Add String.prototype.normalize
http://wiki.ecmascript.org/doku.php?id=strawman:unicode_normalization#add_normalize_method
Resolution:
- total
- deterministic
- idempotent normalization (normalizing the result of normalization again will return the first result)
WH: Note that Unicode got this wrong a while back (their normalization algorithm wasn't idempotent, and it didn't even form proper equivalence relations). They fixed it since then and now explicitly state that it's idempotent.
So this presumably means that if we have:
var ̍ϓ = 5; //u+03D3
var Ύ = 6; //u+038E
we will have two distinct variable even though, the unicode code points logically represent the same character. (and assuming both codepoints are both valid identifier characters).
Right?
If so, it seems we should explicitly state this, probably somewhere in section 7.
Norbert, do you have any language you would like to suggest that would clarify that after your proposed deletions are made?
fixed in rev 14 editor's draft
in Rev 14 draft