I received answers to my questions off-list, so I'll reply to myself
and share them with the rest of everyone. :)
> describe CJB_ESHOP_GR Contains reference to e-shop.gr
> header CJB_ESHOP_GR ALL =~ /e-shop\.gr/i
> score CJB_ESHOP_GR 0.01
>
> The efficiency and wisdom of matching against any headers aside,
> is this correct so far? I'm just thinking simple, I just want to
> match the e-shop.gr string.
While not the best way to catch one of these emails, this will
indeed do what I had intended - match the string 'e-shop.gr' in
any of the email headers.
> Assuming that's OK, I moved on to creating the rule. Now, I store
> my user preferences in a PostgreSQL database, so this is where I'm
> slightly unsure... Can I simply add my rule above to my userprefs
> table, one line per row? Like so (pardon the line wrap):
> Is this the correct way to do it? Or do I *have* to add them to
> local.cf or another such local file?
Nope, this isn't possible - rules cannot be stored in a database,
nor will most installations allow user-defined rules. I went
ahead and created my own .cf file in /etc/mail/spamassassin,
and that works great.
> I'm also unsure if I need to restart spamd to pick up on this new
> rule, if indeed it can be stored in SQL - my other userprefs do not
> need a restart, but rules I'm very unsure about. I tried cat'ing a
> sample spam and piping it to spamc (cat sample.txt | spamc), but
> it did not hit my test rule. And yes, e-shop.gr shows up in the
> Received and From headers. :)
Yes, a restart is necessary. Always do a 'spamassassin --lint'
first, too.
Thank you!
Benny
--
Whoever said that hell hath no fury like a women scorned never
owned a cat.
-- bash.org
