We've got a list that's plagued by vacation autoresponders, which is something Mailman 2.0x doesn't handle well. When I look back through the archives of the mailman-users list, I see that it crops up time and again without any meaningful solution (I don't count unsubbing the villains, manual moderation, whinging about broken autoresponders or re-setting Mailman reply-to as solutions).
Of course, a part-solution is to capture vacation-style subject lines using the 'header value' box on the Privacy page of the web administration. This works. But doesn't capture responses that only mention offices and vacations in the message body.
I believe I have the answer. in Utils.py lurks the list of administrivia phrases that cause mail to be held for moderation. I see no reason why I shouldn't add the worst offending vacation replies to this list. Indeed, I have and it works.
HOWEVER, it only captures text patterns at the *beginning* of the message body. If the body begins with "unsubscribe" or "options" (etc) or, as I now have it, "I will be out of the office" etc., the offending post is held for approval. But if any of those phrases (including the original administrivia phrases) appear further down the message body, the script misses them and the message goes out to the list.
SO ... anyone know how to extend the administrivia settings to capture any recurrence of a pattern in, say, the first 3 to 5 lines of a message? (NB. Normal Perl-style regexp pattern-matching doesn't seem to work here.)
Specifically: each pattern in Utils.py follows the form
"who": (0,0),
What do those numbers in the brackets mean and are they the clue to my answer? Or is the answer elsewhere?
Solve this and you've solved the whole vacation autoresponders problem, IMO.
Any help truly gratefully received.
TIA
Robert