mirror of https://github.com/explosion/spaCy.git
23 lines
818 B
Python
23 lines
818 B
Python
# coding: utf8
|
|
from __future__ import unicode_literals
|
|
|
|
from spacy.language_data import strings_to_exc, update_exc
|
|
from .punctuation import *
|
|
from .stop_words import STOP_WORDS
|
|
from .tokenizer_exceptions import ABBREVIATIONS
|
|
from .tokenizer_exceptions import OTHER_EXC
|
|
from .. import language_data as base
|
|
|
|
STOP_WORDS = set(STOP_WORDS)
|
|
|
|
TOKENIZER_EXCEPTIONS = strings_to_exc(base.EMOTICONS)
|
|
update_exc(TOKENIZER_EXCEPTIONS, strings_to_exc(base.ABBREVIATIONS))
|
|
update_exc(TOKENIZER_EXCEPTIONS, strings_to_exc(OTHER_EXC))
|
|
update_exc(TOKENIZER_EXCEPTIONS, strings_to_exc(ABBREVIATIONS))
|
|
|
|
TOKENIZER_PREFIXES = TOKENIZER_PREFIXES
|
|
TOKENIZER_SUFFIXES = TOKENIZER_SUFFIXES
|
|
TOKENIZER_INFIXES = TOKENIZER_INFIXES
|
|
|
|
__all__ = ["TOKENIZER_EXCEPTIONS", "STOP_WORDS", "TOKENIZER_PREFIXES", "TOKENIZER_SUFFIXES", "TOKENIZER_INFIXES"]
|