explosion/spaCy
1
1
Fork 0
mirror of https://github.com/explosion/spaCy.git synced 2024-12-25 17:36:30 +03:00
Code Issues Packages Projects Releases Wiki Activity

Add infixes and abbreviation exceptions (fr)

Browse Source
This commit is contained in:
Raphaël Bournhonesque 2017-01-24 09:51:29 +01:00
parent cf8474401b
commit 1faaf698ca
1 changed files with 51 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

51
spacy/fr/language_data.py
View File

@ -4,6 +4,9 @@ from __future__ import unicode_literals
from .. import language_data as base
from ..language_data import strings_to_exc, update_exc
from .punctuation import ELISION
from ..symbols import *
from .stop_words import STOP_WORDS
@ -13,5 +16,53 @@ STOP_WORDS = set(STOP_WORDS)
TOKENIZER_EXCEPTIONS = strings_to_exc(base.EMOTICONS)
update_exc(TOKENIZER_EXCEPTIONS, strings_to_exc(base.ABBREVIATIONS))
ABBREVIATIONS = {
"janv.": [
{LEMMA: "janvier", ORTH: "janv."}
],
"févr.": [
{LEMMA: "février", ORTH: "févr."}
],
"avr.": [
{LEMMA: "avril", ORTH: "avr."}
],
"juill.": [
{LEMMA: "juillet", ORTH: "juill."}
],
"sept.": [
{LEMMA: "septembre", ORTH: "sept."}
],
"oct.": [
{LEMMA: "octobre", ORTH: "oct."}
],
"nov.": [
{LEMMA: "novembre", ORTH: "nov."}
],
"déc.": [
{LEMMA: "décembre", ORTH: "déc."}
],
}
INFIXES_EXCEPTIONS_BASE = ["aujourd'hui",
"prud'homme", "prud'hommes",
"prud'homal", "prud'homaux", "prud'homale",
"prud'homales",
"prud'hommal", "prud'hommaux", "prud'hommale",
"prud'hommales",
"prud'homie", "prud'homies",
"prud'hommesque", "prud'hommesques",
"prud'hommesquement"]
INFIXES_EXCEPTIONS = []
for elision_char in ELISION:
INFIXES_EXCEPTIONS += [infix.replace("'", elision_char)
for infix in INFIXES_EXCEPTIONS_BASE]
INFIXES_EXCEPTIONS += [word.capitalize() for word in INFIXES_EXCEPTIONS]
update_exc(TOKENIZER_EXCEPTIONS, strings_to_exc(INFIXES_EXCEPTIONS))
update_exc(TOKENIZER_EXCEPTIONS, ABBREVIATIONS)
__all__ = ["TOKENIZER_EXCEPTIONS", "STOP_WORDS"]