Bug 265414 – REGRESSION (270602@main): [ Monterey / Ventura wk2 ] imported/w3c/web-platform-tests/accname/name/comp_host_language_label.html is a constant failure

• Status: NEW
• Component: New Bugs

[Marta Darbinyan]

Description:
imported/w3c/web-platform-tests/accname/name/comp_host_language_label.html is constantly failing on macOS Ventura and Monterey wk2 since 270602@main was committed.


This issue can be bisected to 270602@main using the command: 
run-webkit-tests --root  --exit-after-n-failures 1 --debug imported/w3c/web-platform-tests/accname/name/comp_host_language_label.html

History:
https://results.webkit.org/?suite=layout-tests&test=imported%2Fw3c%2Fweb-platform-tests%2Faccname%2Fname%2Fcomp_host_language_label.html&version_name=Monterey&version_name=Ventura&flavor=wk2

Diff Log:
 PASS html: label input[type=checkbox] encapsulation
 PASS html: label input[type=color] encapsulation
 FAIL html: label input[type=date] encapsulation assert_equals: <input type="date" data-expectedlabel="date label" data-testname="html: label input[type=date] encapsulation" class="ex"> expected "date label" but got "month / day / year date label"
-FAIL html: label input[type=datetime-local] encapsulation assert_equals: <input type="datetime-local" data-expectedlabel="datetime-local label" data-testname="html: label input[type=datetime-local] encapsulation" class="ex"> expected "datetime-local label" but got "month / day / year , 12 : 30   PM datetime-local label"
+FAIL html: label input[type=datetime-local] encapsulation assert_equals: <input type="datetime-local" data-expectedlabel="datetime-local label" data-testname="html: label input[type=datetime-local] encapsulation" class="ex"> expected "datetime-local label" but got "month / day / year , 12 : 30 PM datetime-local label"
 PASS html: label input[type=email] encapsulation
 FAIL html: label input[type=month] encapsulation assert_equals: <input type="month" data-expectedlabel="month label" data-testname="html: label input[type=month] encapsulation" class="ex"> expected "month label" but got "month / year month label"
 PASS html: label input[type=number] encapsulation
@@ -95,7 +95,7 @@
 FAIL html: label input[type=search] encapsulation assert_equals: <input type="search" data-expectedlabel="search label" data-testname="html: label input[type=search] encapsulation" class="ex"> expected "search label" but got "cancel search label"
 PASS html: label input[type=tel] encapsulation
 PASS html: label[for] input[type=text] encapsulation
-FAIL html: label input[type=time] encapsulation assert_equals: <input type="time" data-expectedlabel="time label" data-testname="html: label input[type=time] encapsulation" class="ex"> expected "time label" but got "12 : 30   PM time label"
+FAIL html: label input[type=time] encapsulation assert_equals: <input type="time" data-expectedlabel="time label" data-testname="html: label input[type=time] encapsulation" class="ex"> expected "time label" but got "12 : 30 PM time label"


Link:
https://build.webkit.org/results/Apple-Monterey-Release-AppleSilicon-WK2-Tests/270602@main%20(12091)/imported/w3c/web-platform-tests/accname/name/comp_host_language_label-pretty-diff.html

[Radar WebKit Bug Importer]

<rdar://problem/118857276>

[EWS]

Test gardening commit 271180@main (a9bd8c3d8d2a): <https://commits.webkit.org/271180@main>

Reviewed commits have been landed. Closing PR #20966 and removing active labels.

[Alexey Proskuryakov]

This seems like a new test, not a regression.

[Ryan Haddad]

(In reply to Alexey Proskuryakov from comment #3)
> This seems like a new test, not a regression.

The test isn't new, but 270602@main rebaselined the test to pass on Sonoma at the expense of Ventura and Monterey.

Seems like the "fix" here is to add a baseline for this test in platform/mac-ventura so it will start passing for Ventura and Monterey again.

[EWS]

Test gardening commit 271685@main (9c3976743660): <https://commits.webkit.org/271685@main>

Reviewed commits have been landed. Closing PR #21453 and removing active labels.

[Jonathan Bedard]

I think we need to revisit this.

The test is "passing" on Ventura/Monterey WK2, but "failing" on Ventura/Monterey WK1. Wee could "fix" this with a rebaseline, but I don't think we expect this test to behave differently on WK1 and WK2, I think there may be a genuine bug here.