Neil Mitchell wrote:
Hi
Want to try also the
Data.ByteString.Base.isSpaceWord8 :: Word8 -> Bool
isspace: 0.375
iswspace: 0.400
ByteString: 0.460
Char: 0.672
Not as fast as isspace/iswspace, but quite a bit faster than Char.
Perhaps someone needs to take a peek at the generated ASM for each of
these routines and see where the Haskell one is falling down.
Depending on the platform, iswspace might not be checking the full Unicode
character space here: on Linux it doesn't unless you have LANG set, I think.
The Data.Char version does support full Unicode.
That's not to say we couldn't improve things here, I'm sure we can.
Cheers,
Simon
_______________________________________________
Glasgow-haskell-users mailing list
Glasgow-haskell-users@haskell.org
http://www.haskell.org/mailman/listinfo/glasgow-haskell-users
