spaCy/tests/test_span.py
2015-04-07 04:52:25 +02:00

24 lines
446 B
Python

from __future__ import unicode_literals
from spacy.en import English
from spacy.spans import Spans
import pytest
import re
EN = English()
@pytest.fixture
def doc():
return EN('This is a sentence. This is another sentence. And a third.')
def test_sent_spans(doc):
sents = list(doc.sents)
assert sents[0].start == 0
assert sents[0].end == 5
assert len(sents) == 3
assert sum(len(sent) for sent in sents) == len(doc)