Automatic update from web-platform-tests
Implement a decision task for taskcluster jobs
A decision task is the first task run, and ensures that we only
schedule subsequent tasks which are relevant to the current push or
pull request. This dynamic scheduling helps reduce load since we avoid
spinning up workers for tasks that ultimately don't run, and unlocks
new possibilities since we are able to schedule tasks that are
dependent on other tasks.
The tasks, their scheduling criteria and their dependnecies are
specified in a a YAML format configuration file in
tools/ci/tc/tasks/test.yml. This has a bespoke format, adopting some
ideas from Azure and Gecko's taskcluster integration. The format is
documented in `tools/ci/tc/README.md`. The data in that file undergoes
trandformations to produce a set of tasks, which are then filtered
according to the event that caused the decision task to run.
To initially prove out the implementation of dependent tasks we make
the Firefox tasks depend on a download task. But this does not yet
include the work to actually make the dependent tasks download Firefox
from the parent task.
--
wpt-commits: f57ead1924c978dc9256d8f572614731a1a51678
wpt-pr: 19401
--HG--
rename : testing/web-platform/tests/tools/taskcluster/testdata/master_push_event.json => testing/web-platform/tests/tools/ci/tc/testdata/master_push_event.json
rename : testing/web-platform/tests/tools/taskcluster/testdata/pr_event.json => testing/web-platform/tests/tools/ci/tc/testdata/pr_event.json
Automatic update from web-platform-tests
[css-sizing-3] Add tests for percentage resolution (#5280)
This just tests that percentages don't resolve against min-content,
which should be indefinite.
--
wpt-commits: 86097c04c60a22d7eecb6ea524e898f0c3cb90ed
wpt-pr: 5280
Automatic update from web-platform-tests
Fix Chromium logger to properly handle expected test status
When `test_end` passes through mozlog, the `expected` status will be
removed if it's the same as the `actual` status.
This was already fixed for `test_status` calls, but it also affects
`test_end`. This PR refactors the expected status extraction into a
common place and adds more unit tests.
--
wpt-commits: b2acffa9b2460e810be492fd115437656e5e289b
wpt-pr: 20363
Automatic update from web-platform-tests
[WebNFC] Add permission prompt for NDEFWriter.push
This CL makes sure NDEFWriter.push shows a permission request if user
didn't previously allow Web NFC. Follow-up CL will do something
similar for NDEFReader.
Screenshots: https://photos.app.goo.gl/k7wz55v9ELVTS19i8
Bug: 520391
Change-Id: I4a1ac1edcd39d5d8a27f891a37044de8e9d0587a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1895345
Commit-Queue: François Beaufort <beaufort.francois@gmail.com>
Reviewed-by: Rijubrata Bhaumik <rijubrata.bhaumik@intel.com>
Reviewed-by: Balazs Engedy <engedy@chromium.org>
Reviewed-by: Peter Conn <peconn@chromium.org>
Reviewed-by: Charlie Harrison <csharrison@chromium.org>
Reviewed-by: Andrew Grieve <agrieve@chromium.org>
Cr-Commit-Position: refs/heads/master@{#718123}
--
wpt-commits: 50fa7d656b2848dedb9821d3f84386d9b9501634
wpt-pr: 20125
Automatic update from web-platform-tests
[css-fonts] Add tests for standard font families (#19917)
e083fe61cd
--
wpt-commits: 21f50fc43b7e291163ba513c59714ad385f7a2e1
wpt-pr: 19917
Automatic update from web-platform-tests
Revert "[sms] Implement AbortController for SMSReceiver API"
This reverts commit 331d40099623dd97e57ddd3361decbfdd9a0e8b1.
Reason for revert: Seems to fail leak tests (see crbug.com/1027386)
Original change's description:
> [sms] Implement AbortController for SMSReceiver API
>
> This change allows developers to use the SMS Receiver API with the Abort
> Controller to cancel once the API is called. This will help developers with
> flows for retrying and sending a new code and they want to abort the previous
> call they made to the API. If the InfoBar is up during abort, it will
> leave the InfoBar open while returning an AbortError to the website. If
> another request is made while the InfoBar for 1st request is still open (after
> abort), upon clicking 'Verify' the SMS retrieved for the 1st request will
> return to the 2nd request.
>
> Bug: 976401
> Change-Id: I132800f2954fdc370172dcd03789d3f08bcc4040
> Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866914
> Reviewed-by: Daniel Cheng <dcheng@chromium.org>
> Reviewed-by: Reilly Grant <reillyg@chromium.org>
> Commit-Queue: Ayu Ishii <ayui@chromium.org>
> Cr-Commit-Position: refs/heads/master@{#717710}
TBR=dcheng@chromium.org,reillyg@chromium.org,goto@chromium.org,ayui@chromium.org
Change-Id: I21501bd728a8b3679b5b594751ee0ea20a1d8e83
No-Presubmit: true
No-Tree-Checks: true
No-Try: true
Bug: 976401
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1930275
Reviewed-by: Darren Shen <shend@chromium.org>
Commit-Queue: Darren Shen <shend@chromium.org>
Cr-Commit-Position: refs/heads/master@{#718000}
--
wpt-commits: 0d75262db8c6973b697a1d930dbf234e1324901f
wpt-pr: 20384
Automatic update from web-platform-tests
[IntersectionObserver] Check containing block chain for stop node
When doing a hit test to measure occlusion, the code previously
assumed that the hit test logic would always prevent descending into
the subtree of the stop node. That's not the case when parent and
child positioned objects are in the same stacking context and the
parent positioned object is the stop node. In that case the subtree
of the child positioned object will be traversed before the parent
positioned object.
When a hit is identified, we need to explicitly check whether it's a
containing block descendant of the stop node; and if so, reject the
hit.
NOTE: this is a clone of
https://chromium-review.googlesource.com/c/chromium/src/+/1914537
that I'm committing on behalf of szager@
BUG=1020784
Change-Id: Ic9270d7b6491be1b3364131fb20cd120fbb6d272
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1929643
Reviewed-by: Stefan Zager <szager@chromium.org>
Commit-Queue: Chris Harrelson <chrishtr@chromium.org>
Cr-Commit-Position: refs/heads/master@{#717879}
--
wpt-commits: dff3a6865c67b469be3b9bc1f4b318554008931b
wpt-pr: 20374
Automatic update from web-platform-tests
Rewrite WPT WebAudio tests to work on Chrome
Chrome times out on these tests because the test throws an exception
in a promise resolver. Rewrite the tests to use the promise_test
method which handles these.
Manually verified that Firefox still passes these tests, and that
Chrome fails them instead of timing out. TestExpectations updated
appropriately.
Bug: 1011080, 1021295
Change-Id: Ib65dec0c49ad6c9764b4432c1e28282f8a7d268a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1898606
Commit-Queue: Raymond Toy <rtoy@chromium.org>
Reviewed-by: Hongchan Choi <hongchan@chromium.org>
Cr-Commit-Position: refs/heads/master@{#717833}
--
wpt-commits: 6d736a0960660daf86c19c4ee92f876ce1503db1
wpt-pr: 20078
Automatic update from web-platform-tests
Null-dereference Read in blink::TransformOperations::ApplyRemaining fix
null_ptr get inserted into the Interpolatedtransform as from_ops via Blend function.
This would cause the functions that call the Apply in Interpolatedtransform class to crash.
Added the test case for testing Interpolatedtransform Blend.
Bug: 998938
Change-Id: I34a2e44d0fadf1c0ea848167714adbd0aa656910
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1887235
Reviewed-by: Ian Vollick <vollick@chromium.org>
Reviewed-by: Kevin Ellis <kevers@chromium.org>
Commit-Queue: Hao Sheng <haozhes@chromium.org>
Cr-Commit-Position: refs/heads/master@{#717829}
--
wpt-commits: 74dbbd220fc9a072c3c14d8ceacda9af11046fb9
wpt-pr: 20183
Automatic update from web-platform-tests
[css-grid-1] Add test for exclusion of cyclic-sizing-dependency items from baseline alignment.
--
Update grid-baseline-align-cycles-001.html
--
Update grid-baseline-align-cycles-001-ref.html
--
Update grid-baseline-align-cycles-001.html
--
Applied some suggested improvements in grid-baseline-align-cycles-001.html test
--
Fixing lint errors in grid-baseline-align-cycles-001.html test
--
Separate percent and orthogonal items in grid-baseline-align-cycles-001.html
--
wpt-commits: c8a55dc00cd0fd76ac57dab436222a06be67b9a1, dcdb03d2abe564c782246f0c45995106f2a9d746, b475390b12d218948eb67249e15ee45db4d63608, 54a856308626a481a6439178524e4e20a6ef31c3, bcb0e1b057d53b15a2979bc0b1d90424dad53a82, a9c9d23d6226aa3620b7c165772116c9d4445695, e743a0e20c64ca187c0c35539ff0748120125071
wpt-pr: 8580
Automatic update from web-platform-tests
[sms] Implement AbortController for SMSReceiver API
This change allows developers to use the SMS Receiver API with the Abort
Controller to cancel once the API is called. This will help developers with
flows for retrying and sending a new code and they want to abort the previous
call they made to the API. If the InfoBar is up during abort, it will
leave the InfoBar open while returning an AbortError to the website. If
another request is made while the InfoBar for 1st request is still open (after
abort), upon clicking 'Verify' the SMS retrieved for the 1st request will
return to the 2nd request.
Bug: 976401
Change-Id: I132800f2954fdc370172dcd03789d3f08bcc4040
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1866914
Reviewed-by: Daniel Cheng <dcheng@chromium.org>
Reviewed-by: Reilly Grant <reillyg@chromium.org>
Commit-Queue: Ayu Ishii <ayui@chromium.org>
Cr-Commit-Position: refs/heads/master@{#717710}
--
wpt-commits: 82ea18f340f977bc8f42e503043f4eb054a6d2e7
wpt-pr: 19853
Automatic update from web-platform-tests
Cookie Store: Move some WPT tests out of resources/.
This conversion is a step towards unifying the structure of the Cookie Store API
WPT tests.
Change-Id: Ib6c8f93d407c7f749652e699cd9c336ac5cb85ac
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1928233
Commit-Queue: Victor Costan <pwnall@chromium.org>
Auto-Submit: Staphany Park <staphany@chromium.org>
Reviewed-by: Victor Costan <pwnall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#717742}
--
wpt-commits: 3bc65de31115836422955f5748b604c9fa09241d
wpt-pr: 20370
Automatic update from web-platform-tests
Follow NDEFScanOptions dictionary spec
NDEFScanOptions dictionary spec was changed.
So we should follow new spec for it.
I updated the dictionary.
Bug: 520391,1001507
Change-Id: I427b572e597bd55c719c18922b92c6419b473a22
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1921930
Reviewed-by: Kinuko Yasuda <kinuko@chromium.org>
Reviewed-by: Reilly Grant <reillyg@chromium.org>
Reviewed-by: Rijubrata Bhaumik <rijubrata.bhaumik@intel.com>
Commit-Queue: Byoungkwon Ko <codeimpl@gmail.com>
Cr-Commit-Position: refs/heads/master@{#717623}
--
wpt-commits: 8e98be7324cb0d839041cca41e9a22869c1e5347
wpt-pr: 20308
Automatic update from web-platform-tests
Use spread syntax in WebIDL/current-realm.html. (#20358)
In [1], it is reported that Chrome is failing the test calling
`CanvasRenderingContext2D.createImageData()` with
FAIL message: Failed to execute 'createImageData' on
'CanvasRenderingContext2D': The provided value is not of type
'(Uint8ClampedArray or Uint16Array or Float32Array)'
which happens because Chrome implements [2], which introduces a 4-argument
`createImageData()` overload. This overload ends up being used by the test
due to the following call:
CanvasRenderingContext2D.createImageData(5, 5, undefined, undefined)
Avoid passing a constant number of arguments when we are testing multiple
calls that take a variable number of arguments by using the spread syntax
instead.
[1] https://wpt.fyi/results/WebIDL/current-realm.html?run_id=353010001&run_id=376180001&run_id=381940001&run_id=376170001
[2] https://github.com/WICG/canvas-color-space/
--
wpt-commits: 6ac023bb7f669cb9be340e1c184dfd4d4226e661
wpt-pr: 20358
Automatic update from web-platform-tests
Add web-nfc test for ArrayBufferView of external type (#20355)
* Add web-nfc test for ArrayBufferView of external type
* Addressed @leonhsl's comments.
--
wpt-commits: e0361b12a78dade131348b19d1dff57f4aad19fb
wpt-pr: 20355
Automatic update from web-platform-tests
Restrict [un]registerProtocolHandler to secure contexts
The registerProtocolHandler() and unregisterProtocolHandler() functions
should only be permitted for secure contexts. When called on non-secure
contexts, a SecurityError should be thrown.
Bug: 882284
Change-Id: Iacf3d31f80f5118e9e9aacad2c99a0111d6e7cc5
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1892213
Reviewed-by: Gyuyoung Kim <gyuyoung@igalia.com>
Reviewed-by: Tarun Bansal <tbansal@chromium.org>
Reviewed-by: Kentaro Hara <haraken@chromium.org>
Reviewed-by: Mike West <mkwst@chromium.org>
Commit-Queue: Eric Lawrence [MSFT] <ericlaw@microsoft.com>
Cr-Commit-Position: refs/heads/master@{#717501}
--
wpt-commits: 6ac865ab791bb845839701ac1c99c2ed5ab10c4e
wpt-pr: 20073
Automatic update from web-platform-tests
[webnfc] Fix code trying to access a null pointer
This CL fixes a bug introduced by crrev.com/c/1824395 that attempts to
access null pointer, also adds a wpt to cover it.
BUG=520391,1026541
Change-Id: I62d5b28c3c6d39cbb27c2f21ca4d5ca91e54d892
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1925494
Commit-Queue: Leon Han <leon.han@intel.com>
Reviewed-by: Rijubrata Bhaumik <rijubrata.bhaumik@intel.com>
Reviewed-by: Reilly Grant <reillyg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#717487}
--
wpt-commits: 5ee699a2d83f07e88b0dbfd7938711bab1fe6cfc
wpt-pr: 20332
Automatic update from web-platform-tests
Cookie Store: Convert some WPT tests from .window to .any.
This conversion is a step towards unifying the structure of the Cookie Store API
WPT tests.
Change-Id: Ib5ddffd79471c4eecacc85be8b172db0d70911fb
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1927579
Commit-Queue: Victor Costan <pwnall@chromium.org>
Auto-Submit: Staphany Park <staphany@chromium.org>
Reviewed-by: Victor Costan <pwnall@chromium.org>
Cr-Commit-Position: refs/heads/master@{#717402}
--
wpt-commits: 9daf4f72027ba9fc5be8b698980e608166b90420
wpt-pr: 20353
--HG--
rename : testing/web-platform/tests/cookie-store/cookieStore_event_arguments.tenative.https.window.js => testing/web-platform/tests/cookie-store/cookieStore_event_arguments.tentative.https.window.js
Automatic update from web-platform-tests
Indexed DB: Verify array proxies are not valid keys (#20352)
Arrays of keys are valid keys, but implementations don't allow proxies of arrays to be keys. Add a test, corresponding to pending spec PR:
Bug: w3c/IndexedDB#309
PR: w3c/IndexedDB#311
--
wpt-commits: c65519066cf95e902607cce03a9057b859a50242
wpt-pr: 20352
Automatic update from web-platform-tests
Fix paint worklet hidpi
Right now in the PaintRenderingContext2D::setTransform, we account for
the device_scale_factor_. This is wrong. It appears that internally
the recording canvas handles the device scale factor in the correct way.
Instead it should account for browser zoom.
A layout test has been changed to expect the correct result.
Screen recordings here to show that the demo is rendered correctly:
https://drive.google.com/drive/folders/1NbHlFgxBvJDvtN3wrq61uyzzIPk83FXR?usp=sharing
Bug: 970783
Change-Id: I8f2b6287fe5b0c443d7de5c8c61bfb62ae1ef42b
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1841530
Commit-Queue: Xida Chen <xidachen@chromium.org>
Reviewed-by: Yi Gu <yigu@chromium.org>
Cr-Commit-Position: refs/heads/master@{#717322}
--
wpt-commits: 1a0e1500a99c6e5742246aa75e45e54a7dab0a51
wpt-pr: 20345
Automatic update from web-platform-tests
[ci] Remove malfunctioning GitHub Workflow (#20351)
This GitHub Workflow creates GitHub Deployments as expected, but those
Deployments no longer trigger further GitHub Workflows. This is critical
behavior for the Pull Request preview solution; without it, the GitHub
UI presents a misleading state (claiming the Deployments are "pending,"
when in reality, they will never complete).
Remove the Workflow file in order to disable the automation until the
original behavior is restored or the current behavior is confirmed to be
intentional:
https://github.community/t5/GitHub-Actions/Github-Actions-workflow-on-deployment/m-p/39010#M3623
--
wpt-commits: 462f3c4680dc9d4c5dd4740742947cd16d6fc1df
wpt-pr: 20351
Automatic update from web-platform-tests
[ci] Change name of environment variable (#20348)
The GitHub Workflow documentation warns against environment variable
names that begin with the sequence `GITHUB_`:
> **Note:** GitHub reserves the `GITHUB_` environment variable prefix
> for internal use by GitHub. Setting an environment variable or secret
> with the `GITHUB_` prefix will result in an error.
Source: https://help.github.com/en/actions/automating-your-workflow-with-github-actions/using-environment-variables#naming-conventions-for-environment-variables
Although early experiments have shown that it was recently possible to
override the `GITHUB_TOKEN` environment variable, the latest behavior of
the modified GitHub Workflows suggests that the overridden value is
being silently rejected. (Additionally, the disparity between GitHub
Secret name and environment variable name is somewhat confusing.)
Update the configuration to define an environment variable with an
allowed name--one that matches the name of the corresponding GitHub
Secret. Update the relevant scripts to reference this new variable.
Change the access pattern to a mechanism which will trigger a runtime
exception when the variable is not defined.
--
wpt-commits: 3f9fbdbefd8db5fdacd930d31e1402be61438b17
wpt-pr: 20348
Automatic update from web-platform-tests
[scroll-snap] Add keyboard test to WPT
Add a basic keyboard snapping to WPT. This is based on the blink internal test:
fast/scroll-snap/snaps-after-keyboard-scrolling.html
Changes from original test:
- The new tests uses test_driver to send key events.
- Add common style and script that I expect to be used in upcoming tests.
- Add clean up logic with add_cleanup to ensure they run even if a test fail.
# Keeping Original Test
This CL does not remove the original test. This is because current CQ wpt runner
does not support testdriver.js fully so the new test fails to run in CQ/trybots.
Also the original test uses gpuBenchmarking which injects the input at the
browser process while current wpt runner input injection uses devtool input
protocol which injects directly in renderer thus it covers a slightly different
code path.
There is ongoing work to support testdriver.js in CQ, and have it inject input
from browser process. Until those changes land keeping both version of the
tests makes sense.
BUG: 1011066
Change-Id: I83db1708277d2d469c940a32ea549edec8c62d23
Fixed: 1011066
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1893142
Reviewed-by: Yi Gu <yigu@chromium.org>
Commit-Queue: Majid Valipour <majidvp@chromium.org>
Cr-Commit-Position: refs/heads/master@{#716192}
--
Add to whitelist
--
wpt-commits: 71ad888773feb302767ccae3bd20bdbe758e5e1d, 5d71a0085efb9464d714dea6198db6273bc31b1c
wpt-pr: 20027
Automatic update from web-platform-tests
[infra] Integrate with external PR preview system (#19838)
Introduce a GitHub Action to monitor Pull Requests, storing relevant
information in the project's git repository (thus allowing the external
wptpr.live system to publish previews) and creating GitHub Deployments
(thus alerting contributors to the status of the preview). This Action
is triggered on a regular interval.
Introduce a second GitHub Action to monitor the state of the preview
system and communicate the relevant status to contributors via the Pull
Request UI. This Action is triggered for every GitHub Deployment created
in the previously-described Action.
For example, if three Pull Requests are updated, the first GitHub Action
will inspect them all. It will create GitHub Deployments only for the
"trusted" Pull Requests. A new GitHub Action will run for each of the
Deployments, polling the preview website until either the preview is
available or a timeout is reached. This Action will update the
deployment accordingly so that the author of each Pull Request author is
aware of the status of the preview site.
The following flow chart visually describes the same sequence:
sync
gh-101 (trusted) ---> |
gh-102 (untrusted) -> |
gh-103 (trusted) ---> |
.----.
|sync|--------+---------------------.
'----' | |
.-------------. .-------------.
|deploy gh-101| |deploy gh-103|
'-------------' '-------------'
| |
poll for preview poll for preview
| |
gh-101 <------ success ----- preview available poll for preview
|
gh-103 <------- error --------------------------------- timeout
--
wpt-commits: 990bd8b2b89c5d6a1e58db3eb903d489c2f36f39
wpt-pr: 19838
Automatic update from web-platform-tests
[Azure Pipelines] drop references to 'Hosted Windows Client Next' (#20334)
We don't use this anymore.
--
wpt-commits: 41c8e340efa48f39526da219d4fedebe1ec5ff1f
wpt-pr: 20334
Automatic update from web-platform-tests
reference all css grid tests to css/support/grid.css and eliminate local duplicates of grid.css (#20337)
--
wpt-commits: 4e30ebf34fd076f7b44ef4bfa6165a61121270eb
wpt-pr: 20337
Automatic update from web-platform-tests
Reassign snap areas when adding snap container
Making a container scrollable should make it a valid snap container.
In doing so, any descendant snap area that should be assigned to the
container should be reassigned as such.
Bug: 1007456
Change-Id: Iccd220311ce592f921dc8424780632e499d3201c
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1900920
Reviewed-by: David Bokan <bokan@chromium.org>
Reviewed-by: Majid Valipour <majidvp@chromium.org>
Reviewed-by: Yi Gu <yigu@chromium.org>
Commit-Queue: Kaan Alsan <alsan@google.com>
Cr-Commit-Position: refs/heads/master@{#717011}
--
wpt-commits: 5b9f9a0d1be1e14cc222cfbc6069fc70cd5d2f8e
wpt-pr: 20151
Automatic update from web-platform-tests
[Background Sync] Update WPT tests.
This CL:
1. Factors out Periodic Background Sync WPT tests in a separate directory
2. Adds a separate idl harness test for Periodic Background Sync,
3. Fixes a bug in periodicsync.https.window.js, in the setting of the scope.
Bug: 996664
Change-Id: I8653e8030880aa893d5d01e053b248828f88bbfc
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1924196
Reviewed-by: Mike West <mkwst@chromium.org>
Reviewed-by: Rayan Kanso <rayankans@chromium.org>
Commit-Queue: Mugdha Lakhani <nator@chromium.org>
Auto-Submit: Mugdha Lakhani <nator@chromium.org>
Cr-Commit-Position: refs/heads/master@{#717007}
--
wpt-commits: c42db5468b46bbf47b52a74fbcc9073980a39560
wpt-pr: 20319
Automatic update from web-platform-tests
Media Capabilities: make framerate a double instead of a DOMString.
Spec change: https://github.com/w3c/media-capabilities/pull/128
Bug: 994017
Change-Id: Iab036264fe19a6676c97bb12648321408d91f283
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1755046
Commit-Queue: Mounir Lamouri <mlamouri@chromium.org>
Reviewed-by: Steven Holte <holte@chromium.org>
Reviewed-by: Chrome Cunningham <chcunningham@chromium.org>
Cr-Commit-Position: refs/heads/master@{#716773}
--
set timeout to long
--
wpt-commits: 36691d0c57c84b80236ed53105ea15afaa7d88aa, 21423802c95e1b800b07e71c6f19a923f7d16616
wpt-pr: 18440
Automatic update from web-platform-tests
Chromedriver] Fix PerformAction to support multi touch points cases (#20276)
* Chromedriver] Fix PerformAction to support multi touch points cases
For the multiple touch points case, if we have one point move and the
other point stay still, such as
point 1: pointerdown, pause, pointerup
point 2: pointerdown, pointermove, pointerup
We should send five touch point events of "pointerdown", "pointerdown",
"pointermove", "pointerup", "pointerup" with pointer id 1, 2, 2, 1, 2.
Bug: 1020674
Change-Id: I7bdb49b7b37d4f83ef976750e106cbd27afc9e74
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1900632
Reviewed-by: John Chen <johnchen@chromium.org>
Commit-Queue: Lan Wei <lanwei@chromium.org>
Cr-Commit-Position: refs/heads/master@{#715739}
* add expectation data
* rename to ini
--
wpt-commits: df29f11cf0686caced7f382bcaff059c4dbfc070
wpt-pr: 20276
Automatic update from web-platform-tests
service worker: Use the correct spelling for referrer in a wpt test (#20341)
--
wpt-commits: 58ecdcd7891483de0942c4e017d4e9bf202d78c7
wpt-pr: 20341
Automatic update from web-platform-tests
bindings: Do not access @@iterator twice when converting union to sequence
A rough approximation of https://heycam.github.io/webidl/#es-union when
trying to convert an ES value to an IDL sequence or frozen array is:
1. Call the GetMethod abstract ES operation to retrieve the @@iterator
property.
2. If it is not undefined, invoke
https://heycam.github.io/webidl/#create-sequence-from-iterable with the
ES value and the method obtained above.
So far, we were implementing it roughly like this:
1. Call the GetMethod abstract ES operation to retrieve the @@iterator
property.
2. Invoke https://heycam.github.io/webidl/#es-sequence with the ES value,
which disregards the method obtained in step 1 and goes through the whole
process of calling GetMethod again and using that to create a sequence
from an interable.
This is obviously not compliant with the spec, and the multiple calls to the
GetMethod abstract operation are user-visible, as can be seen in the test
case attached to the bug 1024388.
Properly fixing this requires changes to a few different classes and
functions:
* Make HasCallableIteratorSymbol(), which is called by the overload
resolution algorithm implementation, be a small wrapper around
GetEsIteratorMethod(). While HasCallableIteratorSymbol() precedes
GetEsIteratorMethod(), the latter's implementation is more spec-compliant.
* Add a move constructor to ScriptIterator, as well as a static method that
can create a ScriptIterator out of an ES value by invoking
GetEsIteratorMethod() and GetEsIteratorWithMethod() underneath.
* Make NativeValueTraits<IDLSequence<T>>::ConvertSequenceSlow() take a
ScriptIterator rather than a v8::Local<v8::Object>, which allows us to
delegate getting an iterator out of an iterable object to the static
ScriptIterator method mentioned above. This also helps reduce the size of
the generated per-type NativeValueTraits<IDLSequence<T>> template code.
* Analogously, invoke ScriptIterator::FromIterable() in the union conversion
code, and invoke a new NativeValueTraits<IDLSequence<T>>::NativeValue()
overload that takes a ScriptIterator, so that we do not access the
@@iterator property multiple times in the union conversion code.
Bug: 1024388
Change-Id: I47e6c0ca881e6e77f883ee78b5e1611138cdcc4a
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1917169
Commit-Queue: Raphael Kubo da Costa <raphael.kubo.da.costa@intel.com>
Reviewed-by: Yuki Shiino <yukishiino@chromium.org>
Cr-Commit-Position: refs/heads/master@{#716960}
--
wpt-commits: 6261dbb7e6c573eaac487db671d8ee8dc76f0dea
wpt-pr: 20318
Automatic update from web-platform-tests
[webnfc] Make NDEFReader#scan() return a Promise
Previously NDEFReader#scan() returns void and an NDEFErrorEvent will be
dispatched in case that the scan operation cannot be started
successfully.
This is not a well accepted pattern, and also to align with
NDEFWriter#push(), this CL makes NDEFReader#scan() return a Promise
instead.
Note that now NDEFErrorEvent is only used to notify Mojo disconnection,
a follow-up CL will remove it completely by using ErrorEvent instead.
The spec change:
https://github.com/w3c/web-nfc/pull/398https://github.com/w3c/web-nfc/pull/432
BUG=520391
Change-Id: I1477258ab70f7e40da31ea8795d63125b6a13af0
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1899586
Commit-Queue: Leon Han <leon.han@intel.com>
Reviewed-by: Reilly Grant <reillyg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#716904}
--
wpt-commits: 581a056bf7a743722a2e83cb80e1a8ce250e4f9d
wpt-pr: 20193
Automatic update from web-platform-tests
[webnfc] Use one nfc mock object per test file
Previously, each run of nfc_test() creates a new nfc mock object in the
start and destroyes it in the end.
Considering a test file containing multiple nfc_test() runs, in Blink we
have just one NFCProxy instance per page context to talk to these mocks,
that means our NFCProxy gets a broken NFC Mojo connection each time the
previous nfc_test() runs to the end then has to reconnect to another new
nfc mock for the coming nfc_test().
This is unnecessary and does not align with the product code logic that
always keeps one NFC Mojo connection alive per page context, unless
something unexpected happens in Device Service.
Also, this makes it easier for a nfc_test() to interfere with the next
one, e.g. a nfc_test() ends and closes the NFC Mojo connection, the
disconnection message will arrive at our NFCProxy asynchronously and we
can not have any assumptions on the arrival time, assuming it's still on
the way and the next nfc_test() has already started, at that point our
NFCProxy has no idea that the current Mojo connection was already closed
but continues to talk to it.
This CL makes one nfc mock object be reused throughout one test file to
address the above concerns.
BUG=1019663
Change-Id: I7e2ae9816936081e1f59688652e34b35a5295af6
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1913588
Commit-Queue: Leon Han <leon.han@intel.com>
Reviewed-by: François Beaufort <beaufort.francois@gmail.com>
Reviewed-by: Reilly Grant <reillyg@chromium.org>
Cr-Commit-Position: refs/heads/master@{#716874}
--
wpt-commits: 5980fe965255f86a1ee2fa698e06047576840640
wpt-pr: 20227
Automatic update from web-platform-tests
Add an extra test to check that inert propagates into Shadow DOM as well as out via slots
Bug: 692360
Change-Id: I8e24c36fbc6b476cfb1f08c0c07b890487995a39
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1918793
Commit-Queue: Alice Boxhall <aboxhall@chromium.org>
Reviewed-by: Meredith Lane <meredithl@chromium.org>
Cr-Commit-Position: refs/heads/master@{#716831}
--
wpt-commits: 3753e6678af4e2440b625cd104f6330c1e93a16d
wpt-pr: 20267
Automatic update from web-platform-tests
Plumb outside settings object for ServiceWorkerRegistration#update()
This is similar to crrev.com/c/1861294 but for update(). Script
fetches for update check happen on the browser and are managed
by ServiceWorkerUpdateChecker so pass some information to
ServiceWorkerUpdateChecker too.
This CL has a behavioral change. The 'Referer' header of the request
triggered by update() in a document/service worker will be changed.
Before this CL it was the url of the top-level script but after this
CL it is document/service worker's url. This behavior is consistent
with the spec.
Bug: 937177
Change-Id: I57e9ce7c91a748ad549a49e4e46c7890dcca2903
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1905028
Commit-Queue: Kenichi Ishibashi <bashi@chromium.org>
Reviewed-by: Hiroki Nakagawa <nhiroki@chromium.org>
Reviewed-by: Kinuko Yasuda <kinuko@chromium.org>
Cr-Commit-Position: refs/heads/master@{#716516}
--
Mark referer-toplevel-script-fetch.https.html as slow
--
wpt-commits: b9fcfe91bc700f21ef08b44fb9dac562eb68495f, 295f22bfeb7e6db748e6b0e28c31adfeb498b55d
wpt-pr: 20170
Automatic update from web-platform-tests
[ResourceTiming] Add test for new TAO check (#20320)
This CL modifies the tests as follows:
* '*' is now the default TAO header. This means most sandwich tests use '*'.
* Another resource is added to crossorigin-sandwich-TAO to test the case in
which the header is the page origin, and this should now fail (this was the
test before this CL).
* The resources in crossorigin-sandwich-TAO are changed to images because it's
not super clear how iframes will be handled.
Bug: 1022816
Change-Id: I2110653a5240cff825d3130cfe6dc64e8d2f8e23
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1924635
Commit-Queue: Nicolás Peña Moreno <npm@chromium.org>
Reviewed-by: Yoav Weiss <yoavweiss@chromium.org>
Cr-Commit-Position: refs/heads/master@{#716652}
--
wpt-commits: cd51095ccff19a22f742a4081db89fd015db5d7b
wpt-pr: 20320
Automatic update from web-platform-tests
Use existing external resource for <use> while revalidating
When we load a previously cached resource for <use> and it requires
revalidation - and an earlier <use> that references the same resource
hasn't yet generated its shadow tree - we could end up failing to
generate a shadow tree for the the "first" <use> since the resource
appears to be "invalid" while the revalidation is in progress. This
"first" <use> would then not be invalidated when the revalidation
completes.
Consider the resource to be valid while revalidating if it has a
Document already - meaning it has previously finished loading and
successfully parsed a document. This means the "first" <use> might end
up being stale until it's invalidated if the revalidation returned a
newer version of the resource - which is wrong but arguably better than
failing to display at all like was the case previously. (This should
also be consistent with how other types of resources are handled in
cases like this.)
Bug: 593261, 1021530
Change-Id: I9d68a1a60685757a990c2ec811c5ca53afc4fb25
Reviewed-on: https://chromium-review.googlesource.com/c/chromium/src/+/1903200
Commit-Queue: Fredrik Söderquist <fs@opera.com>
Reviewed-by: Nate Chapin <japhet@chromium.org>
Cr-Commit-Position: refs/heads/master@{#716678}
--
wpt-commits: b281a06110c196a820b789201ab8793bbef0219a
wpt-pr: 20150
Automatic update from web-platform-tests
Removed unused `platform_ini` member in browser.py
--
wpt-commits: a93b1f7d58ed6456636ab3af5ef178f2d5d51893
wpt-pr: 20316