concatMapAccumC.hs

#!/usr/bin/env stack
-- stack script --resolver lts-8.12
{-# LANGUAGE OverloadedStrings #-}
import Conduit
import Text.Regex (matchRegex,mkRegex,Regex)
import qualified Data.Text as T


loghead = mkRegex "^([0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}:[0-9]{2},[0-9]{3} )"
-- "2015-01-25 00:04:18,840"

logMerge
  :: Regex
  -> Maybe String -- ^ new line
  -> String -- ^ accumulator
  -> (String, [String])
logMerge _ Nothing accum = ("", [accum ++ "\n"])
logMerge logregex (Just str) accum =
    case  matchRegex logregex  str of
        Just _ ->  (str,[(accum++"\n")])
        Nothing -> case null accum of
            True ->  (str,[])
            False ->  (accum ++ "<br>" ++ str,[])


runMerge :: String -> String -> IO ()
runMerge infile outfile =
    runConduitRes
  $ ((sourceFile infile .| decodeUtf8C .| linesUnboundedC .| mapC (Just . T.unpack)) >> yield Nothing)
 .| concatMapAccumC (logMerge loghead ) ""
 .| mapC T.pack
 .| encodeUtf8C
 .| sinkFile outfile

main :: IO ()
main = runMerge "infile.txt" "outfile.txt"