]> zze-MARCHEGAY Michael stagiaire FTRD/DTL/LAN wrote: > In fact, if an object 'a' contains an attribute 'b' that is another > object, Oo.copy on 'a' gives a copy of 'a' in which the attribute > 'b' is **physically** equal to the copied attribute from 'a'. I agree so far. But did you actually test the code below? I have reason to believe you are mistaken in believing that mutable fields are shared between Oo.copied objects, in such a way that assignment to such a field in one object will result in a modification in the value of the same field in all copies. I would consider such a behavior a major design flaw in the language. The following toplevel session proves my point: # class a i = object val mutable c = i method m = c <- c + 1 ; print_string ((string_of_int c) ^ " ") end;; class a : int -> object method m : unit val mutable c : int end # let a1 = new a 0;; val a1 : a = <obj> # let a2 = Oo.copy a1;; val a2 : a = <obj> # for i = 1 to 10 do a1 # m done;; 1 2 3 4 5 6 7 8 9 10 - : unit = () # for i = 1 to 15 do a2 # m done;; 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 - : unit = () # for i = 11 to 15 do a1 # m done;; 11 12 13 14 15 - : unit = () Alex > Example: > class a = > object > val mutable b = new b > method get_b = b > end > and b = > object > val mutable c = "1" > method set_c x = c <- x > method print_c = print_string ("c=" ^ c ^ "\n") > end > > let aa = new a > let aa' = Oo.copy aa > > let _ = > aa#get_b#set_c "3"; > aa#get_b#print_c; > aa'#get_b#print_c > > gives: > c=3 > c=3 > > whereas a deep copy would give: > c=3 > c=1 ------------------- 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