[selectors] Define how "legacy aliases" work on selectors.

[emilio]

There are various bits that mention legacy aliases on selectors as this bit of the draft:

As previous drafts of this specification used the name :matches() for this pseudo-class, UAs may additionally implement this obsolete name as an alias for :is() if needed for backwards-compatibility.

As far as I can tell, browsers don't implement these kinds of things the same way (some preserve the legacy version during serialization, but only on some selectors). I don't think it matters much either way, but it'd be good if moving forward we'd define a "preferred" way to do these kinds of things.

I think for simplicity, doing it as a parse-time operation is better (i.e., .selectorText for :-webkit-autofill will return :autofill). It'd be good to put some guidance in the spec about this at least.

[emilio]

This discussion was brought up in web-platform-tests/wpt#27098

[tabatkins]

Agree that it doesn't matter much, and also that I prefer the parse-level translation if possible.

[css-meeting-bot]

The CSS Working Group just discussed [selectors] Define how "legacy aliases" work on selectors., and agreed to the following:

• RESOLVED: convert at parse time

The full IRC log of that discussion

<dael> Topic: [selectors] Define how "legacy aliases" work on selectors.
<dael> Rossen_: Is emilio on?
<dael> [waiting for emilio to join]
<dael> github: https://github.com//issues/5847
<dael> emilio: Have a couple places where we define a selector where for legacy reason can be impl as alias but no definion of that.
<dael> emilio: Came up as I did compat work on other pseudo classes.
<dael> emilio: Compat can change what needs to be done but seems as if we should have guidance on how this should work
<dael> emilio: afaict two options are keep zeroizing the non-standard or convert them at parse time
<dael> emilio: Some inconsistencies in Blink. Gecko should always turn it intot he stanrdard at parse time. I think that's what we do for properties in general
<dael> Rossen_: Prop is align more closely with converting at parse time
<dael> Rossen_: Questions or should we resolve on that?
<dael> Rossen_: I'll take silence as no
<dael> Rossen_: No comments or objections?
<dael> Rossen_: I see support in GH from TabAtkins
<dael> RESOLVED: convert at parse time