mirror of
https://github.com/explosion/spaCy.git
synced 2024-11-14 13:47:13 +03:00
709c9e75af
* return first sentence when span contains sentence boundary * docs fix * small fixes * cleanup
19 lines
644 B
Python
19 lines
644 B
Python
from spacy.lang.en import English
|
|
|
|
|
|
def test_issue7065():
|
|
text = "Kathleen Battle sang in Mahler 's Symphony No. 8 at the Cincinnati Symphony Orchestra 's May Festival."
|
|
nlp = English()
|
|
nlp.add_pipe("sentencizer")
|
|
ruler = nlp.add_pipe("entity_ruler")
|
|
patterns = [{"label": "THING", "pattern": [{"LOWER": "symphony"}, {"LOWER": "no"}, {"LOWER": "."}, {"LOWER": "8"}]}]
|
|
ruler.add_patterns(patterns)
|
|
|
|
doc = nlp(text)
|
|
sentences = [s for s in doc.sents]
|
|
assert len(sentences) == 2
|
|
sent0 = sentences[0]
|
|
ent = doc.ents[0]
|
|
assert ent.start < sent0.end < ent.end
|
|
assert sentences.index(ent.sent) == 0
|