mirror of
https://github.com/explosion/spaCy.git
synced 2024-12-25 09:26:27 +03:00
Add LEX_ATTRS
This commit is contained in:
parent
bca2ea9c72
commit
c13b3fa052
|
@ -3,6 +3,7 @@ from __future__ import unicode_literals
|
||||||
|
|
||||||
from .tokenizer_exceptions import TOKENIZER_EXCEPTIONS
|
from .tokenizer_exceptions import TOKENIZER_EXCEPTIONS
|
||||||
from .stop_words import STOP_WORDS
|
from .stop_words import STOP_WORDS
|
||||||
|
from .lex_attrs import LEX_ATTRS
|
||||||
from .lemmatizer import LOOKUP
|
from .lemmatizer import LOOKUP
|
||||||
|
|
||||||
from ..tokenizer_exceptions import BASE_EXCEPTIONS
|
from ..tokenizer_exceptions import BASE_EXCEPTIONS
|
||||||
|
@ -18,6 +19,7 @@ class Portuguese(Language):
|
||||||
class Defaults(Language.Defaults):
|
class Defaults(Language.Defaults):
|
||||||
lex_attr_getters = dict(Language.Defaults.lex_attr_getters)
|
lex_attr_getters = dict(Language.Defaults.lex_attr_getters)
|
||||||
lex_attr_getters[LANG] = lambda text: 'pt'
|
lex_attr_getters[LANG] = lambda text: 'pt'
|
||||||
|
lex_attr_getters.update(LEX_ATTRS)
|
||||||
|
|
||||||
tokenizer_exceptions = update_exc(BASE_EXCEPTIONS, TOKENIZER_EXCEPTIONS)
|
tokenizer_exceptions = update_exc(BASE_EXCEPTIONS, TOKENIZER_EXCEPTIONS)
|
||||||
stop_words = set(STOP_WORDS)
|
stop_words = set(STOP_WORDS)
|
||||||
|
|
Loading…
Reference in New Issue
Block a user