WARNING: This server is unstable and will be retired in the next days. If you want to keep this forum available, please request immediately a migration on the Nabble Support forum. Forums that don't receive any migration request will be deleted forever.

Implementation for a (nearly) typesafe shallow option type and a compiler bug?

View: New views
5 Messages — Rating Filter:   Alert me  

Implementation for a (nearly) typesafe shallow option type and a compiler bug?

by Goswin von Brederlow-2 :: Rate this Message:

| View Threaded | Show Only this Message

Hi,

as mentioned earlier I wanted to have a

    type 'a shallow = NULL | 'a

that is like an 'a option but without the indirection.

The first idea was that no ocaml value can be the C NULL pointer (usualy
all bits set to 0) and the C NULL pointer points outside the ocaml heap
so the GC is fine with that too. So NULL could be encoded as C NULL
pointer and 'a used as is.

First problem is that this does not work for an 'a shallow shallow. And
there is no way to constrain 'a to not be a 'b shallow in the type
definition.

Second problem is that someone else might define a 'a shallow2 type with
the same idea and 'a shallow2 shallow would also not work. And with
abstract types it would be impossible to test for that even if ocaml
would allow negative constraints.


The common problem here is that NULL is not unique for each type. So I
thought of a way to make it unique. What I came up with is using a
functor to create a unique NULL value for each instance. Source code at
the end.

Output:
NULL -> None
Some 5 -> Some 5
NULL -> None
Some NULL -> Some None
Some Some 5 -> Some Some 5
1 2 3 4
a b c d
NULL -> Some 70221466564508

As you can see from the output this solves the problem of 'a shallow
shallow and would also solve the 'a shallow2 shallow case.

But in the last line a new problem appears. Each instance of the functor
has a unique NULL element. But instances of the functor with the same
type are compatible. So an IntShallow2.t can be used instead of an
IntShallow.t but then the NULL does not match.

To me this looks like a bug in ocaml because the resulting type of the
functor application does not match the type I specified for the functor:

  module type SHALLOW =
    sig
      type -'a t
      val null : 'a t
      val make : 'a -> 'a t
      val as_option : 'a t -> 'a option
    end
  module Make : functor (Type : TYPE) -> SHALLOW

module IntShallow2 :
  sig
    type 'a t = 'a Shallow.Make(Int).t
    val null : 'a t
    val make : 'a -> 'a t
    val as_option : 'a t -> 'a option
  end

The type I specified has 'a t abstract while the IntShallow2 has the
type 'a t more concret.

Restricting the type to what it should already be results in the correct
error:

module IntShallow2 = (Shallow.Make(Int) : Shallow.SHALLOW)
let () =
  Printf.printf "NULL -> %s\n" (to_string (IntShallow2.null))

File "shallow.ml", line 91, characters 42-60:
Error: This expression has type 'a IntShallow2.t
       but an expression was expected of type
         int IntShallow.t = int Shallow.Make(Int).t

Why is that?

MfG
        Goswin

======================================================================
module Shallow : sig
  module type TYPE = sig type 'a t end
  module type SHALLOW =
    sig
      type -'a t
      val null : 'a t
      val make : 'a -> 'a t
      val as_option : 'a t -> 'a option
    end
  module type SHALLOWFUNCTOR = functor (Type : TYPE) -> SHALLOW
  module Make : functor (Type : TYPE) -> SHALLOW
end = struct
  module type TYPE = sig type 'a t end
  module type SHALLOW =
    sig
      type -'a t
      val null : 'a t
      val make : 'a -> 'a t
      val as_option : 'a t -> 'a option
    end
  module type SHALLOWFUNCTOR = functor (Type : TYPE) -> SHALLOW

  module Make_intern =
    functor (Type : TYPE) ->
      struct
        type -'a t
        (* Dummy object unique to the type *)
        let null = Obj.magic (ref 0)
        let make x = Obj.magic x
        let as_option x = if x == null then None else Some (Obj.magic x)
      end
  module Make = (Make_intern :  functor (Type : TYPE) -> SHALLOW)
end

module Int = struct type 'a t = int end
 
module IntShallow = Shallow.Make(Int)
module IntShallowShallow = Shallow.Make(IntShallow)

let to_string x =
  match IntShallow.as_option x with
    | None -> "None"
    | Some x -> Printf.sprintf "Some %d" x

let to_string2 x =
  match IntShallowShallow.as_option x with
    | None -> "None"
    | Some x -> Printf.sprintf "Some %s" (to_string x)

module List = struct
  module Item = struct
    type 'a t = 'a
  end

  module Shallow = Shallow.Make(Item)

  type 'a item = { mutable next : 'a list; data : 'a; }
  and 'a list = 'a item Shallow.t

  let null = Shallow.null
  let cons x y = Shallow.make { next = y; data = x; }
  let rec iter fn x =
    match Shallow.as_option x with
      | None -> ()
      | Some { next; data; } -> fn data; iter fn next
end

let () =
  Printf.printf "NULL -> %s\n" (to_string (IntShallow.null));
  Printf.printf "Some 5 -> %s\n" (to_string (IntShallow.make 5));
  Printf.printf "NULL -> %s\n" (to_string2 (IntShallowShallow.null));
  Printf.printf "Some NULL -> %s\n"
    (to_string2 (IntShallowShallow.make IntShallow.null));
  Printf.printf "Some Some 5 -> %s\n"
    (to_string2 (IntShallowShallow.make (IntShallow.make 5)));
  let x = List.null in
  let x = List.cons 4 x in
  let x = List.cons 3 x in
  let x = List.cons 2 x in
  let x = List.cons 1 x in
  List.iter (Printf.printf "%d ") x;
  print_newline ();
  let y = List.null in
  let y = List.cons "d" y in
  let y = List.cons "c" y in
  let y = List.cons "b" y in
  let y = List.cons "a" y in
  List.iter (Printf.printf "%s ") y;
  print_newline ()

module IntShallow2 = Shallow.Make(Int)

let () =
  Printf.printf "NULL -> %s\n" (to_string (IntShallow2.null))

--
Caml-list mailing list.  Subscription management and archives:
https://sympa-roc.inria.fr/wws/info/caml-list
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
Bug reports: http://caml.inria.fr/bin/caml-bugs


Re: Implementation for a (nearly) typesafe shallow option type and a compiler bug?

by Gabriel Scherer :: Rate this Message:

| View Threaded | Show Only this Message

What you observe is the so-called "strengthening" of type equalities
in functor applications. See papers 5 or 6 in this list:
  http://caml.inria.fr/about/papers.en.html

It is not a bug, but a feature: you can write functors F such that
applying F(X) twice yields compatible, rather than incompatible,
types. If you want to recover incompatible types, you can seal the
functor result as you did in your workaround, or pass a non-path
functor expression (that behave in a more generative way): F(struct
include X end).

On your more general code:
- I do not understand why you specify the abstract type ('a t) to be
contravariant, and I suspect this will be unsound (is an (< m : int >
t) also an (< m : int; s : string > t)?)
- I am not sure using (Obj.magic (ref 0)) is safe wrt. types whose
representation is not always a pointer (ie. floats)

On Sun, May 6, 2012 at 2:53 PM, Goswin von Brederlow <goswin-v-b@...> wrote:

> Hi,
>
> as mentioned earlier I wanted to have a
>
>    type 'a shallow = NULL | 'a
>
> that is like an 'a option but without the indirection.
>
> The first idea was that no ocaml value can be the C NULL pointer (usualy
> all bits set to 0) and the C NULL pointer points outside the ocaml heap
> so the GC is fine with that too. So NULL could be encoded as C NULL
> pointer and 'a used as is.
>
> First problem is that this does not work for an 'a shallow shallow. And
> there is no way to constrain 'a to not be a 'b shallow in the type
> definition.
>
> Second problem is that someone else might define a 'a shallow2 type with
> the same idea and 'a shallow2 shallow would also not work. And with
> abstract types it would be impossible to test for that even if ocaml
> would allow negative constraints.
>
>
> The common problem here is that NULL is not unique for each type. So I
> thought of a way to make it unique. What I came up with is using a
> functor to create a unique NULL value for each instance. Source code at
> the end.
>
> Output:
> NULL -> None
> Some 5 -> Some 5
> NULL -> None
> Some NULL -> Some None
> Some Some 5 -> Some Some 5
> 1 2 3 4
> a b c d
> NULL -> Some 70221466564508
>
> As you can see from the output this solves the problem of 'a shallow
> shallow and would also solve the 'a shallow2 shallow case.
>
> But in the last line a new problem appears. Each instance of the functor
> has a unique NULL element. But instances of the functor with the same
> type are compatible. So an IntShallow2.t can be used instead of an
> IntShallow.t but then the NULL does not match.
>
> To me this looks like a bug in ocaml because the resulting type of the
> functor application does not match the type I specified for the functor:
>
>  module type SHALLOW =
>    sig
>      type -'a t
>      val null : 'a t
>      val make : 'a -> 'a t
>      val as_option : 'a t -> 'a option
>    end
>  module Make : functor (Type : TYPE) -> SHALLOW
>
> module IntShallow2 :
>  sig
>    type 'a t = 'a Shallow.Make(Int).t
>    val null : 'a t
>    val make : 'a -> 'a t
>    val as_option : 'a t -> 'a option
>  end
>
> The type I specified has 'a t abstract while the IntShallow2 has the
> type 'a t more concret.
>
> Restricting the type to what it should already be results in the correct
> error:
>
> module IntShallow2 = (Shallow.Make(Int) : Shallow.SHALLOW)
> let () =
>  Printf.printf "NULL -> %s\n" (to_string (IntShallow2.null))
>
> File "shallow.ml", line 91, characters 42-60:
> Error: This expression has type 'a IntShallow2.t
>       but an expression was expected of type
>         int IntShallow.t = int Shallow.Make(Int).t
>
> Why is that?
>
> MfG
>        Goswin
>
> ======================================================================
> module Shallow : sig
>  module type TYPE = sig type 'a t end
>  module type SHALLOW =
>    sig
>      type -'a t
>      val null : 'a t
>      val make : 'a -> 'a t
>      val as_option : 'a t -> 'a option
>    end
>  module type SHALLOWFUNCTOR = functor (Type : TYPE) -> SHALLOW
>  module Make : functor (Type : TYPE) -> SHALLOW
> end = struct
>  module type TYPE = sig type 'a t end
>  module type SHALLOW =
>    sig
>      type -'a t
>      val null : 'a t
>      val make : 'a -> 'a t
>      val as_option : 'a t -> 'a option
>    end
>  module type SHALLOWFUNCTOR = functor (Type : TYPE) -> SHALLOW
>
>  module Make_intern =
>    functor (Type : TYPE) ->
>      struct
>        type -'a t
>        (* Dummy object unique to the type *)
>        let null = Obj.magic (ref 0)
>        let make x = Obj.magic x
>        let as_option x = if x == null then None else Some (Obj.magic x)
>      end
>  module Make = (Make_intern :  functor (Type : TYPE) -> SHALLOW)
> end
>
> module Int = struct type 'a t = int end
>
> module IntShallow = Shallow.Make(Int)
> module IntShallowShallow = Shallow.Make(IntShallow)
>
> let to_string x =
>  match IntShallow.as_option x with
>    | None -> "None"
>    | Some x -> Printf.sprintf "Some %d" x
>
> let to_string2 x =
>  match IntShallowShallow.as_option x with
>    | None -> "None"
>    | Some x -> Printf.sprintf "Some %s" (to_string x)
>
> module List = struct
>  module Item = struct
>    type 'a t = 'a
>  end
>
>  module Shallow = Shallow.Make(Item)
>
>  type 'a item = { mutable next : 'a list; data : 'a; }
>  and 'a list = 'a item Shallow.t
>
>  let null = Shallow.null
>  let cons x y = Shallow.make { next = y; data = x; }
>  let rec iter fn x =
>    match Shallow.as_option x with
>      | None -> ()
>      | Some { next; data; } -> fn data; iter fn next
> end
>
> let () =
>  Printf.printf "NULL -> %s\n" (to_string (IntShallow.null));
>  Printf.printf "Some 5 -> %s\n" (to_string (IntShallow.make 5));
>  Printf.printf "NULL -> %s\n" (to_string2 (IntShallowShallow.null));
>  Printf.printf "Some NULL -> %s\n"
>    (to_string2 (IntShallowShallow.make IntShallow.null));
>  Printf.printf "Some Some 5 -> %s\n"
>    (to_string2 (IntShallowShallow.make (IntShallow.make 5)));
>  let x = List.null in
>  let x = List.cons 4 x in
>  let x = List.cons 3 x in
>  let x = List.cons 2 x in
>  let x = List.cons 1 x in
>  List.iter (Printf.printf "%d ") x;
>  print_newline ();
>  let y = List.null in
>  let y = List.cons "d" y in
>  let y = List.cons "c" y in
>  let y = List.cons "b" y in
>  let y = List.cons "a" y in
>  List.iter (Printf.printf "%s ") y;
>  print_newline ()
>
> module IntShallow2 = Shallow.Make(Int)
>
> let () =
>  Printf.printf "NULL -> %s\n" (to_string (IntShallow2.null))
>
> --
> Caml-list mailing list.  Subscription management and archives:
> https://sympa-roc.inria.fr/wws/info/caml-list
> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> Bug reports: http://caml.inria.fr/bin/caml-bugs
>


--
Caml-list mailing list.  Subscription management and archives:
https://sympa-roc.inria.fr/wws/info/caml-list
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
Bug reports: http://caml.inria.fr/bin/caml-bugs


Re: Implementation for a (nearly) typesafe shallow option type and a compiler bug?

by Goswin von Brederlow-2 :: Rate this Message:

| View Threaded | Show Only this Message

Gabriel Scherer <gabriel.scherer@...> writes:

> What you observe is the so-called "strengthening" of type equalities
> in functor applications. See papers 5 or 6 in this list:
>   http://caml.inria.fr/about/papers.en.html
>
> It is not a bug, but a feature: you can write functors F such that
> applying F(X) twice yields compatible, rather than incompatible,
> types. If you want to recover incompatible types, you can seal the
> functor result as you did in your workaround, or pass a non-path
> functor expression (that behave in a more generative way): F(struct
> include X end).

Good to know. I found that surprising. I think it is bad that you can't
specify the type of the functor so that both compatible and incompatible
types would be an option. Just like you can use 'a, +'a and -'a to fine
tune variance in types there could be some syntax to make the functor
type strengthened or not.

> On your more general code:
> - I do not understand why you specify the abstract type ('a t) to be
> contravariant, and I suspect this will be unsound (is an (< m : int >
> t) also an (< m : int; s : string > t)?)

Left over from trying to make the functor type not strengthened.

> - I am not sure using (Obj.magic (ref 0)) is safe wrt. types whose
> representation is not always a pointer (ie. floats)

(Obj.magic (ref 0)) is always a pointer that is unique to the instance
of the functor. No other value can legally have this bit pattern.

As for floats: Manual 18.3.1 Atomic types

Caml type      Encoding
float          Blocks with tag Double_tag.

A float is always a pointer and that can't be legally pointing to our
(Obj.magic (ref 0)).

MfG
        Goswin

--
Caml-list mailing list.  Subscription management and archives:
https://sympa-roc.inria.fr/wws/info/caml-list
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
Bug reports: http://caml.inria.fr/bin/caml-bugs


Re: Implementation for a (nearly) typesafe shallow option type and a compiler bug?

by Richard W.M. Jones-3 :: Rate this Message:

| View Threaded | Show Only this Message

On Sun, May 06, 2012 at 02:53:14PM +0200, Goswin von Brederlow wrote:
> The common problem here is that NULL is not unique for each type. So I
> thought of a way to make it unique.

I don't know whether or not it'll solve your problem, but there is
another way to generate unique pointers that the OCaml runtime won't
follow.  That is, in C create a pointer to some static item:

  static char foo;
  value get_null (value ignored) { return &foo; }

  external get_null : unit -> null = "get_null" "noalloc"

Rich.

--
Richard Jones
Red Hat

--
Caml-list mailing list.  Subscription management and archives:
https://sympa-roc.inria.fr/wws/info/caml-list
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
Bug reports: http://caml.inria.fr/bin/caml-bugs


Re: Implementation for a (nearly) typesafe shallow option type and a compiler bug?

by Richard W.M. Jones-3 :: Rate this Message:

| View Threaded | Show Only this Message

OK!  I guess I should read back over the earlier thread first ...

Rich.

--
Richard Jones
Red Hat

--
Caml-list mailing list.  Subscription management and archives:
https://sympa-roc.inria.fr/wws/info/caml-list
Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
Bug reports: http://caml.inria.fr/bin/caml-bugs