Hi Mark,
| I was the person who created the ferret mode for emacs. I haven't
done
| anything with it for a while now, but I have started running it
| again. Maybe it is time to revisit the ferret mode and update it.
That
| is an interesting idea about running ferret from within emacs, I am
| not currently aware of how to accomplish that — but it sounds like a
| great feature!
I haven't figured out everything, but here are key pieces which I've
tested and found to work:
;(start-process name-of-the-process buffer-for-output name-of-command)
(start-process "ferret-process" "*ferret-output*" "/usr/local/bin/
ferret")
(defun ferret-send-buffer ()
(interactive)
(process-send-region "ferret-process" 1 (point-max)))
; send the process the entire buffer (from position 1 to the end).
Execute (evaluate) these expressions, go to a buffer that is
displaying a ferret script, then evaluate the following expression
to define the keys
(local-set-key "\C-c\C-c" 'ferret-send-buffer)
and type C-c C-c. In this way, I was able to send Ferret
the contents of the current buffer by C-c C-c.
I'm a novice in emacs lisp and I don't know how to
fill the remaining gaps at the moment:
1) Bind the keys only for Ferret buffers;
2) C-c C-c should start a Ferret process if none
has been started;
3) Different Ferret buffers should use separate
Ferret processes.
and I know there is one non-emacs problem:
4) Ferret cannot handle (as far as I know) long input lines from
the standard input, so that we have to wrap the ferret around
in a small shell script that feeds its standard input to Ferret
through a named pipe "ferret -script namedpipe".
I hope these ideas are useful in extending your Ferret mode.
Regards,
Ryo