On Sun, Jan 15, 2012 at 19:09, Ricardo Signes <perl....@rjbs.manxome.org> wrote:
> * Ruslan Zakirov <r...@bestpractical.com> [2012-01-15T08:22:23]
>> May be it's proper from RFC point of view, but it breaks Request
>> Tracker application and it's going to cost a lot of development to get
>> it fixed.
> I've updated Github's repo with a change to only reject non-ASCII in the email
> address, which really is a problem. My guess is that you were having a
> with the decoded phrase legally containing non-ASCII.
Right guess. Is it legal? I don't think it is legal according to the
spec to have non-ascii phrase. As far as I recall it should be encoded
with Q/B. Proper client should first parse structured field and then
decode. We decode before parsing. It worked with earlier versions in
most cases. We even have workaround where we quote phrase after
decoding in some cases to preserve structure of fields.
It would be cool to have unicode phrases back.
> Please let me know if this helps, and if so, I will add some tests, adjust the
> docs, and release.
I can test later tonight.
Best regards, Ruslan.