caml-list - the Caml user's mailing list
 help / color / mirror / Atom feed
From: Pierre Weis <Pierre.Weis@inria.fr>
To: orodeh@cs.huji.ac.il (Ohad Rodeh)
Cc: caml-list@inria.fr
Subject: Re: Ref polymorphism restriction
Date: Mon, 25 Dec 2000 23:09:40 +0100 (MET)	[thread overview]
Message-ID: <200012252209.XAA11293@pauillac.inria.fr> (raw)
In-Reply-To: <Pine.LNX.4.10.10012250930420.7348-100000@zigzag.cs.huji.ac.il> from Ohad Rodeh at "Dec 25, 100 09:34:13 am"

> List,
>   The following program does not compile, with a 
> rather strange error message. 
> 
> module type SIG = 
>   sig
>     val l : '_a list ref
>     val f : '_a -> unit
>   end

Warning, there is a strange feature here that makes your '_a as type
schemes paramters (generalized variables) and not type variables:

        Objective Caml version 3.00

#  module type SIG = sig
     val l : '_a list ref
     val f : '_a -> unit
   end;;
module type SIG = sig val l : 'a list ref val f : 'a -> unit end
#

In effect 'ident means a type scheme parameters. When we introduced
type variables, we denote them by an _ prefix, that was not allowed in
identifiers (as the first character) at that time. Now, identifiers
are allowed to start with an _. Hence '_foo is a perfectly valid
polymorphic type parameter!

> module M : SIG = 
>   struct
>     let l = ref [] 
>       
>     let f a = l:= a::!l
>   end
> 
> 
> Modules do not match:
>   sig val l : '_a list ref val f : '_a -> unit end
> is not included in
>   SIG
> Values do not match:
>   val l : '_a list ref
> is not included in
>   val l : 'a list ref
> 
> Note that the signature for list l IS '_a list ref. 

As mentioned above, no: it is 'a list ref. This explains the
typechecker message.

> I suppose this is due to the restriction on polymorphic references. 
> Is the OCaml restriction such that references may only be local to
> a single module? Is there a different restriction? 
> 
> 	Ohad.
> 
> ----------------------------------------------------------
> orodeh@cs.huji.ac.il
> www.cs.huji.ac.il/~orodeh

Anyway, your right: type variables ('_a types) should not escape the
scope of a module.

Merry Christmas,

Pierre Weis

INRIA, Projet Cristal, Pierre.Weis@inria.fr, http://pauillac.inria.fr/~weis/




      reply	other threads:[~2000-12-25 22:09 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2000-12-25  7:34 Ohad Rodeh
2000-12-25 22:09 ` Pierre Weis [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=200012252209.XAA11293@pauillac.inria.fr \
    --to=pierre.weis@inria.fr \
    --cc=caml-list@inria.fr \
    --cc=orodeh@cs.huji.ac.il \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).