At 04:16 PM 11/29/99 -0500, [EMAIL PROTECTED] wrote:
>I added this snippet to my prj.el some of you might find it usefull, tho it
>isn't working quite right.
>
>(defun geoffs-jde-tab()
> "Inteligent tabbing in JDE"
> (interactive)
> (let* ((keys (recent-keys)) // this stolen from pc-keys....
> (len (length keys))
> (key1 (if (> len 0) (elt keys (- len 1)) nil))
> (key2 (if (> len 1) (elt keys (- len 2)) nil))
> (key3 (if (> len 2) (elt keys (- len 3)) nil))
> (key-equal-1 (eq key1 key2))
> (key-equal-2 (and key-equal-1 (eq key2 key3))))
> (cond ;(key-equal-2 (call-interactively 'dabbrev-expand))
> (key-equal-1 (call-interactively 'jde-complete-at-point))
> (t (call-interactively 'c-indent-command)))
> )
> )
>(provide 'geoffs-jde-tab)
>(define-key jde-mode-map [tab] 'geoffs-jde-tab)
>
>
>The original idea was that the first time you hit [tab] it would indent,
>the second time would call jde-complete-at-point.
>the third time would call dabbr-expand (or hippie-expand).
>
>For some reason dabbr-expand/hippie-expand seem to work fine the first time
>they are called, but then screw up majorly the second time. It works fine
>when I map hippie-expand to [f3].
>
>I know I'm calling jde-complete-at-point as I get "No completion at this
>point." and "Can't find any declaration..." messages. Unfortunately, I
>can't ever get it to complete anything. (even using C-c,C-v,C-.) Has
>anyone had any luck with jde-complete-at-point?
>
>Any help/ideas would be appreciated.
>
jde-complete-at-point is not very robust in this incarnation. It works only
for compiled classes in jde-global-classpath and as far as I can tell, it's
still not very smart. For example, it doesn't seem to know about inherited
methods and fields. I've gotten it to work for things like
String z = "a"
z.len
which it will complete to z.length()
jde-complete's author, Rodrigo Reyes, and I plan to improve it as time permits.
- Paul
