The timing of "deactivate an EditContext" when HTMLElement.editContext is replaced with another one or null

[masayuki-nakano]

https://w3c.github.io/edit-context/#extensions-to-the-htmlelement-interface

In the list:

4. If oldEditContext is not null, then:
   4.1 Assert: oldEditContext's associated element is equal to this.
   4.2 Set oldEditContext's associated element to null.
5. Set this's internal [[EditContext]] slot to be editContext.
6. If oldEditContext is not null and oldEditContext is this's node document's active EditContext, then run the steps to deactivate an EditContext with oldEditContext.

I think that the Step.6 should be between Step 4.1 and 4.2 because it's odd and may be error prune if it happens after disassociated from an element.

[dandclark]

It is a good point that firing compositionend after the EditContexts are switched could be confusing.

If we make this change though, I would propose that step 6 be moved even earlier, before step 2.3. After 2.3 things are in an inconsistent state because we've changed the EditContext's associated element but not yet updated the Element's [[EditContext]] internal slot, so we shouldn't fire compositionend and run page author JS at that time.

[masayuki-nakano]

@dandclark

If we make this change though, I would propose that step 6 be moved even earlier, before step 2.3

Yeah, sounds reasonable!

[dandclark]

We resolved during today's Editing WG call to do this reordering.

[08:56] next: #75
[08:57] dandclark: as the spec is written now, if you have an EC that's being composed in and you disconnect it, we fire compositionend on the disembodied EC
[08:57] proposal: reverse these steps so that we fire compositionEnd before the EC is disconnected
[08:57] q+
[08:58] whsieh: would this allow script to reconnect an EC in the middle of disconnecting?
[08:59] dandclark: should make sure this edge case is not ambiguous
[08:59] sanketj_: do any of the known users of EC use compositionEnd
[09:00] dandclark: Google Docs, maybe?
[09:00] johanneswilm: follow masayuki's proposal, check edge case of reentrant EC reconnect