src/Gargantext/Text/Terms/Mono.hs

   1 {-|
   2 Module      : Gargantext.Text.Terms.Mono
   3 Description : Mono Terms module
   4 Copyright   : (c) CNRS, 2017 - present
   5 License     : AGPL + CECILL v3
   6 Maintainer  : team@gargantext.org
   7 Stability   : experimental
   8 Portability : POSIX
   9
  10 Mono-terms are Nterms where n == 1.
  11
  12 -}
  13
  14 {-# LANGUAGE NoImplicitPrelude #-}
  15
  16 module Gargantext.Text.Terms.Mono (monoterms, monoterms')
  17   where
  18
  19 import Data.Text (Text, toLower, split, splitOn, pack)
  20 import qualified Data.Set as S
  21
  22 import Gargantext.Core
  23 import Gargantext.Core.Types
  24 import Gargantext.Text.Terms.Mono.Stem (stem)
  25
  26 import Gargantext.Prelude
  27 import Data.Char (isAlphaNum, isSpace)
  28
  29 monoterms' :: Lang -> Text -> [Terms]
  30 monoterms' l txt = map (text2terms l) $ monoterms txt
  31
  32 monoterms :: Text -> [Text]
  33 monoterms txt = map toLower $ split isWord txt
  34   where
  35     isWord c = c `elem` [' ', '\'', ',', ';']
  36
  37 text2terms :: Lang -> Text -> Terms
  38 text2terms lang txt = Terms label stems
  39   where
  40     label = splitOn (pack " ") txt
  41     stems = S.fromList $ map (stem lang) label
  42
  43   --monograms :: Text -> [Text]
  44 --monograms xs = monograms $ toLower $ filter isGram xs
  45
  46 isGram :: Char -> Bool
  47 isGram  c  = isAlphaNum c || isSpace c || c `elem` ['-','/','\'']
  48