Could you export from (match) library match+ and trace-match+

This bug is related to: https://bugs.launchpad.net/ikarus/+bug/188607

Perhaps we can find out what it does! :)

Oh, sorry for the duplication. I am quite new to launchpad, and don't know how to get around :)

match+ is for passing multiple values, to do some paralell accumulation, e.g.:

(define expr '(1 2 (3) (5 (6)) 4))
(define x '()) ;; x must be bound

(match+ (x) expr
   ((,(a*) ...) (values a* a*)) ; the first arg passed to values whould be bound to x, second whould be normally returned just like in match
   (,a (values a a)))

whould return reconstructed list twice (multiple values):

(1 2 (3) (5 (6)) 4)
(1 2 (3) (5 (6)) 4)

hope that helps, I am not very good at explaining how things work :)

Also this can be extended to any number of values:

(match+ (x y z ...) expr clauses ...)

I just had a look.

From what I can see there is a problem matching ",()" .

Can you elaborate, I did not understand what you mean?

Sorry that was more of a note to Aziz.

I did try some simple examples, and as long at they do not make use of the ",()" construct the "+" derivatives seem to work correctly.

I think I found the bug :) Andmap returns #f on empty list, which should be #t.

The correct code (round line 290):

(define (andmap f ls)
  (cond
    ((null? ls) #t) ; was #f here
    ((null? (cdr ls)) (f (car ls)))
    (else (and (f (car ls)) (andmap f (cdr ls))))))

Can you apply the change and verify? (I only tried the collect-leaves sample).

On Nov 25, 2008, at 3:18 AM, leppie wrote:

> I think I found the bug :) Andmap returns #f on empty list, which
> should be #t.

That's correct. My fault. Thanks!
Any other issues with match+ and trace-match+?

Exported and fixed in revision 1692. Thanks for explaining match+! :-)

Thank you too both :)