mirror of
https://github.com/explosion/spaCy.git
synced 2024-11-11 12:18:04 +03:00
15 lines
410 B
Python
15 lines
410 B
Python
# coding: utf-8
|
|
from __future__ import unicode_literals
|
|
|
|
import pytest
|
|
|
|
|
|
@pytest.mark.parametrize(
|
|
"text,norms",
|
|
[("пн.", ["понедельник"]), ("пт.", ["пятница"]), ("дек.", ["декабрь"])],
|
|
)
|
|
def test_ru_tokenizer_abbrev_exceptions(ru_tokenizer, text, norms):
|
|
tokens = ru_tokenizer(text)
|
|
assert len(tokens) == 1
|
|
assert [token.norm_ for token in tokens] == norms
|