Version française
Home     About     Download     Resources     Contact us    
Browse thread
[Caml-list] Doubt about function delaration parameter
[ Home ] [ Index: by date | by threads ]
[ Search: ]

[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
Date: -- (:)
From: bluestorm <bluestorm.dylc@g...>
Subject: Re: [Caml-list] Doubt about function delaration parameter
In Caml Light, the rule is the following :
- if you declare a *new* type, such as a sum type ( Foo | Bar | Baz ) or a
record type ( { foo : t; bar : t' } ), use "="
- if you just add an alias, a new name for an existing type, use "=="

It is an alias if the right-hand side is a "type expression" (something you
could use as is in a type annotation)

It is interesting to note that, though less syntactically visible, the
difference has not vanished in OCaml.
OCaml type declarations may have two components, a type "equation" (~ alias
for a type expression) and a "representation" (~ new ADT). For example the
following is valid :

  type test = A.ty = A | B | C

The type is declared to be an alias to the type "ty" in the module A, and
also represented by the sum (A | B | C) : the definition is "re-exported",
and an error will be thrown if A.ty is different from (A | B | C). It is
occasionnaly nice to make representation explicit, while still keeping
compatibility at the type level (the two types are considered equal by the
type checker), and can also be avoid to "strenghten" an external dependency
: if the upstream type definition changes, I'm sure my code will fail here
instead of producing weird errors later.
See the OCaml manual :
http://caml.inria.fr/pub/docs/manual-ocaml/manual016.html

The difference between type declarations and alias/abbreviation is also
significant for the semantics of "private types", which are an extension (or
rather several successive extensions) to the OCaml language described here
: http://caml.inria.fr/pub/docs/manual-ocaml/manual021.html#toc76 .



On Wed, Dec 15, 2010 at 5:56 PM, Raphael Proust <raphlalou@gmail.com> wrote:

> Le mercredi 15 décembre 2010 17:15:23, sieira a écrit :
> > Thanks for your replies. I'm now having some issue with Raphael's
> > suggestion of using (string * string) list;; as the menu type.
> >
> >
> > type menu = (string*string) list;;
> >
> > Results in a syntax error at the first parenthesis, while
> >
> > type menu = string*string;;
> >
> > fails too (at the asterisk)
> >
> > It seems like I'm missing something. Since according to the
> > http://caml.inria.fr/pub/docs/manual-caml-light/node3.5.html
>  documentantion
> > , this sintax should be right.
> >
> > I'm using Camllight 0.81 by François Boisson running in Ubuntu lucid lynx
>
> I just installed Camllight and I encountered the problem as you. Because
> the
> type menu is just a shorter name for the complete type (string * string)
> list,
> you have to use:
> type menu == ( string * string ) list ;;
>
> (with two '=' signs)
> I don't know the exact rule about this. It might be close to the difference
> between 'type' and 'data' in Haskell.
>
> Good luck!
> --
> _______
> Raphaël
>
> _______________________________________________
> Caml-list mailing list. Subscription management:
> http://yquem.inria.fr/cgi-bin/mailman/listinfo/caml-list
> Archives: http://caml.inria.fr
> Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> Bug reports: http://caml.inria.fr/bin/caml-bugs
>
>