diff --git a/spacy/tests/regression/test_issue595.py b/spacy/tests/regression/test_issue595.py index 205d62bb1..97270c54f 100644 --- a/spacy/tests/regression/test_issue595.py +++ b/spacy/tests/regression/test_issue595.py @@ -1,12 +1,41 @@ +from __future__ import unicode_literals import pytest -import spacy +from ...symbols import POS, VERB, VerbForm_inf +from ...tokens import Doc +from ...vocab import Vocab +from ...lemmatizer import Lemmatizer -@pytest.mark.models -def test_not_lemmatize_base_forms(): - nlp = spacy.load('en', parser=False) - doc = nlp(u"Don't feed the dog") +@pytest.fixture +def index(): + return {'verb': {}} + +@pytest.fixture +def exceptions(): + return {'verb': {}} + +@pytest.fixture +def rules(): + return {"verb": [["ed", "e"]]} + +@pytest.fixture +def lemmatizer(index, exceptions, rules): + return Lemmatizer(index, exceptions, rules) + + +@pytest.fixture +def tag_map(): + return {'VB': {POS: VERB, 'morph': VerbForm_inf}} + + +@pytest.fixture +def vocab(lemmatizer, tag_map): + return Vocab(lemmatizer=lemmatizer, tag_map=tag_map) + + +def test_not_lemmatize_base_forms(vocab, lemmatizer): + doc = Doc(vocab, words=["Do", "n't", "feed", "the", "dog"]) feed = doc[2] feed.tag_ = u'VB' assert feed.text == u'feed'