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

   1 {-|
   2 Module      : Gargantext.Core.Text.Corpus.Parsers.RIS
   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
  11 RIS is a standardized tag format developed by Research Information
  12 Systems, Incorporated (the format name refers to the company) to enable
  13 citation programs to exchange data.
  14
  15 [More](https://en.wikipedia.org/wiki/RIS_(file_format))
  16
  17 -}
  18
  19 module Gargantext.Core.Text.Corpus.Parsers.RIS (parser, onField, fieldWith, lines) where
  20
  21 import Data.List (lookup)
  22 import Control.Applicative
  23 import Data.Attoparsec.ByteString (Parser, try, takeTill, take, many1)
  24 import Data.Attoparsec.ByteString.Char8 (isEndOfLine)
  25 import Data.ByteString (ByteString, intercalate)
  26 import Gargantext.Prelude hiding (takeWhile, take)
  27 import qualified Data.List as DL
  28
  29 -------------------------------------------------------------
  30 parser :: Parser [[(ByteString, ByteString)]]
  31 parser = do
  32     n  <- notice "TY  -"
  33     ns <- many1 (notice "\nTY  -")
  34     pure $ [n] <> ns
  35
  36 notice :: Parser ByteString -> Parser [(ByteString, ByteString)]
  37 notice s = start *> many (fieldWith field)  <* end
  38     where
  39       field :: Parser ByteString
  40       field = "\n" *> take 2 <* "  - "
  41
  42       start :: Parser ByteString
  43       start = s *> takeTill isEndOfLine
  44
  45       end :: Parser ByteString
  46       end =  "\nER  -" *> takeTill isEndOfLine
  47
  48
  49 fieldWith :: Parser ByteString -> Parser (ByteString, ByteString)
  50 fieldWith n = do
  51     name  <- n
  52     txt   <- takeTill isEndOfLine
  53     txts  <- try lines
  54     let txts' = case DL.length txts > 0 of
  55             True  -> txts
  56             False -> []
  57     pure (name, intercalate ";" ([txt] <> txts'))
  58
  59
  60 lines :: Parser [ByteString]
  61 lines = many line
  62     where
  63         line :: Parser ByteString
  64         line = "\n " *> takeTill isEndOfLine
  65
  66 -------------------------------------------------------------
  67 -- Field for First elem of a Tuple, Key for corresponding Map
  68 onField :: ByteString -> (ByteString -> [(ByteString, ByteString)])
  69        -> [(ByteString, ByteString)] -> [(ByteString, ByteString)]
  70 onField k f m = m <> ( maybe [] f (lookup k m) )
  71