1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
|
(* If type abbreviations contain type variables that are not present
on the right-hand side we have to reduce the abbreviation before
unification and applying the value restriction.
The first part of the example was provided by Phil Clayton. *)
signature A =
sig
type ('a, 'b) t
val mkT : int -> ('a, int) t
end
structure A :> A =
struct
type ('a, 'b) t = 'b
fun mkT n = n
end
signature B =
sig
type 'a t
val x : ('a t, int) A.t
end
structure B :> B =
struct
type 'a t = unit
val x = A.mkT 2
end;
type 'a t = int;
val x: 'a t = 1 +1;
(* A further example from Phil. *)
signature A =
sig
type ('a, 'b) t
val mkT : int -> ('a, int) t
end
structure A :> A =
struct
type ('a, 'b) t = 'b
fun mkT n = n
end
signature B =
sig
type 'a t
val x : ('a t, int) A.t
end
structure B :> B =
struct
type 'a t1 = unit
type 'a t = 'a t1 * 'a t1 (* two levels of phantom types *)
val x = A.mkT 2
end;
(* A further check. This has worked correctly for a long time because of the
way equality is handled. Include it here just in case it gets broken by
a change in the future. *)
fun f (x: real t) y = x = y;
|
