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 VAA19521 for caml-red; Sun, 20 Aug 2000 21:50:57 +0200 (MET DST) 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 VAA18535 for ; Sun, 20 Aug 2000 21:50:39 +0200 (MET DST) Received: from pauillac.inria.fr (pauillac.inria.fr [128.93.11.35]) by concorde.inria.fr (8.10.0/8.10.0) with ESMTP id e7KJoaD03186; Sun, 20 Aug 2000 21:50:37 +0200 (MET DST) Received: (from weis@localhost) by pauillac.inria.fr (8.7.6/8.7.3) id VAA19121; Sun, 20 Aug 2000 21:50:36 +0200 (MET DST) From: Pierre Weis Message-Id: <200008201950.VAA19121@pauillac.inria.fr> Subject: Re: [newbie] Caml Light vs Objective Caml In-Reply-To: from serge777 at "Aug 18, 100 08:35:37 am" To: serge777@my-Deja.com Date: Sun, 20 Aug 2000 21:50:36 +0200 (MET DST) Cc: caml-list@inria.fr X-Mailer: ELM [version 2.4ME+ PL28 (25)] MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: weis@pauillac.inria.fr [...] > Some Caml Light code from this book (and form others) does not work > for Ocaml, for example, Ocaml report a syntax > error for a definition from page 25 > > let neg = fun true -> false | false -> true ;; Yes, you should write function instead of fun here. > 1) Could anyone point out other known basic discrepancy between Caml > Light and Ocaml that were not mentioned in the FAQ? There are a few of them, namely: - predefined identifiers that has changed (module vect is now module Array) - module access conventions (module__ident is now Module.ident), hence String.sub, instead of sub_string, and String.blit instead of blit_string - case conventions have been enforced in Objective Caml. > 2) I know there is a French book "Diveloppement d'applications avec > Objective Caml". Is it suits for studying functional programming and > OCaml for peoples without any background in the functional programming? It is a good book with many interesting examples, worth reading. There are a lot of other french books for beginners, have a look at http://pauillac.inria.fr/caml/books-eng.html > 3) Why did the meaning of "fun" change? The raw ``meaning'' is still the same: fun introduces a function definition; however the syntax of the construction, and the kind of function definitions the fun keyword introduces has changed. Hence, this is one of the more difficult changes you have to make when you want to translate to O'Caml some code that heavily uses this construct. I summarize here the differences: What is perfectly equivalent: ----------------------------- Introducing a curried function with multiple arguments, as in fun x y z -> expression What has slightly changed: -------------------------- Introducing a pattern matching function with 1 argument; you must change the keyword fun to the keyword function. Hence fun true -> false | false -> true becomes function true -> false | false -> true What has completely changed: ---------------------------- In Objective Caml, the notion of ``curried pattern matching'' (as defined in older versions of Caml) has been removed, because this feature has been considered as having a too questionable (and complex to understand) semantics. The ``curried pattern matching'' is the ability to performs multiple pattern matching in parallel, here multiple pattern matching when defining a function. Instead of the O'Caml code fun x y -> match x, y with | 1, 1 -> bli | _, 0 -> bla | _, _ -> blu you could write the Caml Light specific fragment fun | 1 1 -> bli | _ 0 -> bla | _ _ -> blu So you need to perform the converse transformation, which is fortunately easy if not completely mechanical. > Thank you in advance. > > Sergei. You're welcome. Pierre Weis INRIA, Projet Cristal, Pierre.Weis@inria.fr, http://cristal.inria.fr/~weis/