spaCy/spacy/tests/regression/test_issue636.py
2017-03-13 12:08:31 +01:00

18 lines
506 B
Python

# coding: utf8
from __future__ import unicode_literals
from ...tokens.doc import Doc
import pytest
@pytest.mark.xfail
@pytest.mark.models
@pytest.mark.parametrize('text', ["I cant do this."])
def test_issue636(EN, text):
"""Test that to_bytes and from_bytes don't change the token lemma."""
doc1 = EN(text)
doc2 = Doc(EN.vocab)
doc2.from_bytes(doc1.to_bytes())
print([t.lemma_ for t in doc1], [t.lemma_ for t in doc2])
assert [t.lemma_ for t in doc1] == [t.lemma_ for t in doc2]