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 SAA02022 for caml-redistribution; Fri, 14 May 1999 18:23:53 +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 SAA20691 for <caml-list@pauillac.inria.fr>; Fri, 14 May 1999 18:14:40 +0200 (MET DST)
Received: from mail2.microsoft.com (mail2.microsoft.com [131.107.3.124])
	by nez-perce.inria.fr (8.8.7/8.8.7) with ESMTP id SAA11151
	for <caml-list@inria.fr>; Fri, 14 May 1999 18:14:37 +0200 (MET DST)
Received: by mail2.microsoft.com with Internet Mail Service (5.5.2524.0)
	id <K6VQCS6W>; Fri, 14 May 1999 09:14:35 -0700
Message-ID: <783D93998201D311B0CF00805FEAA07B7E8DAE@RED-MSG-42>
From: Manuel Fahndrich <maf@microsoft.com>
To: "'caml-list@inria.fr'" <caml-list@inria.fr>
Subject: RE: stream pattern [< pat=exp >] and extensions to ordinary patte
	rn m atching
Date: Fri, 14 May 1999 09:14:34 -0700
X-Mailer: Internet Mail Service (5.5.2524.0)
Sender: weis


I have to reply to my own earlier (faulty) comments...

  [...]
  I believe that with such an extension to
  OCAML, the pattern language we explored in the ICFP paper could be
expressed
  completely (and more).
  [...]

This is not entirely true. The pattern language explored in the ICFP paper
allows one to write non-deterministic patterns that get influenced by the
context. For example:

	pat leftOrRight(x) =
            (x, _)
          | (_, x)

If I then write

	match (1,2) with

        leftOrRight(2) ->
      | leftOrRight(y) ->

the first clause would apply. Using the pat=exp approach from stream
parsers, such choices in function of pat would not be possible. I consider
this feature not essential, and having pat=exp patterns whose semantics
evaluate exp first without regard to pat would still be very useful. 


-Manuel Fahndrich