Timeline
Jun 25, 2016:
- 9:34 PM Changeset in webkit [202470] by
-
- 5 edits4 adds in trunk
The active state of elements can break when focus changes
https://bugs.webkit.org/show_bug.cgi?id=159112
Patch by Benjamin Poulain <bpoulain@apple.com> on 2016-06-25
Reviewed by Antti Koivisto.
Source/WebCore:
The pseudo class :active was behaving weirdly when used
with label elements with an associated form element.
The form element would get the :active state on the first click
then no longer get the state until the focus changes.
What was happenning is setFocusedElement() was clearing active
for some unknown reason. When you really do that on an active element,
you end up in an inconsistent state where no invalidation works.
The two tests illustrates 2 ways this breaks.
The test "pseudo-active-on-labeled-element-not-canceled-by-focus" clicks
several time on a lable element. The first time, the input element gets
the focus. The second time, it already has the focus, setFocusedElement()
clears :active before finding the focusable element and end up clearing
the active state on a target in the active chain.
The test "pseudo-active-with-programmatic-focus.html" shows how to invalidate
arbitrary elements using JavaScript. This can cause severely broken active
chains where invalidation never cleans some ancestors.
Tests: fast/css/pseudo-active-on-labeled-element-not-canceled-by-focus.html
fast/css/pseudo-active-with-programmatic-focus.html
- dom/Document.cpp:
(WebCore::Document::setFocusedElement): Deleted.
- page/EventHandler.cpp:
(WebCore::EventHandler::handleMouseDoubleClickEvent):
This is WebKit1 specific. The double click event was dispatching
the mouseUp and Click with after doing an Active hit test.
This causes us to have :active state in and after mouseUp in WebKit1.
LayoutTests:
- fast/css/pseudo-active-on-labeled-element-not-canceled-by-focus-expected.txt: Added.
- fast/css/pseudo-active-on-labeled-element-not-canceled-by-focus.html: Added.
- fast/css/pseudo-active-with-programmatic-focus-expected.txt: Added.
- fast/css/pseudo-active-with-programmatic-focus.html: Added.
- 12:07 PM Changeset in webkit [202469] by
-
- 4 edits in trunk/Tools
REGRESSION: Weak symbol warning when linking TestWebKitAPI
https://bugs.webkit.org/show_bug.cgi?id=159104
Reviewed by Dan Bernstein.
This uses a better approach to turn on -fvisility=hidden. All of the other projects set
this flag in xcconfig, using "GCC_SYMBOLS_PRIVATE_EXTERN = YES;".
- TestWebKitAPI/Configurations/TestWTFLibrary.xcconfig: Use GCC_SYMBOLS_PRIVATE_EXTERN here.
- TestWebKitAPI/Configurations/TestWebKitAPILibrary.xcconfig: Use GCC_SYMBOLS_PRIVATE_EXTERN here.
- TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj: Revert r202461.
- 2:00 AM Changeset in webkit [202468] by
-
- 2 edits in trunk/Tools
[Win] The test fast/events/wheel-event-destroys-frame.html is timing out.
https://bugs.webkit.org/show_bug.cgi?id=159086
Reviewed by Alex Christensen.
Implement required wheel event function in event sender.
- DumpRenderTree/win/EventSender.cpp:
(mouseScrollByWithWheelAndMomentumPhasesCallback):
- 12:07 AM MathML/Early_2016_Refactoring edited by
- (diff)
Jun 24, 2016:
- 7:33 PM Changeset in webkit [202467] by
-
- 3 edits13 adds in trunk/Websites/webkit.org
More examples for the wide-gamut demo.
- blog-files/color-gamut/Iceland-P3.jpg: Added.
- blog-files/color-gamut/Iceland-oog.jpg: Added.
- blog-files/color-gamut/Iceland-sRGB.jpg: Added.
- blog-files/color-gamut/Italy-P3.jpg: Added.
- blog-files/color-gamut/Italy-oog.jpg: Added.
- blog-files/color-gamut/Italy-sRGB.jpg: Added.
- blog-files/color-gamut/Sunset-AdobeRGB.jpg: Added.
- blog-files/color-gamut/Sunset-P3.jpg: Added.
- blog-files/color-gamut/Sunset-oog.jpg: Added.
- blog-files/color-gamut/Sunset-sRGB.jpg: Added.
- blog-files/color-gamut/Webkit-logo-sRGB.png:
- blog-files/color-gamut/YellowFlower-P3.jpg: Added.
- blog-files/color-gamut/YellowFlower-oog.jpg: Added.
- blog-files/color-gamut/YellowFlower-sRGB.jpg: Added.
- blog-files/color-gamut/index.html:
- 5:46 PM Changeset in webkit [202466] by
-
- 4 edits in trunk
Consider exposing or hiding knowledge of a redirect from clients of WebCoreNSURLSession
https://bugs.webkit.org/show_bug.cgi?id=156722
<rdar://problem/25780035>
Reviewed by Alex Christensen.
Source/WebCore:
Fixes tests: http/tests/security/contentSecurityPolicy/audio-redirect-allowed2.html
http/tests/security/contentSecurityPolicy/video-redirect-allowed2.html
When receieving a NSURLResponse containing a redirected URL, AVFoundadtion will use the
URL in the response for subsequent requests. This violates the HTTP specification if the
redirect was temporary, and it also breaks two CSP tests by bypassing the redirect step
for subsequent requests.
Work around this behavior in AVFoundation by recreating the NSURLResponse with the original
request URL in the case of a temporary redirect.
- platform/network/cocoa/WebCoreNSURLSession.mm:
(-[WebCoreNSURLSessionDataTask resource:receivedResponse:]):
(-[WebCoreNSURLSessionDataTask resource:receivedRedirect:request:]):
LayoutTests:
Un-skip http/tests/security/contentSecurityPolicy/audio-redirect-allowed2.html &
http/tests/security/contentSecurityPolicy/video-redirect-allowed2.html.
- platform/mac/TestExpectations:
- 5:40 PM Changeset in webkit [202465] by
-
- 1 edit in trunk/Websites/webkit.org/ChangeLog
Change something in the webkit.org directory in the hope
it will kick off a website update.
- 5:36 PM Changeset in webkit [202464] by
-
- 2 edits in trunk/Source/WebKit2
REGRESSION(r201171): CRASH at WebKit::WebInspectorProxy::open() + 31 when running inspector layout tests
https://bugs.webkit.org/show_bug.cgi?id=159070
<rdar://problem/26768628>
Reviewed by Joseph Pecoraro.
We have been seeing a few crashes underneath WebInspectorProxy::bringToFront() on the bots.
Previously, this code didn't use m_inspectorPage so there was nothing to null-dereference.
However, it doesn't make sense that we would hit the null dereference here:
- The only caller of bringToFront() on the WebProcess side is InspectorController::show(). It only tries to bring to front if there is already a local frontend connection, which shouldn't be the case if m_inspectorPage is null.
- It's guarded by m_underTest, which should have been set to true in createInspectorPage().
These clues lead me to believe that we may be improperly tearing down the inspector between tests.
For example, it seems possible that a local frontend connection is not being torn down, so
InspectorController never asks to create a inspector page when the next test calls showWebInspector.
Since this crash is not easy to reproduce, we don't have much to go on. For now, this patch
adds an early return in the case where m_inspectorPage is null when calling open().
- UIProcess/WebInspectorProxy.cpp:
(WebKit::WebInspectorProxy::open):
- 5:03 PM Changeset in webkit [202463] by
-
- 4 edits in trunk/Source/JavaScriptCore
B3 should die sooner if a Value has the wrong number of children
https://bugs.webkit.org/show_bug.cgi?id=159108
Reviewed by Mark Lam.
I've been looking at a bug (rdar://problem/26500743) that's about a Vector OOB crash in
ReduceStrength::rangeFor(). The only Vector accesses are to Value::m_children, and all of
the accesses in rangeFor() are for child(0) or child(1) of binary arithmetic opcodes.
Clearly those should never go out-of-bounds.
Maybe we have horrible memory corruption. Or maybe some path creates a Value with the
wrong number of children, and that path is not tested by any of our tests. This patch adds
release assertions that will catch the latter.
I've tested this a lot. It's not a regression on our benchmarks.
- b3/B3Opcode.h:
- b3/B3Value.cpp:
(JSC::B3::Value::dumpMeta):
(JSC::B3::Value::typeFor):
(JSC::B3::Value::badOpcode):
(JSC::B3::Value::checkOpcode): Deleted.
- b3/B3Value.h:
- 4:54 PM Changeset in webkit [202462] by
-
- 2 edits in trunk/Source/WebCore
MSE gets confused by in-band text tracks
https://bugs.webkit.org/show_bug.cgi?id=159107
<rdar://problem/26871330>
Reviewed by Eric Carlson.
We can't currently handle text track samples in SourceBufferPrivateAVFObjC,
so don't pass them up to SourceBuffer.
- platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:
(WebCore::SourceBufferPrivateAVFObjC::didParseStreamDataAsAsset):
(WebCore::SourceBufferPrivateAVFObjC::processCodedFrame):
- 4:49 PM Changeset in webkit [202461] by
-
- 2 edits in trunk/Tools
REGRESSION: Weak symbol warning when linking TestWebKitAPI
https://bugs.webkit.org/show_bug.cgi?id=159104
Reviewed by Mark Lam.
The problem is that the tests were not compiled with -fvisibility=hidden like the rest of
WK.
- TestWebKitAPI/TestWebKitAPI.xcodeproj/project.pbxproj:
- 4:34 PM Changeset in webkit [202460] by
-
- 6 edits5 adds in trunk
[JSC] Error prototypes are called on remote scripts.
https://bugs.webkit.org/show_bug.cgi?id=52192
Reviewed by Keith Miller.
Source/JavaScriptCore:
Added a sanitizedToString() to the Error instance object so that it can be used
to get an error string without invoking getters and proxies.
- runtime/ErrorInstance.cpp:
(JSC::ErrorInstance::finishCreation):
(JSC::ErrorInstance::sanitizedToString):
- runtime/ErrorInstance.h:
(JSC::ErrorInstance::createStructure):
(JSC::ErrorInstance::runtimeTypeForCause):
(JSC::ErrorInstance::clearRuntimeTypeForCause):
Source/WebCore:
Test: http/tests/security/regress-52192.html
Parsing errors are reported to the main script's window.onerror function. AFAIK,
both Chrome and Firefox have the error reporting mechanism use an internal
sanitized version of Error.prototype.toString() that will not invoke any getters
or proxies instead.
This patch fixes this issue by matching Chrome and Firefox's behavior.
Note: we did not choose to make error objects and prototypes read-only because
that was observed to have broken the web.
See https://bugs.chromium.org/p/chromium/issues/detail?id=69187#c73
Credit for reporting this issue goes to Daniel Divricean (http://divricean.ro).
- bindings/js/JSDOMBinding.cpp:
(WebCore::reportException):
- ForwardingHeaders/runtime/ErrorInstance.h: Added.
LayoutTests:
The added test will test the following combinations of factors:
- Explicitly throwing an error of each of the types of JS errors i.e. Error, EvalError, RangeError, ReferenceError, SyntaxError, TypeError, URIError.
- Validating that the error received in window.onerror does not leak any info.
- Validating that the 'name' and 'toString' getters set on the error prototype object does not get invoked for creating the error message to be passed to window.error.
- http/tests/security/regress-52192-expected.txt: Added.
- http/tests/security/regress-52192.html: Added.
- http/tests/security/resources/regress-52192-syntax-error.js: Added.
- http/tests/security/resources/regress-52192-throw-error.js: Added.
(catch):
- 4:01 PM Changeset in webkit [202459] by
-
- 2 edits in trunk/Source/WebCore
Media elements should not lose playback controls when muted by a user gesture
https://bugs.webkit.org/show_bug.cgi?id=159078
<rdar://problem/26925904>
Reviewed by Beth Dakin.
Rearrange canControlControlsManager() so that the muted check only occurs if
a user gesture is required.
- html/MediaElementSession.cpp:
(WebCore::MediaElementSession::canControlControlsManager):
- 3:52 PM Changeset in webkit [202458] by
-
- 3 edits in trunk/Source/JavaScriptCore
Unreviewed, rolling out r202443.
https://bugs.webkit.org/show_bug.cgi?id=159105
Introduced memory corruption crashes (Requested by ap on
#webkit).
Reverted changeset:
"Web Inspector: CRASH in backend at
Inspector::HeapFrontendDispatcher::garbageCollected + 552 when
closing frontend/inspected page"
https://bugs.webkit.org/show_bug.cgi?id=159075
http://trac.webkit.org/changeset/202443
- 3:44 PM Changeset in webkit [202457] by
-
- 1 edit6 adds in trunk/Websites/webkit.org/blog-files/color-gamut
More examples for a blog post on color.
- blog-files/color-gamut/Shoes-oog.jpg: Added.
- blog-files/color-gamut/Shoes-sRGB.jpg: Added.
- blog-files/color-gamut/Webkit-logo-P3.png: Added.
- blog-files/color-gamut/Webkit-logo-oog.png: Added.
- blog-files/color-gamut/Webkit-logo-sRGB.png: Added.
- blog-files/color-gamut/index.html:
- blog-files/color-gamut/shoes-AdobeRGB.jpg: Added.
- 3:39 PM Changeset in webkit [202456] by
-
- 2 edits in trunk/LayoutTests
Update to test gardening in r202300
Unreviewed.
- platform/mac/TestExpectations:
- 3:25 PM Changeset in webkit [202455] by
-
- 2 edits in trunk/Source/WebKit2
Vimeo.com videos do not get playback controls
https://bugs.webkit.org/show_bug.cgi?id=159082
<rdar://problem/26586630>
Reviewed by Eric Carlson.
When the HasAudioOrVideo media state changes, let the video controls recalculate
by calling videoControlsManagerDidChange().
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::isPlayingMediaDidChange):
- 3:25 PM Changeset in webkit [202454] by
-
- 7 edits in trunk/Source
Include enclosingListType in EditorState
https://bugs.webkit.org/show_bug.cgi?id=159102
-and corresponding-
rdar://problem/26932490
Reviewed by Enrica Casucci.
Source/WebCore:
Make HTMLOListElement.h and HTMLUListElement.h Private instead of Project.
- WebCore.xcodeproj/project.pbxproj:
Export enclosingList(Node*)
- editing/htmlediting.h:
Source/WebKit2:
- Shared/EditorState.cpp:
(WebKit::EditorState::PostLayoutData::encode):
(WebKit::EditorState::PostLayoutData::decode):
- Shared/EditorState.h:
- WebProcess/WebPage/WebPage.cpp:
(WebKit::WebPage::editorState):
- 3:00 PM Changeset in webkit [202453] by
-
- 5 edits in branches/safari-602.1.38-branch/Source
Versioning.
- 2:59 PM Changeset in webkit [202452] by
-
- 1 copy in tags/Safari-602.1.38.1
New tag.
- 2:58 PM Changeset in webkit [202451] by
-
- 2 edits in trunk/Source/WebCore
Another Windows build fix.
- platform/network/BlobRegistry.h:
- 2:54 PM Changeset in webkit [202450] by
-
- 5 edits in branches/safari-602.1.38-branch/Source
Versioning.
- 2:54 PM Changeset in webkit [202449] by
-
- 2 edits in trunk/Websites/webkit.org
Rewrite the WebKit Nightly build download URLs to use the correct secure URL.
https://bugs.webkit.org/show_bug.cgi?id=159100.
Reviewed by Joseph Pecoraro.
- wp-content/plugins/sync-nightly-builds.php:
- 2:33 PM Changeset in webkit [202448] by
-
- 2 edits in trunk/Source/WebCore
Yeat another Windows build fix.
- dom/ActiveDOMCallbackMicrotask.h:
- 2:09 PM Changeset in webkit [202447] by
-
- 6 edits1 delete in trunk/LayoutTests
Removed tests of STYLE_SCOPED which was removed in r156683.
https://bugs.webkit.org/show_bug.cgi?id=159098
Reviewed by Ryosuke Niwa.
- fast/css/style-scoped/basic-attribute-expected.txt: Removed.
- fast/css/style-scoped/basic-attribute.html: Removed.
- fast/css/style-scoped/registering-expected.txt: Removed.
- fast/css/style-scoped/registering-shadowroot-expected.txt: Removed.
- fast/css/style-scoped/registering-shadowroot.html: Removed.
- fast/css/style-scoped/registering.html: Removed.
- fast/css/style-scoped/style-scoped-apply-author-styles-expected.txt: Removed.
- fast/css/style-scoped/style-scoped-apply-author-styles.html: Removed.
- fast/css/style-scoped/style-scoped-attach-expected.txt: Removed.
- fast/css/style-scoped/style-scoped-attach.html: Removed.
- fast/css/style-scoped/style-scoped-basic-expected.txt: Removed.
- fast/css/style-scoped/style-scoped-basic.html: Removed.
- fast/css/style-scoped/style-scoped-change-scoped-in-shadow-expected.txt: Removed.
- fast/css/style-scoped/style-scoped-change-scoped-in-shadow.html: Removed.
- fast/css/style-scoped/style-scoped-detach-expected.txt: Removed.
- fast/css/style-scoped/style-scoped-detach.html: Removed.
- fast/css/style-scoped/style-scoped-in-shadow-expected.txt: Removed.
- fast/css/style-scoped/style-scoped-in-shadow.html: Removed.
- fast/css/style-scoped/style-scoped-nested-expected.txt: Removed.
- fast/css/style-scoped/style-scoped-nested.html: Removed.
- fast/css/style-scoped/style-scoped-remove-scoped-expected.txt: Removed.
- fast/css/style-scoped/style-scoped-remove-scoped.html: Removed.
- fast/css/style-scoped/style-scoped-set-scoped-expected.txt: Removed.
- fast/css/style-scoped/style-scoped-set-scoped.html: Removed.
- fast/css/style-scoped/style-scoped-with-dom-operation-expected.txt: Removed.
- fast/css/style-scoped/style-scoped-with-dom-operation.html: Removed.
- fast/css/style-scoped/style-scoped-with-important-rule-expected.txt: Removed.
- fast/css/style-scoped/style-scoped-with-important-rule.html: Removed.
- platform/efl/TestExpectations: Don't skip fast/css/style-scoped.
- platform/gtk/TestExpectations: Ditto.
- platform/ios-simulator/TestExpectations: Ditto.
- platform/mac/TestExpectations: Ditto.
- platform/win/TestExpectations: Ditto.
- 1:59 PM Changeset in webkit [202446] by
-
- 2 edits13 adds in trunk/PerformanceTests
Add a ES6 generator benchmark
https://bugs.webkit.org/show_bug.cgi?id=159101
Rubber stamped by Keith Miller.
This adds a Basic interpreter loosely based on ECMA-55:
http://www.ecma-international.org/publications/files/ECMA-ST-WITHDRAWN/ECMA-55,%201st%20Edition,%20January%201978.pdf
It includes a lexer that is a generator, a parser that uses regular expressions, and an AST
walk interpreter where the walking functions for statements are generators that call each
other with yield*. This enables the interpreter to look like an AST walk even though it can
yield at INPUT and PRINT statements.
This also uses for-of, classes, Map, and WeakMap. It also uses deprecated-but-awesome RegExp
features like RegExp.lastMatch and RegExp.rightContext. I did it that way because this is
how I've always written lexers in dynamic languages; see offlineasm's lex() method in
parser.rb for example.
The benchmark runs a handful of simple Basic programs. The longest-running one computes
prime numbers.
Includes a command-line and web harness. On my machine it runs in 2-3 seconds.
- Basic: Added.
- Basic/ast.js: Added.
(Basic.NumberApply):
(Basic.Variable):
(Basic.Const):
(Basic.NumberPow):
(Basic.NumberMul):
(Basic.NumberDiv):
(Basic.NumberNeg):
(Basic.NumberAdd):
(Basic.NumberSub):
(Basic.StringVar):
(Basic.Equals):
(Basic.NotEquals):
(Basic.LessThan):
(Basic.GreaterThan):
(Basic.LessEqual):
(Basic.GreaterEqual):
(Basic.GoTo):
(Basic.GoSub):
(Basic.Def):
(Basic.Let):
(Basic.If):
(Basic.Return):
(Basic.Stop):
(Basic.On):
(sideState.shouldStop):
(Basic.For):
(Basic.Next):
(Basic.Print):
(Basic.Input):
(Basic.Read):
(Basic.Restore):
(Basic.Dim):
(Basic.Randomize):
(Basic.End):
(Basic.Program):
- Basic/basic.js: Added.
(prepare):
(simulate):
- Basic/benchmark.js: Added.
(Benchmark):
(Benchmark.prototype.runIteration.expect):
(Benchmark.prototype.runIteration):
(runBenchmark):
- Basic/caseless_map.js: Added.
(CaselessMap):
- Basic/lexer.js: Added.
(lex.consumeWhitespace):
(lex.consume):
(lex):
- Basic/lexer_test.js: Added.
- Basic/number.js: Added.
(NumberValue):
(NumberValue.prototype.get value):
(NumberValue.prototype.apply):
(NumberValue.prototype.leftApply):
(NumberValue.prototype.assign):
(NumberArray.):
(NumberArray):
(NumberArray.prototype.apply):
(NumberArray.prototype.leftApply):
(NumberFunction):
(NumberFunction.prototype.apply):
(NumberFunction.prototype.leftApply):
(NativeFunction):
(NativeFunction.prototype.apply):
(NativeFunction.prototype.leftApply):
- Basic/parser.js: Added.
(parse):
(parse.pushToken):
(parse.peekToken):
(parse.consumeKind):
(parse.consumeToken):
(parse.parseVariable):
(parse.parseNumericExpression.parsePrimary):
(parse.parseNumericExpression.parseFactor):
(parse.parseNumericExpression.parseTerm):
(parse.parseNumericExpression):
(parse.parseConstant):
(parse.parseStringExpression):
(parse.isStringExpression):
(parse.parseRelationalExpression):
(parse.parseNonNegativeInteger):
(parse.parseGoToStatement):
(parse.parseGoSubStatement):
(parse.parseStatement):
(parse.parseStatements):
- Basic/random.js: Added.
(createRNG):
(createRNGWithFixedSeed):
(createRNGWithRandomSeed):
- Basic/state.js: Added.
(State):
(State.prototype.getValue):
(State.prototype.getSideState):
(State.prototype.abort):
(State.prototype.validate):
- Basic/test.html: Added.
- Basic/test.js: Added.
- Basic/util.js: Added.
(this.performance.performance.now.currentTime):
(else.this.preciseTime.currentTime):
(else.currentTime):
- Skipped: Make sure that we don't run Basic yet.
- 1:53 PM Changeset in webkit [202445] by
-
- 2 edits in trunk/Source/WebCore
Another Windows build fix.
- page/FrameView.h:
- 1:49 PM Changeset in webkit [202444] by
-
- 13 edits1 delete in trunk/Source
Inline more of the Apple Pay source code
https://bugs.webkit.org/show_bug.cgi?id=159099
Reviewed by Andreas Kling.
Source/WebCore:
- page/Settings.h:
(WebCore::Settings::applePayEnabled):
(WebCore::Settings::setApplePayEnabled):
(WebCore::Settings::applePayCapabilityDisclosureAllowed):
(WebCore::Settings::setApplePayCapabilityDisclosureAllowed):
Source/WebKit2:
- Shared/Cocoa/WebCoreArgumentCodersCocoa.h: Removed.
- Shared/WebCoreArgumentCoders.h:
- Shared/WebPreferencesDefinitions.h:
- UIProcess/API/Cocoa/WKPreferences.mm:
(-[WKPreferences _applePayCapabilityDisclosureAllowed]):
(-[WKPreferences _setApplePayCapabilityDisclosureAllowed:]):
- UIProcess/API/Cocoa/WKWebView.mm:
(-[WKWebView _initializeWithConfiguration:]):
- UIProcess/API/Cocoa/WKWebViewConfiguration.mm:
(-[WKWebViewConfiguration copyWithZone:]):
(-[WKWebViewConfiguration _applePayEnabled]):
(-[WKWebViewConfiguration _setApplePayEnabled:]):
(-[WKWebViewConfiguration _paymentsEnabled]):
(-[WKWebViewConfiguration _setPaymentsEnabled:]):
- UIProcess/API/Cocoa/WKWebViewConfigurationPrivate.h:
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::WebPageProxy):
(WebKit::WebPageProxy::reattachToWebProcess):
(WebKit::WebPageProxy::resetState):
- UIProcess/WebPageProxy.h:
- WebKit2.xcodeproj/project.pbxproj:
- WebProcess/WebPage/WebPage.cpp:
(WebKit::m_userInterfaceLayoutDirection):
(WebKit::WebPage::updatePreferences):
- 1:30 PM Changeset in webkit [202443] by
-
- 3 edits in trunk/Source/JavaScriptCore
Web Inspector: CRASH in backend at Inspector::HeapFrontendDispatcher::garbageCollected + 552 when closing frontend/inspected page
https://bugs.webkit.org/show_bug.cgi?id=159075
<rdar://problem/26094341>
Reviewed by Joseph Pecoraro.
Move the asynchronous work to a task class that can be cancelled when the
heap agent is reset, disabled or destroyed.
- inspector/agents/InspectorHeapAgent.cpp:
(Inspector::SendGarbageCollectionEventsTask::SendGarbageCollectionEventsTask):
(Inspector::SendGarbageCollectionEventsTask::addGarbageCollection):
(Inspector::SendGarbageCollectionEventsTask::reset):
(Inspector::SendGarbageCollectionEventsTask::timerFired):
Added. This holds onto GarbageCollection objects that need to be sent asynchronously.
It uses the RunLoop variant of Timer and can queue multiple pending objects to be sent.
(Inspector::InspectorHeapAgent::InspectorHeapAgent):
(Inspector::InspectorHeapAgent::~InspectorHeapAgent):
(Inspector::InspectorHeapAgent::disable):
Reset the task when disabling or tearing down the agent so the timer doesn't fire after destruction.
(Inspector::InspectorHeapAgent::didGarbageCollect):
Send the object to the task to be dispatched asynchronously.
- inspector/agents/InspectorHeapAgent.h:
- 1:18 PM Changeset in webkit [202442] by
-
- 2 edits in trunk/Source/WebCore
Windows build fix.
- platform/GenericTaskQueue.h:
(WebCore::TaskDispatcher::postTask):
- 1:17 PM Changeset in webkit [202441] by
-
- 9 edits in trunk/Source/WebCore
Use auto* for MathML elements and renderers when possible
https://bugs.webkit.org/show_bug.cgi?id=159090
Patch by Frederic Wang <fwang@igalia.com> on 2016-06-24
Reviewed by Alex Christensen.
No new tests, behavior is unchanged.
- mathml/MathMLElement.cpp:
(WebCore::MathMLElement::attributeChanged):
- mathml/MathMLSelectElement.cpp:
(WebCore::MathMLSelectElement::getSelectedActionChildAndIndex):
(WebCore::MathMLSelectElement::getSelectedActionChild):
(WebCore::MathMLSelectElement::getSelectedSemanticsChild):
(WebCore::MathMLSelectElement::updateSelectedChild):
- rendering/mathml/RenderMathMLFraction.cpp:
(WebCore::RenderMathMLFraction::isValid):
- rendering/mathml/RenderMathMLMenclose.cpp:
(WebCore::RenderMathMLMenclose::layoutBlock):
- rendering/mathml/RenderMathMLRoot.cpp:
(WebCore::RenderMathMLRoot::isValid):
- rendering/mathml/RenderMathMLRow.cpp:
(WebCore::RenderMathMLRow::firstLineBaseline):
(WebCore::RenderMathMLRow::computeLineVerticalStretch):
(WebCore::RenderMathMLRow::computePreferredLogicalWidths):
(WebCore::RenderMathMLRow::layoutRowItems):
- rendering/mathml/RenderMathMLScripts.cpp:
(WebCore::RenderMathMLScripts::unembellishedOperator):
(WebCore::RenderMathMLScripts::getBaseAndScripts):
(WebCore::RenderMathMLScripts::computePreferredLogicalWidths):
(WebCore::RenderMathMLScripts::getScriptMetricsAndLayoutIfNeeded):
(WebCore::RenderMathMLScripts::layoutBlock):
(WebCore::RenderMathMLScripts::firstLineBaseline):
- rendering/mathml/RenderMathMLUnderOver.cpp:
(WebCore::RenderMathMLUnderOver::firstLineBaseline):
(WebCore::RenderMathMLUnderOver::isValid):
(WebCore::RenderMathMLUnderOver::over):
- 1:14 PM Changeset in webkit [202440] by
-
- 3 edits in trunk/Source/WebCore
Remove unused and static return value from InspectorStyle::populateAllProperties
https://bugs.webkit.org/show_bug.cgi?id=159069
Patch by Joseph Pecoraro <Joseph Pecoraro> on 2016-06-24
Reviewed by Andreas Kling.
- inspector/InspectorStyleSheet.cpp:
(WebCore::InspectorStyle::populateAllProperties):
- inspector/InspectorStyleSheet.h:
- 12:40 PM Changeset in webkit [202439] by
-
- 65 edits1 move in trunk/Source
Rename NoncopyableFunction to Function
https://bugs.webkit.org/show_bug.cgi?id=158354
Reviewed by Chris Dumez.
Source/WebCore:
- Modules/mediastream/MediaEndpointPeerConnection.cpp:
(WebCore::MediaEndpointPeerConnection::runTask):
- Modules/mediastream/MediaEndpointPeerConnection.h:
- Modules/webaudio/AudioDestinationNode.h:
(WebCore::AudioDestinationNode::resume):
(WebCore::AudioDestinationNode::suspend):
(WebCore::AudioDestinationNode::close):
- Modules/webaudio/DefaultAudioDestinationNode.cpp:
(WebCore::DefaultAudioDestinationNode::resume):
(WebCore::DefaultAudioDestinationNode::suspend):
(WebCore::DefaultAudioDestinationNode::close):
- Modules/webaudio/DefaultAudioDestinationNode.h:
- dom/ActiveDOMCallbackMicrotask.cpp:
(WebCore::ActiveDOMCallbackMicrotask::ActiveDOMCallbackMicrotask):
- dom/ActiveDOMCallbackMicrotask.h:
- dom/ScriptExecutionContext.h:
(WebCore::ScriptExecutionContext::Task::Task):
- fileapi/AsyncFileStream.cpp:
(WebCore::callOnFileThread):
(WebCore::AsyncFileStream::perform):
- fileapi/AsyncFileStream.h:
- page/FrameView.cpp:
(WebCore::FrameView::queuePostLayoutCallback):
(WebCore::FrameView::flushPostLayoutTasksQueue):
- page/FrameView.h:
- page/scrolling/ScrollingThread.cpp:
(WebCore::ScrollingThread::dispatch):
(WebCore::ScrollingThread::dispatchBarrier):
(WebCore::ScrollingThread::dispatchFunctionsFromScrollingThread):
- page/scrolling/ScrollingThread.h:
- platform/GenericTaskQueue.cpp:
(WebCore::TaskDispatcher<Timer>::postTask):
- platform/GenericTaskQueue.h:
(WebCore::TaskDispatcher::postTask):
- platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaStreamAVFObjC.h:
- platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaStreamAVFObjC.mm:
(WebCore::MediaPlayerPrivateMediaStreamAVFObjC::scheduleDeferredTask):
- platform/mediastream/MediaStreamPrivate.cpp:
(WebCore::MediaStreamPrivate::scheduleDeferredTask):
- platform/mediastream/MediaStreamPrivate.h:
- platform/mediastream/mac/AVMediaCaptureSource.h:
- platform/mediastream/mac/AVMediaCaptureSource.mm:
(WebCore::AVMediaCaptureSource::scheduleDeferredTask):
- style/StyleTreeResolver.cpp:
(WebCore::Style::postResolutionCallbackQueue):
(WebCore::Style::queuePostResolutionCallback):
- style/StyleTreeResolver.h:
Source/WebKit:
- Storage/StorageSyncManager.cpp:
(WebCore::StorageSyncManager::dispatch):
- Storage/StorageSyncManager.h:
- Storage/StorageThread.cpp:
(WebCore::StorageThread::dispatch):
(WebCore::StorageThread::terminate):
- Storage/StorageThread.h:
Source/WebKit2:
- NetworkProcess/NetworkProcess.cpp:
(WebKit::fetchDiskCacheEntries):
(WebKit::NetworkProcess::fetchWebsiteData):
(WebKit::clearDiskCacheEntries):
- NetworkProcess/NetworkProcess.h:
- NetworkProcess/cache/NetworkCache.cpp:
(WebKit::NetworkCache::Cache::traverse):
- NetworkProcess/cache/NetworkCache.h:
- NetworkProcess/cache/NetworkCacheIOChannelSoup.cpp:
(WebKit::NetworkCache::runTaskInQueue):
- NetworkProcess/cache/NetworkCacheStorage.h:
Source/WTF:
- WTF.xcodeproj/project.pbxproj:
- wtf/CrossThreadTask.h:
(WTF::CrossThreadTask::CrossThreadTask):
- wtf/Function.h: Renamed from Source/WTF/wtf/NoncopyableFunction.h.
- wtf/FunctionDispatcher.h:
- wtf/MainThread.cpp:
(WTF::functionQueue):
(WTF::dispatchFunctionsFromMainThread):
(WTF::callOnMainThread):
- wtf/MainThread.h:
- wtf/RunLoop.cpp:
(WTF::RunLoop::performWork):
(WTF::RunLoop::dispatch):
- wtf/RunLoop.h:
- wtf/WorkQueue.h:
- wtf/cocoa/WorkQueueCocoa.cpp:
(WTF::WorkQueue::dispatch):
(WTF::WorkQueue::dispatchAfter):
- wtf/efl/DispatchQueueWorkItemEfl.h:
(WorkItem::WorkItem):
(TimerWorkItem::create):
(TimerWorkItem::TimerWorkItem):
- wtf/efl/WorkQueueEfl.cpp:
(WTF::WorkQueue::dispatch):
(WTF::WorkQueue::dispatchAfter):
- wtf/generic/RunLoopGeneric.cpp:
(WTF::RunLoop::TimerBase::ScheduledTask::create):
(WTF::RunLoop::TimerBase::ScheduledTask::ScheduledTask):
(WTF::RunLoop::dispatchAfter):
- wtf/generic/WorkQueueGeneric.cpp:
(WorkQueue::dispatch):
(WorkQueue::dispatchAfter):
- wtf/glib/RunLoopGLib.cpp:
(WTF::DispatchAfterContext::DispatchAfterContext):
(WTF::RunLoop::dispatchAfter):
- wtf/win/WorkItemWin.cpp:
(WTF::WorkItemWin::WorkItemWin):
(WTF::WorkItemWin::create):
(WTF::HandleWorkItem::HandleWorkItem):
(WTF::HandleWorkItem::createByAdoptingHandle):
- wtf/win/WorkItemWin.h:
(WTF::WorkItemWin::function):
- wtf/win/WorkQueueWin.cpp:
(WTF::WorkQueue::dispatch):
(WTF::WorkQueue::dispatchAfter):
- 12:26 PM Changeset in webkit [202438] by
-
- 2 edits in trunk/Source/WebCore
Use _CFHTTPCookieStorageGetDefault directly instead of NSHTTPCookieStorage to get default cookie storage
https://bugs.webkit.org/show_bug.cgi?id=159095
rdar://problem/26630073
Patch by Amir Alavi <aalavi@apple.com> on 2016-06-24
Reviewed by Brent Fulgham.
No new tests, it isn't possible to test this in a LayoutTest.
- platform/network/mac/CookieJarMac.mm:
(WebCore::httpCookiesForURL): Get a CFHTTPCookieStorageRef when no cookie storage is provided to match the case when cookie storage is provided.
- 12:15 PM Changeset in webkit [202437] by
-
- 2 edits in trunk/LayoutTests
Update test expectations for rdar://problem/27000853
- platform/mac/TestExpectations:
- 11:52 AM Changeset in webkit [202436] by
-
- 2 edits in trunk/Source/WebCore
Do not use iOS specific telephone detection on macOS.
https://bugs.webkit.org/show_bug.cgi?id=159096
rdar://problem/25870571
Reviewed by Anders Carlsson.
Adding platform guard.
- platform/cocoa/TelephoneNumberDetectorCocoa.cpp:
(WebCore::TelephoneNumberDetector::phoneNumbersScanner):
- 11:51 AM Changeset in webkit [202435] by
-
- 44 edits48 deletes in trunk
Unreviewed, rolling out r202413.
https://bugs.webkit.org/show_bug.cgi?id=159097
Broke many JSC tests (Requested by ap on #webkit).
Reverted changeset:
"[JSC] Implement isFinite / isNaN in JS and make DFG ToNumber
accept non number values"
https://bugs.webkit.org/show_bug.cgi?id=154022
http://trac.webkit.org/changeset/202413
- 11:38 AM Changeset in webkit [202434] by
-
- 2 edits in trunk/Source/WebCore
Unreviewed build fix after r202429; AVStreamDataParser does not exist on iOS.
- platform/spi/mac/AVFoundationSPI.h:
- 11:34 AM Changeset in webkit [202433] by
-
- 2 edits in trunk/Tools
parser_unittests.pl should not hardcode list of tests
<https://webkit.org/b/159074>
Reviewed by Daniel Bates.
- Scripts/webkitperl/prepare-ChangeLog_unittest/parser_unittests.pl:
Update to read the list of tests from the 'resources' directory,
and instead map file extensions to subroutine names. Switch to
use File::Basename::fileparse() to get basename and file
extension for each test.
(readTestFiles): Added. Reads files from the directory passed
in and ignores: hidden files, anything that isn't a plain file,
and expected test results files.
- 11:31 AM Changeset in webkit [202432] by
-
- 6 edits17 adds1 delete in trunk/Source/WebKit2
Move WebKit2 Apple Pay code to the open source repository
https://bugs.webkit.org/show_bug.cgi?id=159041
Reviewed by Sam Weinig.
- DerivedSources.make:
- Shared/Cocoa/WebCoreArgumentCodersCocoa.h: Added.
- Shared/Cocoa/WebCoreArgumentCodersCocoa.mm: Added.
- UIProcess/ApplePay/WebPaymentCoordinatorProxy.cpp: Added.
- UIProcess/ApplePay/WebPaymentCoordinatorProxy.h: Added.
- UIProcess/ApplePay/WebPaymentCoordinatorProxy.messages.in: Added.
- UIProcess/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.h: Added.
- UIProcess/ApplePay/cocoa/WebPaymentCoordinatorProxyCocoa.mm: Added.
- UIProcess/ApplePay/ios/WebPaymentCoordinatorProxyIOS.mm: Added.
- UIProcess/ApplePay/mac/WebPaymentCoordinatorProxyMac.mm: Added.
- UIProcess/WebPageProxy.cpp:
- UIProcess/WebPageProxy.h:
- WebKit2.xcodeproj/project.pbxproj:
- WebProcess/ApplePay/WebPaymentCoordinator.cpp: Added.
- WebProcess/ApplePay/WebPaymentCoordinator.h: Added.
- WebProcess/ApplePay/WebPaymentCoordinator.messages.in: Added.
- 11:21 AM Changeset in webkit [202431] by
-
- 3 edits in trunk/Source/WebCore
Unreviewed build fix after r202429; Fix the type of the delegate property on AVStreamDataParser.
- platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:
- platform/spi/mac/AVFoundationSPI.h:
- 11:19 AM Changeset in webkit [202430] by
-
- 3 edits in branches/safari-602.1.38-branch/Source/WebKit2
Merged r202389. rdar://problem/26975138
- 10:46 AM Changeset in webkit [202429] by
-
- 5 edits in trunk/Source/WebCore
[MSE] Adopt +[AVStreamDataParser outputMIMECodecParameterForInputMIMECodecParameter:]
https://bugs.webkit.org/show_bug.cgi?id=158312
Reviewed by Eric Carlson.
- platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaSourceAVFObjC.mm:
(WebCore::MediaPlayerPrivateMediaSourceAVFObjC::supportsType):
Move the declaration of AVStreamDataParser into AVFoundationSPI.h:
- platform/graphics/avfoundation/objc/CDMSessionAVStreamSession.mm:
(WebCore::CDMSessionAVStreamSession::update):
- platform/graphics/avfoundation/objc/MediaPlayerPrivateMediaSourceAVFObjC.mm:
- platform/graphics/avfoundation/objc/SourceBufferPrivateAVFObjC.mm:
(-[WebAVStreamDataParserListener streamDataParser:didProvideMediaData:forTrackID:mediaType:flags:]):
- platform/spi/mac/AVFoundationSPI.h:
- 10:34 AM Changeset in webkit [202428] by
-
- 6 edits in trunk/LayoutTests
Make window.open layout tests for about: URLs work on iOS
https://bugs.webkit.org/show_bug.cgi?id=159072
Reviewed by Alex Christensen.
Fix two window.open layout tests for iOS. They previously used click events.
Now they instead turn off the popup blocker and just call window.open() on
the load event.
- http/tests/dom/window-open-about-blank-and-access-document-expected.txt:
- http/tests/dom/window-open-about-blank-and-access-document.html:
- http/tests/dom/window-open-about-webkit-org-and-access-document-expected.txt:
- http/tests/dom/window-open-about-webkit-org-and-access-document.html:
- platform/ios-simulator/TestExpectations:
Removed the skipping of these two test cases.
- 10:32 AM Changeset in webkit [202427] by
-
- 2 edits in trunk/Source/WebCore
[iOS, Mac] Assume a media file has audio during AirPlay
https://bugs.webkit.org/show_bug.cgi?id=159088
<rdar://problem/24616592>
Reviewed by Jer Noble.
No new tests, it isn't possible to test this in a LayoutTest.
- html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::mediaPlayerCurrentPlaybackTargetIsWirelessChanged): Call
mediaSession->setCanProduceAudio(true) when AirPlay becomes active.
- 10:13 AM Changeset in webkit [202426] by
-
- 2 edits in trunk/Source/WebKit2
[WK2] PageConfiguration.m_relatedPage keeps WebPageProxy alive after its WebView has been destroyed
https://bugs.webkit.org/show_bug.cgi?id=159080
<rdar://problem/26982642>
Reviewed by Anders Carlsson.
PageConfiguration.m_relatedPage keeps WebPageProxy alive after its WebView has been destroyed.
This leads to WebPageProxy object leaks when the client keeps chaining related views.
In this patch, we null our the PageConfiguration's relatedPage in WebPageProxy::close()
to break the chain and make sure we only keep alive 1 relatedPage rather than a chain of
them.
- UIProcess/WebPageProxy.cpp:
(WebKit::WebPageProxy::close):
- 10:12 AM Changeset in webkit [202425] by
-
- 18 edits in trunk/Source
Playback controls refer to wrong element when playing multiple items in a page.
https://bugs.webkit.org/show_bug.cgi?id=159076
<rdar://problem/26953532>
Reviewed by Beth Dakin.
Source/WebCore:
Use a new method PlatformMediaSessionManager::currentSessionMatching() to get
the most recently active media element which qualifies for playback controls.
- html/HTMLMediaElement.cpp:
(WebCore::HTMLMediaElement::updatePlaybackControlsManager): Get the most recently active session.
- html/MediaElementSession.cpp:
(WebCore::MediaElementSession::canControlControlsManager): Make virtual; no longer takes an element.
- html/MediaElementSession.h:
(isType): Allow downcasting from PlatformMediaSession -> MediaElementSession.
- page/ChromeClient.h:
- platform/audio/PlatformMediaSession.h:
(WebCore::PlatformMediaSession::canControlControlsManager): Defaults to false;
- platform/audio/PlatformMediaSessionManager.cpp:
(WebCore::PlatformMediaSessionManager::currentSessionMatching): Added.
- platform/audio/PlatformMediaSessionManager.h:
Source/WebKit/mac:
clearPlaybackControlsManager() no longer takes a parameter.
- WebCoreSupport/WebChromeClient.h:
- WebCoreSupport/WebChromeClient.mm:
(WebChromeClient::clearPlaybackControlsManager):
- WebView/WebView.mm:
(-[WebView _clearPlaybackControlsManager]): Renamed from _clearPlaybackControlsManagerForMediaElement:.
- WebView/WebViewInternal.h:
Source/WebKit2:
clearPlaybackControlsManager() no longer takes a parameter.
- WebProcess/WebCoreSupport/WebChromeClient.cpp:
(WebKit::WebChromeClient::clearPlaybackControlsManager):
- WebProcess/WebCoreSupport/WebChromeClient.h:
- WebProcess/cocoa/WebPlaybackSessionManager.h:
- WebProcess/cocoa/WebPlaybackSessionManager.mm:
(WebKit::WebPlaybackSessionManager::setUpPlaybackControlsManager): Make sure to removeClientForContext()
if the old context refers to a live media element.
(WebKit::WebPlaybackSessionManager::clearPlaybackControlsManager):
- 10:05 AM Changeset in webkit [202424] by
-
- 2 edits in trunk/Source/WebKit2
IDB tests crashing attempting to register sandbox extensions for blob files.
https://bugs.webkit.org/show_bug.cgi?id=159094.
Reviewed by Sam Weinig.
Already tested by storage/indexeddb/modern/handle-user-delete.html
- WebProcess/Databases/IndexedDB/WebIDBConnectionToServer.cpp:
(WebKit::preregisterSandboxExtensionsIfNecessary): Check whether the IDBResultData actually contains
a getResult before trying to access it.
- 9:57 AM Changeset in webkit [202423] by
-
- 2 edits in trunk/Source/WebCore
Fixed the macOS build.
- platform/spi/cocoa/DataDetectorsCoreSPI.h:
- 9:24 AM Changeset in webkit [202422] by
-
- 5 edits in trunk/Source/WebCore
[iOS] Inline DataDetectorsAdditions.h
https://bugs.webkit.org/show_bug.cgi?id=159093
Reviewed by Anders Carlsson.
- editing/cocoa/DataDetection.mm:
(WebCore::constructURLStringForResult): Use soft-linked constant directly.
- platform/cocoa/DataDetectorsCoreSoftLink.h: Declare soft-linked constant.
- platform/cocoa/DataDetectorsCoreSoftLink.mm: Define soft-linked constant.
- platform/spi/cocoa/DataDetectorsCoreSPI.h: Declare constant.
- 9:05 AM Changeset in webkit [202421] by
-
- 4 edits2 adds in trunk
[GTK][EFL] ImageBufferCairo should accept resolution factor
https://bugs.webkit.org/show_bug.cgi?id=157848
Reviewed by Martin Robinson.
Source/WebCore:
ImageBufferCairo ignored the resolution factor passed in its constructor.
This resolution factor is originally introduced for HiDPI Canvas,
and since HiDPI canvas is not enabled in the ports using Cairo,
the lack of this implementation does not cause any problems.
And now, HiDPI Canvas is removed from the tree.
However, WebKit CSS filter uses this path.
The missing implementation is required under the HiDPI environment.
Since Cairo surface can have the device scale factor transparently,
the operations onto the surface is correctly done in the logical coordinate system.
So all we need to handle carefully is the direct surface modification done
in filter effects.
In this patch, we extend the image buffer size according to the resolution factor,
as the same to the CoreGraphics' implementation (ImageBufferCG). And by setting the
device scale factor of the surface correctly, we ensure that the rest of the Cairo
painting stack works with the existing logical coordinate system. And in ImageBufferCairo,
we carefully handle the logical and backing store coordinate system.
The attached test applies the CSS filter onto the svg image. And we resize the image size,
and perform scrolling. It incurs the paint, and filter effect recalcuation.
In that path, the filter effect side assumes that the image buffer size is scaled with the
resolution factor. So without this patch, it incurs buffer overflow and leads WebProcess crash.
- platform/graphics/IntPoint.h:
(WebCore::IntPoint::scale):
- platform/graphics/cairo/ImageBufferCairo.cpp:
(WebCore::ImageBufferData::createCompositorBuffer):
(WebCore::ImageBuffer::ImageBuffer):
(WebCore::ImageBuffer::copyImage):
(WebCore::ImageBuffer::platformTransformColorSpace):
(WebCore::getImageData):
(WebCore::logicalUnit):
(WebCore::backingStoreUnit):
(WebCore::ImageBuffer::getUnmultipliedImageData):
(WebCore::ImageBuffer::getPremultipliedImageData):
(WebCore::ImageBuffer::putByteArray):
(WebCore::ImageBuffer::copyToPlatformTexture):
LayoutTests:
- fast/hidpi/filters-and-image-buffer-resolution-expected.html: Added.
- fast/hidpi/filters-and-image-buffer-resolution.html: Added.
- 7:39 AM Changeset in webkit [202420] by
-
- 55 edits in trunk
Refactor RenderMathMLOperator and RenderMathMLToken to avoid using anonymous renderers.
https://bugs.webkit.org/show_bug.cgi?id=155018
Patch by Frederic Wang <fwang@igalia.com> on 2016-06-24
Reviewed by Martin Robinson.
Source/WebCore:
No new tests, already covered by existing tests.
We use MathOperator for RenderMathMLOperator to avoid creating anonymous text nodes again
and again. We reimplement implicit mathvariant="italic" on single-char mi in a way that does
not rely on creating anonymous text nodes. Finally, we improve the determination/update of
when mathvariant is italic to avoid breaking foreign-mi-dynamic test.
The change in the render tree structure breaks mfenced accessibility support but that will
be fixed in follow-up patches. The simplifications made here will also allow to simplify the
accessibility code.
- css/mathml.css:
(mo): Deleted. This flexbox rule is no longer needed.
- rendering/mathml/RenderMathMLBlock.cpp:
(WebCore::RenderMathMLBlock::createAnonymousMathMLBlock): Deleted. We no longer need to
create anonymous renderer with this function.
- rendering/mathml/RenderMathMLBlock.h: Delete createAnonymousMathMLBlock.
- rendering/mathml/RenderMathMLOperator.cpp: Implement layout functions without relying on
flexbox or anonymous.
(WebCore::RenderMathMLOperator::computePreferredLogicalWidths): Handle the case of !useMathOperator()
for which we need to add extra operator spacing after the RenderMathMLToken layout.
(WebCore::RenderMathMLOperator::layoutBlock): Ditto.
(WebCore::RenderMathMLOperator::isChildAllowed): Deleted. We allow the non-anonymous text.
(WebCore::RenderMathMLOperator::rebuildTokenContent): No longer destroy and rebuild
anonymous wrapper. Remove updateStyle call.
(WebCore::RenderMathMLOperator::updateStyle): Deleted. We no longer need anonymous style for the spacing.
- rendering/mathml/RenderMathMLOperator.h: Remove updateStyle() and isChildAllowed().
Make textContent() public so that it can be accessed from the accessibility code.
- rendering/mathml/RenderMathMLToken.cpp: Reimplement implicit mathvariant="italic" by
painting MATHEMATICAL ITALIC characters instead of styling an anonymous wrapper.
(WebCore::RenderMathMLToken::RenderMathMLToken): Init m_mathVariantGlyph and m_mathVariantGlyphDirty
(WebCore::RenderMathMLToken::updateTokenContent): Set mathvariant glyph dirty when the content changes.
(WebCore::transformToItalic): Helper function to map latin and greek alphabets to their
MATHEMATICAL ITALIC counterpart.
(WebCore::RenderMathMLToken::computePreferredLogicalWidths): Implement this function to
handle the case where the mathvariant glyph is used.
(WebCore::RenderMathMLToken::updateMathVariantGlyph): Helper function to update the mathvariant glyph.
For now, we try and keep with the old (and limited) implementation: a mathvariant glyph may
only used for single-char <mi> without mathvariant attribute attached to it.
(WebCore::RenderMathMLToken::styleDidChange): Set the mathvariant glyph dirty when the style
changes.
(WebCore::RenderMathMLToken::updateFromElement): Remove updateStyle call and set mathvariant
glyph dirty.
(WebCore::RenderMathMLToken::firstLineBaseline): Implement this function to handle the case
where the mathvariant glyph is used.
(WebCore::RenderMathMLToken::layoutBlock): Ditto.
(WebCore::RenderMathMLToken::paint): Ditto.
(WebCore::RenderMathMLToken::paintChildren): Ditto.
(WebCore::RenderMathMLToken::addChild): Deleted. No need to bother with anonymous renderer
or style.
(WebCore::RenderMathMLToken::createWrapperIfNeeded): Deleted. Ditto.
(WebCore::RenderMathMLToken::updateStyle): Deleted. Ditto.
- rendering/mathml/RenderMathMLToken.h: Update declarations of functions.
(WebCore::RenderMathMLToken::setMathVariantGlyphDirty): Helper function to indicate that the
mathvariant glyph will need to be updated.
LayoutTests:
We update the test expectations. Most of the adjustments are irrelevant visual
modifications (small shift and size differences) or due to changes in the renderer tree
structure (removal of anonymous, operator text appearing in the dumped text).
We disable the accessibility for mfenced for now, but this will be enabled again and improved
in follow-up accessibility patches.
- platform/ios-simulator/TestExpectations: We disable two reftests that use a character from
the Mathematical Alphanumeric Symbols block as a reference, since iOS does not have
appropriate fonts pre-installed.
- accessibility/mac/mathml-elements.html: Disable the mfenced test for now.
- accessibility/mac/mathml-elements-expected.txt: Ditto.
- accessibility/ios-simulator/math-expected.txt: Update the reference now that the text of
the operator appears in the render tree.
- accessibility/mac/math-alttext-expected.txt: Ditto.
- mathml/opentype/large-operators-italic-correction-expected.txt: Ditto.
- mathml/opentype/large-operators-munderover-expected.txt: Ditto.
- mathml/presentation/inferred-mrow-stretchy-expected.txt: Ditto.
- mathml/presentation/mfrac-nested-expected.txt: Ditto.
- mathml/presentation/stretchy-depth-height-expected.txt: Ditto.
- mathml/presentation/stretchy-depth-height-symmetric-expected.txt: Ditto.
- mathml/presentation/sup-nested-expected.txt: Ditto.
- mathml/very-large-stretchy-operators-expected.txt: Ditto.
- platform/mac/accessibility/math-text-expected.txt: Ditto.
- platform/gtk/accessibility/math-text-expected.txt: Ditto.
- fast/css/readonly-pseudoclass-common-element-expected.txt: Ditto.
- fast/css/readwrite-pseudoclass-editable-expected.txt: Ditto.
- imported/w3c/web-platform-tests/html/dom/documents/dom-tree-accessors/document.getElementsByName/document.getElementsByName-namespace-expected.txt: Ditto.
- imported/w3c/web-platform-tests/html/dom/documents/dom-tree-accessors/document.getElementsByName/document.getElementsByName-namespace-xhtml-expected.txt: Ditto.
- mathml/presentation/attributes-background-color.html: Add some padding to workaround some
small rendering failures.
- mathml/presentation/attributes-background-color-expected.html: Ditto.
- mathml/presentation/attributes-display.html: Ditto.
- mathml/presentation/attributes-display-expected.html: Ditto.
- mathml/presentation/bug97990-expected.html: Ditto.
- mathml/presentation/bug97990.html: Ditto.
- mathml/presentation/attributes-mathvariant-expected.html: Update the reftest to use italic
characters from the Mathematical Alphanumeric Symbols.
- mathml/presentation/tokenElements-mathvariant-expected.html: Ditto.
- mathml/presentation/mo-whitespaces-expected.html: Adjust one edge case since we now only
collapse whitespace for operators that use MathOperator.
- accessibility/math-multiscript-attributes-expected.txt: Small size/coordinate adjustment.
- platform/gtk/accessibility/math-multiscript-attributes-expected.txt: Dottp.
- platform/gtk/mathml/presentation/roots-expected.txt: Size/coordinate adjustment, removal
of anonymous renderers and use italic characters from the Mathematical Alphanumeric Symbols.
- platform/gtk/mathml/presentation/roots-expected.png: Ditto.
- platform/ios-simulator/mathml/presentation/roots-expected.txt: Ditto.
- platform/gtk/mathml/opentype/horizontal-expected.txt: Size/coordinate adjustment and
removal of anonymous renderers.
- platform/gtk/mathml/opentype/horizontal-munderover-expected.txt: Ditto.
- platform/gtk/mathml/opentype/large-operators-expected.txt: Ditto.
- platform/gtk/mathml/opentype/opentype-stretchy-expected.txt: Ditto.
- platform/gtk/mathml/opentype/opentype-stretchy-horizontal-expected.txt: Ditto.
- platform/gtk/mathml/opentype/vertical-expected.txt: Ditto.
- platform/gtk/mathml/presentation/mo-stretch-expected.txt: Ditto.
- platform/ios-simulator/mathml/opentype/horizontal-munderover-expected.txt: Ditto.
- platform/ios-simulator/mathml/opentype/opentype-stretchy-expected.txt: Ditto.
- platform/mac/mathml/opentype/large-operators-expected.txt: Ditto.
- platform/mac/mathml/opentype/opentype-stretchy-expected.txt: Ditto.
- platform/mac/mathml/opentype/opentype-stretchy-horizontal-expected.txt: Ditto.
- platform/mac/mathml/opentype/vertical-expected.txt: Ditto.
- 12:39 AM Changeset in webkit [202419] by
-
- 2 edits in trunk/Source/WebCore
Unreviewed EFL build fix.
There is forward declaration build error on EFL port.
- platform/graphics/texmap/coordinated/CompositingCoordinator.cpp: Include DOMWindow.h and Document.h.
