Hello,



we have noticed a strange (at least to us) behavior of the coercion
operator over phantom types.

Here is an example:

# type n = [`Pos];;   (* non-negative number *)
type n = [ `Pos]
# type z = [`Neg | `Pos];;   (* integer *)
type z = [ `Neg | `Pos]
# type (+'ph) expr = Expr of int;;  (* phantom type *)
type 'a expr = Expr of int

# let x = ((Expr 1) : n expr);;
val x : n expr = Expr 1
# (x:> z expr);;          (* okay *)
- : z expr = Expr 1
# let y  = ((Expr 1) : z expr);;  
val y : z expr = Expr 1
# (y :> n expr);;        (* ??? *)
- : n expr = Expr 1

So, we observe the same behavior whatever the variance specification is.


# type 'ph expr = Expr of int;;  (* without variance *)
type 'a expr = Expr of int
# let x = ((Expr 1) : n expr);;
val x : n expr = Expr 1
# (x:> z expr);;      (* ??? *)
- : z expr = Expr 1
# let y  = ((Expr 1) : z expr);;
val y : z expr = Expr 1
# (y :> n expr);;       (* ??? *)
- : n expr = Expr 1

# type (-'ph) expr = Expr of int;;  (* contravariance *)
type 'a expr = Expr of int
# let x = ((Expr 1) : n expr);;
val x : n expr = Expr 1
# (x:> z expr);;
- : z expr = Expr 1
# let y  = ((Expr 1) : z expr);;
val y : z expr = Expr 1
# (y :> n expr);;
- : n expr = Expr 1


We wonder whether it is a bug or not (anyway, this is rather
counter-intuitive)...?

dc


-- David Chemouil    [mailto:david.chemouil@irit.fr]
--
-- Zeno group              [http://www.irit.fr/zeno]
-- Institut de recherche en informatique de Toulouse
-------------------
Bug reports: http://caml.inria.fr/bin/caml-bugs  FAQ: http://caml.inria.fr/FAQ/
To unsubscribe, mail caml-list-request@inria.fr  Archives: http://caml.inria.fr