|>> What I'd like to do is run an external filter program from my .forward
>> over incoming email to decide if its spam and use the exit code (or
>> output?) in a conditional. This is to decide wether to junk the message
>> or use the remaining rules in my .forward.
>
>I seem to recall that this sort of thing isn't possible with the present
>filter system, as all the output of the filter is, is a list of
>addresses (including files and pipes) to direct, rather than actually
>executing the pipe /during/ the filter processing. If you needed this,
>you'd probably have to set a global procmail somehow (not sure precisely
>how though).
Can you not make the spam filter add a header? For example, SpamBeGone, a
very popular filter, adds an X-Priority: header with a value of 1 thru 5; 1
is good stuff 5 is spam.
The nice thing about SpamBeGone (
http://www.internz.com/SpamBeGone/") is
that you train it what your idea of spam is. And, from personal experience,
it's pretty effective - in the last 3 months no spam has hit my inbox, and
no real mail has been misplaced. It took about 12 - 20 messages to train,
and only a handful more to make the prioritising of other mails useful.
Oliver
--
*** Exim information can be found at
http://www.exim.org/ ***
