Add tests for Joint Iteration proposal #4528
Conversation
…roposal. r=dminor Test262 coverage: tc39/test262#4528 Differential Revision: https://phabricator.services.mozilla.com/D255924
…roposal. r=dminor Test262 coverage: tc39/test262#4528 Differential Revision: https://phabricator.services.mozilla.com/D255924
|
Tests have been validated with the polyfill in tc39/proposal-joint-iteration#44. I'll do a review for completeness soon and then ask for Stage 3. Thanks for implementing these, @anba. |
There was a problem hiding this comment.
I didn't do a real review, just skimmed most of it. As a reviewer, it'd be very helpful if you could cross-reference these with the testing plan. I hate asking for extra work, but anything that makes it easier for us to do the review in small chunks without having to swap in extra context is appreciated.
One test that occurred to me that I didn't see in this PR or the testing plan (although as I said, I only skimmed it so I may have missed it) is verifying the actual behaviour when mode is missing. In other words I see we test thoroughly that absent or undefined mode is handled, but not what it actually does.
There was a problem hiding this comment.
I would prefer we don't compute a large list of expected results inline in a test using what is essentially a mini-polyfill. When the test fails, it's unclear whether the implementation or the mini-polyfill is wrong.
There was some more discussion on this re. locale-sensitive testing in the talk I presented in the April TC39 meeting. slides, discussion
Same for zipKeyed/basic.js.
See #4112 (comment) for the testing plan.