<div>Sorry I was not clear.  I want to specify a function like map, where the arity of the mapping function argument is dependent on the rest of the arguments to map. Your solution does that. But I also want to specify the return type of the mapping function. I did not see a way to describe the mapping function return type that does not also restrict the way I can describe its arguments. </div> <div> </div> <div>More concretly, consider:</div> <div> (define (build/join-strings mapping-f separator . lsts)         (foldl (lambda (st so-far)                  (format "~a~a~a"                          so-far                          (if (equal? so-far "")                              ""                              separator)                          st))                ""                (apply map (cons mapping-f lsts))))) I want to state that mapping-f has the same arity has the rest arguments to build/join-strings which your solution does, but I also want to say that mapping-f must return a string. -pp  </div> <div> </div> <div>  </div> <div>On 1/24/06, Robby Findler <<a href="mailto:robby@cs.uchicago.edu">robby@cs.uchicago.edu</a>> wrote: <blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">You have to pin down the arity enough that users of your function know what the legal argument lengths are. (Is that what you're asking?) Robby At Mon, 23 Jan 2006 23:58:08 +0000, pedro pinto wrote: >  One more question, how would I go about saying something about the return > type of proc/func? From my understanding of the docs I need to write a > function contract, as opposed to what I think is a flat-contract like > arity-matches-lists, but I did not see a way to write one that does not also > pin down the arity. > > -pp > > > > >  On 1/23/06, Robby Findler <<a href="mailto:robby@cs.uchicago.edu">robby@cs.uchicago.edu</a>> wrote: > > > > > > You have to use the more complex combinators to express that. Here's > > > how ->r can be used to express map's contract: > > > > > > (module map-c mzscheme > > > (require (lib "contract.ss")) > > > > > > (define map-c map) > > > > > > (provide/contract > > >   [map-c > > >    (->r ([func > > >           (and/c > > >            procedure? > > >            (let ([arity-matches-lists > > >                   (ë (f) > > >                     (procedure-arity-includes? f (+ 1 (length > > > rest-args))))]) > > >              arity-matches-lists))] > > >          [lst-one (listof any/c)]) > > >         rest-args > > >         (and/c (listof (listof any/c)) > > >                (let ([same-length-lists (ë (x) (apply equal? (length > > > lst-one) (map length x)))]) > > >                  same-length-lists)) > > >         (listof any/c))])) > > > > > > (require map-c) > > > > > > (define (print-msg exn) > > > (display (exn-message exn)) > > > (newline)) > > > > > > (map-c (lambda (x y) x) (list 1 2 3) (list 4 5 6)) > > > (with-handlers ([values print-msg]) > > > (map-c (lambda (x) x) (list 1 2 3) (list 4 5 6))) > > > (with-handlers ([values print-msg]) > > > (map-c (lambda (x y) x) (list 1 2 3) (list 4 5 6 7))) > > > > > > At Mon, 23 Jan 2006 19:49:34 +0000, pedro pinto wrote: > > > > Hi there, > > > > > > > > I just started fumbling around with contracts and I was trying to > > > write a > > > > contract for a function similar to map: > > > > > > > > map proc clist1 clist2 ... -> list > > > > > > > > Is there a way to say that proc is a procedure whose arity should > > > match the > > > > number of clist arguments? > > > > > > > > TIA, > > > > -pp > > > > _________________________________________________ > > > >   For list-related administrative tasks: > > > >   <a href="http://list.cs.brown.edu/mailman/listinfo/plt-scheme">http://list.cs.brown.edu/mailman/listinfo/plt-scheme</a> > > > > > > > </blockquote></div>