Richard Salts schrieb:
[...]
> b. Add the badwords into spamassassin rules and search $spam_report
> for matches on these rules. It is a bit of an abuse of spamassasin but it
> should work.
I'd not chech $spam_report for matches but would assign extreme scores
to those tests. Then you could reject based on score, arguing "a score
of 1000000 can never be reached without bad words".
In that case all spamassin-exim-exampls should work, like
http://wiki.exim.org/ExiscanExamples#Defining_multiple_spam_thresholds_with_different_actions
I just don't like string parsing and matching...
Oh, and don't forget...
Usually, people don't feed all messages to spamassassin but assign a
size-limit. Without a size-limit spamassassin might cause problems like
huge memory-consumption.
Basically, if you are worried about bad words after 32k, can you justify
spamassassin only scanning something like the messages with less than
100k? Or less than 300k, the limit I use here? With such a limit, a
bigger message wouldn't be scanned at all, even if it started with a bad
word.
And don't forget that you won't catch messages with attachments like
doc, where the bad word is hidden in the doc-file...
--
CU,
Patrick.
