Open hdgarrood opened 7 years ago
@paf31 I think we could address this by defining a type alias ->
for Function
in Prim
inside the compiler. I think we might even be able to remove a fair bit of special-casing code that deals with (->)
in e.g. the parser and the pretty-printer if we did this. Does that sound sensible to you?
Sounds good!
Ok great, I'll make an issue in the compiler repo and start looking at that :)
Prim.Function
has started coming out asFunction
rather than(->)
. This might need addressing in the compiler.This isn't actually new behaviour. Pretty-printing of saturated applications of the type constructor
(->)
still work and pretty much always have, but partial applications of it (e.g.instance Category (->)
orinstance Functor ((->) r)
have always been rendered asFunction
.