Haskell CSV histogram printer

csv_histogram.hs

import qualified Data.Text as Text
import qualified Data.Text.IO
import qualified Data.Map as Map
import Data.List
import Data.Ord

-- Restrict histogram values to Int (shouldn't be too big and removes default constraint warnings)
type TextHistogramValue = Int
type TextHistogramKey = Text.Text
type TextHistogramEntry = (TextHistogramKey, TextHistogramValue)
type TextHistogram = Map.Map TextHistogramKey TextHistogramValue

-- Reduce a list of Text into a histogram of the occurences of each Text element
countOccurrences :: [Text.Text] -> TextHistogram
countOccurrences = foldr (\t m -> Map.insertWith (+) t 1 m) Map.empty

-- Take CSV rows (i.e. copy/pasted from Google Docs), join lines w/ ',', split all on ',' and strip whitespace
joinAndCleanCSVLines :: Text.Text -> Text.Text -> [Text.Text]
joinAndCleanCSVLines d t = map Text.strip $ Text.splitOn d $ (Text.intercalate d . Text.lines) t

-- Transform TextHistogramEntry into a bullet point
toBulletedListItem :: TextHistogramEntry -> String
toBulletedListItem (t, n) = "- " ++ Text.unpack t ++ ": " ++ show n

-- CSV Text delimiter
delim :: Text.Text
delim = Text.pack ","

-- Read all lines at once, separate into a single CSV row, then convert into bulleted histogram
main :: IO ()
main = do
  csv <- Data.Text.IO.getContents
  let topicHistogram = countOccurrences $ joinAndCleanCSVLines delim csv
      sortedTopicCount = sortBy (flip (comparing snd)) $ Map.toList topicHistogram
  putStr $ unlines $ map toBulletedListItem sortedTopicCount