spaCy/spacy/tests/regression/test_issue736.py

13 lines
393 B
Python
Raw Normal View History

2017-01-12 10:12:44 +00:00
# coding: utf-8
from __future__ import unicode_literals
import pytest
@pytest.mark.parametrize('text,number', [("7am", "7"), ("11p.m.", "11")])
def test_issue736(en_tokenizer, text, number):
2017-01-12 21:56:20 +00:00
"""Test that times like "7am" are tokenized correctly and that numbers are converted to string."""
2017-01-12 10:12:44 +00:00
tokens = en_tokenizer(text)
assert len(tokens) == 2
assert tokens[0].text == number