I have a similiar approach, I noticed though that when the email is copied across to a public folder OR an IMAP folder, the email header/body is changed with ms-application/tnef. So whatever you feed sa-learn, it's not going to be effective as the body of the email has changed since from the original email.
I get my exchange users to drop all the spams (undetected by spamassassin) to a public folder and I use imap (exchange is imap enabled) to retrieve the email from the exchange to my linux box, and scp && ssh to copy and run sa-learn on the remote mailgateway. I am still trying to figure out how to fix this tnef problem, has anyone experience / noticed the tnef / winmail.dat thing in the email message? apparently this only happens if the email has been read by a Microsoft email client who is tnef aware, and it automaticaly changes the email!!! CP Subject: Re: Exchange/Outlook - how do you learn spam? Date: Tue, Jun 21, 2005 at 02:17:13PM -0500 Quoting E. Falk ([EMAIL PROTECTED]): : Easiest way to get them out of the Exchange public folder without : messing up the headers is via IMAP. There are some scripts available to : open the folder and read the messages (can't recall exactly where, but : if you can't find them let me know and I'll pass mine onto you - they're : modified versions of the scripts available online). : : Works very nicely. : : Evan : : Jon Dossey wrote: : >I'm sure a lot of us have a similar setup, linux/bsd mx gateways : >(running SA) relaying mail to Exchange, and Outlook clients. I'm just : >curious how everyone handles learning? : > : >It seems like a lot of people recommend a public folder for users to : >dump spam in, but how do you get it back out into a useable format that : >sa-learn will understand? Saving messages out of Outlook (for me : >anyway) into a txt file removes all the internet headers. : > : >So how else do you handle getting your messages back out of : >exchange/outlook, and sa-learn'ed? : > : >.jon