spaCy/spacy/lang/pt/__init__.py
Filipe Caixeta 6c498f9ff4 Update Portuguese Language (#2790)
* Add words to portuguese language _num_words

* Add words to portuguese language _num_words

* Portuguese - Add/remove stopwords, fix tokenizer, add currency symbols

* Extended punctuation and norm_exceptions in the Portuguese language
2018-09-29 09:51:45 +02:00

37 lines
1.2 KiB
Python

# coding: utf8
from __future__ import unicode_literals
from .tokenizer_exceptions import TOKENIZER_EXCEPTIONS
from .stop_words import STOP_WORDS
from .lex_attrs import LEX_ATTRS
from .lemmatizer import LOOKUP
from .tag_map import TAG_MAP
from .norm_exceptions import NORM_EXCEPTIONS
from ..tokenizer_exceptions import BASE_EXCEPTIONS
from .punctuation import TOKENIZER_INFIXES, TOKENIZER_PREFIXES
from ..norm_exceptions import BASE_NORMS
from ...language import Language
from ...attrs import LANG, NORM
from ...util import update_exc, add_lookups
class PortugueseDefaults(Language.Defaults):
lex_attr_getters = dict(Language.Defaults.lex_attr_getters)
lex_attr_getters[LANG] = lambda text: 'pt'
lex_attr_getters[NORM] = add_lookups(Language.Defaults.lex_attr_getters[NORM], BASE_NORMS, NORM_EXCEPTIONS)
lex_attr_getters.update(LEX_ATTRS)
tokenizer_exceptions = update_exc(BASE_EXCEPTIONS, TOKENIZER_EXCEPTIONS)
stop_words = STOP_WORDS
lemma_lookup = LOOKUP
tag_map = TAG_MAP
infixes = TOKENIZER_INFIXES
prefixes = TOKENIZER_PREFIXES
class Portuguese(Language):
lang = 'pt'
Defaults = PortugueseDefaults
__all__ = ['Portuguese']