Source/WebCore/ChangeLog

   1 2015-12-01  Carlos Garcia Campos  <cgarcia@igalia.com>
   2
   3         Unreviewed. Fix GTK+ build after r192849.
   4
   5         * bindings/scripts/CodeGeneratorGObject.pm:
   6         (GenerateFunction):
   7         * bindings/scripts/test/GObject/WebKitDOMTestInterface.cpp:
   8         (webkit_dom_test_interface_supplemental_method1):
   9         (webkit_dom_test_interface_supplemental_method2):
  10         (webkit_dom_test_interface_set_supplemental_str2):
  11         (webkit_dom_test_interface_get_supplemental_node):
  12         (webkit_dom_test_interface_set_supplemental_node):
  13
  14 2015-12-01  Youenn Fablet  <youenn.fablet@crf.canon.fr>
  15
  16         [Streams API] Remove use of @catch for exposed promises
  17         https://bugs.webkit.org/show_bug.cgi?id=151625
  18
  19         Reviewed by Darin Adler.
  20
  21         Promise @catch is calling "then" which may be controlled by user scripts.
  22         This patch simply replaces @catch by calling @then directly.
  23
  24         Covered by modified tests.
  25
  26         * Modules/streams/ReadableStream.js:
  27         (pipeTo):
  28         * Modules/streams/ReadableStreamInternals.js:
  29         (teeReadableStream):
  30
  31 2015-11-30  Jaehun Lim  <ljaehun.lim@samsung.com>
  32
  33         Unreviewed, fix build after r192848 and r192849
  34
  35         - Rename canSuspendForPageCache to canSuspendForDocumentSuspension
  36         - Use references instead of pointers
  37
  38         * Modules/battery/BatteryManager.cpp:
  39         (WebCore::BatteryManager::canSuspendForDocumentSuspension):
  40         (WebCore::BatteryManager::canSuspendForPageCache): Deleted.
  41         * Modules/battery/BatteryManager.h:
  42         * Modules/battery/NavigatorBattery.cpp:
  43         (WebCore::NavigatorBattery::webkitBattery):
  44         * Modules/battery/NavigatorBattery.h:
  45         * Modules/gamepad/deprecated/NavigatorGamepad.cpp:
  46         (WebCore::NavigatorGamepad::webkitGetGamepads):
  47         * Modules/gamepad/deprecated/NavigatorGamepad.h:
  48         * Modules/navigatorcontentutils/NavigatorContentUtils.cpp:
  49         (WebCore::NavigatorContentUtils::registerProtocolHandler):
  50         (WebCore::NavigatorContentUtils::isProtocolHandlerRegistered):
  51         (WebCore::NavigatorContentUtils::unregisterProtocolHandler):
  52         * Modules/navigatorcontentutils/NavigatorContentUtils.h:
  53         * Modules/vibration/NavigatorVibration.cpp:
  54         (WebCore::NavigatorVibration::vibrate):
  55         * Modules/vibration/NavigatorVibration.h:
  56
  57 2015-11-30  Jiewen Tan  <jiewen_tan@apple.com>
  58
  59         Amazon.com Additional Information links aren't clickable
  60         https://bugs.webkit.org/show_bug.cgi?id=151401
  61         <rdar://problem/23454261>
  62
  63         Reviewed by Darin Adler.
  64
  65         The cause of this issue is that the painting order is different from the hittest order so we can end up
  66         with visible but unreachable content. To fix this, the executation flow of hittest has been reordered.
  67         According to the paint system, which renders the webpage from the bottom RenderLayer to the top, contents
  68         are rendered before floats. Hence, for the hittest, which determines the hitted location from top RenderLayer
  69         to the bottom, should do it reversedly. Now, hittest will first test floats then contents.
  70
  71         Test: fast/block/float/hit-test-on-overlapping-floats.html
  72
  73         * rendering/RenderBlock.cpp:
  74         (WebCore::RenderBlock::nodeAtPoint):
  75
  76 2015-11-30  Simon Fraser  <simon.fraser@apple.com>
  77
  78         Fix possible crash with animated layers in reflections
  79         https://bugs.webkit.org/show_bug.cgi?id=151689
  80         rdar://problem/23018612
  81
  82         Reviewed by Darin Adler.
  83
  84         Reflections create additional PlatformCALayers whose owner is set to the GraphicsLayerCA.
  85         Those PlatformCALayers need their owner pointer cleared out when the GraphicsLayerCA
  86         is destroyed.
  87
  88         Tested by compositing/reflections/nested-reflection-transition.html
  89
  90         * platform/graphics/ca/GraphicsLayerCA.cpp:
  91         * platform/graphics/ca/GraphicsLayerCA.h:
  92
  93 2015-11-30  Brady Eidson  <beidson@apple.com>
  94
  95         Modern IDB: Iterating index cursors to a specific key is busted.
  96         https://bugs.webkit.org/show_bug.cgi?id=151684
  97
  98         Reviewed by Darin Adler.
  99
 100         No new tests (At least one failing test now passes).
 101
 102         * Modules/indexeddb/server/MemoryIndexCursor.cpp:
 103         (WebCore::IDBServer::MemoryIndexCursor::iterate):
 104
 105 2015-11-30  Brady Eidson  <beidson@apple.com>
 106
 107         Modern IDB: ObjectStore cursors should not be able to iterate out of their range.
 108         https://bugs.webkit.org/show_bug.cgi?id=151683
 109
 110         Reviewed by Darin Adler.
 111
 112         No new tests (Covered by at least one failing test that now passes).
 113
 114         * Modules/indexeddb/server/MemoryObjectStoreCursor.cpp:
 115         (WebCore::IDBServer::MemoryObjectStoreCursor::incrementForwardIterator):
 116         (WebCore::IDBServer::MemoryObjectStoreCursor::incrementReverseIterator):
 117
 118 2015-11-30  Anders Carlsson  <andersca@apple.com>
 119
 120         CTTE autogenerated bindings code
 121         https://bugs.webkit.org/show_bug.cgi?id=151682
 122
 123         Reviewed by Darin Adler.
 124
 125         Make sure that JS bindings pass a reference to the object when calling static member functions.
 126
 127         * Modules/gamepad/NavigatorGamepad.cpp:
 128         (WebCore::NavigatorGamepad::getGamepads):
 129         * Modules/gamepad/NavigatorGamepad.h:
 130         * Modules/geolocation/NavigatorGeolocation.cpp:
 131         (WebCore::NavigatorGeolocation::geolocation):
 132         * Modules/geolocation/NavigatorGeolocation.h:
 133         * Modules/mediasource/AudioTrackMediaSource.h:
 134         (WebCore::AudioTrackMediaSource::sourceBuffer):
 135         * Modules/mediasource/TextTrackMediaSource.h:
 136         (WebCore::TextTrackMediaSource::sourceBuffer):
 137         * Modules/mediasource/VideoTrackMediaSource.h:
 138         (WebCore::VideoTrackMediaSource::sourceBuffer):
 139         * Modules/mediastream/HTMLMediaElementMediaStream.cpp:
 140         (WebCore::HTMLMediaElementMediaStream::srcObject):
 141         (WebCore::HTMLMediaElementMediaStream::setSrcObject):
 142         * Modules/mediastream/HTMLMediaElementMediaStream.h:
 143         * Modules/mediastream/NavigatorMediaDevices.cpp:
 144         (WebCore::NavigatorMediaDevices::mediaDevices):
 145         * Modules/mediastream/NavigatorMediaDevices.h:
 146         * Modules/notifications/DOMWindowNotifications.cpp:
 147         (WebCore::DOMWindowNotifications::webkitNotifications):
 148         * Modules/notifications/DOMWindowNotifications.h:
 149         * Modules/notifications/Notification.cpp:
 150         (WebCore::Notification::Notification):
 151         * Modules/notifications/WorkerGlobalScopeNotifications.cpp:
 152         (WebCore::WorkerGlobalScopeNotifications::webkitNotifications):
 153         * Modules/notifications/WorkerGlobalScopeNotifications.h:
 154         * Modules/speech/DOMWindowSpeechSynthesis.cpp:
 155         (WebCore::DOMWindowSpeechSynthesis::speechSynthesis):
 156         * Modules/speech/DOMWindowSpeechSynthesis.h:
 157         * Modules/webdatabase/DOMWindowWebDatabase.cpp:
 158         (WebCore::DOMWindowWebDatabase::openDatabase):
 159         * Modules/webdatabase/DOMWindowWebDatabase.h:
 160         * bindings/scripts/CodeGeneratorJS.pm:
 161         (GenerateImplementation):
 162         (GenerateParametersCheck):
 163         * testing/Internals.cpp:
 164         (WebCore::Internals::enableMockSpeechSynthesizer):
 165
 166 2015-11-30  Katlyn Graff  <kgraff@apple.com>
 167
 168         Rename ActiveDOMObject/DOMWindow PageCacheSuspension code to support more reasons for suspension
 169         https://bugs.webkit.org/show_bug.cgi?id=151677
 170
 171         Reviewed by Ryosuke Niwa.
 172
 173         Simply a refactoring patch, so no new tests.
 174
 175         * Modules/encryptedmedia/MediaKeySession.cpp:
 176         (WebCore::MediaKeySession::canSuspendForDocumentSuspension):
 177         (WebCore::MediaKeySession::canSuspendForPageCache): Deleted.
 178         * Modules/encryptedmedia/MediaKeySession.h:
 179         * Modules/geolocation/Geolocation.cpp:
 180         (WebCore::Geolocation::canSuspendForDocumentSuspension):
 181         (WebCore::Geolocation::canSuspendForPageCache): Deleted.
 182         * Modules/geolocation/Geolocation.h:
 183         * Modules/indexeddb/DOMWindowIndexedDatabase.cpp:
 184         (WebCore::DOMWindowIndexedDatabase::disconnectFrameForDocumentSuspension):
 185         (WebCore::DOMWindowIndexedDatabase::reconnectFrameFromDocumentSuspension):
 186         (WebCore::DOMWindowIndexedDatabase::disconnectFrameForPageCache): Deleted.
 187         (WebCore::DOMWindowIndexedDatabase::reconnectFrameFromPageCache): Deleted.
 188         * Modules/indexeddb/DOMWindowIndexedDatabase.h:
 189         * Modules/indexeddb/client/IDBDatabaseImpl.cpp:
 190         (WebCore::IDBClient::IDBDatabase::canSuspendForDocumentSuspension):
 191         (WebCore::IDBClient::IDBDatabase::canSuspendForPageCache): Deleted.
 192         * Modules/indexeddb/client/IDBDatabaseImpl.h:
 193         * Modules/indexeddb/client/IDBRequestImpl.cpp:
 194         (WebCore::IDBClient::IDBRequest::canSuspendForDocumentSuspension):
 195         (WebCore::IDBClient::IDBRequest::canSuspendForPageCache): Deleted.
 196         * Modules/indexeddb/client/IDBRequestImpl.h:
 197         * Modules/indexeddb/client/IDBTransactionImpl.cpp:
 198         (WebCore::IDBClient::IDBTransaction::canSuspendForDocumentSuspension):
 199         (WebCore::IDBClient::IDBTransaction::canSuspendForPageCache): Deleted.
 200         * Modules/indexeddb/client/IDBTransactionImpl.h:
 201         * Modules/indexeddb/legacy/LegacyDatabase.cpp:
 202         (WebCore::LegacyDatabase::canSuspendForDocumentSuspension):
 203         (WebCore::LegacyDatabase::canSuspendForPageCache): Deleted.
 204         * Modules/indexeddb/legacy/LegacyDatabase.h:
 205         * Modules/indexeddb/legacy/LegacyRequest.cpp:
 206         (WebCore::LegacyRequest::canSuspendForDocumentSuspension):
 207         (WebCore::LegacyRequest::canSuspendForPageCache): Deleted.
 208         * Modules/indexeddb/legacy/LegacyRequest.h:
 209         * Modules/indexeddb/legacy/LegacyTransaction.cpp:
 210         (WebCore::LegacyTransaction::canSuspendForDocumentSuspension):
 211         (WebCore::LegacyTransaction::canSuspendForPageCache): Deleted.
 212         * Modules/indexeddb/legacy/LegacyTransaction.h:
 213         * Modules/mediasource/MediaSource.cpp:
 214         (WebCore::MediaSource::canSuspendForDocumentSuspension):
 215         (WebCore::MediaSource::canSuspendForPageCache): Deleted.
 216         * Modules/mediasource/MediaSource.h:
 217         * Modules/mediasource/SourceBuffer.cpp:
 218         (WebCore::SourceBuffer::canSuspendForDocumentSuspension):
 219         (WebCore::SourceBuffer::canSuspendForPageCache): Deleted.
 220         * Modules/mediasource/SourceBuffer.h:
 221         * Modules/mediastream/MediaStreamTrack.cpp:
 222         (WebCore::MediaStreamTrack::canSuspendForDocumentSuspension):
 223         (WebCore::MediaStreamTrack::canSuspendForPageCache): Deleted.
 224         * Modules/mediastream/MediaStreamTrack.h:
 225         * Modules/mediastream/RTCDTMFSender.cpp:
 226         (WebCore::RTCDTMFSender::canSuspendForDocumentSuspension):
 227         (WebCore::RTCDTMFSender::canSuspendForPageCache): Deleted.
 228         * Modules/mediastream/RTCDTMFSender.h:
 229         * Modules/mediastream/RTCPeerConnection.cpp:
 230         (WebCore::RTCPeerConnection::canSuspendForDocumentSuspension):
 231         (WebCore::RTCPeerConnection::canSuspendForPageCache): Deleted.
 232         * Modules/mediastream/RTCPeerConnection.h:
 233         * Modules/notifications/DOMWindowNotifications.cpp:
 234         (WebCore::DOMWindowNotifications::disconnectFrameForDocumentSuspension):
 235         (WebCore::DOMWindowNotifications::reconnectFrameFromDocumentSuspension):
 236         (WebCore::DOMWindowNotifications::disconnectFrameForPageCache): Deleted.
 237         (WebCore::DOMWindowNotifications::reconnectFrameFromPageCache): Deleted.
 238         * Modules/notifications/DOMWindowNotifications.h:
 239         * Modules/notifications/Notification.cpp:
 240         (WebCore::Notification::canSuspendForDocumentSuspension):
 241         (WebCore::Notification::canSuspendForPageCache): Deleted.
 242         * Modules/notifications/Notification.h:
 243         * Modules/notifications/NotificationCenter.cpp:
 244         (WebCore::NotificationCenter::canSuspendForDocumentSuspension):
 245         (WebCore::NotificationCenter::canSuspendForPageCache): Deleted.
 246         * Modules/notifications/NotificationCenter.h:
 247         * Modules/webaudio/AudioContext.cpp:
 248         (WebCore::AudioContext::canSuspendForDocumentSuspension):
 249         (WebCore::AudioContext::canSuspendForPageCache): Deleted.
 250         * Modules/webaudio/AudioContext.h:
 251         * Modules/webdatabase/DatabaseContext.cpp:
 252         (WebCore::DatabaseContext::canSuspendForDocumentSuspension):
 253         (WebCore::DatabaseContext::canSuspendForPageCache): Deleted.
 254         * Modules/webdatabase/DatabaseContext.h:
 255         * Modules/websockets/WebSocket.cpp:
 256         (WebCore::WebSocket::canSuspendForDocumentSuspension):
 257         (WebCore::WebSocket::canSuspendForPageCache): Deleted.
 258         * Modules/websockets/WebSocket.h:
 259         * css/FontLoader.cpp:
 260         (WebCore::FontLoader::canSuspendForDocumentSuspension):
 261         (WebCore::FontLoader::canSuspendForPageCache): Deleted.
 262         * css/FontLoader.h:
 263         * dom/ActiveDOMObject.cpp:
 264         (WebCore::ActiveDOMObject::canSuspendForDocumentSuspension):
 265         (WebCore::ActiveDOMObject::canSuspendForPageCache): Deleted.
 266         * dom/ActiveDOMObject.h:
 267         * dom/Document.cpp:
 268         (WebCore::Document::~Document):
 269         * dom/ScriptExecutionContext.cpp:
 270         (WebCore::ScriptExecutionContext::canSuspendActiveDOMObjectsForTabSuspension):
 271         (WebCore::ScriptExecutionContext::canSuspendActiveDOMObjectsForPageCache): Deleted.
 272         * dom/ScriptExecutionContext.h:
 273         * fileapi/FileReader.cpp:
 274         (WebCore::FileReader::canSuspendForDocumentSuspension):
 275         (WebCore::FileReader::canSuspendForPageCache): Deleted.
 276         * fileapi/FileReader.h:
 277         * history/CachedFrame.cpp:
 278         (WebCore::CachedFrame::CachedFrame):
 279         * history/PageCache.cpp:
 280         (WebCore::canCacheFrame):
 281         * html/HTMLMarqueeElement.cpp:
 282         (WebCore::HTMLMarqueeElement::canSuspendForDocumentSuspension):
 283         (WebCore::HTMLMarqueeElement::canSuspendForPageCache): Deleted.
 284         * html/HTMLMarqueeElement.h:
 285         * html/HTMLMediaElement.cpp:
 286         (WebCore::HTMLMediaElement::canSuspendForDocumentSuspension):
 287         (WebCore::HTMLMediaElement::canSuspendForPageCache): Deleted.
 288         * html/HTMLMediaElement.h:
 289         * html/HTMLSourceElement.cpp:
 290         (WebCore::HTMLSourceElement::canSuspendForDocumentSuspension):
 291         (WebCore::HTMLSourceElement::canSuspendForPageCache): Deleted.
 292         * html/HTMLSourceElement.h:
 293         * html/PublicURLManager.cpp:
 294         (WebCore::PublicURLManager::canSuspendForDocumentSuspension):
 295         (WebCore::PublicURLManager::canSuspendForPageCache): Deleted.
 296         * html/PublicURLManager.h:
 297         * html/canvas/WebGLRenderingContextBase.cpp:
 298         (WebCore::WebGLRenderingContextBase::canSuspendForDocumentSuspension):
 299         (WebCore::WebGLRenderingContextBase::canSuspendForPageCache): Deleted.
 300         * html/canvas/WebGLRenderingContextBase.h:
 301         * loader/FrameLoader.cpp:
 302         (WebCore::FrameLoader::clear):
 303         (WebCore::FrameLoader::open):
 304         * loader/appcache/DOMApplicationCache.cpp:
 305         (WebCore::DOMApplicationCache::disconnectFrameForDocumentSuspension):
 306         (WebCore::DOMApplicationCache::reconnectFrameFromDocumentSuspension):
 307         (WebCore::DOMApplicationCache::disconnectFrameForPageCache): Deleted.
 308         (WebCore::DOMApplicationCache::reconnectFrameFromPageCache): Deleted.
 309         * loader/appcache/DOMApplicationCache.h:
 310         * page/DOMWindow.cpp:
 311         (WebCore::DOMWindow::DOMWindow):
 312         (WebCore::DOMWindow::~DOMWindow):
 313         (WebCore::DOMWindow::resetUnlessSuspendedForDocumentSuspension):
 314         (WebCore::DOMWindow::suspendForDocumentSuspension):
 315         (WebCore::DOMWindow::resumeFromDocumentSuspension):
 316         (WebCore::DOMWindow::disconnectDOMWindowProperties):
 317         (WebCore::DOMWindow::reconnectDOMWindowProperties):
 318         (WebCore::DOMWindow::resetUnlessSuspendedForPageCache): Deleted.
 319         (WebCore::DOMWindow::suspendForPageCache): Deleted.
 320         (WebCore::DOMWindow::resumeFromPageCache): Deleted.
 321         * page/DOMWindow.h:
 322         * page/DOMWindowExtension.cpp:
 323         (WebCore::DOMWindowExtension::disconnectFrameForDocumentSuspension):
 324         (WebCore::DOMWindowExtension::reconnectFrameFromDocumentSuspension):
 325         (WebCore::DOMWindowExtension::disconnectFrameForPageCache): Deleted.
 326         (WebCore::DOMWindowExtension::reconnectFrameFromPageCache): Deleted.
 327         * page/DOMWindowExtension.h:
 328         * page/DOMWindowProperty.cpp:
 329         (WebCore::DOMWindowProperty::disconnectFrameForDocumentSuspension):
 330         (WebCore::DOMWindowProperty::reconnectFrameFromDocumentSuspension):
 331         (WebCore::DOMWindowProperty::disconnectFrameForPageCache): Deleted.
 332         (WebCore::DOMWindowProperty::reconnectFrameFromPageCache): Deleted.
 333         * page/DOMWindowProperty.h:
 334         * page/EventSource.cpp:
 335         (WebCore::EventSource::canSuspendForDocumentSuspension):
 336         (WebCore::EventSource::canSuspendForPageCache): Deleted.
 337         * page/EventSource.h:
 338         * page/SuspendableTimer.cpp:
 339         (WebCore::SuspendableTimer::canSuspendForDocumentSuspension):
 340         (WebCore::SuspendableTimer::canSuspendForPageCache): Deleted.
 341         * page/SuspendableTimer.h:
 342         * workers/Worker.cpp:
 343         (WebCore::Worker::canSuspendForDocumentSuspension):
 344         (WebCore::Worker::canSuspendForPageCache): Deleted.
 345         * workers/Worker.h:
 346         * xml/XMLHttpRequest.cpp:
 347         (WebCore::XMLHttpRequest::canSuspendForDocumentSuspension):
 348         (WebCore::XMLHttpRequest::canSuspendForPageCache): Deleted.
 349         * xml/XMLHttpRequest.h:
 350
 351 2015-11-30  Brady Eidson  <beidson@apple.com>
 352
 353         Modern IDB: "prevunique" cursors should point at the lowest primary key that matches, not the highest.
 354         https://bugs.webkit.org/show_bug.cgi?id=151675.
 355
 356         Reviewed by Darin Adler.
 357
 358         No new tests (Covered by at least one failing test that now passes, and updates to previously incorrect tests).
 359
 360         * Modules/indexeddb/server/IndexValueEntry.cpp:
 361         (WebCore::IDBServer::IndexValueEntry::reverseBegin): If CursorDuplicity is NoDuplicates, start at the lowest
 362           entry instead of the highest.
 363         (WebCore::IDBServer::IndexValueEntry::reverseFind):
 364         * Modules/indexeddb/server/IndexValueEntry.h:
 365
 366         * Modules/indexeddb/server/IndexValueStore.cpp:
 367         (WebCore::IDBServer::IndexValueStore::reverseFind):
 368         (WebCore::IDBServer::IndexValueStore::Iterator::Iterator):
 369         (WebCore::IDBServer::IndexValueStore::Iterator::nextIndexEntry):
 370         * Modules/indexeddb/server/IndexValueStore.h:
 371
 372         * Modules/indexeddb/server/MemoryIndexCursor.cpp:
 373         (WebCore::IDBServer::MemoryIndexCursor::MemoryIndexCursor):
 374         (WebCore::IDBServer::MemoryIndexCursor::iterate):
 375
 376         * Modules/indexeddb/shared/IDBCursorInfo.cpp:
 377         (WebCore::IDBCursorInfo::duplicity):
 378         (WebCore::IDBCursorInfo::isDirectionNoDuplicate): Deleted.
 379         * Modules/indexeddb/shared/IDBCursorInfo.h:
 380
 381 2015-11-30  Jiewen Tan  <jiewen_tan@apple.com>
 382
 383         Null dereference loading Blink layout test http/tests/misc/detach-during-notifyDone.html
 384         https://bugs.webkit.org/show_bug.cgi?id=149309
 385         <rdar://problem/22748363>
 386
 387         Reviewed by Brent Fulgham.
 388
 389         A weird order of event execution introduced by the test case will kill the webpage in a
 390         subframe of the page while executing its |frame.loader().checkLoadCompleteForThisFrame()|.
 391         Therefore, any frames comes after the failing subframe will have no page. Check it before
 392         calling to those frames' |frame.loader().checkLoadCompleteForThisFrame()|, otherwise the
 393         assertion in |frame.loader().checkLoadCompleteForThisFrame()| will fail.
 394
 395         Test: http/tests/misc/detach-during-notifyDone.html
 396
 397         * loader/FrameLoader.cpp:
 398         (WebCore::FrameLoader::checkLoadComplete):
 399
 400 2015-11-30  Commit Queue  <commit-queue@webkit.org>
 401
 402         Unreviewed, rolling out r192819.
 403         https://bugs.webkit.org/show_bug.cgi?id=151681
 404
 405         This change broke existing layout tests on Windows (Requested
 406         by ryanhaddad on #webkit).
 407
 408         Reverted changeset:
 409
 410         "Unify font-variant-* with font-variant shorthand"
 411         https://bugs.webkit.org/show_bug.cgi?id=149773
 412         http://trac.webkit.org/changeset/192819
 413
 414 2015-11-30  Darin Adler  <darin@apple.com>
 415
 416         Use Optional instead of isNull out argument for nullable getters
 417         https://bugs.webkit.org/show_bug.cgi?id=151676
 418
 419         Reviewed by Anders Carlsson.
 420
 421         No behavior change, just cleaner code.
 422
 423         * Modules/geolocation/Coordinates.cpp:
 424         (WebCore::Coordinates::altitude): Return an Optional.
 425         (WebCore::Coordinates::altitudeAccuracy): Ditto.
 426         (WebCore::Coordinates::heading): Ditto.
 427         (WebCore::Coordinates::speed): Ditto.
 428         * Modules/geolocation/Coordinates.h: Ditto.
 429
 430         * Modules/indexeddb/IDBVersionChangeEvent.cpp:
 431         (WebCore::IDBVersionChangeEvent::create): Added. The code before was calling
 432         through to Event::create, which is clearly not what was wanted. Also removed
 433         unneeded explicit destructor.
 434         * Modules/indexeddb/IDBVersionChangeEvent.h: Changed return type of newVersion
 435         to Optional and updated for above change.
 436
 437         * Modules/indexeddb/client/IDBVersionChangeEventImpl.cpp:
 438         (WebCore::IDBClient::IDBVersionChangeEvent::newVersion): Changed to return
 439         an Optional.
 440         * Modules/indexeddb/client/IDBVersionChangeEventImpl.h: Removed unused
 441         default argument values; the event type one, at least, was clearly incorrect.
 442         Made more things private, got rid of unneeded destructor, marked class final
 443         instead of marking all functions final.
 444
 445         * Modules/indexeddb/legacy/LegacyVersionChangeEvent.cpp:
 446         (WebCore::LegacyVersionChangeEvent::newVersion): Same as above.
 447         * Modules/indexeddb/legacy/LegacyVersionChangeEvent.h: Ditto.
 448
 449         * Modules/mediastream/MediaTrackConstraints.cpp:
 450         (WebCore::MediaTrackConstraints::optional): Removed bogus bool value. If we
 451         come back to finish later we will have to implement optional return values
 452         for arrays in the JavaScript bindings generator, which should be straightforward.
 453         * Modules/mediastream/MediaTrackConstraints.h: Ditto.
 454
 455         * bindings/js/JSDOMBinding.h:
 456         (WebCore::toNullableJSNumber): Added. This function template is used for
 457         return values that are nullable numbers.
 458
 459         * bindings/scripts/CodeGeneratorGObject.pm:
 460         (GenerateFunction): Replaced some existing bogus code to handle nullables with
 461         new equally-bogus code that should be no worse and will compile.
 462
 463         * bindings/scripts/CodeGeneratorJS.pm:
 464         (GenerateImplementation): Removed old support for nullables.
 465         (NativeToJSValue): Added new support for nullable numbers.
 466
 467         * bindings/scripts/CodeGeneratorObjC.pm:
 468         (GenerateImplementation): Removed support for nullables. We almost certainly
 469         won't need it for Objective-C bindings.
 470
 471         * bindings/scripts/test/GObject/WebKitDOMTestObj.cpp: Updated.
 472         * bindings/scripts/test/JS/JSTestObj.cpp: Updated.
 473         * bindings/scripts/test/ObjC/DOMTestObj.mm: Updated.
 474
 475 2015-11-30  Wenson Hsieh  <wenson_hsieh@apple.com>
 476
 477         Split platform-independent logic in AVCaptureDeviceManager out into a new class
 478         https://bugs.webkit.org/show_bug.cgi?id=151388
 479         <rdar://problem/23593980>
 480
 481         Reviewed by Eric Carlson.
 482
 483         To prepare for creating a MockCaptureDeviceManager to be able to test
 484         MediaDevices.getUserMedia, we create a platform-independent capture device manager
 485         which all platforms should extend and add platform-specific logic to.
 486
 487         The methods CaptureDeviceManager::createMediaSourceForCaptureDeviceWithConstraints and
 488         CaptureDeviceManager::captureDeviceList should be overridden by each platform
 489         CaptureDeviceManager to respectively create a RealtimeMediaSource and return a list of
 490         capture devices. createMediaSourceForCaptureDeviceWithConstraints attempts to create
 491         a media source for a given device with some constraints; if the contraints cannot be
 492         satisfied, this returns null.
 493
 494         The refactored capture device manager also introduces the notion of a platform-
 495         independent capture session which may be extended by platform device managers for
 496         determining whether a given constraint name, value and media type is valid.
 497
 498         A platform-independent CaptureDeviceInfo now represents either the video or audio
 499         component of a capture device, but not both at once. This means a capture device that
 500         supports both video and audio will emit two separate capture devices.
 501
 502         No new tests, since there should be no behavior change.
 503
 504         * Modules/mediastream/CaptureDeviceInfo.h: Added.
 505         (WebCore::CaptureSessionInfo::~CaptureSessionInfo):
 506         (WebCore::CaptureSessionInfo::supportsVideoSize):
 507         (WebCore::CaptureSessionInfo::bestSessionPresetForVideoDimensions):
 508         * Modules/mediastream/CaptureDeviceManager.cpp: Added.
 509         (CaptureDeviceManager::~CaptureDeviceManager):
 510         (CaptureDeviceManager::getSourcesInfo):
 511         (CaptureDeviceManager::captureDeviceFromDeviceID):
 512         (CaptureDeviceManager::verifyConstraintsForMediaType):
 513         (CaptureDeviceManager::bestSourcesForTypeAndConstraints):
 514         (CaptureDeviceManager::sourceWithUID):
 515         (CaptureDeviceManager::bestDeviceForFacingMode):
 516         (facingModeFromString):
 517         (CaptureDeviceManager::sessionSupportsConstraint):
 518         (CaptureDeviceManager::isSupportedFrameRate):
 519         * Modules/mediastream/CaptureDeviceManager.h: Added.
 520         (WebCore::CaptureDeviceManager::refreshCaptureDeviceList):
 521         (WebCore::CaptureDeviceManager::defaultCaptureSession):
 522         * WebCore.xcodeproj/project.pbxproj:
 523         * platform/mediastream/RealtimeMediaSourceSupportedConstraints.cpp: Added.
 524         (WebCore::RealtimeMediaSourceSupportedConstraints::nameForConstraint):
 525         (WebCore::RealtimeMediaSourceSupportedConstraints::constraintFromName):
 526         (WebCore::RealtimeMediaSourceSupportedConstraints::supportsConstraint):
 527         * platform/mediastream/RealtimeMediaSourceSupportedConstraints.h:
 528         * platform/mediastream/mac/AVCaptureDeviceManager.h:
 529         * platform/mediastream/mac/AVCaptureDeviceManager.mm:
 530         (WebCore::AVCaptureSessionInfo::AVCaptureSessionInfo):
 531         (WebCore::AVCaptureSessionInfo::supportsVideoSize):
 532         (WebCore::AVCaptureSessionInfo::bestSessionPresetForVideoDimensions):
 533         (WebCore::AVCaptureDeviceManager::captureDeviceList):
 534         (WebCore::shouldConsiderDeviceInDeviceList):
 535         (WebCore::AVCaptureDeviceManager::refreshCaptureDeviceList):
 536         (WebCore::AVCaptureDeviceManager::AVCaptureDeviceManager):
 537         (WebCore::AVCaptureDeviceManager::bestSourcesForTypeAndConstraints):
 538         (WebCore::AVCaptureDeviceManager::sourceWithUID):
 539         (WebCore::AVCaptureDeviceManager::getSourcesInfo):
 540         (WebCore::AVCaptureDeviceManager::verifyConstraintsForMediaType):
 541         (WebCore::AVCaptureDeviceManager::defaultCaptureSession):
 542         (WebCore::AVCaptureDeviceManager::sessionSupportsConstraint):
 543         (WebCore::AVCaptureDeviceManager::createMediaSourceForCaptureDeviceWithConstraints):
 544         (WebCore::AVCaptureDeviceManager::deviceDisconnected):
 545         (WebCore::AVCaptureDeviceManager::isSupportedFrameRate):
 546         (WebCore::CaptureDevice:::m_enabled): Deleted.
 547         (WebCore::captureDeviceList): Deleted.
 548         (WebCore::captureDeviceFromDeviceID): Deleted.
 549         (WebCore::refreshCaptureDeviceList): Deleted.
 550         (WebCore::AVCaptureDeviceManager::bestSessionPresetForVideoSize): Deleted.
 551         (WebCore::AVCaptureDeviceManager::deviceSupportsFacingMode): Deleted.
 552         (WebCore::AVCaptureDeviceManager::bestDeviceForFacingMode): Deleted.
 553         (WebCore::AVCaptureDeviceManager::isValidConstraint): Deleted.
 554         (WebCore::AVCaptureDeviceManager::validConstraintNames): Deleted.
 555         (WebCore::AVCaptureDeviceManager::validFacingModes): Deleted.
 556         * platform/mediastream/mac/AVVideoCaptureSource.mm:
 557         (WebCore::AVVideoCaptureSource::applyConstraints):
 558         * platform/mediastream/mac/RealtimeMediaSourceCenterMac.cpp:
 559         (WebCore::RealtimeMediaSourceCenterMac::validateRequestConstraints):
 560         (WebCore::RealtimeMediaSourceCenterMac::createMediaStream):
 561
 562 2015-11-30  Brady Eidson  <beidson@apple.com>
 563
 564         Modern IDB: Set the correct source on the IDBRequest for cursor updates
 565         https://bugs.webkit.org/show_bug.cgi?id=151665
 566
 567         Reviewed by Andy Estes.
 568
 569         No new tests (At least one existing failing test now passes).
 570
 571         * Modules/indexeddb/client/IDBCursorImpl.cpp:
 572         (WebCore::IDBClient::IDBCursor::update):
 573
 574         * Modules/indexeddb/client/IDBObjectStoreImpl.cpp:
 575         (WebCore::IDBClient::IDBObjectStore::putForCursorUpdate):
 576         (WebCore::IDBClient::IDBObjectStore::putOrAdd):
 577         * Modules/indexeddb/client/IDBObjectStoreImpl.h:
 578
 579         * Modules/indexeddb/client/IDBRequestImpl.cpp:
 580         (WebCore::IDBClient::IDBRequest::IDBRequest):
 581         (WebCore::IDBClient::IDBRequest::setSource):
 582         * Modules/indexeddb/client/IDBRequestImpl.h:
 583
 584 2015-11-30  Tim Horton  <timothy_horton@apple.com>
 585
 586         Get rid of the legacy TextIndicatorWindow style
 587         https://bugs.webkit.org/show_bug.cgi?id=151674
 588
 589         Reviewed by Anders Carlsson.
 590
 591         * page/TextIndicator.h:
 592         * page/mac/TextIndicatorWindow.mm:
 593         (-[WebTextIndicatorView initWithFrame:textIndicator:margin:offset:]): Deleted.
 594
 595 2015-11-30  Brady Eidson  <beidson@apple.com>
 596
 597         Modern IDB: Correct handling of cursors finishing iteration.
 598         https://bugs.webkit.org/show_bug.cgi?id=151664
 599
 600         Reviewed by Andy Estes.
 601
 602         No new tests (At least one previously failing test now passes).
 603
 604         * Modules/indexeddb/client/IDBCursorImpl.cpp:
 605         (WebCore::IDBClient::IDBCursor::setGetResult):
 606
 607         * Modules/indexeddb/client/IDBObjectStoreImpl.cpp:
 608         (WebCore::IDBClient::IDBObjectStore::putOrAdd):
 609
 610 2015-11-30  Myles C. Maxfield  <mmaxfield@apple.com>
 611
 612         Unify font-variant-* with font-variant shorthand
 613         https://bugs.webkit.org/show_bug.cgi?id=149773
 614
 615         Reviewed by Darin Adler.
 616
 617         This patch makes font-variant a shorthand for the following properties:
 618         font-variant-ligatures
 619         font-variant-position
 620         font-variant-caps
 621         font-variant-numeric
 622         font-variant-alternates
 623         font-variant-east-asian
 624
 625         This is consistent with the CSS Fonts Level 3 spec.
 626
 627         This patch also migrates the "font" longhand to use the font-variant-caps
 628         property.
 629
 630         Test: fast/text/font-variant-shorthand.html
 631
 632         * css/CSSComputedStyleDeclaration.cpp:
 633         (WebCore::fontVariantEastAsianPropertyValue): Rename FontVariantEastAsian values.
 634         (WebCore::fontVariantFromStyle): We must consult with the longhand properties to determine
 635         font-variant computed style.
 636         (WebCore::ComputedStyleExtractor::propertyValue): Don't put any-old font-variant-caps inside
 637         the font shorthand.
 638         * css/CSSFontSelector.cpp:
 639         (WebCore::CSSFontSelector::addFontFaceRule): Guard against incorrect downcasts (due to inherit
 640         of the new shorthand property).
 641         * css/CSSParser.cpp: Parse font-variant as a shorthand. Also implement its "normal" and "none" values.
 642         (WebCore::CSSParser::parseValue):
 643         (WebCore::CSSParser::parseFont):
 644         (WebCore::CSSParser::parseSystemFont):
 645         (WebCore::CSSParser::parseFontVariantLigatures):
 646         (WebCore::CSSParser::parseFontVariantNumeric):
 647         (WebCore::CSSParser::parseFontVariantEastAsian):
 648         (WebCore::CSSParser::parseFontVariant):
 649         (WebCore::isValidKeywordPropertyAndValue): Deleted.
 650         (WebCore::isKeywordPropertyID): Deleted.
 651         * css/CSSParser.h:
 652         * css/CSSPropertyNames.in: Turn font-variant into a shorthand property.
 653         * css/FontVariantBuilder.h: Guard against incorrect downcasts. Also update for renamed
 654         FontVariantEastAsian type.
 655         (WebCore::applyValueFontVariantLigatures):
 656         (WebCore::applyValueFontVariantNumeric):
 657         (WebCore::applyValueFontVariantEastAsian):
 658         * css/StyleProperties.cpp: Update to use the more specific property.
 659         (WebCore::StyleProperties::appendFontLonghandValueIfExplicit):
 660         (WebCore::StyleProperties::fontValue):
 661         (WebCore::StyleProperties::asText):
 662         * css/StyleResolver.cpp: Ditto.
 663         (WebCore::StyleResolver::isValidCueStyleProperty):
 664         * editing/EditingStyle.cpp: Ditto.
 665         * editing/cocoa/HTMLConverter.mm: Ditto.
 666         (HTMLConverterCaches::propertyValueForNode):
 667         (HTMLConverter::computedAttributesForElement):
 668         * editing/ios/EditorIOS.mm: Ditto.
 669         (WebCore::Editor::removeUnchangeableStyles):
 670         * html/canvas/CanvasRenderingContext2D.cpp: Ditto.
 671         (WebCore::CanvasRenderingContext2D::font):
 672         (WebCore::CanvasRenderingContext2D::setFont):
 673         * platform/graphics/FontCache.h: Removing duplicate cache key value.
 674         (WebCore::FontDescriptionKey::makeFlagsKey):
 675         * platform/graphics/FontCascade.cpp: Migrate to the new font-variant-caps from the old member variable.
 676         (WebCore::FontCascade::glyphDataForCharacter):
 677         * platform/graphics/FontCascade.h: Ditto.
 678         (WebCore::FontCascade::isSmallCaps):
 679         * platform/graphics/FontDescription.cpp: Ditto.
 680         (WebCore::FontDescription::FontDescription):
 681         * platform/graphics/FontDescription.h: Ditto.
 682         (WebCore::FontCascadeDescription::equalForTextAutoSizing):
 683         (WebCore::FontDescription::smallCaps): Deleted.
 684         (WebCore::FontDescription::setSmallCaps): Deleted.
 685         (WebCore::FontDescription::setIsSmallCaps): Deleted.
 686         (WebCore::FontDescription::operator==): Deleted.
 687         * platform/graphics/cocoa/FontCacheCoreText.cpp: Rename FontVariantEastAsianWidth.
 688         (WebCore::computeFeatureSettingsFromVariants):
 689         * platform/text/TextFlags.h: Ditto.
 690         (WebCore::FontVariantSettings::operator==):
 691         * rendering/RenderText.cpp: Migrage to the new font-variant-caps from the old member variable.
 692         (WebCore::RenderText::widthFromCache):
 693
 694 2015-11-30  Brady Eidson  <beidson@apple.com>
 695
 696         Modern IDB: After versionchange transactions abort, fire onerror on the original IDBOpenDBRequest.
 697         https://bugs.webkit.org/show_bug.cgi?id=151648
 698
 699         Reviewed by Andy Estes.
 700
 701         No new tests. Covered by at least one existing failing test which now passes, and many
 702         other tests updated to fix their incorrect behavior.
 703
 704         * Modules/indexeddb/client/IDBDatabaseImpl.cpp:
 705         (WebCore::IDBClient::IDBDatabase::transaction):
 706         (WebCore::IDBClient::IDBDatabase::willAbortTransaction):
 707         (WebCore::IDBClient::IDBDatabase::didAbortTransaction):
 708
 709         * Modules/indexeddb/client/IDBOpenDBRequestImpl.cpp:
 710         (WebCore::IDBClient::IDBOpenDBRequest::fireErrorAfterVersionChangeAbort):
 711         * Modules/indexeddb/client/IDBOpenDBRequestImpl.h:
 712
 713         * Modules/indexeddb/client/IDBTransactionImpl.cpp:
 714         (WebCore::IDBClient::IDBTransaction::notifyDidAbort):
 715         (WebCore::IDBClient::IDBTransaction::didAbort):
 716         (WebCore::IDBClient::IDBTransaction::didCommit):
 717         * Modules/indexeddb/client/IDBTransactionImpl.h:
 718
 719 2015-11-30  Chris Dumez  <cdumez@apple.com>
 720
 721         location.origin is undefined in a web worker
 722         https://bugs.webkit.org/show_bug.cgi?id=151614
 723
 724         Reviewed by Darin Adler.
 725
 726         Expose location.origin to web workers, as per:
 727         https://html.spec.whatwg.org/multipage/workers.html#workerlocation
 728
 729         This behavior is consistent with the behavior of Firefox and Chrome.
 730
 731         Test: fast/workers/worker-location.html
 732
 733         * workers/WorkerLocation.cpp:
 734         (WebCore::WorkerLocation::origin):
 735         * workers/WorkerLocation.h:
 736         * workers/WorkerLocation.idl:
 737
 738 2015-11-30  Brady Eidson  <beidson@apple.com>
 739
 740         Modern IDB: Support updating cursor values when the object store uses inline keys.
 741         https://bugs.webkit.org/show_bug.cgi?id=151647
 742
 743         Reviewed by Andy Estes.
 744
 745         No new tests (At least two previously failing tests now pass and are unskipped).
 746
 747         * Modules/indexeddb/client/IDBCursorImpl.cpp:
 748         (WebCore::IDBClient::IDBCursor::update): Use putForCursorUpdate() instead of put()
 749
 750         * Modules/indexeddb/client/IDBObjectStoreImpl.cpp:
 751         (WebCore::IDBClient::IDBObjectStore::add):
 752         (WebCore::IDBClient::IDBObjectStore::put):
 753         (WebCore::IDBClient::IDBObjectStore::putForCursorUpdate): Use the flag to skip the inline-key check.
 754         (WebCore::IDBClient::IDBObjectStore::putOrAdd): Add a flag to skip the inline-key check.
 755         * Modules/indexeddb/client/IDBObjectStoreImpl.h:
 756
 757 2015-11-30  Brady Eidson  <beidson@apple.com>
 758
 759         Modern IDB: Resolve flaky GC-vs-wrapper issue with IDBOpenDBRequest.
 760         https://bugs.webkit.org/show_bug.cgi?id=151645
 761
 762         Reviewed by Andy Estes.
 763
 764         No new tests (Resolves flakiness with hundreds of existing IDB tests).
 765
 766         Do to improper management of the m_hasPendingActivity flag on IDBRequestImpl,
 767         the request wrapper for an IDBOpenDBRequest might be garbage collected in between the
 768         onUpgradeNeeded event and onSuccess event.
 769
 770         This manifested as flakiness in many tests, some more than others.
 771
 772         I tried to write a targeted 100% reproducible case manually forcing GC, but could not get
 773         the timing right.
 774
 775         * Modules/indexeddb/client/IDBOpenDBRequestImpl.cpp:
 776         (WebCore::IDBClient::IDBOpenDBRequest::fireSuccessAfterVersionChangeCommit):
 777         * Modules/indexeddb/client/IDBOpenDBRequestImpl.h:
 778
 779         * Modules/indexeddb/client/IDBRequestImpl.cpp:
 780         (WebCore::IDBClient::IDBRequest::dispatchEvent):
 781         (WebCore::IDBClient::IDBRequest::willIterateCursor):
 782         * Modules/indexeddb/client/IDBRequestImpl.h:
 783         (WebCore::IDBClient::IDBRequest::isOpenDBRequest):
 784
 785 2015-11-30  Per Arne Vollan  <peavo@outlook.com>
 786
 787         [WinCairo][MediaFoundation] Implement seek.
 788         https://bugs.webkit.org/show_bug.cgi?id=151609
 789
 790         Reviewed by Alex Christensen.
 791
 792         The methods maxTimeSeekable() and buffered() needs to be implemented.
 793
 794         * platform/graphics/win/MediaPlayerPrivateMediaFoundation.cpp:
 795         (WebCore::MediaPlayerPrivateMediaFoundation::seekDouble):
 796         (WebCore::MediaPlayerPrivateMediaFoundation::durationDouble):
 797         (WebCore::MediaPlayerPrivateMediaFoundation::readyState):
 798         (WebCore::MediaPlayerPrivateMediaFoundation::maxTimeSeekable):
 799         (WebCore::MediaPlayerPrivateMediaFoundation::buffered):
 800         (WebCore::MediaPlayerPrivateMediaFoundation::didLoadingProgress):
 801         (WebCore::MediaPlayerPrivateMediaFoundation::CustomVideoPresenter::currentTime):
 802         (WebCore::MediaPlayerPrivateMediaFoundation::CustomVideoPresenter::isActive):
 803         * platform/graphics/win/MediaPlayerPrivateMediaFoundation.h:
 804         (WebCore::MediaPlayerPrivateMediaFoundation::CustomVideoPresenter::maxTimeLoaded):
 805
 806 2015-11-30  Brady Eidson  <beidson@apple.com>
 807
 808         Modern IDB: openCursor() fix resulting in at least 4 more passing tests.
 809         https://bugs.webkit.org/show_bug.cgi?id=151630
 810
 811         Reviewed by Andy Estes.
 812
 813         No new tests (At least 4 failing tests now pass, and 9 other incorrect tests updated).
 814
 815         * Modules/indexeddb/IDBGetResult.cpp:
 816         (WebCore::IDBGetResult::isolatedCopy):
 817         * Modules/indexeddb/IDBGetResult.h:
 818         (WebCore::IDBGetResult::IDBGetResult):
 819         (WebCore::IDBGetResult::isDefined):
 820
 821         * Modules/indexeddb/client/IDBRequestImpl.cpp:
 822         (WebCore::IDBClient::IDBRequest::didOpenOrIterateCursor): If the IDBGetResult is undefined,
 823           do not expose the cursor as the result property of the IDBRequest.
 824
 825 2015-11-30  Brady Eidson  <beidson@apple.com>
 826
 827         Modern IDB: Support keyPath injection into object store records.
 828         https://bugs.webkit.org/show_bug.cgi?id=151640
 829
 830         Reviewed by Andy Estes.
 831
 832         No new tests (At least one existing failure now passes and is unskipped,
 833         while many other existing failures are now closer to passing).
 834
 835         * Modules/indexeddb/server/MemoryObjectStore.cpp:
 836         (WebCore::IDBServer::MemoryObjectStore::updateIndexesForPutRecord): Use the new UniqueIDBDatabase VM/ExecState.
 837         (WebCore::IDBServer::MemoryObjectStore::populateIndexWithExistingRecords): Ditto
 838         (WebCore::IDBServer::indexVM): Deleted.
 839         (WebCore::IDBServer::indexGlobalExec): Deleted.
 840
 841         * Modules/indexeddb/server/UniqueIDBDatabase.cpp:
 842         (WebCore::IDBServer::UniqueIDBDatabase::databaseThreadVM):
 843         (WebCore::IDBServer::UniqueIDBDatabase::databaseThreadExecState):
 844         (WebCore::IDBServer::UniqueIDBDatabase::performPutOrAdd): If appropriate, inject the key that will be used into
 845           the value before storing the record.
 846         * Modules/indexeddb/server/UniqueIDBDatabase.h:
 847
 848         Add modern JSValue/ExecState& version of some binding utilities, for use today and in preparation of getting
 849         rid of the DOMRequestState and Deprecated::ScriptValue versions later:
 850         * bindings/js/IDBBindingUtilities.cpp:
 851         (WebCore::idbKeyToJSValue):
 852         (WebCore::injectIDBKeyIntoScriptValue):
 853         (WebCore::deserializeIDBValueData):
 854         (WebCore::deserializeIDBValueDataToJSValue):
 855         * bindings/js/IDBBindingUtilities.h:
 856
 857 2015-11-25  Andy Estes  <aestes@apple.com>
 858
 859         [Content Filtering] Avoid creating a ContentFilter when loading the empty document
 860         https://bugs.webkit.org/show_bug.cgi?id=151615
 861
 862         Reviewed by Daniel Bates.
 863
 864         It's expensive to create the first ContentFilter since two frameworks must be soft-linked. There's no reason to
 865         pay this cost if we're just loading the empty document.
 866
 867         No new tests. It's not possible to write a test that would fail without this change since ContentFilter is not
 868         notified of empty document loads.
 869
 870         * loader/DocumentLoader.cpp:
 871         (WebCore::DocumentLoader::startLoadingMainResource): Don't initialize m_contentFilter until we know we aren't
 872         loading the empty document.
 873         (WebCore::DocumentLoader::DocumentLoader):
 874
 875 2015-11-29  Antoine Quint  <graouts@apple.com>
 876
 877         Browser does not fall back to SVG attribute value when CSS style value is invalid or not supported
 878         https://bugs.webkit.org/show_bug.cgi?id=147932
 879
 880         Reviewed by Dean Jackson.
 881
 882         Instead of returning an SVGPaint object of type SVG_PAINTTYPE_UNKNOWN when we encounter an SVG paint
 883         value that cannot be parsed, we now return `nullptr` which will cause that value to be ignored and
 884         let another paint value in the cascade be used instead. This is the same approach used for SVGColor.
 885         Since we're removing the only call site for `SVGPaint::createUnknown()`, we remove that function entirely.
 886
 887         Tests: svg/css/invalid-color-cascade.svg
 888                svg/css/invalid-paint-cascade.svg
 889
 890         * css/SVGCSSParser.cpp:
 891         (WebCore::CSSParser::parseSVGPaint):
 892         * svg/SVGPaint.h:
 893         (WebCore::SVGPaint::createUnknown): Deleted.
 894
 895 2015-11-29  Simon Fraser  <simon.fraser@apple.com>
 896
 897         Use SVGTransform::SVGTransformType instead of an unsigned short
 898         https://bugs.webkit.org/show_bug.cgi?id=151637
 899
 900         Reviewed by Brady Eidson.
 901
 902         Make 'type' more strongly typed.
 903
 904         * svg/SVGTransformable.cpp:
 905         (WebCore::SVGTransformable::parseTransformValue):
 906         (WebCore::parseAndSkipType):
 907         (WebCore::SVGTransformable::parseTransformType):
 908         (WebCore::SVGTransformable::parseTransformAttribute):
 909         * svg/SVGTransformable.h:
 910
 911 2015-11-27  Brady Eidson  <beidson@apple.com>
 912
 913         Modern IDB: Class-ify IDBGetResult making it impossible to get the data members wrong.
 914         https://bugs.webkit.org/show_bug.cgi?id=151627
 915
 916         Reviewed by Alexey Proskuryakov.
 917
 918         No new tests (No change in behavior).
 919
 920         * CMakeLists.txt:
 921         * WebCore.xcodeproj/project.pbxproj:
 922
 923         * Modules/indexeddb/IDBGetResult.cpp: Added.
 924         (WebCore::IDBGetResult::dataFromBuffer):
 925         (WebCore::IDBGetResult::isolatedCopy):
 926         * Modules/indexeddb/IDBGetResult.h:
 927         (WebCore::IDBGetResult::IDBGetResult):
 928         (WebCore::IDBGetResult::valueBuffer):
 929         (WebCore::IDBGetResult::keyData):
 930         (WebCore::IDBGetResult::primaryKeyData):
 931         (WebCore::IDBGetResult::keyPath):
 932         (WebCore::IDBGetResult::setValueBuffer):
 933         (WebCore::IDBGetResult::setKeyData):
 934         (WebCore::IDBGetResult::setPrimaryKeyData):
 935         (WebCore::IDBGetResult::setKeyPath):
 936         (WebCore::IDBGetResult::dataFromBuffer): Deleted.
 937         (WebCore::IDBGetResult::isolatedCopy): Deleted.
 938
 939         * Modules/indexeddb/client/IDBCursorImpl.cpp:
 940         (WebCore::IDBClient::IDBCursor::setGetResult):
 941         * Modules/indexeddb/client/IDBCursorImpl.h:
 942
 943         * Modules/indexeddb/client/IDBTransactionImpl.cpp:
 944         (WebCore::IDBClient::IDBTransaction::didGetRecordOnServer):
 945
 946         * Modules/indexeddb/legacy/IDBTransactionBackendOperations.cpp:
 947         (WebCore::GetOperation::perform):
 948
 949         * Modules/indexeddb/server/IDBBackingStore.h:
 950
 951         * Modules/indexeddb/server/MemoryCursor.h:
 952
 953         * Modules/indexeddb/server/MemoryIndexCursor.cpp:
 954         (WebCore::IDBServer::MemoryIndexCursor::currentData):
 955
 956         * Modules/indexeddb/server/MemoryObjectStoreCursor.cpp:
 957         (WebCore::IDBServer::MemoryObjectStoreCursor::currentData):
 958
 959         * platform/CrossThreadCopier.h:
 960
 961 2015-11-27  Carlos Garcia Campos  <cgarcia@igalia.com>
 962
 963         [GTK] Do not use the WebCore garbage collector timer
 964         https://bugs.webkit.org/show_bug.cgi?id=151623
 965
 966         Reviewed by Martin Robinson.
 967
 968         Now that garbage collector timers have been implemented in
 969         JavaScriptCore for glib, we don't need to use another Timer in WebCore.
 970
 971         * bindings/js/GCController.cpp:
 972         (WebCore::GCController::garbageCollectSoon):
 973         (WebCore::GCController::garbageCollectNowIfNotDoneRecently):
 974
 975 2015-11-18  Andy Estes  <aestes@apple.com>
 976
 977         [Content Filtering] Crash in DocumentLoader::notifyFinished() when allowing a media document to load
 978         https://bugs.webkit.org/show_bug.cgi?id=151433
 979         rdar://problem/23506594
 980
 981         Reviewed by Alexey Proskuryakov.
 982
 983         When the main resource of a media document commits, WebKit cancels its load since the plug-in or media engine
 984         will do its own loading. If content filtering is enabled, and the filter waits allow the load until the entire
 985         resource is downloaded, then ContentFilter will attempt to call DocumentLoader::notifyFinished() immediately
 986         after delivering the buffered resource data to DocumentLoader. However, delivering the data will have nulled out
 987         DocumentLoader's m_mainResource when the load was cancelled, leading to a crash in notifyFinished().
 988
 989         To resolve this, add a new Stopped state to ContentFilter. Set this state if DocumentLoader clears its main
 990         resource or detaches from its frame. If ContentFilter is in the Stopped state after calling
 991         DocumentLoader::dataReceived(), do not proceed to call DocumentLoader::notifyFinished().
 992
 993         Test: contentfiltering/allow-media-document.html
 994
 995         * loader/ContentFilter.cpp:
 996         (WebCore::ContentFilter::stopFilteringMainResource): Set m_state to Stopped. If m_mainResource is non-null,
 997         removed ContentFilter as a client and set m_mainResource to null.
 998         (WebCore::ContentFilter::notifyFinished): Stopped calling DocumentLoader::notifyFinished() if m_state is Stopped
 999         after calling DocumentLoader::dataReceived().
1000         * loader/ContentFilter.h:
1001         * loader/DocumentLoader.cpp:
1002         (WebCore::DocumentLoader::detachFromFrame): Called ContentFilter::stopFilteringMainResource() instead of setting
1003         m_contentFilter to null.
1004         (WebCore::DocumentLoader::clearMainResource): Ditto.
1005
1006 2015-11-25  Pranjal Jumde  <pjumde@apple.com>
1007
1008         Checks for buffer-overflows when reading characters from textRun
1009         https://bugs.webkit.org/show_bug.cgi?id=151055
1010         <rdar://problem/23251789>
1011
1012         Reviewed by Myles C. Maxfield.
1013
1014         Prevents an off by one error when adding the last font data to the GlyphBuffer.
1015
1016         * Source/WebCore/platform/graphics/WidthIterator.cpp:
1017         * Source/WebCore/platform/graphics/FontCascade.cpp:
1018
1019 2015-11-24  Xabier Rodriguez Calvar  <calvaris@igalia.com>
1020
1021         [Streams API] Implement pipeTo method in readable Stream
1022         https://bugs.webkit.org/show_bug.cgi?id=151588
1023
1024         Reviewed by Darin Adler.
1025
1026         Implemented pipeTo method according to the reference implementation in the spec as the spec is not written
1027         yet. It can be found at https://github.com/whatwg/streams/blob/632b26a05f3106650b1ec91239ad5b012e6c64af/reference-implementation/lib/readable-stream.js#L75.
1028
1029         Tests: streams/pipe-to.html
1030                streams/reference-implementation/brand-checks.html
1031                streams/reference-implementation/pipe-through.html
1032                streams/reference-implementation/pipe-to.html
1033                streams/reference-implementation/pipe-to-options.html
1034                streams/reference-implementation/readable-stream-templated
1035
1036         * Modules/streams/ReadableStream.js:
1037         (doPipe): Internal function of pipeTo.
1038         (closeDest): Internal function of pipeTo.
1039         (abortDest): Internal function of pipeTo.
1040         (pipeTo): Implemented as per spec with some other internal functions as helpers.
1041
1042 2015-11-24  Antti Koivisto  <antti@apple.com>
1043
1044         REGRESSION (r190983): Non-element, non-text nodes should not be distributed to slots
1045         https://bugs.webkit.org/show_bug.cgi?id=151566
1046         rdar://problem/23430177
1047
1048         Reviewed by Zalan Bujtas.
1049
1050         We don't invalidate slot assignments except for text or element children. Fix by not
1051         not assigning other nodes to slots as it is not useful.
1052
1053         Test: fast/html/details-comment-crash.html
1054
1055         * dom/SlotAssignment.cpp:
1056         (WebCore::slotNameFromSlotAttribute):
1057         (WebCore::SlotAssignment::findAssignedSlot):
1058         (WebCore::SlotAssignment::assignSlots):
1059
1060 2015-11-23  David Kilzer  <ddkilzer@apple.com>
1061
1062         Hardening against CSSSelector double frees
1063         <http://webkit.org/b/56124>
1064         <rdar://problem/9119036>
1065
1066         Reviewed by Antti Koivisto.
1067
1068         Add some security assertions to catch this issue if it ever
1069         happens in Debug builds, and make changes in
1070         CSSSelector::~CSSSelector() and
1071         CSSSelectorList::deleteSelectors() to prevent obvious issues if
1072         they're ever called twice in Release builds.
1073
1074         No new tests because we don't know how to reproduce this.
1075
1076         * css/CSSSelector.cpp:
1077         (WebCore::CSSSelector::CSSSelector): Initialize
1078         m_destructorHasBeenCalled.
1079         * css/CSSSelector.h:
1080         (WebCore::CSSSelector::m_destructorHasBeenCalled): Add bitfield.
1081         (WebCore::CSSSelector::CSSSelector): Initialize
1082         m_destructorHasBeenCalled.
1083         (WebCore::CSSSelector::~CSSSelector): Add security assertion
1084         that this is never called twice.  Clear out any fields that
1085         would have caused us to dereference an object twice.
1086
1087         * css/CSSSelectorList.cpp:
1088         (WebCore::CSSSelectorList::deleteSelectors): Clear
1089         m_selectorArray when freeing the memory to which it was
1090         pointing.  This prevents re-entrancy issues or calling this
1091         method twice on the same thread.  Also restructure the for()
1092         loop to prevent calling CSSSelector::isLastInSelectorList()
1093         after CSSSelector::~CSSSelector() has been called (via CRBug
1094         241892).
1095
1096 2015-11-23  Brian Burg  <bburg@apple.com>
1097
1098         Web Inspector: inspector settings should not be shared between different inspection levels
1099         https://bugs.webkit.org/show_bug.cgi?id=151151
1100
1101         Reviewed by Timothy Hatcher.
1102
1103         InspectorController and InspectorFrontendClient should know how to compute their
1104         inspection levels, so that this level can be exposed to the Inspector frontend for
1105         preference disambiguation by inspection level.
1106
1107          - A Page with normal web content has an inspection level of 0.
1108          - The first Web Inspector has an inspection level of 1.
1109          - The second Web Inspector that inspects the Inspector has an inspection level of 2.
1110          - And so forth...
1111
1112         For local frontend clients (WK1 and the ProtocolTestStub), the inspection level of
1113         a Page computed as follows:
1114
1115         - If the page's inspector controller has no frontend client, then the Page
1116         is normal web content (level 0).
1117         - If the page's inspector controller has a frontend client, ask for its inspection level.
1118         A frontend client looks at the inspection level reported by inspected page's inspector
1119         controller and increments by one to account for the crossed inspection boundary.
1120
1121         No new tests, this is blocked by test infrastructure fixes (https://webkit.org/b/151573).
1122
1123         * inspector/InspectorController.cpp:
1124         (WebCore::InspectorController::inspectionLevel): Added.
1125         (WebCore::InspectorController::hasInspectorFrontendClient): Deleted.
1126         * inspector/InspectorController.h:
1127         * inspector/InspectorFrontendClient.h:
1128         * inspector/InspectorFrontendClientLocal.cpp:
1129         (WebCore::InspectorFrontendClientLocal::canAttachWindow): Check inspectionLevel().
1130         (WebCore::InspectorFrontendClientLocal::inspectionLevel): Added.
1131         * inspector/InspectorFrontendClientLocal.h:
1132         * inspector/InspectorFrontendHost.cpp:
1133         (WebCore::InspectorFrontendHost::inspectionLevel): Added.
1134         * inspector/InspectorFrontendHost.h:
1135         * inspector/InspectorFrontendHost.idl:
1136         * page/ContextMenuController.cpp:
1137         (WebCore::ContextMenuController::populate): Check inspectionLevel().
1138
1139 2015-11-20  Brian Burg  <bburg@apple.com>
1140
1141         Web Inspector: RemoteInspector should track targets and connections for remote automation
1142         https://bugs.webkit.org/show_bug.cgi?id=151042
1143
1144         Reviewed by Joseph Pecoraro.
1145
1146         Use the new RemoteControllableTarget API method names.
1147
1148         No new tests, no behavior change.
1149
1150         * page/PageDebuggable.cpp:
1151         (WebCore::PageDebuggable::connect):
1152         (WebCore::PageDebuggable::dispatchMessageFromRemote):
1153         (WebCore::PageDebuggable::dispatchMessageFromRemoteFrontend): Deleted.
1154         * page/PageDebuggable.h:
1155
1156 2015-11-23  Brady Eidson  <beidson@apple.com>
1157
1158         Modern IDB: Unskip all indexeddb/mozilla tests that pass.
1159         https://bugs.webkit.org/show_bug.cgi?id=151568
1160
1161         Reviewed by Alex Christensen.
1162
1163         No new tests (Lots of skipped existing tests now pass).
1164
1165         * Modules/indexeddb/shared/IDBError.h:
1166
1167 2015-11-23  Nikos Andronikos <nikos.andronikos-webkit@cisra.canon.com.au>
1168
1169         Add WebCore namespace for generated bindings for supplemental method calls
1170         https://bugs.webkit.org/show_bug.cgi?id=151431
1171
1172         Reviewed by Alex Christensen.
1173
1174         Address build failures for generated supplemental method calls which are in the WebCore namespace.
1175         The bindings are outside the WebCore namespace - so add WebCore namespace specifier where required.
1176
1177         * bindings/scripts/CodeGeneratorJS.pm:
1178         (GenerateHeader):
1179         (GenerateImplementation):
1180         (GenerateParametersCheck):
1181         * bindings/scripts/CodeGeneratorObjC.pm:
1182         (GenerateImplementation):
1183         * bindings/scripts/test/JS/JSTestInterface.cpp:
1184         (WebCore::jsTestInterfaceConstructorSupplementalStaticReadOnlyAttr):
1185         (WebCore::jsTestInterfaceConstructorSupplementalStaticAttr):
1186         (WebCore::jsTestInterfaceSupplementalStr1):
1187         (WebCore::jsTestInterfaceSupplementalStr2):
1188         (WebCore::jsTestInterfaceSupplementalNode):
1189         (WebCore::setJSTestInterfaceConstructorSupplementalStaticAttr):
1190         (WebCore::setJSTestInterfaceSupplementalStr2):
1191         (WebCore::setJSTestInterfaceSupplementalNode):
1192         (WebCore::jsTestInterfacePrototypeFunctionSupplementalMethod1):
1193         (WebCore::jsTestInterfacePrototypeFunctionSupplementalMethod2):
1194         (WebCore::jsTestInterfaceConstructorFunctionSupplementalMethod4):
1195         * bindings/scripts/test/ObjC/DOMTestInterface.mm:
1196         (-[DOMTestInterface supplementalStr1]):
1197         (-[DOMTestInterface supplementalStr2]):
1198         (-[DOMTestInterface setSupplementalStr2:]):
1199         (-[DOMTestInterface supplementalStr3]):
1200         (-[DOMTestInterface setSupplementalStr3:]):
1201         (-[DOMTestInterface supplementalNode]):
1202         (-[DOMTestInterface setSupplementalNode:]):
1203         (-[DOMTestInterface builtinAttribute]):
1204         (-[DOMTestInterface setBuiltinAttribute:]):
1205         (-[DOMTestInterface supplementalMethod1]):
1206         (-[DOMTestInterface supplementalMethod2:objArg:]):
1207         (-[DOMTestInterface supplementalMethod3]):
1208         (-[DOMTestInterface supplementalMethod4]):
1209         (-[DOMTestInterface builtinFunction]):
1210
1211 2015-11-23  Brady Eidson  <beidson@apple.com>
1212
1213         Modern IDB: When a transaction is aborted, call onerror handlers for all in-progress requests.
1214         https://bugs.webkit.org/show_bug.cgi?id=151550
1215
1216         Reviewed by Alex Christensen.
1217
1218         Test: storage/indexeddb/modern/abort-requests-cancelled.html
1219               storage/indexeddb/modern/idbtransaction-objectstore-failures.html (with changes)
1220               storage/indexeddb/modern/index-5.html (with changes)
1221               Various (currently skipped) legacy IDB tests.
1222
1223         * Modules/indexeddb/client/IDBDatabaseImpl.cpp:
1224         (WebCore::IDBClient::IDBDatabase::transaction):
1225         (WebCore::IDBClient::IDBDatabase::startVersionChangeTransaction):
1226         (WebCore::IDBClient::IDBDatabase::didStartTransaction):
1227         (WebCore::IDBClient::IDBDatabase::willCommitTransaction):
1228         (WebCore::IDBClient::IDBDatabase::didCommitTransaction):
1229         (WebCore::IDBClient::IDBDatabase::willAbortTransaction):
1230         (WebCore::IDBClient::IDBDatabase::didAbortTransaction):
1231         (WebCore::IDBClient::IDBDatabase::didCommitOrAbortTransaction):
1232
1233         * Modules/indexeddb/client/IDBTransactionImpl.cpp:
1234         (WebCore::IDBClient::IDBTransaction::abort):
1235         (WebCore::IDBClient::IDBTransaction::abortOnServerAndCancelRequests):
1236         (WebCore::IDBClient::IDBTransaction::didCreateObjectStoreOnServer):
1237         (WebCore::IDBClient::IDBTransaction::didCreateIndexOnServer):
1238         (WebCore::IDBClient::IDBTransaction::didGetRecordOnServer):
1239         (WebCore::IDBClient::IDBTransaction::didDeleteObjectStoreOnServer):
1240         (WebCore::IDBClient::IDBTransaction::didDeleteIndexOnServer):
1241         (WebCore::IDBClient::IDBTransaction::immediateAbort): Deleted.
1242         (WebCore::IDBClient::IDBTransaction::abortOnServer): Deleted.
1243         * Modules/indexeddb/client/IDBTransactionImpl.h:
1244
1245         * Modules/indexeddb/server/MemoryIDBBackingStore.cpp:
1246         (WebCore::IDBServer::MemoryIDBBackingStore::abortTransaction):
1247         (WebCore::IDBServer::MemoryIDBBackingStore::commitTransaction):
1248
1249         * Modules/indexeddb/shared/IDBError.cpp:
1250         (WebCore::idbErrorName):
1251         (WebCore::idbErrorDescription):
1252         * Modules/indexeddb/shared/IDBError.h:
1253
1254         * Modules/indexeddb/shared/IDBResourceIdentifier.cpp:
1255         (WebCore::IDBResourceIdentifier::loggingString):
1256         * Modules/indexeddb/shared/IDBResourceIdentifier.h:
1257
1258 2015-11-23  Brady Eidson  <beidson@apple.com>
1259
1260         Modern IDB: Unskip storage/indexeddb/mozilla/global-data.html.
1261         https://bugs.webkit.org/show_bug.cgi?id=151557
1262
1263         Reviewed by Alex Christensen.
1264
1265         No new tests (Unskipping existing test storage/indexeddb/mozilla/global-data.html).
1266
1267         - Reworking some invalid ASSERTS
1268         - Actually opening pending open-database-requests after a version change transaction completes
1269         - Allow starting new transactions when the version change transaction has *started* finishing,
1270           but before it finishes finishing.
1271
1272         * Modules/indexeddb/client/IDBDatabaseImpl.cpp:
1273         (WebCore::IDBClient::IDBDatabase::transaction):
1274
1275         * Modules/indexeddb/client/IDBTransactionImpl.h:
1276
1277         * Modules/indexeddb/server/MemoryObjectStore.cpp:
1278         (WebCore::IDBServer::MemoryObjectStore::~MemoryObjectStore):
1279
1280         * Modules/indexeddb/server/MemoryObjectStoreCursor.cpp:
1281         (WebCore::IDBServer::MemoryObjectStoreCursor::keyAdded): Deleted.
1282
1283         * Modules/indexeddb/server/UniqueIDBDatabase.cpp:
1284         (WebCore::IDBServer::UniqueIDBDatabase::UniqueIDBDatabase):
1285         (WebCore::IDBServer::UniqueIDBDatabase::handleOpenDatabaseOperations):
1286         (WebCore::IDBServer::UniqueIDBDatabase::commitTransaction):
1287         * Modules/indexeddb/server/UniqueIDBDatabase.h:
1288
1289 2015-11-23  Youenn Fablet  <youenn.fablet@crf.canon.fr>
1290
1291         Remove DOMPromiseWithCallback
1292         https://bugs.webkit.org/show_bug.cgi?id=151565
1293
1294         Reviewed by Eric Carlson.
1295
1296         DOMPromiseWithCallback is no longer used as WebCore JS built-ins are used to the cases where DOMPromiseWithCallback made sense.
1297
1298         No change in behavior.
1299
1300         * Modules/mediastream/MediaDevices.h: Moved from DOMPromiseWithCallback to DOMPromise since the API is now called by JS built-ins.
1301         * bindings/js/JSDOMPromise.h:
1302         (WebCore::DOMPromiseWithCallback::DOMPromiseWithCallback): Deleted.
1303         (WebCore::DOMPromiseIteratorWithCallback::DOMPromiseIteratorWithCallback): Deleted.
1304         (WebCore::Error>::resolve): Deleted.
1305         (WebCore::Error>::reject): Deleted.
1306         (WebCore::Error>::resolveEnd): Deleted.
1307
1308 2015-11-23  Brian Burg  <bburg@apple.com>
1309
1310         Web Inspector: when inspecting the inspector, add the inspection level to the title bar
1311         https://bugs.webkit.org/show_bug.cgi?id=151555
1312
1313         Reviewed by Timothy Hatcher.
1314
1315         * English.lproj/Localizable.strings: add new localized string for alternate inspector title.
1316
1317 2015-11-23  Zan Dobersek  <zdobersek@igalia.com>
1318
1319         [GStreamer] No need to assert the pipeline's bus presence in MediaPlayerPrivateGStreamerBase dtor
1320         https://bugs.webkit.org/show_bug.cgi?id=151558
1321
1322         Reviewed by Carlos Garcia Campos.
1323
1324         In the MediaPlayerPrivateGStreamerBase destructor, there's some leftover code
1325         that acquires a reference to the GStreamer pipeline's bus object and then
1326         just asserts that it's not null. Not very useful, so remove it.
1327
1328         Also use nullptr to null out the m_player member.
1329
1330         * platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
1331         (WebCore::MediaPlayerPrivateGStreamerBase::~MediaPlayerPrivateGStreamerBase):
1332
1333 2015-11-17  Sergio Villar Senin  <svillar@igalia.com>
1334
1335         ASSERTION FAILED: freeSpace >= 0 in WebCore::RenderGrid::computeTrackSizesForDirection
1336         https://bugs.webkit.org/show_bug.cgi?id=151254
1337
1338         Reviewed by Darin Adler.
1339
1340         When in quirks mode, abnormally huge margins could lead to
1341         negative computations of available logical sizes. We could add
1342         an !document.inQuirksMode() check in the assertion but since
1343         negative freeSpace values are valid (i.e. the condition is not
1344         really part of the contract and everything will work as
1345         expected) I've decided to simply remove it.
1346
1347         Test: fast/css-grid-layout/grid-quirks-mode-huge-margin-crash.html
1348
1349         * rendering/RenderGrid.cpp:
1350         (WebCore::RenderGrid::computeTrackSizesForDirection): Deleted.
1351
1352 2015-11-22  Myles C. Maxfield  <mmaxfield@apple.com>
1353
1354         Font selection should not consult font-variant property
1355         https://bugs.webkit.org/show_bug.cgi?id=151537
1356
1357         Reviewed by Simon Fraser.
1358
1359         In section 4.7 of the CSS Fonts Level 3 spec, it says "[The font-variant and
1360         font-feature-settings] do not affect font selection."
1361
1362         All the other browsers (Chrome, Firefox, and Edge) all obey the spec here. We
1363         are the only one who misbehaves. This patch aligns our behavior with the other
1364         browsers.
1365
1366         Test: fast/text/font-selection-font-variant.html
1367
1368         * css/CSSFontSelector.cpp:
1369         (WebCore::computeTraitsMask): Deleted.
1370         (WebCore::compareFontFaces): Deleted.
1371         (WebCore::CSSFontSelector::getFontFace): Deleted.
1372         * css/CSSParser.cpp:
1373         (WebCore::isValidKeywordPropertyAndValue):
1374         (WebCore::isKeywordPropertyID):
1375         (WebCore::CSSParser::createFontFaceRule):
1376         (WebCore::CSSParser::CSSParser): Deleted.
1377         (WebCore::CSSParser::parseValue): Deleted.
1378         (WebCore::CSSParser::parseDeclaration): Deleted.
1379         (WebCore::CSSParser::clearProperties): Deleted.
1380         (WebCore::CSSParser::parseFontVariant): Deleted.
1381         (WebCore::CSSParser::createStyleRule): Deleted.
1382         (WebCore::CSSParser::deleteFontFaceOnlyValues): Deleted.
1383         * css/CSSParser.h:
1384         * platform/graphics/FontDescription.cpp:
1385         (WebCore::FontDescription::traitsMask): Deleted.
1386         * platform/graphics/win/FontCacheWin.cpp:
1387         (WebCore::traitsInFamilyEnumProc):
1388         * platform/graphics/cocoa/FontCacheCoreText.cpp:
1389         (WebCore::toTraitsMask):
1390         * platform/text/TextFlags.h:
1391
1392 2015-11-22  Carlos Garcia Campos  <cgarcia@igalia.com>
1393
1394         Unreviewed, rolling out r192727.
1395
1396         It made the selections transparent again and broke
1397         /webkit2/WebKitWebView/snapshot
1398
1399         Reverted changeset:
1400
1401         "[GTK] RenderThemeGtk::platformActiveSelectionBackgroundColor,
1402         et. al. should not clobber state of cached GtkStyleContexts"
1403         https://bugs.webkit.org/show_bug.cgi?id=151533
1404         http://trac.webkit.org/changeset/192727
1405
1406 2015-11-22  Michael Catanzaro  <mcatanzaro@igalia.com>
1407
1408         [GTK] RenderThemeGtk::platformActiveSelectionBackgroundColor, et. al. should not clobber state of cached GtkStyleContexts
1409         https://bugs.webkit.org/show_bug.cgi?id=151533
1410
1411         Reviewed by Carlos Garcia Campos.
1412
1413         platformActiveSelectionBackgroundColor(), platformInactiveSelectionBackgroundColor(), etc.
1414         are const functions intended only to return a color used for painting, but since r174929
1415         they also change the state of the cached style contexts we use for GTK_TYPE_ENTRY and
1416         GTK_TYPE_TREE_VIEW. That's wrong; those style contexts should not have any state set. This
1417         could cause theme colors returned by those GtkStyleContexts to change unexpectedly,
1418         depending on whether the state is explicitly set before each use, or whether the theme
1419         actually uses the states.
1420
1421         This didn't cause any regression only because every place using these style contexts
1422         explicitly sets the state of the style contexts before use. In fact, the GtkTreeView style
1423         context is not used anywhere else, and the GtkEntry style context is only used in
1424         paintTextField, which does set the state before use (and then reverts it using
1425         save/restore), so this cannot have broken anything in practice. But it's a landmine waiting
1426         for the next programmer to trip it.
1427
1428         Fix this with a gtk_style_context_save()/gtk_style_context_restore() pair.
1429
1430         * rendering/RenderThemeGtk.cpp:
1431         (WebCore::styleColor):
1432
1433 2015-11-21  Myles C. Maxfield  <mmaxfield@apple.com>
1434
1435         Tiny cleanup in ComplexTextController::collectComplexTextRuns()
1436         https://bugs.webkit.org/show_bug.cgi?id=151534
1437
1438         Reviewed by Zalan Bujtas.
1439
1440         The isMissingGlyph boolean is completely unnecessary. Its entire
1441         responsiblity is duplicated by the "font" pointer.
1442
1443         No new tests because there is no behavior change.
1444
1445         * platform/graphics/mac/ComplexTextController.cpp:
1446         (WebCore::ComplexTextController::collectComplexTextRuns):
1447
1448 == Rolled over to ChangeLog-2015-11-21 ==