Re: [PATCH] Make some local functions public (was: Re: lily-library.scm question)

[Jay Anderson]

On Mon, Jun 22, 2009 at 6:03 PM, Mark Polesky<address@hidden> wrote:
> I know, I know, I just won't let this die. I rewrote this silly
> function yet again (by now it's more of a programming exercise
> than anything else). But the things I've learned from everyone so
> far have found their way into my other LilyPond work, so I don't
> think it's a waste of time.
>
> I discovered 2 procedures from srfi-1 that can work in tandem for
> this: list-index and split-at. I'm left wondering if this is an
> improvement or not. I don't know the implications of use-modules;
> perhaps there's some overhead that I don't know about that isn't
> worth it. Or maybe list-index and split-at are less optimal than
> than Jay's loop-and-reverse solution.
>
> Just wanting to learn...
>
> Thanks,
> Mark
> ___________________________
>
> (use-modules (srfi srfi-1))
>
> (define-public (split-at-predicate predicate lst)
>  "Split a list into 2 lists at the first element that returns #f for
>  (PREDICATE previous_element element). Return the two parts as a pair.
>  Example: (split-at-predicate < '(1 2 3 2 1)) ==> ((1 2 3) . (2 1))"
>  (let ((i (list-index predicate(cdr lst) lst)))
>    (if i
>        (call-with-values (lambda () (split-at lst (1+ i)))
>                          cons)
>        `(,lst))))
> ___________________________
>

Actually, I like this much better. A couple things:
- It doesn't handle an empty list as input. Or is an error the correct behavior?
- I'm not the biggest fan of multiple return values. You could do
(cons (take lst (1+ i)) (drop lst (1+ i))) instead (unless there are
efficiencies in the split-at approach).
- I think (list lst) is clearer than `(,lst).

-----Jay