User loginNavigation |
archivesfoundations for J, APL etci have been playing around with Haskell, and finding ADTs a little *choke* restrictive. it seems every evaluation has to be done in the order of creation. i am wondering if there is a non-functional approach to pattern-matching that can handle abstract operations on arrays and other complex data structures. i am thinking in the direction of APL, but more a language where the APL primitives can be built from scratch. from a scan of his pubs, Barry Jay's pattern calculus might be a good fit. maybe i can do this with Haskell and just don't realise? perhaps 'active patterns' will save me? Alix |
Browse archivesActive forum topics |