def test_entities_that_would_overlap_keeps_longer_earlier_match(
nlp: Language, patterns: List[Dict[str, Any]], doc: Doc) -> None:
"""It matches the longest/earliest entities."""
ruler = SpaczzRuler(nlp, spaczz_patterns=patterns)
ruler.add_patterns([{"label": "TEST", "pattern": "Fake", "type": "fuzzy"}])
doc = ruler(doc)
assert "FAKE" not in [ent.label_ for ent in doc.ents]
def test_add_patterns_warns_if_spaczz_type_unrecognized(
nlp: Language, ) -> None:
"""It raises a ValueError if patterns not correct format."""
ruler = SpaczzRuler(nlp)
with pytest.warns(PatternTypeWarning):
ruler.add_patterns([{
"label": "GPE",
"pattern": "Montana",
"type": "invalid"
}])
def test_seeing_tokens_again(
nlp: Language, patterns: list[dict[str, Any]], doc: Doc
) -> None:
"""If ruler has already seen tokens, it ignores them."""
ruler = SpaczzRuler(nlp, spaczz_patterns=patterns)
ruler.add_patterns(
[{"label": "ADDRESS", "pattern": "122 Fake St, Apt 54", "type": "fuzzy"}]
)
doc = ruler(doc)
assert "ADDRESS" in [ent.label_ for ent in doc.ents]
def test_add_patterns_raises_error_pattern_not_iter_of_dict(
nlp: Language) -> None:
"""It raises a TypeError if pattern not iterable of dicts."""
ruler = SpaczzRuler(nlp)
with pytest.raises(TypeError):
ruler.add_patterns({"label": "GPE", "pattern": "Montana"})
def test_add_patterns_raises_error_if_not_spaczz_pattern(
nlp: Language, ) -> None:
"""It raises a ValueError if patterns not correct format."""
ruler = SpaczzRuler(nlp)
with pytest.raises(ValueError):
ruler.add_patterns([{"label": "GPE", "pattern": "Montana"}])
