(Apologies for previous incomplete post: Here's the whole thing)
One thing that bugs me about the named record syntax, is that with a datatype:
data T = T {
t_f1 :: X,
t_f2 :: X
};
the function t_f1 has the type of an "accessor", ie
t_f1 :: T -> X
but there doesn't any tidy way to get at the "mutator" function
t_f1' :: X -> T -> T
without actually having to write the following
t_f1' x t = t{t_f1=x}
or is there?
Actually mutators of the more general form
t_f1' :: (X->X) -> T -> T
t_f1' f t = t{t_f1=f (t_f1 t)}
would seem to be quite useful, (even though they are tedious to write)
as you can compose them for nested structures, ie if
data X = X {
x_f1 :: Int,
x_f2 :: Int
};
Then...
(t_f1'.x_f2') (const 5)
would be the functional equivalent of the imperative t.f1.f2 = 5
Tim
_______________________________________________
Haskell mailing list
[EMAIL PROTECTED]
http://www.haskell.org/mailman/listinfo/haskell
