Ruud H.G. van Tol said:
Mike Peeler schreef:
* $ ^Subject:(.*\<)?\/.*
That $ can go.
Thanks, Ruud. The production recipe I snarfed that from had used $NW:
* $ ^Subject:.*\/$NW.*
Not to bring $NW for non-whitespace into this discussion, I made a wee
tactical change. I overlooked the orphaned $. Good catch.
Variant:
OR = 9876543210^0
...
*$ $OR SUBJ ?? ()\<ports/x11/xterm( |$)
That's nice. The use of $GO^0 is amid a run of scoring conditions one
of which decides the outcome if tripped. $OR could be good sometimes.
But, $OR is slightly misleading, as the short-circuited score is still
ANDed with any non-scoring conditions. IOW, ORed conditions are bound
together--in effect parenthesized to flip the usual and/or precedence.
FWIW, the rationale I learned for disjunctive normal form was to group
one set of true conditions, vs. another group, vs. another. Alternate
possible worlds.
-mdp
____________________________________________________________
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