GAUD-9420: delay event to try and fix flaky test

[dlockhart]

Looks like the original fix in #6559 didn't do the trick, so this is another attempt.

My new theory is that because dispatching events is synchronous, that the test doing a clickElem and then await oneEvent one line after the other is missing the re-firing of the new event. So just like we do in many other places, this wraps the dispatching of the new event in a timeout.

[github-actions]

Thanks for the PR! 🎉

We've deployed an automatic preview for this PR - you can see your changes here:

URL	https://live.d2l.dev/prs/BrightspaceUI/core/pr-6566/

Note

The build needs to finish before your changes are deployed.
Changes to the PR will automatically update the instance.

[dbatiste]

Does this mean we might end up delaying more events?

I always find this logic non-intuitive, in that it's relying on the internals of clickElem delaying things long enough for oneEvent to do its wire-up before the event is dispatched. But I appreciate the simplicity of it.

clickElem(closeButton);
await oneEvent(alert, 'd2l-alert-close');

This is more intuitive to me since it eliminates that internal timing dependency.

const p = new Promise(resolve => {
    closeButton.addEventListener('some-event', resolve, { once: true });
});
clickElem(closeButton);
await p;

Maybe there is a way we can set up the helpers to do the wire-up in this order, without consumers needing to write all this code...

[dbatiste]

Or, maybe we should add the setTimeout in clickElem instead, since it seems like it's just to support the order of ops in the test.