spaCy/spacy/lang/pl/__init__.py

25 lines
631 B
Python
Raw Normal View History

2017-05-12 07:25:37 +00:00
# coding: utf8
from __future__ import unicode_literals
2017-06-26 21:38:28 +00:00
from .tokenizer_exceptions import TOKENIZER_EXCEPTIONS
2017-05-12 07:25:37 +00:00
from .stop_words import STOP_WORDS
from ..tokenizer_exceptions import BASE_EXCEPTIONS
from ...language import Language
from ...attrs import LANG
from ...util import update_exc
class Polish(Language):
lang = 'pl'
class Defaults(Language.Defaults):
lex_attr_getters = dict(Language.Defaults.lex_attr_getters)
lex_attr_getters[LANG] = lambda text: 'pl'
2017-06-26 21:38:28 +00:00
tokenizer_exceptions = update_exc(BASE_EXCEPTIONS, TOKENIZER_EXCEPTIONS)
2017-05-12 07:25:37 +00:00
stop_words = set(STOP_WORDS)
__all__ = ['Polish']