On 01/09/2016 07:42 PM, Tom Ellis wrote:
On Sat, Jan 09, 2016 at 07:22:48PM +0100, Arjen wrote:
I was thinking of exps as a value (having a non-function type). Maybe I'm wrong or just not understanding the issue fully. How do you differentiate between expression that are values and those that are function applications on arguments?
I'm not sure you really "differentiate" between them. The evaluation of function arguments just has different consequences when those argument values are integers than it does when those argument values are functions.
If I were to print the value of exps, like main = print exps. How would I express this? Or is it print's responsibility to evaluate the argument? Say, exps has type Integer. How does print differentiate between a actual value (print 42) and unevaluated expressions (print exps)?
I'm not quite sure what you're asking, but consider the difference between these two calls to the function id in OCaml.
# let id x = x;; val id : 'a -> 'a = <fun>
# id (Printf.printf "Hello");; Hello- : unit = ()
# id (fun () -> Printf.printf "Hello");; - : unit -> unit = <fun>
The correspond to the following in an imaginary impure Haskell-like language:
id (print "Hello") "Hello"
id (\() -> print "Hello") <no output>
I think I see what you mean. Then in OCaml exps would have type unit->Int? And to get the value, you would use exps ()? kind regards, Arjen