Browse thread
'_a
[
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: | skaller <skaller@u...> |
| Subject: | Re: [Caml-list] Re: '_a |
On Sat, 2005-01-29 at 00:42, Christophe TROESTLER wrote: > On 28 Jan 2005, skaller <skaller@users.sourceforge.net> wrote: > > > > let x = > > let rec f l -> | [] -> raise Not_found > > | h :: t -> if h == v then (raise Found; 0) else f t > > in try f l with Found -> 1 | Not_found -> 2 > > in print_endline (string_of_int x) > > > > where the compiler doesn't know f l cannot return, so it needs a > > useless '0' after the Found case to get the typing correct. > > Not quite, [] > Maybe you mean something like > > let cl file = > let fh = open_in file in > let nl = ref 0 in > try > while true do > let _ = input_line fh in > incr nl > done > with End_of_file -> !nl Thanks for a better example (well a right one has to better :) > where the [while] has type [unit] while [!nl] has type [int] and the > two cannot be unified. But this is because there is no way for the > compiler to know that a loop indeed never ends, so one has to tell: > > let cl file = > let fh = open_in file in > let nl = ref 0 in > try > while true do > let _ = input_line fh in > incr nl > done; > assert false > with End_of_file -> !nl > > With that everything is fine. Yes, that's a better solution too. -- John Skaller, mailto:skaller@users.sf.net voice: 061-2-9660-0850, snail: PO BOX 401 Glebe NSW 2037 Australia Checkout the Felix programming language http://felix.sf.net