On 7/21/06, S C Kuo
Not totally relevant to what the discussion has evolved to, but I wrote a factorial function using STRefs (in the spirit of the Evolution of a Haskell programmer) and I think it qualifies as a really simple example. Code follows:
import Data.STRef import Control.Monad.ST
foreach :: (Monad m) => [a] -> (a -> m b) -> m () foreach = flip mapM_ -- Bryn Keller's foreach, but with type restrictions
fac :: (Num a, Enum a) => a -> a fac n = runST (fac' n)
fac' :: (Num a, Enum a) => a -> ST s a fac' n = do r <- newSTRef 1 foreach [1..n] (\x -> modifySTRef r (*x)) x <- readSTRef r return x
Forgive me for not understanding, but I was hoping you would explain a choice you made in your code. Why did you define foreach and then use
foreach [1..n] (\x -> modifySTRef r (*x))
Instead of simply using
mapM_ (\x -> modifySTRef r (*x)) [1..n]
? I tried it out in GHCi, and it worked fine, and I have seen code that has been defined as a flip to take advantage of partial application. But your code doesn't seem to take advantage of partial application, so why did you define 'foreach' and then use it instead of using 'mapM_'? I am just curious, and have always been interested in reasons behind coding style. Bryan Burgers