Re: [Caml-list] Confusing behaviour of type inference for polymorphic classes.

[Alain Frisch <alain@frisch.fr>]

On 12/02/2013 04:24 PM, Jeremy Yallop wrote:
> There's good news and bad news.  The good news is that the problem can
> be solved for your particular example.  The bad news is that the
> approach doesn't work in the general case.

A more general solution is based on recursive modules:

=======================================================================
type ('a, 'b) t =
     A of 'a * ('b, 'a) t
   | B of 'a

module rec X : sig
   class ['a, 'b, 'ta, 'tb] m : object
     method t : ('a -> 'ta) -> ('b -> 'tb) -> ('a, 'b) t -> ('ta, 'tb) t
   end
end = struct
   class ['a, 'b, 'ta, 'tb] m =
     object
       method t : ('a -> 'ta) -> ('b -> 'tb) -> ('a, 'b) t -> ('ta, 'tb) t =
         fun fa fb s ->
           match s with
           | A (a, bat) -> A (fa a, (new X.m)#t fb fa bat)
           | B  a       -> B (fa a)
     end
end
=======================================================================


This technique also works e.g. for defining mutually recursive classes 
and nominal types.


Alain