src/Gargantext/Core/Text/Corpus/Parsers/Iramuteq.hs

   1 {-|
   2 Module      : Gargantext.Core.Text.Corpus.Parsers.WOS
   3 Description :
   4 Copyright   : (c) CNRS, 2017-Present
   5 License     : AGPL + CECILL v3
   6 Maintainer  : team@gargantext.org
   7 Stability   : experimental
   8 Portability : POSIX
   9
  10 Here is a longer description of this module, containing some
  11 commentary with @some markup@.
  12 -}
  13
  14
  15 module Gargantext.Core.Text.Corpus.Parsers.Iramuteq (parseIramuteqFile, parser, keys) where
  16
  17 import Control.Applicative
  18 import Data.Attoparsec.ByteString (Parser, takeTill, parseOnly)
  19 import Data.Attoparsec.ByteString.Char8 (isEndOfLine, takeWhile, endOfLine)
  20 import Data.ByteString (ByteString)
  21 import Prelude hiding (takeWhile, take, concat, readFile, lines, concat)
  22 import qualified Data.ByteString as DB
  23
  24 parseIramuteqFile :: FilePath -> IO (Either String [[(ByteString, ByteString)]])
  25 parseIramuteqFile fp = do
  26   txts <- DB.readFile fp
  27   pure $ parseOnly parser txts
  28
  29 -------------------------------------------------------------
  30 parser :: Parser [[(ByteString, ByteString)]]
  31 parser = do
  32   ns <- (many notice)
  33   pure ns
  34
  35 notice :: Parser [(ByteString, ByteString)]
  36 notice = do
  37     hs <- headers
  38     ns <- takeWhile (/= '*')
  39     pure $ hs <> [("text", ns)]
  40
  41 -----------------------------------------------------------------
  42 headers :: Parser [(ByteString, ByteString)]
  43 headers = parseOf header fields
  44
  45 header :: Parser ByteString
  46 header = "**** " *> takeTill isEndOfLine <* endOfLine
  47
  48 -----------------------------------------------------------------
  49 fields :: Parser [(ByteString, ByteString)]
  50 fields = many (parseOf field fieldTuple)
  51
  52 field :: Parser ByteString
  53 field =   "*" *> takeWhile (/= ' ') <* " "
  54       <|> "*" *> takeWhile (/= '\n')
  55
  56 fieldTuple :: Parser (ByteString, ByteString)
  57 fieldTuple = do
  58   name <- takeWhile (/= '_') <* "_"
  59   rest <- takeWhile (/= '\n')
  60   pure (name,rest)
  61
  62 -----------------------------------------------------------------
  63 constP :: Parser a -> ByteString -> Parser a
  64 constP p t = case parseOnly p t of
  65   Left _ -> empty
  66   Right a -> return a
  67
  68 parseOf :: Parser ByteString -> Parser a -> Parser a
  69 parseOf ptxt pa = bothParse <|> empty
  70   where
  71     bothParse = ptxt >>= constP pa
  72
  73 -----------------------------------------------------------------
  74 -- These keys may not be constant for Iramuteq files formats
  75 keys :: ByteString -> ByteString
  76 keys f
  77       | f == "id"    = "doi"
  78       | f == "qui"   = "authors"
  79       | f == "quand" = "PY"
  80       | f == "type"  = "source"
  81       | f == "titre" = "title"
  82       | f == "ou"    = "institutes"
  83       | f == "text"  = "abstract"
  84       | otherwise  = f