Instead of having a long, fixed grammar (which would not be very Scheme-like if we were talking about the language itself), I have something like
(define-pattern and <pattern-transformer>) in mind to define, in this case, the semantics of "and" as a derived pattern form. While SRFI 204 was still warm, I started with a design. I have to locate it in the depths of my hard drive. Compare it with SRFI 166; its core is very small and it is highly extensible. I would like to have something like that. (And compare with SRFI 115, which is, as SRFI 204 was, highly monolithic. It should also be extensible in terms of a small set of core matchers.) But before thinking about it further, I should take care of my current SRFIs. :) Am Fr., 11. Nov. 2022 um 05:46 Uhr schrieb John Cowan <[email protected]>: > > SRFI 241 says: "The author of this SRFI suggests that any reviving attempt > of SRFI 204 should make the resulting pattern matcher extensible so that > there are only few primitives with clear semantics allowing the rest of the > specification defined in terms of them." > > What would this extensibility look like?
