Using csv-conduit to answer "How many kms of fences are on public land in the Gold Coast?"

bigdata.cabal

name:                fences
version:             0.1.0.0
-- synopsis:            
-- description:         
license:             BSD3
license-file:        LICENSE
author:              Maxwell Swadling
maintainer:          [email protected]
-- copyright:           
-- category:            
build-type:          Simple
-- extra-source-files:  
cabal-version:       >=1.10

executable bigdata
  main-is:             Fences.hs
  -- other-modules:       
  -- other-extensions:    
  build-depends:       base >=4.6 && <4.7, conduit >=1.0 && <1.1, csv-conduit == 0.5.1, text >= 1.1 && < 1.2
  -- hs-source-dirs:      
  default-language:    Haskell2010

Fences.hs

import Data.Conduit (Conduit, runResourceT, ($$), ($=), (=$=))
import Data.Conduit.Binary (sourceFile)
import Data.Conduit.List as CL (consume, map)
import Data.CSV.Conduit (Row, intoCSV, defCSVSettings)
import Data.Text (Text, unpack)

-- data source: http://data.gov.au/dataset/fences-on-public-land

mkValuable :: Monad m => Conduit (Row Text) m Double
mkValuable = CL.map extractVal
  where
    extractVal = read . unpack . flip (!!) 7

getBig :: IO [Double]
getBig = fmap (drop 1) -- the power of lazy
       $ runResourceT
       $ sourceFile "fences.csv"
       $= intoCSV defCSVSettings
       $= mkValuable
       $$ consume

main = do
  xs <- getBig
  putStrLn $ show (round (sum xs / 1000)) ++ "km of fences"

Setup.hs

import Distribution.Simple
main = defaultMain