[TriLUG] Re: more spamassassin, procmail, damaged 'from' line and beyond

James Manning jmm at sublogic.com
Tue Apr 15 17:02:20 EDT 2003


> [Turnpike Man]
>  Otherwise, could I make this change to .procmailrc :
> 
> From:
> :0:
> * ^X-Spam-Status: Yes 
> caughtspam
> 
> To:
> :0:
> * ^X-Spam-Status: Yes 
> /dev/null
> 
> With that change, no longer would messages accumulate in caughtspam, but they
> would literally disappear, right??

yes, but false positives will happen at some point.

Also, you wanna keep around spam (and ham) for bayes learning.  You
can set some auto-learn settings (or keep the defaults), but training
uncaught spam is critical to SA getting better over time.

I have a couple of mutt lines to take current folder or current
message and file it over into a spam folder

macro index \cs "T.\n;s=spam\nD.\nc"
macro pager \cs "s=spam\n"

and a once-a-night cronjob to learn any of this filed spam

jmm at xanex:/home/jmm> cat bin/process_spam.sh
#!/bin/sh

PATH=/usr/bin:/bin
SPAM_MBOX=~/mail/spam

if [ -s $SPAM_MBOX ]; then
    sa-learn --spam --mbox $SPAM_MBOX
    cp /dev/null $SPAM_MBOX
    sa-learn --rebuild
fi

Since I'm pretty diligent about checking for false positives and
negatives, I allow autolearn to be a little greedier than default:

jmm at xanex:/home/jmm> cat .spamassassin/user_prefs
auto_learn_threshold_spam 7.0
auto_learn_threshold_nonspam 0.0

my standard SA .procmailrc entry's like most:

:0fw
| spamassassin

:0:
* ^X-Spam-Status: Yes
incoming-spamassassin.mbox

-- 
James Manning <http://www.sublogic.com/james/>
GPG Key fingerprint = B913 2FBD 14A9 CE18 B2B7  9C8E A0BF B026 EEBB F6E4



More information about the TriLUG mailing list