class JaccardTestCase(unittest.TestCase):
def setUp(self):
self.threshold = 0.3
self.matches_using_cart_prod = sim_match(
table_A, table_B, tokenized_table_A, tokenized_table_B, l_attr,
r_attr, get_jaccard_fn(), self.threshold, ['id'], ['id'])
self.size_filter = SizeFilter(table_A, tokenized_table_A, l_attr, tok)
self.size_filter.build_index()
self.prefix_filter = PrefixFilter(table_A, tokenized_table_A, l_attr,
tok, self.threshold, token_ordering)
self.prefix_filter.build_index()
self.position_filter = PositionFilter(table_A, tokenized_table_A,
l_attr, tok, self.threshold,
token_ordering)
self.position_filter.build_index()
self.suffix_filter = SuffixFilter(table_A, tokenized_table_A, l_attr,
tok, self.threshold, token_ordering)
def test_jaccard_match(self):
# test jaccard with position filter, size filter, suffix filter
matches = jaccard_match(
table_A, table_B, tokenized_table_A, tokenized_table_B, l_attr,
r_attr, self.threshold,
[self.position_filter, self.size_filter, self.suffix_filter],
['id'], ['id'])
self.assertTrue(compare_matches(self.matches_using_cart_prod, matches))
# test jaccard with prefix filter, size filter, suffix filter
matches = jaccard_match(
table_A, table_B, tokenized_table_A, tokenized_table_B, l_attr,
r_attr, self.threshold,
[self.prefix_filter, self.size_filter, self.suffix_filter], ['id'],
['id'])
self.assertTrue(compare_matches(self.matches_using_cart_prod, matches))
class SizeFilterTestCase(unittest.TestCase):
def setUp(self):
self.size_filter = SizeFilter(A, A_tokenized, 'str', tok)
self.size_filter.build_index()
def test_apply_filter(self):
# size filter satisfies
l_tokens = ['aa', 'bb', 'cd', 'ef', 'fg']
r_tokens = ['xx', 'yy', 'aa', 'bb']
self.assertTrue(
self.size_filter.apply_filter(l_tokens, r_tokens, len(l_tokens),
len(r_tokens), 0.8))
# size filter doesn't satisfy
l_tokens = ['aa', 'bb', 'cd', 'ef', 'fg']
r_tokens = ['xx']
self.assertFalse(
self.size_filter.apply_filter(l_tokens, r_tokens, len(l_tokens),
len(r_tokens), 0.8))
# test empty list of tokens
l_tokens = ['aa', 'bb', 'cd', 'ef', 'fg']
r_tokens = []
self.assertFalse(
self.size_filter.apply_filter(l_tokens, r_tokens, len(l_tokens),
len(r_tokens), 0.8))
self.assertFalse(
self.size_filter.apply_filter(r_tokens, l_tokens, len(r_tokens),
len(l_tokens), 0.8))
def test_find_candidates(self):
# test default case (presence of candidates)
tokens = ['aa', 'xx', 'yy', 'uu']
self.assertSetEqual(
self.size_filter.find_candidates(tokens, len(tokens), 0.8),
set([0, 3, 4]))
# test empty set of candidates
tokens = ['aa', 'op', 'xx', 'yy', 'uu', 'yu', 'iu', 'lp']
self.assertSetEqual(
self.size_filter.find_candidates(tokens, len(tokens), 0.8), set())
# test empty list of probe tokens
tokens = []
self.assertSetEqual(
self.size_filter.find_candidates(tokens, len(tokens), 0.8), set())
