Changes between Initial Version and Version 1 of Ticket #27029, comment 40


Ignore:
Timestamp:
Jul 3, 2026, 3:02:51 PM (3 weeks ago)
Author:
Mike Edmunds

Legend:

Unmodified
Added
Removed
Modified

  • Ticket #27029, comment 40

    initial v1  
    33But if that simplified proposal—or something like it—is ''not'' accepted (and this ticket remains accepted), then this ticket will need to be addressed by following RFC 6532 to allow UTF8-non-ascii characters in the EmailValidator user_regex, in both the dot-atom and quoted-string parts.
    44
    5 Since this is a potentially breaking change, it would need to be opt-in, e.g. via a new `accept_eai_user` EmailValidator keyword param (like DomainNameValidator's `accept_idna`, but defaulting to False).
     5Since this is a potentially breaking change, it would need to be opt-in, e.g. via a new `accept_eai_username` EmailValidator keyword param (like DomainNameValidator's `accept_idna`, but defaulting to False).
    66
    77To address concerns about regular expression complexity, it would probably be helpful to break up EmailValidator's user_regex into simpler components, similar to what was done in DomainNameValidator. (I suggest using the RFC 5322 terms: atom, dot_atom, qtext, quoted_pair, quoted_string. Since RFC 6532 is specified using those same terms, it should help with reviewing a PR for this ticket.)
Back to Top