From mboxrd@z Thu Jan  1 00:00:00 1970
Received: (from weis@localhost) by pauillac.inria.fr (8.7.6/8.7.3) id UAA22538 for caml-redistribution; Sat, 10 Jul 1999 20:05:35 +0200 (MET DST)
Received: from nez-perce.inria.fr (nez-perce.inria.fr [192.93.2.78]) by pauillac.inria.fr (8.7.6/8.7.3) with ESMTP id OAA27536 for <caml-list@pauillac.inria.fr>; Fri, 9 Jul 1999 14:51:53 +0200 (MET DST)
Received: from mail3.microsoft.com (mail3.microsoft.com [131.107.3.123])
	by nez-perce.inria.fr (8.8.7/8.8.7) with SMTP id OAA19264
	for <caml-list@inria.fr>; Fri, 9 Jul 1999 14:51:52 +0200 (MET DST)
Received: from 157.54.9.100 by mail3.microsoft.com (InterScan E-Mail VirusWall NT); Fri, 09 Jul 1999 05:50:32 -0700 (Pacific Daylight Time)
Received: by INET-IMC-03 with Internet Mail Service (5.5.2524.0)
	id <3NFXCM5K>; Fri, 9 Jul 1999 05:50:32 -0700
Message-ID: <39ADCF833E74D111A2D700805F1951EF0F00BCB6@RED-MSG-06>
From: Don Syme <dsyme@microsoft.com>
To: caml-list@inria.fr
Subject: RE: parametric types in simple functors?
Date: Fri, 9 Jul 1999 05:50:25 -0700 
X-Mailer: Internet Mail Service (5.5.2524.0)
Sender: weis

Yes, I had the same problem just yesterday when wanting to generate the
natural, parametric version of "Map" using Pervasives.compare.  In the end I
just copied to code in stdlib.map, substituting "Pervasives" for "Ord".  Is
there some trick we're overlooking?
Don


-----Original Message-----
From: STARYNKEVITCH Basile [mailto:Basile.Starynkevitch@cea.fr]
Sent: 05 July 1999 11:57
To: caml-list@inria.fr
Subject: Q: parametric types in simple functors?



The subject title is poor. I have a concrete simple question.

(All CAML code is indented to the right here)
Consider

   type 'a symbol = { sy_name: string; sy_hash: int; mutable sy_prop: 'a
option };;

This is essentially a symbol, memoizing its hash code for fast
access. Some structures are naturally hashed map, whose key are
symbols (eg bindings). We make symbols with

  let get_symbol n =
    try Hashtbl.find symdict n 
    with Not_found ->
     let s = { sy_name=n; sy_hash=Hashtbl.hash n; sy_prop=None } in
      Hashtbl.add symdict n s;
      s
  ;;


The property field of symbols could be used the way you like it.

I want to define a binding as an hashtable mapping symbols to (say)
integers - or even any type 'b. But I can't code:

  module SymbolHash = Hashtbl.Make(
     struct 
       type t = symbol
       let equal s1 s2 = s1 == s2
       let hash {sy_hash=h} = h
     end)
;;

Because the type constructor t has no argument in the
Hashtbl.HashedType signature.

Any comments or suggestions (could define a class abstract_symbol
with a subclass symbol holding the properties. But I don't want
classes because I don't need inheritance).


The intent is that the symbol properties would be defined in another
module. I don't want to fix the 'a in symbol_t here, yet I want to
make an hashtable whose keys are symbols.

In other words, 

  why the stdlib doesn't give parametrized types for many modules
  (Hashtbl, Set...)

  when should I use a type constructor with an argument, or a functor?


Any comments?


N.B. Any opinions expressed here are only mine, and not of my organization.
N.B. Les opinions exprimees ici me sont personnelles et n engagent pas le
CEA.

---------------------------------------------------------------------
Basile STARYNKEVITCH   ----  Commissariat à l Energie Atomique 
DTA/LETI/DEIN/SLA * CEA/Saclay b.528 (p111f) * 91191 GIF/YVETTE CEDEX *
France
phone: 1,69.08.60.55; fax: 1.69.08.83.95 home: 1,46.65.45.53
email: Basile point Starynkevitch at cea point fr