Automatic update from web-platform-tests
Allow hiding test state for first-contentful-paint tests (#24170)
Some tests are sensitive to the content of the document, thus
show_status may interfere the test results.
We add a new setup property to allow hiding the test state output.
--
wpt-commits: ec9f61df8671488160d20a560ee7579126f8a24d
wpt-pr: 24170
Automatic update from web-platform-tests
Update picture-in-picture idlharness test (#24235)
The actual idl update already happened in
1449b69c39,
but the associated test was not updated.
Closes https://github.com/web-platform-tests/wpt/pull/24226
--
wpt-commits: b70f4c4663087bd441962e54ee6d062d39fc2ab5
wpt-pr: 24235
Automatic update from web-platform-tests
Python 3: port some tests in common and infrastructure (#24482)
--
wpt-commits: 924cdb5f242893239b3835a5c78a5a7a8deec240
wpt-pr: 24482
Automatic update from web-platform-tests
[NativeFS] Stop marking WPT tests as tentative.
All these tests test what is in the spec, so these shouldn't be marked
as tentative anymore.
Bug: none
Change-Id: Ib494c57b3c258e9a7e80ac3fd2a4df25abc4783a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2283735
Commit-Queue: Victor Costan <pwnall@chromium.org>
Auto-Submit: Marijn Kruisselbrink <mek@chromium.org>
Reviewed-by: Victor Costan <pwnall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#785593}
--
wpt-commits: ff124a192986ad53d0af652f99b277e3019701ad
wpt-pr: 24472
Automatic update from web-platform-tests
Switch WPT to use Taskcluster Checks integration (#24151)
See https://github.com/web-platform-tests/wpt/issues/14846; this should
only be landed once wpt.fyi is ready to receive check_suite events.
Co-authored-by: Robert Ma <robertma@chromium.org>
--
wpt-commits: 972cc74af2d621fb45e5097350f4ee7d3b7c6ef4
wpt-pr: 24151
Automatic update from web-platform-tests
Move six import after path is set (#24484)
Co-authored-by: Stephen McGruer <smcgruer@chromium.org>
--
wpt-commits: 47821bc96914cfb6a42832490c987e16408e58be
wpt-pr: 24484
Automatic update from web-platform-tests
Move sourcefile construction to the manifest update function
Instead of having the tree iterator return either paths or constructed
sourcefiles, just have it always return paths, and construct the
sourcefiles in the caller. This has the advantage that we can better
handle some error conditions that may be caused by having an invalid
cache. It is also somewhat clearer since we don't end up with an
iterator that can return quite different types depending on whether it
thinks the file was updated or not.
--
Fix a bunch of Python 3 issues in the manifest code
The goal here is to get mypy passing in Python 3; prior to this patch
there were a lot of cases where Python 3 behaviour was broken but not
picked up by the tests (e.g. because it only happened when updating
the manifest or with certain caches or similar). Many of these issues
were seen by mypy, but that had been disabled for Py3 specfically to
avoid having to fix all the issues at the time the annotations were
added. But now we do need that support.
The general approach here is to use the same types for Py 2 and Py
3. In particular we use Text where possible (with a couple of
exceptions where it was easier not to), rather than bytes. The native
str type is avoided wherever possible since this introduces difficult
to debug differences between Python versions.
The preference for Text includes paths, which means we no longer
support running in non-unicode directories (in practice it is likely
to be broken if the path isn't UTF-8 compatible on Unix or UTF-16
compatible on Windows).
--
Fix passing arguments from wpt to Python 3
--
Update six in docker image
--
Run pip3 when installing in Python 3
--
Ensure deps are installed for py3
--
Update vendored six to latest
--
Use pip2 rather than just pip
Seems like bare pip started defaulting to python 3
--
Ensure that html5lib test build uses Python 2
--
Make canvas tests use python2 for build
--
wpt-commits: b923853fec706d8d448010caab33583026ae4749, 6539f1e3433e77e23a33e293007f2da993c96abc, cbe30a2fb78294ae530856d5174a47065b4f1958, 4e07dbd21b1ef0bf749bd38f3b8ea2e954abdf27, 96ea8bb0f93a29a09dfa181a65eebe871d5b6d95, da9f11f0610f1ab52f2d85f40a8b58ecd72571d7, 394d3fb5f647fa2416a95d105fb82f9269f73586, 40a49406d70c9d3bb73f38e2e0576efd301d75a4, 76a0eee0d2b697a4bcc599b4f7ad0522ade4d0e8, 6a531642611c81e83bed58421fdc555e2db1f77a
wpt-pr: 23733
Automatic update from web-platform-tests
[css-grid] Check that 'align-self: last baseline' does not interfere with baseline alignment in first row. https://github.com/w3c/csswg-drafts/issues/5293
--
wpt-commits: e3698c7bb3c309df69134e9bc0a375f00535e226
wpt-pr: 24477
Automatic update from web-platform-tests
[css-grid] Test autorepeat with max,min. https://github.com/w3c/csswg-drafts/issues/4043
--
[css-grid] Use background vs. background-color consistently. (editorial/lint)
--
wpt-commits: 9520e0643c5171ff97157469ec342636f0f1cf88, 7f77bdd0d1a3d7ae933026b74342ee0de1afed46
wpt-pr: 24470
Automatic update from web-platform-tests
Reland "Don't reset computed 'float' for flex and grid items"
This is a reland of 906abd41e4f6c786a1927f6edd69d6926e20317b
Original change :
https://chromium-review.googlesource.com/c/chromium/src/+/2157889
Revert of original change :
https://chromium-review.googlesource.com/c/chromium/src/+/2256772
-----
The original change was reverted due to a DCHECK being triggered. See
the following CRBug for more details:
https://bugs.chromium.org/p/chromium/issues/detail?id=1097595
The original change no longer adjusted the ComputedStyle for floats
based on whether or not it was a flex/grid item. This adjustment was
left for the LayoutBox to handle. However, there was code on the layout
side that depended on IsFloating() getting adjusted for the style, as
well. This resulted in the DCHEK mentioned above getting hit.
To fix this, I updated all cases that used the style's IsFloating()
function and either updated it to also take the style's
IsFlexOrGridItem() function into account or updated the code to use
LayoutObject's IsFloating() function instead.
-----
Original change's description:
> Don't reset computed 'float' for flex and grid items
>
> Floated flex and grid items have their 'float' incorrectly computed to
> "none". This change fixes this by adjusting IsFloating() for flex and
> grid items inside LayoutBox::UpdateFromStyle() instead of inside
> StyleAdjuster::AdjustStyleForDisplay().
>
> Beyond this, legend elements are not allowed to be rendered legends
> if they are floating. However, if legends are flex items, we
> adjust IsFloating() to be false. This causes legends to be
> used as rendered legends when they are flex items, even if they
> were styled as a float.
>
> This is fixed by checking ComputedStyle::IsFloating() instead
> of LayoutObject::IsFloating() when finding the rendered legend in
> order to get its non-adjusted floating value.
>
> Bug: 875235,350505
> Change-Id: Ia1e7f7c244cc0c443cd58be42854866884b7f7e7
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2157889
> Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
> Reviewed-by: Morten Stenshorne <mstensho@chromium.org>
> Reviewed-by: Christian Biesinger <cbiesinger@chromium.org>
> Commit-Queue: Alison Maher <almaher@microsoft.com>
> Cr-Commit-Position: refs/heads/master@{#763988}
Bug: 875235, 350505
Change-Id: I915e045db7fda94d99cfdae33911d8b5f6df26de
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2261412
Reviewed-by: Morten Stenshorne <mstensho@chromium.org>
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Alison Maher <almaher@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#785456}
--
wpt-commits: d33bc572d46eab7d3db2f67c24ab7e132344e22f
wpt-pr: 24316
Automatic update from web-platform-tests
Add another variant that explicitly uses minmax(0, 1fr), rather than relying on the auto size. (#24464)
--
wpt-commits: 1f53c57f1b474d7877184d45b32b583b88b5c6af
wpt-pr: 24464
Automatic update from web-platform-tests
[FragmentItem] Clear inline fragment for positioned objects
This is a refix of r784685 <crrev.com/c/2275820>.
When floating objects become positioned, |NGFragmentItems|
can no longer track further changes or destructions. r784685
<crrev.com/c/1101277> fixed this in pre-layout, but it is
possible that changes can occur before next layout cycle.
This patch marks |NGFragmentItem| as moved and clears the
index in |LayoutBox::StyleWillChange| instead.
Logically speaking, this could happen to non-floating objects
such as inline-block. Our current code reattaches when non-
floating objects become positioned, so only floating objects
can cause this after r784297 <crrev.com/c/2275373>.
Bug: 1101818, 1101986, 1101277
Change-Id: I56e53aaf4e0b6b5b0ef78c3f75eaf3ffd3568596
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2279425
Reviewed-by: Morten Stenshorne <mstensho@chromium.org>
Reviewed-by: Ian Kilpatrick <ikilpatrick@chromium.org>
Commit-Queue: Koji Ishii <kojii@chromium.org>
Cr-Commit-Position: refs/heads/master@{#785453}
--
wpt-commits: bc36542dc6e20d502f593295a989895e7cbc1960
wpt-pr: 24442
Automatic update from web-platform-tests
Unify most clip-path code paths of pre-CAP and CAP
Previously in pre-CAP, because we always created a composited mask
layer for any mask and/or clip-path (either path-based or mask-based)
if the object needed to be composited, we needed to handle clip-path
differently in pre-CAP and CAP,
Now always apply path-based clip-path after paint. No longer create
a mask layer for path-based clip-path in pre-CAP.
Rename ClipPath to ClipPathMask to make it clear that this effect node
is for the mask created for a mask-based clip-path.
If there are both mask and mask-based clip-path, always let ClipPathMask
effect node be the child of the Mask effect node. This is required by
pre-CAP (unchanged) to apply the clip-path in the mask layer, but also
works in CAP (previously both Mask ClipPathMask were directly under
Effect) because:
DstIn(DstIn(A, Mask), ClipPathMask)
== DstIn(A, DstIn(Mask, ClipPathMask)
where ClipPathMask is black-and-white.
If there is only a mask-based clip-path, we'll only create ClipPathMask
for both pre-CAP (now no Mask and MaskClip nodes) and CAP (unchanged).
The pre-CAP mask layer will use ClipPathMask as the effect state, which
avoids the need of filling a black mask before applying the clip path
mask.
Now the only difference between pre-CAP and CAP about clip-path is that
that when a LayerObject with mask-based clip-path is composited, pre-CAP
creates a GraphicsLayer for the clip-path mask before paint, while CAP
creates a cc::Layer for the mask after paint.
Bug: 1100224
Change-Id: I243e463b3c94e09b64ed3c171ece12145c9deec9
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2274177
Reviewed-by: Philip Rogers <pdr@chromium.org>
Commit-Queue: Xianzhu Wang <wangxianzhu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#785435}
--
wpt-commits: 3d92704499a679683fb03500ceeeaef6ad652676
wpt-pr: 24440
Automatic update from web-platform-tests
Duplicate -001, but with the item getting its size from a child rather than explicitly. (#24463)
* Duplicate -001, but with the item getting its size from a child rather than explicitly.
* Whoops, remove the <base> I was using for testing.
--
wpt-commits: e8401319ee32642fa1fd77b50f097ee9c5199487
wpt-pr: 24463
Automatic update from web-platform-tests
Add metadata to preserve-fragment.html WPT test
Adds a long timeout since the single file tests hundreds of combinations
which timesout when manually run on mobile devices and cuts it close on
some desktop browsers.
Also added help links that refer to relevant spec text as well as bugs
that motivated this test.
Bug: 1055455
Change-Id: I8923bdd8201818dddb066ccd6d40790b416db16c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2281223
Reviewed-by: Stephen McGruer <smcgruer@chromium.org>
Commit-Queue: David Bokan <bokan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#785433}
--
wpt-commits: 08c4b3cb1535fedc32a7ced2aef987d3d797b3fc
wpt-pr: 24439
Automatic update from web-platform-tests
Remove .tentative flag from Split Cache Web Platform Test
Now that the spec for HTTP cache partitioning is about to be landed,
this flag must be correspondingly removed.
Bug: 1064765
Change-Id: I4ee646b4e3a48c33ed6a025132d4e9a10940f675
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2238349
Commit-Queue: Shivani Sharma <shivanisha@chromium.org>
Reviewed-by: Shivani Sharma <shivanisha@chromium.org>
Cr-Commit-Position: refs/heads/master@{#785405}
--
wpt-commits: b785f83c799379ac64c8d0836e56caa350ff95f8
wpt-pr: 24070
Automatic update from web-platform-tests
Use the style being computed as rem base for root element.
If we are resolving style for the root element, return the style being
built as the root style. That style is not available at the point we are
initialising the root style in ElementResolveContext.
Bug: 918480
Change-Id: If04aa2db7de9f74b9089b29cfda2e20c43762d65
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2279802
Reviewed-by: Anders Hartvoll Ruud <andruud@chromium.org>
Commit-Queue: Rune Lillesveen <futhark@chromium.org>
Cr-Commit-Position: refs/heads/master@{#785359}
--
wpt-commits: c269b86af5e687cc6e9df00a96b64d9165c22ba0
wpt-pr: 24443
Automatic update from web-platform-tests
[FragmentItem] Limit reusing cached lines only when top-level child is clean
This patch simplifies |DirtyLinesFromNeedsLayout| to check
|NeedsLayout| of top-level children only.
The previous code tried to reuse more lines, e.g.:
<div><span>many lines of text</span></div>
Most lines are reusable when appending to the end of "text".
But supporting this case complicates the logic, especially
when culled inline is involved.
The new logic can't reuse lines in the above case, but still
can reuse liens if the `<span>` does not exist, and should
cover most common cases.
Bug: 1102083
Change-Id: I4f76e154b834c8c00e5ce04ab251c4f1fcdabab0
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2282821
Reviewed-by: Yoshifumi Inoue <yosin@chromium.org>
Commit-Queue: Yoshifumi Inoue <yosin@chromium.org>
Commit-Queue: Koji Ishii <kojii@chromium.org>
Cr-Commit-Position: refs/heads/master@{#785345}
--
wpt-commits: eae9c1b62560eedbccc8efa3a1f6b6ce1ba97b30
wpt-pr: 24453
Automatic update from web-platform-tests
[FragmentItem] Fix DirtyLinesFromNeedsLayout
This patch fixes a case where |DirtyLinesFromNeedsLayout|
fails to mark items dirty for reusing cached lines.
When searching for |LayoutObject| needing layout, checking
|NormalChildNeedsLayout| and |PosChildNeedsLayout| were not
enough. In some specific cases, only
|NeedsSimplifiedNormalFlowLayout| was set for the parent
inline box.
This patch changes the condition to check descendants to
|NeedsLayout|.
Bug: 1101883
Change-Id: Ia73ace058c9ea1ebaaffdaff793b56d68099cd5c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2281781
Reviewed-by: Yoshifumi Inoue <yosin@chromium.org>
Reviewed-by: Kent Tamura <tkent@chromium.org>
Commit-Queue: Koji Ishii <kojii@chromium.org>
Cr-Commit-Position: refs/heads/master@{#785287}
--
wpt-commits: aef9da08fe2fa0b9fa473e2e126cb3cf29fae43e
wpt-pr: 24444
Automatic update from web-platform-tests
[TablesNG] New test suite: colspan redistribution
Investigation into how cells with colspan > 1
redistribute their min/max/% widths.
Browsers often disagree, and spec allows for variations.
Bug: 958381
Change-Id: I588053d1392e00e9d199cdd1095b93a148101fc4
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2219093
Reviewed-by: David Grogan <dgrogan@chromium.org>
Commit-Queue: Aleks Totic <atotic@chromium.org>
Cr-Commit-Position: refs/heads/master@{#785260}
--
wpt-commits: 6f5097d7cb5bcedef39dcdd5331064eadfdd0283
wpt-pr: 23807
Automatic update from web-platform-tests
[COOP] Report-only navigation tests (#24379)
This adds basic tests of the report-only features for the navigation
case, where the report-only headers would cause a browsing context
group switch.
Bug: 1099208
Change-Id: Ia5261d5d1ddac4a83943e0a48b5ef5f2cdb47b7b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/2266000
Commit-Queue: Pâris Meuleman <pmeuleman@chromium.org>
Reviewed-by: Arthur Sonzogni <arthursonzogni@chromium.org>
Auto-Submit: Pâris Meuleman <pmeuleman@chromium.org>
Cr-Commit-Position: refs/heads/master@{#785021}
Co-authored-by: Pâris MEULEMAN <pmeuleman@chromium.org>
--
wpt-commits: b7938227eacdc8d6d475c5a7bff061724d6ebd3f
wpt-pr: 24379
moz-wptsync-bot
Stephen McGruer
Sean Feng
ziransun
Marijn Kruisselbrink
jgraham
fantasai
Tab Atkins Jr
Blink WPT Bot
Simon Pieters
autofoolip
Alison Maher
Koji Ishii
Xianzhu Wang
David Bokan
pyup.io bot
Alex Turner
Frédéric Wang
Rune Lillesveen
Aleks Totic