Use std::unique_ptr<>|std::make_unique_ptr in RenderThemeEfl::ThemePartCacheEntry::create()
3180         https://bugs.webkit.org/show_bug.cgi?id=138558
3181
3182         Reviewed by Anders Carlsson.
3183
3184         This patch removes PassOwnPtr and OwnPtr in RenderThemeEfl::ThemePartCacheEntry::create().
3185         It would be good if we also remove create() factory function though, in this case, it is
3186         a little hard to remove it since the create() factory function does many works to create
3187         a ThemePartCacheEntry instance. Thus this patch just replaces PassOwnPtr|OwnPtr with
3188         std::unique_ptr<> and std::make_unique<> in RenderThemeEfl::ThemePartCacheEntry::create().
3189
3190         Besides the theme entries have been maintained by Eina_List*. Unfortunately it doesn't support
3191         std::unique_ptr<> yet. Thus the Eina_List* should be replaced with Vector<std::unique_ptr<>>
3192         so that m_partCache has ownship of all EFL theme entries. This patch tries to keep original
3193         cache algorithm based on Vector class.
3194
3195         No new tests, no behavior changes.
3196
3197         * platform/efl/RenderThemeEfl.cpp:
3198         (WebCore::RenderThemeEfl::ThemePartCacheEntry::create):
3199         (WebCore::RenderThemeEfl::getThemePartFromCache):
3200         * platform/efl/RenderThemeEfl.h:
3201
3202 2014-11-09  Darin Adler  <darin@apple.com>
3203
3204         Minor tweaks to HTMLCollection
3205         https://bugs.webkit.org/show_bug.cgi?id=138556
3206
3207         Reviewed by Chris Dumez.
3208
3209         * accessibility/AccessibilityRenderObject.cpp:
3210         (WebCore::AccessibilityRenderObject::addRadioButtonGroupMembers):
3211         Update for namedElements function that now returns a Vector; also use
3212         a new style for loop to iterate it.
3213
3214         * bindings/js/JSHTMLAllCollectionCustom.cpp: Removed some unneeded includes.
3215         (WebCore::namedItems): Updated name to match WebKit coding style, and also
3216         updated to use the return value from namedItems, which now returns a Vector.
3217         (WebCore::callHTMLAllCollection): Updated for namedItems name change.
3218         Also removed explicit Node* type from result of namedItemWithIndex, since that
3219         function now returns a more specific type.
3220         (WebCore::JSHTMLAllCollection::nameGetter): Update for namedItems name change.
3221         (WebCore::JSHTMLAllCollection::item): Ditto.
3222         (WebCore::JSHTMLAllCollection::namedItem): Ditto.
3223
3224         * bindings/js/JSHTMLFormControlsCollectionCustom.cpp: Removed some unneeded includes.
3225         (WebCore::namedItems): Updated name to match WebKit coding style, and also
3226         updated to use the return value from namedItems, which now returns a Vector.
3227         (WebCore::JSHTMLFormControlsCollection::nameGetter): Update for namedItems name change.
3228         (WebCore::JSHTMLFormControlsCollection::namedItem): Ditto.
3229
3230         * bindings/js/JSHTMLFormElementCustom.cpp:
3231         (WebCore::JSHTMLFormElement::nameGetter): Updated to use the return value from
3232         namedItems, which now returns a Vector.
3233
3234         * html/HTMLAllCollection.cpp:
3235         (WebCore::HTMLAllCollection::HTMLAllCollection): Marked the constructor inline,
3236         since it's only used in one place, the create function.
3237         (WebCore::HTMLAllCollection::~HTMLAllCollection): Deleted. No need to have an
3238         explicit destructor since there's nothing special to implement, and includers of
3239         the header file have everything they ened to compile the compiler-generated one.
3240         (WebCore::HTMLAllCollection::namedItemWithIndex): Changed return type to Element.
3241
3242         * html/HTMLAllCollection.h: Removed unneeded explicit declaration of destructor.
3243         Chagned return type of namedItemWithIndex to Element.
3244
3245         * html/HTMLCollection.cpp:
3246         (WebCore::HTMLCollection::rootTypeFromCollectionType): Marked this inline. Also
3247         changed this to be a static member function so it can use the RootType enum, which
3248         is now private to the class.
3249         (WebCore::isMatchingHTMLElement): Marked this function static so it will get
3250         internal linkage.
3251         (WebCore::isMatchingElement): Ditto.
3252         (WebCore::previousElement): Marked this function inline since it's called in only
3253         one place. Changed argument type to a reference since it can never be null.
3254         (WebCore::HTMLCollection::iterateForPreviousElement): Changed argument name and
3255         also updated for above changes.
3256         (WebCore::firstMatchingElement): Marked this function static so it will get
3257         internal linkage.
3258         (WebCore::nextMatchingElement): Ditto. Changed argument type to a reference
3259         since it can never be null.
3260         (WebCore::HTMLCollection::item): Changed return type to Element.
3261         (WebCore::nameShouldBeVisibleInDocumentAll): Added an overload that takes an
3262         Element. This streamlines the code below that calls it so it fits on one line.
3263         (WebCore::firstMatchingChildElement): Marked this function static so it will get
3264         internal linkage.
3265         (WebCore::nextMatchingSiblingElement): Ditto. Changed argument type to a reference
3266         since it can never be null.
3267         (WebCore::HTMLCollection::usesCustomForwardOnlyTraversal): Moved here from the
3268         header since, although it's marked inline, it's only used inside this file.
3269         (WebCore::HTMLCollection::traverseForward): Restructured the code a little bit
3270         to make the function smaller and possibly easier to read. This does add one
3271         redundant null check, but it seems OK to do that.
3272         (WebCore::HTMLCollection::collectionTraverseBackward): Tweaked foramtting a bit.
3273         (WebCore::HTMLCollection::namedItem): Changed return type to Element. Tightened
3274         the code that calls nameShouldBeVisibleInDocumentAll so it fits better on one line.
3275         Changed code that handles m_shouldOnlyIncludeDirectChildren to use a nested if
3276         instead of an && since it makes the code a little easier to read.
3277         (WebCore::HTMLCollection::updateNamedElementCache): Tweaked code a little bit,
3278         using shorter variable names, and using references instead of pointers. Also removed
3279         the call to didPopulate, since setNamedItemCache now takes care of that.
3280         (WebCore::HTMLCollection::namedItems): Changed to return a Vector instead of
3281         appending to an existing one. Also use reserveInitialCapacity and uncheckedAppend
3282         for better performance. Added a FIXME, because there seems to be something wrong
3283         here about this being non-virtual. Made other small tweaks to streamline the code.
3284         (WebCore::HTMLCollection::customElementAfter): Moved this here from the header.
3285         There is no reason to need to inline this.
3286
3287         * html/HTMLCollection.h: Removed unneeded includes. Moved function bodies out
3288         of the class definitions so the class definitions are easier to read. Made some
3289         functions that were formerly public or protected be private instead. Added a call
3290         to didPopulate to setNamedItemCache so the callers don't have to do it.
3291
3292         * html/HTMLFormControlsCollection.cpp:
3293         (WebCore::HTMLFormControlsCollection::namedItem): Changed return value to Element.
3294         Tweaked coding style a little bit.
3295         (WebCore::HTMLFormControlsCollection::updateNamedElementCache): Rearranged to
3296         simplify a bit. Don't build the foundInputElements set when the owner is not a
3297         form element, since we don't use the set in that case. Use shorter variable names,
3298         and modern for loops. Also removed the call to didPopulate, since setNamedItemCache
3299         now takes care of that.
3300
3301         * html/HTMLFormControlsCollection.h: Removed some uneeded forward declarations.
3302         Updated return type for namedItem, and also made the override private.
3303
3304         * html/HTMLFormElement.cpp:
3305         (WebCore::HTMLFormElement::namedElements): Changed to return a Vector and updated
3306         function name accordingly.
3307         * html/HTMLFormElement.h: Ditto.
3308
3309         * html/HTMLNameCollection.h: Removed a stray blank line.
3310
3311         * html/HTMLSelectElement.cpp:
3312         (WebCore::HTMLSelectElement::namedItem): Changed return value to Element.
3313         (WebCore::HTMLSelectElement::item): Ditto.
3314         * html/HTMLSelectElement.h: Ditto.
3315
3316         * page/scrolling/AxisScrollSnapOffsets.cpp:
3317         (WebCore::appendChildSnapOffsets): Rewrote loop as a for loop rather than a while
3318         loop. Removed unwanted use of children()->collectionBegin() to get the first element
3319         child of an HTMLElement. This function uses a mix of DOM and rendering functions that
3320         is probably incorrect, but I did not tackle fixing that at this time.
3321
3322 2014-11-11  Dan Bernstein  <mitz@apple.com>
3323
3324         [Mac] WebCore includes unused cursor image resources
3325         https://bugs.webkit.org/show_bug.cgi?id=138605
3326
3327         Reviewed by Anders Carlsson.
3328
3329         * Resources/aliasCursor.png: Removed.
3330         * Resources/cellCursor.png: Removed.
3331         * Resources/contextMenuCursor.png: Removed.
3332         * Resources/copyCursor.png: Removed.
3333         * Resources/eastResizeCursor.png: Removed.
3334         * Resources/eastWestResizeCursor.png: Removed.
3335         * Resources/helpCursor.png: Removed.
3336         * Resources/linkCursor.png: Removed.
3337         * Resources/moveCursor.png: Removed.
3338         * Resources/noDropCursor.png: Removed.
3339         * Resources/noneCursor.png: Removed.
3340         * Resources/northEastResizeCursor.png: Removed.
3341         * Resources/northEastSouthWestResizeCursor.png: Removed.
3342         * Resources/northResizeCursor.png: Removed.
3343         * Resources/northSouthResizeCursor.png: Removed.
3344         * Resources/northWestResizeCursor.png: Removed.
3345         * Resources/northWestSouthEastResizeCursor.png: Removed.
3346         * Resources/progressCursor.png: Removed.
3347         * Resources/southEastResizeCursor.png: Removed.
3348         * Resources/southResizeCursor.png: Removed.
3349         * Resources/southWestResizeCursor.png: Removed.
3350         * Resources/verticalTextCursor.png: Removed.
3351         * Resources/waitCursor.png: Removed.
3352         * Resources/westResizeCursor.png: Removed.
3353         * Resources/zoomInCursor.png: Removed.
3354         * Resources/zoomOutCursor.png: Removed.
3355
3356         * WebCore.xcodeproj/project.pbxproj: Removed references to the above files.
3357
3358         * platform/mac/CursorMac.mm:
3359         (WebCore::Cursor::ensurePlatformCursor): For Help, Cell, ZoomIn and ZoomOut, don’t check
3360         for a nil return value from wkCursor(), which can’t occur in any supported configuration.
3361         For None, create a cursor with an empty NSImage instead of loading a transparent image from
3362         the file system.
3363         (WebCore::createNamedCursor): Deleted.
3364
3365 2014-11-11  Gyuyoung Kim  <gyuyoung.kim@samsung.com>
3366
3367         Unreviewed, rolling out r175844.
3368
3369         Win debug build has broken since r175844.
3370
3371         Reverted changeset:
3372
3373         "Use std::unique_ptr<> and std::make_unique<> in WebCore"
3374         https://bugs.webkit.org/show_bug.cgi?id=138560
3375         http://trac.webkit.org/changeset/175844
3376
3377 2014-11-11  Philippe Normand  <pnormand@igalia.com>
3378
3379         Qt/Chromium remnants in MediaPlayer
3380         https://bugs.webkit.org/show_bug.cgi?id=138611
3381
3382         Reviewed by Jer Noble.
3383
3384         * platform/graphics/MediaPlayer.h: Remove unused Qt/Chromium stuff.
3385
3386 2014-11-11  Chris Dumez  <cdumez@apple.com>
3387
3388         Move '-webkit-marquee-speed' CSS property to the new StyleBuilder
3389         https://bugs.webkit.org/show_bug.cgi?id=138595
3390
3391         Reviewed by Antti Koivisto.
3392
3393         Move '-webkit-marquee-speed' CSS property from DeprecatedStyleBuilder
3394         to the new StyleBuilder by adding a new MarqueeSpeed Converter.
3395
3396         No new tests, no behavior change.
3397
3398         * css/CSSPropertyNames.in:
3399         * css/DeprecatedStyleBuilder.cpp:
3400         (WebCore::DeprecatedStyleBuilder::DeprecatedStyleBuilder):
3401         (WebCore::ApplyPropertyMarqueeSpeed::applyValue): Deleted.
3402         (WebCore::ApplyPropertyMarqueeSpeed::createHandler): Deleted.
3403         * css/StyleBuilderConverter.h:
3404         (WebCore::StyleBuilderConverter::convertMarqueeSpeed):
3405
3406 2014-10-07  Sergio Villar Senin  <svillar@igalia.com>
3407
3408         [CSS Grid Layout] Limit the size of explicit/implicit grid
3409         https://bugs.webkit.org/show_bug.cgi?id=136217
3410
3411         Reviewed by Andreas Kling.
3412
3413         A recent change in the specs allow us to set a limit (maximum
3414         size) for both the explicit and implicit grids so we can protect
3415         ourselves from absurdly huge grid specifications. It was decided
3416         to use the recommended limit which is 1 million.
3417
3418         This means that we can remove the old limitation for the number of
3419         repetitions in repeat(). Instead we now clamp the number of
3420         repetitions to the maximum value that allow us t