David W. Tamkin schreef:
Ruud:
It contains:
* MATCH ?? ^^\/([^-]|-[^-])+
which I think does the same as
* MATCH ?? ^^\/.*[^-]
Is there a special reason for your (or Bart's) approach?
There is a difference, Ruud; your simplification can match a string
with two or more successive hyphens, as long as there is at least one
non-hyphen at the end. Sean's (or Bart's) cannot match a string with
two adjacent hyphens.
Ah, hadn't seen that. Variant:
* MATCH ?? ^^\/.*[^-]-
* MATCH ?? ^^\/([^-]|-[^-])+
I think ^^\/(-?[^-])+ is equivalent to it, but
there are always those considerations of the form that's most
efficient for procmail to parse, and I never know how to determine
that.
Yes, I had thought of that variant too. Well, let's just benchmark it,
by timing it against a header of a 1000 long lines that only has a
match at the last line, and several almost-matches earlier on. And
also with 1000 matching lines (scoring).
$ time procmail ...
--
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