Normative: add "canonical" to AvailableCanonical(Calendars|Collations|NumberingSystems)

[FrankYFTang]

Address #37
@ljharb @anba @sffc @gibson042

[sffc]

CC @ptomato since you recently commented on the linked issue

[ptomato]

Thanks for tagging me in!

I would actually prefer that we go the other way with this. That is, the Available___ operations return both canonical and non-canonical names, and we also define Canonicalize___Name operations, so that in supportedValuesOf we can filter out the non-canonical ones.

The reason for this is because I'd like to bring AvailableCalendars and AvailableTimeZones into 262 along with Temporal, and implement IsValidTimeZoneName in terms of those. The current definition of AvailableTimeZones uses IsValidTimeZoneName and CanonicalizeTimeZoneName, so that wouldn't be possible as written.

I can write a PR for this proposal demonstrating what I mean, if that would be helpful.

[FrankYFTang]

I can write a PR for this proposal demonstrating what I mean, if that would be helpful.

Please show us your counterplan and we can discuss that. Thanks!

[FrankYFTang]

@ptomato ping- We are waiting for your " PR for this proposal demonstrating what I mean."

[ptomato]

I have opened #49 as a counterproposal, please let me know what you think.

[FrankYFTang]

I thiink #49 is beyond the scope of this proposal. The goal of this proposal is to provide functionality for the userland code to ues, not to make the future spec writer reuse AOs beyond what this spec intend to provide- especially not to change the rest of the ECMA402 to use those AOs. Those operations needs a separate discussion beyond what this proposal intend to solve.

[FrankYFTang]

@anba - do you think this PR would address the issue you mentioned in #37 sorry for the delay. @sffc @ryzokuken @gibson042 please review

[anba]

@anba - do you think this PR would address the issue you mentioned in #37 sorry for the delay.

Yes, I think that should be okay. We can later integrate the algorithms from #49 to more formally specify what "canonical" means.

[sffc]

As a reminder, please make sure there is a follow-up issue filed in the ecma402 repo to continue this discussion after this proposal lands as stage 4.

[ptomato]

As a reminder, please make sure there is a follow-up issue filed in the ecma402 repo to continue this discussion after this proposal lands as stage 4.

→ tc39/ecma402#726