src/Gargantext/Text/Context.hs

   1 {-|
   2 Module      : Gargantext.Text.Context
   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 Context of text management tool
  11
  12 -}
  13
  14 {-# LANGUAGE NoImplicitPrelude #-}
  15 {-# LANGUAGE OverloadedStrings #-}
  16
  17 module Gargantext.Text.Context
  18   where
  19
  20 import Data.Text (Text, pack, unpack)
  21 import Data.String (IsString)
  22
  23 import Text.HTML.TagSoup (parseTags, isTagText, Tag(..))
  24 import Gargantext.Text
  25 import Gargantext.Prelude hiding (length)
  26
  27
  28 data SplitContext = Chars Int | Sentences Int | Paragraphs Int
  29
  30 tag :: Text -> [Tag Text]
  31 tag = parseTags
  32
  33 -- | splitBy contexts of Chars or Sentences or Paragraphs
  34 -- >> splitBy (Chars 0) "abcde"
  35 -- ["a","b","c","d","e"]
  36 -- >> splitBy (Chars 1) "abcde"
  37 -- ["ab","bc","cd","de"]
  38 -- >> splitBy (Chars 2) "abcde"
  39 -- ["abc","bcd","cde"]
  40 splitBy :: SplitContext -> Text -> [Text]
  41 splitBy (Chars     n)  = map pack        . chunkAlong (n+1) 1 . unpack
  42 splitBy (Sentences n)  = map unsentences . chunkAlong (n+1) 1 . sentences
  43 splitBy (Paragraphs _) = map unTag       . filter isTagText   . tag
  44   where
  45     unTag :: IsString p => Tag p -> p
  46     unTag (TagText x) = x
  47     unTag _           = ""
  48
  49