From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from mail2-relais-roc.national.inria.fr (mail2-relais-roc.national.inria.fr [192.134.164.83])
	by walapai.inria.fr (8.13.6/8.13.6) with ESMTP id p54Fgoou017971
	for <caml-list@sympa-roc.inria.fr>; Sat, 4 Jun 2011 17:42:50 +0200
X-IronPort-Anti-Spam-Filtered: true
X-IronPort-Anti-Spam-Result: ApACAH5R6k3UTWUIi2dsb2JhbAAyCxaESqF8FAEBAQoLCwcSBiG4EIJ6AY0eBYErg2yBCpU3iwI
X-IronPort-AV: E=Sophos;i="4.65,320,1304287200"; 
   d="scan'208";a="96055990"
Received: from mx4.wp.pl ([212.77.101.8])
  by mail2-smtp-roc.national.inria.fr with ESMTP/TLS/DHE-RSA-AES256-SHA; 04 Jun 2011 17:42:45 +0200
Received: (wp-smtpd smtp.wp.pl 22196 invoked from network); 4 Jun 2011 17:42:43 +0200
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=wp.pl; s=1024a;
          t=1307202163; bh=WHoQcX5BBJgVB9IdzCWfrmR+MQDgiFvkizgFWNDZ7YQ=;
          h=From:To:Subject;
          b=gEJXaZPfyLANYhUu+4z2q5o4HUf3xfKQsJ2luXhErqfr4I+zK/etXzSbfoWBBGoiZ
           h75LZc0sHernSzE6UNItoPjD9xA6eyVBIUs5HdtBDbP6wgvxPAUft/2hmvxWQj3w+d
           bqE6yctMG9uAZBo1djIrSfT7PbfVwM/N5e4zCyGs=
Received: from 18-161.2-0.pl (HELO [192.168.1.101]) (d0@[91.189.18.161])
          (envelope-sender <d0@wp.pl>)
          by smtp.wp.pl (WP-SMTPD) with AES256-SHA encrypted SMTP
          for <caml-list@yquem.inria.fr>; 4 Jun 2011 17:42:43 +0200
Message-ID: <4DEA5272.9040905@wp.pl>
Date: Sat, 04 Jun 2011 17:42:42 +0200
From: Dawid Toton <d0@wp.pl>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.2.17) Gecko/20110424 Thunderbird/3.1.10
MIME-Version: 1.0
To: caml-list <caml-list@yquem.inria.fr>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-WP-AV: skaner antywirusowy poczty Wirtualnej Polski S. A.
X-WP-SPAM: NO 0000000 [4eOU]                               
Subject: [Caml-list] Constrain module type to reuse a sum type

I have a functor F which itself applies its own arguments to some other 
functors. Results R of these applications contain sum types. These R are 
exposed in the outcome of F in two ways: as types carried by the 
functions defined by F and all R packed into a wrapper module. I'd like 
to tell the compiler that each sum type is equal in both contexts.

The problem boils down to the following:

B.ml:

module type A = sig type a = A end
module A = struct type a = A end

module E = A

type want_equal = A.a as 'a constraint 'a = E.a

B.mli:

module type A = sig type a = A end
module A : A

module type E = A (*with type a = A.a*)
module E : E

type want_equal = A.a as 'a constraint 'a = E.a


There are no problems with the module implementation. I can see right 
types in the output of ocamlc -i .
But the interface is wrong, since it is forgotten that the sum types in 
E and A are the same.

I have put a constraint in a comment - this is what I'm trying to achieve.

In my real case, the module type A is a big signature that comes from 
somewhere else. The mentioned functor F should require as little changes 
as possible for each change of the module type that corresponds to the 
module type A in the above example.

This is why I'm looking for a way to constrain A to obtain E. How to do 
this?
I would be grateful for any comments.
If the code is simplified too much, I can post something more similar to 
my real code.

Dawid