Michelle Konzack schreef:
Udi Mottelo:
Michelle Konzack:
:0:
* ^Subject:.*=\?[^?]*\?[bq]\?[^?]*\?=
in.suspect
This would work for those peoples using only US-ASCII.
If =?big5? ... It works for chinese too :-)
and =?iso-2022-jp? for japanese
I mean, this allow only messages in 100% us-ascii
and droping all the rest.
(1) The recipe is only testing against the Subject, not against the
whole message.
(2) The message is not dropPed but stored in a mailbox.
(3) Everybody is free to make the recipe more specific, such as checking
for specific languages.
Some kind of spam has always more than one encoding part in the Subject:
:0:
* ^Subject:\/.*
* -1 ^ 0
* 1 ^ 1 MATCH ?? =\?[^?]*\?[bq]\?[^?]*\?=
in.suspect
(untested)
If I write you a message and put
äöüß in the subject, this message would automaticaly filtered.
Many clients allow for 8-bit characters in the Subject, with the ecoding
as mentioned in a Content header field (that actually are about the
body).
AFAIK that is not in any RFC yet, but I think it should be.
--
Groet, Ruud
____________________________________________________________
procmail mailing list Procmail homepage: http://www.procmail.org/
procmail(_at_)lists(_dot_)RWTH-Aachen(_dot_)DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/procmail