Re: [SLUG] Re: Killfile howto

From: Paul M Foster (paulf@quillandmouse.com)
Date: Tue May 29 2001 - 01:52:48 EDT


On Tue, May 29, 2001 at 12:15:40AM -0400, Russ Herrold wrote:

> On Mon, 28 May 2001, Paul M Foster wrote:
>
> > > and then later, I use:
> > >
> > >:0 h
> > > * ^X-killfile: cheseburger
> > > * !^X-passfile: olive
> > > | ~/$MAILDIR/killfile-dump.$DOM
>
> ... oops -- correction -- this is a deliver recipe -- no
> 'pipe' needed or wanted (that's what I get for sanitizing my
> recipes)
>
> > If the last recipe is rewritten as:
> >
> > :0 h
> > * ^X-killfile: cheeseburger
> > killfile-dump.$DOM
> >
> > the _headers only_ of _all_ messages passed to it will end up in your
> > killfile folder.
>
> ... no -- the _match_ option (lowercase h or b) is against the
> _headers_ -- the default delivery is of the entire message,
> both header and body, regardless of whether a filter or a
> delivery recipe (those are the Capital H or B options, to get
> just that part of a message)
>

Actually, my description above of what occurs _did_ occur when I used
that recipe.

Just to be clear, the h and b flags apply to the _action_ taken by the
recipe. The H and B flags apply to the _egrep_ done by the recipe. This
is per man procmailrc and per your previously cited URL, at:

http://www.hcs.harvard.edu/~thurston/ua/pm-tips.html#flags_h_and_b

Thus, a recipe that starts as:

:0Hb

will egrep the header only, and (if deliverable per the egrep) deliver
the body only to whatever pipe, file or email address is in the action
part of the recipe. This behavior could be mediated by the f or c flags,
but my understanding is as above.

I struggled with this for a long time (prior to this thread). If this is
incorrect in any way, please correct me and point me to the appropriate
docs.

Paul



This archive was generated by hypermail 2.1.3 : Fri Aug 01 2014 - 20:12:23 EDT