Re: Open Issue: Splitting of local-part into labels and where?

At 19:48 03/02/13 +0100, Marc Mutz wrote:
>Hi!
>
>The question of whether or not to split the dequoted local-part into
>labels and if so, at which delimiter characters, remains unanswered in
>the -00 draft.

>The more interesting point is _where_ to split. There are some very
>obvious characters (mainly full stops and hyphens), but apart from
>that, the rest of the candidate chars is much less clear.

My understanding is that this is very much a slippery slope.
So we better stop as soon as possible. I see some argument
for making '.' a delimiter, because that would make it easier
to apply the same function to a whole email address as to
a domain name only. But that on the other hand prohibits
us to make '-' a delimiter.



>In addition, all such separators should be recognized in all their
>variants (for full-stops, see IDNA, for @ see draft, for others I admit
>to not know the Unicode repertoire by heart ;-)) and be replaced with
>their US-ASCII equivalents.

There are a lot of variants, and they very much depend on
circumstances. IDNA recognizes the full-width and ideographic
variants of the full stop, but as far as I remember, it does
not recognize any other variants or equivalents.

Making sure that a user enters a '@' in the right form can and
should be a quality of implementation issue under the responsibility
of the application. Such issues can easily be integrated into
the input architecture e.g. for East Asian languages and scripts.


If we treat all non-alphanum ASCII characters special, then what
about non-alphabetic/syllabic/ideographic symbols,... in other
scripts? Shouldn't they also be treated as special separators?


Regards,    Martin.