bin/gargantext-cli/Main.hs

   1 {-|
   2 Module      : Main.hs
   3 Description : Gargantext starter
   4 Copyright   : (c) CNRS, 2017-Present
   5 License     : AGPL + CECILL v3
   6 Maintainer  : team@gargantext.org
   7 Stability   : experimental
   8 Portability : POSIX
   9
  10 Main specifications to index a corpus with a term list
  11
  12  -}
  13
  14 {-# LANGUAGE DataKinds         #-}
  15 {-# LANGUAGE DeriveGeneric     #-}
  16 {-# LANGUAGE FlexibleInstances #-}
  17 {-# LANGUAGE NoImplicitPrelude #-}
  18 {-# LANGUAGE OverloadedStrings #-}
  19 {-# LANGUAGE StandaloneDeriving #-}
  20 {-# LANGUAGE TypeOperators     #-}
  21 {-# LANGUAGE Strict             #-}
  22
  23 module Main where
  24
  25 import qualified Data.Vector as DV
  26
  27 import Data.Text (Text)
  28 import System.Environment
  29 --import Control.Concurrent.Async as CCA (mapConcurrently)
  30
  31 import Gargantext.Prelude
  32 import Gargantext.Text.Context
  33 import Gargantext.Text.Terms
  34 import Gargantext.Text.Terms.WithList
  35 import Gargantext.Text.Parsers.CSV (readCsv, csv_title, csv_abstract)
  36 import Gargantext.Text.List.CSV (csvGraphTermList)
  37 import Gargantext.Text.Terms (terms)
  38 import Gargantext.Text.Metrics.Count (cooc)
  39
  40 main :: IO ()
  41 main = do
  42   [corpusFile, termListFile, outputFile] <- getArgs
  43
  44   -- corpus :: [Text]
  45   corpus <- DV.toList <$> map (\n -> (csv_title n) <> " " <> (csv_abstract n))
  46                       <$> snd
  47                       <$> readCsv corpusFile
  48
  49   putStrLn $ show $ length corpus
  50   -- termListMap :: [Text]
  51   termList <- csvGraphTermList termListFile
  52
  53   putStrLn $ show $ length termList
  54
  55   corpusIndexed <- mapM (terms (WithList $ buildPatterns termList)) corpus
  56
  57   putStrLn $ show corpusIndexed
  58   let myCooc = cooc corpusIndexed
  59
  60   putStrLn $ show myCooc
  61   --writeFile outputFile cooc