[GTK][WPE] Incorrect rendering of layers whose backingStore hasn't changed

[WebKit-https.git] / Source / WebCore / ChangeLog

2018-10-03  Miguel Gomez  <magomez@igalia.com>

        [GTK][WPE] Incorrect rendering of layers whose backingStore hasn't changed
        https://bugs.webkit.org/show_bug.cgi?id=190249

        Reviewed by Žan Doberšek.

        Do not overwrite m_nicosia.performLayerSync when updating the content buffers or we lose
        the value calculated during the layer flush. Use an OR instead to keep the old value.

        * platform/graphics/texmap/coordinated/CoordinatedGraphicsLayer.cpp:
        (WebCore::CoordinatedGraphicsLayer::updateContentBuffers):

2018-10-03  Youenn Fablet  <youenn@apple.com>

        Enable H264 simulcast
        https://bugs.webkit.org/show_bug.cgi?id=190167

        Reviewed by Eric Carlson.

        Activate H264 simulcast trial field.
        Make track.getSettings() expose width and height for incoming tracks.

        Test: webrtc/simulcast-h264.html

        * Configurations/WebCore.xcconfig:
        * Modules/mediastream/libwebrtc/LibWebRTCMediaEndpoint.cpp:
        (WebCore::LibWebRTCMediaEndpoint::LibWebRTCMediaEndpoint):
        * platform/mediastream/RealtimeIncomingVideoSource.cpp:
        (WebCore::RealtimeIncomingVideoSource::RealtimeIncomingVideoSource):

2018-10-03  Michael Catanzaro  <mcatanzaro@igalia.com>

        -Wunused-variable in RenderLayer::updateScrollableAreaSet
        https://bugs.webkit.org/show_bug.cgi?id=190200

        Reviewed by Yusuke Suzuki.

        Pass it through UNUSED_VARIABLE().

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::calculateClipRects const):

2018-10-03  Zan Dobersek  <zdobersek@igalia.com>

        Ref<FetchResponse> use-after-move in DOMCache::put()
        https://bugs.webkit.org/show_bug.cgi?id=190239

        Reviewed by Youenn Fablet.

        Retrieve reference from the Ref<FetchResponse> object before it's
        move-captured in the lambda that's passed to the
        FetchResponse::consumeBodyReceivedByChunk() method that is invoked on
        that very same object. This is a classic use-after-move bug that pops
        up on compilers with different C++ calling convention.

        * Modules/cache/DOMCache.cpp:
        (WebCore::DOMCache::put):

2018-10-03  Ryosuke Niwa  <rniwa@webkit.org>

        Enable selectionAcrossShadowBoundariesEnabled by default in WebKitLegacy
        https://bugs.webkit.org/show_bug.cgi?id=190238

        Reviewed by Antti Koivisto.

        Enable the feature by default.

        * page/Settings.yaml:

2018-10-02  Ryosuke Niwa  <rniwa@webkit.org>

        Copying content with shadow DOM doesn't copy any contents
        https://bugs.webkit.org/show_bug.cgi?id=157443

        Reviewed by Wenson Hsieh.

        This patch adds the support for copying and pasting content across shadow boundaries in HTML and plain text,
        which is enabled whenever selection across shadow boundaries is enabled.

        To do this, TextIterator now has a constructor which takes two Positions, and the node traversal code in
        StyledMarkupAccumulator has been abstracted via helper functions as done for TextIterator.

        When serializing a HTMl slot element, serialize it as a span with "display: contents" to make sure when
        the content is pasted into a shadow tree, it wouldn't affect the slot assignment of the shadow tree.

        Tests: editing/pasteboard/copy-paste-across-shadow-boundaries-1.html
               editing/pasteboard/copy-paste-across-shadow-boundaries-2.html
               editing/pasteboard/copy-paste-across-shadow-boundaries-3.html
               editing/pasteboard/copy-paste-across-shadow-boundaries-4.html
               editing/pasteboard/copy-paste-across-shadow-boundaries-with-style-1.html
               editing/pasteboard/copy-paste-across-shadow-boundaries-with-style-2.html
               editing/pasteboard/copy-paste-with-shadow-content.html

        * dom/ComposedTreeIterator.h:
        (WebCore::assignedSlotIgnoringUserAgentShadow): Moved from TextIterator.cpp.
        (WebCore::shadowRootIgnoringUserAgentShadow): Ditto.
        (WebCore::firstChildInComposedTreeIgnoringUserAgentShadow): Ditto.
        (WebCore::nextSiblingInComposedTreeIgnoringUserAgentShadow): Ditto.
        * dom/Position.h:
        (WebCore::Position::treeScope const): Added.
        * editing/EditingStyle.cpp:
        (WebCore::EditingStyle::addDisplayContents): Added.
        * editing/EditingStyle.h:
        * editing/Editor.cpp:
        (WebCore::Editor::selectedText const): Use the new behavior when selectionAcrossShadowBoundariesEnabled is set.
        (WebCore::Editor::selectedTextForDataTransfer const): Ditto.
        * editing/MarkupAccumulator.cpp:
        (WebCore::MarkupAccumulator::appendEndElement): Renamed from appendEndTag. Now takes StringBuilder.
        * editing/MarkupAccumulator.h:
        (WebCore::MarkupAccumulator::appendEndTag):
        * editing/TextIterator.cpp:
        (WebCore::TextIterator::TextIterator): Added a new variant which takes two positions.
        (WebCore::TextIterator::init):
        (WebCore::firstChild):
        (WebCore::nextSibling):
        (WebCore::plainText): Ditto.
        * editing/TextIterator.h:
        * editing/cocoa/EditorCocoa.mm:
        (WebCore::Editor::selectionInHTMLFormat): Use the new behavior if selectionAcrossShadowBoundariesEnabled is set.
        * editing/gtk/EditorGtk.cpp:
        (WebCore::Editor::writeSelectionToPasteboard): Ditto.
        * editing/markup.cpp:
        (WebCore::StyledMarkupAccumulator::parentNode): Added.
        (WebCore::StyledMarkupAccumulator::firstChild): Added.
        (WebCore::StyledMarkupAccumulator::nextSibling): Added.
        (WebCore::StyledMarkupAccumulator::nextSkippingChildren): Added.
        (WebCore::StyledMarkupAccumulator::hasChildNodes): Added.
        (WebCore::StyledMarkupAccumulator::isDescendantOf): Added.
        (WebCore::StyledMarkupAccumulator::StyledMarkupAccumulator):
        (WebCore::StyledMarkupAccumulator::appendElement): Serialize a slot element as a span with display: contents.
        (WebCore::StyledMarkupAccumulator::appendEndElement): Added. Ditto.
        (WebCore::StyledMarkupAccumulator::serializeNodes):
        (WebCore::StyledMarkupAccumulator::traverseNodesForSerialization): Use the newly added helper functions to
        traverse the composed tree when m_useComposedTree is set.
        (WebCore::commonShadowIncludingAncestor): Added.
        (WebCore::serializePreservingVisualAppearanceInternal): Added SerializeComposedTree as an argument. Also use
        StyledMarkupAccumulator::parentNode to serialize special common ancestors; e.g. to preserve b, i, etc...
        (WebCore::serializePreservingVisualAppearance): Ditto to the variant which takes VisibleSelection.
        (WebCore::sanitizedMarkupForFragmentInDocument):
        * editing/markup.h:
        * editing/wpe/EditorWPE.cpp:
        (WebCore::Editor::writeSelectionToPasteboard):
        * loader/archive/cf/LegacyWebArchive.cpp:
        (WebCore::LegacyWebArchive::createFromSelection):
        * page/PageSerializer.cpp:
        (WebCore::PageSerializer::SerializerMarkupAccumulator::appendEndElement):
        * testing/Internals.cpp:
        (WebCore::Internals::setSelectionWithoutValidation): Added. A helper function to create a selection across
        shadow boundaries for testing purposes.
        * testing/Internals.h:
        * testing/Internals.idl:

2018-10-02  Chris Dumez  <cdumez@apple.com>

        MessageEvent.ports should return the same object
        https://bugs.webkit.org/show_bug.cgi?id=190151

        Reviewed by Darin Adler.

        MessageEvent.ports should return the same object it was initialized to instead of
        constructing a new JSValue every time.

        No new tests, rebaselined existing test.

        * bindings/js/JSMessageEventCustom.cpp:
        (WebCore::JSMessageEvent::ports const):
        (WebCore::JSMessageEvent::visitAdditionalChildren):
        * dom/MessageEvent.cpp:
        (WebCore::MessageEvent::initMessageEvent):
        * dom/MessageEvent.h:
        * dom/MessageEvent.idl:

2018-10-01  Ryosuke Niwa  <rniwa@webkit.org>

        GC can collect JS wrappers of nodes in the mutation records waiting to be delivered
        https://bugs.webkit.org/show_bug.cgi?id=190115

        Reviewed by Geoffrey Garen.

        Fixed the bug by retaining JS wrappers of elements in mutation records using GCReachableRef.

        This patch deploys GCReachableRef in two places: MutationObserver where each mutation record's
        target is kept alive and MutationObserverRegistration where each node which had been removed
        from an observed tree is kept alive for a subtree observation.

        Test: fast/dom/MutationObserver/mutation-observer-retains-js-wrappers-of-targets-alive.html

        * dom/GCReachableRef.h:
        (WebCore::GCReachableRef): Made it work with hash table.
        (WebCore::GCReachableRef::operator T& const):
        (WebCore::GCReachableRef::GCReachableRef):
        (WebCore::GCReachableRef::isHashTableDeletedValue const):
        (WebCore::GCReachableRef::isHashTableEmptyValue const):
        (WebCore::GCReachableRef::ptrAllowingHashTableEmptyValue const):
        (WebCore::GCReachableRef::ptrAllowingHashTableEmptyValue):
        (WebCore::GCReachableRef::assignToHashTableEmptyValue):
        (WTF::HashTraits<WebCore::GCReachableRef<P>>::emptyValue):
        (WTF::HashTraits<WebCore::GCReachableRef<P>>::constructEmptyValue):
        (WTF::HashTraits<WebCore::GCReachableRef<P>>::isEmptyValue):
        (WTF::HashTraits<WebCore::GCReachableRef<P>>::assignToEmpty):
        (WTF::HashTraits<WebCore::GCReachableRef<P>>::peek):
        (WTF::HashTraits<WebCore::GCReachableRef<P>>::take):
        * dom/MutationObserver.cpp:
        (WebCore::MutationObserver::takeRecords): Don't clear m_pendingTargets because that would allow wrappers
        to be collected before elements in mutation records are accessed. We delay until the end of the current
        microtask at which point deliver() function is called.
        (WebCore::MutationObserver::disconnect):
        (WebCore::MutationObserver::enqueueMutationRecord): Add the target to the list of elements to keep alive.
        This is needed for a newly inserted node, a node with attribute change, etc...
        (WebCore::MutationObserver::deliver): Keep the set of transient registration targets alive until mutation
        records are delivered to each observer. These are nodes which had been removed from a tree and whose
        subtree had still been obsreved up until this point.
        * dom/MutationObserver.h:
        * dom/MutationObserverRegistration.cpp:
        (WebCore::MutationObserverRegistration::observedSubtreeNodeWillDetach):
        (WebCore::MutationObserverRegistration::takeTransientRegistrations): Return the hash set of elemenets
        that need to be kept alive so that MutationObserver::deliver can keep them alive until the deliver
        function had been called.
        (WebCore::MutationObserverRegistration::addRegistrationNodesToSet const):
        * dom/MutationObserverRegistration.h:

2018-10-02  Chris Dumez  <cdumez@apple.com>

        radio / checkbox inputs should fire "click, input, change" events in order when clicked
        https://bugs.webkit.org/show_bug.cgi?id=190223

        Reviewed by Ryosuke Niwa.

        radio / checkbox inputs should fire "click, input, change" events in order when clicked:
        - https://html.spec.whatwg.org/#radio-button-state-(type=radio)
        - https://html.spec.whatwg.org/#checkbox-state-(type=checkbox)
        - https://dom.spec.whatwg.org/#ref-for-eventtarget-activation-behavior③ (step 11)

        Gecko and Blink already behave this way. However, WebKit has the following issues:
        - the input event is not fired
        - the click event is fired after the change event

        No new tests, updated / rebaselined existing tests.

        * html/BaseCheckableInputType.cpp:
        (WebCore::BaseCheckableInputType::fireInputAndChangeEvents):
        * html/BaseCheckableInputType.h:
        * html/CheckboxInputType.cpp:
        (WebCore::CheckboxInputType::willDispatchClick):
        (WebCore::CheckboxInputType::didDispatchClick):
        * html/HTMLInputElement.cpp:
        (WebCore::HTMLInputElement::setChecked):
        * html/HTMLInputElement.h:
        * html/RadioInputType.cpp:
        (WebCore::RadioInputType::willDispatchClick):
        (WebCore::RadioInputType::didDispatchClick):

2018-10-02  Chris Dumez  <cdumez@apple.com>

        fieldset.elements should return an HTMLCollection instead of an HTMLFormControlsCollection
        https://bugs.webkit.org/show_bug.cgi?id=190218

        Reviewed by Alex Christensen.

        fieldset.elements should return an HTMLCollection instead of an HTMLFormControlsCollection:
        - https://github.com/whatwg/html/commit/8beedf0c2ffd38853caddec67490288f47afc8eb

        Gecko has always behaved this way. Blink aligned with Gecko and the HTML specification in December 2016:
        - https://bugs.chromium.org/p/chromium/issues/detail?id=665291

        This simplifies our HTMLFieldSetElement code a lot.

        Test: fast/forms/fieldset/fieldset-elements-htmlcollection.html

        * html/CollectionType.h:
        * html/GenericCachedHTMLCollection.cpp:
        (WebCore::GenericCachedHTMLCollection<traversalType>::elementMatches const):
        * html/HTMLCollection.cpp:
        (WebCore::HTMLCollection::rootTypeFromCollectionType):
        (WebCore::invalidationTypeExcludingIdAndNameAttributes):
        * html/HTMLFieldSetElement.cpp:
        (WebCore::HTMLFieldSetElement::elements):
        * html/HTMLFieldSetElement.h:
        * html/HTMLFieldSetElement.idl:
        * html/HTMLFormControlsCollection.cpp:
        (WebCore::HTMLFormControlsCollection::HTMLFormControlsCollection):
        (WebCore:: const):
        (WebCore::HTMLFormControlsCollection::copyFormControlElementsVector const):
        (WebCore::HTMLFormControlsCollection::ownerNode const):
        (WebCore::HTMLFormControlsCollection::updateNamedElementCache const):
        * html/HTMLFormControlsCollection.h:

2018-10-02  Devin Rousso  <drousso@apple.com>

        Web Inspector: prevent layer events from firing until the layer information is re-requested
        https://bugs.webkit.org/show_bug.cgi?id=190159

        Reviewed by Joseph Pecoraro.

        Test: inspector/layers/layerTreeDidChange.html

        * inspector/agents/InspectorLayerTreeAgent.h:
        * inspector/agents/InspectorLayerTreeAgent.cpp:
        (WebCore::InspectorLayerTreeAgent::reset):
        (WebCore::InspectorLayerTreeAgent::layerTreeDidChange):
        (WebCore::InspectorLayerTreeAgent::layersForNode):

2018-10-02  Brian Burg  <bburg@apple.com>

        Web Automation: tab default key handler should always cycle focus when page is controlled by automation
        https://bugs.webkit.org/show_bug.cgi?id=190221
        <rdar://problem/44914534>

        Reviewed by Joseph Pecoraro.

        This change progresses WPT WebDriver test special_keys.py::test_webdriver_special_key_sends_keydown[TAB-expected24].

        * page/FocusController.cpp:
        (WebCore::FocusController::advanceFocusInDocumentOrder):
        Always cycle focus if the page is controlled by automation. If the chrome takes
        focus, then the first responder will be something other than the WebView, which
        causes subsequent WebDriver commands to hang.

2018-10-01  Dean Jackson  <dino@apple.com>

        [macOS] Switching to discrete GPU should be done in the UI process
        https://bugs.webkit.org/show_bug.cgi?id=189361
        <rdar://problem/43949622>

        Reviewed by Simon Fraser.

        Based on an earlier patch by Per Arne Vollan.

        Due to the fact we can't talk to the Window Server, the Web Process can
        no longer muxing to the discrete GPU directly. Instead we have to get the
        UI Process to process the change. Do this by adding a new Chrome client
        called GPUClient, that will have implementations provided by both WebKit
        and legacy WebKit.

        Unfortunately this can't be tested by a regular WKTR since:
        - it requires specific hardware
        - swapping to/from the discrete GPU takes about 20 seconds
        - running concurrent tests could confuse the tests into thinking
          the wrong GPU is active

        Instead we'll write a specific test for this functionality and
        run it on a separate bot.

        * WebCore.xcodeproj/project.pbxproj: Add GPUClient files.

        * page/Chrome.cpp: Drive by clean-up.
        (WebCore::Chrome::windowScreenDidChange):

        * platform/graphics/GraphicsContext3D.h: We need to keep track of
        whether we've muxed for this context, in order to not respond to
        the screen change notifications (they are misleading in the case
        of muxing).

        * platform/graphics/GraphicsContext3DManager.cpp: Rather than try
        to mux directly, call into GPUClient.
        (WebCore::GraphicsContext3DManager::displayWasReconfigured):
        (WebCore::GraphicsContext3DManager::updateHighPerformanceState):
        (WebCore::GraphicsContext3DManager::disableHighPerformanceGPUTimerFired):
        (WebCore::GraphicsContext3DManager::recycleContextIfNecessary):
        * platform/graphics/GraphicsContext3DManager.h:

        * platform/graphics/cocoa/GraphicsContext3DCocoa.mm: Only reconfigure
        the virtual display if it didn't happen from muxing.
        (WebCore::GraphicsContext3D::GraphicsContext3D):
        (WebCore::GraphicsContext3D::updateCGLContext):
        (WebCore::GraphicsContext3D::screenDidChange):

        * platform/graphics/mac/SwitchingGPUClient.cpp: Added.
        (WebCore::SwitchingGPUClient::singleton):
        (WebCore::SwitchingGPUClient::setSingleton):
        * platform/graphics/mac/SwitchingGPUClient.h: Added.

        * testing/Internals.cpp: Testing helper.
        (WebCore::Internals::hasMuxableGPU):
        * testing/Internals.h:
        * testing/Internals.idl:

2018-10-02  Chris Dumez  <cdumez@apple.com>

        Image.__proto__ should be Function.prototype, not HTMLElement.prototype
        https://bugs.webkit.org/show_bug.cgi?id=190216

        Reviewed by Alex Christensen.

        Properties created for named constructors should always use Function.prototype as prototype, as per:
        - https://heycam.github.io/webidl/#named-constructors

        Gecko and Blink agree with the Web IDL specification. However, WebKit was using the parent interface's
        prototype if such a parent existing. So Image.__proto__ would end up being HTMLElement.prototype
        instead of Function.prototype.

        No new tests, rebaselined existing test.

        * bindings/scripts/CodeGeneratorJS.pm:
        (GenerateConstructorHelperMethods):

2018-10-02  Alex Christensen  <achristensen@webkit.org>

        Prepare WebCoreNSURLExtras for ARC
        https://bugs.webkit.org/show_bug.cgi?id=190219

        Reviewed by Tim Horton.

        ARC doesn't like the explicit sending of -release.
        Use RetainPtr instead.

        * platform/mac/WebCoreNSURLExtras.mm:
        (WebCore::collectRangesThatNeedMapping):
        (WebCore::collectRangesThatNeedEncoding):
        (WebCore::collectRangesThatNeedDecoding):
        (WebCore::applyHostNameFunctionToMailToURLString):
        (WebCore::applyHostNameFunctionToURLString):
        (WebCore::mapHostNames):
        (WebCore::stringByTrimmingWhitespace):
        (WebCore::URLWithUserTypedString):
        (WebCore::userVisibleString):
        (WebCore::rangeOfURLScheme):
        (WebCore::looksLikeAbsoluteURL):
        (WebCore::retain): Deleted.

2018-10-02  Basuke Suzuki  <Basuke.Suzuki@sony.com>

        [Curl] Fix missing values of  resource timing API.
        https://bugs.webkit.org/show_bug.cgi?id=190193

        Reviewed by Alex Christensen.

        The property nextHopProtocol was not returned correctly. It was
        returned only when remote inspector is opened.

        Tests: http/wpt/resource-timing/rt-nextHopProtocol.html
               http/wpt/resource-timing/rt-nextHopProtocol.worker.html

        * platform/network/curl/CurlContext.cpp:
        (WebCore::CurlHandle::getNetworkLoadMetrics):
        (WebCore::CurlHandle::addExtraNetworkLoadMetrics):

2018-10-02  Alex Christensen  <achristensen@webkit.org>

        Remove unused linked-on-or-before-iOS5 check
        https://bugs.webkit.org/show_bug.cgi?id=190164

        Reviewed by Michael Saboff.

        If an app hasn't been updated since iOS5, it can't run supported iOS.
        This value is also only checked in an uninstantiated template function.

        * platform/URL.cpp:
        (WebCore::enableURLSchemeCanonicalization): Deleted.
        (WebCore::equal): Deleted.
        * platform/URL.h:

2018-10-02  Alex Christensen  <achristensen@webkit.org>

        Remove ParsedURLString
        https://bugs.webkit.org/show_bug.cgi?id=190154

        Reviewed by Chris Dumez.

        Before the introduction of URLParser, it would indicate that we should assume the String
        is from a valid URL so we can skip canonicalization and just find the offsets inside the String
        to quickly create a URL.  It was a performance optimization that caused security issues when
        misused.  Since the introduction of URLParser, we have a fast path for all URL parsing, so
        right now it actually doesn't change any behavior.  It's just a relic of the past that complicates
        the URL class, making it harder to express which constructor to use and making it harder to move
        the class.

        * Modules/navigatorcontentutils/NavigatorContentUtils.cpp:
        (WebCore::NavigatorContentUtils::registerProtocolHandler):
        (WebCore::NavigatorContentUtils::isProtocolHandlerRegistered):
        (WebCore::NavigatorContentUtils::unregisterProtocolHandler):
        * dom/Document.cpp:
        (WebCore::Document::updateBaseURL):
        (WebCore::Document::initSecurityContext):
        * dom/ExtensionStyleSheets.cpp:
        (WebCore::ExtensionStyleSheets::updateInjectedStyleSheetCache const):
        * dom/ProcessingInstruction.cpp:
        (WebCore::ProcessingInstruction::checkStyleSheet):
        * editing/markup.cpp:
        (WebCore::completeURLs):
        * fileapi/BlobURL.cpp:
        (WebCore::BlobURL::createBlobURL):
        * history/HistoryItem.cpp:
        (WebCore::HistoryItem::url const):
        (WebCore::HistoryItem::originalURL const):
        * html/HTMLFrameElementBase.cpp:
        (WebCore::HTMLFrameElementBase::location const):
        * html/HTMLMediaElement.cpp:
        (WebCore::HTMLMediaElement::ensureMediaControlsInjectedScript):
        * html/PublicURLManager.cpp:
        (WebCore::PublicURLManager::stop):
        * inspector/InspectorStyleSheet.cpp:
        (WebCore::InspectorStyleSheet::resourceStyleSheetText const):
        * inspector/agents/InspectorPageAgent.cpp:
        (WebCore::InspectorPageAgent::getCookies):
        (WebCore::InspectorPageAgent::deleteCookie):
        (WebCore::InspectorPageAgent::getResourceContent):
        (WebCore::InspectorPageAgent::searchInResource):
        * inspector/agents/page/PageDebuggerAgent.cpp:
        (WebCore::PageDebuggerAgent::sourceMapURLForScript):
        * loader/DocumentLoader.cpp:
        (WebCore::DocumentLoader::subresources const):
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::init):
        (WebCore::FrameLoader::initForSynthesizedDocument):
        * loader/HistoryController.cpp:
        (WebCore::HistoryController::pushState):
        (WebCore::HistoryController::replaceState):
        * loader/appcache/ApplicationCache.cpp:
        (WebCore::ApplicationCache::addResource):
        (WebCore::ApplicationCache::resourceForURL):
        * loader/appcache/ApplicationCacheGroup.cpp:
        (WebCore::ApplicationCacheGroup::startLoadingEntry):
        (WebCore::ApplicationCacheGroup::addEntry):
        * loader/appcache/ApplicationCacheStorage.cpp:
        (WebCore::ApplicationCacheStorage::cacheGroupForURL):
        (WebCore::ApplicationCacheStorage::fallbackCacheGroupForURL):
        (WebCore::ApplicationCacheStorage::loadCache):
        (WebCore::ApplicationCacheStorage::manifestURLs):
        * loader/archive/cf/LegacyWebArchive.cpp:
        (WebCore::LegacyWebArchive::create):
        * page/DOMWindow.cpp:
        (WebCore::DOMWindow::createWindow):
        * page/PageSerializer.cpp:
        (WebCore::PageSerializer::urlForBlankFrame):
        * platform/URL.cpp:
        (WebCore::blankURL):
        * platform/URL.h:
        (): Deleted.
        * platform/graphics/avfoundation/MediaPlayerPrivateAVFoundation.cpp:
        (WebCore::MediaPlayerPrivateAVFoundation::load):
        * platform/network/BlobRegistryImpl.cpp:
        (WebCore::BlobRegistryImpl::populateBlobsForFileWriting):
        * platform/network/ResourceRequestBase.h:
        (WebCore::ResourceRequestBase::decodeBase):
        * platform/network/ResourceResponseBase.cpp:
        (WebCore::ResourceResponseBase::sanitizeSuggestedFilename):
        * platform/network/cf/DNSResolveQueueCFNet.cpp:
        (WebCore::DNSResolveQueueCFNet::updateIsUsingProxy):
        * platform/network/cf/ResourceRequest.h:
        (WebCore::ResourceRequest::ResourceRequest):
        * platform/network/curl/CookieJarDB.cpp:
        (WebCore::CookieJarDB::searchCookies):
        (WebCore::CookieJarDB::setCookie):
        (WebCore::CookieJarDB::deleteCookie):
        * platform/network/curl/ResourceRequest.h:
        (WebCore::ResourceRequest::ResourceRequest):
        * platform/network/soup/ResourceRequest.h:
        (WebCore::ResourceRequest::ResourceRequest):
        * xml/XSLTProcessorLibxslt.cpp:
        (WebCore::docLoaderFunc):

2018-10-02  Per Arne Vollan  <pvollan@apple.com>

        [WebVTT] Cue with line setting is not rendered correctly
        https://bugs.webkit.org/show_bug.cgi?id=190168

        Reviewed by Eric Carlson.

        When the line setting contains an optional alignment value, the cue is not rendered at the correct position,
        see https://w3c.github.io/webvtt/#webvtt-line-cue-setting. This patch does not implement correct handling of
        the line setting alignment values, it only makes sure parsing does not fail when the cue has line alignment
        settings.

        Test: media/track/track-cue-line-position.html

        * html/track/VTTCue.cpp:
        (WebCore::VTTCueBox::applyCSSProperties):
        (WebCore::VTTCue::getPositionCoordinates const):
        (WebCore::VTTCue::setCueSettings):

2018-10-02  Antti Koivisto  <antti@apple.com>

        User installed fonts are not always disabled when they should be
        https://bugs.webkit.org/show_bug.cgi?id=190195

        Reviewed by Geoffrey Garen.

        SVG images and some theme cases fail to respect the setting. Besides the obvious problem this
        is also a performance issue as various font caches include this setting in the key.

        * platform/graphics/FontDescription.cpp:
        (WebCore::m_shouldAllowUserInstalledFonts):

        Initialize to 'No' by default. All paths where user fonts make sense already set the bit from
        settings. This fixes some cases in system themes that construct FontDescriptions from scratch.

        * rendering/RenderElement.cpp:
        (WebCore::RenderElement::styleWillChange):

        Add assertion. This verified the change with the existing tests.

        * svg/graphics/SVGImage.cpp:
        (WebCore::SVGImage::dataChanged):

        Always disallow user fonts in SVG used as images.

2018-10-01  Dean Jackson  <dino@apple.com>

        Remove CSS Animation Triggers
        https://bugs.webkit.org/show_bug.cgi?id=190175
        <rdar://problem/44925626>

        Reviewed by Simon Fraser.

        Remove the never-properly specified CSS Animation Triggers.

        * Configurations/FeatureDefines.xcconfig:
        * Sources.txt:
        * WebCore.xcodeproj/project.pbxproj:
        * css/CSSAnimationTriggerScrollValue.cpp: Removed.
        * css/CSSAnimationTriggerScrollValue.h: Removed.
        * css/CSSComputedStyleDeclaration.cpp:
        (WebCore::ComputedStyleExtractor::valueForPropertyinStyle):
        (WebCore::createAnimationTriggerValue): Deleted.
        (WebCore::animationTriggerValue): Deleted.
        * css/CSSProperties.json:
        * css/CSSToStyleMap.cpp:
        (WebCore::CSSToStyleMap::mapAnimationTrigger): Deleted.
        * css/CSSToStyleMap.h:
        * css/CSSValue.cpp:
        (WebCore::CSSValue::equals const):
        (WebCore::CSSValue::cssText const):
        (WebCore::CSSValue::destroy):
        * css/CSSValue.h:
        (WebCore::CSSValue::isAnimationTriggerScrollValue const): Deleted.
        * css/parser/CSSPropertyParser.cpp:
        (WebCore::consumeAnimationValue):
        (WebCore::CSSPropertyParser::parseSingleValue):
        (WebCore::consumeWebkitAnimationTrigger): Deleted.
        * page/FrameView.cpp:
        (WebCore::FrameView::sendScrollEvent):
        * page/RuntimeEnabledFeatures.h:
        (WebCore::RuntimeEnabledFeatures::setAnimationTriggersEnabled): Deleted.
        (WebCore::RuntimeEnabledFeatures::animationTriggersEnabled const): Deleted.
        * page/animation/AnimationBase.cpp:
        (WebCore::AnimationBase::updateStateMachine):
        (WebCore::AnimationBase::fireAnimationEventsIfNeeded):
        (WebCore::AnimationBase::timeToNextService):
        (WebCore::AnimationBase::getElapsedTime const):
        * page/animation/CSSAnimationController.cpp:
        (WebCore::CSSAnimationControllerPrivate::animationWillBeRemoved):
        (WebCore::CSSAnimationControllerPrivate::addToAnimationsDependentOnScroll): Deleted.
        (WebCore::CSSAnimationControllerPrivate::removeFromAnimationsDependentOnScroll): Deleted.
        (WebCore::CSSAnimationControllerPrivate::scrollWasUpdated): Deleted.
        (WebCore::CSSAnimationController::wantsScrollUpdates const): Deleted.
        (WebCore::CSSAnimationController::scrollWasUpdated): Deleted.
        * page/animation/CSSAnimationController.h:
        * page/animation/CSSAnimationControllerPrivate.h:
        (WebCore::CSSAnimationControllerPrivate::wantsScrollUpdates const): Deleted.
        (WebCore::CSSAnimationControllerPrivate::scrollPosition const): Deleted.
        * page/animation/CompositeAnimation.cpp:
        (WebCore::CompositeAnimation::updateKeyframeAnimations):
        * page/animation/CompositeAnimation.h:
        (WebCore::CompositeAnimation::hasScrollTriggeredAnimation const): Deleted.
        * platform/animation/Animation.cpp:
        (WebCore::Animation::Animation):
        (WebCore::Animation::operator=):
        (WebCore::Animation::animationsMatch const):
        * platform/animation/Animation.h:
        (WebCore::Animation::isTimingFunctionSet const):
        (WebCore::Animation::isEmpty const):
        (WebCore::Animation::clearTimingFunction):
        (WebCore::Animation::clearAll):
        (WebCore::Animation::animationMode const):
        (WebCore::Animation::setAnimationMode):
        (WebCore::Animation::initialTimingFunction):
        (WebCore::Animation::isTriggerSet const): Deleted.
        (WebCore::Animation::clearTrigger): Deleted.
        (WebCore::Animation::trigger const): Deleted.
        (WebCore::Animation::setTrigger): Deleted.
        (WebCore::Animation::initialTrigger): Deleted.
        * platform/animation/AnimationTrigger.h: Removed.
        * platform/graphics/ca/GraphicsLayerCA.cpp:
        (WebCore::GraphicsLayerCA::animationCanBeAccelerated const):
2018-10-02  Commit Queue  <commit-queue@webkit.org>

        Unreviewed, rolling out r236624 and r236671.
        https://bugs.webkit.org/show_bug.cgi?id=190207

        The change in r236624 introduced crashes on the bots
        (Requested by ryanhaddad on #webkit).

        Reverted changesets:

        "Refactoring: eliminate raw pointer usage in Fullscreen code"
        https://bugs.webkit.org/show_bug.cgi?id=188747
        https://trac.webkit.org/changeset/236624

        "Unify implementation in VideoFullscreenInterfaceAVKit"
        https://bugs.webkit.org/show_bug.cgi?id=190091
        https://trac.webkit.org/changeset/236671

2018-10-02  Sihui Liu  <sihui_liu@apple.com>

        Add release assertion to ensure m_owningPointerForClose is null in UniqueIDBDatabase::invokeOperationAndTransactionTimer()
        https://bugs.webkit.org/show_bug.cgi?id=190178

        Reviewed by Chris Dumez.

        This would help debug rdar://problem/44902833.

        * Modules/indexeddb/server/UniqueIDBDatabase.cpp:
        (WebCore::IDBServer::UniqueIDBDatabase::invokeOperationAndTransactionTimer):

2018-10-02  Commit Queue  <commit-queue@webkit.org>

        Unreviewed, rolling out r236719.
        https://bugs.webkit.org/show_bug.cgi?id=190197

        this revision caused 39 layout test failures that tested for
        scrolling, a bug was also not present in the commit or change
        log. (Requested by Truitt on #webkit).

        Reverted changeset:

        "Unreviewed, fix unused variable in
        RenderLayer::updateScrollableAreaSet"
        https://trac.webkit.org/changeset/236719

2018-10-02  Alicia Boya García  <aboya@igalia.com>

        [MSE][GStreamer] Add h264parse to accept MP4 without stss
        https://bugs.webkit.org/show_bug.cgi?id=190143

        Reviewed by Xabier Rodriguez-Calvar.

        The MP4 file used in this URL does not contain a stss (Sync Sample
        Box). In consequence, in acordance with the ISO BMFF spec, all samples
        are assumed to be sync frames... But in this case that is not true,
        it's just that the file is wrong (e.g. created with a buggy muxer).

        http://orange-opensource.github.io/hasplayer.js/1.2.0/player.html?url=http://playready.directtaps.net/smoothstreaming/SSWSS720H264/SuperSpeedway_720.ism/Manifest

        The way it works in other browsers is because instead of trusting the
        MP4 stss table, they rely on parsing the h264 frames. We can do that
        too.

        This patch also changes RELEASE_ASSERT() when creating the parsers
        to GLib criticals.

        * platform/graphics/gstreamer/mse/AppendPipeline.cpp:
        (WebCore::createOptionalParserForFormat):

2018-10-02  Eric Carlson  <eric.carlson@apple.com>

        [MediaStream] RealtimeMediaSource should be able to vend hashed IDs
        https://bugs.webkit.org/show_bug.cgi?id=190142
        <rdar://problem/44911109>

        Reviewed by Youenn Fablet.

        No new tests, covered by existing tests.

        * Modules/mediastream/CanvasCaptureMediaStreamTrack.cpp:
        (WebCore::CanvasCaptureMediaStreamTrack::Source::Source): Update order of parameters passed
        to base class.

        * Modules/mediastream/MediaDevicesRequest.cpp:
        (WebCore::MediaDevicesRequest::start): ASSERT if document.deviceIDHashSalt is not the same
        as passed salt.

        * Modules/mediastream/MediaStreamTrack.cpp:
        (WebCore::MediaStreamTrack::getSettings const): Don't need to hash ID.
        (WebCore::MediaStreamTrack::getCapabilities const): Ditto.
        * Modules/mediastream/MediaStreamTrack.h:
        * Modules/mediastream/MediaStreamTrack.idl:

        * Modules/mediastream/UserMediaRequest.cpp:
        (WebCore::UserMediaRequest::allow): Pass hash salt to createMediaStream.

        * Modules/mediastream/libwebrtc/LibWebRTCPeerConnectionBackend.cpp:
        (WebCore::LibWebRTCPeerConnectionBackend::createReceiverForSource): Update order of parameters passed
        to base class.

        * Modules/webaudio/MediaStreamAudioSource.cpp:
        (WebCore::MediaStreamAudioSource::MediaStreamAudioSource): Ditto.
        * platform/mediastream/MediaConstraints.h:

        * platform/mediastream/RealtimeIncomingAudioSource.cpp:
        (WebCore::RealtimeIncomingAudioSource::RealtimeIncomingAudioSource): Ditto.

        * platform/mediastream/RealtimeIncomingVideoSource.cpp:
        (WebCore::RealtimeIncomingVideoSource::RealtimeIncomingVideoSource): Ditto.

        * platform/mediastream/RealtimeMediaSource.cpp:
        (WebCore::RealtimeMediaSource::RealtimeMediaSource): Calculate hashed ID.
        (WebCore::RealtimeMediaSource::selectSettings): Use m_hashedID.
        (WebCore::RealtimeMediaSource::hashedId const): New.
        (WebCore::RealtimeMediaSource::deviceIDHashSalt const): New.
        * platform/mediastream/RealtimeMediaSource.h:

        * platform/mediastream/RealtimeMediaSourceCenter.cpp:
        (WebCore::RealtimeMediaSourceCenter::createMediaStream): Take hash salt, pass it when creating
        a source.
        (WebCore::RealtimeMediaSourceCenter::getUserMediaDevices): Ditto.
        (WebCore::RealtimeMediaSourceCenter::validateRequestConstraints): Ditto.
        * platform/mediastream/RealtimeMediaSourceCenter.h:

        * platform/mediastream/RealtimeMediaSourceFactory.h:
        * platform/mediastream/RealtimeVideoSource.cpp:
        (WebCore::RealtimeVideoSource::RealtimeVideoSource): Update parameters.
        * platform/mediastream/RealtimeVideoSource.h:

        * platform/mediastream/gstreamer/GStreamerAudioCaptureSource.cpp:
        (WebCore::GStreamerAudioCaptureSource::create): Ditto.
        (WebCore::GStreamerAudioCaptureSource::GStreamerAudioCaptureSource): Ditto.
        * platform/mediastream/gstreamer/GStreamerAudioCaptureSource.h:

        * platform/mediastream/gstreamer/GStreamerVideoCaptureSource.cpp:
        (WebCore::GStreamerVideoCaptureSource::create): Ditto.
        (WebCore::GStreamerVideoCaptureSource::GStreamerVideoCaptureSource): Ditto.
        * platform/mediastream/gstreamer/GStreamerVideoCaptureSource.h:

        * platform/mediastream/gstreamer/MockGStreamerAudioCaptureSource.cpp:
        (WebCore::WrappedMockRealtimeAudioSource::WrappedMockRealtimeAudioSource): Ditto.
        (WebCore::MockRealtimeAudioSource::create): Ditto.
        (WebCore::MockGStreamerAudioCaptureSource::MockGStreamerAudioCaptureSource): Ditto.
        * platform/mediastream/gstreamer/MockGStreamerAudioCaptureSource.h:

        * platform/mediastream/gstreamer/MockGStreamerVideoCaptureSource.cpp:
        (WebCore::MockRealtimeVideoSource::create): Ditto.
        (WebCore::MockGStreamerVideoCaptureSource::MockGStreamerVideoCaptureSource): Ditto.
        * platform/mediastream/gstreamer/MockGStreamerVideoCaptureSource.h:

        * platform/mediastream/mac/AVVideoCaptureSource.h:
        * platform/mediastream/mac/AVVideoCaptureSource.mm:
        (WebCore::AVVideoCaptureSource::create): Ditto.
        (WebCore::AVVideoCaptureSource::AVVideoCaptureSource): Ditto.
        (WebCore::AVVideoCaptureSource::settings): Use hashedId to set device ID.
        (WebCore::AVVideoCaptureSource::capabilities): Ditto.

        * platform/mediastream/mac/CoreAudioCaptureSource.cpp:
        (WebCore::CoreAudioCaptureSource::create): Update parameters.
        (WebCore::CoreAudioCaptureSource::CoreAudioCaptureSource): Ditto.
        (WebCore::CoreAudioCaptureSource::capabilities): Use hashedId to set device ID.
        (WebCore::CoreAudioCaptureSource::settings): Ditto.
        * platform/mediastream/mac/CoreAudioCaptureSource.h:

        * platform/mediastream/mac/DisplayCaptureSourceCocoa.cpp:
        (WebCore::DisplayCaptureSourceCocoa::DisplayCaptureSourceCocoa): Update parameters.
        * platform/mediastream/mac/DisplayCaptureSourceCocoa.h:

        * platform/mediastream/mac/MockRealtimeAudioSourceMac.h:
        * platform/mediastream/mac/MockRealtimeAudioSourceMac.mm:
        (WebCore::MockRealtimeAudioSource::create): Ditto.
        (WebCore::MockRealtimeAudioSourceMac::MockRealtimeAudioSourceMac): Ditto.

        * platform/mediastream/mac/MockRealtimeVideoSourceMac.h:
        * platform/mediastream/mac/MockRealtimeVideoSourceMac.mm:
        (WebCore::MockRealtimeVideoSource::create): Ditto.
        (WebCore::MockRealtimeVideoSourceMac::MockRealtimeVideoSourceMac): Ditto.

        * platform/mediastream/mac/RealtimeMediaSourceCenterMac.cpp:

        * platform/mediastream/mac/ScreenDisplayCaptureSourceMac.h:
        * platform/mediastream/mac/ScreenDisplayCaptureSourceMac.mm:
        (WebCore::ScreenDisplayCaptureSourceMac::create): Ditto.
        (WebCore::ScreenDisplayCaptureSourceMac::ScreenDisplayCaptureSourceMac): Ditto.
        (WebCore::ScreenDisplayCaptureSourceMac::createDisplayStream): Update logging.
        (WebCore::ScreenDisplayCaptureSourceMac::startDisplayStream): Ditto.
        (WebCore::ScreenDisplayCaptureSourceMac::frameAvailable): Ditto.

        * platform/mediastream/mac/WindowDisplayCaptureSourceMac.h:
        * platform/mediastream/mac/WindowDisplayCaptureSourceMac.mm:
        (WebCore::WindowDisplayCaptureSourceMac::create): Update parameters.

        * platform/mock/MockRealtimeAudioSource.cpp:
        (WebCore::MockRealtimeAudioSource::create): Ditto.
        (WebCore::MockRealtimeAudioSource::MockRealtimeAudioSource): Ditto.
        (WebCore::MockRealtimeAudioSource::settings): Use hashedId to set device ID.
        (WebCore::MockRealtimeAudioSource::capabilities): Ditto.
        * platform/mock/MockRealtimeAudioSource.h:

        * platform/mock/MockRealtimeMediaSourceCenter.cpp:

        * platform/mock/MockRealtimeVideoSource.cpp:
        (WebCore::MockRealtimeVideoSource::create): Update parameters.
        (WebCore::MockRealtimeVideoSource::MockRealtimeVideoSource): Ditto.
        (WebCore::MockRealtimeVideoSource::capabilities): Use hashedId to set device ID.
        (WebCore::MockRealtimeVideoSource::settings): Ditto.
        * platform/mock/MockRealtimeVideoSource.h:

2018-10-02  Philippe Normand  <pnormand@igalia.com>

        [GStreamer][playbin3] Stream tag lists leaks
        https://bugs.webkit.org/show_bug.cgi?id=190192

        Reviewed by Xabier Rodriguez-Calvar.

        The gst_stream_get_tags() result is transfer-full, so needs to be adopted to prevent a leak.
        Also check the tags list pointer which might be NULL in some cases.

        * platform/graphics/gstreamer/AudioTrackPrivateGStreamer.cpp:
        (WebCore::AudioTrackPrivateGStreamer::AudioTrackPrivateGStreamer):
        * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
        (WebCore::MediaPlayerPrivateGStreamer::naturalSize const):
        * platform/graphics/gstreamer/VideoTrackPrivateGStreamer.cpp:
        (WebCore::VideoTrackPrivateGStreamer::VideoTrackPrivateGStreamer):

2018-10-01  Michael Catanzaro  <mcatanzaro@igalia.com>

        Unreviewed, fix unused variable in RenderLayer::updateScrollableAreaSet

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::calculateClipRects const): This is a prepare-ChangeLog bug. I don't
        have any changes in this function....

2018-10-02  Alicia Boya García  <aboya@igalia.com>

        [MSE][GStreamer] Make same thread assert non-release
        https://bugs.webkit.org/show_bug.cgi?id=189924

        Reviewed by Xabier Rodriguez-Calvar.

        * platform/graphics/gstreamer/mse/AppendPipeline.cpp:
        (WebCore::AppendPipeline::handleNewAppsinkSample):

2018-10-01  Ryosuke Niwa  <rniwa@webkit.org>

        Add a new variant of serializePreservingVisualAppearance which takes VisibleSelection
        https://bugs.webkit.org/show_bug.cgi?id=190108

        Reviewed by Wenson Hsieh.

        Added a version of serializePreservingVisualAppearance which takes VisibleSelection so that we can avoid creating
        a range simply to get the first node and the end node of the selection later. This simple change also fixes a bug
        demonstrated in editing/pasteboard/paste-table-003.html.

        Test: editing/pasteboard/paste-table-003.html

        * editing/cocoa/EditorCocoa.mm:
        (WebCore::Editor::selectionInHTMLFormat): Adopt the new variant.
        * editing/gtk/EditorGtk.cpp:
        (WebCore::Editor::writeSelectionToPasteboard): Ditto.
        * editing/markup.cpp:
        (WebCore::serializePreservingVisualAppearance): Added.
        * editing/markup.h:
        * editing/wpe/EditorWPE.cpp:
        (WebCore::Editor::writeSelectionToPasteboard): Ditto.
        * loader/archive/cf/LegacyWebArchive.cpp:
        (WebCore::LegacyWebArchive::createFromSelection): Ditto.
        * platform/win/PasteboardWin.cpp:
        (WebCore::Pasteboard::writeSelection): Ditto.

2018-10-01  Alex Christensen  <achristensen@webkit.org>

        Don't read from WebCore's bundle for IDNScriptWhiteList
        https://bugs.webkit.org/show_bug.cgi?id=190157

        Reviewed by Dan Bernstein.

        No change in behavior.  This increases performance by not reading from the WebCore bundle,
        and it makes it so that URL-related functionality can be moved to a place without
        a bundle for resources.

        * Resources/IDNScriptWhiteList.txt: Removed.
        * WebCore.xcodeproj/project.pbxproj:
        * platform/mac/WebCoreNSURLExtras.mm:
        (WebCore::whiteListIDNScripts):
        (WebCore::allCharactersInIDNScriptWhiteList):
        (WebCore::readIDNScriptWhiteListFile): Deleted.

2018-10-01  Alex Christensen  <achristensen@webkit.org>

        Unreviewed, rolling out r236551.

        Fails URL validating too aggressively

        Reverted changeset:

        "URLWithUserTypedString should return nil for URLs deemed to
        be invalid by WebCore::URL"
        https://bugs.webkit.org/show_bug.cgi?id=189979
        https://trac.webkit.org/changeset/236551

2018-10-01  Keith Miller  <keith_miller@apple.com>

        Create a RELEASE_AND_RETURN macro for ExceptionScopes
        https://bugs.webkit.org/show_bug.cgi?id=190163

        Reviewed by Mark Lam.

        The new RELEASE_AND_RETURN does all the work for cases
        where you want to return the result of some expression
        without explicitly checking for an exception. This is
        much like the existing RETURN_IF_EXCEPTION macro.

        No new tests since this is a refactor.

        * bridge/runtime_array.cpp:
        (JSC::RuntimeArray::put):

2018-10-01  Daniel Bates  <dabates@apple.com>

        Attempt to fix the watchOS build after <https://trac.webkit.org/changeset/236678>
        (https://bugs.webkit.org/show_bug.cgi?id=189974)

        Explicitly cast index to unsigned to make the operator[] call unambiguous.

        * platform/ios/KeyEventIOS.mm:
        (WebCore::PlatformKeyboardEvent::disambiguateKeyDownEvent):

2018-10-01  Ryosuke Niwa  <rniwa@webkit.org>

        ASAN failure in ~GCReachableRef()
        https://bugs.webkit.org/show_bug.cgi?id=190113

        Reviewed by Darin Adler.

        The bug was caused by ~GCReachableRef accessing Ref after it had been poisoned for ASAN
        in Ref::leakRef via Ref(Ref&& other). Fixed the bug by using RefPtr instead since that's
        the simplest solution here although we could unpoison Ref temporarily as done in ~Ref.

        * dom/GCReachableRef.h:
        (WebCore::GCReachableRef::GCReachableRef):
        (WebCore::GCReachableRef::~GCReachableRef):
        (WebCore::GCReachableRef::operator-> const):
        (WebCore::GCReachableRef::get const):
        (WebCore::GCReachableRef::operator T& const):
        (WebCore::GCReachableRef::operator! const):
        (WebCore::GCReachableRef::isNull const): Deleted.

2018-10-01  Sihui Liu  <sihui_liu@apple.com>

        Remove StorageProcess
        https://bugs.webkit.org/show_bug.cgi?id=189975

        Reviewed by Geoffrey Garen.

        Clean up code. No behavior change.

        * English.lproj/Localizable.strings:

2018-10-01  Alicia Boya García  <aboya@igalia.com>

        [MSE][GStreamer] Set a minimum sample duration
        https://bugs.webkit.org/show_bug.cgi?id=190125

        Reviewed by Xabier Rodriguez-Calvar.

        The last sample of the audio track in the asset used in this test
        player has a tiny duration (100 ns):

        http://orange-opensource.github.io/hasplayer.js/1.2.0/player.html?url=http://playready.directtaps.net/smoothstreaming/SSWSS720H264/SuperSpeedway_720.ism/Manifest

        So small, we were truncating it to zero. We're not supposed to have
        frames with zero duration. Instead, lets set a minimum frame duration
        for those fringe cases.

        * platform/graphics/gstreamer/MediaSampleGStreamer.cpp:
        (WebCore::MediaSampleGStreamer::MediaSampleGStreamer):

2018-10-01  Daniel Bates  <dabates@apple.com>

        [iOS] Special keys are misidentified in DOM keyboard events
        https://bugs.webkit.org/show_bug.cgi?id=189974

        Reviewed by Wenson Hsieh.

        This patch fixes two issues:
            1. Special keyboard keys would be misidentified in dispatched DOM keyboard events.
            2. DOM keypress events may not be dispatched for some special keys.

        UIKit uses special input strings to identify the Page Up, Page Down, Escape, Up Arrow, Down Arrow,
        Left Arrow, and Right Arrow keys. It also uses ASCII control characters to represent some other
        special keys, including Num Lock / Clear, Home, End, Forward Delete, and F1, ..., F24. We need
        to explicitly handle these special keyboard keys in order to be able to identify the key that
        was pressed as well as to correctly disambiguate a key down to know whether to dispatch a DOM
        keypress event for the key.

        Unlike UIKit, AppKit reserves Unicode Private Use Area (PUA) code points in 0xF700–0xF8FF to
        represent special keyboard keys. This makes it straightforward to disambiguate such keys using
        the input string of the keyboard event alone. To simplify the implementation for iOS
        we normalize the input string be AppKit compatible. See the explaination for WebCore::windowsKeyCodeForCharCode()
        below for more details on why this is done.

        Tests: fast/events/ios/keydown-keyup-arrow-keys-in-non-editable-element.html
               fast/events/ios/keypress-keys-in-non-editable-element.html

        * SourcesCocoa.txt:
        * WebCore.xcodeproj/project.pbxproj:
        Do not use unified source build strategy when building WebEvent.mm as it makes
        use of SoftLinking macros that are incompatible with this strategy.

        * platform/ios/KeyEventIOS.mm:
        (WebCore::windowsKeyCodeForCharCode): Recognize some special AppKit special char codes.
        These special char codes are generated by WebKit. WebKit uses the same special char codes
        as AppKit as a convenience instead of defining our own constants for the same purpose.
        Encoding the special UIKit input strings (e.g. up arrow) as distinct char codes allows us
        to use integer arithmetic and switch blocks to map characters to Windows virtual key
        codes as opposed to special cased branches to perform pointer or string comparisions.
        The latter would be necessary in Modern WebKit in order for key down events to be properly
        disambiguated to dispatch a DOM keypress event because pointers are not perserved, though
        what they point to is, when sending the WebEvent from UIProcess to the WebProcess and
        vice versa.
        (WebCore::isFunctionKey): Convenience function that determines whether the specified char
        code corresponds to a function key on the keyboard. The term "function key" is taken from
        AppKit parlance to describe a special keyboard key. These keys include F1, F2, ..., F24,
        and cursor keys among other special keyboard keys.
        (WebCore::PlatformKeyboardEvent::disambiguateKeyDownEvent): Write in terms of isFunctionKey().
        * platform/ios/PlatformEventFactoryIOS.h:
        * platform/ios/PlatformEventFactoryIOS.mm:
        (WebCore::keyIdentifierForKeyEvent): Remove code to handle UIKit special input strings as
        we now map such special input strings to char codes and hence can use the default code path.
        (WebCore::keyForKeyEvent): Ditto.
        (WebCore::codeForKeyEvent): Remove code to compute the Window virtual key code corresponding
        to a UIKit special key command now that we map such special input strings to char codes and
        subsequently map the char codes to the Windows virtual key code (see -[WebEvent initWithKeyEventType:...]
        constructors). So, we can now use WebEvent.keyCode directly to compute the DOM UIEvents code
        for the event.
        (WebCore::PlatformKeyboardEventBuilder::PlatformKeyboardEventBuilder): Remove code to fix up
        WebEvent.keyCode to account for UIKit special input strings now that we map such special key
        commands to char codes and subsequently map the char codes to the Windows virtual key code (see -[WebEvent initWithKeyEventType:...]
        constructors). So, we can now take WebEvent.keyCode verbatim to be the Window virtual key code.
        (WebCore::convertSpecialKeyToCharCode): Deleted.
        (WebCore::keyCodeForEvent): Deleted.
        * platform/ios/WebEvent.mm:
        (normalizedStringWithAppKitCompatibilityMapping): Added; converts a UIKit character string
        to the corresponding AppKit-compatible one (if not already compatible). See the explaination
        for WebCore::windowsKeyCodeForCharCode() above for more details on why this is done.

        (-[WebEvent initWithKeyEventType:timeStamp:characters:charactersIgnoringModifiers:modifiers:isRepeating:withFlags:keyCode:isTabKey:characterSet:]):
        (-[WebEvent initWithKeyEventType:timeStamp:characters:charactersIgnoringModifiers:modifiers:isRepeating:withFlags:withInputManagerHint:keyCode:isTabKey:]):
        Normalize the character strings to be AppKit compatible.

2018-10-01  Simon Fraser  <simon.fraser@apple.com>

        Optimize RenderStyle::diff() and clean up the code
        https://bugs.webkit.org/show_bug.cgi?id=190104

        Reviewed by Dan Bernstein.

        RenderStyle::changeRequiresLayout() and related should only check values on 
        m_rareNonInheritedData and m_rareInheritedData after checking for pointer equality.
        To reduce the chances of future changes regressing this, move code comparing values
        on StyleRare[Non]InheritedData into dedication functions.
        
        In addition, the transform comparison double-compared the transformOperations,
        because m_rareNonInheritedData->transform != other.m_rareNonInheritedData->transform
        is a deep comparison, and it was followed by *m_rareNonInheritedData->transform != *other.m_rareNonInheritedData->transform.
        Change the first to be a pointer comparison.

        * rendering/style/RenderStyle.cpp:
        (WebCore::rareNonInheritedDataChangeRequiresLayout):
        (WebCore::rareInheritedDataChangeRequiresLayout):
        (WebCore::RenderStyle::changeRequiresLayout const):
        (WebCore::rareNonInheritedDataChangeRequiresLayerRepaint):
        (WebCore::RenderStyle::changeRequiresLayerRepaint const):
        (WebCore::rareNonInheritedDataChangeRequiresRepaint):
        (WebCore::rareInheritedDataChangeRequiresRepaint):
        (WebCore::RenderStyle::changeRequiresRepaint const):

2018-10-01  Alex Christensen  <achristensen@webkit.org>

        URL should not use TextEncoding internally
        https://bugs.webkit.org/show_bug.cgi?id=190111

        Reviewed by Andy Estes.

        That dependency makes it impossible to move or use elsewhere.
        Using TextEncoding was overkill because we know the credentials are UTF-8 percent-encoded in a parsed URL.
        No change in behavior as verified by new API tests.

        * page/SecurityOrigin.cpp:
        * page/csp/ContentSecurityPolicySourceList.cpp:
        * platform/URL.cpp:
        (WebCore::decodeEscapeSequencesFromParsedURL):
        (WebCore::URL::user const):
        (WebCore::URL::pass const):
        (WebCore::URL::fileSystemPath const):
        (WebCore::decodeURLEscapeSequences): Deleted.
        * platform/URL.h:
        * platform/network/DataURLDecoder.cpp:
        * platform/text/TextEncoding.cpp:
        (WebCore::decodeURLEscapeSequences):
        * platform/text/TextEncoding.h:

2018-10-01  Simon Pieters  <zcorpan@gmail.com>

        <form> in quirks mode should have margin-block-end: 1em
        https://bugs.webkit.org/show_bug.cgi?id=157788

        Reviewed by Simon Fraser.

        Change the default style for forms to take writing-mode into account
        in quirks mode. Matches the behavior of Gecko and Edge and the HTML
        standard.

        Spec: https://html.spec.whatwg.org/multipage/rendering.html#flow-content-3

        Test: imported/w3c/web-platform-tests/html/rendering/non-replaced-elements/flow-content-0/form-margin-quirk.html

        * css/quirks.css:
        (form):

2018-10-01  Jeremy Jones  <jeremyj@apple.com>

        Unify implementation in VideoFullscreenInterfaceAVKit
        https://bugs.webkit.org/show_bug.cgi?id=190091
        rdar://problem/44734523

        Reviewed by Jer Noble.

        No new tests because no behavior change.

        Unified code in VideoFullscreenInterfaceAVKit now that new code path is proven and include
        any changes that had been made in the old path.

        * platform/ios/VideoFullscreenInterfaceAVKit.h:
        * platform/ios/VideoFullscreenInterfaceAVKit.mm:
        (-[WebAVPlayerViewControllerDelegate playerViewControllerShouldStartPictureInPictureFromInlineWhenEnteringBackground:]):
        (VideoFullscreenInterfaceAVKit::preparedToExitFullscreen):
        (VideoFullscreenInterfaceAVKit::shouldExitFullscreenWithReason):
        * platform/ios/WebVideoFullscreenControllerAVKit.mm:
        (VideoFullscreenControllerContext::requestUpdateInlineRect):
        (VideoFullscreenControllerContext::requestVideoContentLayer):
        (VideoFullscreenControllerContext::returnVideoContentLayer):
        (VideoFullscreenControllerContext::didSetupFullscreen):
        (VideoFullscreenControllerContext::didExitFullscreen):

2018-10-01  Antoine Quint  <graouts@apple.com>

        [Web Animations] Ensure renderers with accelerated animations have layers
        https://bugs.webkit.org/show_bug.cgi?id=189990

        Reviewed by Simon Fraser.

        In r236501 we added code that would make a RenderBox and a RenderInline query the document timeline for whether a given element has
        accelerated animations running on it. Since the calls to requiresLayer() are in a hot path, we instead keep a list of elements with
        exclusively accelerated animations running.

        No new tests, this is already covered by webanimations/accelerated-animation-with-delay.html and webanimations/opacity-animation-yields-compositing-span.html
        which respectively check that we can apply an accelerated animation to a non-positioned block and an inline element.

        * animation/AnimationTimeline.h:
        * animation/DocumentTimeline.cpp:
        (WebCore::DocumentTimeline::detachFromDocument):
        (WebCore::DocumentTimeline::animationWasAddedToElement):
        (WebCore::DocumentTimeline::animationWasRemovedFromElement):
        (WebCore::DocumentTimeline::animationAcceleratedRunningStateDidChange):
        (WebCore::DocumentTimeline::updateListOfElementsWithRunningAcceleratedAnimationsForElement): Iterate over an element's animations to determine
        whether all of its animations are running accelerated, then update the HashSet containing elements running accelerated animations to remove or
        add this element.
        (WebCore::DocumentTimeline::runningAnimationsForElementAreAllAccelerated const): Make a simple contains() call on the HashSet containing elements
        running accelerated animations.
        * animation/DocumentTimeline.h:
        * animation/KeyframeEffectReadOnly.cpp:
        (WebCore::KeyframeEffectReadOnly::updateAcceleratedAnimationState):
        (WebCore::KeyframeEffectReadOnly::applyPendingAcceleratedActions):
        * rendering/RenderBoxModelObject.h:

2018-10-01  Alicia Boya García  <aboya@igalia.com>

        [GStreamer] Fix abort in gst_sample_get_info()
        https://bugs.webkit.org/show_bug.cgi?id=190135

        Reviewed by Philippe Normand.

        A flush can occur before any frame has finished decoding -- especially
        in tests, where actions on the player often occur in quick succession.

        Therefore, the code must not assume by the time a flush occurs any
        frame has reached the sink. This patch fixes a case when such wrong
        assumption was causing gst_sample_get_info() to abort (crashing
        WebKit).

        * platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
        (WebCore::MediaPlayerPrivateGStreamerBase::flushCurrentBuffer):
        (WebCore::MediaPlayerPrivateGStreamerBase::createGLAppSink):

2018-10-01  Olivier Blin  <olivier.blin@softathome.com>

        [WPE] fix buffer over-read in RenderThemeWPE::mediaControlsStyleSheet()
        https://bugs.webkit.org/show_bug.cgi?id=190139

        Reviewed by Michael Catanzaro.

        Like done upstream for EFL in r210213
        https://bugs.webkit.org/show_bug.cgi?id=166622

        This has been detected by a charactersAreAllASCII() assert failure.

        This is because ASCIILiteral() is wrongly used in mediaControlsStyleSheet().
        mediaControlsBaseUserAgentStyleSheet is a char array, not a null-terminated string.
        It is thus incorrect to use StringImpl::createFromLiteral() that calls
        strlen() to get the string length.

        The String::ConstructFromLiteral constructor can not be used, since it
        skips the last character.

        * platform/wpe/RenderThemeWPE.cpp:
        (WebCore::RenderThemeWPE::mediaControlsStyleSheet):
        Explicitely pass the size to the String constructor.

2018-10-01  Rob Buis  <rbuis@igalia.com>

        Align XMLHttpRequest's overrideMimeType() with the standard
        https://bugs.webkit.org/show_bug.cgi?id=169276

        Reviewed by Chris Dumez.

        Implement the overrideMimeType() as specified in that standard, i.e.
        add a check that the passed mime type is valid and if not fallback
        to application/octet-stream.

        In order for this patch to have any effect, I went ahead and
        made an improvement to the ContentType parsing, parseContentType now
        will reject mime types that do not match the type / subtype format, I
        believe this is required by both RFC2045 and mimesniff specs.

        This behavior matches Chrome and Firefox.

        Test: web-platform-tests/xhr/overridemimetype-invalid-mime-type.htm

        * platform/network/ParsedContentType.cpp:
        (WebCore::parseContentType):
        * xml/XMLHttpRequest.cpp:
        (WebCore::XMLHttpRequest::overrideMimeType):


2018-10-01  Chris Dumez  <cdumez@apple.com>

        Make crossOriginObject.then undefined for promises
        https://bugs.webkit.org/show_bug.cgi?id=190094

        Reviewed by Darin Adler.

        Make crossOriginObject.then undefined for promises. This allows promises to work better with cross-origin WindowProxy
        and Location objects.

        Specification:
        - https://github.com/whatwg/html/pull/3242
        - https://github.com/whatwg/dom/issues/536

        This aligns our behavior with Blink and Gecko.

        No new tests, rebaselined existing test.

        * bindings/js/JSDOMWindowCustom.cpp:
        (WebCore::jsDOMWindowGetOwnPropertySlotRestrictedAccess):
        (WebCore::addCrossOriginWindowOwnPropertyNames):
        * bindings/js/JSLocationCustom.cpp:
        (WebCore::getOwnPropertySlotCommon):
        (WebCore::addCrossOriginLocationOwnPropertyNames):

2018-10-01  Xan Lopez  <xan@igalia.com>

        [SOUP] Fix the build for libsoup > 2.61.90
        https://bugs.webkit.org/show_bug.cgi?id=190126

        Reviewed by Michael Catanzaro.

        * platform/network/soup/SocketStreamHandleImplSoup.cpp:

2018-10-01  Alicia Boya García  <aboya@igalia.com>

        [MSE][GStreamer] Reset running time in PlaybackPipeline::flush()
        https://bugs.webkit.org/show_bug.cgi?id=190076

        Reviewed by Philippe Normand.

        Test: media/media-source/media-source-seek-redundant-append.html

        PlaybackPipeline::flush() is called when already enqueued frames are
        appended again. This may be caused by a quality change or just a
        redundant append. Either way, the pipeline has to be flushed and
        playback begin again, but without changing the player position by
        much.

        There are two kinds of time to consider here: stream time (i.e. the
        time of a frame as written in the file, e.g. a frame may have stream
        time 0:01:00), and running time (i.e. how much time since playback
        started should pass before the frame should be played, e.g. if we
        started playing at 0:00:59 that same frame would have a running time
        of just 1 second).

        Notice how running time depends on where and when playback starts.
        Running time can also be optionally resetted after a flush. (This is
        indeed done currently by most demuxers after a seek.)

        Instead of resetting running time, PlaybackPipeline used to modify the
        first GstSegment emitted after the flush. A GstSegment declares the
        mapping between stream time and running time for the following frames.
        There, PlaybackPipeline used to set `base` (the running time at which
        the segment starts) to the position reported by a position query
        (which is stream time).

        This, of course, only worked when playback (or the last seek) started
        at stream time 0:00:00, since that's the only case where running time
        equals stream time. In other cases delays as long as the difference
        between these timelines would appear. This is demonstrated in the
        attached test, where seeks and appends are made in such an order that
        the difference is more than 5 minutes, making the playback stall for
        >5 minutes before playing 1 second of audio.

        This patch fixes the problem by resetting running time with the flush
        and not modifying GstSegment.base anymore (it will be left as zero,
        which is now correct since the running time has been reset).

        * platform/graphics/gstreamer/mse/PlaybackPipeline.cpp:
        (WebCore::PlaybackPipeline::flush):
        (WebCore::segmentFixerProbe): Deleted.

2018-09-30  Ryosuke Niwa  <rniwa@webkit.org>

        Use Position instead of Range in createMarkupInternal
        https://bugs.webkit.org/show_bug.cgi?id=190107

        Reviewed by Darin Adler.

        Use two Position's indicating start and end instead of Range in createMarkupInternal and StylizedMarkupAccumulator
        in order to support copy & paste across shadow boundaries in the bug 157443. This patch also removes the use of
        Range in MarkupAccumulator since all uses of range is via StylizedMarkupAccumulator.

        Also renamed createMarkupInternal to serializePreservingVisualAppearanceInternal to match the rename in r236612.

        * dom/Position.cpp:
        (WebCore::Position::firstNode const):  Added.
        * dom/Position.h:
        * editing/MarkupAccumulator.cpp:
        (WebCore::MarkupAccumulator::MarkupAccumulator): No longer takes Range.
        (WebCore::MarkupAccumulator::appendText): Removed the code to truncate string at the boundary points of the range.
        * editing/MarkupAccumulator.h:
        (WebCore::MarkupAccumulator): Made this class non-copyable.
        * editing/markup.cpp:
        (WebCore::StyledMarkupAccumulator::StyledMarkupAccumulator): Now takes and stores two positions.

        (WebCore::StyledMarkupAccumulator::appendText): Use textContentRespectingRange in the case annotation is disabled
        instead of calling to MarkupAccumulator::appendText, which no longer respects boundary offsets.

        (WebCore::StyledMarkupAccumulator::renderedTextRespectingRange): Renamed from renderedText. Updated to respect
        boundary offsets defined by m_start and m_end Positions instead of m_range Range.

        (WebCore::StyledMarkupAccumulator::textContentRespectingRange): Renamed from stringValueForRange. Ditto.

        (WebCore::StyledMarkupAccumulator::serializeNodes): Now computes startNode and pastEnd nodes from start and end
        Positions. Note that the end position is always the next node in the tree order  for a character node
        and computeNodeAfterPosition returns nullptr for a character data.

        (WebCore::highestAncestorToWrapMarkup): Now takes two positions instead of a range.

        (WebCore::serializePreservingVisualAppearanceInternal): Renamed from createMarkupInternal. Removed the obsolete
        comments which were added for DOMRange in WebKitLegacy.

        (WebCore::serializePreservingVisualAppearance):

        (WebCore::sanitizedMarkupForFragmentInDocument): Create positions instead of a range to pass to
        serializePreservingVisualAppearanceInternal.

        (WebCore::serializeFragment):

        * editing/markup.h:
        * page/PageSerializer.cpp:
        (WebCore::PageSerializer::SerializerMarkupAccumulator): Removed the unnecessary WebCore namespace qualifier.

2018-09-30  Walker Henderson  <wjahenderson@gmail.com>

        AudioNode.connect should return passed destination node
        https://bugs.webkit.org/show_bug.cgi?id=188834

        Reviewed by Eric Carlson.

        No new tests, rebaselined existing test.

        * Modules/webaudio/AudioBasicInspectorNode.cpp:
        (WebCore::AudioBasicInspectorNode::connect): Deleted.
        * Modules/webaudio/AudioBasicInspectorNode.h:
        * Modules/webaudio/AudioNode.cpp:
        * Modules/webaudio/AudioNode.h:
        * Modules/webaudio/AudioNode.idl:

2018-09-30  Eric Carlson  <eric.carlson@apple.com>

        [MediaStream] Clean up RealtimeMediaSource settings change handling
        https://bugs.webkit.org/show_bug.cgi?id=189998
        <rdar://problem/44797884>

        Reviewed by Youenn Fablet.

        No new tests, updated webrtc/video-disabled-black.html.

        * Modules/mediastream/CanvasCaptureMediaStreamTrack.cpp:
        (WebCore::CanvasCaptureMediaStreamTrack::Source::Source):
        (WebCore::CanvasCaptureMediaStreamTrack::Source::settings):
        (WebCore::CanvasCaptureMediaStreamTrack::Source::settingsDidChange):
        (WebCore::CanvasCaptureMediaStreamTrack::Source::canvasResized):
        * Modules/mediastream/CanvasCaptureMediaStreamTrack.h:
        * platform/mediastream/RealtimeIncomingVideoSource.cpp:
        (WebCore::RealtimeIncomingVideoSource::RealtimeIncomingVideoSource):
        (WebCore::RealtimeIncomingVideoSource::settings):
        (WebCore::RealtimeIncomingVideoSource::settingsDidChange):
        * platform/mediastream/RealtimeIncomingVideoSource.h:
        * platform/mediastream/RealtimeMediaSource.cpp:
        (WebCore::RealtimeMediaSource::settingsDidChange):
        (WebCore::RealtimeMediaSource::notifySettingsDidChangeObservers):
        (WebCore::RealtimeMediaSource::setSize):
        (WebCore::RealtimeMediaSource::setFrameRate):
        (WebCore::RealtimeMediaSource::setAspectRatio):
        (WebCore::RealtimeMediaSource::setFacingMode):
        (WebCore::RealtimeMediaSource::setVolume):
        (WebCore::RealtimeMediaSource::setSampleRate):
        (WebCore::RealtimeMediaSource::setSampleSize):
        (WebCore::RealtimeMediaSource::setEchoCancellation):
        * platform/mediastream/RealtimeMediaSource.h:
        * platform/mediastream/gstreamer/GStreamerAudioCaptureSource.cpp:
        (WebCore::GStreamerAudioCaptureSource::settingsDidChange):
        * platform/mediastream/gstreamer/GStreamerVideoCaptureSource.cpp:
        (WebCore::GStreamerVideoCaptureSource::settingsDidChange):
        * platform/mediastream/mac/AVVideoCaptureSource.mm:
        (WebCore::AVVideoCaptureSource::settingsDidChange):
        * platform/mediastream/mac/CoreAudioCaptureSource.cpp:
        (WebCore::CoreAudioCaptureSource::settingsDidChange):
        * platform/mediastream/mac/DisplayCaptureSourceCocoa.cpp:
        (WebCore::DisplayCaptureSourceCocoa::settingsDidChange):
        * platform/mediastream/mac/RealtimeIncomingVideoSourceCocoa.mm:
        (WebCore::RealtimeIncomingVideoSourceCocoa::processNewSample):
        * platform/mock/MockRealtimeAudioSource.cpp:
        (WebCore::MockRealtimeAudioSource::settingsDidChange):
        * platform/mock/MockRealtimeVideoSource.cpp:
        (WebCore::MockRealtimeVideoSource::settingsDidChange):

2018-09-30  Eric Carlson  <eric.carlson@apple.com>

        [MediaStream] Use display-specific capture factories
        https://bugs.webkit.org/show_bug.cgi?id=190043
        <rdar://problem/44834412>

        Reviewed by Youenn Fablet.

        No new tests, no change in functionality.

        * Sources.txt:
        * WebCore.xcodeproj/project.pbxproj:
        * platform/mediastream/RealtimeMediaSource.cpp:
        (WebCore::RealtimeMediaSource::AudioCaptureFactory::~AudioCaptureFactory): Deleted.
        (WebCore::RealtimeMediaSource::VideoCaptureFactory::~VideoCaptureFactory): Deleted.
        * platform/mediastream/RealtimeMediaSource.h:
        * platform/mediastream/RealtimeMediaSourceCenter.cpp:
        (WebCore::RealtimeMediaSourceCenter::createMediaStream):
        (WebCore::RealtimeMediaSourceCenter::getDisplayMediaDevices):
        * platform/mediastream/RealtimeMediaSourceCenter.h:
        (WebCore::RealtimeMediaSourceCenter::setAudioFactory):
        (WebCore::RealtimeMediaSourceCenter::unsetAudioFactory):
        * platform/mediastream/RealtimeMediaSourceFactory.cpp: Added.
        (WebCore::SingleSourceFactory::setActiveSource):
        (WebCore::SingleSourceFactory::unsetActiveSource):
        * platform/mediastream/RealtimeMediaSourceFactory.h: Added.
        (WebCore::SingleSourceFactory::activeSource):
        (WebCore::VideoCaptureFactory::setVideoCapturePageState):
        (WebCore::DisplayCaptureFactory::setDisplayCapturePageState):
        * platform/mediastream/gstreamer/GStreamerVideoCaptureSource.cpp:
        (WebCore::libWebRTCVideoCaptureSourceFactory):
        (WebCore::libWebRTCDisplayCaptureSourceFactory):
        (WebCore::GStreamerVideoCaptureSource::factory):
        (WebCore::GStreamerVideoCaptureSource::displayFactory):
        * platform/mediastream/gstreamer/GStreamerVideoCaptureSource.h:
        * platform/mediastream/gstreamer/RealtimeMediaSourceCenterLibWebRTC.cpp:
        (WebCore::RealtimeMediaSourceCenterLibWebRTC::audioCaptureSourceFactory):
        (WebCore::RealtimeMediaSourceCenterLibWebRTC::audioFactory):
        (WebCore::RealtimeMediaSourceCenterLibWebRTC::videoFactory):
        (WebCore::RealtimeMediaSourceCenterLibWebRTC::displayCaptureFactory):
        * platform/mediastream/gstreamer/RealtimeMediaSourceCenterLibWebRTC.h:
        * platform/mediastream/mac/AVVideoCaptureSource.mm:
        (WebCore::AVVideoCaptureSource::~AVVideoCaptureSource):
        (WebCore::AVVideoCaptureSource::setupCaptureSession):
        * platform/mediastream/mac/CoreAudioCaptureSource.cpp:
        (WebCore::CoreAudioCaptureSource::factory):
        * platform/mediastream/mac/CoreAudioCaptureSource.h:
        * platform/mediastream/mac/RealtimeMediaSourceCenterMac.cpp:
        (WebCore::RealtimeMediaSourceCenterMac::videoCaptureSourceFactory):
        (WebCore::RealtimeMediaSourceCenterMac::displayCaptureSourceFactory):
        (WebCore::RealtimeMediaSourceCenterMac::audioCaptureSourceFactory):
        (WebCore::RealtimeMediaSourceCenterMac::audioFactory):
        (WebCore::RealtimeMediaSourceCenterMac::videoFactory):
        (WebCore::RealtimeMediaSourceCenterMac::displayCaptureFactory):
        * platform/mediastream/mac/RealtimeMediaSourceCenterMac.h:
        * platform/mock/MockRealtimeAudioSource.cpp:
        (WebCore::MockRealtimeAudioSource::~MockRealtimeAudioSource):
        (WebCore::MockRealtimeAudioSource::startProducingData):
        (): Deleted.
        (WebCore::mockAudioCaptureSourceFactory): Deleted.
        (WebCore::MockRealtimeAudioSource::factory): Deleted.
        * platform/mock/MockRealtimeAudioSource.h:
        * platform/mock/MockRealtimeMediaSourceCenter.cpp:
        (WebCore::MockRealtimeVideoSourceFactory::setVideoCapturePageState):
        (WebCore::MockRealtimeMediaSourceCenter::audioFactory):
        (WebCore::MockRealtimeMediaSourceCenter::videoFactory):
        (WebCore::MockRealtimeMediaSourceCenter::displayCaptureFactory):
        * platform/mock/MockRealtimeMediaSourceCenter.h:
        * platform/mock/MockRealtimeVideoSource.cpp:
        (): Deleted.
        (WebCore::MockRealtimeVideoSourceFactory::setVideoCapturePageState): Deleted.
        (WebCore::mockVideoCaptureSourceFactory): Deleted.
        (WebCore::MockRealtimeVideoSource::factory): Deleted.
        * platform/mock/MockRealtimeVideoSource.h:

2018-09-29  Oriol Brufau  <obrufau@igalia.com>

        [css-grid] Properly align items next to collapsed tracks with gutters
        https://bugs.webkit.org/show_bug.cgi?id=190089

        Reviewed by Manuel Rego Casasnovas.

        gridAreaPositionForInFlowChild could return a wrong end position for
        grid items adjacent to a collapsed track, because it didn't take into
        account that gutters collapse in that case. Therefore, "center" or
        "end" alignments displayed the item at the wrong position.

        Test: imported/w3c/web-platform-tests/css/css-grid/alignment/grid-gutters-013.html

        * rendering/RenderGrid.cpp:
        (WebCore::RenderGrid::gridAreaPositionForInFlowChild const):

2018-09-29  Alicia Boya García  <aboya@igalia.com>

        [GStreamer][MSE] Use GObject for GST_TRACE_OBJECT
        https://bugs.webkit.org/show_bug.cgi?id=190045

        Reviewed by Philippe Normand.

        Passing a non-GObject object to GST_TRACE_OBJECT() can be
        theoretically misunderstood by the GStreamer logging function, so this
        patch avoids that.

        * platform/graphics/gstreamer/mse/AppendPipeline.cpp:
        (WebCore::AppendPipeline::appsrcEndOfAppendCheckerProbe):
        (WebCore::AppendPipeline::handleEndOfAppend):
        (WebCore::AppendPipeline::consumeAppsinkAvailableSamples):
        (WebCore::AppendPipeline::pushNewBuffer):

2018-09-28  Zamiul Haque  <zhaque@apple.com>

        Angled gradient backgrounds in body render vertically when body height is 0
        https://bugs.webkit.org/show_bug.cgi?id=177232
        <rdar://problem/34548230>.

        Reviewed by Tim Horton.

        Specifically, gradients displayed at an angle (ie. 45 degrees) are rendered
        as if they are vertical when the body tag containing the gradient
        has a height of 0. Other browsers do not render under these circumstances,
        so WebKit was modified to follow in suit. The problem was due to layout sizes for
        fill tiles being calculated with a minimum height of 1px. A simple change of the
        minimum height and width to 0px was enough to bring about the desired behavior.

        Tests: angled-background-repeating-gradient-rendering-vertical.html

        * rendering/RenderBoxModelObject.cpp:
        (WebCore::RenderBoxModelObject::calculateFillTileSize const):

2018-09-28  Wenson Hsieh  <wenson_hsieh@apple.com>

        No DOM API to instantiate an attachment for an img element
        https://bugs.webkit.org/show_bug.cgi?id=189934
        <rdar://problem/44743222>

        Reviewed by Ryosuke Niwa.

        Adds support for HTMLAttachmentElement.getAttachmentIdentifier, a function that internal WebKit clients can use
        to ensure that an image element is backed by a unique _WKAttachment. See below for more details.

        Tests:  WKAttachmentTests.AddAttachmentToConnectedImageElement
                WKAttachmentTests.ChangeFileWrapperForPastedImage
                WKAttachmentTests.ConnectImageWithAttachmentToDocument

        * dom/Document.cpp:
        (WebCore::Document::registerAttachmentIdentifier):

        Add a new hook to register an empty _WKAttachment in the UI process with a given identifier. Used when creating
        a new empty attachment to back an image element.

        * dom/Document.h:
        * editing/Editor.cpp:
        (WebCore::Editor::registerAttachmentIdentifier):
        (WebCore::Editor::notifyClientOfAttachmentUpdates):
        * editing/Editor.h:
        * html/HTMLAttachmentElement.cpp:
        (WebCore::HTMLAttachmentElement::getAttachmentIdentifier):

        Creates an attachment element to back the image element, if an attachment does not already exist, and returns
        the unique identifier. This also causes an empty corresponding _WKAttachment to be created in the client, whose
        file wrapper determines the contents of the image.

        (WebCore::HTMLAttachmentElement::ensureUniqueIdentifier):
        (WebCore::HTMLAttachmentElement::hasEnclosingImage const):
        (WebCore::HTMLAttachmentElement::updateEnclosingImageWithData):

        Add a helper that updates the source of the enclosing image element given a content type and image data, by
        creating a new blob and blob URL.

        * html/HTMLAttachmentElement.h:
        * html/HTMLAttachmentElement.idl:
        * html/HTMLImageElement.idl:

        Rename webkitAttachmentIdentifier to just attachmentIdentifier.

        * page/EditorClient.h:
        (WebCore::EditorClient::registerAttachmentIdentifier):
        (WebCore::EditorClient::didInsertAttachmentWithIdentifier):

2018-09-28  Chris Dumez  <cdumez@apple.com>

        The return value of an OnBeforeUnloadEventHandler should always be coerced into a DOMString
        https://bugs.webkit.org/show_bug.cgi?id=190090

        Reviewed by Ryosuke Niwa.

        The return value of an OnBeforeUnloadEventHandler should always be coerced into a DOMString:
        - https://html.spec.whatwg.org/#onbeforeunloadeventhandler
        - https://html.spec.whatwg.org/#the-event-handler-processing-algorithm (Step 5)

        In particular, this means that returning false in an OnBeforeUnloadEventHandler should NOT
        cancel the event when the event is a CustomEvent (and not a BeforeUnloadEvent). This is
        because the return value cannot be false at:
        - https://html.spec.whatwg.org/#the-event-handler-processing-algorithm (Step 5. Otherwise case).

        No new tests, rebaselined existing test.

        * bindings/js/JSEventListener.cpp:
        (WebCore::JSEventListener::handleEvent):

2018-09-28  Simon Fraser  <simon.fraser@apple.com>

        RenderLayer::removeOnlyThisLayer() should not call updateLayerPositions()
        https://bugs.webkit.org/show_bug.cgi?id=190093

        Reviewed by Dean Jackson and Zalan Bujtas.
        
        It's wrong for RenderLayer::removeOnlyThisLayer() to call updateLayerPositions(),
        because this is called at style update time, and layout will be stale.
        
        It was added (see webkit.org/b/25252) so that opacity changes, which can destroy layers, correctly update
        descendants. However, RenderStyle::changeRequiresLayout() checks for opacity <=> no opacity
        changes and triggers layout accordingly, which will result in a full post-layout
        updateLayerPositions().
        
        This also revealed that changes to the "isolate" property fail to trigger any kind of style recalc or layout;
        we need it to trigger layout (for now) because it affects z-order.

        Covered by existing tests.

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::removeOnlyThisLayer):
        * rendering/style/RenderStyle.cpp:
        (WebCore::RenderStyle::changeRequiresLayout const):

2018-09-28  Jiewen Tan  <jiewen_tan@apple.com>

        [WebAuthN] Polish WebAuthN auto-test environment
        https://bugs.webkit.org/show_bug.cgi?id=189283
        <rdar://problem/44117828>

        Reviewed by Chris Dumez.

        This patch removes the old mocking mechanism.

        Tests: http/wpt/webauthn/public-key-credential-create-with-invalid-parameters.https.html
               http/wpt/webauthn/public-key-credential-get-with-invalid-parameters.https.html
               http/wpt/webauthn/public-key-credential-same-origin-with-ancestors.https.html

        * DerivedSources.make:
        * WebCore.xcodeproj/project.pbxproj:
        * testing/Internals.cpp:
        (WebCore::Internals::Internals):
        (WebCore::Internals::mockAuthenticatorCoordinator const): Deleted.
        * testing/Internals.h:
        * testing/Internals.idl:
        * testing/MockAuthenticatorCoordinator.cpp: Removed.
        * testing/MockAuthenticatorCoordinator.h: Removed.
        * testing/MockAuthenticatorCoordinator.idl: Removed.

2018-09-28  Jer Noble  <jer.noble@apple.com>

        Refactoring: eliminate raw pointer usage in Fullscreen code
        https://bugs.webkit.org/show_bug.cgi?id=188747
        <rdar://problem/43541164>

        Reviewed by Alex Christensen.

        Two sources of raw pointers in the Fullscreen code:
        - Model classes (PlaybackSessionModel and VideoFullscreenModel) aren't ref-able, so
          they are passed around as raw references.
        - Observer classes (PlaybackSessionModelClient and VideoFullscreenModelClient, and
          VideoFullscreenChangeObserver) are also passed around as raw pointers, but shouldn't
          be ref-able.

        Make Model classes ref-able by adding ref() and deref() which call virtual refModel and
        derefModel methods, overridden by implementing subclasses. Make every concrete observer
        inherit from CanMakeWeakPtr, and every registration method take WeakPtr wrappers around
        the client interface.

        Since every Interface class now holds a strong reference to its Model classes, and each
        Model class holds a weak reference to all its clients, no explicit invalidate() method
        is necessary.

        Notes:

        - Since the weak pointer methods need to be able to downcast to the abstract base class,
          observers need to inherit publically (rather than privately) from those base classes.
        - Media element Models should compose EventListener rather than inheriting from it, since
          EventListener has its own RefCount.
        - WeakPtrs can't be held in HashSets (because they change value, and therefore hash, when
          their underlying object is destroyed), so clients should be stored in a Vector instead.
        - Interfaces should be given all required Refs at creation time, so that they can store
          those parameters as Refs instead of RefPtrs.

        * platform/cocoa/PlaybackSessionInterface.h:
        (WebCore::PlaybackSessionInterface::~PlaybackSessionInterface): Deleted.
        * platform/cocoa/PlaybackSessionModel.h:
        (WebCore::PlaybackSessionModel::ref):
        (WebCore::PlaybackSessionModel::deref):
        (WebCore::PlaybackSessionModel::~PlaybackSessionModel): Deleted.
        * platform/cocoa/PlaybackSessionModelMediaElement.h:
        * platform/cocoa/PlaybackSessionModelMediaElement.mm:
        (WebCore::PlaybackSessionModelMediaElement::PlaybackSessionModelMediaElement):
        (WebCore::PlaybackSessionModelMediaElement::~PlaybackSessionModelMediaElement):
        (WebCore::PlaybackSessionModelMediaElement::setMediaElement):
        (WebCore::PlaybackSessionModelMediaElement::updateForEventName):
        (WebCore::PlaybackSessionModelMediaElement::addClient):
        (WebCore::PlaybackSessionModelMediaElement::removeClient):
        (WebCore::PlaybackSessionModelMediaElement::updateMediaSelectionOptions):
        (WebCore::PlaybackSessionModelMediaElement::updateMediaSelectionIndices):
        (WebCore::PlaybackSessionModelMediaElement::handleEvent): Deleted.
        * platform/cocoa/VideoFullscreenChangeObserver.h:
        (WebCore::VideoFullscreenChangeObserver::~VideoFullscreenChangeObserver): Deleted.
        * platform/cocoa/VideoFullscreenModel.h:
        (WebCore::VideoFullscreenModel::ref):
        (WebCore::VideoFullscreenModel::deref):
        (WebCore::VideoFullscreenModel::~VideoFullscreenModel): Deleted.
        * platform/cocoa/VideoFullscreenModelVideoElement.h:
        * platform/cocoa/VideoFullscreenModelVideoElement.mm:
        (VideoFullscreenModelVideoElement::VideoFullscreenModelVideoElement):
        (VideoFullscreenModelVideoElement::setVideoElement):
        (VideoFullscreenModelVideoElement::addClient):
        (VideoFullscreenModelVideoElement::removeClient):
        (VideoFullscreenModelVideoElement::setHasVideo):
        (VideoFullscreenModelVideoElement::setVideoDimensions):
        (VideoFullscreenModelVideoElement::willEnterPictureInPicture):
        (VideoFullscreenModelVideoElement::didEnterPictureInPicture):
        (VideoFullscreenModelVideoElement::failedToEnterPictureInPicture):
        (VideoFullscreenModelVideoElement::willExitPictureInPicture):
        (VideoFullscreenModelVideoElement::didExitPictureInPicture):
        (VideoFullscreenModelVideoElement::handleEvent): Deleted.
        * platform/ios/PlaybackSessionInterfaceAVKit.h:
        (WebCore::PlaybackSessionInterfaceAVKit::create):
        (WebCore::PlaybackSessionInterfaceAVKit::playbackSessionModel const):
        (): Deleted.
        * platform/ios/PlaybackSessionInterfaceAVKit.mm:
        (WebCore::PlaybackSessionInterfaceAVKit::PlaybackSessionInterfaceAVKit):
        (WebCore::PlaybackSessionInterfaceAVKit::~PlaybackSessionInterfaceAVKit):
        (WebCore::PlaybackSessionInterfaceAVKit::invalidate): Deleted.
        * platform/ios/VideoFullscreenInterfaceAVKit.h:
        * platform/ios/VideoFullscreenInterfaceAVKit.mm:
        (-[WebAVPlayerLayer layoutSublayers]):
        (-[WebAVPlayerLayer resolveBounds]):
        (-[WebAVPlayerLayer setVideoGravity:]):
        (VideoFullscreenInterfaceAVKit::create):
        (VideoFullscreenInterfaceAVKit::VideoFullscreenInterfaceAVKit):
        (VideoFullscreenInterfaceAVKit::~VideoFullscreenInterfaceAVKit):
        (VideoFullscreenInterfaceAVKit::setVideoFullscreenChangeObserver):
        (VideoFullscreenInterfaceAVKit::applicationDidBecomeActive):
        (VideoFullscreenInterfaceAVKit::setupFullscreen):
        (VideoFullscreenInterfaceAVKit::presentingViewController):
        (VideoFullscreenInterfaceAVKit::requestHideAndExitFullscreen):
        (VideoFullscreenInterfaceAVKit::preparedToExitFullscreen):
        (VideoFullscreenInterfaceAVKit::willStartPictureInPicture):
        (VideoFullscreenInterfaceAVKit::didStartPictureInPicture):
        (VideoFullscreenInterfaceAVKit::failedToStartPictureInPicture):
        (VideoFullscreenInterfaceAVKit::willStopPictureInPicture):
        (VideoFullscreenInterfaceAVKit::didStopPictureInPicture):
        (VideoFullscreenInterfaceAVKit::shouldExitFullscreenWithReason):
        (VideoFullscreenInterfaceAVKit::doSetup):
        (VideoFullscreenInterfaceAVKit::setMode):
        (VideoFullscreenInterfaceAVKit::clearMode):
        (VideoFullscreenInterfaceAVKit::setVideoFullscreenModel): Deleted.
        (VideoFullscreenInterfaceAVKit::invalidate): Deleted.
        * platform/ios/WebAVPlayerController.h:
        * platform/ios/WebAVPlayerController.mm:
        (-[WebAVPlayerController delegate]):
        (-[WebAVPlayerController playbackSessionInterface]):
        (-[WebAVPlayerController setPlaybackSessionInterface:]):
        * platform/ios/WebVideoFullscreenControllerAVKit.mm:
        (VideoFullscreenControllerContext::didCleanupFullscreen):
        (VideoFullscreenControllerContext::addClient):
        (VideoFullscreenControllerContext::removeClient):
        (VideoFullscreenControllerContext::willEnterPictureInPicture):
        (VideoFullscreenControllerContext::didEnterPictureInPicture):
        (VideoFullscreenControllerContext::failedToEnterPictureInPicture):
        (VideoFullscreenControllerContext::willExitPictureInPicture):
        (VideoFullscreenControllerContext::didExitPictureInPicture):
        (VideoFullscreenControllerContext::setUpFullscreen):
        * platform/mac/PlaybackSessionInterfaceMac.h:
        * platform/mac/PlaybackSessionInterfaceMac.mm:
        (WebCore::PlaybackSessionInterfaceMac::create):
        (WebCore::PlaybackSessionInterfaceMac::PlaybackSessionInterfaceMac):
        (WebCore::PlaybackSessionInterfaceMac::playbackSessionModel const):
        (WebCore::PlaybackSessionInterfaceMac::rateChanged):
        (WebCore::PlaybackSessionInterfaceMac::beginScrubbing):
        (WebCore::PlaybackSessionInterfaceMac::endScrubbing):
        (WebCore::PlaybackSessionInterfaceMac::setPlayBackControlsManager):
        (WebCore::PlaybackSessionInterfaceMac::updatePlaybackControlsManagerTiming):
        (WebCore::PlaybackSessionInterfaceMac::~PlaybackSessionInterfaceMac): Deleted.
        (WebCore::PlaybackSessionInterfaceMac::invalidate): Deleted.
        * platform/mac/VideoFullscreenInterfaceMac.h:
        (WebCore::VideoFullscreenInterfaceMac::create):
        (WebCore::VideoFullscreenInterfaceMac::videoFullscreenModel const):
        (WebCore::VideoFullscreenInterfaceMac::playbackSessionModel const):
        (WebCore::VideoFullscreenInterfaceMac::videoFullscreenChangeObserver const):
        * platform/mac/VideoFullscreenInterfaceMac.mm:
        (-[WebVideoFullscreenInterfaceMacObjC setUpPIPForVideoView:withFrame:inWindow:]):
        (-[WebVideoFullscreenInterfaceMacObjC boundsDidChangeForVideoViewContainer:]):
        (-[WebVideoFullscreenInterfaceMacObjC pipDidClose:]):
        (-[WebVideoFullscreenInterfaceMacObjC pipActionPlay:]):
        (-[WebVideoFullscreenInterfaceMacObjC pipActionPause:]):
        (-[WebVideoFullscreenInterfaceMacObjC pipActionStop:]):
        (WebCore::VideoFullscreenInterfaceMac::VideoFullscreenInterfaceMac):
        (WebCore::VideoFullscreenInterfaceMac::~VideoFullscreenInterfaceMac):
        (WebCore::VideoFullscreenInterfaceMac::setVideoFullscreenChangeObserver):
        (WebCore::VideoFullscreenInterfaceMac::setMode):
        (WebCore::VideoFullscreenInterfaceMac::clearMode):
        (WebCore::VideoFullscreenInterfaceMac::invalidate):
        (WebCore::VideoFullscreenInterfaceMac::requestHideAndExitPiP):
        (WebCore::VideoFullscreenInterfaceMac::setVideoFullscreenModel): Deleted.
        * platform/mac/WebPlaybackControlsManager.mm:
        (-[WebPlaybackControlsManager seekToTime:toleranceBefore:toleranceAfter:]):
        (-[WebPlaybackControlsManager setCurrentAudioTouchBarMediaSelectionOption:]):
        (-[WebPlaybackControlsManager setCurrentLegibleTouchBarMediaSelectionOption:]):
        (-[WebPlaybackControlsManager togglePlayback]):
        (-[WebPlaybackControlsManager setPlaying:]):
        (-[WebPlaybackControlsManager isPlaying]):
        (-[WebPlaybackControlsManager togglePictureInPicture]):

2018-09-28  Chris Dumez  <cdumez@apple.com>

        Drop support for cross-origin-window-policy header
        https://bugs.webkit.org/show_bug.cgi?id=190081

        Reviewed by Ryosuke Niwa.

        Drop support for cross-origin-window-policy header as this was never enabled and its design has
        some issues we have not resolved. An alternative is being worked on but will be substantially
        different so there is not much value in keeping this code around.

        * bindings/js/JSDOMBindingSecurity.cpp:
        * bindings/js/JSDOMBindingSecurity.h:
        * bindings/js/JSDOMWindowCustom.cpp:
        (WebCore::jsDOMWindowGetOwnPropertySlotRestrictedAccess):
        (WebCore::JSDOMWindow::getOwnPropertySlotByIndex):
        (WebCore::addCrossOriginWindowPropertyNames):
        (WebCore::addScopedChildrenIndexes):
        (WebCore::addCrossOriginWindowOwnPropertyNames):
        (WebCore::JSDOMWindow::getOwnPropertyNames):
        * bindings/js/JSDOMWindowCustom.h:
        * bindings/js/JSRemoteDOMWindowCustom.cpp:
        * bindings/scripts/CodeGeneratorJS.pm:
        (GenerateAttributeGetterBodyDefinition):
        (GenerateAttributeSetterBodyDefinition):
        (GenerateOperationBodyDefinition):
        * bindings/scripts/IDLAttributes.json:
        * dom/Document.cpp:
        (WebCore::Document::canNavigate):
        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::didBeginDocument):
        * page/AbstractDOMWindow.cpp:
        (WebCore::AbstractDOMWindow::AbstractDOMWindow):
        * page/AbstractDOMWindow.h:
        * page/DOMWindow.idl:
        * page/Settings.yaml:
        * platform/network/HTTPParsers.cpp:
        * platform/network/HTTPParsers.h:

2018-09-28  Daniel Bates  <dabates@apple.com>

        [iOS] Allow programmatic focus when hardware keyboard is attached
        https://bugs.webkit.org/show_bug.cgi?id=190017
        <rdar://problem/42270463>

        Reviewed by Wenson Hsieh.

        Add support for checking if the embedding client is WebKitTestRunner and export isDumpRenderTree()
        so that we can make use of it from WebKit. We will make use of these functions to keep the current
        behavior of disallowing programmatic focus when running tests in these apps. This is needed to
        keep testing deterministic. Otherwise, test results would be dependent on whether a hardware
        keyboard is attached. When running tests in Simulator.app the hardware keyboard may also not be
        connected (i.e. Hardware > Keyboard > Connect Hardware Keyboard is disabled).

        * platform/RuntimeApplicationChecks.h:
        * platform/cocoa/RuntimeApplicationChecksCocoa.mm:
        (WebCore::IOSApplication::isWebKitTestRunner): Added.

2018-09-28  Ryosuke Niwa  <rniwa@webkit.org>

        REGRESSION(r236609): API tests for mso list preservation are failing
        https://bugs.webkit.org/show_bug.cgi?id=190095

        Reviewed by Wenson Hsieh.

        The regression was caused by appendNodeToPreserveMSOList called after an early return for not having renderer.
        Clearly, comment & style elements coming from a MS word document wouldn't have a renderer.

        Fixed the bug by changing the order.

        * editing/markup.cpp:
        (WebCore::StyledMarkupAccumulator::traverseNodesForSerialization):

2018-09-28  Ryosuke Niwa  <rniwa@webkit.org>

        Build fix after r236612.

        * platform/win/PasteboardWin.cpp:
        (WebCore::Pasteboard::writeSelection):

2018-09-28  Andy Estes  <aestes@apple.com>

        [Apple Pay] Remove the "in-store" button type
        https://bugs.webkit.org/show_bug.cgi?id=190079

        Reviewed by Tim Horton.

        According to <https://developer.apple.com/design/human-interface-guidelines/apple-pay/buttons-and-marks/buttons/>,
        this button is meant only for certain kinds of native apps. It shouldn't be available on the web.

        Updated http/tests/ssl/applepay/ApplePayButton.html.

        * css/CSSPrimitiveValueMappings.h:
        (WebCore::CSSPrimitiveValue::CSSPrimitiveValue):
        (WebCore::CSSPrimitiveValue::operator ApplePayButtonType const):
        * css/CSSValueKeywords.in:
        * css/parser/CSSParserFastPaths.cpp:
        (WebCore::CSSParserFastPaths::isValidKeywordPropertyAndValue):
        * rendering/RenderThemeCocoa.mm:
        (WebCore::toPKPaymentButtonType):
        * rendering/style/RenderStyleConstants.h:

2018-09-28  Chris Dumez  <cdumez@apple.com>

        document.open() should throw errors for cross-origin calls
        https://bugs.webkit.org/show_bug.cgi?id=189371
        <rdar://problem/44282700>

        Reviewed by Youenn Fablet.

        document.open() / document.write() should throw errors for cross-origin calls as per:
        - https://html.spec.whatwg.org/#document-open-steps (Step 4)

        No new tests, rebaselined existing tests.

        * dom/Document.cpp:
        (WebCore::Document::open):
        (WebCore::Document::write):
        (WebCore::Document::writeln):
        * dom/Document.h:

2018-09-28  Ryosuke Niwa  <rniwa@webkit.org>

        Rename createMarkup to serializePreservingVisualAppearance
        https://bugs.webkit.org/show_bug.cgi?id=190086

        Reviewed by Wenson Hsieh.

        Renamed the function to clarify what it does. Also removed the unused Range::toHTML.

        * dom/Range.cpp:
        (WebCore::Range::toHTML const): Deleted.
        * dom/Range.h:
        * editing/CompositeEditCommand.cpp:
        (WebCore::CompositeEditCommand::moveParagraphs):
        * editing/cocoa/EditorCocoa.mm:
        (WebCore::Editor::selectionInHTMLFormat):
        * editing/gtk/EditorGtk.cpp:
        (WebCore::Editor::writeSelectionToPasteboard):
        * editing/markup.cpp:
        (WebCore::serializePreservingVisualAppearance):
        (WebCore::createMarkup): Deleted.
        * editing/markup.h:
        * editing/wpe/EditorWPE.cpp:
        (WebCore::Editor::writeSelectionToPasteboard):
        * loader/archive/cf/LegacyWebArchive.cpp:
        (WebCore::LegacyWebArchive::create):
        (WebCore::LegacyWebArchive::createFromSelection):
        * platform/win/PasteboardWin.cpp:
        (WebCore::Pasteboard::writeRangeToDataObject):

2018-09-28  Simon Fraser  <simon.fraser@apple.com>

        Remove some unused RenderLayer code
        https://bugs.webkit.org/show_bug.cgi?id=190078

        Reviewed by Zalan Bujtas.

        The 'outOfFlowDescendantContainingBlocks' code was related to the accelerated overflow scrolling code that
        I removed recently.
        
        updateDescendantsLayerListsIfNeeded() is never called.

        * rendering/RenderLayer.cpp:
        (WebCore::RenderLayer::updateDescendantDependentFlags):
        (WebCore::RenderLayer::calculateClipRects const):
        * rendering/RenderLayer.h:

2018-09-28  Commit Queue  <commit-queue@webkit.org>

        Unreviewed, rolling out r236605.
        https://bugs.webkit.org/show_bug.cgi?id=190087

        caused three API test timeouts (Requested by jernoble on
        #webkit).

        Reverted changeset:

        "Refactoring: eliminate raw pointer usage in Fullscreen code"
        https://bugs.webkit.org/show_bug.cgi?id=188747
        https://trac.webkit.org/changeset/236605

2018-09-28  Ryosuke Niwa  <rniwa@webkit.org>

        Simplify StyledMarkupAccumulator::traverseNodesForSerialization
        https://bugs.webkit.org/show_bug.cgi?id=190073

        Reviewed by Antti Koivisto.

        Simplified the range traversal algorithm in traverseNodesForSerialization as it was too complicated
        to support shadow DOM for copy and paste.

        Instead of using NodeTraversal::next to traverse past ancestors and then figuring out which ancestor
        must be closed or to wrap the existing markup with, new code collects the list of ancestors as we
        traverse out of them.

        Also extracted lambdas for generating markup and deciding whether to skip a node as well as keeping
        track of the depth of the current markup. This further reduces the code complexity of the actual
        node traversal algorithm. Keeping track of the depth allows us to now generate ancestor elements'
        closing tags without keeping a stack of ancestor nodes we opened at all times.

        * editing/markup.cpp:
        (WebCore::StyledMarkupAccumulator::traverseNodesForSerialization):

2018-09-27  Ryosuke Niwa  <rniwa@webkit.org>

        Replace every use of Node::offsetInCharacters() by Node::isCharacterDataNode()
        https://bugs.webkit.org/show_bug.cgi?id=190069

        Reviewed by Zalan Bujtas.

        Removed Node::offsetInCharacters() and replaced every use of it by isCharacterDataNode()
        because their implementations are identical.

        Note that offsetInCharacters() sounds like a function which returns some kind of an offset
        but it doesn't. It returns true when called on a CharacterData and false elsewhere.

        * accessibility/AXObjectCache.cpp:
        (WebCore::characterOffsetsInOrder):
        (WebCore::AXObjectCache::startOrEndCharacterOffsetForRange):
        (WebCore::AXObjectCache::characterOffsetFromVisiblePosition):
        * dom/CharacterData.cpp:
        (WebCore::CharacterData::offsetInCharacters const): Deleted.
        * dom/CharacterData.h:
        * dom/Node.cpp:
        (WebCore::Node::offsetInCharacters const): Deleted.
        * dom/Node.h:
        * dom/Position.cpp:
        (WebCore::Position::parentAnchoredEquivalent const):
        * dom/Position.h:
        (WebCore::lastOffsetInNode):
        (WebCore::minOffsetForNode):
        (WebCore::offsetIsBeforeLastNodeOffset):
        * dom/Range.cpp:
        (WebCore::Range::firstNode const):
        (WebCore::Range::pastLastNode const):
        * dom/RangeBoundaryPoint.h:
        (WebCore::RangeBoundaryPoint::setOffset):
        (WebCore::RangeBoundaryPoint::setToEndOfNode):
        * editing/Editing.cpp:
        (WebCore::lastOffsetForEditing):
        * editing/TextIterator.cpp:
        (WebCore::nextInPreOrderCrossingShadowBoundaries):
        (WebCore::TextIterator::node const):
        (WebCore::SimplifiedBackwardsTextIterator::SimplifiedBackwardsTextIterator):
        * page/DOMSelection.cpp:
        (WebCore::DOMSelection::extend):

2018-09-28  Jer Noble  <jer.noble@apple.com>

        Refactoring: eliminate raw pointer usage in Fullscreen code
        https://bugs.webkit.org/show_bug.cgi?id=188747
        <rdar://problem/43541164>

        Reviewed by Alex Christensen.

        Two sources of raw pointers in the Fullscreen code:
        - Model classes (PlaybackSessionModel and VideoFullscreenModel) aren't ref-able, so
          they are passed around as raw references.
        - Observer classes (PlaybackSessionModelClient and VideoFullscreenModelClient, and
          VideoFullscreenChangeObserver) are also passed around as raw pointers, but shouldn't
          be ref-able.

        Make Model classes ref-able by adding ref() and deref() which call virtual refModel and
        derefModel methods, overridden by implementing subclasses. Make every concrete observer
        inherit from CanMakeWeakPtr, and every registration method take WeakPtr wrappers around
        the client interface.

        Since every Interface class now holds a strong reference to its Model classes, and each
        Model class holds a weak reference to all its clients, no explicit invalidate() method
        is necessary.

        Notes:

        - Since the weak pointer methods need to be able to downcast to the abstract base class,
          observers need to inherit publically (rather than privately) from those base classes.
        - Media element Models should compose EventListener rather than inheriting from it, since
          EventListener has its own RefCount.
        - WeakPtrs can't be held in HashSets (because they change value, and therefore hash, when
          their underlying object is destroyed), so clients should be stored in a Vector instead.
        - Interfaces should be given all required Refs at creation time, so that they can store
          those parameters as Refs instead of RefPtrs.

        * platform/cocoa/PlaybackSessionInterface.h:
        (WebCore::PlaybackSessionInterface::~PlaybackSessionInterface): Deleted.
        * platform/cocoa/PlaybackSessionModel.h:
        (WebCore::PlaybackSessionModel::ref):
        (WebCore::PlaybackSessionModel::deref):
        (WebCore::PlaybackSessionModel::~PlaybackSessionModel): Deleted.
        * platform/cocoa/PlaybackSessionModelMediaElement.h:
        * platform/cocoa/PlaybackSessionModelMediaElement.mm:
        (WebCore::PlaybackSessionModelMediaElement::PlaybackSessionModelMediaElement):
        (WebCore::PlaybackSessionModelMediaElement::~PlaybackSessionModelMediaElement):
        (WebCore::PlaybackSessionModelMediaElement::setMediaElement):
        (WebCore::PlaybackSessionModelMediaElement::updateForEventName):
        (WebCore::PlaybackSessionModelMediaElement::addClient):
        (WebCore::PlaybackSessionModelMediaElement::removeClient):
        (WebCore::PlaybackSessionModelMediaElement::updateMediaSelectionOptions):
        (WebCore::PlaybackSessionModelMediaElement::updateMediaSelectionIndices):
        (WebCore::PlaybackSessionModelMediaElement::handleEvent): Deleted.
        * platform/cocoa/VideoFullscreenChangeObserver.h:
        (WebCore::VideoFullscreenChangeObserver::~VideoFullscreenChangeObserver): Deleted.
        * platform/cocoa/VideoFullscreenModel.h:
        (WebCore::VideoFullscreenModel::ref):
        (WebCore::VideoFullscreenModel::deref):
        (WebCore::VideoFullscreenModel::~VideoFullscreenModel): Deleted.
        * platform/cocoa/VideoFullscreenModelVideoElement.h:
        * platform/cocoa/VideoFullscreenModelVideoElement.mm:
        (VideoFullscreenModelVideoElement::VideoFullscreenModelVideoElement):
        (VideoFullscreenModelVideoElement::setVideoElement):
        (VideoFullscreenModelVideoElement::addClient):
        (VideoFullscreenModelVideoElement::removeClient):
        (VideoFullscreenModelVideoElement::setHasVideo):
        (VideoFullscreenModelVideoElement::setVideoDimensions):
        (VideoFullscreenModelVideoElement::willEnterPictureInPicture):
        (VideoFullscreenModelVideoElement::didEnterPictureInPicture):
        (VideoFullscreenModelVideoElement::failedToEnterPictureInPicture):
        (VideoFullscreenModelVideoElement::willExitPictureInPicture):
        (VideoFullscreenModelVideoElement::didExitPictureInPicture):
        (VideoFullscreenModelVideoElement::handleEvent): Deleted.
        * platform/ios/PlaybackSessionInterfaceAVKit.h:
        (WebCore::PlaybackSessionInterfaceAVKit::create):
        (WebCore::PlaybackSessionInterfaceAVKit::playbackSessionModel const):
        (): Deleted.
        * platform/ios/PlaybackSessionInterfaceAVKit.mm:
        (WebCore::PlaybackSessionInterfaceAVKit::PlaybackSessionInterfaceAVKit):
        (WebCore::PlaybackSessionInterfaceAVKit::~PlaybackSessionInterfaceAVKit):
        (WebCore::PlaybackSessionInterfaceAVKit::invalidate): Deleted.
        * platform/ios/VideoFullscreenInterfaceAVKit.h:
        * platform/ios/VideoFullscreenInterfaceAVKit.mm:
        (-[WebAVPlayerLayer layoutSublayers]):
        (-[WebAVPlayerLayer resolveBounds]):
        (-[WebAVPlayerLayer setVideoGravity:]):
        (VideoFullscreenInterfaceAVKit::create):
        (VideoFullscreenInterfaceAVKit::VideoFullscreenInterfaceAVKit):
        (VideoFullscreenInterfaceAVKit::~VideoFullscreenInterfaceAVKit):
        (VideoFullscreenInterfaceAVKit::setVideoFullscreenChangeObserver):
        (VideoFullscreenInterfaceAVKit::applicationDidBecomeActive):
        (VideoFullscreenInterfaceAVKit::setupFullscreen):
        (VideoFullscreenInterfaceAVKit::presentingViewController):
        (VideoFullscreenInterfaceAVKit::requestHideAndExitFullscreen):
        (VideoFullscreenInterfaceAVKit::preparedToExitFullscreen):
        (VideoFullscreenInterfaceAVKit::willStartPictureInPicture):
        (VideoFullscreenInterfaceAVKit::didStartPictureInPicture):
        (VideoFullscreenInterfaceAVKit::failedToStartPictureInPicture):
        (VideoFullscreenInterfaceAVKit::willStopPictureInPicture):
        (VideoFullscreenInterfaceAVKit::didStopPictureInPicture):
        (VideoFullscreenInterfaceAVKit::shouldExitFullscreenWithReason):
        (VideoFullscreenInterfaceAVKit::doSetup):
        (VideoFullscreenInterfaceAVKit::setMode):
        (VideoFullscreenInterfaceAVKit::clearMode):
        (VideoFullscreenInterfaceAVKit::setVideoFullscreenModel): Deleted.
        (VideoFullscreenInterfaceAVKit::invalidate): Deleted.
        * platform/ios/WebAVPlayerController.h:
        * platform/ios/WebAVPlayerController.mm:
        (-[WebAVPlayerController delegate]):
        (-[WebAVPlayerController playbackSessionInterface]):
        (-[WebAVPlayerController setPlaybackSessionInterface:]):
        * platform/ios/WebVideoFullscreenControllerAVKit.mm:
        (VideoFullscreenControllerContext::didCleanupFullscreen):
        (VideoFullscreenControllerContext::addClient):
        (VideoFullscreenControllerContext::removeClient):
        (VideoFullscreenControllerContext::willEnterPictureInPicture):
        (VideoFullscreenControllerContext::didEnterPictureInPicture):
        (VideoFullscreenControllerContext::failedToEnterPictureInPicture):
        (VideoFullscreenControllerContext::willExitPictureInPicture):
        (VideoFullscreenControllerContext::didExitPictureInPicture):
        (VideoFullscreenControllerContext::setUpFullscreen):
        * platform/mac/PlaybackSessionInterfaceMac.h:
        * platform/mac/PlaybackSessionInterfaceMac.mm:
        (WebCore::PlaybackSessionInterfaceMac::create):
        (WebCore::PlaybackSessionInterfaceMac::PlaybackSessionInterfaceMac):
        (WebCore::PlaybackSessionInterfaceMac::playbackSessionModel const):
        (WebCore::PlaybackSessionInterfaceMac::rateChanged):
        (WebCore::PlaybackSessionInterfaceMac::beginScrubbing):
        (WebCore::PlaybackSessionInterfaceMac::endScrubbing):
        (WebCore::PlaybackSessionInterfaceMac::setPlayBackControlsManager):
        (WebCore::PlaybackSessionInterfaceMac::updatePlaybackControlsManagerTiming):
        (WebCore::PlaybackSessionInterfaceMac::~PlaybackSessionInterfaceMac): Deleted.
        (WebCore::PlaybackSessionInterfaceMac::invalidate): Deleted.
        * platform/mac/VideoFullscreenInterfaceMac.h:
        (WebCore::VideoFullscreenInterfaceMac::create):
        (WebCore::VideoFullscreenInterfaceMac::videoFullscreenModel const):
        (WebCore::VideoFullscreenInterfaceMac::playbackSessionModel const):
        (WebCore::VideoFullscreenInterfaceMac::videoFullscreenChangeObserver const):
        * platform/mac/VideoFullscreenInterfaceMac.mm:
        (-[WebVideoFullscreenInterfaceMacObjC setUpPIPForVideoView:withFrame:inWindow:]):
        (-[WebVideoFullscreenInterfaceMacObjC boundsDidChangeForVideoViewContainer:]):
        (-[WebVideoFullscreenInterfaceMacObjC pipDidClose:]):
        (-[WebVideoFullscreenInterfaceMacObjC pipActionPlay:]):
        (-[WebVideoFullscreenInterfaceMacObjC pipActionPause:]):
        (-[WebVideoFullscreenInterfaceMacObjC pipActionStop:]):
        (WebCore::VideoFullscreenInterfaceMac::VideoFullscreenInterfaceMac):
        (WebCore::VideoFullscreenInterfaceMac::~VideoFullscreenInterfaceMac):
        (WebCore::VideoFullscreenInterfaceMac::setVideoFullscreenChangeObserver):
        (WebCore::VideoFullscreenInterfaceMac::setMode):
        (WebCore::VideoFullscreenInterfaceMac::clearMode):
        (WebCore::VideoFullscreenInterfaceMac::invalidate):
        (WebCore::VideoFullscreenInterfaceMac::requestHideAndExitPiP):
        (WebCore::VideoFullscreenInterfaceMac::setVideoFullscreenModel): Deleted.
        * platform/mac/WebPlaybackControlsManager.mm:
        (-[WebPlaybackControlsManager seekToTime:toleranceBefore:toleranceAfter:]):
        (-[WebPlaybackControlsManager setCurrentAudioTouchBarMediaSelectionOption:]):
        (-[WebPlaybackControlsManager setCurrentLegibleTouchBarMediaSelectionOption:]):
        (-[WebPlaybackControlsManager togglePlayback]):
        (-[WebPlaybackControlsManager setPlaying:]):
        (-[WebPlaybackControlsManager isPlaying]):
        (-[WebPlaybackControlsManager togglePictureInPicture]):

2018-09-28  Chris Dumez  <cdumez@apple.com>

        Drop iOS specific quirk in SettingsBase::scriptEnabledChanged()
        https://bugs.webkit.org/show_bug.cgi?id=190077
        <rdar://problem/44812613>

        Reviewed by Zalan Bujtas.

        Drop iOS specific quirk in SettingsBase::scriptEnabledChanged() that would dirty style after the
        "JavaScriptEnabled" setting's state is toggled. I do not see a good reason to do this given that
        scripts would not get executed until a reload.

        If we find out after dropping this that this is actually useful for some reason, then we can
        always bring it back and consider making this non-iOS specific, as well as documenting why this
        it is needed.

        * page/Settings.yaml:
        * page/SettingsBase.cpp:
        (WebCore::SettingsBase::scriptEnabledChanged): Deleted.
        * page/SettingsBase.h:

2018-09-27  Basuke Suzuki  <Basuke.Suzuki@sony.com>

        [Curl] Fix priority issue with multiple cookies with different level of path.
        https://bugs.webkit.org/show_bug.cgi?id=189920

        Reviewed by Fujii Hironori.

        When multiple cookies are stored in the database for same site, the priority of
        multiple cookies which matches path criteria was not defined. The backend
        implementation `sqlite` then returns the first matching result, which is the one
        stored earlier.

        Test: http/tests/cookies/cookie-with-multiple-level-path.html

        * platform/network/curl/CookieJarDB.cpp:
        (WebCore::CookieJarDB::searchCookies):

2018-09-26  Ryosuke Niwa  <rniwa@webkit.org>

        Use enum class in createMarkup arguments
        https://bugs.webkit.org/show_bug.cgi?id=190028

        Reviewed by Wenson Hsieh.

        Replaced enums used by createMarkup with equivalent enum classes: EChildrenOnly with SerializedNodes,
        EAbsoluteURLs with ResolveURLs, and EFragmentSerialization with SerializationSyntax.

        Also replaced the boolean convertBlocksToInlines with an enum class of the same name.

        Finally, renamed the createMarkup variant which doesn't serialize style and used for innerHTML and XMLSerializer
        to serializeFragment.

        * dom/Element.cpp:
        (WebCore::Element::innerHTML const):
        (WebCore::Element::outerHTML const):
        * dom/ShadowRoot.cpp:
        (WebCore::ShadowRoot::innerHTML const):
        * editing/CompositeEditCommand.cpp:
        (WebCore::CompositeEditCommand::moveParagraphs):
        * editing/HTMLInterchange.h:
        (WebCore::AnnotateForInterchange): Renamed from EAnnotateForInterchange.
        * editing/MarkupAccumulator.cpp:
        (WebCore::MarkupAccumulator::MarkupAccumulator):
        (WebCore::MarkupAccumulator::serializeNodes):
        (WebCore::MarkupAccumulator::serializeNodesWithNamespaces):
        (WebCore::MarkupAccumulator::resolveURLIfNeeded const):
        * editing/MarkupAccumulator.h:
        (WebCore::MarkupAccumulator::inXMLFragmentSerialization const):
        * editing/ReplaceRangeWithTextCommand.cpp:
        (WebCore::ReplaceRangeWithTextCommand::inputEventDataTransfer const):
        * editing/ReplaceSelectionCommand.cpp:
        (WebCore::ReplaceSelectionCommand::willApplyCommand):
        * editing/SpellingCorrectionCommand.cpp:
        (WebCore::SpellingCorrectionCommand::inputEventDataTransfer const):
        * editing/cocoa/EditorCocoa.mm:
        (WebCore::Editor::selectionInHTMLFormat):
        * editing/cocoa/WebContentReaderCocoa.mm:
        (WebCore::WebContentMarkupReader::readRTFD):
        (WebCore::WebContentMarkupReader::readRTF):
        * editing/gtk/EditorGtk.cpp:
        (WebCore::Editor::writeImageToPasteboard):
        (WebCore::Editor::writeSelectionToPasteboard):
        * editing/markup.cpp:
        (WebCore::StyledMarkupAccumulator::StyledMarkupAccumulator):
        (WebCore::StyledMarkupAccumulator::serializeNodes):
        (WebCore::StyledMarkupAccumulator::traverseNodesForSerialization):
        (WebCore::highestAncestorToWrapMarkup):
        (WebCore::createMarkupInternal):
        (WebCore::createMarkup):
        (WebCore::sanitizedMarkupForFragmentInDocument):
        (WebCore::serializeFragment): Renamed from createMarkup. This is used for innerHTML and XMLSerializer,
        which faithfully serializes the fragment without any computed style as inline styles.
        (WebCore::documentTypeString):
        (WebCore::createFullMarkup): Deleted two varinats used in WebKitLegacy.
        * editing/markup.h:
        (WebCore::ResolveURLs): Renamed from EAbsoluteURLs.
        (WebCore::ConvertBlocksToInlines): Added.
        (WebCore::SerializedNodes): Renamed from EChildrenOnly.
        (WebCore::SerializationSyntax): Renamed from EFragmentSerialization.
        * editing/wpe/EditorWPE.cpp:
        (WebCore::Editor::writeSelectionToPasteboard):
        * inspector/DOMEditor.cpp:
        * inspector/agents/InspectorDOMAgent.cpp:
        (WebCore::InspectorDOMAgent::getOuterHTML):
        * loader/archive/cf/LegacyWebArchive.cpp:
        (WebCore::LegacyWebArchive::create):
        (WebCore::LegacyWebArchive::createFromSelection):
        * page/PageSerializer.cpp:
        (WebCore::PageSerializer::SerializerMarkupAccumulator::SerializerMarkupAccumulator):
        (WebCore::PageSerializer::serializeFrame):
        * page/win/DragControllerWin.cpp:
        (WebCore::DragController::declareAndWriteDragImage):
        * Source/WebCore/platform/win/PasteboardWin.cpp:
        (WebCore::Pasteboard::writeRangeToDataObject):
        (WebCore::Pasteboard::writeSelection):
        * xml/XMLHttpRequest.cpp:
        (WebCore::XMLHttpRequest::send):
        * xml/XMLSerializer.cpp:
        (WebCore::XMLSerializer::serializeToString):
        * xml/XSLTProcessorLibxslt.cpp:
        (WebCore::xsltStylesheetPointer):
        (WebCore::xmlDocPtrFromNode):

2018-09-27  John Wilander  <wilander@apple.com>

        Resource Load Statistics: Non-redirected top frame navigation should not get captured in statistics
        https://bugs.webkit.org/show_bug.cgi?id=190055
        <rdar://problem/44843460>

        Reviewed by Chris Dumez.

        Test: http/tests/resourceLoadStatistics/do-not-capture-statistics-for-simple-top-navigations.html

        * loader/ResourceLoadStatistics.cpp:
        (WebCore::ResourceLoadStatistics::decode):
            Corrects legacy statistics for frames and triggers a re-classification.

2018-09-27  Jer Noble  <jer.noble@apple.com>

        Unreviewed watchOS build fix; Fix declaration for ports which USE(ENCRYPTED_MEDIA) but don't
        HAVE(AVCONTENTKEYSESSION).

        * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:

2018-09-27  Justin Michaud  <justin_michaud@apple.com>

        Remove duplicate CSS Properties and Values feature on status page
        https://bugs.webkit.org/show_bug.cgi?id=189909

        Reviewed by Simon Fraser.

        Update CSS properties and values api feature in features.json

        * features.json:

2018-09-27  Jer Noble  <jer.noble@apple.com>

        MediaPlayer should have mediaPlayerWaitingForKeyChanged() / bool waitingForKey() accessor
        https://bugs.webkit.org/show_bug.cgi?id=189951

        Reviewed by Eric Carlson.

        In order to implement the "Resume Playback" section of EME, part 4, we need to be able
        to query whether the MediaPlayer is still waiting for a key after attemptToDecrypt()
        has been called. Currently this involves no behavioral changes, as all modern EME ports
        will still just notify the media element that they no longer need keys after one has
        been added, but future ports may be able to wait for multiple keys before reporting
        that it is no longer waiting for keys.

        * html/HTMLMediaElement.cpp:
        (WebCore::HTMLMediaElement::mediaPlayerWaitingForKeyChanged):
        (WebCore::HTMLMediaElement::attemptToResumePlaybackIfNecessary):
        (WebCore::HTMLMediaElement::mediaPlayerWaitingForKey): Deleted.
        * html/HTMLMediaElement.h:
        * platform/graphics/MediaPlayer.cpp:
        (WebCore::MediaPlayer::waitingForKeyChanged):
        (WebCore::MediaPlayer::waitingForKey const):
        (WebCore::MediaPlayer::waitingForKey): Deleted.
        * platform/graphics/MediaPlayer.h:
        (WebCore::MediaPlayerClient::mediaPlayerWaitingForKeyChanged):
        (WebCore::MediaPlayerClient::mediaPlayerWaitingForKey): Deleted.
        * platform/graphics/MediaPlayerPrivate.h:
        (WebCore::MediaPlayerPrivateInterface::waitingForKey const):
        * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.h:
        * platform/graphics/avfoundation/objc/MediaPlayerPrivateAVFoundationObjC.mm:
        (WebCore::MediaPlayerPrivateAVFoundationObjC::shouldWaitForLoadingOfResource):
        (WebCore::MediaPlayerPrivateAVFoundationObjC::attemptToDecryptWithInstance):
        * platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaSourceAVFObjC.h:
        * platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaSourceAVFObjC.mm:
        (WebCore::MediaPlayerPrivateMediaSourceAVFObjC::attemptToDecryptWithInstance):
        (WebCore::MediaPlayerPrivateMediaSourceAVFObjC::waitingForKey const):
        (WebCore::MediaPlayerPrivateMediaSourceAVFObjC::waitingForKeyChanged):
        (WebCore::MediaPlayerPrivateMediaSourceAVFObjC::initializationDataEncountered):
        * platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.h:
        * platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:
        (WebCore::SourceBufferPrivateAVFObjC::didProvideContentKeyRequestInitializationDataForTrackID):
        (WebCore::SourceBufferPrivateAVFObjC::attemptToDecrypt):
        * platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:
        (WebCore::MediaPlayerPrivateGStreamer::handleMessage):
        * platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.cpp:
        (WebCore::MediaPlayerPrivateGStreamerBase::reportWaitingForKey):
        (WebCore::MediaPlayerPrivateGStreamerBase::setWaitingForKey):
        (WebCore::MediaPlayerPrivateGStreamerBase::waitingForKey const):
        * platform/graphics/gstreamer/MediaPlayerPrivateGStreamerBase.h:
        * platform/graphics/gstreamer/eme/WebKitCommonEncryptionDecryptorGStreamer.cpp:
        (webkitMediaCommonEncryptionDecryptSinkEventHandler):

2018-09-27  Alicia Boya García  <aboya@igalia.com>

        [MSE] Fix unwanted sample erase from the decode queue
        https://bugs.webkit.org/show_bug.cgi?id=180643

        Reviewed by Jer Noble.

        Test: media/media-source/media-source-append-acb-no-frame-lost.html

        This bug reproduced when unordered appends were made. For instance, if
        the application appended [0, 10) and then [20, 30), the frame at 20
        would be wrongly discarded from the decode queue.

        Later the application could append [10, 20) and the gap at [20, 21)
        would persist in the decode queue, even if the frame remained in the
        track buffer table.

        Thanks to Daniel Zhang for reporting the issue.

        * Modules/mediasource/SourceBuffer.cpp:
        (WebCore::SourceBuffer::provideMediaData):

2018-09-27  Alex Christensen  <achristensen@webkit.org>

        URLParser should use TextEncoding through an abstract class
        https://bugs.webkit.org/show_bug.cgi?id=190027

        Reviewed by Andy Estes.

        URLParser uses TextEncoding for one call to encode, which is only used for encoding the query of URLs in documents with non-UTF encodings.
        There are 3 call sites that specify the TextEncoding to use from the Document, and even those call sites use a UTF encoding most of the time.
        All other URL parsing is done using a well-optimized path which assumes UTF-8 encoding and uses macros from ICU headers, not a TextEncoding.
        Moving the logic in this way breaks URL and URLParser's dependency on TextEncoding, which makes it possible to use in a lower-level project
        without also moving TextEncoding, TextCodec, TextCodecICU, ThreadGlobalData, and the rest of WebCore and JavaScriptCore.

        There is no observable change in behavior.  There is now one virtual function call in a code path in URLParser that is not performance-sensitive,
        and TextEncodings now have a vtable, which uses a few more bytes of memory total for WebKit.

        * css/parser/CSSParserContext.h:
        (WebCore::CSSParserContext::completeURL const):
        * css/parser/CSSParserIdioms.cpp:
        (WebCore::completeURL):
        * dom/Document.cpp:
        (WebCore::Document::completeURL const):
        * html/HTMLBaseElement.cpp:
        (WebCore::HTMLBaseElement::href const):
        Move the call to encodingForFormSubmission from the URL constructor to the 3 call sites that specify the encoding from the Document.
        * loader/FormSubmission.cpp:
        (WebCore::FormSubmission::create):
        * loader/TextResourceDecoder.cpp:
        (WebCore::TextResourceDecoder::encodingForURLParsing):
        * loader/TextResourceDecoder.h:
        * platform/URL.cpp:
        (WebCore::URL::URL):
        * platform/URL.h:
        (WebCore::URLTextEncoding::~URLTextEncoding):
        * platform/URLParser.cpp:
        (WebCore::URLParser::encodeNonUTF8Query):
        (WebCore::URLParser::copyURLPartsUntil):
        (WebCore::URLParser::URLParser):
        (WebCore::URLParser::parse):
        (WebCore::URLParser::encodeQuery): Deleted.
        A pointer replaces the boolean isUTF8Encoding and the TextEncoding& which had a default value of UTF8Encoding.
        Now the pointer being null means that we use UTF8, and the pointer being non-null means we use that encoding.
        * platform/URLParser.h:
        (WebCore::URLParser::URLParser):
        * platform/text/TextEncoding.cpp:
        (WebCore::UTF7Encoding):
        (WebCore::TextEncoding::encodingForFormSubmissionOrURLParsing const):
        (WebCore::ASCIIEncoding):
        (WebCore::Latin1Encoding):
        (WebCore::UTF16BigEndianEncoding):
        (WebCore::UTF16LittleEndianEncoding):
        (WebCore::UTF8Encoding):
        (WebCore::WindowsLatin1Encoding):
        (WebCore::TextEncoding::encodingForFormSubmission const): Deleted.
        Use NeverDestroyed because TextEncoding now has a virtual destructor.
        * platform/text/TextEncoding.h:
        Rename encodingForFormSubmission to encodingForFormSubmissionOrURLParsing to make it more clear that we are intentionally using it for both.

2018-09-27  John Wilander  <wilander@apple.com>

        Resource Load Statistics: Remove temporary compatibility fix for auto-dismiss popups
        https://bugs.webkit.org/show_bug.cgi?id=189980
        <rdar://problem/44780645>

        Reviewed by Alex Christensen.

        Test: http/tests/storageAccess/deny-storage-access-under-opener-if-auto-dismiss.html

        The change in https://bugs.webkit.org/show_bug.cgi?id=183620 was a temporary
        compatibility fix as explained in:
        https://webkit.org/blog/8311/intelligent-tracking-prevention-2-0/. We should
        remove it.

        Most of these changes remove the parameter isTriggeredByUserGesture since it's no longer needed.

        * loader/ResourceLoadObserver.cpp:
        (WebCore::ResourceLoadObserver::setRequestStorageAccessUnderOpenerCallback):
        (WebCore::ResourceLoadObserver::logUserInteractionWithReducedTimeResolution):
        (WebCore::ResourceLoadObserver::requestStorageAccessUnderOpener):
        (WebCore::ResourceLoadObserver::logWindowCreation): Deleted.
        * loader/ResourceLoadObserver.h:
        * page/DOMWindow.cpp:
        (WebCore::DOMWindow::createWindow):
            Now no longer logs anything to ResourceLoadObserver.

2018-09-27  Ryan Haddad  <ryanhaddad@apple.com>

        Unreviewed, rolling out r236557.

        Really roll out r236557 this time because it breaks internal
        builds.

        Reverted changeset:

        "Add VP8 support to WebRTC"
        https://bugs.webkit.org/show_bug.cgi?id=189976
        https://trac.webkit.org/changeset/236557

2018-09-27  Chris Dumez  <cdumez@apple.com>

        Fragment should be stripped from document URL during document.open() URL propagation
        https://bugs.webkit.org/show_bug.cgi?id=189374
        <rdar://problem/44282736>

        Reviewed by Alex Christensen.

        Strip the Document URL fragment during document.open() URL propagation if the entry document
        is not the current document, as per:
        - https://html.spec.whatwg.org/#document-open-steps (Step 11.2.)

        No new tests, rebaselined existing test.

        * dom/Document.cpp:
        (WebCore::Document::open):

2018-09-27  Youenn Fablet  <youenn@apple.com>

        Add VP8 support to WebRTC
        https://bugs.webkit.org/show_bug.cgi?id=189976

        Reviewed by Eric Carlson.

        Add a runtime flag to control activation of VP8 codec.
        Bind this runtime flag to the video codec factories.
        Test: webrtc/video-mute-vp8.html

        * Modules/mediastream/libwebrtc/LibWebRTCPeerConnectionBackend.cpp:
        (WebCore::createLibWebRTCPeerConnectionBackend):
        * page/RuntimeEnabledFeatures.h:
        (WebCore::RuntimeEnabledFeatures::webRTCVP8CodecEnabled const):
        (WebCore::RuntimeEnabledFeatures::setWebRTCVP8CodecEnabled):
        * platform/mediastream/libwebrtc/LibWebRTCProvider.h:
        * platform/mediastream/libwebrtc/LibWebRTCProviderCocoa.cpp:
        (WebCore::LibWebRTCProviderCocoa::createDecoderFactory):
        (WebCore::LibWebRTCProviderCocoa::createEncoderFactory):
        * testing/Internals.cpp:
        (WebCore::Internals::resetToConsistentState):
        Enable VP8 codec for tests.

2018-09-27  Chris Dumez  <cdumez@apple.com>

        Crash under WebCore::deleteCookiesForHostnames()
        https://bugs.webkit.org/show_bug.cgi?id=190040
        <rdar://problem/38020368>

        Reviewed by Alex Christensen.

        Update NetworkStorageSession::deleteCookiesForHostnames() to properly deal with the fact
        that NSHTTPCookie.domain can return nil.

        * platform/network/cocoa/NetworkStorageSessionCocoa.mm:
        (WebCore::NetworkStorageSession::deleteCookiesForHostnames):

2018-09-27  Youenn Fablet  <youenn@apple.com>

        Use kCVPixelFormatType_420YpCbCr8Planar for capturing frames
        https://bugs.webkit.org/show_bug.cgi?id=190014

        Reviewed by Eric Carlson.

        On Mac, rely on the monoplanar format which can be displayed without any issue.
        Once rendering is fixed, we should change it back to biplanar as it is closer to what libwebrtc consumes.
        Covered by manual testing.

        * platform/mediastream/mac/AVVideoCaptureSource.mm:
        (WebCore::avVideoCapturePixelBufferFormat):
        (WebCore::AVVideoCaptureSource::setSizeAndFrameRateWithPreset):
        (WebCore::AVVideoCaptureSource::setupCaptureSession):
        (WebCore::AVVideoCaptureSource::captureOutputDidOutputSampleBufferFromConnection):

2018-09-27  Andy Estes  <aestes@apple.com>

        [Apple Pay] Support granular errors in PaymentDetailsUpdate
        https://bugs.webkit.org/show_bug.cgi?id=189938

        Reviewed by Youenn Fablet.

        Implemented the shippingAddressErrors, payerErrors, and paymentMethodErrors properties on
        PaymentDetailsUpdate, as specified in the Payment Request API W3C Editor's Draft of
        26 September 2018.

        When these errors are specified in a PaymentDetailsUpdate, map them to PaymentErrors. For
        shippingAddressErrors and payerErrors, we use the "shippingContactInvalid" code and a
        contact field that matches the shippingAddressError or payerError property specified.

        For paymentMethodErrors, we interpret this as a sequence of ApplePayErrors, which are
        converted to PaymentErrors as in Apple Pay JS.

        Tests: http/tests/ssl/applepay/ApplePayShippingAddressChangeEventErrors.https.html
               http/tests/ssl/applepay/ApplePayShippingAddressChangeEventErrorsV3.https.html

        * DerivedSources.make: Removed some tabs and added new .idl files.
        * Modules/applepay/ApplePayError.idl: Moved ApplePayErrorCode and ApplePayErrorContactField
        into their own .idl files so they can be used in MockPaymentError.
        * Modules/applepay/ApplePayErrorCode.h: Added.
        * Modules/applepay/ApplePayErrorCode.idl: Added.
        * Modules/applepay/ApplePayErrorContactField.h: Added.
        * Modules/applepay/ApplePayErrorContactField.idl: Added.

        * Modules/applepay/paymentrequest/ApplePayPaymentHandler.cpp:
        (WebCore::appendShippingContactInvalidError): Appended a "shippingContactInvalid"
        PaymentError to errors if the message is non-null.
        (WebCore::ApplePayPaymentHandler::computeErrors const):
        (WebCore::ApplePayPaymentHandler::detailsUpdated):
        (WebCore::ApplePayPaymentHandler::shippingAddressUpdated): Computed a vector of PaymentErrors
        based on shippingAddressErrors, payerErrors, and paymentMethodErrors.

        * Modules/applepay/paymentrequest/ApplePayPaymentHandler.h:

        * Modules/paymentrequest/PaymentDetailsUpdate.h:
        * Modules/paymentrequest/PaymentDetailsUpdate.idl: Defined shippingAddressErrors,
        payerErrors, and paymentMethodErrors.

        * Modules/paymentrequest/PaymentHandler.h:
        * Modules/paymentrequest/PaymentRequest.cpp:
        (WebCore::PaymentRequest::paymentMethodChanged):
        (WebCore::PaymentRequest::settleDetailsPromise): Passed shippingAddressErrors, payerErrors,
        and paymentMethodErrors to the payment handler.

        * SourcesCocoa.txt:
        * WebCore.xcodeproj/project.pbxproj:

        * testing/MockPaymentCoordinator.cpp:
        (WebCore::MockPaymentCoordinator::completeShippingContactSelection): Stored errors in m_errors.
        * testing/MockPaymentCoordinator.h:
        * testing/MockPaymentCoordinator.idl: Added an errors attribute.

        * testing/MockPaymentError.h: Added.
        * testing/MockPaymentError.idl: Added.

2018-09-27  Alex Christensen  <achristensen@webkit.org>

        URLWithUserTypedString should return nil for URLs deemed to be invalid by WebCore::URL
        https://bugs.webkit.org/show_bug.cgi?id=189979

        Reviewed by Youenn Fablet.

        * platform/mac/WebCoreNSURLExtras.mm:
        (WebCore::URLWithUserTypedString):
        (WebCore::dataForURLComponentType):
        (WebCore::URLByRemovingComponentAndSubsequentCharacter):
        (WebCore::URLByCanonicalizingURL):
        (WebCore::originalURLData):
        (WebCore::userVisibleString):

2018-09-27  Chris Dumez  <cdumez@apple.com>

        document.open() should not propagate URLs to non-fully active documents
        https://bugs.webkit.org/show_bug.cgi?id=189375
        <rdar://problem/44282755>

        Reviewed by Youenn Fablet.

        Update our document.open() to not propagate URLs to non-fully active documents, as per:
        - https://html.spec.whatwg.org/#document-open-steps (Step 11)

        A "fully active" document is defined by at:
        - https://html.spec.whatwg.org/#fully-active

        No new tests, rebaselined existing test.

        * dom/Document.cpp:
        (WebCore::Document::open):
        (WebCore::Document::isFullyActive const):
        * dom/Document.h:
        * dom/Document.idl:

2018-09-27  Alicia Boya García  <aboya@igalia.com>

        [MSE][GStreamer] Use sentinel buffer to detect end of append
        https://bugs.webkit.org/show_bug.cgi?id=189924

        Reviewed by Philippe Normand.

        This patch introduces a new mechanism to detect when an append has
        been consumed completely by the demuxer. It takes advantage of the
        fact that buffer pushing is synchronous: both the appsrc and the
        demuxer live in the same streaming thread. When appsrc pushes a
        buffer, it's actually making a qtdemux function call (it calls its
        "chain" function). The demuxer will return from that call when it has
        finished processing that buffer; only then the control returns to
        appsrc, that can push the next buffer.

        By pushing an additional buffer and capturing it in a probe we can
        detect reliably when the previous buffer has been processed.
        Because the pipeline only has one thread, at this point no more frames
        can arrive to the appsink.

        This replaces the old method of detecting end of append which relied
        on the `need-data` event, which is more difficult to handle correctly
        because it fires whenever the appsrc is empty (or below a given
        level), which also happens when a buffer has not been pushed yet or
        in response to a flush.

        * platform/graphics/gstreamer/mse/AppendPipeline.cpp:
        (WebCore::EndOfAppendMeta::init):
        (WebCore::EndOfAppendMeta::transform):
        (WebCore::EndOfAppendMeta::free):
        (WebCore::AppendPipeline::staticInitialization):
        (WebCore::AppendPipeline::AppendPipeline):
        (WebCore::AppendPipeline::~AppendPipeline):
        (WebCore::AppendPipeline::appsrcEndOfAppendCheckerProbe):
        (WebCore::AppendPipeline::handleApplicationMessage):
        (WebCore::AppendPipeline::handleEndOfAppend):
        (WebCore::AppendPipeline::consumeAppsinkAvailableSamples):
        (WebCore::AppendPipeline::resetPipeline):
        (WebCore::AppendPipeline::pushNewBuffer):
        (WebCore::AppendPipeline::handleAppsrcNeedDataReceived): Deleted.:
        (WebCore::AppendPipeline::handleAppsrcAtLeastABufferLeft): Deleted.
        (WebCore::AppendPipeline::checkEndOfAppend): Deleted.
        (WebCore::AppendPipeline::setAppsrcDataLeavingProbe): Deleted.
        (WebCore::AppendPipeline::removeAppsrcDataLeavingProbe): Deleted.
        (WebCore::AppendPipeline::reportAppsrcAtLeastABufferLeft): Deleted.
        (WebCore::AppendPipeline::reportAppsrcNeedDataReceived): Deleted.
        (WebCore::appendPipelineAppsrcDataLeaving): Deleted.
        (WebCore::appendPipelineAppsrcNeedData): Deleted.
        * platform/graphics/gstreamer/mse/AppendPipeline.h:

2018-09-27  Chris Dumez  <cdumez@apple.com>

        The WebContent process should not process incoming IPC while waiting for a sync IPC reply
        https://bugs.webkit.org/show_bug.cgi?id=184183
        <rdar://problem/36800576>

        Reviewed by Ryosuke Niwa.

        Drop test infrastructure for the DoNotProcessIncomingMessagesWhenWaitingForSyncReply IPC::SendOption
        given that this SendOption was removed from this patch.

        * page/ChromeClient.h:
        * testing/Internals.cpp:
        * testing/Internals.h:
        * testing/Internals.idl:

2018-09-27  Philippe Normand  <pnormand@igalia.com>

        Unreviewed, GTK Ubuntu LTS build fix attempt after r236396.

        * platform/graphics/gstreamer/GStreamerCommon.h:

2018-09-27  Antoine Quint  <graouts@apple.com>

        [Web Animations] Turn Web Animations with CSS integration on
        https://bugs.webkit.org/show_bug.cgi?id=184819
        <rdar://problem/39597337>

        Reviewed by Dean Jackson.

        * page/RuntimeEnabledFeatures.h:

2018-09-26  YUHAN WU  <yuhan_wu@apple.com>

        Video track clone cannot preserve original property
        https://bugs.webkit.org/show_bug.cgi?id=189872
        <rdar://problem/44706579>

        Reviewed by Youenn Fablet.

        Fix the issue that the cloned track created by canvas.captureStream().getVideoTracks() cannot keep some attributes, such as enabled.
        Updated a testcase expected result:
        LayoutTests/imported/w3c/web-platform-tests/mst-content-hint/MediaStreamTrack-contentHint.html

        * Modules/mediastream/CanvasCaptureMediaStreamTrack.cpp:
        (WebCore::CanvasCaptureMediaStreamTrack::CanvasCaptureMediaStreamTrack):
        (WebCore::CanvasCaptureMediaStreamTrack::create):
        (WebCore::CanvasCaptureMediaStreamTrack::clone):
        * Modules/mediastream/CanvasCaptureMediaStreamTrack.h:
        * Modules/mediastream/MediaStreamTrack.h:

2018-09-26  Alex Christensen  <achristensen@webkit.org>

        Unreviewed, rolling out r236524.

        Broke API tests

        Reverted changeset:

        "URLWithUserTypedString should return nil for URLs deemed to
        be invalid by WebCore::URL"
        https://bugs.webkit.org/show_bug.cgi?id=189979
        https://trac.webkit.org/changeset/236524

2018-09-26  Per Arne Vollan  <pvollan@apple.com>

        WebVTT cue alignment broken
        https://bugs.webkit.org/show_bug.cgi?id=190004

        Reviewed by Eric Carlson.

        If the position of the queue is unspecified, the default value of 50 was used, which is incorrect.
        This patch also updates the API according to https://w3c.github.io/webvtt/#the-vttcue-interface.
        The position attribute should not be a double, but either a double or the "auto" keyword. Parts
        of this patch is inspired by the associated code in the Chromium project.

        Test: media/track/track-cue-left-align.html

        * html/track/TextTrackCueGeneric.cpp:
        (WebCore::TextTrackCueGenericBoxElement::applyCSSProperties):
        (WebCore::TextTrackCueGeneric::setPosition):
        * html/track/TextTrackCueGeneric.h:
        * html/track/VTTCue.cpp:
        (WebCore::VTTCueBox::applyCSSProperties):
        (WebCore::VTTCue::initialize):
        (WebCore::VTTCue::position const):
        (WebCore::VTTCue::setPosition):
        (WebCore::VTTCue::textPositionIsAuto const):
        (WebCore::VTTCue::calculateComputedTextPosition const):
        (WebCore::VTTCue::calculateDisplayParameters):
        (WebCore::VTTCue::toJSON const):
        * html/track/VTTCue.h:
        (WebCore::VTTCue::position const): Deleted.
        * html/track/VTTCue.idl:

2018-09-26  James Savage  <james.savage@apple.com>

        Allow override of viewport configuration.
        https://bugs.webkit.org/show_bug.cgi?id=188772.
        <rdar://problem/43538892>.

        Reviewed by Simon Fraser.

        * page/Settings.yaml:
        * page/ViewportConfiguration.cpp:
        (WebCore::ViewportConfiguration::nativeWebpageParameters): Provide a viewport configuration
        similar to width=device-width, with initial scale set to 1.
        * page/ViewportConfiguration.h:

2018-09-26  Alex Christensen  <achristensen@webkit.org>

        URLs with mismatched surrogate pairs in the host should fail to parse
        https://bugs.webkit.org/show_bug.cgi?id=190005

        Reviewed by Chris Dumez.

        Elsewhere in the URLParser, when we encounter mismatched surrogate pairs we use the replacement character,
        but that just fails later on in domainToASCII, so we may as well just fail.
        This behavior matches Chrome, but is unclear in the spec.  There are no valid uses of hosts containing mismatched surrogate pairs.
        Covered by new API tests.

        * platform/URLParser.cpp:
        (WebCore::URLParser::parseHostAndPort):

2018-09-26  Alex Christensen  <achristensen@webkit.org>

        uidna_nameToASCII only needs a buffer capacity of 64
        https://bugs.webkit.org/show_bug.cgi?id=190006

        Reviewed by Chris Dumez.

        This is specified in https://www.unicode.org/reports/tr46/#ToASCII
        This is how Chrome and Firefox also behave with long unicode hosts.

        * platform/URLParser.cpp:
        (WebCore::URLParser::domainToASCII):

2018-09-26  Alex Christensen  <achristensen@webkit.org>

        URLWithUserTypedString should return nil for URLs deemed to be invalid by WebCore::URL
        https://bugs.webkit.org/show_bug.cgi?id=189979
        <rdar://problem/44119696>

        Reviewed by Chris Dumez.

        Covered by an API test.

        * platform/mac/WebCoreNSURLExtras.mm:
        (WebCore::URLWithData):

2018-09-26  Ryosuke Niwa  <rniwa@webkit.org>

        Selection should work across shadow boundary when initiated by a mouse drag
        https://bugs.webkit.org/show_bug.cgi?id=151380
        <rdar://problem/24363872>

        Revert the change that I said I would from r236519.

        * editing/VisibleSelection.cpp:
        (WebCore::VisibleSelection::adjustSelectionToAvoidCrossingShadowBoundaries):

2018-09-26  Commit Queue  <commit-queue@webkit.org>

        Unreviewed, rolling out r236511.
        https://bugs.webkit.org/show_bug.cgi?id=190008

        It is breaking some WebRTC tests (Requested by youenn on
        #webkit).

        Reverted changeset:

        "[MediaStream] Clean up RealtimeMediaSource settings change
        handling"
        https://bugs.webkit.org/show_bug.cgi?id=189998
        https://trac.webkit.org/changeset/236511

2018-09-26  Ryosuke Niwa  <rniwa@webkit.org>

        Selection should work across shadow boundary when initiated by a mouse drag
        https://bugs.webkit.org/show_bug.cgi?id=151380
        <rdar://problem/24363872>

        Reviewed by Antti Koivisto and Wenson Hsieh.

        This patch adds the basic support for selecting content across shadow DOM boundaries to VisibleSelection,
        which is enough to allow users to select content across shadow DOM boundaries via a mouse drag.

        This is the first step in allowing users to select, copy and paste content across shadow DOM boundaries,
        which is a serious user experience regression right now. The new behavior is disabled by default under
        an interal debug feature flag: selectionAcrossShadowBoundariesEnabled.

        Like Chrome, we are not going to support selecting editable content across shadow DOM boundaries since
        we'd have to generalize every editing commands to make that work, and there aren't any HTML editors that
        use shadow DOM boundaries within an editable region yet. For simplicity, we also don't support extending
        a selection out of a shadow root which resides inside an editing region.

        The keyboard based navigation & manipulation of selection as well as allowing copy & paste of content
        across shadow DOM boundaries will be implemented by separate patches. DOMSelection will not expose this new
        behavior either. This is tracked in the spec as https://github.com/w3c/webcomponents/issues/79

        Tests: editing/selection/selection-across-shadow-boundaries-mixed-editability-1.html
               editing/selection/selection-across-shadow-boundaries-mixed-editability-2.html
               editing/selection/selection-across-shadow-boundaries-mixed-editability-3.html
               editing/selection/selection-across-shadow-boundaries-mixed-editability-4.html
               editing/selection/selection-across-shadow-boundaries-mixed-editability-5.html
               editing/selection/selection-across-shadow-boundaries-readonly-1.html
               editing/selection/selection-across-shadow-boundaries-readonly-2.html
               editing/selection/selection-across-shadow-boundaries-readonly-3.html
               editing/selection/selection-across-shadow-boundaries-user-select-all-1.html

        * editing/VisibleSelection.cpp:
        (WebCore::isInUserAgentShadowRootOrHasEditableShadowAncestor): Added.
        (WebCore::VisibleSelection::adjustSelectionToAvoidCrossingShadowBoundaries): When the feature is enabled,
        allow crossing shadow DOM boundaries except when either end is inside an user agent shadow root, or one of
        its shadow includign ancestor is inside an editable region. The latter check is needed to disallow
        an extension of a selection starting in a shadow tree inside a non-editable region inside an editable region
        to outside the editable region. The rest of the editing code is not ready to deal with selection like that.
        * page/Settings.yaml: Added an internal debug feature to enable this new behavior.

2018-09-26  Chris Dumez  <cdumez@apple.com>

        Ignore-opens-during-unload counter of a parent should apply to its children during beforeunload event
        https://bugs.webkit.org/show_bug.cgi?id=189376
        <rdar://problem/44282754>

        Reviewed by Ryosuke Niwa.

        Make sure the Ignore-opens-during-unload counter of a parent stays incremented while we are firing the
        beforeunload event for its descendants, as per:
        - https://html.spec.whatwg.org/multipage/browsing-the-web.html#prompt-to-unload-a-document

        No new tests, rebaselined existing tests.

        * loader/FrameLoader.cpp:
        (WebCore::FrameLoader::shouldClose):
        (WebCore::FrameLoader::dispatchBeforeUnloadEvent):

2018-09-26  Justin Fan  <justin_fan@apple.com>

        WebGL 2: Adding logging to in-progress features
        https://bugs.webkit.org/show_bug.cgi?id=189978

        Reviewed by Jon Lee.

        Logging will allow us to better identify the most in-demand WebGL 2 features and prioritize our efforts.

        No tests as no change in WebGL functionality.

        * html/canvas/WebGL2RenderingContext.cpp: Added logging statement to every function not yet implemented.
        (WebCore::WebGL2RenderingContext::blitFramebuffer):
        (WebCore::WebGL2RenderingContext::framebufferTextureLayer):
        (WebCore::WebGL2RenderingContext::invalidateFramebuffer):
        (WebCore::WebGL2RenderingContext::invalidateSubFramebuffer):
        (WebCore::WebGL2RenderingContext::readBuffer):
        (WebCore::WebGL2RenderingContext::texStorage3D):
        (WebCore::WebGL2RenderingContext::texImage2D):
        (WebCore::WebGL2RenderingContext::texImage3D):
        (WebCore::WebGL2RenderingContext::texSubImage2D):
        (WebCore::WebGL2RenderingContext::texSubImage3D):
        (WebCore::WebGL2RenderingContext::copyTexSubImage3D):
        (WebCore::WebGL2RenderingContext::compressedTexImage2D):
        (WebCore::WebGL2RenderingContext::compressedTexImage3D):
        (WebCore::WebGL2RenderingContext::compressedTexSubImage3D):
        (WebCore::WebGL2RenderingContext::getFragDataLocation):
        (WebCore::WebGL2RenderingContext::uniform1ui):
        (WebCore::WebGL2RenderingContext::uniform2ui):
        (WebCore::WebGL2RenderingContext::uniform3ui):
        (WebCore::WebGL2RenderingContext::uniform4ui):
        (WebCore::WebGL2RenderingContext::uniform1uiv):
        (WebCore::WebGL2RenderingContext::uniform2uiv):
        (WebCore::WebGL2RenderingContext::uniform3uiv):
        (WebCore::WebGL2RenderingContext::uniform4uiv):
        (WebCore::WebGL2RenderingContext::uniformMatrix2x3fv):
        (WebCore::WebGL2RenderingContext::uniformMatrix3x2fv):
        (WebCore::WebGL2RenderingContext::uniformMatrix2x4fv):
        (WebCore::WebGL2RenderingContext::uniformMatrix4x2fv):
        (WebCore::WebGL2RenderingContext::uniformMatrix3x4fv):
        (WebCore::WebGL2RenderingContext::uniformMatrix4x3fv):
        (WebCore::WebGL2RenderingContext::vertexAttribI4i):
        (WebCore::WebGL2RenderingContext::vertexAttribI4iv):
        (WebCore::WebGL2RenderingContext::vertexAttribI4ui):
        (WebCore::WebGL2RenderingContext::vertexAttribI4uiv):
        (WebCore::WebGL2RenderingContext::vertexAttribIPointer):
        (WebCore::WebGL2RenderingContext::drawRangeElements):
        (WebCore::WebGL2RenderingContext::createQuery):
        (WebCore::WebGL2RenderingContext::deleteQuery):
        (WebCore::WebGL2RenderingContext::isQuery):
        (WebCore::WebGL2RenderingContext::beginQuery):
        (WebCore::WebGL2RenderingContext::endQuery):
        (WebCore::WebGL2RenderingContext::getQuery):
        (WebCore::WebGL2RenderingContext::getQueryParameter):
        (WebCore::WebGL2RenderingContext::createSampler):
        (WebCore::WebGL2RenderingContext::deleteSampler):
        (WebCore::WebGL2RenderingContext::isSampler):
        (WebCore::WebGL2RenderingContext::bindSampler):
        (WebCore::WebGL2RenderingContext::samplerParameteri):
        (WebCore::WebGL2RenderingContext::samplerParameterf):
        (WebCore::WebGL2RenderingContext::getSamplerParameter):
        (WebCore::WebGL2RenderingContext::fenceSync):
        (WebCore::WebGL2RenderingContext::isSync):
        (WebCore::WebGL2RenderingContext::deleteSync):
        (WebCore::WebGL2RenderingContext::clientWaitSync):
        (WebCore::WebGL2RenderingContext::waitSync):
        (WebCore::WebGL2RenderingContext::getSyncParameter):
        (WebCore::WebGL2RenderingContext::createTransformFeedback):
        (WebCore::WebGL2RenderingContext::deleteTransformFeedback):
        (WebCore::WebGL2RenderingContext::isTransformFeedback):
        (WebCore::WebGL2RenderingContext::bindTransformFeedback):
        (WebCore::WebGL2RenderingContext::beginTransformFeedback):
        (WebCore::WebGL2RenderingContext::endTransformFeedback):
        (WebCore::WebGL2RenderingContext::transformFeedbackVaryings):
        (WebCore::WebGL2RenderingContext::getTransformFeedbackVarying):
        (WebCore::WebGL2RenderingContext::pauseTransformFeedback):
        (WebCore::WebGL2RenderingContext::resumeTransformFeedback):
        (WebCore::WebGL2RenderingContext::bindBufferBase):
        (WebCore::WebGL2RenderingContext::bindBufferRange):
        (WebCore::WebGL2RenderingContext::getUniformIndices):
        (WebCore::WebGL2RenderingContext::getUniformBlockIndex):
        (WebCore::WebGL2RenderingContext::getActiveUniformBlockParameter):
        (WebCore::WebGL2RenderingContext::getActiveUniformBlockName):
        (WebCore::WebGL2RenderingContext::uniformBlockBinding):

2018-09-26  Eric Carlson  <eric.carlson@apple.com>

        [MediaStream] Clean up RealtimeMediaSource settings change handling
        https://bugs.webkit.org/show_bug.cgi?id=189998
        <rdar://problem/44797884>

        Reviewed by Youenn Fablet.

        No new tests, no change in functionality.

        * platform/mediastream/RealtimeMediaSource.cpp:
        (WebCore::RealtimeMediaSource::notifySettingsDidChangeObservers):
        (WebCore::RealtimeMediaSource::setSize):
        (WebCore::RealtimeMediaSource::setFrameRate):
        (WebCore::RealtimeMediaSource::setAspectRatio):
        (WebCore::RealtimeMediaSource::setFacingMode):
        (WebCore::RealtimeMediaSource::setVolume):
        (WebCore::RealtimeMediaSource::setSampleRate):
        (WebCore::RealtimeMediaSource::setSampleSize):
        (WebCore::RealtimeMediaSource::setEchoCancellation):
        (WebCore::RealtimeMediaSource::settingsDidChange): Deleted.
        * platform/mediastream/RealtimeMediaSource.h:
        * platform/mediastream/mac/AVVideoCaptureSource.mm:
        (WebCore::AVVideoCaptureSource::settingsDidChange):
        * platform/mediastream/mac/CoreAudioCaptureSource.cpp:
        (WebCore::CoreAudioCaptureSource::settingsDidChange):
        * platform/mediastream/mac/DisplayCaptureSourceCocoa.cpp:
        (WebCore::DisplayCaptureSourceCocoa::settingsDidChange):
        * platform/mock/MockRealtimeAudioSource.cpp:
        (WebCore::MockRealtimeAudioSource::settingsDidChange):
        * platform/mock/MockRealtimeVideoSource.cpp:
        (WebCore::MockRealtimeVideoSource::settingsDidChange):

2018-09-26  Antoine Quint  <graouts@apple.com>

        [Web Animations] Ensure renderers with accelerated animations have layers
        https://bugs.webkit.org/show_bug.cgi?id=189990
        <rdar://problem/44791222>

        Reviewed by Zalan Bujtas.

        We have done some work already in webkit.org/b/189784 to prevent never-ending calls to DocumentTimeline::updateAnimations(). This was due to
        the change made for webkit.org/b/186930 where we queued calls to updateAnimations() in KeyframeEffectReadOnly::applyPendingAcceleratedActions()
        while we were waiting for a renderer with a layer backing for a given animation target. Instead of doing this, we now ensure renderers always
        have a layer when they have an accelerated animation applied.

        No new tests, this is already covered by webanimations/accelerated-animation-with-delay.html and webanimations/opacity-animation-yields-compositing-span.html
        which respectively check that we can apply an accelerated animation to a non-positioned block and an inline element.

        * animation/DocumentTimeline.cpp:
        (WebCore::DocumentTimeline::runningAnimationsForElementAreAllAccelerated const): This method should have been marked const all along and it is
        now required so it can be called through RenderBox::requiresLayer() and RenderInline::requiresLayer().
        (WebCore::DocumentTimeline::runningAnimationsForElementAreAllAccelerated): Deleted.
        * animation/DocumentTimeline.h:
        * animation/KeyframeEffectReadOnly.cpp:
        (WebCore::KeyframeEffectReadOnly::applyPendingAcceleratedActions): Stop enqueuing the accelerated actions in case we're lacking a composited renderer
        since this situation should no longer arise.
        * rendering/RenderBox.h: Make requiresLayer() return true if this renderer's element is the target of accelerated animations.
        * rendering/RenderBoxModelObject.cpp:
        (WebCore::RenderBoxModelObject::hasRunningAcceleratedAnimations const): Query the document timeline, if it exists, to check that this renderer's element
        has accelerated animations applied.
        * rendering/RenderBoxModelObject.h:
        * rendering/RenderInline.h: Make requiresLayer() return true if this renderer's element is the target of accelerated animations.

2018-09-25  Eric Carlson  <eric.carlson@apple.com>

        [MediaStream] Add Mac window capture source
        https://bugs.webkit.org/show_bug.cgi?id=189958
        <rdar://problem/44767616>

        Reviewed by Youenn Fablet.

        * SourcesCocoa.txt: Add WindowDisplayCaptureSourceMac.
        * WebCore.xcodeproj/project.pbxproj: Ditto.

        * platform/mediastream/mac/DisplayCaptureManagerCocoa.cpp:
        (WebCore::DisplayCaptureManagerCocoa::captureDevices): Include window "devices".
        (WebCore::DisplayCaptureManagerCocoa::updateWindowCaptureDevices): New.
        (WebCore::DisplayCaptureManagerCocoa::windowCaptureDeviceWithPersistentID): New.
        (WebCore::DisplayCaptureManagerCocoa::captureDeviceWithPersistentID): Include window devices.
        * platform/mediastream/mac/DisplayCaptureManagerCocoa.h:

        * platform/mediastream/mac/DisplayCaptureSourceCocoa.cpp:
        (WebCore::DisplayCaptureSourceCocoa::DisplayCaptureSourceCocoa):
        (WebCore::DisplayCaptureSourceCocoa::settings): Use frameSize, report surface type and 
        logical surface.
        (WebCore::DisplayCaptureSourceCocoa::settingsDidChange): Clear m_lastSampleBuffer when size changes.
        (WebCore::DisplayCaptureSourceCocoa::frameSize const): New, return size() or intrinsic size.
        (WebCore::DisplayCaptureSourceCocoa::setIntrinsicSize): New.
        (WebCore::DisplayCaptureSourceCocoa::emitFrame): generateFrame now returns a CVPixelBuffer
        so derived classes don't have to deal with resizing/transforming.
        * platform/mediastream/mac/DisplayCaptureSourceCocoa.h:

        * platform/mediastream/mac/RealtimeMediaSourceCenterMac.cpp:

        * platform/mediastream/mac/ScreenDisplayCaptureSourceMac.h:
        * platform/mediastream/mac/ScreenDisplayCaptureSourceMac.mm:
        (WebCore::ScreenDisplayCaptureSourceMac::ScreenDisplayCaptureSourceMac): Add fixme.
        (WebCore::ScreenDisplayCaptureSourceMac::createDisplayStream): Update intrinsic size when
        width/height changes.
        (WebCore::ScreenDisplayCaptureSourceMac::generateFrame): Return a CVPixelBuffer.

        * platform/mediastream/mac/WindowDisplayCaptureSourceMac.h: Added.
        * platform/mediastream/mac/WindowDisplayCaptureSourceMac.mm: Added.
        (WebCore::anyOfCGWindow):
        (WebCore::windowDescription):
        (WebCore::WindowDisplayCaptureSourceMac::create):
        (WebCore::WindowDisplayCaptureSourceMac::WindowDisplayCaptureSourceMac):
        (WebCore::WindowDisplayCaptureSourceMac::windowImage):
        (WebCore::WindowDisplayCaptureSourceMac::generateFrame):
        (WebCore::WindowDisplayCaptureSourceMac::pixelBufferFromCGImage):
        (WebCore::WindowDisplayCaptureSourceMac::windowCaptureDeviceWithPersistentID):
        (WebCore::WindowDisplayCaptureSourceMac::windowCaptureDevices):

2018-09-25  Justin Fan  <justin_fan@apple.com>

        WebGL 2 Conformance: primitive restart and draw_primitive_restart WebGL2 sample
        https://bugs.webkit.org/show_bug.cgi?id=189625
        <rdar://problem/42882620>

        Reviewed by Dean Jackson.

        Ref test: webgl/webgl2-primitive-restart.html.

        Implement support for Primitive Restart Fixed Index as expected 
        by the WebGL 2 specifications. 

        * html/canvas/WebGL2RenderingContext.cpp:
        (WebCore::WebGL2RenderingContext::validateIndexArrayConservative):
        * html/canvas/WebGLRenderingContextBase.h:
        (WebCore::WebGLRenderingContextBase::getLastIndex): Template that must be defined in header.
        * html/canvas/WebGLRenderingContextBase.cpp:
        (WebCore::WebGLRenderingContextBase::validateIndexArrayPrecise):
        * platform/graphics/GraphicsContext3D.h:
        * platform/graphics/cocoa/GraphicsContext3DCocoa.mm:
        (WebCore::GraphicsContext3D::GraphicsContext3D):
        * platform/graphics/opengl/GraphicsContext3DOpenGLCommon.cpp:
        (WebCore::GraphicsContext3D::primitiveRestartIndex):

2018-09-25  John Wilander  <wilander@apple.com>

        Change from HAVE(CFNETWORK_STORAGE_PARTITIONING) to ENABLE(RESOURCE_LOAD_STATISTICS)
        https://bugs.webkit.org/show_bug.cgi?id=189959
        <rdar://problem/44767642>

        Reviewed by Chris Dumez.

        No new tests because of no code change.

        We no longer make use of CFNetwork's cookie partitioning so we should
        change the compile-time flag to something that makes sense. This should
        also make it easier/cleaner for other ports.

        * dom/Document.cpp:
        (WebCore::Document::hasStorageAccess):
        (WebCore::Document::requestStorageAccess):
        (WebCore::Document::setHasRequestedPageSpecificStorageAccessWithUserInteraction):
        * dom/Document.h:
        * loader/EmptyFrameLoaderClient.h:
        * loader/FrameLoaderClient.h:
        * loader/ResourceLoadObserver.cpp:
        (WebCore::ResourceLoadObserver::logUserInteractionWithReducedTimeResolution):
        (WebCore::ResourceLoadObserver::logWindowCreation):
        * loader/ResourceLoadObserver.h:
        * page/DOMWindow.cpp:
        (WebCore::DOMWindow::createWindow):
        * platform/network/NetworkStorageSession.h:
        * platform/network/cf/NetworkStorageSessionCFNet.cpp:
        * platform/network/cocoa/NetworkStorageSessionCocoa.mm:
        (WebCore::cookiesForURL):
        (WebCore::NetworkStorageSession::setCookiesFromDOM const):

2018-09-25  Jiewen Tan  <jiewen_tan@apple.com>

        [WebAuthN] Make AuthenticatorManager
        https://bugs.webkit.org/show_bug.cgi?id=189279
        <rdar://problem/44116792>