readR

module Reader
       ( -- * Functions
         readReviews
       , countWords
       ) where


import System.Environment
import System.IO
import System.IO.Error
import System.Directory
import Control.DeepSeq
import Data.List
import Data.List.Split
import Data.Char
import Types

main = do
    (pos,neg) <- readReviews "../../resources/dvd/"
    let keys = ["batman"]
    let x = map (countWords keys) pos
    mapM_ print  ( zipWith (++) (map show [1..]) (map show x))

readReviews :: String -> IO RawReview
readReviews path = do
    pos <- readDir (path ++ "/pos/")
    neg <- readDir (path ++ "/neg/")
    return (pos,neg)


readDir :: String -> IO [String]
readDir p = do
    files <- getDirectoryContents p
    let pathAndFiles = map (p ++) . drop 2 $ files
    contents <- mapM (readFile' latin1) pathAndFiles
    return contents


readFile' :: TextEncoding -> FilePath -> IO String
readFile' enc path = do
    h <- openFile path ReadMode
    hSetEncoding h enc
    s <- hGetContents h
    deepseq s (hClose h)
    return s


countWords :: [String] -> String -> [Double]
countWords ws s = [ genericLength (filter (==w) (filterSuffix (words (filterNonChar s)) )) | w <- ws ]
{-countWords ws s = [ genericLength (filter (==w) (words (filterNonChar s) )) | w <- ws ]-}

filterSuffix :: String -> String
filterSuffix s = do
    head (split (oneOf delimeter) s)
    where "'\\"

filterNonChar :: String -> String
filterNonChar s = do
    filter (`notElem` nonChar ) (map toLower s)
    where nonChar = ",.;"