Re: Bidi issues

Adam Costello wrote:

 > I'd be interested in a bidi check similar to the Stringprep bidi check,
 > but which avoids allowing two distinct strings that display the same,
 > while still allowing all pure-ASCII strings.
 > 
 > Such a check could take care of single-component anomalies (like
 > A-123,456B) but not multiple-component anomalies (like 123.ABC.com).

Unfortuately, as far as I can see, the additional check on single
components is probably too complex to go into any protocol spec
(essentially it's a regular expression).  It's possible to simplify
it, of course, at the expense of disallowing perfectly safe labels.
For instance, if we restrict the discussion to labels that follow
hostname rules, then it is sufficient, I think, to disallow ARABIC
COMMA (being the only character of class ES or CS that survives
nameprep).  But this would be a little unfair on users of arabic
languages, given ARABIC COMMA is in fact perfectly safe when preceded
by arabic characters.

I hope to post my full analysis in a week or two.

	-roy