module DetectEnglish where

import Data.Char
import Data.HashSet (HashSet)
import qualified Data.HashSet as HashSet
import Data.List
import System.IO

importDictionary :: IO (HashSet String)
importDictionary = do
  wholeFile <- readFile "dictionary.txt"
  return (HashSet.fromList (words wholeFile))

-- |
-- >>> isEnglish (HashSet.fromList ["HI", "THERE"]) 0.5 "hi there English"
-- True

-- |
-- >>> isEnglish (HashSet.fromList ["HI", "THERE"]) 0.2 "hi there English"
-- True
isEnglish :: HashSet String -> Float -> String -> Bool
isEnglish hashDict threshold message =
  let wordsList = words (map toUpper message)
      numEnglishWords = genericLength . filter id $ map (`HashSet.member` hashDict) wordsList
      numTotalWords = genericLength wordsList
   in threshold < numEnglishWords / numTotalWords