I've written a couple scripts to re-process false positives that get
marked up by spamassassin (after I have manually quarantined them into
an MBOX named "false_positives").
This runs from the commandline:
formail -s procmail legitimize_false_positive.pmr < false_positives
mv false_positives false_positives.delete_me
This is the procmail script "legitimize_false_positive.pmr":
:0 hbfw:
| spamassassin --remove-markup --auto-whitelist --add-to-whitelist
INCLUDERC=$HOME/.procmailrc
The goal here is to clean up the spamassassin ("SA") markups, and then
put the email back through my normal chain of procmail scripts -- and
doing so in such a way that the message does not get falsely accused
again.
This script nearly works. The problem is that the spamassassin
auto-whitelisting capability is broken. It fails to whitelist the
sender, so it is then flagged and filtered as spam.
I'm about to give up on this somewhat elegant approach, and go with a
different strategy; that is, using formail to insert a header field
(such as x-spam-status: immunity) that will later prevent the message
from being SA processed at all. But I'm not sure I like that
approach, so I'd first like to discover why the script above doesn't
work as it should.
____________________________________________________________
procmail mailing list Procmail homepage: http://www.procmail.org/
procmail(_at_)lists(_dot_)RWTH-Aachen(_dot_)DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/procmail