WebKit export of https://bugs.webkit.org/show_bug.cgi?id=252373

[sideshowbarker]

No description provided.

[sideshowbarker]

@nt1m well I only now notice that I’d left out the old/newState asserts from the details6 “Calling open twice on 'details' fires only one toggle event”

To fix that, what should I do first?

• First do a follow-up WebKit PR for Use ToggleEvent for <details> element toggle event WebKit/WebKit#16795 that adds the assertions to LayoutTests/imported/w3c/web-platform-tests in WebKit — and then (re)export them here, in a follow-up WPT PR? or
• First do a follow-up WPT PR that adds the assertions here, then (re)import then to WebKit, in a follow-up WebKit PR?

I guess the advantage of doing the WebKit PR first would be that the updated tests get run by the EWS bots, so we know that they’re passing in WebKit.

(I do already know that “Calling open twice on 'details' fires only one toggle event” is in fact passing with those asserts included — I think I’d already included those asserts in an interim commit in the WebKit PR branch, but then while I was troubleshooting locally to try to isolate what might be causing the flakiness for that particular test, I’d inadvertently ended up amending and force-pushing a local version in a state that omitted those assertions.)

[sideshowbarker]

@nt1m Also by the way, when I ran export-w3c-test-changes, it didn’t work completely as expected.

I ran it like this:

export-w3c-test-changes -c -g HEAD -b 252373 --no-linter -u git@github.com:web-platform-tests/wpt.git -r wpt

…and that created the PR here as expected — except that it didn’t cause wpt-pr-bot to add the expected “This patch has been exported from WebKit; it will be approved automatically once the downstream patch is r+” comment.

I guess maybe that part gets done correctly if export-w3c-test-changes detects that I’m a WebKit committer? Or maybe at least a WebKit contributor? (I’m not actually even in the WebKit contributors.json file yet.)

Or maybe it failed because I used --no-linter and that had the effect of suppressing some necessary step?

I had added --no-linter because when I tried without adding that, I got this error:

ERROR:lint:/opt/webkit/WebKitBuild/w3c-tests/web-platform-tests/html/semantics/interactive-elements/the-details-element/toggleEvent.html: /opt/webkit/WebKitBuild/w3c-tests/web-platform-tests/html/semantics/interactive-elements/the-details-element/toggleEvent.html matches an ignore filter in .gitignore - please add a .gitignore exception (IGNORED PATH)

…which as far as I could see was a spurious error — because that test doesn’t actually match anything in .gitignore, and I think we definitely don’t want to add an IGNORED PATH exception for it…

[nt1m]

To fix that, what should I do first?

Either works!

…which as far as I could see was a spurious error — because that test doesn’t actually match anything in .gitignore, and I think we definitely don’t want to add an IGNORED PATH exception for it…

Yeah, I usually use --no-linter to workaround this.