src/Gargantext/Ngrams/Lang/Fr.hs

   1 {-|
   2 Module      : Gargantext.Ngrams.Lang.Fr
   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.Ngrams.Lang.Fr (selectNgrams, groupNgrams, textTest)
  18     where
  19
  20 import Gargantext.Prelude
  21 import Data.Text (Text)
  22 import Data.Monoid ((<>))
  23
  24 selectNgrams :: [(Text, Text, Text)] -> [(Text, Text, Text)]
  25 selectNgrams xs = filter selectNgrams' xs
  26     where
  27         selectNgrams' (_,"N"    ,_       ) = True
  28         selectNgrams' (_,"NC"   ,_       ) = True
  29         selectNgrams' (_,"NN+CC",_       ) = True
  30         selectNgrams' (_,_      ,"PERSON"      ) = True
  31         selectNgrams' (_,_      ,"ORGANIZATION") = True
  32         selectNgrams' (_,_      ,"LOCATION"    ) = True
  33         selectNgrams' (_,_      ,_       ) = False
  34
  35
  36 groupNgrams :: [(Text, Text, Text)] -> [(Text, Text, Text)]
  37 groupNgrams []       = []
  38
  39 --groupNgrams ((_,"DET",_):xs) = groupNgrams xs
  40
  41 -- "Groupe : nom commun et adjectifs avec conjonction"
  42 groupNgrams ((n,"NC",n'):(j1,"ADJ",_):(_,"CC",_):(j2,"ADJ",_):xs) = groupNgrams (n1:n2:xs)
  43     where
  44         n1 = (n <> " " <> j1, "NC", n')
  45         n2 = (n <> " " <> j2, "NC", n')
  46
  47 -- /!\ sometimes N instead of NC (why?)
  48 groupNgrams ((n,"N",n'):(j1,"ADJ",_):(_,"CC",_):(j2,"ADJ",_):xs) = groupNgrams (n1:n2:xs)
  49     where
  50         n1 = (n <> " " <> j1, "N", n')
  51         n2 = (n <> " " <> j2, "N", n')
  52
  53 -- Groupe : Adjectif + Conjonction de coordination + Adjectif
  54 -- groupNgrams ((j1,"ADJ",_):(_,"CC",_):(j2,"ADJ",j2'):xs) = groupNgrams ((j1 <> " " <> j2, "ADJ", j2'):xs)
  55
  56 -- Groupe : Nom commun + préposition + Nom commun
  57 groupNgrams ((n1,"NC",_):(p,"P",_):(n2,"NC",n2'):xs)  = groupNgrams ((n1 <> " " <> p <> " " <> n2, "NC", n2'):xs)
  58 groupNgrams ((n1,"NC",_):(p,"P",_):(n2,"NPP",n2'):xs) = groupNgrams ((n1 <> " " <> p <> " " <> n2, "NC", n2'):xs)
  59 groupNgrams ((n1,"NC",_):(prep,"P",_):(det,"DET",_):(n2,"NPP",n2'):xs) = groupNgrams ((n1 <> " " <> prep <> " " <> det <> " " <> n2, "NC", n2'):xs)
  60
  61 -- Groupe : Plusieurs adjectifs successifs
  62 groupNgrams ((x,"ADJ",_):(y,"ADJ",yy):xs) = groupNgrams ((x <> " " <> y, "ADJ", yy):xs)
  63
  64 -- Groupe : nom commun et adjectif
  65 groupNgrams ((x,"NC",_):(y,"ADJ",yy):xs)  = groupNgrams ((x <> " " <> y, "NC", yy):xs)
  66 -- /!\ sometimes N instead of NC (why?)
  67 groupNgrams ((x,"N",_):(y,"ADJ",yy):xs)   = groupNgrams ((x <> " " <> y, "NC", yy):xs)
  68
  69 -- Groupe : adjectif et nom commun
  70 groupNgrams ((x,"ADJ",_):(y,"NC",yy):xs)  = groupNgrams ((x <> " " <> y, "NC", yy):xs)
  71 -- /!\ sometimes N instead of NC (why?)
  72 groupNgrams ((x,"ADJ",_):(y,"N",yy):xs)   = groupNgrams ((x <> " " <> y, "NC", yy):xs)
  73
  74
  75 groupNgrams ((x,_,"PERSON"):(y,yy,"PERSON"):xs)             = groupNgrams ((x <> " " <> y,yy,"PERSON"):xs)
  76 groupNgrams ((x,_,"ORGANIZATION"):(y,yy,"ORGANIZATION"):xs) = groupNgrams ((x <> " " <> y,yy,"ORGANIZATION"):xs)
  77
  78
  79 -- Si aucune des règles précédentes n'est remplie
  80 groupNgrams (x:xs)                        = (x:(groupNgrams xs))
  81
  82
  83 textTest :: [Text]
  84 textTest = [ "L'heure d'arrivée des coureurs dépend de la météo du jour."]
  85