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
  20 module Gargantext.Core.Text.Corpus.Parsers.RIS (parser, onField, fieldWith, lines) where
  21
  22 import Data.List (lookup)
  23 import Control.Applicative
  24 import Data.Attoparsec.ByteString (Parser, try, takeTill, take, many1)
  25 import Data.Attoparsec.ByteString.Char8 (isEndOfLine)
  26 import Data.ByteString (ByteString, intercalate)
  27 import Gargantext.Prelude hiding (takeWhile, take)
  28 import qualified Data.List as DL
  29 -------------------------------------------------------------
  30
  31 parser :: Parser [[(ByteString, ByteString)]]
  32 parser = do
  33     n  <- notice "TY  -"
  34     ns <- many1 (notice "\nTY  -")
  35     pure $ [n] <> ns
  36
  37 notice :: Parser ByteString -> Parser [(ByteString, ByteString)]
  38 notice s = start *> many (fieldWith field)  <* end
  39     where
  40       field :: Parser ByteString
  41       field = "\n" *> take 2 <* "  - "
  42
  43       start :: Parser ByteString
  44       start = s *> takeTill isEndOfLine
  45
  46       end :: Parser ByteString
  47       end =  "\nER  -" *> takeTill isEndOfLine
  48
  49
  50 fieldWith :: Parser ByteString -> Parser (ByteString, ByteString)
  51 fieldWith n = do
  52     name  <- n
  53     txt   <- takeTill isEndOfLine
  54     txts  <- try lines
  55     let txts' = case DL.length txts > 0 of
  56             True  -> txts
  57             False -> []
  58     pure (name, intercalate ";" ([txt] <> txts'))
  59
  60
  61 lines :: Parser [ByteString]
  62 lines = many line
  63     where
  64         line :: Parser ByteString
  65         line = "\n " *> takeTill isEndOfLine
  66
  67 -------------------------------------------------------------
  68 -- Field for First elem of a Tuple, Key for corresponding Map
  69 onField :: ByteString -> (ByteString -> [(ByteString, ByteString)])
  70        -> [(ByteString, ByteString)] -> [(ByteString, ByteString)]
  71 onField k f m = m <> ( maybe [] f (lookup k m) )
  72