Pgnus-0.33: bug in "mailcap-possible-viewers"

Pgnus-0.33: bug in "mailcap-possible-viewers"

[Lloyd Zusman]

I found a small bug in the `mailcap-possible-viewers' routine in the
`mailcap.el' file in pgnus-0.33.  The `minor' parameter sometimes gets
passed in as `nil', and this causes `string-match' to fail.  I'm not
sure under what conditions this parameter gets set to `nil' ... this
might have to do with badly formed MIME headers in the message being
parsed ... ???

But whatever the case, the enclosed patch seems to fix the problem.
Use the `-b' option with `patch', since my `diff' program might have
have messed with the leading white space.

By the way, today is the first day that I tried anything from the
pgnus series.  My congratulations to you, Lars ... your MIME-handling
code works great!

*** mailcap.el.orig	Thu Oct  8 22:41:00 1998
--- mailcap.el	Thu Oct  8 22:41:07 1998
***************
*** 469,475 ****
        (cond
         ((equal (car (car major)) minor)
  	(setq exact (cons (cdr (car major)) exact)))
!        ((string-match (car (car major)) minor)
  	(setq wildcard (cons (cdr (car major)) wildcard))))
        (setq major (cdr major)))
      (nconc (nreverse exact) (nreverse wildcard))))
--- 469,475 ----
        (cond
         ((equal (car (car major)) minor)
  	(setq exact (cons (cdr (car major)) exact)))
!        ((and minor (string-match (car (car major)) minor))
  	(setq wildcard (cons (cdr (car major)) wildcard))))
        (setq major (cdr major)))
      (nconc (nreverse exact) (nreverse wildcard))))

-- 
 Lloyd Zusman
 ljz@asfast.com

Correction! (Re: Pgnus-0.33: bug in "mailcap-possible-viewers")

[Lloyd Zusman]

Lloyd Zusman <ljz@asfast.com> writes:

> I found a small bug in the `mailcap-possible-viewers' routine in the
> `mailcap.el' file in pgnus-0.33.  The `minor' parameter sometimes gets
> passed in as `nil', and this causes `string-match' to fail.  I'm not
> sure under what conditions this parameter gets set to `nil' ... this
> might have to do with badly formed MIME headers in the message being
> parsed ... ???
> 
> But whatever the case, the enclosed patch seems to fix the problem.
> Use the `-b' option with `patch', since my `diff' program might have
> have messed with the leading white space.

Correction: I mean, you might have to use the `-l' option with `patch'.

> [ ... ] 

-- 
 Lloyd Zusman
 ljz@asfast.com