Re: nnmail-split-it problem

Re: nnmail-split-it problem

[David Moore]

The following message is a courtesy copy of an article
that has been posted as well.

<crossposted to the Gnus mailing list, as it applies to red gnus as well>

Hunter Kelly <retnuh@corona.pixar.com> writes:
> I'm a new Gnus user, and was having problems getting
> nnmail-split-fancy to work correctly.
> 
> Digging through the source it seems that nnmail-split-it does
> something really annoying:
> When you have a list that is something like 
> ("From"  "baz" "wherever")
> The regular expression it forms is 
> "^\\(From\\):.*\\<\\(baz\\)\\>"
> Why does it put the \< in there?  The restriction of only being able
> to match on word boundries is really lame.  If I want word boudries,
> I'd put them in the regexp myself, dangit!  :)

	Yes, it does do that.  It should probably be documented better,
I remember being confused by that at first also.  (mostly because I had
to change some of my patterns from "foo@" to "foo@.*".)  However, since
then, I've come to like the decision, since it helps protect in
situations like:  I have a friend jack@foo.com, but there might also be
mjack@foo.com.  With 250 or so splits now, the majority all set to
crosspost, it can help.  Word alignment helps also with subject
patterns, although I guess one might argue, people can stick \<\> in
there themselves.

> Anyway, included is a patch that will fix this rudeness.

	I guess people might want to comment on whether this should be
changed or not.  I'd be concerned about this sort of a change
potentially breaking people's mail filing in unexpected ways.  But then
again, it is a little strange.



-- 
David Moore <dmoore@ucsd.edu>       | Computer Systems Lab      __o
UCSD Dept. Computer Science - 0114  | Work: (619) 534-8604    _ \<,_
La Jolla, CA 92093-0114             | Fax:  (619) 534-1445   (_)/ (_)
<URL:http://oj.egbt.org/dmoore/>    | Solo Furnace Creek 508 -- 1996!

Re: nnmail-split-it problem

[Per Abrahamsen]

>>>>> "DM" == David Moore <dmoore@UCSD.EDU> writes:

DM> 	I guess people might want to comment on whether this should be
DM> changed or not.  I'd be concerned about this sort of a change
DM> potentially breaking people's mail filing in unexpected ways.  But then
DM> again, it is a little strange.

Changing it would silently break useful documented behaviour.  If you
want to change it, you should add a `nnmail-split-fancy-not-so-fancy'
variable to do so, and the default should be the current behaviour.
Or perhaps add a new `nnmail-split-recursive' method for splitting
mail.

Re: nnmail-split-it problem

[Mark Eichin]

> 	Yes, it does do that.  It should probably be documented better,
> I remember being confused by that at first also.  (mostly because I had

I was just explaining that to a friend who was also confused by it; I
was confused when I first set up splits, until I stepped through the
code and found the \\< use... then went back and read the docs, and
The Light Dawned -- seriously, I read the doc and thought "OHHH,
*that's* what he was talking about..."  The original wording, 

:	VALUE must match a complete word according to the fundamental
:	mode syntax table.

didn't make sense -- perhaps if it had said "VALUE *is matched as* a
complete word" or something like that.  I don't know, really -- every
one I've talked to about it was someone who would have understood
"VALUE is wrapped in \< \>" *perfectly* :-)

:	You can use `.*' in the regexps to match partial field names or words.

That looks like new text, and might have been enough of a hint to me
first time around; then again, maybe it wasn't...

Re: nnmail-split-it problem

[Lars Magne Ingebrigtsen]

Mark Eichin <eichin@cygnus.com> writes:

> :	VALUE must match a complete word according to the fundamental
> :	mode syntax table.
> 
> didn't make sense -- perhaps if it had said "VALUE *is matched as* a
> complete word" or something like that.  I don't know, really -- every
> one I've talked to about it was someone who would have understood
> "VALUE is wrapped in \< \>" *perfectly* :-)

I've now added this.

> :	You can use `.*' in the regexps to match partial field names or words.
> 
> That looks like new text, and might have been enough of a hint to me
> first time around; then again, maybe it wasn't...

It's new text.

-- 
(domestic pets only, the antidote for overdose, milk.)
  larsi@ifi.uio.no * Lars Ingebrigtsen