From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <Pascal.Zimmer@sophia.inria.fr>
Delivered-To: caml-list@yquem.inria.fr
Received: from nez-perce.inria.fr (nez-perce.inria.fr [192.93.2.78])
	by yquem.inria.fr (Postfix) with ESMTP id B1E11BC84
	for <caml-list@yquem.inria.fr>; Mon, 14 Mar 2005 11:31:12 +0100 (CET)
Received: from daimi.au.dk (daimi.au.dk [130.225.16.1])
	by nez-perce.inria.fr (8.13.0/8.13.0) with ESMTP id j2EAVBhQ011879
	(version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO)
	for <caml-list@yquem.inria.fr>; Mon, 14 Mar 2005 11:31:12 +0100
Received: from [10.11.41.210] (arwen [10.11.41.210])
	(authenticated bits=0)
	by daimi.au.dk (8.12.11/8.12.11) with ESMTP id j2EAUvO0001061;
	Mon, 14 Mar 2005 11:30:57 +0100
Message-ID: <423567E1.1030909@sophia.inria.fr>
Date: Mon, 14 Mar 2005 11:30:57 +0100
From: Pascal Zimmer <Pascal.Zimmer@sophia.inria.fr>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7.3) Gecko/20041020
X-Accept-Language: fr, fr-fr, en
MIME-Version: 1.0
To: Oliver Bandel <oliver@first.in-berlin.de>
Cc: caml-list@yquem.inria.fr
Subject: Re: [Caml-list] immutable Strings?
References: <20050312205011.GA2244@first.in-berlin.de>
In-Reply-To: <20050312205011.GA2244@first.in-berlin.de>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit
X-DAIMI-Spam-Score: 0 () 
X-Scanned-By: MIMEDefang 2.44
X-Miltered: at nez-perce with ID 423567EF.001 by Joe's j-chkmail (http://j-chkmail.ensmp.fr)!
X-Spam: no; 0.00; caml-list:01 oliver:01 bandel:01 mutable:01 mutable:01 ocaml:01 datatype:01 sig:01 val:01 val:01 char:01 struct:01 abstr:01 char:01 ...:98 
X-Spam-Checker-Version: SpamAssassin 3.0.2 (2004-11-16) on yquem.inria.fr
X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=disabled 
	version=3.0.2
X-Spam-Level: 

Oliver Bandel wrote:
> for records it is possible to say "mutable" to change a normally
> non mutable value into a mutable one.
> 
> It would be nice to have the possibility to turn on
> immutability fpr strings with a keyword like "immutable"
> or so.
> 
> So it could be forbidden to modify strings in cases,
> where it makes sense; otherwise it must be provided
> String.copy at a lot of cases in a program, that
> want's to forbid modyfiing the original data.
> 
> And disallowing a modification is much stronger/more strict
> than only allowing modyfiing a copy.
> So, it would be nice to have such a feature in newer versions
> of OCaml.

You can achieve the same effect quite easily using modules to provide a 
new interface to String, where you restrict the set of possible 
operations and abstract the datatype:

module type IMMSTRING =
  sig
   type t    (* abstract type *)

   val create : string -> t
   val copy : t -> t
   val copy_to_string : t -> string

   val get : t -> int -> char
   val cat : t -> t -> t
   val print_string : t -> unit

   (* ... *)
  end;;

module ImmString : IMMSTRING =
  struct
   type t = string

   let create = String.copy
   let copy = String.copy
   let copy_to_string = String.copy

   let get s i = s.[i]
   let cat s t = s ^ t
   let print_string = print_string

   (* ... *)
  end;;

Now you can write:

# let s = ImmString.create "abcde";;
val s : ImmString.t = <abstr>
# ImmString.get s 0;;
- : char = 'a'
# s.[0] <- 'f';;
This expression has type ImmString.t but is here used with type string

The bad points:
- you cannot use the shortcuts s.[i] and (^) anymore (this gets the code 
less readable especially for the first one)
- you have to make a copy when creating an immutable string; if you were 
not planning to keep the original string to modify it, this copy is in 
fact useless

Pascal