# 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]