From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <jon@jdh30.plus.com>
Delivered-To: caml-list@yquem.inria.fr
Received: from concorde.inria.fr (concorde.inria.fr [192.93.2.39])
	by yquem.inria.fr (Postfix) with ESMTP id 62CC1BB91
	for <caml-list@yquem.inria.fr>; Sun, 23 Jan 2005 22:48:09 +0100 (CET)
Received: from ptb-relay03.plus.net (ptb-relay03.plus.net [212.159.14.214])
	by concorde.inria.fr (8.13.0/8.13.0) with ESMTP id j0NLm8um008708
	(version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-SHA bits=256 verify=NO)
	for <caml-list@yquem.inria.fr>; Sun, 23 Jan 2005 22:48:09 +0100
Received: from [80.229.56.224] (helo=chetara)
	 by ptb-relay03.plus.net with esmtp (Exim) id 1CspsI-000Dho-2y
	for caml-list@yquem.inria.fr; Sun, 23 Jan 2005 22:06:26 +0000
From: Jon Harrop <jon@jdh30.plus.com>
Organization: University of Cambridge
To: caml-list@yquem.inria.fr
Subject: Re: [Caml-list] Strange syntax error
Date: Sun, 23 Jan 2005 21:49:54 +0000
User-Agent: KMail/1.7.1
References: <20050123211112.GA14067@furbychan.cocan.org>
In-Reply-To: <20050123211112.GA14067@furbychan.cocan.org>
MIME-Version: 1.0
Content-Type: text/plain;
  charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
Message-Id: <200501232149.54437.jon@jdh30.plus.com>
X-Miltered: at concorde with ID 41F41B98.000 by Joe's j-chkmail (http://j-chkmail.ensmp.fr)!
X-Spam: no; 0.00; caml-list:01 syntax:01 wrote:01 parsing:01 bug:01 failwith:01 failwith:01 cheers:01 ...:98 inefficient:01 patterns:02 tree:02 guess:02 clause:03 pattern:03 
X-Spam-Checker-Version: SpamAssassin 3.0.0 (2004-09-13) on yquem.inria.fr
X-Spam-Status: No, score=0.0 required=5.0 tests=none autolearn=disabled 
	version=3.0.0
X-Spam-Level: 

On Sunday 23 January 2005 21:11, Richard Jones wrote:
> ...
> Is this a parsing bug?

I assume a "when" clause is only allowed at the end of the list of alternative 
patterns. For example:

let f = function
  { keyword = keyword } -> keyword
| { keyword = keyword; cost = None } when keyword.[0] = '-' -> keyword
| _ -> failwith "whatever"

is allowed (but not what you wanted).

I guess that allowing "when" clauses on each alternative pattern would obviate 
the advantages of using a decision tree to represent pattern matching and, 
therefore, would make pattern matches significantly more inefficient. Hence 
they don't do it.

In this case, I'd go for:

let f = function
  { keyword = keyword } when keyword.[0] = '-' -> keyword
| { cost = None } as t -> t.keyword
| _ -> failwith "whatever"

Say, this wouldn't be related to that arg-passing problem would it? ;-)

Cheers,
Jon.