[pmg-devel] [PATCH pmg-api 0/5] ruledb - improve experience for non-ascii tests and mails

[Dominik Csapak]

ok tested a bit around with this series

generally "works" as in mail flows and reaches the right things
(recipient/quarantine/etc) AFAICS

some things are a bit broken:

* using notification/modify field with smtputf8 has not the desired result:
   sending an email with smtputf8 and an utf8 encoded subject results in
   the subject being \x "encoded", in the quarantine the notifications
   and the resulting mail on delivery
   (unicode characters configured in the rule themselves show properly)

   ideally this would be detected and properly de/encoded

* still some issues with the statistics database
   (talked to stoiko off list about that)

* the quarantine ui is rather broken with this:
   neither the sender/recipient nor the mail/subject are correctly (en?)decoded
   such that the utf-8 bytes are double encoded

   we may want to save the info if the mail came from a 'smtputf8' source somewhere
   so that we can properly de/encode the info again?

   also i'm not sure if we want to release it, with the quarantine in this state.
   i guess it'll be one of the first bug reports then..

What worked well:

* using unicode characters in the rule system (where appropriate):
   - rule names
   - rule comments
   - rule values

   i tested as many rules as i could find where it would make sense:
   match field, attachment replacement, notify text, modify field, and so on

* sending / receiving mails with unicode characters in the sender/recipient


What's missing:

* ldap and who objects are a big one -> we should soon think about how we can do that
* statistics entries


all in all a good step in the right direction, thanks :)