mirror of
https://github.com/explosion/spaCy.git
synced 2024-11-11 20:28:20 +03:00
0b44ea23bd
* modifying FR lemmatization for nouns * modifying FR lemmatization for nouns * adding contributor agreement for amperinet * adding rules for words with inclusive parentheses wrongly tokenized * adding contributor agreement for amperinet * adding a missing comma
147 lines
2.5 KiB
Python
147 lines
2.5 KiB
Python
# coding: utf8
|
|
from __future__ import unicode_literals
|
|
|
|
|
|
ADJECTIVE_RULES = [
|
|
["s", ""],
|
|
["e", ""],
|
|
["es", ""]
|
|
]
|
|
|
|
|
|
NOUN_RULES = [
|
|
["a", "a"],
|
|
["à", "à"],
|
|
["â", "â"],
|
|
["b", "b"],
|
|
["c", "c"],
|
|
["ç", "ç"],
|
|
["d", "d"],
|
|
["e", "e"],
|
|
["é", "é"],
|
|
["è", "è"],
|
|
["ê", "ê"],
|
|
["ë", "ë"],
|
|
["f", "f"],
|
|
["g", "g"],
|
|
["h", "h"],
|
|
["i", "i"],
|
|
["î", "î"],
|
|
["ï", "ï"],
|
|
["j", "j"],
|
|
["k", "k"],
|
|
["l", "l"],
|
|
["m", "m"],
|
|
["n", "n"],
|
|
["o", "o"],
|
|
["ô", "ö"],
|
|
["ö", "ö"],
|
|
["p", "p"],
|
|
["q", "q"],
|
|
["r", "r"],
|
|
["t", "t"],
|
|
["u", "u"],
|
|
["û", "û"],
|
|
["v", "v"],
|
|
["w", "w"],
|
|
["y", "y"],
|
|
["z", "z"],
|
|
|
|
["as", "a"],
|
|
["aux", "au"],
|
|
["cs", "c"],
|
|
["chs", "ch"],
|
|
["ds", "d"],
|
|
["és", "é"],
|
|
["es", "e"],
|
|
["eux", "eu"],
|
|
["fs", "f"],
|
|
["gs", "g"],
|
|
["hs", "h"],
|
|
["is", "i"],
|
|
["ïs", "ï"],
|
|
["js", "j"],
|
|
["ks", "k"],
|
|
["ls", "l"],
|
|
["ms", "m"],
|
|
["ns", "n"],
|
|
["oux", "ou"],
|
|
["os", "o"],
|
|
["ps", "p"],
|
|
["qs", "q"],
|
|
["rs", "r"],
|
|
["ses", "se"],
|
|
["se", "se"],
|
|
["ts", "t"],
|
|
["us", "u"],
|
|
["vs", "v"],
|
|
["ws", "w"],
|
|
["ys", "y"],
|
|
|
|
["nt(e", "nt"],
|
|
["nt(e)", "nt"],
|
|
["al(e", "ale"],
|
|
["é(", "é"],
|
|
["é(e", "é"],
|
|
["é.e", "é"],
|
|
["el(le", "el"],
|
|
["eurs(rices", "eur"],
|
|
["eur(rice", "eur"],
|
|
["eux(se", "eux"],
|
|
["ial(e","ial"],
|
|
["er(ère", "er"],
|
|
["eur(se", "eur"],
|
|
["teur(trice", "teur"],
|
|
["teurs(trices", "teur"]
|
|
]
|
|
|
|
|
|
VERB_RULES = [
|
|
["é", "er"],
|
|
["és", "er"],
|
|
["ée", "er"],
|
|
["ées", "er"],
|
|
["é", "er"],
|
|
["es", "er"],
|
|
["ons", "er"],
|
|
["ez", "er"],
|
|
["ent", "er"],
|
|
["ais", "er"],
|
|
["ait", "er"],
|
|
["ions", "er"],
|
|
["iez", "er"],
|
|
["aient", "er"],
|
|
["ai", "er"],
|
|
["as", "er"],
|
|
["a", "er"],
|
|
["âmes", "er"],
|
|
["âtes", "er"],
|
|
["èrent", "er"],
|
|
["erai", "er"],
|
|
["eras", "er"],
|
|
["era", "er"],
|
|
["erons", "er"],
|
|
["erez", "er"],
|
|
["eront", "er"],
|
|
["erais", "er"],
|
|
["erait", "er"],
|
|
["erions", "er"],
|
|
["eriez", "er"],
|
|
["eraient", "er"],
|
|
["asse", "er"],
|
|
["asses", "er"],
|
|
["ât", "er"],
|
|
["assions", "er"],
|
|
["assiez", "er"],
|
|
["assent", "er"],
|
|
["ant", "er"],
|
|
["ante", "er"],
|
|
["ants", "er"],
|
|
["antes", "er"],
|
|
|
|
["u(er", "u"],
|
|
["és(ées", "er"],
|
|
["é()e", "er"],
|
|
["é()", "er"]
|
|
]
|