`/u` flag vs. `/v` flag

[rauschma]

There is no way of escaping that works in all scenarios:

	[<]	[\<]
Neither /u nor /v	✅	✅
/u	✅	❌
/v	❌	✅

What’s the best approach then? Only support /v?

More information: https://github.com/tc39/proposal-regexp-v-flag/issues/71

[lionel-rowe]

For completeness — I know it's been discussed in other issues as a solution to various issues of context-sensitivity, though it's suboptimal in terms of conciseness and readability:

	[<]	[\<]	[\x3c]
Neither /u nor /v	✅	✅	✅
/u	✅	❌	✅
/v	❌	✅	✅

[bakkot]

One option (mentioned elsewhere and elsewhere) is to just make all of those escapes legal (as identity escapes) in u-mode regexes. I think that's the easiest - we wouldn't want to give those different meanings than in v-mode, I think.

[eemeli]

It seems like adding a second flags argument to RegExp.escape() would provide a pretty ergonomic solution (as in the RegExp constructor), if there are differences in required escapes.