Skip to content

Instantly share code, notes, and snippets.

@adinapoli
Created October 16, 2012 12:39
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 1 You must be signed in to fork a gist
  • Save adinapoli/3899017 to your computer and use it in GitHub Desktop.
Save adinapoli/3899017 to your computer and use it in GitHub Desktop.
Iteratee attempt
{-# LANGUAGE OverloadedStrings #-}
import Data.ByteString (ByteString)
import qualified Data.ByteString.Char8 as B
import Data.Enumerator (Iteratee, Enumeratee, ($=), (=$), (=$=), (==<<))
import qualified Data.Enumerator as E
import qualified Data.Enumerator.Binary as EB
import qualified Data.Enumerator.Text as ET
import qualified Data.Enumerator.List as EL
import Data.Text
import System.Environment
-- http://okmij.org/ftp/Haskell/Iteratee/describe.pdf
------------------------------------------------------------------------------
countCharBS :: (Monad m) => Char -> Iteratee ByteString m Integer
countCharBS needle = loop 0
where loop n = EL.head >>= check n
check n Nothing = return n
check n (Just t) = loop (n + toInteger (B.count needle t))
------------------------------------------------------------------------------
countLines :: (Monad m) => Iteratee Text m Integer
countLines = E.length
------------------------------------------------------------------------------
convertUtf8 :: (Monad m) => Enumeratee ByteString Text m b
convertUtf8 = ET.decode ET.utf8
------------------------------------------------------------------------------
main :: IO ()
main = do
args <- getArgs
case args of
(fname:needle:[]) ->
case needle of
(c:"") -> do
print "Switching to single char scan..."
cnt <- E.run $ EB.enumFile fname ==<< countCharBS c
print cnt
_ -> do
cnt <- E.run $ (EB.enumFile fname $= convertUtf8) ==<<
countLines
print cnt
_ -> print "Error in accessing file."
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment