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, here are logic of main types.
  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 ------------------------------------------------------------------------
  29
  30 type Term = Text
  31 type Label = Term
  32
  33 type TermList = [(Label, [[Term]])]
  34
  35 type Sentence  a = [a] -- or a nominal group
  36 type Corpus    a = [Sentence a] -- a list of sentences
  37
  38 -- type ConText a = [Sentence a]
  39 -- type Corpus a = [ConText a]
  40
  41
  42 ------------------------------------------------------------------------
  43
  44 data SplitContext = Chars Int | Sentences Int | Paragraphs Int
  45
  46 tag :: Text -> [Tag Text]
  47 tag = parseTags
  48
  49 -- | splitBy contexts of Chars or Sentences or Paragraphs
  50 -- >> splitBy (Chars 0) "abcde"
  51 -- ["a","b","c","d","e"]
  52 -- >> splitBy (Chars 1) "abcde"
  53 -- ["ab","bc","cd","de"]
  54 -- >> splitBy (Chars 2) "abcde"
  55 -- ["abc","bcd","cde"]
  56 splitBy :: SplitContext -> Text -> [Text]
  57 splitBy (Chars     n)  = map pack        . chunkAlong (n+1) 1 . unpack
  58 splitBy (Sentences n)  = map unsentences . chunkAlong (n+1) 1 . sentences
  59 splitBy (Paragraphs _) = map unTag       . filter isTagText   . tag
  60   where
  61     unTag :: IsString p => Tag p -> p
  62     unTag (TagText x) = x
  63     unTag _           = ""
  64
  65