Am 17.02.2015 um 18:46 schrieb Wietse Venema:
li...@rhsoft.net:
is it intentional that if a message hits more than one Regex that it
creates also more than once BCC like below? it's little bit surprising
Of course. If more than one header line matches the table, then
more than one action will execute(*).
Why is that surprising?
because you can't make decisions who get what BCC and there is a
practical difference if a message got above 5.5 points by whatever rule
or has a BYES_%0
(*) The exceptions are REJECT and DISCARD which terminate further
table lookups because the decision is obviously final.
and DUNNO
because in all known cases the first rule hit's and the evaluation of
the file is stopped
That is absolutely incorrect.
REJECT, OK, DUNNO, DISCARD are final inside *the same* access table
maybe there is a border-case i never faced where it's not so
the intention of the "spamfilter+inbox...@rhsoft.net" is to get
ham-train candiates collected in a seperate folder while the inbox is
checked for spam-candidates
/^X\-Spam\-Flag: Yes/ BCC spamfil...@rhsoft.net
/^X\-Spam\-Report:.*(BAYES_50)/ BCC spamfilter+inbox...@rhsoft.net
