Re: Sensible nnmail-split-fancy filters

[Bjorn Danielsson <bd@sunet.se>]

Arcady Genkin <a.genkin@utoronto.ca> wrote:
> The problem, the way I understand it, is that when a person replies
> within a mailing list with "Reply to all", I get two copies of the
> message. Then gnus deletes one of them, because they are duplicates,
> and keeps the first one to arrive. Which in most cases would be the
> one sent to me directly.
> 
> To avoid that, I use something like this:
> ,-----
> | ("delivered-to\\|cc" "freebsd-questions" "mail.FreeBSD-Questions")
> | (any "owner-rhua@FootPrints.net" "mail.RHUA")
> | ;; *lots* of other stuff, and in the end:
> | (to "\\(genkin\\|agenkin\\)@\\(sympatico\\|utoronto\\|home\\|canada\\)\\.c\\(a\\|om\\)"
> |     "mail.Private")
> `-----
> 
> I believe such usage is not awfully efficient, and often I have to
> wait longer than I would like to for the mail to be sorted. Also, the
> first kind of setup works only in some cases (i.e. won't work if
> mailing lists's address would be in To: header. And "any" seems the
> least efficient.
> 
> So, are there any tricks you guys use?  Just wondering...

I don't use Gnus mail splitting at all. I use procmail to sort stuff
into various mailbox spool files, and then in Emacs I setq mail-sources
to point at the procmail spool directory.  My .procmailrc file is
structured in a similar way to your example above, but the advantage
(IMHO) is that when I start Gnus it picks up the mail very fast since
it's already sorted by procmail. This also lets me run xbuffy to monitor
the spool files in the background while I am doing actual work in some
other window, which is something that happens all too often :-)

I just recently switched to pGnus for handling my mail after having
used Mew for two years (and before that mh-e, and before that RMAIL),
so I was already habituated to getting a quick response when I "unspool"
my new mail, and I wanted to keep it that way...

> Also, is there a nice tutorial for emacs's regexp's?

Not that I am aware of. I assume you have looked at the "Regexps" node
in the Emacs section of the info tree?  Some good examples of how to use
regexps can be found in Perl books, although they are not 100% identical
to the Emacs regexps.

-- 
Björn Danielsson  KTHNOC  <bd@sunet.se>