def test1_smallset(self):
path_input = "test/data/schema_matching/default_matches_cities_input.csv"
df = pd.read_csv(path_input)
pairs_relational = relational_matching(df)[["uri_1","uri_2"]].sort_values(by=["uri_1","uri_2"]).reset_index(drop=True)
pairs_string = string_similarity_matching(df)[["uri_1","uri_2"]].sort_values(by=["uri_1","uri_2"]).reset_index(drop=True)
pairs_schema = label_schema_matching(df)[["uri_1","uri_2"]].sort_values(by=["uri_1","uri_2"]).reset_index(drop=True)
pairs_overlap = value_overlap_matching(df)[["uri_1","uri_2"]].sort_values(by=["uri_1","uri_2"]).reset_index(drop=True)
assert all([pairs_relational.equals(x) for x in [pairs_string, pairs_schema, pairs_overlap]])
def test3_diffpredicate_diffmetric(self):
path_input = "test/data/schema_matching/string_matching_input_t3.csv"
df = pd.read_csv(path_input)
path_expected = "test/data/schema_matching/string_matching_output_t3.csv"
result_expected = pd.read_csv(path_expected)
result = string_similarity_matching(df, predicate="dbo:abstract", to_lowercase=False, remove_prefixes=False, remove_punctuation=False, similarity_metric="token_set_levenshtein")
pd.testing.assert_frame_equal(result, result_expected, check_like=True)
def test2_highthreshold(self):
path_input = "test/data/schema_matching/string_matching_input_t1t2.csv"
df = pd.read_csv(path_input)
path_expected = "test/data/schema_matching/string_matching_output_t2.csv"
result_expected = pd.read_csv(path_expected)
result = string_similarity_matching(df, prefix_threshold=10)
pd.testing.assert_frame_equal(result, result_expected, check_like=True)
def test2_bigset(self):
#WARNING: Takes long to run!
path_input = "test/data/schema_matching/pair_equality_test2_bigset.csv"
df = pd.read_csv(path_input)
pairs_relational = relational_matching(df)[["uri_1","uri_2"]].sort_values(by=["uri_1","uri_2"]).reset_index(drop=True)
pairs_string = string_similarity_matching(df)[["uri_1","uri_2"]].sort_values(by=["uri_1","uri_2"]).reset_index(drop=True)
pairs_schema = label_schema_matching(df)[["uri_1","uri_2"]].sort_values(by=["uri_1","uri_2"]).reset_index(drop=True)
pairs_overlap = value_overlap_matching(df)[["uri_1","uri_2"]].sort_values(by=["uri_1","uri_2"]).reset_index(drop=True)
assert all([pairs_relational.equals(x) for x in [pairs_string, pairs_schema, pairs_overlap]])