[
Home
]
[ Index:
by date
|
by threads
]
[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
| Date: | -- (:) |
| From: | Jean-Christophe Filliatre <Jean-Christophe.Filliatre@l...> |
| Subject: | Re: [Caml-list] Is it possible to implement this? |
Lukasz Lew writes: > > Let's suppose wy have module with such interface > > module X = > type t (*big data structure*) > > add : string -> t > delete : t -> () > > And in the other module i keep X.t list, and i want to keep with each of > them additional data of unknown type [at time of X compilation] You could have a polymorphic type "'a X.t" where 'a stands for the type of the associated data, and provide functions "add_info : 'a -> 'a t -> unit" and "get_info : 'a t -> 'a" in module X. > Other is to keep with every t int identifier as small as possible, to keep > additional data in array, but this solution isn't very good. Following this kind of idea, you could do hash-consing of your X.t values, so that they are allocated only once; doing this, it is easy to associate a unique integer to each value in X.t, which will allow you to build an efficient map over X.t. I already wrote an hash-consing module, available here: http://www.lri.fr/~filliatr/software.en.html (By the way, doing hash-consing has many other advantages such as saving space, providing O(1) equality, ...) Hope this helps, -- Jean-Christophe Filliātre ------------------- To unsubscribe, mail caml-list-request@inria.fr Archives: http://caml.inria.fr Bug reports: http://caml.inria.fr/bin/caml-bugs FAQ: http://caml.inria.fr/FAQ/ Beginner's list: http://groups.yahoo.com/group/ocaml_beginners