fd9fd275c5209b64d163943da33cadeea01b247c,spacy/tests/regression/test_issue1945.py,,test_issue1945,#,9

Before Change

    nlp = English()
    matcher = PhraseMatcher(nlp.vocab)
    matcher.add("MWE", None, *[nlp.tokenizer(item) for item in mw_list])

    assert len(matcher(nlp(text))) == 1

After Change

    text = "a a a"
    matcher = Matcher(Vocab())
    matcher.add("MWE", None, [{"orth": "a"}, {"orth": "a"}])
    doc = Doc(matcher.vocab, words=["a", "a", "a"])
    matches = matcher(doc)
    // We should see two overlapping matches here
    assert len(matches) == 2
    assert matches[0][1:] == (0, 2)
    assert matches[1][1:] == (1, 3)

Instances