spaCy/spacy/lang/bn/__init__.py

36 lines
969 B
Python
Raw Normal View History

2017-03-12 15:07:28 +03:00
# coding: utf8
2017-05-08 16:43:50 +03:00
from __future__ import unicode_literals
2017-02-28 04:48:37 +03:00
2017-05-08 16:43:50 +03:00
from .tokenizer_exceptions import TOKENIZER_EXCEPTIONS
from .punctuation import TOKENIZER_PREFIXES, TOKENIZER_SUFFIXES, TOKENIZER_INFIXES
from .tag_map import TAG_MAP
from .stop_words import STOP_WORDS
from .lemmatizer import LEMMA_RULES
2017-05-08 23:29:04 +03:00
from ..tokenizer_exceptions import BASE_EXCEPTIONS
from ...language import Language
from ...attrs import LANG
from ...util import update_exc
2017-02-28 04:48:37 +03:00
class BengaliDefaults(Language.Defaults):
lex_attr_getters = dict(Language.Defaults.lex_attr_getters)
lex_attr_getters[LANG] = lambda text: 'bn'
tokenizer_exceptions = update_exc(BASE_EXCEPTIONS, TOKENIZER_EXCEPTIONS)
tag_map = TAG_MAP
stop_words = STOP_WORDS
lemma_rules = LEMMA_RULES
2017-02-28 04:48:37 +03:00
prefixes = tuple(TOKENIZER_PREFIXES)
suffixes = tuple(TOKENIZER_SUFFIXES)
infixes = tuple(TOKENIZER_INFIXES)
2017-02-28 04:48:37 +03:00
2017-02-28 18:07:14 +03:00
class Bengali(Language):
lang = 'bn'
Defaults = BengaliDefaults
2017-05-03 12:01:42 +03:00
2017-05-08 16:43:50 +03:00
__all__ = ['Bengali']