src/Gargantext/Text/Parsers/WOS.hs

   1 {-|
   2 Module      : Gargantext.Text.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 {-# LANGUAGE NoImplicitPrelude #-}
  15 {-# LANGUAGE OverloadedStrings #-}
  16
  17 module Gargantext.Text.Parsers.WOS (wosParser) where
  18
  19 -- TOFIX : Should import Gargantext.Prelude here
  20 import Prelude hiding (takeWhile, take, concat, readFile, lines, concat)
  21
  22 import qualified Data.List as DL
  23
  24 import Data.Monoid ((<>))
  25 import Data.Attoparsec.ByteString (Parser, try, string
  26                                   , takeTill, take
  27                                   , manyTill, many1)
  28 import Data.Attoparsec.ByteString.Char8 (anyChar, isEndOfLine)
  29 import Data.ByteString (ByteString, concat)
  30 import Data.ByteString.Char8 (pack)
  31 import Control.Applicative
  32
  33 -------------------------------------------------------------
  34
  35
  36
  37
  38 -- | wosParser parses ISI format from
  39 -- Web Of Science Database
  40 wosParser :: Parser [[(ByteString, ByteString)]]
  41 wosParser = do
  42     -- TODO Warning if version /= 1.0
  43     -- FIXME anyChar (string ..) /= exact string "\nVR 1.0" ?
  44     _  <- manyTill anyChar (string $ pack "\nVR 1.0")
  45     ns <- many1 notice <*  (string $ pack "\nEF"    )
  46     pure ns
  47
  48 notice :: Parser [(ByteString, ByteString)]
  49 notice = start *> fields <* end
  50     where
  51       start :: Parser ByteString
  52       start = "\nPT " *> takeTill isEndOfLine
  53
  54       end :: Parser [Char]
  55       end = manyTill anyChar (string $ pack "\nER\n")
  56
  57
  58 fields :: Parser [(ByteString, ByteString)]
  59 fields = many field
  60     where
  61         field :: Parser (ByteString, ByteString)
  62         field = do
  63             name  <- "\n" *> take 2 <* " "
  64             txt   <- takeTill isEndOfLine
  65             txts  <- try lines
  66             let txts' = case DL.length txts > 0 of
  67                     True  -> txts
  68                     False -> []
  69             pure (translate name, concat ([txt] <> txts'))
  70
  71
  72 lines :: Parser [ByteString]
  73 lines = many line
  74     where
  75         line :: Parser ByteString
  76         line = "\n  " *> takeTill isEndOfLine
  77
  78 translate :: ByteString -> ByteString
  79 translate champs
  80             | champs == "AF" = "authors"
  81             | champs == "TI" = "title"
  82             | champs == "SO" = "source"
  83             | champs == "DI" = "doi"
  84             | champs == "PD" = "publication_date"
  85             | champs == "AB" = "abstract"
  86             | otherwise  = champs
  87 -------------------------------------------------------------
  88