Automatic update from web-platform-tests
Clear ImageLoader image state if load is attempted in destroyed document
When the document an ImageLoader is associated with is detached
(i.e destroyed), the ImageLoader will not get any notification about it.
Thus there could be tasks pending.
If this case is discovered when the microtask (DoUpdateFromElement())
runs clear the ImageLoaders image state.
Bug: 1428079
Change-Id: Ib8bdad6ce71910ddc727efa0903e9e5bc2d35078
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4374008
Commit-Queue: Fredrik Söderquist <fs@opera.com>
Reviewed-by: Dominic Farolino <dom@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1123590}
--
wpt-commits: 24f063acbba441edb1d921673b4fd6bee3f4e628
wpt-pr: 39248
Automatic update from web-platform-tests
[@scope] Propagate proximity from SubResult
This fixes the proximity of nested rules which contain a nesting
selector (&) that refer to a selector with a kScopeActivation.
Fixed: 1428648
Change-Id: Ibb9698062dc4a2ba88802bcbfb924c2854a7e1c1
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4373876
Reviewed-by: Rune Lillesveen <futhark@chromium.org>
Commit-Queue: Anders Hartvoll Ruud <andruud@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1123508}
--
wpt-commits: 3cda307aba9cfa3c4ae8c3c2c2839418ba3768b5
wpt-pr: 39266
Automatic update from web-platform-tests
Remove thread check in RTPSender UnregisterEncoded{A/V}Callback()
Also make the encoded_audio_transformer_ and encoded_video_transformer_
fields actually const to ensure that these methods are, and remain,
threadsafe.
Also add a wpt of an encodedInsertableStream worker transform triggering
an error, which actually exercises this method on the worker thread -
see attached bug.
Bug: 1428006
Change-Id: Ia7b9715c3663e7ba1d753dc504ac3b0597e68289
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4372842
Reviewed-by: Guido Urdaneta <guidou@chromium.org>
Commit-Queue: Tony Herre <toprice@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1123506}
--
wpt-commits: 8835eb21a8c3d67e952ece4c386c17dccbdab441
wpt-pr: 39249
Automatic update from web-platform-tests
initial role verification cases (#38925)
* simple role cases and region special cases
* synonym roles tests, reviewer+, and minor errata
* synonym roles tests, reviewer+, and minor errata
* use assert_equals instead of assert_true
* use tagName with implicit role to test non-generic fallback
* remove synonym role test; will spawn a new issue
* add list roles and break scripts into a shared utils file
* m. whitespace.
* new roles ReadMe
* clarity in attr name and removing code duplication
* m. lint tabs -> spaces
* review feedback from @zcorpan; include a permanent and unique test name for dashboard/logs.
* Gecko-specific fix from @zcorpan
--
wpt-commits: e364f98f0dbb8026f64e6c7e3de7d8bd524d1532
wpt-pr: 38925
Automatic update from web-platform-tests
[FLEDGE] Rename some WPT files.
Use dashes instead of underscores. Dashes are more common in WPT tests.
Most files are already using dashes, but accidentally used underscores
in two of them.
Bug: None
Change-Id: I95dd5d3e326d4280cd82b033b7deac15ee533524
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4378337
Reviewed-by: Maks Orlovich <morlovich@chromium.org>
Commit-Queue: Matt Menke <mmenke@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1123259}
--
wpt-commits: b8dacfea9f4790adbb49f19c888e5fffabe9b039
wpt-pr: 39258
Automatic update from web-platform-tests
[white-space] Turn `white-space` to a shorthand
This patch turns the `white-space` property to a shorthand
under a runtime switch `CSSWhiteSpaceShorthand`.
The parser and serializer in this CL supports the spec change
committed on 1st March, 2023[1], which addressed the
multi-value syntax request[2].
This patch also adds:
* The `white-space-collapse` property.
* The `nowrap` value to the `text-wrap` property.
to map all the existing `white-space` values.
The switch is currently off, because enabling it causes
several test failures in DOM, editing, copy/paste,
accessibility, and more, but when it's turned on, CSS parsing
tests added in this CL should pass. The plan is to fix these
failures in following CLs. When all tests and additional tests
for editing/etc. are all fixed, we can turn the switch on.
The `EWhiteSpace` enum is changed to bit-flags of the
longhands (`white-space-collapse` and `text-wrap`.) See
`white_spae.h`.
Regarding the storage in the `ComputedStyleBase`, having both
storage (for when the flag is on and off) is a waste, both
from memory and from read/write code performance point of view,
so the longhands storage is used even when the flag is off by
customized `ComputedStyle` functions; i.e., the storage for
the longhand version of the `white-space` is no longer used.
But this CL still preserves 1 bit for the `white-space` to
avoid changing the generator only for this transition period.
[1] https://github.com/w3c/csswg-drafts/pull/8274
[2] https://github.com/w3c/csswg-drafts/issues/5382
Bug: 1417543
Change-Id: I47080d0f10a4a5f366e3d855acfa5e584c029595
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4232738
Reviewed-by: Anders Hartvoll Ruud <andruud@chromium.org>
Commit-Queue: Koji Ishii <kojii@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1123192}
--
wpt-commits: c5dc05f232b36134c54031244de60d7bd5c32721
wpt-pr: 39244
Automatic update from web-platform-tests
Make NavigateReaction private to NavigateEvent, attempt #2
Most of the logic now lives on NaviagateEvent itself (in React()
and ReactDone()), which the new NavigateEvent::Reaction is a minimal
callback function wrapper.
Change-Id: Id31152b048fc7166a8ac45bda7e136a2143fd56f
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4369772
Reviewed-by: Domenic Denicola <domenic@chromium.org>
Commit-Queue: Nate Chapin <japhet@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1123074}
--
wpt-commits: 22357f00690d85b0fcf8b569926e19eac48e5d05
wpt-pr: 39231
Automatic update from web-platform-tests
Handle the case where a new navigation happens synchronously inside RenderFrameImpl::CommitSameDocumentNavigation
This is broken after https://chromium-review.googlesource.com/c/chromium/src/+/4300672
Bug: 1427449
Change-Id: Ie6ec0a597eeb45e3a7ae11215fc010d284a997bf
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4375397
Commit-Queue: Nate Chapin <japhet@chromium.org>
Reviewed-by: Rakina Zata Amni <rakina@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1123066}
--
wpt-commits: 94aec983f166c374178fa6761d462342edc62ddf
wpt-pr: 39226
Automatic update from web-platform-tests
LoAF: Expose pauseDuration
Pause duration is time spent in programmatic pauses,
i.e. synchronous XHR or JS dialogs (alert/confirm/prompt).
Added a test for sync XHR, alerts require WPT infra
(TODO: test for alerts, requires changes to WPT infra)
Bug: 1392685
Change-Id: I835871c7ba86f37a4d01d0dfe5bf71ea2083fc67
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4373947
Commit-Queue: Noam Rosenthal <nrosenthal@chromium.org>
Reviewed-by: Yoav Weiss <yoavweiss@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1123036}
--
wpt-commits: 390d91561e2d01c14b10d0986618d98c4d968e30
wpt-pr: 39243
Automatic update from web-platform-tests
Replace TrustedType spec URL
[chrome-security-owp-fixit-week-9]
The TrustedType repository moved toward w3c. As a result, a lot of link
are either invalid or behind a redirect. This patch correct the links.
This is an automated patch. It only update comments and documentation. Some
invalid links are now fixed.
Script:
```
old="github.com\/wicg\/trusted-types"
new="github.com\/w3c\/trusted-types"
old="github.com\/WICG\/trusted-types"
new="github.com\/w3c\/trusted-types"
old="github.com\/w3c\/trusted-types\/#"
new="w3c.github.io\/trusted-types\/dist\/spec\/#"
old="WICG\/trusted-types"
new="w3c\/trusted-types"
old="w3c.github.io\/webappsec-trusted-types\/dist\/spec\/"
new="w3c.github.io\/trusted-types\/dist\/spec\/"
for file in chrome/ third_party/blink content/ services/ docs/ third_party/devtools-frontend
do
echo $file
grep -r -l "${old}" "$file" | sort | uniq | xargs perl -e "s/${old}/${new}/" -pi
done
```
Fixed: 1386115
Bug: 1386115
Change-Id: If9650dec718b9153f6921ab080d2baf60332eae6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4376759
Reviewed-by: Daniel Vogelheim <vogelheim@chromium.org>
Commit-Queue: Arthur Sonzogni <arthursonzogni@chromium.org>
Owners-Override: Arthur Sonzogni <arthursonzogni@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1122970}
--
wpt-commits: dedf646b32e212a97a2d7451c74d778435f05cce
wpt-pr: 39245
Automatic update from web-platform-tests
Add WPT for onstart compliance.
This adds a WPT for the change added in crrev.com/c/4349969.
Bug: chromium:1422582, chromium:1423413
Change-Id: Iec7df62b63397194ede583b20648738243291d31
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4360296
Reviewed-by: Henrik Boström <hbos@chromium.org>
Commit-Queue: Markus Handell <handellm@google.com>
Cr-Commit-Position: refs/heads/main@{#1122952}
--
wpt-commits: 518f585d58c2ec86a0cb50b275fc459746ce6fb0
wpt-pr: 39141
Automatic update from web-platform-tests
fetch: Add a test for SRI check with null body status
a fetch with SRI metadata should abort when the response has a
null body status, per step 22.2 of [1].
[1] https://fetch.spec.whatwg.org/#main-fetch
Bug: 1427319
Change-Id: I4635c77639589ce17fa4cd215a65792a7c330494
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4369033
Commit-Queue: Kenichi Ishibashi <bashi@chromium.org>
Reviewed-by: Adam Rice <ricea@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1122899}
--
wpt-commits: 62293ebf20600d4d27c6b6b77df94b49fae18e54
wpt-pr: 39212
Automatic update from web-platform-tests
Consistently encompass fragmentainer overflow in the container.
Monolithic content may cause a fragmentainer to be overflowed, and the
container(s) of the monolithic node should then grow to hold the
monolithic content (and thereby also overflow the fragmentainer), if
allowed.
Some of the tests included here were already passing, but not all.
For block containers we were doing it wrong if the container of the
overflowing (monolithic) fragment had more children after the monolithic
node, whereas we got it right if it ended in the current fragmentainer.
See monolithic-overflow-002.tentative.html vs.
monolithic-overflow-003.tentative.html .
For flex containers we were already doing it right.
For grid containers and tables we were doing it wrong, because we didn't
provide a correct intrinsic block-size (which is used by the
fragmentation machinery to handle unbreakable content correctly).
One additional complication for grid and tables was that we didn't
handle the fact that cells or grid items might (now) be stretched by
monolithic content, whereas sibling cells / items aren't. So we had to
add support for that. Flex layout also already handles this part
correctly, in GiveItemsFinalPositionAndSizeForFragmentation(), by
compensating for the difference between space consumed by the flex line
and the space consumed by the flex item. Do the same for grid and
tables.
Although this behavior isn't specified, we're now more compatible with
Gecko. This behavior also makes more sense. With the previous behavior a
child might overflow its container for no reason (the monolithic child
would overflow one fragmentainer, whereas the container would fragment
at the fragmentation line).
This change also makes it much easier to fix crbug.com/1402540 , so that
we can just very easily walk past monolithic content that bleeds into
subsequent pages.
Bug: 1425077, 1378607
Change-Id: I8ebe706b48fa18abb669041843baef74f021a340
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4370497
Reviewed-by: Alison Maher <almaher@microsoft.com>
Commit-Queue: Morten Stenshorne <mstensho@chromium.org>
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1122895}
--
wpt-commits: a2c3d90657a9d510fc80ded0b4c0317e86c3dfa8
wpt-pr: 39224
Automatic update from web-platform-tests
Update ReadableStreamDefaultReader implementation
This change includes updating the ReadableStreamDefaultReader
implementation to match the Streams spec. This means properly
implementing read requests so that they are not just
StreamPromiseResolvers, and have separate chunk, close, and error steps
based on where the read request is used. This CL also uses
ReadableStreamReadResult to match the spec, and finally removes the use
of `for_author_code_` which is no longer used in the standard. These
changes in turn also fix the teeing behavior.
Bug: 1227496
Change-Id: Id43ece0bef5fb5a7588e8adf64b5a4e398d5163e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/3763542
Commit-Queue: Nidhi Jaju <nidhijaju@chromium.org>
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Reviewed-by: Adam Rice <ricea@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1122821}
--
wpt-commits: 08d7b0e170a2a39dc437ecd9b61c6d8b24d0c004
wpt-pr: 38875
Automatic update from web-platform-tests
URL: cleanup urltestdata.json
It had several subtle errors:
* Empty string, "#", and "?" input was not marked with inputCanBeRelative.
* #29579 allowed base to be null, but this wasn't documented and led to weird special casing.
This then resulted in failures results in wrappers, such as failure.html, despite there not actually being a bug in the implementation.
This cleans up the model to be in line with the URL Standard and adjusts the various wrappers accordingly.
--
wpt-commits: 6d461b4ddb2f1b8d226ca6ae92e14bbd464731a5
wpt-pr: 39203
Automatic update from web-platform-tests
Prevent fieldsets from being disabled form controls
Fieldset elements themselves aren't actually supposed to be disabled
form controls, they are just supposed to make their descendant field
control elements disabled:
https://html.spec.whatwg.org/multipage/form-control-infrastructure.html#concept-fe-disabled
This was causing a bug where SendMouseEventsDisabledFormControls
prevents click/mouseup/mousedown events from being propagated at or
above disabled fieldset elements.
Fixed: 1422096, 1422547
Bug: 588760
Change-Id: Ifb76ac92f7f36b097666b4384f6de4bd561547cd
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4321065
Reviewed-by: Mason Freed <masonf@chromium.org>
Commit-Queue: Joey Arhar <jarhar@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1122658}
--
wpt-commits: de685a735ebc0acd8c2b5a06869b79f5f1815863
wpt-pr: 38887
Automatic update from web-platform-tests
[GridNG] Flip rows and columns for orthogonal sugrids
Orthogonal grids flip rows and columns, and subgrids need to support
this when combining lines from the parent grid. This gets us to pass
repeat-auto-fill-003.html, however a slight adjustment was made to the
the test because baselines didn't match the reference. We match WebKit,
but Firefox appears to have a bug with orthogonal baselines. I came
across this WebKit discussion that implies that *they* are incorrect
(and thus we would be too), but I believe Firefox is incorrect here.
See [this discussion](https://bugs.webkit.org/show_bug.cgi?id=236958).
Note that as of the most recent rebase, it picked up
https://chromium-review.googlesource.com/c/chromium/src/+/4355553,
and repeat-auto-fill-003.html no longer passes, so it was added back
to TestExpectations. This should start passing once those regressions
are fixed.
Either way, this behavior is unrelated to the actual test content, so
I made some adjustments to get the baselines to match in all
aforementioned platforms. In particular, I set `line-height`to zero and
set `vertical-align` to `top`.
repeat-auto-fill-003.html is an interesting case, and I believe it's
the only test with nested orthogonal subgrids, so I added a new test
that covers more than auto repeat behaviors.
These changes exposed a bug from my prior auto repeat checkin - the
adjustment for non-repeat lines that come after an auto repeater
shouldn't happen on nested subgrids, as it will shift lines multiple
times for each level of nesting. Ideally, this would happen when we
add the non-repeat named lines. For now, I added a TODO, as this will
require some additional work and testing.
Bug: 618969
Change-Id: Ic8c42189d0563e6b2ed4f322ae89b0583c1fb519
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4355586
Reviewed-by: Alison Maher <almaher@microsoft.com>
Reviewed-by: Ethan Jimenez <ethavar@microsoft.com>
Commit-Queue: Kurt Catti-Schmidt <kschmi@microsoft.com>
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1122607}
--
wpt-commits: 1cfda889f6a15b4d608553f60cdfa6022fcf5e21
wpt-pr: 39130
Automatic update from web-platform-tests
[Topics] add Topics header for iframe navigation request
Implement topics navigation header:
- Before a request or redirect is handled (i.e. NavigationRequest's
constructor, and OnRedirectChecksComplete()), calculate and append
the header.
- After the corresponding response is received (i.e.
OnRedirectChecksComplete() and CommitNavigation()), check and handle
the topics response header.
Also, fixed the checks in HTMLIFrameElement::ParseAttribute() and
HTMLIFrameElement::DidChangeAttributes(): currently, the RuntimeFeature
and SecureContext are not properly checked when the iframe attribute is
set via setAttribute. The annotation in the IDL wasn't sufficient.
Bug: 1352345
Change-Id: Ia84c67e8f9858a98d7161668bc82fa688c563497
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4356935
Reviewed-by: Alex Moshchuk <alexmos@chromium.org>
Reviewed-by: Josh Karlin <jkarlin@chromium.org>
Commit-Queue: Yao Xiao <yaoxia@chromium.org>
Reviewed-by: Dominic Farolino <dom@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1122570}
--
wpt-commits: d21cfea4438a4f82599a8f493d5371cfc29c3a05
wpt-pr: 39097
Automatic update from web-platform-tests
[GridNG] Fix computed style values for non-layout subgrids
This CL makes some changes in order to round-trip specified values when
`subgrid` is specified on a non-subgrid.
This necessitated a complete rewrite of how these values are queried
for non-grids, as I discovered that the existing implementation
improperly expands integer repeats when an auto repeat exists anywhere
in the definition. This was an omission in the test cases, so I added
various examples of this to grid-template-columns-computed-nogrid.html.
The biggest change here is that in order to round-trip the specified
values, we need to store some additional data in `NGGridTrackRepeater`.
This includes creating repeaters of type `kNoRepeat` for subgrids,
which brings them closer to standalone grids.
However, this had some implications elsewhere. Layout needed some extra
checks for subgrid when it builds the track collections, and placement
behaves incorrectly when `track_count_without_auto_repeat_`
incorporates these `kNoRepeat` tracks.
grid-template-computed-nogrid.html didn't have any cases for integer
repeats, so I added a bunch, both with and without auto repeats.
Bug: 618969
Change-Id: If348f6b099a9032f067fa2299c9d3444d2f52a26
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4317507
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Reviewed-by: Alison Maher <almaher@microsoft.com>
Reviewed-by: Ethan Jimenez <ethavar@microsoft.com>
Commit-Queue: Kurt Catti-Schmidt <kschmi@microsoft.com>
Cr-Commit-Position: refs/heads/main@{#1122559}
--
wpt-commits: 8233b6326d0d3cffccebfe4743a3692857d52696
wpt-pr: 38921
Automatic update from web-platform-tests
Check cookie accessibility in requestStorageAccess WPTs
This CL extends the existing WPTs for document.requestStorageAccess
such that they actually verify the accessibility of cookies after
storage access is obtained. This is just a check that the API does
what it's supposed to do, in addition to returning the result that
it's supposed to.
This would have caught https://crbug.com/1409163,
https://crbug.com/1409162, and https://crbug.com/1408969.
Change-Id: Icf53e696a709398f7cf0079d0047a9e5cba20b5e
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/4369340
Commit-Queue: Matt Reichhoff <mreichhoff@chromium.org>
Auto-Submit: Chris Fredrickson <cfredric@chromium.org>
Reviewed-by: Matt Reichhoff <mreichhoff@chromium.org>
Cr-Commit-Position: refs/heads/main@{#1122494}
--
wpt-commits: 331357d448fcfa956ca2ae36ac41802b2ba80631
wpt-pr: 39187
Fredrik Söderquist
Vitor Roriz
Marcos Cáceres
moz-wptsync-bot
Anders Hartvoll Ruud
Tony Herre
James Craig
Matt Menke
Koji Ishii
Nate Chapin
Noam Rosenthal
Arthur Sonzogni
dependabot[bot]
Markus Handell
Kenichi Ishibashi
Morten Stenshorne
Nidhi Jaju
Anne van Kesteren
Joey Arhar
Kurt Catti-Schmidt (SCHMIDT)
Yao Xiao
Chris Fredrickson