Re: Arrays and arrays

ketil@ii.uib.no (Ketil Z. Malde) writes:

ketil@ii.uib.no (Ketil Z. Malde) writes:

...

Hal Daume III writes:

...

...

Yes, (//) is terrible. It *never* tries to update in-place.

replace :: UArray Int Int -> [(Int,Int)] -> UArray Int Int replace a p = runST (thaw a >>= \u -> update u p >> freeze u)

update :: STUArray () Int Int -> [(Int,Int)] -> ST () () update u ps = mapM_ (uncurry (writeArray u)) ps

I've tried rearranging the code in various ways, but I can get no further.

It's funny, you know, how asking questions on the internet gets you the answer quickly. Just after sending this, I scratched my head, and did: update :: STUArray s Int Int -> [(Int,Int)] -> ST s () and, well, that apparently worked. Like a charm. The 's' parameter is apparently just magic, or in Marcin Kowalczyk's words (old (26 Feb 2001) post to the haskell list): | The type variable 's' is used in a very tricky way, to ensure | safety when | runST :: (forall s. ST s a) -> a | is used to wrap the ST-monadic computation in a purely functional | interface. It does not correspond to the type of data being | manipulated. (I'll be right back with the benchmarks.) -kzm -- If I haven't seen further, it is by standing in the footprints of giants