(make-regexp...) in an alist

(make-regexp...) in an alist

[Norman Walsh]

Hello World,

I'm converting over to nnmail-split-methods from my old
home-grown perl+procmail splitting system.  (Hey, gnus does
everything else, why deprive it of this one pleasure ;-)

So I started building the nnmail-split-methods alist and discovered
that there are some parts I just don't want to build by hand.
Then I found (make-regexp) and was happy again.

But now the question is, how do I build the nnmail-split-methods
alist with the _value_ of (make-regexp) in it?

I'm looking for something to replace this sort of thing (which
is clearly wrong):

(setq nnmail-split-methods
      '(("letterpress" "Sender:.*letpress@")
	("minolta" "To:.*minolta-l")
	("emacs.ding" "To:.*ding@")
	("spam" (make-regexp '("From:.*@savetrees.com"
			       "From:.*@ispam.net"
			       "From:.*@earthlink.net"
			       ...etc. add nauseum here
			       "Subject:.*are you being investigated")))
	("dssslist" "To:.*dssslist")))

Sorry this is only a quasi-gnus question, I'm just hoping for
the kindness of elisp wizards ;-)

--norm

Re: (make-regexp...) in an alist

[William M. Perry]

Norman Walsh <norm@berkshire.net> writes:

> Hello World,
> 
> I'm converting over to nnmail-split-methods from my old
> home-grown perl+procmail splitting system.  (Hey, gnus does
> everything else, why deprive it of this one pleasure ;-)
> 
> So I started building the nnmail-split-methods alist and discovered
> that there are some parts I just don't want to build by hand.
> Then I found (make-regexp) and was happy again.
> 
> But now the question is, how do I build the nnmail-split-methods
> alist with the _value_ of (make-regexp) in it?

  Backquote:

(setq nnmail-split-methods
      (list '("letterpress" "Sender:.*letpress@")
	    '("minolta" "To:.*minolta-l")
	    '("emacs.ding" "To:.*ding@")
	    `("spam" ,@(make-regexp '("From:.*@savetrees.com"
				      "From:.*@ispam.net"
				      "From:.*@earthlink.net"
				      "Subject:.*are you being investigated")))
	    '("dssslist" "To:.*dssslist")))

-Bill P.

Re: (make-regexp...) in an alist

[Paul Franklin]

>>>>> Norman Walsh writes:

 > (setq nnmail-split-methods
 >       '(("letterpress" "Sender:.*letpress@")
 > 	("spam" (make-regexp '("From:.*@savetrees.com"
 > 			       "From:.*@ispam.net"
 > 			       "Subject:.*are you being investigated")))
 > 	("dssslist" "To:.*dssslist")))

I've attached a complicated example which doesn't use backquotes.
Basically, the idea is to use "(list" to get yourself one level down
(or cons/concat/make-regexp), or "'(" to avoid interpretation of
everything below it.

(Yes, my split methods are way out of hand.)

--Paul

(setq pdf-nnmail-split-methods
      (let ((not-paul "\\([^p]aul\\|[^a]ul\\|[^u]l\\|[^l]\\)")
	    (not-paul-pdf
	     "\\([^p]aul\\|[^a]ul\\|[^u]l\\|[^p]df\\|[^d]f\\|[^lf]\\)"))
	(list

	 '((gnus-warning)
	   ("-mail.duplicates" . "\\<duplicate\\>"))

	 '((x-from-line)
	   ("-demons.mail" . "\\<postmaster\\>"))

	 '((a)
	   ("-demons.mail" . "\\<MAILER.?DA?EMON\\>")
	   ("-demons.listproc" . "\\<listproc\\>")
	   ("-demons.majordomo.uw.cs" . "\\<Majordomo@cs.washington.edu\\>")
	   ("-demons.majordomo" . "\\<Majordomo\\>"))

	 (list '(a)
	       (list "-mail.ads"
		     "[^a-z0-9][0-9][0-9][0-9][0-9]+@"	; mbox is 4+ digits
		     "abuse@"		; typically my complaints
		     (concat
		      "@[-.a-zA-Z9-0]*\\." ;subdomains + "."
					; top-level domain errors:
		      "\\([a-zA-Z]*[-0-9][-a-zA-Z9-0]*" ; inval char 
		      "\\|[a-zA-Z]\\|" ; too short (single char)
		      "[a-zA-Z][a-zA-Z][a-zA-Z][a-zA-Z]+\\)" ; too long (4+)
		      "[^-.a-zA-Z9-0]"))) ; notice end of addr

;; lots more simpler stuff here

	 '((l)
	   ("-uw.cac"
	    "@.*\\<cac.washington.edu\\>"
	    "postmaster@u.washington.edu"))

	 (list '(l)
	       (cons "-transmeta.misc"
		     (concat not-paul "@.*\\<transmeta.com\\>")))
	 
	 '((l received message-id references in-reply-to)
	   ("-mail.uw.cs.iws"
	    "\\<grizzly\\>" "\\<lynx\\>" "\\<wolf\\>"))
	 
	 (list '(l)
	       (cons "-mail.uw.cs"
		     (concat not-paul "@.*\\<cs.washington.edu\\>")))

	 (list '(l)
	       (cons "-mail.uw.misc"
		     (concat not-paul-pdf "@.*\\<washington.edu\\>")))

	 (list '(l)
	       (cons "-mail.misc"
		     (concat not-paul-pdf "@.*\\.")))

	 '(nil ("-mail.uw.cs" . "")))))

Re: (make-regexp...) in an alist

[Stefan Waldherr]

>>>>> "Norman" == Norman Walsh <norm@berkshire.net> writes:

    Norman> So I started building the nnmail-split-methods alist and
    Norman> discovered that there are some parts I just don't want to build by
    Norman> hand.  Then I found (make-regexp) and was happy again.

Can some kind soul explain me, why I would want to use it at all? It as in
`make-regexp', not `nnmail-split-methods'. Why not using multiple rules

       ("spam" "From:.*@savetrees.com")
       ("spam" "From:.*@ispam.net")
       ...etc. add nauseum here 
       ("spam" "Subject:.*are you being investigated")

Thanks,
Stefan.
-- 
Stefan Waldherr                office +1 (412) 268-3837
                                  fax +1 (412) 268-5576
                               e-Mail swa@cs.cmu.edu
                                  www http://mind.learning.cs.cmu.edu/

Re: (make-regexp...) in an alist

[Hrvoje Niksic]

wmperry@aventail.com (William M. Perry) writes:

>   Backquote:
> 
> (setq nnmail-split-methods
>       (list '("letterpress" "Sender:.*letpress@")
> 	    '("minolta" "To:.*minolta-l")
> 	    '("emacs.ding" "To:.*ding@")
> 	    `("spam" ,@(make-regexp '("From:.*@savetrees.com"
> 				      "From:.*@ispam.net"
> 				      "From:.*@earthlink.net"
> 				      "Subject:.*are you being investigated")))
> 	    '("dssslist" "To:.*dssslist")))

Ha!  Unnecessary consing!  How about this:

(setq nnmail-split-methods
      `(("letterpress" "Sender:.*letpress@")
	("minolta" "To:.*minolta-l")
	("emacs.ding" "To:.*ding@")
	("spam" ,@(make-regexp '("From:.*@savetrees.com"
			         "From:.*@ispam.net"
				 "From:.*@earthlink.net"
				 "Subject:.*are you being investigated")))
	("dssslist" "To:.*dssslist")))

-- 
Hrvoje Niksic <hniksic@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
"Beware of bugs in the above code; I have only proved it correct,
not tried it."                                    -- Donald Knuth