You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by Erik Wickstrom <er...@gmail.com> on 2004/09/29 20:07:43 UTC
Setting up a filter for sa-learn --spam
I'm trying to setup an email address where I can redirect false
negatives to so they will be learned as spam by Bayes without me
having to login to the shell.
This is what I have so far, but it doesn't seem to work. (note I am
using cPanel with exim) Right now I am using a forwarder, but I can
also setup a cPanel/exim filter if that is a better way to go.
spam@mydomain.com => | sa-learn --spam
Why isn't it working? (NOTE: I can't use the procmail script located
on the SA website because procmail doesn't really get along with
exim....).
Does anyone have any other solutions for this? I think it would be
great if SA put something in the header that you could click and have
a webpage open up and retrain your Bayes on that message, but until
that happens - workarounds are the way to go... :-)
Thanks!
Erik
Re: Setting up a filter for sa-learn --spam
Posted by Keith Hackworth <ke...@rpemail.com>.
you have to be careful doing this. It'll learn your new headers from your
forwarding as spam.
Your problem may be pathing. My sa-learn is in /usr/local/bin, which is
not in my default path when my startup scripts run. Try "|
/path/to/sa-learn --spam".
Keith
> I'm trying to setup an email address where I can redirect false
> negatives to so they will be learned as spam by Bayes without me
> having to login to the shell.
>
> This is what I have so far, but it doesn't seem to work. (note I am
> using cPanel with exim) Right now I am using a forwarder, but I can
> also setup a cPanel/exim filter if that is a better way to go.
>
> spam@mydomain.com => | sa-learn --spam
>
> Why isn't it working? (NOTE: I can't use the procmail script located
> on the SA website because procmail doesn't really get along with
> exim....).
>
> Does anyone have any other solutions for this? I think it would be
> great if SA put something in the header that you could click and have
> a webpage open up and retrain your Bayes on that message, but until
> that happens - workarounds are the way to go... :-)
>
> Thanks!
> Erik
>