From mboxrd@z Thu Jan 1 00:00:00 1970 Received: (from majordomo@localhost) by pauillac.inria.fr (8.7.6/8.7.3) id JAA29098; Tue, 5 Feb 2002 09:57:56 +0100 (MET) X-Authentication-Warning: pauillac.inria.fr: majordomo set sender to owner-caml-list@pauillac.inria.fr using -f Received: from concorde.inria.fr (concorde.inria.fr [192.93.2.39]) by pauillac.inria.fr (8.7.6/8.7.3) with ESMTP id JAA29105 for ; Tue, 5 Feb 2002 09:57:55 +0100 (MET) Received: from verdot.inria.fr (verdot.inria.fr [128.93.11.7]) by concorde.inria.fr (8.11.1/8.11.1) with ESMTP id g158vtH03774 for ; Tue, 5 Feb 2002 09:57:55 +0100 (MET) Received: (from ddr@localhost) by verdot.inria.fr (8.9.3/8.9.3) id JAA23647 for caml-list@inria.fr; Tue, 5 Feb 2002 09:57:55 +0100 Date: Tue, 5 Feb 2002 09:57:55 +0100 From: Daniel de Rauglaudre To: caml-list@inria.fr Subject: Re: [Caml-list] camlp4o problem (was: otags problem) Message-ID: <20020205095755.B23442@verdot.inria.fr> References: <20020204155242.B2338@verdot.inria.fr> <20020204222850.N52713-100000@fledge.watson.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.2.5i In-Reply-To: <20020204222850.N52713-100000@fledge.watson.org>; from patrick@watson.org on Mon, Feb 04, 2002 at 10:40:04PM -0500 Sender: owner-caml-list@pauillac.inria.fr Precedence: bulk Hi, On Mon, Feb 04, 2002 at 10:40:04PM -0500, Patrick M Doane wrote: > Why? The ocaml syntax has changed before - there is no technical reason > that prevents identical behavior between the two systems. Just experience. This case is more complicated than "fun x -> x+x" vs. "fun x -> 2*x". > I just don't see camlp4 as being practical until it can correctly > parse standard ocaml syntax. I have done it, little by little. But with the current ("standard") syntax I have sometimes problems, in particular lastly with class types. But there are ways to turn around by adding specific grammar entries with simple parsers, but loosing the "extensibility" for these cases. In the revised syntax, there are no such hacks. > From what I can tell, the revised syntax makes more changes to the > language than are necessary to fit into camlp4 parsing technology. Indeed. I was not obliged to use the constructors with currification or list cons between brackets: I just considered this as better (read the chapter about revised syntax in the Camlp4 tutorial), more logical. Yes this syntax is not only to make Camlp4 work, it is an "opinion", my opinion (built in part from my ideas, but not only). > If the revised syntax only changed the syntax as necessary to > support camlp4, and the standard compiler would be modified to > reject anything not accepted by this new syntax, then the following > proposal seems perfectly reasonable. You propose that I add a third syntax? A middle between the normal and the revised syntax. A norsed (or revmal) syntax? Mmmm... there is also the problem of unclosed constructions (match, try) which results in problems sometimes. It is not necessary to add that to be able to parse with LL(1) Camlp4. The revised syntax is supposed to fix some problems of the normal syntax. -- Daniel de RAUGLAUDRE daniel.de_rauglaudre@inria.fr http://cristal.inria.fr/~ddr/ ------------------- Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/ To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr