On Tue, Jul 15, 2003 at 01:59:49AM -0700, Virtual University wrote:
Is there a FAQ somewhere, or does anyone know, which is more efficient
insofar as reducing server load when filtering spam -- having a set of
recipes with multiple match checks (aaa|bbb|ccc|ddd) or using scoring?
I haven't seen anything like that in a FAQ, but I would think that
matching as much as possible in a single regexp would take less CPU than
processing multiple regexps and calculating a score for each of them.
So I would prefer to use:
:0
* (aaa|bbb|ccc)
{ do something }
than:
:0
* 1^0 aaa
* 1^0 bbb
* 1^0 ccc
{ do something }
Also, can a procmail score be assigned to a variable such as $SCORE so
the result can be reported in a mail header?
The procmailsc man page's MISCELLANEOUS section mentions $= which you
can insert into a header using formail, or use elsewhere, thus:
:0
* 1^1 aaa
* 3^1 bbb
* 2^1 ccc
| formail -A"X-letter-factor: $="
should work, or you can use a previous score in later recipes:
:0
* 1^1 aaa
* 3^1 bbb
* 2^1 ccc
{ LETTERFACTOR=$= }
...
:0
* $ -${LETTERFACTOR}^0
* 5^1 ddd
* 1^1 eee
{ do something }
--
Paul Chvostek
<paul(_at_)it(_dot_)ca>
Operations / Abuse / Whatever
it.canada, hosting and development http://www.it.ca/
_______________________________________________
procmail mailing list
procmail(_at_)lists(_dot_)RWTH-Aachen(_dot_)DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/procmail