Re: ffap -vs- RFCs

[Kevin Ryde]

Glenn Morris <address@hidden> writes:
>
> I think it's hard to get ffap to handle things with spaces, so I just
> removed the space from the RFC regexp.

I struck this too and massaged ffap-string-at-point to take a space.
Operative part below, whole thing at
http://www.geocities.com/user42_kevin/ffap/index.html#ffap-rfc-space

If the space in the alist regexp is not actively harmful perhaps it
could be left in.

For this and a couple of other ffap bits I thought it might be nice to
have ffap test some thing-at-point types as a way of extending what it
can recognise at point.  But I never got to any actual code -- it'd
probably for instance want to tie in to ffap-alist so some "things"
could be mode-specific, then there'd be a matter of how to make it work
with ffap-next.


(put 'rfc 'bounds-of-thing-at-point
     (lambda ()
       ;; this regexp is the same as in `ffap-alist'
       (and (thing-at-point-looking-at "[Rr][Ff][Cc][- #]?\\([0-9]+\\)")
            (cons (match-beginning 0) (match-end 0)))))

(defadvice ffap-string-at-point (around ffap-rfc-space activate)
  "Recognise RFCs with a space, like \"RFC 1234\"."
  (or (let ((bounds (bounds-of-thing-at-point 'rfc)))
        (and bounds
             (setq ffap-string-at-point-region (list (car bounds)
                                                     (cdr bounds)))
             (setq ad-return-value (buffer-substring (car bounds)
                                                     (cdr bounds)))))
      ad-do-it))