mirror of
https://github.com/explosion/spaCy.git
synced 2024-09-21 19:39:13 +03:00
44 lines
791 B
Python
44 lines
791 B
Python
"""Test that tokens are created correctly for whitespace."""
|
|
from __future__ import unicode_literals
|
|
|
|
from spacy.en import English
|
|
import pytest
|
|
|
|
|
|
@pytest.fixture
|
|
def EN():
|
|
return English()
|
|
|
|
|
|
def test_single_space(EN):
|
|
tokens = EN('hello possums')
|
|
assert len(tokens) == 2
|
|
|
|
|
|
def test_double_space(EN):
|
|
tokens = EN('hello possums')
|
|
assert len(tokens) == 3
|
|
assert tokens[1].string == ' '
|
|
|
|
|
|
def test_newline(EN):
|
|
tokens = EN('hello\npossums')
|
|
assert len(tokens) == 3
|
|
|
|
|
|
def test_newline_space(EN):
|
|
tokens = EN('hello \npossums')
|
|
assert len(tokens) == 3
|
|
|
|
|
|
def test_newline_double_space(EN):
|
|
tokens = EN('hello \npossums')
|
|
assert len(tokens) == 3
|
|
|
|
|
|
def test_newline_space_wrap(EN):
|
|
tokens = EN('hello \n possums')
|
|
assert len(tokens) == 3
|
|
|
|
|