1 2016-11-09 Alex Christensen <achristensen@webkit.org>
3 URLParser should not consider path of URLs with no host to start at the first slash after the colon
4 https://bugs.webkit.org/show_bug.cgi?id=164555
6 Reviewed by Tim Horton.
8 When we see a url that is only scheme:// we treated the // as the path. Firefox did this with unrecognized schemes,
9 but based on https://github.com/whatwg/url/issues/148 they seem willing to change. We had added similar behavior to
10 URL::parse, and I added this to URLParser in r206783 which this effectively reverts.
12 Covered by API and layout tests.
14 * platform/URLParser.cpp:
15 (WebCore::URLParser::parse):
16 Don't move m_userStart to m_pathStart back by two when we see an empty host.
18 2016-11-09 Alex Christensen <achristensen@webkit.org>
20 Simplify logic of SecurityOrigin::databaseIdentifier
21 https://bugs.webkit.org/show_bug.cgi?id=164565
23 Reviewed by Brady Eidson.
25 No change in behavior.
27 SecurityOrigins with the file scheme need a special database identifier to be backwards-compatible with existing storage.
28 Instead of determining whether this is a file SecurityOrigin at parsing time and only using that information when
29 making the database identifier, just determine whether we need this quirk when making the database identifier.
30 I'm planning to move this logic to SecurityOriginData in another patch.
32 * page/SecurityOrigin.cpp:
33 (WebCore::SecurityOrigin::SecurityOrigin):
34 (WebCore::SecurityOrigin::create):
35 (WebCore::SecurityOrigin::databaseIdentifier):
36 * page/SecurityOrigin.h:
38 2016-11-09 Jaehun Lim <ljaehun.lim@samsung.com>
40 Unreviewed, build fix after r208460
42 isValidColorString() was renamed isValidSimpleColorString().
44 * html/ColorInputType.cpp:
45 (WebCore::ColorInputType::suggestions):
47 2016-11-09 Anders Carlsson <andersca@apple.com>
53 2016-11-09 Simon Fraser <simon.fraser@apple.com>
55 Implement visual-viewport based position:fixed handling for Mac async scrolling
56 https://bugs.webkit.org/show_bug.cgi?id=164495
58 Reviewed by Tim Horton.
60 Educate the scrolling tree about visual and layout viewports. This is runtime-switchable,
61 so we push the enable flag to via the root state node, then push the layout viewport,
62 and the min/max scroll position that contstrain it, through frame state nodes.
64 When a scroll happens, we compute a new layout viewport when the visual viewport hits
65 an edge, and push that down through setScrollLayerPosition() since it's used to position
66 fixed and sticky layers.
68 When the main thread gets notified about an async scroll, we set the new layout viewport
69 on the FrameView, but do so in such a way that does not trigger layout. This is OK because
70 we do a RenderLayer update which udpates all the layoutViewport-dependent state, and is
71 necessary to avoid repaints every main thread update.
73 The iOS code is made to compile, but not work yet.
75 Tests: compositing/tiling/visiblerect-accumulated-offset.html
76 fast/visual-viewport/tiled-drawing/zoomed-fixed-scrolled-down-then-up.html
77 fast/visual-viewport/tiled-drawing/zoomed-fixed-scrolled-down.html
78 fast/visual-viewport/tiled-drawing/zoomed-fixed-scrolling-layers-state.html
81 (WebCore::FrameView::setLayoutViewportOrigin):
82 (WebCore::FrameView::updateLayoutViewport):
83 (WebCore::FrameView::visualViewportRect):
84 (WebCore::FrameView::unscaledMinimumScrollPosition):
85 (WebCore::FrameView::scrollPositionChanged):
87 * page/scrolling/AsyncScrollingCoordinator.cpp:
88 (WebCore::AsyncScrollingCoordinator::frameViewLayoutUpdated):
89 (WebCore::AsyncScrollingCoordinator::requestScrollPositionUpdate):
90 (WebCore::AsyncScrollingCoordinator::scheduleUpdateScrollPositionAfterAsyncScroll):
91 (WebCore::AsyncScrollingCoordinator::updateScrollPositionAfterAsyncScrollTimerFired):
92 (WebCore::AsyncScrollingCoordinator::updateScrollPositionAfterAsyncScroll):
93 (WebCore::AsyncScrollingCoordinator::visualViewportEnabled):
94 * page/scrolling/AsyncScrollingCoordinator.h:
95 (WebCore::AsyncScrollingCoordinator::ScheduledScrollUpdate::ScheduledScrollUpdate):
96 * page/scrolling/ScrollingStateFrameScrollingNode.cpp:
97 (WebCore::ScrollingStateFrameScrollingNode::ScrollingStateFrameScrollingNode):
98 (WebCore::ScrollingStateFrameScrollingNode::setLayoutViewport):
99 (WebCore::ScrollingStateFrameScrollingNode::setMinLayoutViewportOrigin):
100 (WebCore::ScrollingStateFrameScrollingNode::setMaxLayoutViewportOrigin):
101 (WebCore::ScrollingStateFrameScrollingNode::setVisualViewportEnabled):
102 (WebCore::ScrollingStateFrameScrollingNode::dumpProperties):
103 * page/scrolling/ScrollingStateFrameScrollingNode.h:
104 * page/scrolling/ScrollingTree.cpp:
105 (WebCore::ScrollingTree::viewportChangedViaDelegatedScrolling):
106 (WebCore::ScrollingTree::scrollPositionChangedViaDelegatedScrolling):
107 (WebCore::ScrollingTree::commitTreeState):
108 * page/scrolling/ScrollingTree.h:
109 (WebCore::ScrollingTree::visualViewportEnabled):
110 (WebCore::ScrollingTree::setVisualViewportEnabled):
111 * page/scrolling/ScrollingTreeFrameScrollingNode.cpp:
112 (WebCore::ScrollingTreeFrameScrollingNode::commitStateBeforeChildren):
113 (WebCore::ScrollingTreeFrameScrollingNode::layoutViewportForScrollPosition):
114 * page/scrolling/ScrollingTreeFrameScrollingNode.h:
115 (WebCore::ScrollingTreeFrameScrollingNode::layoutViewport):
116 (WebCore::ScrollingTreeFrameScrollingNode::minLayoutViewportOrigin):
117 (WebCore::ScrollingTreeFrameScrollingNode::maxLayoutViewportOrigin):
118 * page/scrolling/ScrollingTreeScrollingNode.cpp:
119 (WebCore::ScrollingTreeScrollingNode::setScrollPositionWithoutContentEdgeConstraints):
120 * page/scrolling/ScrollingTreeScrollingNode.h:
121 * page/scrolling/ThreadedScrollingTree.cpp:
122 (WebCore::ThreadedScrollingTree::scrollingTreeNodeDidScroll):
123 * page/scrolling/ThreadedScrollingTree.h:
124 * page/scrolling/ios/ScrollingTreeFrameScrollingNodeIOS.h:
125 * page/scrolling/ios/ScrollingTreeFrameScrollingNodeIOS.mm:
126 (WebCore::ScrollingTreeFrameScrollingNodeIOS::setScrollPositionWithoutContentEdgeConstraints):
127 (WebCore::ScrollingTreeFrameScrollingNodeIOS::setScrollLayerPosition):
128 * page/scrolling/ios/ScrollingTreeIOS.cpp:
129 (WebCore::ScrollingTreeIOS::scrollingTreeNodeDidScroll):
130 * page/scrolling/ios/ScrollingTreeIOS.h:
131 * page/scrolling/mac/ScrollingTreeFrameScrollingNodeMac.h:
132 * page/scrolling/mac/ScrollingTreeFrameScrollingNodeMac.mm:
133 (WebCore::ScrollingTreeFrameScrollingNodeMac::setScrollPositionWithoutContentEdgeConstraints):
134 (WebCore::ScrollingTreeFrameScrollingNodeMac::setScrollLayerPosition):
136 2016-11-09 Brady Eidson <beidson@apple.com>
138 IndexedDB 2.0: W3C test IndexedDB/idbtransaction_objectStoreNames.html fails.
139 https://bugs.webkit.org/show_bug.cgi?id=164528
141 Reviewed by Alex Christensen.
143 No new tests (Covered by existing test).
145 * Modules/indexeddb/IDBDatabase.cpp:
146 (WebCore::IDBDatabase::transaction): De-dupe the input names.
148 2016-11-09 Brady Eidson <beidson@apple.com>
150 IndexedDB 2.0: Implement new IDBCursor.continuePrimaryKey function.
151 https://bugs.webkit.org/show_bug.cgi?id=164404
153 Reviewed by Alex Christensen.
155 Tests: storage/indexeddb/modern/idbcursor-continue-primary-key-1-private.html
156 storage/indexeddb/modern/idbcursor-continue-primary-key-1.html
157 Also covered by existing tests.
159 * Modules/indexeddb/IDBCursor.cpp:
160 (WebCore::IDBCursor::continuePrimaryKey):
161 (WebCore::IDBCursor::uncheckedIterateCursor):
162 * Modules/indexeddb/IDBCursor.h:
163 * Modules/indexeddb/IDBCursor.idl:
165 * Modules/indexeddb/IDBKeyData.h:
166 (WebCore::IDBKeyData::operator>):
167 (WebCore::IDBKeyData::operator<=):
168 (WebCore::IDBKeyData::operator>=):
170 * Modules/indexeddb/server/MemoryCursor.h:
172 * Modules/indexeddb/server/MemoryIDBBackingStore.cpp:
173 (WebCore::IDBServer::MemoryIDBBackingStore::iterateCursor):
175 * Modules/indexeddb/server/MemoryIndexCursor.cpp:
176 (WebCore::IDBServer::MemoryIndexCursor::iterate):
177 * Modules/indexeddb/server/MemoryIndexCursor.h:
179 * Modules/indexeddb/server/MemoryObjectStoreCursor.cpp:
180 (WebCore::IDBServer::MemoryObjectStoreCursor::iterate):
181 * Modules/indexeddb/server/MemoryObjectStoreCursor.h:
183 * Modules/indexeddb/server/SQLiteIDBBackingStore.cpp:
184 (WebCore::IDBServer::SQLiteIDBBackingStore::iterateCursor):
186 * Modules/indexeddb/server/SQLiteIDBCursor.cpp:
187 (WebCore::IDBServer::SQLiteIDBCursor::iterate):
188 * Modules/indexeddb/server/SQLiteIDBCursor.h:
190 * Modules/indexeddb/shared/IDBIterateCursorData.cpp:
191 (WebCore::IDBIterateCursorData::isolatedCopy):
192 * Modules/indexeddb/shared/IDBIterateCursorData.h:
193 (WebCore::IDBIterateCursorData::encode):
194 (WebCore::IDBIterateCursorData::decode):
196 2016-11-09 Antoine Quint <graouts@apple.com>
198 [Modern Media Controls] Media Controller: set status label according to media state
199 https://bugs.webkit.org/show_bug.cgi?id=164557
200 <rdar://problem/29184097>
202 Reviewed by Dean Jackson.
204 Correctly set the StatusLabel text based on the media loading and network state.
206 Tests: http/tests/media/modern-media-controls/status-support/status-support-live-broadcast.html
207 http/tests/media/modern-media-controls/status-support/status-support-loading.html
208 media/modern-media-controls/status-support/status-support-error.html
210 * Modules/modern-media-controls/js-files:
211 * Modules/modern-media-controls/media/media-controller.js:
213 * Modules/modern-media-controls/media/status-support.js: Added.
214 (StatusSupport.prototype.get control):
215 (StatusSupport.prototype.get mediaEvents):
216 (StatusSupport.prototype.syncControl):
218 * WebCore.xcodeproj/project.pbxproj:
220 2016-11-09 Zalan Bujtas <zalan@apple.com>
222 No need to set setFlowThreadState on RenderText in createTextRenderer.
223 https://bugs.webkit.org/show_bug.cgi?id=164559
225 Reviewed by Antti Koivisto.
227 setFlowThreadState in create*Renderer ensures that by the time we issue the initial call to
228 styleWillChange/styleDidChange through initializeStyle, the state is already set.
229 However since RenderText does not have its own style, it's sufficient to have the flow state set
230 through the normal RenderElement::insertChildInternal code path.
232 No change in functionality.
234 * style/RenderTreeUpdater.cpp:
235 (WebCore::createTextRenderer):
237 2016-11-09 Brady Eidson <beidson@apple.com>
239 IndexedDB 2.0: Encapsulate cursor iteration parameters for easy future expansion.
240 https://bugs.webkit.org/show_bug.cgi?id=164504
242 Reviewed by Darin Adler.
244 No new tests (Refactor, no behavior change).
246 This patch literally just takes the "key" and "count" arguments and encapsulates them in a struct.
247 That struct will then be easily expandable in the future (e.g. bug 164404).
249 * Modules/indexeddb/IDBCursor.cpp:
250 (WebCore::IDBCursor::uncheckedIterateCursor):
252 * Modules/indexeddb/IDBTransaction.cpp:
253 (WebCore::IDBTransaction::iterateCursor):
254 (WebCore::IDBTransaction::iterateCursorOnServer):
255 * Modules/indexeddb/IDBTransaction.h:
257 * Modules/indexeddb/client/IDBConnectionProxy.cpp:
258 (WebCore::IDBClient::IDBConnectionProxy::iterateCursor):
259 * Modules/indexeddb/client/IDBConnectionProxy.h:
261 * Modules/indexeddb/client/IDBConnectionToServer.cpp:
262 (WebCore::IDBClient::IDBConnectionToServer::iterateCursor):
263 * Modules/indexeddb/client/IDBConnectionToServer.h:
264 * Modules/indexeddb/client/IDBConnectionToServerDelegate.h:
266 * Modules/indexeddb/server/IDBBackingStore.h:
268 * Modules/indexeddb/server/IDBServer.cpp:
269 (WebCore::IDBServer::IDBServer::iterateCursor):
270 * Modules/indexeddb/server/IDBServer.h:
272 * Modules/indexeddb/server/MemoryIDBBackingStore.cpp:
273 (WebCore::IDBServer::MemoryIDBBackingStore::iterateCursor):
274 * Modules/indexeddb/server/MemoryIDBBackingStore.h:
276 * Modules/indexeddb/server/SQLiteIDBBackingStore.cpp:
277 (WebCore::IDBServer::SQLiteIDBBackingStore::iterateCursor):
278 * Modules/indexeddb/server/SQLiteIDBBackingStore.h:
280 * Modules/indexeddb/server/UniqueIDBDatabase.cpp:
281 (WebCore::IDBServer::UniqueIDBDatabase::iterateCursor):
282 (WebCore::IDBServer::UniqueIDBDatabase::performIterateCursor):
283 * Modules/indexeddb/server/UniqueIDBDatabase.h:
285 * Modules/indexeddb/server/UniqueIDBDatabaseTransaction.cpp:
286 (WebCore::IDBServer::UniqueIDBDatabaseTransaction::iterateCursor):
287 * Modules/indexeddb/server/UniqueIDBDatabaseTransaction.h:
289 * Modules/indexeddb/shared/IDBIterateCursorData.cpp: Added.
290 (WebCore::IDBIterateCursorData::isolatedCopy):
291 * Modules/indexeddb/shared/IDBIterateCursorData.h: Added.
292 (WebCore::IDBIterateCursorData::encode):
293 (WebCore::IDBIterateCursorData::decode):
295 * Modules/indexeddb/shared/InProcessIDBServer.cpp:
296 (WebCore::InProcessIDBServer::iterateCursor):
297 * Modules/indexeddb/shared/InProcessIDBServer.h:
300 * WebCore.xcodeproj/project.pbxproj:
302 2016-11-09 Ryosuke Niwa <rniwa@webkit.org>
304 StyledElement::attributeChanged shouldn't do any work when the attribute value didn't change
305 https://bugs.webkit.org/show_bug.cgi?id=129476
307 Reviewed by Andreas Kling.
309 Avoid calling styleAttributeChanged and setPresentationAttributeStyleIsDirty
310 when the attribute value didn't change as in r164856.
312 * dom/StyledElement.cpp:
313 (WebCore::StyledElement::attributeChanged):
315 2016-11-09 Yusuke Suzuki <utatane.tea@gmail.com>
317 [DOMJIT] Implement Node::ownerDocument
318 https://bugs.webkit.org/show_bug.cgi?id=164004
320 Reviewed by Darin Adler.
322 Test: js/dom/domjit-accessor-owner-document.html
324 Still I cannot reproduce this crash in x64 environment, according to the crash log, it accesses 0x8 address.
325 This can happen if document() accidentally returns nullptr. In the C++ ownerDocument implementation,
326 if document() returns nullptr, it just returns nullptr. But in the DOMJIT implementation, we assume that
327 document() won't return nullptr and access the member of it.
329 This patch aligns the DOMJIT implementation strictly to the C++ one.
332 * domjit/JSNodeDOMJIT.cpp:
333 (WebCore::NodeOwnerDocumentDOMJIT::checkDOM):
334 (WebCore::NodeOwnerDocumentDOMJIT::callDOMGetter):
336 2016-11-09 Sam Weinig <sam@webkit.org>
338 [SVG] Start moving special casing of SVG out of the bindings - SVGAngle
339 https://bugs.webkit.org/show_bug.cgi?id=164496
341 Reviewed by Darin Adler.
343 There is quite a bit of special casing of SVG types in the bindings that adds
344 a lot of complexity and is relatively fragile, as it is based on type naming.
346 Instead of keeping the complexity in the bindings, I am going to move it into
347 the implementation, where it has also longed to be.
349 Starting small, with just SVGAngle. It has been split in two, with the existing
350 SVGAngle being renamed SVGAngleValue, and the bound instance, which used to be name
351 SVGPropertyTearOff<SVGAngle>, taking the name SVGAngle (and inheriting from
352 SVGPropertyTearOff<SVGAngleValue>).
355 * WebCore.xcodeproj/project.pbxproj:
356 Add SVGAngleValue.cpp
358 * bindings/scripts/CodeGenerator.pm:
359 Remove SVGAngle as a special case.
361 * svg/SVGAngle.cpp: Removed.
363 Added. Implements the SVGAngle interface explicitly, getting to
364 the SVGAngleValue through propertyReference().
367 * svg/SVGAngleValue.cpp: Copied from Source/WebCore/svg/SVGAngle.cpp.
368 * svg/SVGAngleValue.h: Copied from Source/WebCore/svg/SVGAngle.h.
369 Move old SVGAngle implementation to SVGAngleValue.
371 * svg/SVGAnimatedAngle.cpp:
372 Replace SVGAngle usage with SVGAngleValue.
374 * svg/SVGAnimatedAngle.h:
375 Switch SVGAnimatedAngle to be a type alias. This remains SVGAnimatedPropertyTearOff<SVGAngle>
376 as SVGAnimatedPropertyTearOff has been changed to take the TearOff type as its parameter.
378 * svg/SVGAnimatedLength.h:
379 * svg/SVGAnimatedPreserveAspectRatio.h:
380 * svg/SVGAnimatedRect.h:
381 Switch to using type aliases and pass the TearOff to SVGAnimatedPropertyTearOff.
383 * svg/SVGAnimatedType.cpp:
384 (WebCore::SVGAnimatedType::createAngleAndEnumeration):
385 * svg/SVGAnimatedType.h:
386 (WebCore::SVGAnimatedType::angleAndEnumeration):
389 * svg/SVGComponentTransferFunctionElement.h:
390 Add missing include of SVGElement.h (need because it removed from SVGPropertyTearOff).
392 * svg/SVGMarkerElement.cpp:
393 (WebCore::SVGMarkerElement::parseAttribute):
394 (WebCore::SVGMarkerElement::setOrient):
395 Switch to take an SVGAngleValue.
397 (WebCore::SVGMarkerElement::setOrientToAngle):
398 Update to pull the value out via propertyReference().
400 * svg/SVGMarkerElement.h:
401 Switch to take an SVGAngleValue.
403 * svg/SVGLengthList.h:
404 * svg/SVGNumberList.h:
405 * svg/SVGPathSegList.h:
406 * svg/SVGPointList.h:
407 * svg/SVGStringList.h:
408 * svg/SVGTransformList.h:
409 Switch to using type aliases in SVGPropertyTraits and add an alias for
412 * svg/SVGSVGElement.cpp:
413 (WebCore::SVGSVGElement::createSVGAngle):
414 * svg/SVGSVGElement.h:
415 Change createSVGAngle to return a Ref<SVGAngle> and create one.
417 * svg/SVGSVGElement.idl:
418 Annotate IDL to indicate that a new value is being returned.
420 * svg/SVGTransform.cpp:
421 Remove unnecessary include of SVGAngle.h.
423 * svg/SVGViewSpec.cpp:
424 Add missing include of SVGElement.h (need because it removed from SVGPropertyTearOff).
426 * svg/properties/SVGAnimatedPropertyTearOff.h:
427 Change to be parameterized on the TearOffType, rather than the PropertyType itself. Get the
428 Property type from the TearOffType.
430 * svg/properties/SVGListProperty.h:
431 * svg/properties/SVGListPropertyTearOff.h:
432 Fix assumption that all TearOffTypes are just a SVGPropertyTearOff templatized on a property
433 type. This is no longer true for SVGAngle. Instead, get the TearOffType for lists via SVGPropertyTraits.
435 * svg/properties/SVGPropertyTearOff.h:
436 Make the PropertyType available by exposing it as a type alias.
438 2016-11-09 Darin Adler <darin@apple.com>
440 Move Range from ExceptionCode to ExceptionOr
441 https://bugs.webkit.org/show_bug.cgi?id=164457
443 Reviewed by Alex Christensen.
445 * accessibility/AXObjectCache.cpp:
446 (WebCore::AXObjectCache::rangeForNodeContents): Update to use ExceptionOr,
447 keeping behavior the same.
448 (WebCore::characterOffsetsInOrder): Ditto.
449 (WebCore::setRangeStartOrEndWithCharacterOffset): Changed argument to a
450 reference instead of a pointer. Use a boolean return value to indicate
451 success rather than an exception, since the callers don't need to know
452 which exception it is.
453 (WebCore::AXObjectCache::rangeForUnorderedCharacterOffsets): Updated for
455 (WebCore::AXObjectCache::nextBoundary): Ditto.
456 (WebCore::AXObjectCache::previousBoundary): Ditto.
458 * accessibility/AccessibilityObject.cpp:
459 (WebCore::AccessibilityObject::rangeOfStringClosestToRangeInDirection):
460 Update to use ExceptionOr, keeping behavior the same.
461 * accessibility/AccessibilityRenderObject.cpp:
462 (WebCore::AccessibilityRenderObject::documentBasedSelectedTextRange): Ditto.
463 * accessibility/atk/WebKitAccessibleUtil.cpp:
464 (selectionBelongsToObject): Ditto.
465 * accessibility/ios/WebAccessibilityObjectWrapperIOS.mm:
466 (-[WebAccessibilityObjectWrapper _convertToNSRange:]): Ditto.
468 (WebCore::Node::textRects): Ditto.
471 (WebCore::Range::~Range): Remove old comment that no longer makes sense now
472 that the detach function no longer does anything.
473 (WebCore::checkForDifferentRootContainer): Updated to use ExceptionOr,
474 keeping behavior the same.
475 (WebCore::Range::setStart): Ditto.
476 (WebCore::Range::setEnd): Ditto.
477 (WebCore::Range::isPointInRange): Ditto.
478 (WebCore::Range::comparePoint): Ditto.
479 (WebCore::Range::compareNode): Ditto.
480 (WebCore::top): Added helper function so that compareBoundaryPoints doesn't
481 need to have two identical loops in it.
482 (WebCore::Range::compareBoundaryPoints): Updated to use ExceptionOr,
483 keeping behavior the same.
484 (WebCore::Range::compareBoundaryPointsForBindings): Ditto. Also use a switch
485 instead of relying on the order of the values to check for unsupported values.
486 (WebCore::Range::boundaryPointsValid): Ditto.
487 (WebCore::Range::deleteContents): Ditto.
488 (WebCore::Range::intersectsNode): Ditto.
489 (WebCore::Range::processContents): Ditto.
490 (WebCore::deleteCharacterData): Ditto.
491 (WebCore::processContentsBetweenOffsets): Ditto. Also changed to be a
492 non-member function and private to this file instead of in the class.
493 (WebCore::processNodes): Ditto. Also changed one argument to be a RefPtr
494 since the code relies on using it after mutating the DOM.
495 (WebCore::processAncestorsAndTheirSiblings): Ditto. Changed one argument type
496 to use ExceptionOr so the caller doesn't have to check the exception first.
497 (WebCore::Range::extractContents): Ditto.
498 (WebCore::Range::cloneContents): Ditto.
499 (WebCore::Range::insertNode): Ditto. Also fixed to only call nodeType once
500 instead of three times.
501 (WebCore::Range::toString): Ditto. Also fixed to call nodeType only once
502 per node instead of twice, to use downcast instead of static_cast, and to
503 use the word "node" instead of "n" for the local variable name.
504 (WebCore::Range::createContextualFragment): Ditto.
505 (WebCore::Range::checkNodeWOffset): Ditto.
506 (WebCore::Range::setStartAfter): Ditto.
507 (WebCore::Range::setEndBefore): Ditto.
508 (WebCore::Range::setEndAfter): Ditto.
509 (WebCore::Range::selectNode): Ditto.
510 (WebCore::Range::selectNodeContents): Ditto.
511 (WebCore::Range::surroundContents): Ditto.
512 (WebCore::Range::setStartBefore): Ditto.
513 (WebCore::Range::contains): Ditto. Except added code to handle exception
514 case to return false without asserting because I saw at least one crash
515 that seemed to imply this behavior was needed.
516 (WebCore::rangesOverlap): Ditto.
517 (WebCore::rangeOfContents): Ditto.
518 (WebCore::Range::expand): Ditto.
519 (WebCore::Range::getClientRects): Ditto.
520 (WebCore::Range::getBoundingClientRect): Ditto.
521 (WebCore::Range::borderAndTextQuads): Changed to use return value
522 instead of out argument, since it's a private function used only
523 within this class so it was easy to update all call sites.
524 (WebCore::Range::boundingRect): Updated for above. Also renamed since
525 there was no need for the name "internal" in this.
526 (WebCore::Range::absoluteBoundingRect): Ditto.
527 * dom/Range.h: Updated for above.
528 * dom/Range.idl: Use non-legacy exceptions. Also changed the default value
529 of the string argument to the expand function to the empty string rather
530 than "undefined", because the function silently does nothing when passed
531 any unrecognized string, and so this leaves behavior unchanged. I removed
532 the comment saying that the "undefined" default is wrong.
534 * editing/AlternativeTextController.cpp:
535 (WebCore::AlternativeTextController::applyAlternativeTextToRange): Updated
536 to use ExceptionOr but behave the same.
537 * editing/Editor.cpp:
538 (WebCore::Editor::advanceToNextMisspelling): Ditto.
539 (WebCore::Editor::markAndReplaceFor): Ditto.
540 (WebCore::isFrameInRange): Ditto. Also made a few style tweaks.
541 (WebCore::Editor::countMatchesForText): Ditto.
542 * editing/EditorCommand.cpp:
543 (WebCore::unionDOMRanges): Ditto.
544 * editing/FrameSelection.cpp:
545 (WebCore::FrameSelection::respondToNodeModification): Ditto.
546 * editing/InsertListCommand.cpp:
547 (WebCore::InsertListCommand::doApplyForSingleParagraph): Ditto.
548 * editing/TextCheckingHelper.cpp:
549 (WebCore::TextCheckingParagraph::offsetTo): Ditto.
550 * editing/TextCheckingHelper.h: Updated for above and also deleted
551 unneeded private function checkingRange, which just churned the
552 reference count unnecessarily; instead use m_checkingRange directly.
553 * editing/TextIterator.cpp:
554 (WebCore::TextIterator::getLocationAndLengthFromRange): Ditto.
555 * editing/VisiblePosition.cpp:
556 (WebCore::setStart): Ditto.
557 (WebCore::setEnd): Ditto.
558 * editing/VisibleSelection.cpp:
559 (WebCore::makeSearchRange): Ditto.
561 * editing/VisibleUnits.cpp:
562 (WebCore::suffixLengthForRange): Changed argument from RefPtr to
564 (WebCore::prefixLengthForRange): Ditto.
565 (WebCore::previousBoundary): Updated for ExceptionOr and the change
567 (WebCore::nextBoundary): Ditto.
568 * editing/VisibleUnits.h: Updated for above.
570 * editing/htmlediting.cpp:
571 (WebCore::comparePositions): Updated to use ExceptionOr but behave
573 (WebCore::visiblePositionForIndexUsingCharacterIterator): Ditto.
574 (WebCore::isNodeVisiblyContainedWithin): Ditto.
575 * editing/ios/EditorIOS.mm:
576 (WebCore::Editor::setDictationPhrasesAsChildOfElement): Ditto.
577 (WebCore::Editor::setTextAsChildOfElement): Ditto.
578 * editing/mac/EditorMac.mm:
579 (WebCore::Editor::adjustedSelectionRange): Ditto.
580 * editing/markup.cpp:
581 (WebCore::createMarkupInternal): Ditto.
582 * page/ContextMenuController.cpp:
583 (WebCore::ContextMenuController::contextMenuItemSelected): Ditto.
584 * page/DOMSelection.cpp:
585 (WebCore::DOMSelection::addRange): Ditto.
586 (WebCore::DOMSelection::deleteFromDocument): Ditto.
587 (WebCore::DOMSelection::containsNode): Ditto.
589 * page/EventHandler.cpp:
590 (WebCore::EventHandler::dispatchMouseEvent): Updated for change to
591 use ExceptionOr in Ragne::compareNode. Also refactored the function
592 to make the logic a little mroe straightforward and nest less of it
596 (WebCore::Page::findStringMatchingRanges): Updated for ExceptionOr
597 without changing behavior.
598 * page/TextIndicator.cpp:
599 (WebCore::hasNonInlineOrReplacedElements): Ditto.
600 * rendering/RenderNamedFlowThread.cpp:
601 (WebCore::RenderNamedFlowThread::getRanges): Ditto.
603 2016-11-09 Dave Hyatt <hyatt@apple.com>
605 [CSS Parser] Fix grid layout parsing
606 https://bugs.webkit.org/show_bug.cgi?id=164489
608 Reviewed by Dean Jackson.
610 * css/CSSValueKeywords.in:
611 * css/parser/CSSPropertyParser.cpp:
612 (WebCore::consumeFitContent):
613 (WebCore::isGridTrackFixedSized):
614 (WebCore::consumeGridTrackSize):
615 (WebCore::consumeGridTrackRepeatFunction):
616 (WebCore::consumeGridTrackList):
617 (WebCore::isCustomIdentValue):
618 (WebCore::CSSPropertyParser::consumeGridItemPositionShorthand):
619 (WebCore::CSSPropertyParser::consumeGridAreaShorthand):
620 (WebCore::consumeImplicitGridAutoFlow):
621 (WebCore::CSSPropertyParser::consumeGridShorthand):
623 2016-11-09 Darin Adler <darin@apple.com>
625 Move EventTarget from ExceptionCode to ExceptionOr
626 https://bugs.webkit.org/show_bug.cgi?id=164465
628 Reviewed by Youenn Fablet.
630 * Modules/indexeddb/IDBRequest.h: Added now-needed forward
632 * Modules/webaudio/AudioContext.h: Ditto.
634 * bindings/js/JSEventListener.cpp:
635 (WebCore::eventHandlerAttribute): Updated for name change of the
636 attributeEventListener function.
637 (WebCore::documentEventHandlerAttribute): Ditto.
639 (WebCore::Document::getWindowAttributeEventListener): Ditto.
641 * dom/EventTarget.cpp:
642 (WebCore::EventTarget::setAttributeEventListener): Updated for
644 (WebCore::EventTarget::attributeEventListener): Ditto.
645 (WebCore::EventTarget::dispatchEventForBindings): Use ExceptionOr.
646 (WebCore::legacyType): Use null instead of empty for no type, since
647 it's more efficient to check for null.
648 (WebCore::EventTarget::fireEventListeners): Check for null.
649 Also streamlined logic a little bit and removed a very old comment.
650 (WebCore::EventTarget::eventListeners): Renamed from getEventListeners.
651 * dom/EventTarget.h: Removed lots of unneeded declarations. Renamed
652 some functions to remove get prefix. Updated for above changes.
653 Moved one inline function out of the class header. Made the destructor
654 for EventTarget be inline to make the destructors for derived classes
655 slightly more efficient.
656 * dom/EventTarget.idl: Use non-legacy exception.
659 (WebCore::Node::didMoveToNewDocument): Updated for name change.
660 * editing/ReplaceSelectionCommand.cpp:
661 (WebCore::ReplacementFragment::ReplacementFragment): Ditto.
663 * inspector/InspectorCSSAgent.h: Added now-needed forward declaration.
665 * inspector/InspectorDOMAgent.cpp:
666 (WebCore::InspectorDOMAgent::getEventListeners): Updated for name change.
668 * inspector/InspectorInstrumentation.h: Added now-needed forward declaration.
669 * page/DOMWindow.h: Ditto.
670 * xml/XMLHttpRequest.h: Ditto.
672 2016-11-09 Daniel Bates <dabates@apple.com>
674 Add test infrastructure and tests for existing HTTP 0.9 sandbox machinery
675 https://bugs.webkit.org/show_bug.cgi?id=164389
676 <rdar://problem/29101072>
678 Reviewed by Alex Christensen.
680 Add test infrastructure to support registering an arbitrary port as the default port
681 for a protocol. The behavior of various machinery, including the HTTP 0.9 machinery,
682 can be effected by whether the resource request was made using the default port for
683 the protocol. We expose window.internals.registerDefaultPortForProtocol() to allow
684 a test to override the default port associated with a protocol so as to support
685 testing these code paths using the existing port 8000 server started by run-webkit-httpd.
686 Without window.internals.registerDefaultPortForProtocol() we would need to teach
687 run-webkit-httpd to run a web server on port 80, which requires superuser privileges
688 (since it is a privileged port number) and is more likely to interfere with an
689 existing web server setup.
691 Tests: http/tests/security/http-0.9/default-port-plugin-blocked.html
692 http/tests/security/http-0.9/default-port-script-blocked.html
693 http/tests/security/http-0.9/iframe-blocked.html
694 http/tests/security/http-0.9/image-blocked.html
695 http/tests/security/http-0.9/image-on-HTTP-0.9-default-port-page-allowed-ref-test.html
696 http/tests/security/http-0.9/image-on-HTTP-0.9-default-port-page-allowed.html
697 http/tests/security/http-0.9/image-on-HTTP-0.9-page-blocked.html
698 http/tests/security/http-0.9/worker-connect-src-blocked.html
699 http/tests/security/http-0.9/worker-importScripts-blocked.html
700 http/tests/security/http-0.9/xhr-asynchronous-blocked.html
703 (WebCore::defaultPortForProtocolMapForTesting): Added.
704 (WebCore::registerDefaultPortForProtocolForTesting): Adds the specified (protocol, port) to the
705 mapping used for testing.
706 (WebCore::clearDefaultPortForProtocolMapForTesting): Clears the protocol to default port testing map.
707 We call this function from Internals::resetToConsistentState() so that the mapping is cleared between
709 (WebCore::defaultPortForProtocol): Modified to check the protocol to default port map for testing
710 before consulting URLParser::defaultPortForProtocol().
712 * testing/Internals.cpp:
713 (WebCore::Internals::resetToConsistentState): Clear the default port mapping used for testing.
714 (WebCore::Internals::registerDefaultPortForProtocol): Added.
715 * testing/Internals.h:
716 * testing/Internals.idl: Added declaration for registerDefaultPortForProtocol().
718 2016-11-09 Sam Weinig <sam@webkit.org>
720 [WebIDL] Add proper parsing for Promises
721 https://bugs.webkit.org/show_bug.cgi?id=164497
723 Reviewed by Tim Horton.
725 * bindings/scripts/IDLParser.pm:
728 Require Promise types to declare the type they resolve to.
730 * bindings/js/JSDOMPromise.h:
731 Allow DOMPromise to be be parameterized on void. Add an SFINAE guarded
732 overload of resolve that takes no arguments when in a DOMPromise<void>.
734 * Modules/applepay/ApplePaySession.idl:
735 * Modules/fetch/DOMWindowFetch.idl:
736 * Modules/fetch/FetchBody.idl:
737 * Modules/fetch/FetchResponse.idl:
738 * Modules/fetch/WorkerGlobalScopeFetch.idl:
739 * Modules/mediastream/MediaDevices.idl:
740 * Modules/mediastream/MediaStreamTrack.idl:
741 * Modules/mediastream/RTCPeerConnection.idl:
742 * Modules/mediastream/RTCRtpSender.idl:
743 * Modules/mediastream/RTCStatsReport.idl:
744 * Modules/streams/ReadableStream.idl:
745 * Modules/streams/ReadableStreamDefaultReader.idl:
746 * Modules/streams/ReadableStreamSource.idl:
747 * Modules/streams/WritableStream.idl:
748 * Modules/webaudio/AudioContext.idl:
749 * bindings/scripts/test/TestNode.idl:
750 * bindings/scripts/test/TestObj.idl:
751 * crypto/SubtleCrypto.idl:
752 * crypto/WebKitSubtleCrypto.idl:
754 * css/FontFaceSet.idl:
755 * dom/CustomElementRegistry.idl:
756 * html/HTMLMediaElement.idl:
757 Update IDLs to specify the resolve type of promise types.
759 * Modules/mediastream/MediaEndpointPeerConnection.cpp:
760 (WebCore::MediaEndpointPeerConnection::replaceTrack):
761 (WebCore::MediaEndpointPeerConnection::replaceTrackTask):
762 * Modules/mediastream/MediaStreamTrack.cpp:
763 (WebCore::MediaStreamTrack::applyConstraints):
764 * Modules/mediastream/MediaStreamTrack.h:
765 * Modules/mediastream/PeerConnectionBackend.cpp:
766 (WebCore::PeerConnectionBackend::setLocalDescriptionSucceeded):
767 (WebCore::PeerConnectionBackend::setRemoteDescriptionSucceeded):
768 (WebCore::PeerConnectionBackend::addIceCandidateSucceeded):
769 * Modules/mediastream/PeerConnectionBackend.h:
770 * Modules/streams/ReadableStreamSource.h:
771 (WebCore::ReadableStreamSource::start):
772 (WebCore::ReadableStreamSource::pull):
773 (WebCore::ReadableStreamSource::startFinished):
774 (WebCore::ReadableStreamSource::pullFinished):
775 * Modules/webaudio/AudioContext.cpp:
776 (WebCore::AudioContext::addReaction):
777 (WebCore::AudioContext::setState):
778 (WebCore::AudioContext::suspend):
779 (WebCore::AudioContext::resume):
780 (WebCore::AudioContext::close):
781 * Modules/webaudio/AudioContext.h:
782 * html/HTMLMediaElement.cpp:
783 (WebCore::HTMLMediaElement::rejectPendingPlayPromises):
784 (WebCore::HTMLMediaElement::resolvePendingPlayPromises):
785 (WebCore::HTMLMediaElement::play):
786 * html/HTMLMediaElement.h:
787 Update implementations to use DOMPromise<void> rather than DOMPromise<nullptr_t>
788 and use the new resolve() overload.
790 2016-11-07 Brady Eidson <beidson@apple.com>
792 Followup to https://bugs.webkit.org/show_bug.cgi?id=164466 - Make an IDBTransaction* be an IDBTransaction&
794 Rubberstamped by Alex Christensen.
796 No new tests (No behavior change).
798 * Modules/indexeddb/IDBObjectStore.cpp:
799 (WebCore::IDBObjectStore::IDBObjectStore):
800 (WebCore::IDBObjectStore::~IDBObjectStore):
801 (WebCore::IDBObjectStore::hasPendingActivity):
802 (WebCore::IDBObjectStore::name):
803 (WebCore::IDBObjectStore::setName):
804 (WebCore::IDBObjectStore::keyPath):
805 (WebCore::IDBObjectStore::indexNames):
806 (WebCore::IDBObjectStore::transaction):
807 (WebCore::IDBObjectStore::autoIncrement):
808 (WebCore::IDBObjectStore::openCursor):
809 (WebCore::IDBObjectStore::openKeyCursor):
810 (WebCore::IDBObjectStore::get):
811 (WebCore::IDBObjectStore::putOrAdd):
812 (WebCore::IDBObjectStore::doDelete):
813 (WebCore::IDBObjectStore::clear):
814 (WebCore::IDBObjectStore::createIndex):
815 (WebCore::IDBObjectStore::index):
816 (WebCore::IDBObjectStore::deleteIndex):
817 (WebCore::IDBObjectStore::doCount):
818 (WebCore::IDBObjectStore::getAll):
819 (WebCore::IDBObjectStore::getAllKeys):
820 (WebCore::IDBObjectStore::markAsDeleted):
821 (WebCore::IDBObjectStore::rollbackForVersionChangeAbort):
822 (WebCore::IDBObjectStore::ref):
823 (WebCore::IDBObjectStore::deref):
824 * Modules/indexeddb/IDBObjectStore.h:
826 2016-11-09 Zalan Bujtas <zalan@apple.com>
828 Move RenderNamedFlowThread nextRendererForElement logic to RenderTreeUpdater.
829 https://bugs.webkit.org/show_bug.cgi?id=164503
831 Reviewed by Antti Koivisto.
833 When we insert a renderer into the render tree, we need to know both its parent
834 and its next sibling. Normally the parent and the sibling are based on the DOM, but
835 when this renderer is part of a flow thread, its insertion sibling is not necessarily the DOM sibling.
836 To find the correct sibling, we call RenderNamedFlowThread's nextRendererForElement().
837 RenderNamedFlowThread keeps track of its children so that it can compute the next sibling
838 for the insertion point.
840 This patch eliminates the need for keeping track of the child renderers of each
841 flow by moving the 'next sibling' logic to RenderTreePosition.
843 No change in functionality.
845 * rendering/RenderElement.cpp:
846 (WebCore::RenderElement::insertedIntoTree):
847 (WebCore::RenderElement::willBeDestroyed):
848 (WebCore::RenderElement::removeFromRenderFlowThread):
849 (WebCore::RenderElement::renderNamedFlowThreadWrapper): Deleted.
850 * rendering/RenderElement.h:
851 * rendering/RenderNamedFlowThread.cpp:
852 (WebCore::RenderNamedFlowThread::nextRendererForElement): Deleted.
853 (WebCore::RenderNamedFlowThread::addFlowChild): Deleted.
854 (WebCore::RenderNamedFlowThread::removeFlowChild): Deleted.
855 * rendering/RenderNamedFlowThread.h:
856 * style/RenderTreePosition.cpp:
857 (WebCore::RenderTreePosition::previousSiblingRenderer):
858 (WebCore::RenderTreePosition::flowThreadInsertionContext):
859 * style/RenderTreePosition.h:
860 (WebCore::RenderTreePosition::RenderTreePosition):
861 (WebCore::RenderTreePosition::parent):
862 * style/RenderTreeUpdater.cpp:
863 (WebCore::registerElementForFlowThreadIfNeeded): We need to registed the element even when it does not create renderer (display: none).
864 (WebCore::RenderTreeUpdater::createRenderer):
865 (WebCore::moveToFlowThreadIfNeeded): Deleted.
867 2016-11-09 Per Arne Vollan <pvollan@apple.com>
869 [Win][Direct2D] Incomplete image decoding.
870 https://bugs.webkit.org/show_bug.cgi?id=164511
872 Reviewed by Darin Adler.
874 Create native decoder when all image data has been received.
876 * platform/graphics/win/ImageDecoderDirect2D.cpp:
877 (WebCore::ImageDecoder::setData):
879 2016-11-09 Beth Dakin <bdakin@apple.com>
883 * platform/spi/cocoa/AVKitSPI.h:
885 2016-11-09 Brady Eidson <beidson@apple.com>
887 IndexedDB 2.0: Clean up more transaction abort behavior, including tweaks to Index/ObjectStore lifetime.
888 https://bugs.webkit.org/show_bug.cgi?id=164466
890 Reviewed by Alex Christensen.
892 No new tests (Covered by existing tests that now pass).
894 Previously, IDBIndex ref/deref didn't track a traditional ref count but instead kept the owning object store alive.
895 Now, IDBObjectStore ref/deref do the same thing for the owning transaction.
897 Now when a version change transaction is rolled back, some object stores and indexes get pulled out of the "deleted"
898 set and get promoted back up into the "referenced" set.
900 Now deleted object stores/indexes are considered opaque roots, as live objects in the deleted state *can* get back
901 to the owning objects.
904 * WebCore.xcodeproj/project.pbxproj:
906 * Modules/indexeddb/IDBIndex.cpp:
907 (WebCore::IDBIndex::rollbackInfoForVersionChangeAbort):
909 * Modules/indexeddb/IDBObjectStore.cpp:
910 (WebCore::IDBObjectStore::IDBObjectStore):
911 (WebCore::IDBObjectStore::indexNames):
912 (WebCore::IDBObjectStore::transaction):
913 (WebCore::IDBObjectStore::openCursor):
914 (WebCore::IDBObjectStore::openKeyCursor):
915 (WebCore::IDBObjectStore::deleteIndex):
916 (WebCore::IDBObjectStore::rollbackForVersionChangeAbort):
917 (WebCore::IDBObjectStore::visitReferencedIndexes):
918 (WebCore::IDBObjectStore::ref):
919 (WebCore::IDBObjectStore::deref):
920 (WebCore::IDBObjectStore::create): Deleted.
921 * Modules/indexeddb/IDBObjectStore.h:
923 * Modules/indexeddb/IDBTransaction.cpp:
924 (WebCore::IDBTransaction::objectStore):
925 (WebCore::IDBTransaction::transitionedToFinishing):
926 (WebCore::IDBTransaction::internalAbort):
927 (WebCore::IDBTransaction::createObjectStore):
928 (WebCore::IDBTransaction::deleteObjectStore):
929 (WebCore::IDBTransaction::visitReferencedObjectStores):
930 * Modules/indexeddb/IDBTransaction.h:
931 * Modules/indexeddb/IDBTransaction.idl:
933 * bindings/js/JSIDBTransactionCustom.cpp: Added.
934 (WebCore::JSIDBTransaction::visitAdditionalChildren):
936 2016-11-09 Simon Fraser <simon.fraser@apple.com>
938 Allow customization of TextStream-based logging for geometry types
939 https://bugs.webkit.org/show_bug.cgi?id=164460
941 Reviewed by Zalan Bujtas.
943 TextStream-based logging was constrained by the requirement to maintain compatibility
944 with DRT-style output, which includes cumbersome rect logging ("at (5,0) size 40x40")
945 and dumping LayoutRects as IntRects.
947 Add some formatting flags so that other TextStream clients (e.g. logging) can have
948 more readable output, and opt into automatic FormatNumberRespectingIntegers behavior.
950 TextStreams whose output appears in test results are given flags to avoid behavior
951 changes, but in the longer term test results should be updated.
953 * html/canvas/CanvasRenderingContext2D.cpp:
954 (WebCore::CanvasRenderingContext2D::replayDisplayListAsText):
955 * page/scrolling/ScrollingStateNode.cpp:
956 (WebCore::ScrollingStateNode::scrollingStateTreeAsText):
957 * platform/graphics/FloatPoint.cpp:
958 (WebCore::operator<<):
959 * platform/graphics/FloatRect.cpp:
960 (WebCore::operator<<):
961 * platform/graphics/GraphicsLayer.cpp:
962 (WebCore::GraphicsLayer::layerTreeAsText):
963 * platform/graphics/IntRect.cpp:
964 (WebCore::operator<<):
965 * platform/graphics/LayoutPoint.cpp:
966 (WebCore::operator<<):
967 * platform/graphics/LayoutRect.cpp:
968 (WebCore::operator<<):
969 * platform/graphics/ca/GraphicsLayerCA.cpp:
970 (WebCore::GraphicsLayerCA::replayDisplayListAsText):
971 * platform/graphics/displaylists/DisplayList.cpp:
972 (WebCore::DisplayList::DisplayList::asText):
973 * platform/text/TextStream.cpp:
974 (WebCore::TextStream::operator<<):
975 * platform/text/TextStream.h:
976 (WebCore::TextStream::TextStream):
977 (WebCore::TextStream::formattingFlags):
978 (WebCore::TextStream::setFormattingFlags):
979 (WebCore::TextStream::hasFormattingFlag):
980 (WebCore::TextStream::increaseIndent):
981 (WebCore::TextStream::decreaseIndent):
982 * rendering/RenderTreeAsText.cpp:
983 (WebCore::externalRepresentation):
984 (WebCore::counterValueForElement):
986 2016-11-09 Zalan Bujtas <zalan@apple.com>
988 RenderFlowThread::flowThreadRelativeWillBeRemoved should take RenderObject& instead of RenderObject*
989 https://bugs.webkit.org/show_bug.cgi?id=164543
991 Reviewed by Simon Fraser.
993 No change in functionality.
995 * rendering/RenderBlockFlow.cpp:
996 (WebCore::RenderBlockFlow::removeChild):
997 * rendering/RenderFlowThread.h:
998 * rendering/RenderMultiColumnFlowThread.cpp:
999 (WebCore::RenderMultiColumnFlowThread::handleSpannerRemoval):
1000 (WebCore::RenderMultiColumnFlowThread::flowThreadRelativeWillBeRemoved):
1001 * rendering/RenderMultiColumnFlowThread.h:
1003 2016-11-09 Jer Noble <jer.noble@apple.com>
1005 REGRESSION (r208149): Media scrubber is not displayed in media controls
1006 https://bugs.webkit.org/show_bug.cgi?id=164514
1008 Reviewed by Darin Adler.
1010 Fixes broken Media Controls API tests.
1012 Added a new PlatformMediaSessionType; need to add that same type to the TYPE_TRAITS section of
1013 MediaElementSession.h so that is<> and downcast<> work correctly.
1015 * html/MediaElementSession.h:
1018 2016-11-09 Wenson Hsieh <wenson_hsieh@apple.com>
1020 When editing IME, `compositionend` events should fire after input events
1021 https://bugs.webkit.org/show_bug.cgi?id=164324
1022 <rdar://problem/29050438>
1024 Reviewed by Darin Adler.
1026 Moves where we dispatch `compositionend` events to after applying editing commands that fire `beforeinput` or
1027 `input` events. Also augments existing layout tests to verify the change.
1029 * editing/Editor.cpp:
1030 (WebCore::Editor::setComposition):
1032 2016-11-09 Wenson Hsieh <wenson_hsieh@apple.com>
1034 Setting foreground color when text is selected should fire an input event with color data
1035 https://bugs.webkit.org/show_bug.cgi?id=164241
1036 <rdar://problem/29032759>
1038 Reviewed by Darin Adler.
1040 Refactors Editor::applyStyle and Editor::applyParagraphStyle to handle beforeinput and input event dispatch.
1041 Instead of going through the ApplyStyleCommand to dispatch input events, override shouldDispatchInputEvents to
1042 return false. This strategy also has the effect of unifying the way input events are dispatched in applyStyle,
1043 in both codepaths where we computeAndSetTypingStyle and where we create and then apply a style command.
1045 Test: fast/events/input-events-selection-forecolor-data.html
1047 * editing/ApplyStyleCommand.h:
1048 * editing/Editor.cpp:
1049 (WebCore::inputEventDataForEditingStyleAndAction):
1050 (WebCore::Editor::applyStyle):
1051 (WebCore::Editor::applyParagraphStyle):
1052 (WebCore::Editor::computeAndSetTypingStyle):
1054 2016-11-08 Dean Jackson <dino@apple.com>
1056 Rendering support for ExtendedColors
1057 https://bugs.webkit.org/show_bug.cgi?id=164443
1058 <rdar://problems/29123243>
1060 Reviewed by Simon Fraser and Darin Adler.
1062 Add support for rendering the new color() syntax, which
1063 ends up as an ExtendedColor.
1065 In order to make rendering code a little more readable, I
1066 changed Color::hasAlpha to Color::isOpaque (since an alpha
1067 of 100% is still an alpha), and added a Color::isVisible
1068 helper (the color isn't completely transparent). These new
1069 helpers support ExtendedColor forms.
1071 Support for painting gradients and blending between colors
1072 is still to come. I also added some FIXME comments
1073 to show other places that don't handle ExtendedColors yet.
1075 Tests: css3/color/backgrounds-and-borders.html
1076 css3/color/box-shadows.html
1077 css3/color/canvas.html
1078 css3/color/composited-solid-backgrounds.html
1079 css3/color/text.html
1081 * css/CSSGradientValue.cpp: Add some notes that this is broken.
1082 (WebCore::interpolate):
1083 (WebCore::CSSGradientValue::knownToBeOpaque):
1085 * editing/EditingStyle.cpp: Use new Color helpers.
1086 (WebCore::isTransparentColorValue):
1088 * editing/mac/EditorMac.mm: Use new Color helpers.
1089 (WebCore::Editor::fontAttributesForSelectionStart):
1091 * html/ColorInputType.cpp: No need to use the Color class at all here.
1092 (WebCore::isValidSimpleColorString): Renamed from isValidColorString.
1093 (WebCore::ColorInputType::sanitizeValue):
1094 (WebCore::ColorInputType::typeMismatchFor):
1095 (WebCore::isValidColorString): Deleted.
1097 * html/canvas/CanvasRenderingContext2D.cpp: New helpers.
1098 (WebCore::CanvasRenderingContext2D::shouldDrawShadows):
1099 (WebCore::CanvasRenderingContext2D::didDraw):
1101 * page/FrameView.cpp: Ditto.
1102 (WebCore::FrameView::recalculateScrollbarOverlayStyle):
1103 (WebCore::FrameView::hasOpaqueBackground):
1104 (WebCore::FrameView::setBaseBackgroundColor):
1106 * platform/graphics/Color.cpp:
1107 (WebCore::differenceSquared): Support ExtendedColor, but also
1108 add a note to indicate that this method and its call sites
1110 (WebCore::Color::serialized): New helper.
1111 (WebCore::Color::cssText): Ditto.
1112 (WebCore::Color::blend): Ditto.
1113 (WebCore::Color::blendWithWhite):
1114 (WebCore::Color::colorWithAlphaMultipliedBy): Implementation of new function.
1115 (WebCore::Color::colorWithAlpha): Ditto.
1116 (WebCore::Color::opaqueColor): New method to return an opaque version of the given color.
1118 * platform/graphics/Color.h:
1119 (WebCore::Color::isOpaque): New helper that is !hasAlpha().
1120 (WebCore::Color::isVisible): New helper.
1121 (WebCore::Color::alphaAsFloat): Gets the alpha value as a float. This replaces
1122 a bunch of places that were calculating it manually each time. Meanwhile, we
1123 might consider always exposing the primaries as floats... or at least
1125 (WebCore::isBlackColor): New helper - it was used in a couple of places.
1126 (WebCore::isWhiteColor): Ditto.
1127 (WebCore::Color::hasAlpha): Deleted.
1129 * platform/graphics/Gradient.cpp: Add FIXME.
1130 (WebCore::Gradient::addColorStop):
1131 * platform/graphics/Gradient.h:
1133 * platform/graphics/GraphicsContext.cpp: Use new helpers.
1134 (WebCore::GraphicsContext::computeLineBoundsAndAntialiasingModeForText):
1135 * platform/graphics/GraphicsContext.h:
1136 (WebCore::GraphicsContext::hasVisibleShadow):
1138 * platform/graphics/Image.cpp: Ditto.
1139 (WebCore::Image::fillWithSolidColor):
1141 * platform/graphics/ShadowBlur.cpp: Ditto.
1142 (WebCore::ShadowBlur::updateShadowBlurValues):
1144 * platform/graphics/ca/GraphicsLayerCA.cpp: Ditto.
1145 (WebCore::GraphicsLayerCA::setContentsToSolidColor):
1147 * platform/graphics/cg/GradientCG.cpp:
1148 (WebCore::Gradient::platformGradient): Add a FIXME to note that we can
1149 add ExtendedColor support simply by using CGColors, rather than fetching
1150 the components ourselves.
1152 * platform/graphics/cg/GraphicsContextCG.cpp: New helpers.
1153 (WebCore::calculateDrawingMode):
1155 * platform/graphics/cocoa/FontCascadeCocoa.mm: New helpers.
1156 (WebCore::FontCascade::drawGlyphs):
1158 * platform/graphics/mac/ColorMac.mm: Use the new helpers and Color::hash().
1161 * platform/graphics/texmap/TextureMapperGL.cpp: New helpers.
1162 (WebCore::TextureMapperGL::drawBorder):
1164 * rendering/BorderEdge.cpp: Ditto.
1165 (WebCore::BorderEdge::obscuresBackgroundEdge):
1166 (WebCore::BorderEdge::obscuresBackground):
1168 * rendering/RenderBox.cpp: Ditto.
1169 (WebCore::RenderBox::getBackgroundPaintedExtent):
1170 (WebCore::RenderBox::backgroundIsKnownToBeOpaqueInRect):
1171 (WebCore::RenderBox::backgroundHasOpaqueTopLayer):
1173 * rendering/RenderBoxModelObject.cpp: Ditto.
1174 (WebCore::RenderBoxModelObject::paintFillLayerExtended):
1175 (WebCore::colorNeedsAntiAliasAtCorner):
1176 (WebCore::willBeOverdrawn):
1177 (WebCore::RenderBoxModelObject::paintTranslucentBorderSides):
1178 (WebCore::RenderBoxModelObject::paintBorder):
1179 (WebCore::RenderBoxModelObject::boxShadowShouldBeAppliedToBackground):
1180 (WebCore::RenderBoxModelObject::paintBoxShadow):
1181 * rendering/RenderElement.cpp:
1182 (WebCore::RenderElement::paintOutline):
1183 * rendering/RenderInline.cpp:
1184 (WebCore::RenderInline::paintOutline):
1185 * rendering/RenderLayerBacking.cpp:
1186 (WebCore::canCreateTiledImage):
1187 * rendering/RenderLayerCompositor.cpp:
1188 (WebCore::RenderLayerCompositor::viewHasTransparentBackground):
1189 * rendering/RenderMenuList.cpp:
1190 (RenderMenuList::getItemBackgroundColor):
1191 * rendering/RenderTheme.cpp:
1192 (WebCore::RenderTheme::disabledTextColor):
1193 * rendering/RenderView.cpp:
1194 (WebCore::RenderView::paintBoxDecorations):
1195 * rendering/TextDecorationPainter.cpp:
1196 (WebCore::TextDecorationPainter::paintTextDecoration):
1197 * rendering/TextPainter.cpp:
1198 (WebCore::TextPainter::paintTextWithShadows):
1199 * rendering/style/BorderValue.h:
1200 (WebCore::BorderValue::isTransparent):
1201 * rendering/style/RenderStyle.cpp:
1202 (WebCore::RenderStyle::visitedDependentColor):
1203 * rendering/style/RenderStyle.h:
1204 (WebCore::RenderStyle::hasBackground):
1205 * rendering/svg/RenderSVGResource.cpp:
1206 (WebCore::requestPaintingResource):
1207 * rendering/svg/SVGInlineTextBox.cpp:
1208 (WebCore::SVGInlineTextBox::paintSelectionBackground):
1210 * svg/SVGAnimatedColor.cpp: Add a FIXME to note this is broken.
1211 (WebCore::SVGAnimatedColorAnimator::calculateAnimatedValue):
1213 2016-11-09 Antoine Quint <graouts@apple.com>
1215 [Modern Media Controls] UI Library: StatusLabel
1216 https://bugs.webkit.org/show_bug.cgi?id=164544
1217 <rdar://problem/29179541>
1219 Reviewed by Dean Jackson.
1221 We add a new StatusLabel class to display a string of text in place of the TimeControl.
1222 A followup patch will add the logic to display "Error", "Loading" and "Live Broadcast"
1223 test under the right media state.
1225 Tests: media/modern-media-controls/macos-inline-media-controls/macos-inline-media-controls-status-label.html
1226 media/modern-media-controls/status-label/status-label.html
1228 * Modules/modern-media-controls/controls/macos-inline-media-controls.css:
1229 (.media-controls.mac.inline .time-label,):
1230 (.media-controls.mac.inline .time-label): Deleted.
1231 * Modules/modern-media-controls/controls/macos-inline-media-controls.js:
1232 (MacOSInlineMediaControls.prototype.layout):
1233 * Modules/modern-media-controls/controls/media-controls.js:
1235 * Modules/modern-media-controls/controls/status-label.css: Added.
1237 * Modules/modern-media-controls/controls/status-label.js: Added.
1238 (StatusLabel.prototype.get text):
1239 (StatusLabel.prototype.set text):
1240 (StatusLabel.prototype.commitProperty):
1241 * Modules/modern-media-controls/js-files:
1242 * WebCore.xcodeproj/project.pbxproj:
1244 2016-11-09 Chris Dumez <cdumez@apple.com>
1246 [Mac] Stop using deprecated AppKit enumeration values
1247 https://bugs.webkit.org/show_bug.cgi?id=164494
1249 Reviewed by Darin Adler.
1251 Stop using deprecated AppKit enumeration values.
1253 * editing/cocoa/HTMLConverter.mm:
1254 (HTMLConverter::computedAttributesForElement):
1255 (HTMLConverter::_processElement):
1256 (HTMLConverter::_addMarkersToList):
1257 * page/mac/EventHandlerMac.mm:
1258 (WebCore::EventHandler::keyEvent):
1259 (WebCore::lastEventIsMouseUp):
1260 (WebCore::EventHandler::passSubframeEventToSubframe):
1261 (WebCore::EventHandler::widgetDidHandleWheelEvent):
1262 (WebCore::EventHandler::sendFakeEventsAfterWidgetTracking):
1263 * page/mac/TextIndicatorWindow.mm:
1264 (WebCore::TextIndicatorWindow::setTextIndicator):
1265 * platform/graphics/mac/IconMac.mm:
1266 (WebCore::Icon::paint):
1267 * platform/mac/CursorMac.mm:
1268 (WebCore::createCustomCursor):
1269 * platform/mac/DragImageMac.mm:
1270 (WebCore::dissolveDragImageToFraction):
1271 (WebCore::createDragImageFromImage):
1272 * platform/mac/EventLoopMac.mm:
1273 (WebCore::EventLoop::cycle):
1274 * platform/mac/PasteboardMac.mm:
1275 (WebCore::Pasteboard::setDragImage):
1276 * platform/mac/PlatformEventFactoryMac.mm:
1277 (WebCore::globalPointForEvent):
1278 (WebCore::pointForEvent):
1279 (WebCore::mouseButtonForEvent):
1280 (WebCore::mouseEventTypeForEvent):
1281 (WebCore::clickCountForEvent):
1282 (WebCore::isKeypadEvent):
1283 (WebCore::windowsKeyCodeForKeyEvent):
1284 (WebCore::isKeyUpEvent):
1285 (WebCore::PlatformKeyboardEventBuilder::PlatformKeyboardEventBuilder):
1286 * platform/mac/ScrollbarThemeMac.mm:
1287 (WebCore::scrollbarControlSizeToNSControlSize):
1288 * platform/mac/ThemeMac.mm:
1289 (-[WebCoreThemeView window]):
1290 (WebCore::controlSizeForFont):
1291 (WebCore::controlSizeFromPixelSize):
1292 (WebCore::setUpButtonCell):
1293 (WebCore::stepperControlSizeForFont):
1294 (WebCore::paintStepper):
1295 (WebCore::ThemeMac::minimumControlSize):
1296 * platform/mac/WebVideoFullscreenHUDWindowController.mm:
1297 (-[WebVideoFullscreenHUDWindow initWithContentRect:styleMask:backing:defer:]):
1298 (-[WebVideoFullscreenHUDWindow performKeyEquivalent:]):
1299 (-[WebVideoFullscreenHUDWindowController init]):
1300 (-[WebVideoFullscreenHUDWindowController keyDown:]):
1301 (-[WebVideoFullscreenHUDWindowController windowDidLoad]):
1302 * platform/mac/WebWindowAnimation.mm:
1303 (WebWindowAnimationDurationFromDuration):
1304 * rendering/RenderThemeMac.mm:
1305 (WebCore::RenderThemeMac::updateCachedSystemFontDescription):
1306 (WebCore::RenderThemeMac::controlSizeForFont):
1307 (WebCore::RenderThemeMac::controlSizeForCell):
1308 (WebCore::RenderThemeMac::controlSizeForSystemFont):
1309 (WebCore::RenderThemeMac::paintProgressBar):
1310 (WebCore::RenderThemeMac::popupMenuSize):
1311 (WebCore::RenderThemeMac::sliderThumbHorizontal):
1312 (WebCore::RenderThemeMac::sliderThumbVertical):
1314 2016-11-08 Antoine Quint <graouts@apple.com>
1316 [Modern Media Controls] UI Library: iOS inline controls
1317 https://bugs.webkit.org/show_bug.cgi?id=164513
1318 <rdar://problem/27989475>
1320 Reviewed by Dean Jackson.
1322 We introduce a new IOSInlineMediaControls class which can be used to instantiate media controls
1323 for inline playback on iOS.
1325 Tests: media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-buttons-styles.html
1326 media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-constructor.html
1327 media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-controls-bar-styles.html
1328 media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-layout.html
1329 media/modern-media-controls/ios-inline-media-controls/ios-inline-media-controls-time-control-styles.html
1330 media/modern-media-controls/ios-inline-media-controls/ios-inline-media-dropping-controls.html
1332 * Modules/modern-media-controls/controls/ios-inline-media-controls.css: Added.
1333 (.media-controls.ios.inline > .controls-bar):
1334 (.media-controls.ios.inline .time-control):
1335 (.media-controls.ios.inline button):
1336 (.media-controls.ios.inline button:active):
1337 (.media-controls.ios.inline > .controls-bar button):
1338 (.media-controls.ios.inline .buttons-container.right):
1339 (.media-controls.ios.inline button.play-pause):
1340 (.media-controls.ios.inline button.skip-back):
1341 (.media-controls.ios.inline .scrubber.slider):
1342 (.media-controls.ios.inline button.airplay):
1343 (.media-controls.ios.inline button.pip):
1344 (.media-controls.ios.inline button.fullscreen):
1345 (.media-controls.ios.inline .time-label):
1346 (.media-controls.ios.inline .scrubber.slider > .fill):
1347 (.media-controls.ios.inline .scrubber.slider > input::-webkit-slider-thumb):
1348 * Modules/modern-media-controls/controls/ios-inline-media-controls.js: Added.
1349 (IOSInlineMediaControls.prototype.layout):
1350 (IOSInlineMediaControls):
1351 * Modules/modern-media-controls/images/iOS/slider-thumb@2x.png: Added.
1352 * Modules/modern-media-controls/js-files:
1353 * WebCore.xcodeproj/project.pbxproj:
1355 2016-11-09 Beth Dakin <bdakin@apple.com>
1357 Support TouchBar in WebKit
1358 https://bugs.webkit.org/show_bug.cgi?id=164437
1360 rdar://problem/28876524
1362 Reviewed by Darin Adler.
1364 * WebCore.xcodeproj/project.pbxproj:
1365 * platform/spi/cocoa/AVKitSPI.h:
1366 * platform/spi/cocoa/NSTouchBarSPI.h: Added.
1367 * platform/spi/mac/NSSpellCheckerSPI.h:
1369 2016-11-09 Chris Dumez <cdumez@apple.com>
1371 Use Blob URL instead of webkit-fake-url when pasting an image
1372 https://bugs.webkit.org/show_bug.cgi?id=49141
1374 Reviewed by Darin Adler.
1376 Use Blob URL instead of webkit-fake-url when pasting an image.
1378 Tests: editing/pasteboard/paste-image-as-blob-url.html
1379 editing/pasteboard/paste-image-using-image-data.html
1382 * editing/mac/EditorMac.mm:
1383 (WebCore::Editor::WebContentReader::readImage):
1384 (WebCore::Editor::createFragmentForImageAndURL):
1386 2016-11-09 Michael Catanzaro <mcatanzaro@igalia.com>
1388 Fix error message when SQLite initialization fails
1389 https://bugs.webkit.org/show_bug.cgi?id=164462
1391 Reviewed by Darin Adler.
1393 * platform/sql/SQLiteDatabase.cpp:
1394 (WebCore::initializeSQLiteIfNecessary):
1396 2016-11-08 Antoine Quint <graouts@apple.com>
1398 [Modern Media Controls] UI Library: macOS fullscreen controls
1399 https://bugs.webkit.org/show_bug.cgi?id=164414
1400 <rdar://problem/27989474>
1402 Reviewed by Dean Jackson.
1404 We introduce a new MacOSFullscreenMediaControls class which can be used to instantiate media controls
1405 for fullscreen playback on macOS. These controls can be dragged by the user.
1407 Tests: media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-buttons-containers-styles.html
1408 media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-buttons-styles.html
1409 media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-constructor.html
1410 media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-controls-bar-styles.html
1411 media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-right-container-margin.html
1412 media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-time-control-styles.html
1413 media/modern-media-controls/macos-fullscreen-media-controls/macos-fullscreen-media-controls-volume-styles.html
1415 * Modules/modern-media-controls/controls/button.js:
1416 (Button.prototype.set enabled):
1418 Correctly notify the layoutDelegate when the enabled property changes, regardless of whether
1421 * Modules/modern-media-controls/controls/icon-button.js:
1422 (IconButton.prototype._updateImage):
1425 Correctly notify the layout delegate when the image metrics have changed so that it may perform
1426 a layout. This issues became apparent with the new tests on Yosemite and caused some flakyness.
1428 * Modules/modern-media-controls/controls/icon-service.js:
1429 (const.iconService.new.IconService.prototype._fileNameAndPlatformForIconNameAndLayoutTraits):
1430 (const.iconService.new.IconService):
1434 * Modules/modern-media-controls/controls/macos-fullscreen-media-controls.css: Added.
1435 (.media-controls.mac.fullscreen > .controls-bar):
1436 (.media-controls.mac.fullscreen .volume.slider):
1437 (.media-controls.mac.fullscreen .buttons-container):
1438 (.media-controls.mac.fullscreen .buttons-container.center):
1439 (.media-controls.mac.fullscreen > .controls-bar button):
1440 (.media-controls.mac.fullscreen button.rewind):
1441 (.media-controls.mac.fullscreen button.play-pause):
1442 (.media-controls.mac.fullscreen button.forward):
1443 (.media-controls.mac.fullscreen .buttons-container.right):
1444 (.media-controls.mac.fullscreen button.airplay):
1445 (.media-controls.mac.fullscreen button.aspect-ratio):
1446 (.media-controls.mac.fullscreen button.pip):
1447 (.media-controls.mac.fullscreen button.tracks):
1448 (.media-controls.mac.fullscreen button.fullscreen):
1449 (.media-controls.mac.fullscreen .time-control):
1450 (.media-controls.mac.fullscreen .time-label):
1451 (.media-controls.mac.fullscreen .scrubber):
1452 * Modules/modern-media-controls/controls/macos-fullscreen-media-controls.js: Added.
1453 (MacOSFullscreenMediaControls.prototype.layout):
1454 (MacOSFullscreenMediaControls):
1455 * Modules/modern-media-controls/controls/macos-media-controls.js:
1456 (MacOSMediaControls):
1458 Allow the layoutTraits property to be set to something other than just LayoutTraits.macOS
1459 so that MacOSFullscreenMediaControls may set the LayoutTraits.Fullscreen bit.
1461 * Modules/modern-media-controls/js-files:
1463 Add a reference to the new macos-fullscreen-media-controls.js file.
1465 * WebCore.xcodeproj/project.pbxproj:
1467 Add references to the new macos-fullscreen-media-controls.js and
1468 macos-fullscreen-media-controls.css files.
1470 2016-11-09 Chris Dumez <cdumez@apple.com>
1472 Shave 16 bytes off HTMLInputElement
1473 https://bugs.webkit.org/show_bug.cgi?id=164488
1475 Reviewed by Sam Weinig.
1477 Shave 16 bytes off HTMLInputElement (232 -> 216) by packing data members
1480 * html/HTMLFormControlElement.h:
1481 * html/HTMLTextFormControlElement.cpp:
1482 (WebCore::HTMLTextFormControlElement::HTMLTextFormControlElement):
1483 * html/HTMLTextFormControlElement.h:
1485 2016-11-09 Youenn Fablet <youenn@apple.com>
1487 [WebRTC] Introduce asynchronous backend for other RTCPeerConnection API
1488 https://bugs.webkit.org/show_bug.cgi?id=164409
1490 Reviewed by Eric Carlson.
1492 Covered by existing tests.
1494 Following on createOffer changes, applying the same changes to createAnswer, setLocalDescription, setRemoteDescription and addIceCandidate.
1495 Also refactored ICE candidate event generation (done at PeerConnectionBackend).
1496 Updated stop implementation to clean any promise that may be stored in PeerConnectionBackend.
1498 The goal of this is to be more aligned with https://www.w3.org/TR/webrtc/.
1499 Implementation of the various functions such as //www.w3.org/TR/webrtc/#set-description would be done in PeerConnectionBackend.
1500 This will require additional code moved from MediaEndpointPeerConnection up to PeerConnectionBackend.
1502 * Modules/mediastream/MediaEndpointPeerConnection.cpp:
1503 (WebCore::MediaEndpointPeerConnection::createOfferTask):
1504 (WebCore::MediaEndpointPeerConnection::doCreateAnswer):
1505 (WebCore::MediaEndpointPeerConnection::createAnswerTask):
1506 (WebCore::MediaEndpointPeerConnection::doSetLocalDescription):
1507 (WebCore::MediaEndpointPeerConnection::setLocalDescriptionTask):
1508 (WebCore::MediaEndpointPeerConnection::doSetRemoteDescription):
1509 (WebCore::MediaEndpointPeerConnection::setRemoteDescriptionTask):
1510 (WebCore::MediaEndpointPeerConnection::doAddIceCandidate):
1511 (WebCore::MediaEndpointPeerConnection::addIceCandidateTask):
1512 (WebCore::MediaEndpointPeerConnection::doStop):
1513 (WebCore::MediaEndpointPeerConnection::gotIceCandidate):
1514 (WebCore::MediaEndpointPeerConnection::doneGatheringCandidates):
1515 (WebCore::MediaEndpointPeerConnection::createAnswer): Deleted.
1516 (WebCore::MediaEndpointPeerConnection::setLocalDescription): Deleted.
1517 (WebCore::MediaEndpointPeerConnection::setRemoteDescription): Deleted.
1518 (WebCore::MediaEndpointPeerConnection::addIceCandidate): Deleted.
1519 (WebCore::MediaEndpointPeerConnection::stop): Deleted.
1520 (WebCore::MediaEndpointPeerConnection::localDescriptionTypeValidForState): Deleted.
1521 (WebCore::MediaEndpointPeerConnection::remoteDescriptionTypeValidForState): Deleted.
1522 * Modules/mediastream/MediaEndpointPeerConnection.h:
1523 * Modules/mediastream/PeerConnectionBackend.cpp:
1524 (WebCore::PeerConnectionBackend::createOffer):
1525 (WebCore::PeerConnectionBackend::createOfferFailed):
1526 (WebCore::PeerConnectionBackend::createAnswer):
1527 (WebCore::PeerConnectionBackend::createAnswerSucceeded):
1528 (WebCore::PeerConnectionBackend::createAnswerFailed):
1529 (WebCore::isLocalDescriptionTypeValidForState):
1530 (WebCore::PeerConnectionBackend::setLocalDescription):
1531 (WebCore::PeerConnectionBackend::setLocalDescriptionSucceeded):
1532 (WebCore::PeerConnectionBackend::setLocalDescriptionFailed):
1533 (WebCore::isRemoteDescriptionTypeValidForState):
1534 (WebCore::PeerConnectionBackend::setRemoteDescription):
1535 (WebCore::PeerConnectionBackend::setRemoteDescriptionSucceeded):
1536 (WebCore::PeerConnectionBackend::setRemoteDescriptionFailed):
1537 (WebCore::PeerConnectionBackend::addIceCandidate):
1538 (WebCore::PeerConnectionBackend::addIceCandidateSucceeded):
1539 (WebCore::PeerConnectionBackend::addIceCandidateFailed):
1540 (WebCore::PeerConnectionBackend::fireICECandidateEvent):
1541 (WebCore::PeerConnectionBackend::doneGatheringCandidates):
1542 (WebCore::PeerConnectionBackend::stop):
1543 * Modules/mediastream/PeerConnectionBackend.h:
1545 2016-11-09 Eric Carlson <eric.carlson@apple.com>
1547 [MediaStream][Mac] Mark captured video frames as ready for display immediately
1548 https://bugs.webkit.org/show_bug.cgi?id=164482
1549 <rdar://problem/29139073>
1551 Reviewed by Jer Noble.
1553 * platform/cf/CoreMediaSoftLink.cpp: Add new constant.
1554 * platform/cf/CoreMediaSoftLink.h:
1556 * platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaStreamAVFObjC.h:
1557 * platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaStreamAVFObjC.mm:
1558 (WebCore::MediaPlayerPrivateMediaStreamAVFObjC::requestNotificationWhenReadyForMediaData):
1559 New, ask register for a callback when the sample buffer display layer is ready
1560 for more media data.
1561 (WebCore::MediaPlayerPrivateMediaStreamAVFObjC::enqueueVideoSampleBuffer): Don't change
1562 the sample timestamps, assume the caller has configured the sample correctly.
1563 (WebCore::MediaPlayerPrivateMediaStreamAVFObjC::prepareVideoSampleBufferFromTrack): Don't
1564 drop frames when the display layer isn't ready.
1565 (WebCore::MediaPlayerPrivateMediaStreamAVFObjC::destroyLayer): Call stopRequestingMediaData.
1566 (WebCore::MediaPlayerPrivateMediaStreamAVFObjC::sampleBufferUpdated):
1567 (WebCore::MediaPlayerPrivateMediaStreamAVFObjC::enqueueVideoSampleBufferFromTrack): Deleted.
1569 * platform/mediastream/mac/AVVideoCaptureSource.mm:
1570 (WebCore::AVVideoCaptureSource::setupCaptureSession): Tell the video output to always discard
1571 late video frames, we don't need them.
1572 (WebCore::AVVideoCaptureSource::processNewFrame): Add a kCMSampleAttachmentKey_DisplayImmediately
1575 2016-11-09 Joanmarie Diggs <jdiggs@igalia.com>
1577 AX: [ATK] Wrong selected element at a given index in a list box (redux)
1578 https://bugs.webkit.org/show_bug.cgi?id=164430
1580 Reviewed by Darin Adler.
1582 This essentially undoes the implementation change resulting from r164577.
1583 As stated in the ATK documentation, atk_selection_ref_selection() takes
1584 "a gint specifying the index in the selection set. (e.g. the ith selection
1585 as opposed to the ith child)." r164577 deliberately modified that, causing
1586 atk_selection_ref_selection() to treat the index as if it were the position
1587 with respect to all of the children. There is different API in ATK, namely
1588 atk_object_ref_accessible_child(), when the ith child from the set of all
1591 Tests: accessibility/aria-listbox-no-selection.html
1592 accessibility/native-listbox-no-selection.html
1594 * accessibility/atk/WebKitAccessibleInterfaceSelection.cpp:
1595 (optionFromSelection):
1597 2016-11-09 Gyuyoung Kim <gyuyoung.kim@webkit.org>
1599 [EFL] Use libgcrypt instead of GnuTLS for CryptoDigest
1600 https://bugs.webkit.org/show_bug.cgi?id=164461
1602 Reviewed by Michael Catanzaro.
1604 As GTK port in r208297, EFL port starts to use libgcrypt instead of GnuTLS as well.
1606 No new tests, no behavior change.
1608 * PlatformEfl.cmake:
1610 2016-11-09 Ryan Haddad <ryanhaddad@apple.com>
1612 Unreviewed, rolling out r208422.
1614 Roll r208382 back in since it was not responsible for the API
1615 test failures seen on macOS.
1619 "Unreviewed, rolling out r208382."
1620 https://bugs.webkit.org/show_bug.cgi?id=164319
1621 http://trac.webkit.org/changeset/208422
1623 2016-11-09 Csaba Osztrogonác <ossy@webkit.org>
1625 One more URTBF after r208361.
1627 * PlatformMac.cmake:
1629 2016-11-09 Csaba Osztrogonác <ossy@webkit.org>
1631 Typo fix after r162782
1632 https://bugs.webkit.org/show_bug.cgi?id=164473
1634 Unreviewed trivial fix.
1636 * platform/ThreadGlobalData.cpp:
1638 2016-11-07 Ryan Haddad <ryanhaddad@apple.com>
1640 Unreviewed, rolling out r208382.
1642 This change appears to have caused 3
1643 SerializedCryptoKeyWrapTest API tests to fail on macOS.
1647 "[Readable Streams API] Implement ByteStreamController
1649 https://bugs.webkit.org/show_bug.cgi?id=164319
1650 http://trac.webkit.org/changeset/208382
1652 2016-11-04 Filip Pizlo <fpizlo@apple.com>
1654 WTF::ParkingLot should stop using std::chrono because std::chrono::duration casts are prone to overflows
1655 https://bugs.webkit.org/show_bug.cgi?id=152045
1657 Reviewed by Andy Estes.
1659 No new layout tests because no new behavior. The new WTF time classes have some unit tests
1662 * fileapi/ThreadableBlobRegistry.cpp:
1663 (WebCore::ThreadableBlobRegistry::blobSize):
1664 * platform/MainThreadSharedTimer.h:
1665 * platform/SharedTimer.h:
1666 * platform/ThreadTimers.cpp:
1667 (WebCore::ThreadTimers::updateSharedTimer):
1668 * platform/cf/MainThreadSharedTimerCF.cpp:
1669 (WebCore::MainThreadSharedTimer::setFireInterval):
1670 * platform/efl/MainThreadSharedTimerEfl.cpp:
1671 (WebCore::MainThreadSharedTimer::setFireInterval):
1672 * platform/glib/MainThreadSharedTimerGLib.cpp:
1673 (WebCore::MainThreadSharedTimer::setFireInterval):
1674 * platform/win/MainThreadSharedTimerWin.cpp:
1675 (WebCore::MainThreadSharedTimer::setFireInterval):
1676 * workers/WorkerRunLoop.cpp:
1677 (WebCore::WorkerRunLoop::runInMode):
1679 2016-11-04 Zalan Bujtas <zalan@apple.com>
1681 RenderFlowThread state reset cleanup.
1682 https://bugs.webkit.org/show_bug.cgi?id=164426
1684 Reviewed by Simon Fraser.
1686 RenderFlowThread state reset is spread across several functions. This patch groups them
1687 together in RenderObject::resetFlowThreadState().
1689 No change in functionality.
1691 * rendering/RenderBlock.cpp:
1692 (WebCore::RenderBlock::removeLeftoverAnonymousBlock):
1693 (WebCore::RenderBlock::dropAnonymousBoxChild): This is now part of resetFlowThreadState() since resetFlowThreadState
1694 gets called even when NotifyChildren is false.
1695 * rendering/RenderElement.cpp:
1696 (WebCore::RenderElement::insertChildInternal): Initialize the thread state before we notify the child.
1697 (WebCore::RenderElement::removeChildInternal): Reset the state even when NotifyChildren is false.
1698 (WebCore::RenderElement::willBeRemovedFromTree): This code is moved to removeFromRenderFlowThread().
1699 (WebCore::RenderElement::removeFromRenderFlowThread):
1700 * rendering/RenderObject.cpp:
1701 (WebCore::RenderObject::initializeFlowThreadState): This is in transition for webkit.org/b/164428 (RenderFlowThread state initialization cleanup.)
1702 (WebCore::RenderObject::resetFlowThreadState):
1703 (WebCore::RenderObject::setParent): This was seemingly a random place to put flow state initialization.
1704 (WebCore::RenderObject::willBeRemovedFromTree): resetFlowThreadState() takes care of it now.
1705 * rendering/RenderObject.h:
1707 2016-11-04 Yusuke Suzuki <utatane.tea@gmail.com>
1709 [DOMJIT] Add DOMJIT::Signature annotation to Document::getElementById
1710 https://bugs.webkit.org/show_bug.cgi?id=164356
1712 Reviewed by Filip Pizlo.
1714 This patch implements DOMJIT::Signature annotation for getElementById.
1715 Since getElementById is also implemented in DocumentFragment, we implement
1716 the branchIfDocumentFragment/branchIfNotDocumentFragment for that.
1718 In dromaeo, we have a test like this.
1720 test( "getElementById", function(){
1721 for ( var i = 0; i < num * 30; i++ ) {
1722 ret = document.getElementById("testA" + num).nodeType;
1723 ret = document.getElementById("testB" + num).nodeType;
1724 ret = document.getElementById("testC" + num).nodeType;
1725 ret = document.getElementById("testD" + num).nodeType;
1726 ret = document.getElementById("testE" + num).nodeType;
1727 ret = document.getElementById("testF" + num).nodeType;
1731 In the above test, JSC already knows the following things.
1733 1. Since nodeType is now handled as CallDOMGetter, we know that it is pure.
1734 2. getElementById look up becomes PureGetById since document is impure object. But it is kept as PureGetById. So it does not write DOMState.
1735 3. `"testA" + num` will be converted to constant string.
1736 4. CallDOM for getElementById said it just reads(DOMState:DOM). And it saids that it returns the same value as long as DOMState is not clobbered.
1737 5. CheckCell leading CallDOM ensures the inlined getElementById node. (CallDOM node).
1739 The key thing is that no node clobbers DOMState during the loop. So CallDOM & CallDOMGetter can be hoisted.
1740 This improves dom-query significantly. Dromaeo dom-query getElementById becomes 40x faster (247796 v.s. 6197).
1741 Dromaeo dom-query getElementById (not in document) becomes 89x faster (630317.8 v.s. 7066.).
1743 Tests: js/dom/domjit-function-get-element-by-id-changed.html
1744 js/dom/domjit-function-get-element-by-id-licm.html
1745 js/dom/domjit-function-get-element-by-id.html
1747 * dom/NonElementParentNode.idl:
1748 * domjit/DOMJITCheckDOM.h:
1749 (WebCore::DOMJIT::TypeChecker<DocumentFragment>::branchIfFail):
1750 * domjit/DOMJITHelpers.h:
1751 (WebCore::DOMJIT::branchIfDocumentFragment):
1752 (WebCore::DOMJIT::branchIfNotDocumentFragment):
1754 2016-11-04 Simon Fraser <simon.fraser@apple.com>
1756 Rename unscaledUnobscuredVisibleContentSize and unscaledVisibleContentSizeIncludingObscuredArea for attempted clarity
1757 https://bugs.webkit.org/show_bug.cgi?id=164438
1759 Reviewed by Tim Horton.
1761 unscaledUnobscuredVisibleContentSize() and unscaledVisibleContentSizeIncludingObscuredArea() were an endless source
1764 Functions with "VisibleContent" in the name are usually expected to return document coordinates (affected by zooming),
1765 so unscaledUnobscuredVisibleContentSize caused cognitive dissonance because of "unscaled" vs "visibleContent", and
1766 "unobscured" vs "visible".
1769 unscaledUnobscuredVisibleContentSize -> sizeForUnobscuredContent
1770 unscaledVisibleContentSizeIncludingObscuredArea -> sizeForVisibleContent
1772 sizeForUnobscuredContent() can also be private to ScrollView.
1774 * inspector/InspectorOverlay.cpp:
1775 (WebCore::InspectorOverlay::update):
1776 * platform/ScrollView.cpp:
1777 (WebCore::ScrollView::unobscuredContentRectInternal):
1778 (WebCore::ScrollView::sizeForVisibleContent):
1779 (WebCore::ScrollView::sizeForUnobscuredContent): Don't compute unscaledVisibleContentSizeIncludingObscuredArea
1780 before testing whether we have a platform widget.
1781 (WebCore::ScrollView::layoutSize):
1782 (WebCore::ScrollView::unscaledVisibleContentSizeIncludingObscuredArea): Deleted.
1783 (WebCore::ScrollView::unscaledUnobscuredVisibleContentSize): Deleted.
1784 * platform/ScrollView.h:
1785 * rendering/RenderBoxModelObject.cpp:
1786 (WebCore::RenderBoxModelObject::calculateBackgroundImageGeometry):
1787 * rendering/RenderLayerCompositor.cpp:
1788 (WebCore::RenderLayerCompositor::flushPendingLayerChanges):
1789 (WebCore::RenderLayerCompositor::frameViewDidChangeSize):
1790 (WebCore::RenderLayerCompositor::updateRootLayerPosition):
1791 (WebCore::RenderLayerCompositor::ensureRootLayer):
1793 2016-11-04 Simon Fraser <simon.fraser@apple.com>
1795 Layout viewport wrong with RTL documents
1796 https://bugs.webkit.org/show_bug.cgi?id=164434
1798 Reviewed by Tim Horton.
1800 The layoutViewportRect was computed incorrectly in RTL documents, because
1801 FrameView::unscaledMaximumScrollPosition() was wrong; it erroneously mapped
1802 what it thought was a scrollOffset to a scrollPosition.
1804 Unscaled scroll positions are in the same coordinate space as unscaledDocumentRect,
1805 so we should not call scrollPositionFromOffset() in FrameView::unscaledMaximumScrollPosition().
1807 Changed FrameView::unscaledMinimumScrollPosition() to also just grab the location of
1808 unscaledDocumentRect, for symmetry.
1810 Finally fixed the tiled scrolling indicator's viewport rect for zoomed RTL documents
1811 by using the unscaled scroll origin.
1813 Tests: fast/visual-viewport/rtl-nonzoomed-rects.html
1814 fast/visual-viewport/rtl-zoomed-rects.html
1816 * page/FrameView.cpp:
1817 (WebCore::FrameView::setLayoutViewportOrigin):
1818 (WebCore::FrameView::unscaledScrollOrigin):
1819 (WebCore::FrameView::unscaledMinimumScrollPosition):
1820 (WebCore::FrameView::unscaledMaximumScrollPosition):
1823 2016-11-04 Sam Weinig <sam@webkit.org>
1825 [WebIDL] Add support for modern callback syntax
1826 https://bugs.webkit.org/show_bug.cgi?id=164435
1828 Reviewed by Chris Dumez.
1830 Support new callback syntax:
1831 callback Function = void (DOMString arg1, long arg2);
1833 This replaces "callback interface" types with a Callback=FunctionOnly
1836 * Modules/geolocation/PositionCallback.idl:
1837 * Modules/geolocation/PositionErrorCallback.idl:
1838 * Modules/notifications/NotificationPermissionCallback.idl:
1839 * Modules/quota/StorageErrorCallback.idl:
1840 * Modules/quota/StorageQuotaCallback.idl:
1841 * Modules/quota/StorageUsageCallback.idl:
1842 * Modules/webaudio/AudioBufferCallback.idl:
1843 * Modules/webdatabase/DatabaseCallback.idl:
1844 * Modules/webdatabase/SQLStatementCallback.idl:
1845 * Modules/webdatabase/SQLStatementErrorCallback.idl:
1846 * Modules/webdatabase/SQLTransactionCallback.idl:
1847 * Modules/webdatabase/SQLTransactionErrorCallback.idl:
1848 * dom/RequestAnimationFrameCallback.idl:
1849 * dom/StringCallback.idl:
1850 * html/VoidCallback.idl:
1851 * page/IntersectionObserverCallback.idl:
1852 * css/MediaQueryListListener.idl:
1853 Update to new syntax.
1855 * css/MediaQueryListListener.h:
1856 * css/MediaQueryMatcher.cpp:
1857 (WebCore::MediaQueryMatcher::styleResolverChanged):
1858 Switch to using the now required 'handleEvent' name. This is an implementation detail
1859 that we should change.
1861 * bindings/scripts/CodeGenerator.pm:
1862 Update document processing to allow a callback only file. Update callback
1863 type checks to look for a regex that matches in the new format.
1865 * bindings/scripts/CodeGeneratorJS.pm:
1866 (AddToImplIncludesForIDLType):
1867 (AddToIncludesForIDLType):
1868 (AddToImplIncludes):
1870 Abstract includes functions to allow passing in an include hash.
1872 (GenerateCallbackFunctionHeader):
1873 (GenerateCallbackFunctionImplementation):
1874 (GenerateCallbackInterfaceHeader):
1875 (GenerateCallbackInterfaceImplementation):
1876 (GenerateCallbackHeaderContent):
1877 (GenerateCallbackImplementationContent):
1878 Refactor callback generation code into GenerateCallbackHeaderContent and GenerateCallbackImplementationContent
1879 to allow using it for both the new callbacks as well as the old callback interfaces.
1881 * bindings/scripts/IDLParser.pm:
1884 (applyTypedefsToOperation):
1885 (parseCallbackRest):
1886 Parse callbacks into the new IDLCallbackFunction type. Ensure that typedefs are applied as well.
1888 * bindings/scripts/IDLAttributes.txt:
1889 Remove support for Callback=FunctionOnly.
1891 * bindings/scripts/test/JS/JSTestCallback.cpp: Removed.
1892 * bindings/scripts/test/JS/JSTestCallback.h: Removed.
1893 * bindings/scripts/test/JS/JSTestCallbackFunction.cpp:
1894 * bindings/scripts/test/JS/JSTestCallbackFunction.h:
1895 * bindings/scripts/test/JS/JSTestCallbackFunctionWithTypedefs.cpp: Added.
1896 * bindings/scripts/test/JS/JSTestCallbackFunctionWithTypedefs.h: Added.
1897 * bindings/scripts/test/JS/JSTestCallbackInterface.cpp: Copied from Source/WebCore/bindings/scripts/test/JS/JSTestCallback.cpp.
1898 * bindings/scripts/test/JS/JSTestCallbackInterface.h: Copied from Source/WebCore/bindings/scripts/test/JS/JSTestCallback.h.
1899 * bindings/scripts/test/JS/JSTestObj.cpp:
1900 * bindings/scripts/test/JS/JSTestTypedefs.cpp:
1901 * bindings/scripts/test/TestCallback.idl: Removed.
1902 * bindings/scripts/test/TestCallbackFunction.idl:
1903 * bindings/scripts/test/TestCallbackFunctionWithTypedefs.idl: Added.
1904 * bindings/scripts/test/TestCallbackInterface.idl: Copied from Source/WebCore/bindings/scripts/test/TestCallback.idl.
1905 * bindings/scripts/test/TestObj.idl:
1906 * bindings/scripts/test/TestTypedefs.idl:
1907 Update existing tests and add new ones to test callback functions specifically.
1909 2016-11-04 Alex Christensen <achristensen@webkit.org>
1911 Move isDefaultPortForProtocol from URLParser.cpp back to URL.cpp
1912 https://bugs.webkit.org/show_bug.cgi?id=164439
1914 Reviewed by Daniel Bates.
1916 No change in behaviour.
1919 (WebCore::defaultPortForProtocol):
1920 (WebCore::isDefaultPortForProtocol):
1921 * platform/URLParser.cpp:
1922 (WebCore::URLParser::defaultPortForProtocol):
1923 (WebCore::URLParser::parsePort):
1924 (WebCore::defaultPortForProtocol): Deleted.
1925 (WebCore::isDefaultPortForProtocol): Deleted.
1926 * platform/URLParser.h:
1928 2016-11-04 Wenson Hsieh <wenson_hsieh@apple.com>
1930 Safari does not emit composition end if blurred for dead key / Japanese IME
1931 https://bugs.webkit.org/show_bug.cgi?id=164369
1932 <rdar://problem/29050439>
1934 Reviewed by Ryosuke Niwa.
1936 On Mac, _before_ changing selection, try to finalize the composition by calling Editor::cancelComposition early.
1937 This is because the focused element may have changed after performing the selection change, so we would
1938 otherwise be dispatching the `compositionend` to the new focused element (or no compositionend at all) instead
1939 of the element with the composition.
1941 Doing this allows us to match Chrome and Firefox behavior. After canceling the composition, we then need to also
1942 clear the system IME state. We do this on Mac WK1/WK2 through the cancelComposition() codepath, which ends up
1943 calling into -discardMarkedText, which resets the marked text state. Some minor refactoring was performed to
1944 accomplish this -- currently, discardedComposition sends a CompositionWasCanceled message over to the UI process
1945 that discards the marked text, and then updates the editor state. This patch splits this into two separate
1946 steps -- see the WebKit2 ChangeLog for more details.
1948 Test: fast/events/ime-compositionend-on-selection-change.html
1950 * editing/Editor.cpp:
1951 (WebCore::Editor::selectionWillChange):
1953 * editing/FrameSelection.cpp:
1954 (WebCore::FrameSelection::setSelectionWithoutUpdatingAppearance):
1955 * editing/mac/EditorMac.mm:
1956 (WebCore::Editor::selectionWillChange):
1957 * loader/EmptyClients.h:
1958 * page/EditorClient.h:
1960 2016-11-04 Brady Eidson <beidson@apple.com>
1962 IndexedDB 2.0: Clean up more transaction abort and exception throwing behavior from IDBObjectStore.
1963 https://bugs.webkit.org/show_bug.cgi?id=164424
1965 Reviewed by Alex Christensen.
1967 No new tests (Covered by existing tests).
1969 This patch actually turns a handful of PASS to FAIL in the imported tests, but those are parts of the
1970 spec in flux/under discussion.
1972 We'll update either source or the tests as things are resolved.
1974 * Modules/indexeddb/IDBIndex.cpp:
1975 (WebCore::IDBIndex::rollbackInfoForVersionChangeAbort):
1977 * Modules/indexeddb/IDBObjectStore.cpp:
1978 (WebCore::IDBObjectStore::get):
1979 (WebCore::IDBObjectStore::putOrAdd):
1980 (WebCore::IDBObjectStore::rollbackForVersionChangeAbort):
1982 2016-11-04 Ryosuke Niwa <rniwa@webkit.org>
1984 Load stylesheets in link elements inside a connected shadow tree
1985 https://bugs.webkit.org/show_bug.cgi?id=160683
1986 <rdar://problem/29040652>
1988 Reviewed by Antti Koivisto.
1990 Allow external stylesheets within a shadow tree by storing the appropriate style scope in HTMLLinkElement
1991 when it's connected to a document instead of always talking to document's style scope.
1993 Tests: fast/shadow-dom/link-element-in-shadow-tree.html
1994 fast/shadow-dom/selected-stylesheet-in-shadow-tree.html
1996 * html/HTMLLinkElement.cpp:
1997 (WebCore::HTMLLinkElement::HTMLLinkElement):
1998 (WebCore::HTMLLinkElement::~HTMLLinkElement):
1999 (WebCore::HTMLLinkElement::setDisabledState): Exit early when the element is not in a document as invoking
2000 didChangeActiveStyleSheetCandidates would require having a valid m_styleScope and process() already exits
2001 early when inDocument() is false.
2002 (WebCore::HTMLLinkElement::parseAttribute):
2003 (WebCore::HTMLLinkElement::process): Removed the early exit for when the element is in a shadow tree.
2004 (WebCore::HTMLLinkElement::insertedInto): Exit early unless this element has just become connected to
2005 a document instead of whenever its self-inclusive ancestor is inserted into a container.
2006 (WebCore::HTMLLinkElement::removedFrom): Ditto for removal. Also call removeStyleSheetCandidateNode after
2007 calling removePendingSheet since the latter depends on m_styleScope being not null.
2008 (WebCore::HTMLLinkElement::addPendingSheet):
2009 (WebCore::HTMLLinkElement::removePendingSheet):
2010 * html/HTMLLinkElement.h:
2011 * html/HTMLStyleElement.cpp:
2012 (WebCore::HTMLStyleElement::insertedInto): Only call inline style owner's insertedIntoDocument if this
2013 element has just become connected to a document.
2014 (WebCore::HTMLStyleElement::removedFrom): Ditto for the removal.
2015 * style/StyleScope.h:
2016 * svg/SVGStyleElement.cpp:
2017 (WebCore::SVGStyleElement::insertedInto): Ditto.
2018 (WebCore::SVGStyleElement::removedFrom): Ditto for the removal.
2020 2016-11-04 Said Abou-Hallawa <sabouhallawa@apple.com>
2022 Add a setting and preferences to enable/disable async image decoding
2023 https://bugs.webkit.org/show_bug.cgi?id=164417
2025 Reviewed by Simon Fraser.
2027 Add an asyncImageDecodingEnabled setting. This setting controls whether an
2028 image "can" be asynchronously decoded on a separate thread or not. The
2029 function ImageSource::isAsyncDecodingRequired() will be used in conjunction
2030 with this setting to decide whether an image "should" be asynchronously
2035 2016-11-04 Tim Horton <timothy_horton@apple.com>
2037 Apply post-landing review comments for r208347
2040 (WebCore::Element::findAnchorElementForLink):
2041 Use attributeWithoutSynchronization.
2043 * page/PrintContext.cpp:
2044 (WebCore::PrintContext::spoolPage):
2045 (WebCore::PrintContext::spoolRect):
2046 (WebCore::PrintContext::collectLinkedDestinations):
2047 (WebCore::PrintContext::outputLinkedDestinations):
2048 * page/PrintContext.h:
2049 Pass Document by reference instead of Node by pointer,
2050 use ElementTraversal instead of NodeTraversal to avoid
2051 having to locally check the type, and null-check renderers.
2053 2016-11-04 Myles C. Maxfield <mmaxfield@apple.com>
2055 Implement WebGL2RenderingContext::copyBufferSubData()
2056 https://bugs.webkit.org/show_bug.cgi?id=164008
2058 Reviewed by Dean Jackson.
2060 Similar to previous work regarding WebGL 2 buffers, this method implements
2061 the ability to copy from one buffer to another without the data leaving
2064 Test: fast/canvas/webgl/copyBufferSubData.html
2066 * html/canvas/WebGL2RenderingContext.cpp:
2067 (WebCore::WebGL2RenderingContext::copyBufferSubData):
2068 * html/canvas/WebGLBuffer.cpp:
2069 (WebCore::WebGLBuffer::associateCopyBufferSubData):
2070 * html/canvas/WebGLBuffer.h:
2071 * platform/graphics/GraphicsContext3D.h:
2072 * platform/graphics/opengl/GraphicsContext3DOpenGLCommon.cpp:
2073 (WebCore::GraphicsContext3D::copyBufferSubData):
2075 2016-11-04 Simon Fraser <simon.fraser@apple.com>
2077 Rename some ScrollingTree/Node-related functions to reduce the number of uses of "update"
2078 https://bugs.webkit.org/show_bug.cgi?id=164420
2080 Reviewed by Tim Horton.
2083 commitNewTreeState -> commitTreeState ("new" was redundant)
2084 updateBeforeChildren -> commitStateBeforeChildren
2085 updateAfterChildren -> commitStateAfterChildren
2087 * page/scrolling/ScrollingTree.cpp:
2088 (WebCore::ScrollingTree::commitTreeState):
2089 (WebCore::ScrollingTree::updateTreeFromStateNode):
2090 (WebCore::ScrollingTree::commitNewTreeState): Deleted.
2091 * page/scrolling/ScrollingTree.h:
2092 * page/scrolling/ScrollingTreeFrameScrollingNode.cpp:
2093 (WebCore::ScrollingTreeFrameScrollingNode::commitStateBeforeChildren):
2094 (WebCore::ScrollingTreeFrameScrollingNode::updateBeforeChildren): Deleted.
2095 * page/scrolling/ScrollingTreeFrameScrollingNode.h:
2096 * page/scrolling/ScrollingTreeNode.h:
2097 (WebCore::ScrollingTreeNode::commitStateAfterChildren):
2098 (WebCore::ScrollingTreeNode::updateAfterChildren): Deleted.
2099 * page/scrolling/ScrollingTreeScrollingNode.cpp:
2100 (WebCore::ScrollingTreeScrollingNode::commitStateBeforeChildren):
2101 (WebCore::ScrollingTreeScrollingNode::commitStateAfterChildren):
2102 (WebCore::ScrollingTreeScrollingNode::updateBeforeChildren): Deleted.
2103 (WebCore::ScrollingTreeScrollingNode::updateAfterChildren): Deleted.
2104 * page/scrolling/ScrollingTreeScrollingNode.h:
2105 * page/scrolling/ThreadedScrollingTree.cpp:
2106 (WebCore::ThreadedScrollingTree::commitTreeState):
2107 (WebCore::ThreadedScrollingTree::commitNewTreeState): Deleted.
2108 * page/scrolling/ThreadedScrollingTree.h:
2109 * page/scrolling/ios/ScrollingTreeFrameScrollingNodeIOS.h:
2110 * page/scrolling/ios/ScrollingTreeFrameScrollingNodeIOS.mm:
2111 (WebCore::ScrollingTreeFrameScrollingNodeIOS::commitStateBeforeChildren):
2112 (WebCore::ScrollingTreeFrameScrollingNodeIOS::commitStateAfterChildren):
2113 (WebCore::ScrollingTreeFrameScrollingNodeIOS::updateBeforeChildren): Deleted.
2114 (WebCore::ScrollingTreeFrameScrollingNodeIOS::updateAfterChildren): Deleted.
2115 * page/scrolling/ios/ScrollingTreeIOS.cpp:
2116 (WebCore::ScrollingTreeIOS::commitNewTreeState): Deleted.
2117 * page/scrolling/ios/ScrollingTreeIOS.h:
2118 * page/scrolling/mac/ScrollingCoordinatorMac.mm:
2119 (WebCore::ScrollingCoordinatorMac::commitTreeState):
2120 * page/scrolling/mac/ScrollingTreeFixedNode.h:
2121 * page/scrolling/mac/ScrollingTreeFixedNode.mm:
2122 (WebCore::ScrollingTreeFixedNode::commitStateBeforeChildren):
2123 (WebCore::ScrollingTreeFixedNode::updateBeforeChildren): Deleted.
2124 * page/scrolling/mac/ScrollingTreeFrameScrollingNodeMac.h:
2125 * page/scrolling/mac/ScrollingTreeFrameScrollingNodeMac.mm:
2126 (WebCore::ScrollingTreeFrameScrollingNodeMac::commitStateBeforeChildren):
2127 (WebCore::ScrollingTreeFrameScrollingNodeMac::commitStateAfterChildren):
2128 (WebCore::ScrollingTreeFrameScrollingNodeMac::updateBeforeChildren): Deleted.
2129 (WebCore::ScrollingTreeFrameScrollingNodeMac::updateAfterChildren): Deleted.
2130 * page/scrolling/mac/ScrollingTreeStickyNode.h:
2131 * page/scrolling/mac/ScrollingTreeStickyNode.mm:
2132 (WebCore::ScrollingTreeStickyNode::commitStateBeforeChildren):
2133 (WebCore::ScrollingTreeStickyNode::updateBeforeChildren): Deleted.
2135 2016-11-03 Anders Carlsson <andersca@apple.com>
2137 Add new 'other' Apple Pay button style
2138 https://bugs.webkit.org/show_bug.cgi?id=164384
2139 rdar://problem/28302528
2141 Reviewed by Dean Jackson.
2143 * DerivedSources.make:
2145 Add extension points.
2147 * css/CSSPrimitiveValueMappings.h:
2148 (WebCore::CSSPrimitiveValue::CSSPrimitiveValue):
2149 Add ApplePayButtonType::Other.
2151 (WebCore::CSSPrimitiveValue::operator ApplePayButtonType):
2154 * css/CSSValueKeywords.in:
2157 * css/parser/CSSParser.cpp:
2158 (WebCore::isValidKeywordPropertyAndValue):
2161 * css/parser/CSSParserFastPaths.cpp:
2162 (WebCore::CSSParserFastPaths::isValidKeywordPropertyAndValue):
2165 * rendering/RenderThemeCocoa.mm:
2166 (WebCore::toPKPaymentButtonType):
2167 Handle ApplePayButtonType::Other.
2169 * rendering/style/RenderStyleConstants.h:
2170 Add ApplePayButtonType::Other.
2172 2016-11-04 Antti Koivisto <antti@apple.com>
2174 slotted() pseudo does not work with ID selector
2175 https://bugs.webkit.org/show_bug.cgi?id=160538
2176 <rdar://problem/28534529>
2178 Reviewed by Andreas Kling.
2180 When we saw an id selector while addin rules we immediately threw it into the m_idRules
2181 optimization bucket and bailed out. However selectors containing ::slotted must always end
2182 up in m_slottedPseudoElementRules list no matter what else is there.
2184 Fix by treating id like other selectors and only choosing the bucket after analysing all
2185 the selector components.
2187 Test: fast/shadow-dom/css-scoping-slot-with-id.html
2190 (WebCore::RuleSet::addRule): Also made this use switch instead of a series of ifs.
2192 2016-11-04 Brady Eidson <beidson@apple.com>
2194 IndexedDB 2.0: Handle IDBObjectStore rename behavior properly when version change transaction aborts.
2195 https://bugs.webkit.org/show_bug.cgi?id=164416
2197 Reviewed by Beth Dakin.
2199 No new tests (Covered by existing tests).
2201 * Modules/indexeddb/IDBObjectStore.cpp:
2202 (WebCore::IDBObjectStore::rollbackForVersionChangeAbort):
2204 2016-11-04 Keith Rollin <krollin@apple.com>
2206 NetworkSession: Add NetworkDataTask implementation for blobs
2207 https://bugs.webkit.org/show_bug.cgi?id=163939
2209 Reviewed by Alex Christensen.
2211 * WebCore.xcodeproj/project.pbxproj: Mark HTTPParsers.h and AsyncFileStream.h as private.
2212 * fileapi/AsyncFileStream.h: Add WEBCORE_EXPORT to AsyncFileStream class.
2213 * platform/network/BlobData.h: Add WEBCORE_EXPORT to length().
2214 * platform/network/HTTPParsers.h: Add WEBCORE_EXPORT to parseRange().
2215 * platform/network/ResourceResponseBase.h: Add WEBCORE_EXPORT to setHTTPHeaderField().
2217 2016-11-04 Brady Eidson <beidson@apple.com>
2219 IndexedDB 2.0: Use IDB-specific exceptions in places where the generic exceptions are currently used.
2220 https://bugs.webkit.org/show_bug.cgi?id=164406
2222 Reviewed by Alex Christensen.
2224 No new tests (Covered by existing tests).
2226 * Modules/indexeddb/IDBIndex.cpp:
2227 (WebCore::IDBIndex::setName):
2229 * Modules/indexeddb/IDBObjectStore.cpp:
2230 (WebCore::IDBObjectStore::setName):
2232 2016-11-04 Joanmarie Diggs <jdiggs@igalia.com>
2234 AX: [ATK] Attempting to clear selection on ARIA listboxes results in crash
2235 https://bugs.webkit.org/show_bug.cgi?id=164331
2237 Reviewed by Chris Fleizach.
2239 The ATK code is using is<AccessibilityListBox>() to identify native listboxes.
2240 But is<AccessibilityListBox>() returns the value of isListBox() which returns
2241 true both for AccessibilityListBox instances as well as for AccessibilityObject
2242 instances which have an AccessibilityRole value of ListBoxRole. Because only
2243 native listboxes should be AccessibilityListBoxes, add isNativeListBox() so
2244 that we can distinguish native and ARIA listboxes.
2246 Tests: accessibility/aria-listbox-clear-selection-crash.html
2247 accessibility/listbox-clear-selection.html
2249 * accessibility/AccessibilityListBox.h:
2250 * accessibility/AccessibilityObject.h:
2251 (WebCore::AccessibilityObject::isNativeListBox):
2252 (WebCore::AccessibilityObject::isListBox):
2254 2016-11-04 Brady Eidson <beidson@apple.com>
2256 IndexedDB 2.0: Throw the correct exceptions during IDBObjectStore/IDBIndex renaming.
2257 https://bugs.webkit.org/show_bug.cgi?id=164405
2259 Reviewed by Alex Christensen.
2261 No new tests (Covered by existing tests).
2263 * Modules/indexeddb/IDBIndex.cpp:
2264 (WebCore::IDBIndex::setName):
2266 * Modules/indexeddb/IDBObjectStore.cpp:
2267 (WebCore::IDBObjectStore::setName):
2269 2016-11-04 Romain Bellessort <romain.bellessort@crf.canon.fr>
2271 [Readable Streams API] Implement ByteStreamController error()
2272 https://bugs.webkit.org/show_bug.cgi?id=164319
2274 Reviewed by Youenn Fablet.
2276 Implemented error() method of ReadableByteStreamController.
2278 Updated test expectations for error() and added IDL-related tests.
2280 * Modules/streams/ReadableByteStreamController.js:
2281 (error): Implemented.
2282 * Modules/streams/ReadableByteStreamInternals.js:
2283 (privateInitializeReadableByteStreamController):
2284 (isReadableByteStreamController): Added.
2285 (readableByteStreamControllerError): Added.
2286 (readableByteStreamControllerClearPendingPullIntos): Added.
2287 * Modules/streams/ReadableStream.js:
2288 (initializeReadableStream): More detailed error message.
2289 * Modules/streams/ReadableStreamDefaultController.js:
2290 (error): Removed unnecessary variable declaration.
2291 * bindings/js/WebCoreBuiltinNames.h: Added totalQueuedBytes.
2293 2016-11-03 Brady Eidson <beidson@apple.com>
2295 IndexedDB 2.0: Handle IDBIndex rename behavior properly when version change transaction aborts.
2296 https://bugs.webkit.org/show_bug.cgi?id=164403
2298 Reviewed by Alex Christensen.
2300 No new tests (Covered by existing test).
2302 * Modules/indexeddb/IDBIndex.cpp:
2303 (WebCore::IDBIndex::rollbackInfoForVersionChangeAbort): Only rollback the info if this index
2304 already existed before this version change transaction.
2306 2016-11-03 Youenn Fablet <youenn@apple.com>
2308 [WebRTC] Introduce asynchronous backend createOffer API
2309 https://bugs.webkit.org/show_bug.cgi?id=164365
2311 Reviewed by Sam Weinig.
2313 Covered by existing tests.
2315 Removing PeerEndpointBackendClient as it is only RTCPeerConnection.
2316 This allows removing virtual for some functions.
2318 Moving MediaEndpointPeerClient::m_client to PeerEndpointBackendClient::m_peerConnection and making it a reference.
2320 Implementing createOffer at PeerConnectionBackend by splitting it in four sub-functions:
2321 - main createOffer, implemented at PeerConnectionBackend.
2322 - doCreateOffer implemented by subclasses (MediaEndpointPeerConnection).
2323 - createOfferSucceeded/createOfferFailed implemented by PeerConnectionBackend.
2326 * Modules/mediastream/MediaEndpointPeerConnection.cpp:
2327 (WebCore::createMediaEndpointPeerConnection):
2328 (WebCore::MediaEndpointPeerConnection::MediaEndpointPeerConnection):
2329 (WebCore::MediaEndpointPeerConnection::doCreateOffer):
2330 (WebCore::MediaEndpointPeerConnection::createOfferTask):
2331 (WebCore::MediaEndpointPeerConnection::createAnswerTask):
2332 (WebCore::MediaEndpointPeerConnection::setLocalDescriptionTask):
2333 (WebCore::MediaEndpointPeerConnection::setRemoteDescriptionTask):
2334 (WebCore::MediaEndpointPeerConnection::addIceCandidateTask):
2335 (WebCore::MediaEndpointPeerConnection::createReceiver):
2336 (WebCore::MediaEndpointPeerConnection::replaceTrack):
2337 (WebCore::MediaEndpointPeerConnection::replaceTrackTask):
2338 (WebCore::MediaEndpointPeerConnection::markAsNeedingNegotiation):
2339 (WebCore::MediaEndpointPeerConnection::localDescriptionTypeValidForState):
2340 (WebCore::MediaEndpointPeerConnection::remoteDescriptionTypeValidForState):
2341 (WebCore::MediaEndpointPeerConnection::gotIceCandidate):
2342 (WebCore::MediaEndpointPeerConnection::doneGatheringCandidates):
2343 (WebCore::MediaEndpointPeerConnection::iceTransportStateChanged):
2344 (WebCore::MediaEndpointPeerConnection::createOffer): Deleted.
2345 * Modules/mediastream/MediaEndpointPeerConnection.h:
2346 * Modules/mediastream/PeerConnectionBackend.cpp:
2347 (WebCore::PeerConnectionBackend::createOffer):
2348 (WebCore::PeerConnectionBackend::createOfferSucceeded):
2349 (WebCore::PeerConnectionBackend::createOfferFailed):
2350 (WebCore::createPeerConnectionBackend): Deleted.
2351 * Modules/mediastream/PeerConnectionBackend.h:
2352 (WebCore::PeerConnectionBackend::PeerConnectionBackend):
2353 (WebCore::PeerConnectionBackendClient::~PeerConnectionBackendClient): Deleted.
2354 * Modules/mediastream/RTCPeerConnection.cpp:
2355 (WebCore::RTCPeerConnection::RTCPeerConnection):
2356 * Modules/mediastream/RTCPeerConnection.h:
2357 * WebCore.xcodeproj/project.pbxproj:
2359 2016-11-03 Antti Koivisto <antti@apple.com>
2361 REGRESSION (r207717): DumpRenderTree crashed in com.apple.WebCore: WebCore::Style::Scope::flushPendingUpdate + 16
2362 https://bugs.webkit.org/show_bug.cgi?id=164397
2363 <rdar://problem/29100135>
2365 Reviewed by Ryosuke Niwa.
2367 The problem here was that we were leaving stale pointers to Document::m_inDocumentShadowRoots set when
2368 using fast-path document teardown.
2370 (Patch and stories mostly by rniwa).
2373 (WebCore::Document::~Document):
2374 (WebCore::Document::didInsertInDocumentShadowRoot):
2375 (WebCore::Document::didRemoveInDocumentShadowRoot):
2380 (WebCore::Element::removeShadowRoot):
2382 Remove the superfluous call to notifyChildNodeRemoved in Element::removeShadowRoot to
2383 avoid invoking notifyChildNodeRemoved during a document teardown, which is incorrect. It's sufficient that
2384 ~ShadowRoot calls ContainerNode::removeDetachedChildren(), and in turn removeDetachedChildrenInContainer()
2385 since the latter function tears down nodes via the deletion queue during a document destruction and use
2386 notifyChildNodeRemoved() on nodes that outlive the shadow root.
2388 * dom/ShadowRoot.cpp:
2389 (WebCore::ShadowRoot::~ShadowRoot):
2391 Take care to clean up inDocumentShadowRoots for fast-pathed destruction too.
2393 (WebCore::ShadowRoot::insertedInto):
2394 (WebCore::ShadowRoot::removedFrom):
2396 Improve ShadowRoot's insertedInto and removedFrom so that they only try to add and remove itself from
2397 m_inDocumentShadowRoots when the connected-ness changes.
2399 2016-11-03 Simon Fraser <simon.fraser@apple.com>
2401 Give all the geometry classes a single-argument scale() function for consistency
2402 https://bugs.webkit.org/show_bug.cgi?id=164400
2404 Reviewed by Zalan Bujtas.
2406 Add single-argument scale() to FloatPoint, FloatQuad, FloatSize and LayoutPoint, as well
2407 as adding one to GraphicsContext. Switch callers who passed the same value for sx and sy
2408 to the new functions.
2411 (WebCore::Document::adjustFloatQuadsForScrollAndAbsoluteZoomAndFrameScale):
2412 * dom/MouseRelatedEvent.cpp:
2413 (WebCore::MouseRelatedEvent::init):
2414 (WebCore::MouseRelatedEvent::computeRelativePosition):
2415 * dom/TreeScope.cpp:
2416 (WebCore::TreeScope::nodeFromPoint):
2417 * page/PrintContext.cpp:
2418 (WebCore::PrintContext::spoolPage):
2419 * platform/cocoa/ThemeCocoa.mm:
2420 (WebCore::fitContextToBox):
2421 * platform/graphics/FloatPoint.h:
2422 (WebCore::FloatPoint::scale):
2423 * platform/graphics/FloatQuad.h:
2424 (WebCore::FloatQuad::scale):
2425 * platform/graphics/FloatSize.h:
2426 (WebCore::FloatSize::scale):
2427 * platform/graphics/GraphicsContext.cpp:
2428 (WebCore::GraphicsContext::applyDeviceScaleFactor):
2429 * platform/graphics/GraphicsContext.h:
2430 (WebCore::GraphicsContext::scale):
2431 * platform/graphics/LayoutPoint.h:
2432 (WebCore::LayoutPoint::scale):
2433 * platform/graphics/ca/TileCoverageMap.cpp:
2434 (WebCore::TileCoverageMap::update):
2435 * platform/graphics/ca/TileGrid.cpp:
2436 (WebCore::TileGrid::platformCALayerPaintContents):
2437 * platform/graphics/cg/ImageBufferCG.cpp:
2438 (WebCore::ImageBuffer::drawConsuming):
2439 (WebCore::ImageBuffer::draw):
2440 (WebCore::ImageBuffer::drawPattern):
2441 * platform/mac/ThemeMac.mm:
2442 (WebCore::paintToggleButton):
2443 (WebCore::paintButton):
2444 (WebCore::paintStepper):
2445 * rendering/RenderImage.cpp:
2446 (WebCore::RenderImage::nodeAtPoint):
2447 * rendering/RenderMediaControls.cpp:
2448 (WebCore::getUnzoomedRectAndAdjustCurrentContext):
2449 * rendering/RenderThemeMac.mm:
2450 (WebCore::RenderThemeMac::paintMenuList):
2451 (WebCore::RenderThemeMac::paintSliderThumb):
2452 (WebCore::RenderThemeMac::paintSearchField):
2453 (WebCore::RenderThemeMac::paintSearchFieldCancelButton):
2454 (WebCore::RenderThemeMac::paintSearchFieldResultsButton):
2455 * rendering/svg/SVGInlineTextBox.cpp:
2456 (WebCore::SVGInlineTextBox::selectionRectForTextFragment):
2457 (WebCore::SVGInlineTextBox::paintDecorationWithStyle):
2458 (WebCore::SVGInlineTextBox::paintTextWithShadows):
2459 * svg/SVGPathBlender.cpp:
2460 (WebCore::SVGPathBlender::blendAnimatedFloatPoint):
2461 (WebCore::SVGPathBlender::blendArcToSegment):
2462 * svg/SVGPathParser.cpp:
2463 (WebCore::SVGPathParser::parseCurveToCubicSmoothSegment):
2464 (WebCore::SVGPathParser::parseCurveToQuadraticSegment):
2465 (WebCore::SVGPathParser::parseCurveToQuadraticSmoothSegment):
2466 (WebCore::SVGPathParser::decomposeArcToCubic):
2467 * svg/SVGSVGElement.cpp:
2468 (WebCore::SVGSVGElement::localCoordinateSpaceTransform):
2470 2016-11-03 Antti Koivisto <antti@apple.com>
2472 REGRESSION (r207669): Crash under media controls shadow root construction
2473 https://bugs.webkit.org/show_bug.cgi?id=164381
2474 <rdar://problem/28935401>
2476 Reviewed by Simon Fraser.
2478 The problem is that we are running a script for media control UA shadow tree in HTMLMediaElement::insertedInto.
2479 It is not safe to run scripts in insertedInto as the tree is in inconsistent state. Instead finishedInsertingSubtree
2480 callback should be used.
2482 Test: media/media-controls-shadow-construction-crash.html
2484 Seen on https://www.theguardian.com/artanddesign/video/2013/oct/14/banksy-central-park-new-york-video
2486 * html/HTMLMediaElement.cpp:
2487 (WebCore::HTMLMediaElement::insertedInto):
2488 (WebCore::HTMLMediaElement::finishedInsertingSubtree):
2490 Move configureMediaControls() to finishedInsertingSubtree().
2492 * html/HTMLMediaElement.h:
2493 * style/StyleTreeResolver.cpp:
2494 (WebCore::Style::TreeResolver::resolveComposedTree):
2496 Add an assert to make the bad state easier to hit in tests.
2498 2016-11-03 Ryosuke Niwa <rniwa@webkit.org>
2500 Add an assertion to diagnose stress GC bots test failures
2501 https://bugs.webkit.org/show_bug.cgi?id=164396
2503 Reviewed by Antti Koivisto.
2505 Added an assertion for calling ElementQueue::add while ElementQueue::invokeAll is in progress.
2506 This should never happen as long as all DOM API has an appropriate CEReactions IDL attribute.
2508 * dom/CustomElementReactionQueue.cpp:
2509 (WebCore::CustomElementReactionStack::ElementQueue::add):
2510 (WebCore::CustomElementReactionStack::ElementQueue::invokeAll):
2511 * dom/CustomElementReactionQueue.h:
2513 2016-11-03 Said Abou-Hallawa <sabouhallawa@apple.com>
2515 Add the asynchronous image decoding mode
2516 https://bugs.webkit.org/show_bug.cgi?id=155546
2518 Reviewed by Simon Fraser.
2520 The asynchronous image decoding feature targets enhancing the rendering
2521 in two scenarios: the animated images and scrolling a page which large
2522 images. Enabling this feature for these two scenarios will be landed
2525 The goal of the asynchronous image decoding is to have the decoded image
2526 frame ready before it has to be drawn. Drawing an image does not have to
2527 wait the image frame to be decoded.
2529 * platform/graphics/BitmapImage.cpp:
2530 (WebCore::BitmapImage::frameImageAtIndex): Use the negation of frameHasValidNativeImageAtIndex().
2531 * platform/graphics/BitmapImage.h:
2532 (WebCore::BitmapImage::frameIsBeingDecodedAtIndex): Answers whether a frame is being decoded.
2533 (WebCore::BitmapImage::frameHasValidNativeImageAtIndex): Checks the validity of a frame.
2534 (WebCore::BitmapImage::frameHasInvalidNativeImageAtIndex): Deleted.
2535 * platform/graphics/Image.h:
2536 (WebCore::Image::newFrameNativeImageAvailableAtIndex): Notifies the image with the availability of a frame NativeImage.
2537 * platform/graphics/ImageFrame.h:
2538 (WebCore::ImageFrame::isBeingDecoded): Answers whether the frame is being decoded.
2539 (WebCore::ImageFrame::hasValidNativeImage): Checks the validity of the frame.
2540 (WebCore::ImageFrame::hasInvalidNativeImage): Deleted.
2541 * platform/graphics/ImageFrameCache.cpp:
2542 (WebCore::ImageFrameCache::~ImageFrameCache): Asserts the decoding loop was ended before deleting the ImageFrameCache.
2543 (WebCore::ImageFrameCache::setFrameNativeImageAtIndex): Rename this function to matches the other which take the frame index.
2544 (WebCore::ImageFrameCache::setFrameMetadataAtIndex): Ditto.
2545 (WebCore::ImageFrameCache::replaceFrameNativeImageAtIndex): It setts the ImageFrame's members and updates the decoded size.
2546 (WebCore::ImageFrameCache::cacheFrameNativeImageAtIndex): Replaces the frame NativeImage and notifies the Image with the new frame.
2547 (WebCore::ImageFrameCache::decodingQueue): Ensures the decoding WorkQueue is created and returns it.
2548 (WebCore::ImageFrameCache::startAsyncDecodingQueue): Starts a decoding WorkQueue which loops until m_frameRequestQueue is closed.
2549 (WebCore::ImageFrameCache::requestFrameAsyncDecodingAtIndex): Allows ImageSource to send a request to start asynchronous frame image decoding.
2550 (WebCore::ImageFrameCache::stopAsyncDecodingQueue): Stops the decoding WorkQueue by closing m_frameRequestQueue.
2551 (WebCore::ImageFrameCache::frameAtIndex): Call replaceFrameNativeImageAtIndex().
2552 (WebCore::ImageFrameCache::frameIsBeingDecodedAtIndex): Returns true if a request for the image frame is issued but not finished yet.
2553 (WebCore::ImageFrameCache::frameHasValidNativeImageAtIndex): Checks the validity of a frame.
2554 (WebCore::ImageFrameCache::setFrameNativeImage): Deleted. Was renamed to be setFrameNativeImageAtIndex.
2555 (WebCore::ImageFrameCache::setFrameMetadata): Deleted. Was renamed to be setFrameMetadataAtIndex
2556 (WebCore::ImageFrameCache::frameHasInvalidNativeImageAtIndex): Deleted. Was renamed to be frameHasValidNativeImageAtIndex.
2557 * platform/graphics/ImageFrameCache.h:
2558 (WebCore::ImageFrameCache::create): The decoding queue needs to hold a reference to this class so it can stop decoding safely without blocking.
2559 (WebCore::ImageFrameCache::hasDecodingQueue): Returns true if a decoding queue has started.
2560 * platform/graphics/ImageSource.cpp:
2561 (WebCore::ImageSource::ImageSource): Call ImageFrameCache::create().
2562 (WebCore::ImageSource::clear): Deleting the decoder is unnecessary for asynchronous decoding because ImageFrameCache manages all the memory.
2564 (WebCore::ImageSource::destroyDecodedData):
2565 (WebCore::ImageSource::destroyDecodedDataIfNecessary):
2566 (WebCore::ImageSource::ensureDecoderAvailable):
2567 (WebCore::ImageSource::dataChanged):
2568 (WebCore::ImageSource::isAllDataReceived):
2569 (WebCore::ImageSource::isAsyncDecodingRequired): Answers the question whether the async image decoding is required for this ImageSource.
2570 (WebCore::ImageSource::frameImageAtIndex):
2571 * platform/graphics/ImageSource.h:
2572 (WebCore::ImageSource::decodedSize):
2573 (WebCore::ImageSource::requestFrameAsyncDecodingAtIndex):
2574 (WebCore::ImageSource::stopAsyncDecodingQueue):
2575 (WebCore::ImageSource::isSizeAvailable):
2576 (WebCore::ImageSource::frameCount):
2577 (WebCore::ImageSource::repetitionCount):
2578 (WebCore::ImageSource::filenameExtension):
2579 (WebCore::ImageSource::hotSpot):
2580 (WebCore::ImageSource::size):
2581 (WebCore::ImageSource::sizeRespectingOrientation):
2582 (WebCore::ImageSource::singlePixelSolidColor):
2583 (WebCore::ImageSource::frameIsBeingDecodedAtIndex):
2584 (WebCore::ImageSource::frameIsCompleteAtIndex):
2585 (WebCore::ImageSource::frameHasAlphaAtIndex):
2586 (WebCore::ImageSource::frameHasImageAtIndex):
2587 (WebCore::ImageSource::frameSubsamplingLevelAtIndex):
2588 (WebCore::ImageSource::frameSizeAtIndex):
2589 (WebCore::ImageSource::frameBytesAtIndex):
2590 (WebCore::ImageSource::frameDurationAtIndex):
2591 (WebCore::ImageSource::frameOrientationAtIndex):
2592 Make m_frameCache a type Ref<ImageFrameCache>. Use '->' instead of '.' when accessing its members.
2594 (WebCore::ImageSource::frameHasValidNativeImageAtIndex): Checks the validity of a frame.
2595 (WebCore::ImageSource::frameHasInvalidNativeImageAtIndex): Deleted. Was renamed to be frameHasValidNativeImageAtIndex.
2597 2016-11-03 Myles C. Maxfield <mmaxfield@apple.com>
2599 [WebGL2] Implement getBufferSubData()
2600 https://bugs.webkit.org/show_bug.cgi?id=164111
2602 Reviewed by Dean Jackson.
2604 The call exists in OpenGL 3.2 but in order to have parity with
2605 OpenGL ES 3 we back it with glMapBufferRange() instead.
2607 This patch simply adds surface area to GraphicsContext3D
2608 until we can get an ANGLE implementation of it.
2610 When testing this patch I discovered that r207649 incorrectly
2611 interpreted arguments to bufferData() and bufferSubData() as
2612 byte offsets. Instead, they should be element indices. This
2613 patch fixes those functions to work correctly so that
2614 getBufferSubData() can be tested correctly.
2616 Tests: fast/canvas/webgl/webgl2-buffers.html
2617 fast/canvas/webgl/getBufferSubData-webgl1.html
2619 * html/canvas/WebGL2RenderingContext.cpp:
2620 (WebCore::arrayBufferViewElementSize):
2621 (WebCore::WebGL2RenderingContext::bufferData):
2622 (WebCore::WebGL2RenderingContext::bufferSubData):
2623 (WebCore::WebGL2RenderingContext::getBufferSubData):
2624 * html/canvas/WebGL2RenderingContext.h:
2625 * html/canvas/WebGL2RenderingContext.idl:
2626 * platform/graphics/GraphicsContext3D.h:
2627 * platform/graphics/opengl/GraphicsContext3DOpenGLCommon.cpp:
2628 (WebCore::GraphicsContext3D::getBufferSubData):
2629 (WebCore::GraphicsContext3D::mapBufferRange):
2630 (WebCore::GraphicsContext3D::unmapBuffer):
2632 2016-11-03 Chris Dumez <cdumez@apple.com>
2634 Unreviewed, mark support for the 'download' attribute as 'Done'.
2638 2016-11-03 Chris Dumez <cdumez@apple.com>
2640 [WK2][Cocoa] Implement user interface for HTML form validation
2641 https://bugs.webkit.org/show_bug.cgi?id=164143
2642 <rdar://problem/28944652>
2644 Reviewed by Simon Fraser.
2646 Add ValidationBubble class to show HTML form validation messages
2647 using native dialogs. It currently has an implementation for both
2648 Mac and iOS. It is in WebCore under platform/ so that it can be
2649 used by both WebKit1 and WebKit2.
2651 Update ownership of ValidationMessageClient so that is is owned
2652 by the Page using a unique_ptr<>, which seems to be the modern
2653 way of handling lifetime for page clients.
2655 Test: fast/forms/validation-messages.html
2657 * WebCore.xcodeproj/project.pbxproj:
2658 * html/HTMLFormControlElement.cpp:
2659 (WebCore::HTMLFormControlElement::focusAndShowValidationMessage):
2660 * html/ValidationMessage.cpp:
2661 (WebCore::ValidationMessage::updateValidationMessage):
2663 (WebCore::Page::Page):
2664 (WebCore::Page::~Page):
2666 (WebCore::Page::validationMessageClient):
2667 * page/PageConfiguration.cpp:
2668 * page/PageConfiguration.h:
2669 * platform/ValidationBubble.h: Copied from Tools/DumpRenderTree/mac/UIScriptControllerMac.mm.
2670 (WebCore::ValidationBubble::message):
2671 * platform/ios/ValidationBubbleIOS.mm: Added.
2672 (-[WebValidationBubbleDelegate adaptivePresentationStyleForPresentationController:traitCollection:]):
2673 (WebCore::ValidationBubble::ValidationBubble):
2674 (WebCore::ValidationBubble::~ValidationBubble):
2675 (WebCore::ValidationBubble::show):
2676 (WebCore::ValidationBubble::setAnchorRect):
2677 * platform/mac/ValidationBubbleMac.mm: Added.
2678 (WebCore::ValidationBubble::ValidationBubble):
2679 (WebCore::ValidationBubble::~ValidationBubble):
2680 (WebCore::ValidationBubble::showRelativeTo):
2682 2016-11-03 Brady Eidson <beidson@apple.com>
2684 IndexedDB 2.0: Rename IDBKeyRange.contains to IDBKeyRange.includes.
2685 https://bugs.webkit.org/show_bug.cgi?id=164383
2687 Reviewed by Beth Dakin.
2689 No new tests (Updated existing tests).
2691 * Modules/indexeddb/IDBKeyRange.cpp:
2692 (WebCore::IDBKeyRange::includes):
2693 (WebCore::IDBKeyRange::contains): Deleted.
2694 * Modules/indexeddb/IDBKeyRange.h:
2695 * Modules/indexeddb/IDBKeyRange.idl:
2697 2016-11-03 Ryan Haddad <ryanhaddad@apple.com>
2699 Unreviewed, rolling out r208302.
2701 This change causes LayoutTest crashes under GuardMalloc.
2705 "Load stylesheets in link elements inside a connected shadow
2707 https://bugs.webkit.org/show_bug.cgi?id=160683
2708 http://trac.webkit.org/changeset/208302
2710 2016-11-03 Chris Dumez <cdumez@apple.com>
2712 Unreviewed, add HTML5 specification for HTML Interactive Form Validation feature.
2716 2016-11-03 Chris Dumez <cdumez@apple.com>
2718 Unreviewed, add HTML Interactive Form Validation to features.json
2722 2016-11-03 Yusuke Suzuki <utatane.tea@gmail.com>
2724 Unreviewed, rolling out due to crash in Amazon web site
2725 https://bugs.webkit.org/show_bug.cgi?id=164380
2726 <rdar://problem/29094221>
2729 * domjit/JSNodeDOMJIT.cpp:
2730 (WebCore::NodeOwnerDocumentDOMJIT::checkDOM): Deleted.
2731 (WebCore::NodeOwnerDocumentDOMJIT::callDOMGetter): Deleted.
2733 2016-11-03 Brady Eidson <beidson@apple.com>
2735 IndexedDB 2.0: Support binary keys.
2736 <rdar://problem/28806927> and https://bugs.webkit.org/show_bug.cgi?id=164359
2738 Reviewed by Alex Christensen.
2740 Tests: storage/indexeddb/modern/binary-keys-1-private.html
2741 storage/indexeddb/modern/binary-keys-1.html
2742 Changes to other existing tests.
2744 * Modules/indexeddb/IDBKey.cpp:
2745 (WebCore::IDBKey::createBinary):
2746 (WebCore::IDBKey::IDBKey):
2747 (WebCore::IDBKey::compare):
2748 * Modules/indexeddb/IDBKey.h:
2749 (WebCore::IDBKey::binary):
2750 (WebCore::compareBinaryKeyData):
2752 * Modules/indexeddb/IDBKeyData.cpp:
2753 (WebCore::IDBKeyData::IDBKeyData):
2754 (WebCore::IDBKeyData::maybeCreateIDBKey):
2755 (WebCore::IDBKeyData::isolatedCopy):
2756 (WebCore::IDBKeyData::encode):
2757 (WebCore::IDBKeyData::decode):
2758 (WebCore::IDBKeyData::compare):
2759 (WebCore::IDBKeyData::loggingString):
2760 (WebCore::IDBKeyData::operator==):
2761 * Modules/indexeddb/IDBKeyData.h:
2762 (WebCore::IDBKeyData::hash):
2763 (WebCore::IDBKeyData::encode):
2764 (WebCore::IDBKeyData::decode):
2766 * Modules/indexeddb/IndexedDB.h: Add new enum for the new key type.
2768 * bindings/js/IDBBindingUtilities.cpp:
2770 (WebCore::createIDBKeyFromValue):
2772 * platform/ThreadSafeDataBuffer.h:
2773 (WebCore::ThreadSafeDataBufferImpl::ThreadSafeDataBufferImpl):
2774 (WebCore::ThreadSafeDataBuffer::copyData):
2775 (WebCore::ThreadSafeDataBuffer::size):
2776 (WebCore::ThreadSafeDataBuffer::operator==):
2777 (WebCore::ThreadSafeDataBuffer::ThreadSafeDataBuffer):
2779 2016-11-03 Tim Horton <timothy_horton@apple.com>
2781 Printing to PDF should produce internal links when HTML has internal links
2782 https://bugs.webkit.org/show_bug.cgi?id=112081
2783 <rdar://problem/5955705>
2785 Reviewed by Simon Fraser.
2786 Patch originally by David Lattimore.
2788 No new tests, as it's unclear how to test PDF output.
2791 (WebCore::Element::findAnchorElementForLink):
2793 Add findAnchorElementForLink, which looks up the anchor element corresponding
2794 to the current element's href, and also returns the fragment name as an out parameter.
2796 * page/PrintContext.cpp:
2797 (WebCore::PrintContext::PrintContext):
2798 (WebCore::PrintContext::spoolPage):
2799 (WebCore::PrintContext::spoolRect):
2800 (WebCore::PrintContext::end):
2801 (WebCore::PrintContext::collectLinkedDestinations):
2802 (WebCore::PrintContext::outputLinkedDestinations):
2803 * rendering/RenderObject.cpp:
2804 (WebCore::RenderObject::addPDFURLRect):
2805 Plumb internal links (fragment links) through to GraphicsContext, using
2806 the fragment name from the page.
2808 * page/PrintContext.h:
2809 * platform/graphics/GraphicsContext.cpp:
2810 (WebCore::GraphicsContext::supportsInternalLinks):
2811 (WebCore::GraphicsContext::setDestinationForRect):
2812 (WebCore::GraphicsContext::addDestinationAtPoint):
2813 * platform/graphics/GraphicsContext.h:
2814 * platform/graphics/cg/GraphicsContextCG.cpp:
2815 (WebCore::GraphicsContext::supportsInternalLinks):
2816 (WebCore::GraphicsContext::setDestinationForRect):
2817 (WebCore::GraphicsContext::addDestinationAtPoint):
2818 Plumb internal links through to the CGContext. Apply the CTM, because
2819 these functions expect positions in global coordinates.
2821 * platform/graphics/win/GraphicsContextDirect2D.cpp:
2822 (WebCore::GraphicsContext::setURLForRect):
2823 * platform/graphics/cairo/GraphicsContextCairo.cpp:
2824 (WebCore::GraphicsContext::setURLForRect):
2825 Adjust setURLForRect to take a FloatRect, like everything else, and
2828 2016-11-03 Alex Christensen <achristensen@webkit.org>
2830 Unreviewed, rolling out r208298.
2831 https://bugs.webkit.org/show_bug.cgi?id=163939
2837 "NetworkSession: Add NetworkDataTask implementation for blobs"
2838 https://bugs.webkit.org/show_bug.cgi?id=163939
2839 http://trac.webkit.org/changeset/208298
2841 2016-11-03 Brent Fulgham <bfulgham@apple.com>
2843 [Win][Direct2D] Native Windows widgets are drawn upside-down
2844 https://bugs.webkit.org/show_bug.cgi?id=164370
2846 Reviewed by Alex Christensen.
2848 When we return from drawing in GDI code, we need to flip the resulting
2849 bitmap so that it draws in the proper orientation in Direct2D.
2851 Tested by existing widget tests.
2853 * platform/graphics/win/GraphicsContextDirect2D.cpp:
2854 (WebCore::GraphicsContext::releaseWindowsContext): Flip before drawing
2855 to our Direct2D context.
2857 2016-11-01 Gavin Barraclough <barraclough@apple.com>
2859 Remove PageThrottler & all related code
2860 https://bugs.webkit.org/show_bug.cgi?id=164302
2862 Reviewed by Ryosuke Niwa.
2864 All relevant information now available from the ActivityState.
2867 * WebCore.xcodeproj/project.pbxproj:
2868 * html/HTMLMediaElement.cpp:
2869 (WebCore::HTMLMediaElement::setMuted):
2870 (WebCore::HTMLMediaElement::updateVolume):
2871 (WebCore::HTMLMediaElement::updatePlayState):
2872 (WebCore::HTMLMediaElement::updateAudioAssertionState): Deleted.
2873 * html/HTMLMediaElement.h:
2874 * loader/FrameLoader.cpp:
2875 (WebCore::FrameLoader::completed):
2876 (WebCore::FrameLoader::started):
2877 * loader/FrameLoader.h:
2878 * page/ChromeClient.h:
2880 (WebCore::Page::Page):
2882 (WebCore::Page::pageActivityStateChanged): Deleted.
2883 (WebCore::Page::pageThrottler): Deleted.
2884 * page/PageThrottler.cpp: Removed.
2885 * page/PageThrottler.h: Removed.
2887 2016-11-03 Dan Bernstein <mitz@apple.com>
2889 REGRESSION (r206247): Painting milestones can be delayed until the next layer flush
2890 https://bugs.webkit.org/show_bug.cgi?id=164340
2891 <rdar://problem/29074344>
2893 Reviewed by Tim Horton.
2895 To give WebKit a chance to deliver the painting milestones to its client after the commit,
2896 we must tell it about them before or during the commit. To that end, we should not defer
2897 the call to firePaintRelatedMilestonesIfNeeded until after the commit.
2899 * rendering/RenderLayerCompositor.cpp:
2900 (WebCore::RenderLayerCompositor::RenderLayerCompositor): Removed
2901 m_paintRelatedMilestonesTimer initializer.
2902 (WebCore::RenderLayerCompositor::didPaintBacking): Call
2903 FrameView::firePaintRelatedMilestonesIfNeeded directly from here.
2904 (WebCore::RenderLayerCompositor::paintRelatedMilestonesTimerFired): Deleted.
2905 * rendering/RenderLayerCompositor.h:
2907 2016-11-03 Antti Koivisto <antti@apple.com>
2909 imported/mozilla/svg/paint-order-01.svg and imported/mozilla/svg/paint-order-02.svg are flaky failures
2910 https://bugs.webkit.org/show_bug.cgi?id=164355
2912 Reviewed by Ryosuke Niwa.
2914 Revert one change from https://trac.webkit.org/changeset/207669 to see if it is causing flakiness in
2915 some XML based tests.
2917 * xml/parser/XMLDocumentParser.cpp:
2918 (WebCore::XMLDocumentParser::end):
2920 2016-11-02 Yusuke Suzuki <utatane.tea@gmail.com>
2922 Unreviewed, fix CLoop build after r208320.
2923 https://bugs.webkit.org/show_bug.cgi?id=162980
2925 Guard with ENABLE(JIT).
2927 * bindings/scripts/CodeGeneratorJS.pm:
2928 (GenerateImplementation):
2929 * bindings/scripts/test/JS/JSTestDOMJIT.cpp:
2931 2016-11-02 Brady Eidson <beidson@apple.com>
2933 Expand upon IndexedDB status in features.json.
2937 2016-11-02 Myles C. Maxfield <mmaxfield@apple.com>
2939 CSS.supports("font-variation-settings", "'wght' 500") erroneously returns false
2940 https://bugs.webkit.org/show_bug.cgi?id=164244
2942 Reviewed by Michael Catanzaro.
2944 Because we weren't passing a Document into CSSParserContext's constructor,
2945 there was no way for our parser to know whether the runtime switch was on
2946 or not. Instead, we can use the CallWith attribute in the IDL file to pass
2949 Test: fast/text/variations/css-supports-runtime-switch.html
2951 * css/DOMCSSNamespace.cpp:
2952 (WebCore::DOMCSSNamespace::supports):
2953 * css/DOMCSSNamespace.h:
2954 * css/DOMCSSNamespace.idl:
2956 2016-11-02 Yusuke Suzuki <utatane.tea@gmail.com>
2958 [DOMJIT] Add DOMJIT::Signature
2959 https://bugs.webkit.org/show_bug.cgi?id=162980
2961 Reviewed by Saam Barati and Sam Weinig.
2963 We introduce DOMJIT::Signature. This signature object is automatically generated by IDL code generator.
2964 It holds (1) types, (2) pointer to the unsafe function (the function without checks), and (3) the effect
2965 of the function. We use constexpr to initialize DOMJIT::Signature without invoking global constructors.
2966 Thus the content is embedded into the binary as the constant values.
2968 We also clean up the IDL code generator related to DOMJIT part. Instead of switching things inside IDL
2969 code generator, we use C++ template to dispatch things at compile time. This template meta programming
2970 is highly utilized in IDL these days.
2972 To make DOMJIT::Signature constexpr, we also need to define DOMJIT abstract heap things in the build time.
2973 To do so, we introduce a tiny Ruby script to calculate the range of abstract heaps. We can offer the abstract
2974 heap tree as YAML format and the script will produce a C++ header holding the calculated abstract heap ranges
2977 * DerivedSources.make:
2978 * ForwardingHeaders/bytecode/SpeculatedType.h: Renamed from Source/WebCore/domjit/DOMJITAbstractHeapRepository.h.
2979 * ForwardingHeaders/domjit/DOMJITSignature.h: Renamed from Source/WebCore/domjit/DOMJITAbstractHeapRepository.cpp.
2980 * WebCore.xcodeproj/project.pbxproj:
2981 * bindings/js/JSDOMGlobalObject.h:
2982 * bindings/scripts/CodeGeneratorJS.pm:
2984 (GeneratePropertiesHashTable):
2985 (GetUnsafeArgumentType):
2986 (GetArgumentTypeFilter):
2987 (GetResultTypeFilter):
2988 (GenerateImplementation):
2990 (GenerateHashTableValueArray):
2991 (ComputeFunctionSpecial):
2992 * bindings/scripts/IDLAttributes.txt:
2993 * bindings/scripts/test/JS/JSTestDOMJIT.cpp:
2994 (WebCore::BindingCaller<JSTestDOMJIT>::castForOperation):
2995 (WebCore::TestDOMJITAnyAttrDOMJIT::TestDOMJITAnyAttrDOMJIT):
2996 (WebCore::TestDOMJITBooleanAttrDOMJIT::TestDOMJITBooleanAttrDOMJIT):
2997 (WebCore::TestDOMJITByteAttrDOMJIT::TestDOMJITByteAttrDOMJIT):
2998 (WebCore::TestDOMJITOctetAttrDOMJIT::TestDOMJITOctetAttrDOMJIT):
2999 (WebCore::TestDOMJITShortAttrDOMJIT::TestDOMJITShortAttrDOMJIT):
3000 (WebCore::TestDOMJITUnsignedShortAttrDOMJIT::TestDOMJITUnsignedShortAttrDOMJIT):
3001 (WebCore::TestDOMJITLongAttrDOMJIT::TestDOMJITLongAttrDOMJIT):
3002 (WebCore::TestDOMJITUnsignedLongAttrDOMJIT::TestDOMJITUnsignedLongAttrDOMJIT):
3003 (WebCore::TestDOMJITLongLongAttrDOMJIT::TestDOMJITLongLongAttrDOMJIT):
3004 (WebCore::TestDOMJITUnsignedLongLongAttrDOMJIT::TestDOMJITUnsignedLongLongAttrDOMJIT):
3005 (WebCore::TestDOMJITFloatAttrDOMJIT::TestDOMJITFloatAttrDOMJIT):
3006 (WebCore::TestDOMJITUnrestrictedFloatAttrDOMJIT::TestDOMJITUnrestrictedFloatAttrDOMJIT):
3007 (WebCore::TestDOMJITDoubleAttrDOMJIT::TestDOMJITDoubleAttrDOMJIT):
3008 (WebCore::TestDOMJITUnrestrictedDoubleAttrDOMJIT::TestDOMJITUnrestrictedDoubleAttrDOMJIT):
3009 (WebCore::TestDOMJITDomStringAttrDOMJIT::TestDOMJITDomStringAttrDOMJIT):
3010 (WebCore::TestDOMJITByteStringAttrDOMJIT::TestDOMJITByteStringAttrDOMJIT):
3011 (WebCore::TestDOMJITUsvStringAttrDOMJIT::TestDOMJITUsvStringAttrDOMJIT):
3012 (WebCore::TestDOMJITNodeAttrDOMJIT::TestDOMJITNodeAttrDOMJIT):
3013 (WebCore::TestDOMJITBooleanNullableAttrDOMJIT::TestDOMJITBooleanNullableAttrDOMJIT):
3014 (WebCore::TestDOMJITByteNullableAttrDOMJIT::TestDOMJITByteNullableAttrDOMJIT):
3015 (WebCore::TestDOMJITOctetNullableAttrDOMJIT::TestDOMJITOctetNullableAttrDOMJIT):
3016 (WebCore::TestDOMJITShortNullableAttrDOMJIT::TestDOMJITShortNullableAttrDOMJIT):
3017 (WebCore::TestDOMJITUnsignedShortNullableAttrDOMJIT::TestDOMJITUnsignedShortNullableAttrDOMJIT):
3018 (WebCore::TestDOMJITLongNullableAttrDOMJIT::TestDOMJITLongNullableAttrDOMJIT):
3019 (WebCore::TestDOMJITUnsignedLongNullableAttrDOMJIT::TestDOMJITUnsignedLongNullableAttrDOMJIT):
3020 (WebCore::TestDOMJITLongLongNullableAttrDOMJIT::TestDOMJITLongLongNullableAttrDOMJIT):
3021 (WebCore::TestDOMJITUnsignedLongLongNullableAttrDOMJIT::TestDOMJITUnsignedLongLongNullableAttrDOMJIT):
3022 (WebCore::TestDOMJITFloatNullableAttrDOMJIT::TestDOMJITFloatNullableAttrDOMJIT):
3023 (WebCore::TestDOMJITUnrestrictedFloatNullableAttrDOMJIT::TestDOMJITUnrestrictedFloatNullableAttrDOMJIT):
3024 (WebCore::TestDOMJITDoubleNullableAttrDOMJIT::TestDOMJITDoubleNullableAttrDOMJIT):
3025 (WebCore::TestDOMJITUnrestrictedDoubleNullableAttrDOMJIT::TestDOMJITUnrestrictedDoubleNullableAttrDOMJIT):
3026 (WebCore::TestDOMJITDomStringNullableAttrDOMJIT::TestDOMJITDomStringNullableAttrDOMJIT):
3027 (WebCore::TestDOMJITByteStringNullableAttrDOMJIT::TestDOMJITByteStringNullableAttrDOMJIT):
3028 (WebCore::TestDOMJITUsvStringNullableAttrDOMJIT::TestDOMJITUsvStringNullableAttrDOMJIT):
3029 (WebCore::TestDOMJITNodeNullableAttrDOMJIT::TestDOMJITNodeNullableAttrDOMJIT):
3030 (WebCore::jsTestDOMJITPrototypeFunctionGetAttribute):
3031 (WebCore::jsTestDOMJITPrototypeFunctionGetAttributeCaller):
3032 (WebCore::unsafeJsTestDOMJITPrototypeFunctionGetAttribute):
3033 (WebCore::jsTestDOMJITPrototypeFunctionItem):
3034 (WebCore::jsTestDOMJITPrototypeFunctionItemCaller):
3035 (WebCore::unsafeJsTestDOMJITPrototypeFunctionItem):
3036 (WebCore::jsTestDOMJITPrototypeFunctionHasAttribute):
3037 (WebCore::jsTestDOMJITPrototypeFunctionHasAttributeCaller):
3038 (WebCore::unsafeJsTestDOMJITPrototypeFunctionHasAttribute):
3039 (WebCore::jsTestDOMJITPrototypeFunctionGetElementById):
3040 (WebCore::jsTestDOMJITPrototypeFunctionGetElementByIdCaller):
3041 (WebCore::unsafeJsTestDOMJITPrototypeFunctionGetElementById):
3042 (WebCore::jsTestDOMJITPrototypeFunctionGetElementsByName):
3043 (WebCore::jsTestDOMJITPrototypeFunctionGetElementsByNameCaller):
3044 (WebCore::unsafeJsTestDOMJITPrototypeFunctionGetElementsByName):
3045 * bindings/scripts/test/TestDOMJIT.idl:
3047 * domjit/DOMJITAbstractHeapRepository.yaml: Added.
3048 * domjit/DOMJITIDLConvert.h: Added.
3049 (WebCore::DOMJIT::DirectConverter<IDLDOMString>::directConvert<StringConversionConfiguration::Normal>):
3050 * domjit/DOMJITIDLType.h: Added.
3051 * domjit/DOMJITIDLTypeFilter.h: Added.
3052 * domjit/JSDocumentDOMJIT.cpp:
3053 (WebCore::DocumentDocumentElementDOMJIT::callDOMGetter):
3054 * domjit/JSNodeDOMJIT.cpp:
3055 (WebCore::NodeFirstChildDOMJIT::callDOMGetter):
3056 (WebCore::NodeLastChildDOMJIT::callDOMGetter):
3057 (WebCore::NodeNextSiblingDOMJIT::callDOMGetter):
3058 (WebCore::NodePreviousSiblingDOMJIT::callDOMGetter):
3059 (WebCore::NodeParentNodeDOMJIT::callDOMGetter):
3060 (WebCore::NodeOwnerDocumentDOMJIT::callDOMGetter):
3061 * domjit/generate-abstract-heap.rb: Added.
3063 2016-11-02 Simon Fraser <simon.fraser@apple.com>
3065 Followup after r208314.
3067 The style created for reflections contains transforms and a mask, so needs to get explicit
3068 z-index on it. This doesn't change rendering, since this layer has no children.
3070 Fixes assertions in various reflection tests.
3072 * rendering/RenderLayer.cpp:
3073 (WebCore::RenderLayer::calculateClipRects):
3075 2016-11-02 Simon Fraser <simon.fraser@apple.com>
3077 REGRESSION (r208025) GraphicsContext state stack assertions loading webkit.org
3078 https://bugs.webkit.org/show_bug.cgi?id=164350
3079 rdar://problem/29053414
3081 Reviewed by Dean Jackson.
3083 After r208025 it as possible for KeyframeAnimation::animate() to produce a RenderStyle
3084 with a non-1 opacity, but without the explicit z-index that triggers stacking context.
3085 This confused the RenderLayer paintWithTransparency code, triggering mismsatched GraphicsContext
3088 This occurred when the runningOrFillingForwards state was mis-computed. keyframeAnim->animate()
3089 can spit out a new style when in the StartWaitTimer sometimes, so "!keyframeAnim->waitingToStart() && !keyframeAnim->postActive()"
3090 gave the wrong answser.
3092 Rather than depend on the super-confusing animation state, use a bool out param from animate() to say
3093 when it actually produced a new style, and when true, do the setZIndex(0).
3095 Test: animations/stacking-during-opacity-animation.html
3097 * page/animation/AnimationBase.h:
3098 * page/animation/CSSPropertyAnimation.cpp:
3099 (WebCore::CSSPropertyAnimation::blendProperties): Log after blending so the log shows the blended style.
3100 * page/animation/CompositeAnimation.cpp:
3101 (WebCore::CompositeAnimation::animate):
3102 * page/animation/ImplicitAnimation.cpp:
3103 (WebCore::ImplicitAnimation::animate):
3104 * page/animation/ImplicitAnimation.h:
3105 * page/animation/KeyframeAnimation.cpp:
3106 (WebCore::KeyframeAnimation::animate):
3107 * page/animation/KeyframeAnimation.h:
3108 * platform/graphics/GraphicsContext.cpp:
3109 (WebCore::GraphicsContext::restore):
3110 * platform/graphics/ca/cocoa/PlatformCALayerCocoa.mm:
3111 (PlatformCALayer::drawLayerContents): No functional change, but created scope for the
3112 GraphicsContext so that it didn't outlive the CGContextRestoreGState(context).
3113 * rendering/RenderLayer.cpp:
3114 (WebCore::RenderLayer::beginTransparencyLayers): New assertion that catches the problem earlier.
3116 2016-11-02 Myles C. Maxfield <mmaxfield@apple.com>
3118 [iOS] [WebGL] Multisample resolve step may operate on stale data
3119 https://bugs.webkit.org/show_bug.cgi?id=164347
3121 Reviewed by Dean Jackson.
3123 When antialiasing is enabled, WebKit internally creates a multisampled FBO
3124 and uses that as the target of all the drawing commands. Then, just before
3125 we actually put the image on the glass, we perform a “resolve” step which
3126 averages all the samples to create the final image. However, it appears
3127 that this resolve step only waits for commands to complete which were
3128 already submitted to the hardware. OpenGL is allowed (indeed, expected) to
3129 batch up drawing commands in main memory so it can submit them to the
3130 hardware in fewer batches, but this means that the hardware may not know
3131 about all the commands that the application submitted. Because of this,
3132 the data the resolve step saw is the result of only some of the previous
3133 draw calls - not all of them.
3135 This doesn’t occur on macOS because we have a different code path there
3136 for performing the resolve step. On iOS 9 and below, WebKit didn’t
3137 implement multisampling in WebGL at all, which explains why this only
3140 Luckily, the OpenGL command glFlush() is exactly designed to submit any
3141 pending commands to the hardware.
3143 Test: fast/canvas/webgl/multisample-resolve-consistency.html
3145 * platform/graphics/opengl/GraphicsContext3DOpenGL.cpp:
3146 (WebCore::GraphicsContext3D::resolveMultisamplingIfNecessary):
3148 2016-11-02 Brady Eidson <beidson@apple.com>
3150 Give IDBKey(Data) a WTF::Variant overhaul.
3151 https://bugs.webkit.org/show_bug.cgi?id=164332
3153 Reviewed by Alex Christensen and Andy Estes.
3155 No new tests (Refactor, no behavior change).
3157 * Modules/indexeddb/IDBKey.cpp:
3158 (WebCore::IDBKey::IDBKey):
3159 (WebCore::IDBKey::isValid):
3160 (WebCore::IDBKey::compare):
3161 * Modules/indexeddb/IDBKey.h:
3162 (WebCore::IDBKey::array):
3163 (WebCore::IDBKey::string):
3164 (WebCore::IDBKey::date):
3165 (WebCore::IDBKey::number):
3166 (WebCore::IDBKey::IDBKey): Deleted.
3168 * Modules/indexeddb/IDBKeyData.cpp:
3169 (WebCore::IDBKeyData::IDBKeyData):
3170 (WebCore::IDBKeyData::maybeCreateIDBKey):
3171 (WebCore::IDBKeyData::isolatedCopy):
3172 (WebCore::IDBKeyData::encode):
3173 (WebCore::IDBKeyData::decode):
3174 (WebCore::IDBKeyData::compare):
3175 (WebCore::IDBKeyData::loggingString):
3176 (WebCore::IDBKeyData::setArrayValue):
3177 (WebCore::IDBKeyData::setStringValue):
3178 (WebCore::IDBKeyData::setDateValue):
3179 (WebCore::IDBKeyData::setNumberValue):
3180 (WebCore::IDBKeyData::operator==):
3181 * Modules/indexeddb/IDBKeyData.h:
3182 (WebCore::IDBKeyData::hash):
3183 (WebCore::IDBKeyData::string):
3184 (WebCore::IDBKeyData::date):
3185 (WebCore::IDBKeyData::number):
3186 (WebCore::IDBKeyData::array):
3187 (WebCore::IDBKeyData::encode):
3188 (WebCore::IDBKeyData::decode):
3190 2016-11-01 Sam Weinig <sam@webkit.org>
3192 [WebIDL] Move interfaces and typed arrays over to JSDOMConvert
3193 https://bugs.webkit.org/show_bug.cgi?id=164256
3195 Reviewed by Alex Christensen.
3197 - Add the ability to pass an "exception thrower" functor to the convert functions.
3198 This is only implemented for convert<IDLInterface<T>> and convert<IDLNullable<IDLInterface<T>>>
3199 for now, but can be extended for more types as necessary to improve exception messages.
3200 - Add support for using toJSNewlyCreated in JSDOMConvert.
3202 * bindings/generic/IDLTypes.h:
3203 (WebCore::IDLString::extractValueFromNullable):
3204 Use forwarding to simplify extraction function.
3206 (WebCore::IDLInterface::nullValue):
3207 Update nullValue to work for both RefPtr<T> and T*.
3209 (WebCore::IDLInterface::extractValueFromNullable):
3210 Use forwarding to simplify extraction function.
3212 * bindings/js/JSDOMConvert.h:
3213 (WebCore::DefaultExceptionThrower::operator()):
3214 Add a default "exception thrower" which throws a normal type error.
3217 Add an overload of convert which takes an "exception thrower".
3219 (WebCore::toJSNewlyCreated):
3220 Add new overloaded function toJSNewlyCreated, matching the toJS overload set,
3221 which will return "newly created" values. This only works for types that implement
3222 a toJSNewlyCreated function for themselves.
3224 (WebCore::Converter<IDLNullable<T>>::convert):
3225 Fix the return type of Converter<IDLNullable<T>> to be specialized when
3226 T is an IDLInterface. In that case, we want to match the return type of
3229 Also add implementation of convert overload that takes an "exception thrower".
3231 (WebCore::JSConverter<IDLNullable<T>>::convert):
3232 (WebCore::JSConverter<IDLNullable<T>>::convertNewlyCreated):
3233 Reimplement conversion to use forwarding of the value.
3235 (WebCore::Converter<IDLInterface<T>>::convert):
3236 Add support for an "exception thrower".
3238 (WebCore::Detail::getPtrOrRef):
3239 Add helper functions that extract either a pointer or reference, depending on the type,
3240 and const_casts it allowing the value to be used with toJS functions.
3242 (WebCore::JSConverter<IDLInterface<T>>::convert):
3243 Re-implement to support more varied input values.
3245 (WebCore::JSConverter<IDLInterface<T>>::convertNewlyCreated):
3246 Added. Forwards to overloaded toJSNewlyCreated functions.
3248 * bindings/scripts/CodeGeneratorJS.pm:
3249 (AddToImplIncludesForIDLType):
3250 Add support for adding the right includes for SerializedScriptValue and Dictionary.
3252 (GetArgumentExceptionThrower):
3253 (GetAttributeExceptionThrower):
3254 Add helpers to generate "exception thrower" lambdas for wrappers and typed arrays
3255 being passed to setters and functions.
3257 (GenerateParametersCheck):
3258 Move around special cases so it is clear that it's not wrappers and typed arrays that
3259 need specialization here, it is now just EventListener and XPathNSResolver.
3261 (GetIDLInterfaceName):
3262 Add helper to get the InterfaceName for use in IDLInterface template.
3265 Use new GetIDLInterfaceName helper.
3267 (IsValidContextForJSValueToNative):
3268 Remove IDLOperation as a valid context. It is not.
3271 Move JSDOMConvert based conversion to the bottom, to show that everything above it
3272 is a special case that should be fixed. I have used explicit c-style if-statements
3273 to make it clear what the types of the exceptional cases are.
3275 (NativeToJSValueDOMConvertNeedsState):
3276 (NativeToJSValueDOMConvertNeedsGlobalObject):
3277 Add wrapper types and typed arrays to the list needing state and globalObject.
3280 Move JSDOMConvert based conversion to the bottom, to show that everything above it
3281 is a special case that should be fixed. I have used explicit c-style if-statements
3282 to make it clear what the types of the exceptional cases are.
3284 (JSValueToNativeIsHandledByDOMConvert): Deleted.
3285 (NativeToJSValueIsHandledByDOMConvert): Deleted.
3286 Remove predicates protecting use of JSDOMConvert now that it is the default.
3288 * bindings/scripts/test/JS/JSTestActiveDOMObject.cpp:
3289 * bindings/scripts/test/JS/JSTestCEReactions.cpp:
3290 * bindings/scripts/test/JS/JSTestCallback.cpp:
3291 * bindings/scripts/test/JS/JSTestCallbackFunction.cpp:
3292 * bindings/scripts/test/JS/JSTestDOMJIT.cpp:
3293 * bindings/scripts/test/JS/JSTestEventTarget.cpp:
3294 * bindings/scripts/test/JS/JSTestInterface.cpp:
3295 * bindings/scripts/test/JS/JSTestObj.cpp:
3296 * bindings/scripts/test/JS/JSTestOverloadedConstructors.cpp:
3297 * bindings/scripts/test/JS/JSTestOverrideBuiltins.cpp:
3298 * bindings/scripts/test/JS/JSTestSerialization.cpp:
3299 * bindings/scripts/test/JS/JSTestTypedefs.cpp:
3300 Update test results.
3302 2016-11-02 David Kilzer <ddkilzer@apple.com>
3304 Bug 164333: Add logging for "WebKit encountered an internal error" messages due to Network process crashes
3305 <https://webkit.org/b/164333>
3306 <rdar://problem/29072727>
3308 Reviewed by Alex Christensen.
3310 * page/DiagnosticLoggingKeys.cpp:
3311 (WebCore::DiagnosticLoggingKeys::networkProcessCrashedKey):
3312 - Add implementation for new key method.
3313 * page/DiagnosticLoggingKeys.h:
3314 (WebCore::DiagnosticLoggingKeys::networkProcessCrashedKey):
3315 - Add declaration for new key method.
3317 2016-11-02 Filip Pizlo <fpizlo@apple.com>
3319 The GC should be in a thread
3320 https://bugs.webkit.org/show_bug.cgi?id=163562
3322 Reviewed by Geoffrey Garen and Andreas Kling.
3324 No new tests because existing tests cover this.
3326 We now need to be more careful about using JSLock. This fixes some places that were not
3327 holding it. New assertions in the GC are more likely to catch this than before.
3329 * bindings/js/WorkerScriptController.cpp:
3330 (WebCore::WorkerScriptController::WorkerScriptController):
3332 2016-11-02 Joseph Pecoraro <pecoraro@apple.com>
3334 Web Inspector: Include DebuggerAgent in Workers - see, pause, and step through scripts
3335 https://bugs.webkit.org/show_bug.cgi?id=164136
3336 <rdar://problem/29028462>
3338 Reviewed by Brian Burg.
3340 Tests: inspector/worker/debugger-pause.html
3341 inspector/worker/debugger-scripts.html
3344 * WebCore.xcodeproj/project.pbxproj:
3345 * inspector/InspectorAllInOne.cpp:
3348 * inspector/PageDebuggerAgent.h:
3349 * inspector/WorkerDebuggerAgent.cpp: Added.
3350 (WebCore::WorkerDebuggerAgent::WorkerDebuggerAgent):
3351 (WebCore::WorkerDebuggerAgent::~WorkerDebuggerAgent):
3352 (WebCore::WorkerDebuggerAgent::breakpointActionLog):
3353 (WebCore::WorkerDebuggerAgent::injectedScriptForEval):
3354 * inspector/WorkerDebuggerAgent.h: Added.
3355 DebuggerAgent customizations for Workers.
3357 * inspector/WorkerInspectorController.cpp:
3358 (WebCore::WorkerInspectorController::WorkerInspectorController):
3361 * inspector/WorkerScriptDebugServer.cpp:
3362 (WebCore::WorkerScriptDebugServer::runEventLoopWhilePaused):
3363 Implement the nested run loop for Workers.
3365 2016-11-02 Simon Fraser <simon.fraser@apple.com>
3367 Add Battery Status to features.json, marked as "Removed".
3371 2016-11-02 Ryosuke Niwa <rniwa@webkit.org>
3373 Load stylesheets in link elements inside a connected shadow tree
3374 https://bugs.webkit.org/show_bug.cgi?id=160683
3375 <rdar://problem/29040652>
3377 Reviewed by Antti Koivisto.
3379 Allow external stylesheets within a shadow tree by storing the appropriate style scope in HTMLLinkElement
3380 when it's connected to a document instead of always talking to document's style scope.
3382 Also improve ShadowRoot's insertedInto and removedFrom so that they only try to add and remove itself from
3383 m_inDocumentShadowRoots when the connected-ness changes.
3385 This patch also removes the superfluous call to notifyChildNodeRemoved in Element::removeShadowRoot to
3386 avoid invoking notifyChildNodeRemoved during a document teardown, which is incorrect. It's sufficient that
3387 ~ShadowRoot calls ContainerNode::removeDetachedChildren(), and in turn removeDetachedChildrenInContainer()
3388 since the latter function tears down nodes via the deletion queue during a document destruction and use
3389 notifyChildNodeRemoved() on nodes that outlive the shadow root.
3391 Tests: fast/shadow-dom/link-element-in-shadow-tree.html
3392 fast/shadow-dom/selected-stylesheet-in-shadow-tree.html
3395 (WebCore::Document::didInsertInDocumentShadowRoot): Assert that the shadow root is not in the set.
3396 (WebCore::Document::didRemoveInDocumentShadowRoot): Assert that the shadow root is not in the document as
3397 this function is now called after Node::removedFrom in ShadowRoot::removedFrom.
3399 (WebCore::Element::removeShadowRoot): See the description above.
3400 * dom/ShadowRoot.cpp:
3401 (WebCore::ShadowRoot::insertedInto): Only call didInsertInDocumentShadowRoot when the this shadow root is
3402 newly connected to a document so we can add assertions in didInsertInDocumentShadowRoot.
3403 (WebCore::ShadowRoot::removedFrom): Ditto for the removal.
3404 * html/HTMLLinkElement.cpp:
3405 (WebCore::HTMLLinkElement::HTMLLinkElement):
3406 (WebCore::HTMLLinkElement::~HTMLLinkElement):
3407 (WebCore::HTMLLinkElement::setDisabledState): Exit early when the element is not in a document as invoking
3408 didChangeActiveStyleSheetCandidates would require having a valid m_styleScope and process() already exits
3409 early when inDocument() is false.
3410 (WebCore::HTMLLinkElement::parseAttribute):
3411 (WebCore::HTMLLinkElement::process): Removed the early exit for when the element is in a shadow tree.
3412 (WebCore::HTMLLinkElement::insertedInto): Exit early unless this element has just become connected to
3413 a document instead of whenever its self-inclusive ancestor is inserted into a container.
git://git.webkit.org / WebKit-https.git / blob