Running reconstruct as cyrus user

[Robert Fitzpatrick]

On Fri, 2004-04-02 at 14:37, Craig Ringer wrote:
> On Sat, 2004-04-03 at 01:01, Robert Fitzpatrick wrote:
> 
> > Can someone offer advice on how to get this command the script is trying
> > to execute working? Maybe I don't need to be using the 'cyrus' user to
> > do this?
> 
> I'd be reluctant to manually move messages and reconstruct unless I had
> to. I do currently do this after nightly virus re-scans, if any messages
> are found to be suspect that the scanner missed when the msg was first
> received, but that's only until I get a perl/Python script done to do it
> the proper way.

I got it working with the help of a friend. We were able to write a C
script to setuid prior to running reconstruct.

This Cyrus-IMAP-2.1.16 machine is a mail gateway with SpamAssassin,
Postfix on FreeBSD-5.2.1 and I have two mailboxes called spam and ham.
When messages get by SpamAssassin on this gateway, they are still tested
at the destination mail server by procmail using a different Spam
filtering system called SpamBouncer. If SpamBouncer finds any messages
to be definite Spam, it sends a copy to spam@<the_gateway_host> and if
it finds it to be clean, sends to ham@<the_gateway_host>.

>From the spam and ham mailboxes, this script on the gateway moves those
messages into a created tmp folder and sends them through sa-learn and
removes the tmp folder. After moving the messages, it does reconstruct.
This is working beautifully, the only issue is the cache seems to still
have the messages listed. My IMAP client still shows the messages, but
they are physically not there. Is my client looking somewhere else? I
can open the messages, etc. from within my client. But nothing in
/var/spool/imap/user/spam folder but cyrus.cache, cyrus.index and
cyrus.header. Is my client pulling from these files and if so, can I
refresh them someway?

-- 
Robert

---
Home Page: http://asg.web.cmu.edu/cyrus
Wiki/FAQ: http://cyruswiki.andrew.cmu.edu
List Archives/Info: http://asg.web.cmu.edu/cyrus/mailing-list.html