On Wed, 30 Oct 1996, Alan K. Stebbens wrote:
Another solution is to place all of the regexps into a file (for
example, called "spamsters"), and use an egrep against the possible
source addresses:
:0:spamsters.lock
* ? formail -xFrom -xFrom: -xReply-To: -xSender | \
egrep -s -f spamsters
| $FLUSHFILTER $LIST > /dev/null
An excellent suggestion- thanks much! But it turns out that the egrep
I have laying around has an even smaller BUFSIZ than the builtin egrep
in Procmail, and I couldn't get gnuegrep to work at all for some
reason (it always seemed to return a status of 0 even with a match,
which is odd). I'll just use the LINEBUF variable as you suggested-
it's probably more efficient than forking off the egrep and formail
processes, anyway.
On Oct 30, Wotan wrote:
This has one problem. It doesn't allow for the many spammers who aren't
above forgery.
Cyberpromo was extremely bad when it came to filtering due to the MANY
domains mail could have come from.
True. But I amended my recipe to have a slightly more inclusive
header, as well:
:0
* ^(From|Reply-To|Sender|Message-ID).*(.*ramp.com|\
.*ramp.net|\
199.182.120.3|\
1stamend.com|\
[and so on, deleted...]
Presumably one could also stick in ^Recieved-from lines as well, but I
think that snagging the ^Message-ID will work well enough against our
normal crop of spammists.
We'll see how that works for a while. I think..
-skod
--
Scott Griffith, Sun Microsystems Lumpyware
Who didn't have _anything_ to do with the 386i...
Nope, not a bit. Nothing at all. Not me, nosir. Nope.
Return Path : skod(_at_)Sun(_dot_)COM