[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [r6rs-discuss] [Scheme-reports] Scheme pattern matching: the case for (case)

On Tue, 2010-12-21 at 15:30 -0500, John Cowan wrote:
> > eqv? is immaterial here:
> > 
> > (let ([eqv? equal?]) (case "asd" (("asd") #t)))
> That does not mean what you think it means, for two reasons.  One is that
> with a proper hygienic macro system, rebinding eqv? does not affect any
> uses of eqv? in the expansion of the case macro.  The second reason is
> that implementations are free to make literal strings eqv? if they have
> the same content.

Right. However, this still is an underspecified corner of Scheme. Even
in R6RS I don't see a mandate to always implement (case) via hygienic 
macros (and have the system solve the rebinding). And neither do
implementors, in my experience.

> > Not necessarily so. One can consider a (case) pattern to be implicitly 
> > quasiquoted (just like traditional case is implicitly quoted). So, if
> > one needs "extended" (case) capabilities, one would write:
> > 
> > (case 'b
> >   ((,a) (list a)))
> Quoting, alas, is incompatible with quasiquoting.  (,a) is just syntax
> sugar for ((unquote a)), which is a perfectly legitimate argument for
> a case arm that matches on a list eqv? to it.  Somewhat unlikely that
> you will find such a list, but that's still what it means.

I have no quoting in my example clauses. Read it as follows:

(case 'b
   ((`,a) (list a))) ; explicate quasiquote

BTW, multi-clause case would be:

(case 'b
   ((`x `y `,a) (list 'foo))) ; no a in body due to disjunction

or if you prefer (still matching either 'x, 'y or anything):

(case 'b
   (`(x y ,a) (list 'foo))) ; still no a

Obviously, one can elide explicit quasiquotes around each alternative,
giving back the degenerate semantics of (case). Oleg's matcher does this
also, but for a different reason (to delineate pattern variables).

> > From my perspective, (case) is a degenerate case of (match) and should 
> > be merged - we don't need two switch constructs in a simple and small 
> > language.
> Note that this is a WG2 proposal, ergo not the "simple and small language".

I hope this WG shall not try make it more complex and large that it
needs to be.


r6rs-discuss mailing list