spaCy/spacy/lang/nl/__init__.py

33 lines
853 B
Python
Raw Normal View History

2017-03-12 15:07:28 +03:00
# coding: utf8
2017-05-08 16:51:39 +03:00
from __future__ import unicode_literals
2017-05-08 16:51:39 +03:00
from .stop_words import STOP_WORDS
from .lex_attrs import LEX_ATTRS
2017-11-05 15:48:07 +03:00
from .tag_map import TAG_MAP
2017-05-08 16:51:39 +03:00
2017-05-08 23:29:04 +03:00
from ..tokenizer_exceptions import BASE_EXCEPTIONS
from ..norm_exceptions import BASE_NORMS
2017-05-08 23:29:04 +03:00
from ...language import Language
from ...attrs import LANG, NORM
from ...util import update_exc, add_lookups
2017-05-08 16:51:39 +03:00
class DutchDefaults(Language.Defaults):
lex_attr_getters = dict(Language.Defaults.lex_attr_getters)
lex_attr_getters.update(LEX_ATTRS)
lex_attr_getters[LANG] = lambda text: "nl"
lex_attr_getters[NORM] = add_lookups(
Language.Defaults.lex_attr_getters[NORM], BASE_NORMS
)
tokenizer_exceptions = update_exc(BASE_EXCEPTIONS)
stop_words = STOP_WORDS
2017-11-05 15:48:07 +03:00
tag_map = TAG_MAP
class Dutch(Language):
lang = "nl"
Defaults = DutchDefaults
2017-05-03 12:01:42 +03:00
__all__ = ["Dutch"]