bin/gargantext-cli/CleanCsvCorpus.hs

   1 {-|
   2 Module      : CleanCsvCorpus.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 Given a Gargantext CSV File and its Query This script cleans and
  11 compress the contexts around the main terms of the query.
  12 -}
  13
  14 {-# LANGUAGE NoImplicitPrelude #-}
  15
  16 module CleanCsvCorpus  where
  17
  18 --import GHC.IO (FilePath)
  19 import Data.SearchEngine as S
  20 import qualified Data.Set as S
  21 import Data.Text (pack)
  22 import Data.Vector (Vector)
  23 import qualified Data.Vector as V
  24
  25 import Gargantext.Prelude
  26 import Gargantext.Text.Search
  27 import Gargantext.Text.Parsers.CSV
  28
  29 ------------------------------------------------------------------------
  30
  31 type Query = [S.Term]
  32
  33 filterDocs :: [DocId] -> Vector Doc -> Vector Doc
  34 filterDocs docIds = V.filter (\doc -> S.member (d_docId doc) $ S.fromList docIds )
  35
  36
  37 main :: IO ()
  38 main = do
  39   let rPath = "/tmp/Gargantext_Corpus.csv"
  40   let wPath = "/tmp/Gargantext_Corpus_bis.csv"
  41   --let q = ["water", "scarcity", "morocco", "shortage","flood"]
  42   let q = ["gratuit", "gratuité", "culture", "culturel"]
  43
  44   (h,csvDocs) <- readCsv rPath
  45
  46   putStrLn $ "Number of documents before:" <> show (V.length csvDocs)
  47   putStrLn $ "Mean size of docs:" <> show ( docsSize csvDocs)
  48
  49   let docs   = toDocs csvDocs
  50   let engine = insertDocs docs initialDocSearchEngine
  51   let docIds = S.query engine (map pack q)
  52   let docs'  = fromDocs $ filterDocs docIds (V.fromList docs)
  53
  54   putStrLn $ "Number of documents after:" <> show (V.length docs')
  55   putStrLn $ "Mean size of docs:" <> show (docsSize docs')
  56
  57   writeCsv wPath (h, docs')