def test_sorted_iterable(self):
sorted_data = ['bar', 'foo', 'foobar']
d = simtrie.Set(sorted_data, sorted=True)
assert 'bar' in d
assert 'foo' in d
def test_unicode_sorting(self):
key1 = '\uff72\uff9c\uff90\uff7b\uff9e\uff9c'
key2 = '\U00010345\U0001033f\U00010337\U00010330\U0001033d'
# This apparently depends on Python version:
# assert key1 < key2
assert key1.encode('utf8') < key2.encode('utf8')
# Constructor should sort data according to utf8 values,
# not according to unicode sorting rules. It will raise an exception
# if data is sorted according to unicode rules.
simtrie.Set([key1, key2], sorted=True)
def test_contains_with_null_bytes(self):
d = simtrie.Set(['foo'])
assert b'foo' in d
assert b'foo\x00bar' not in d
def test_build_errors(self):
with pytest.raises(ValueError):
data = [b'foo\x00bar', b'bar']
simtrie.Set(data)
def test_no_segfaults_after_wrong_stream(self):
d = simtrie.Set()
assert 'random-key' not in d
def test_no_segfaults_on_invalid_file(self):
d = simtrie.Set()
with pytest.raises(IOError) as e:
d.frombytes(b"foo")
def _dam_lev(a, b, metric=None):
s = simtrie.Set([a])
for k, cost in s.similar(b, MAX_COST, metric=metric, allow_transpose=True):
assert k == a
return cost
def _lev(a, b, metric):
s = simtrie.Set([a])
for k, cost in s.similar(b, MAX_COST, metric):
assert k == a
return cost
def get_cost(a, b):
s = simtrie.Set([a])
for k, cost in s.similar(b, 10):
assert k == a
return cost
def test_no_segfaults_on_empty_dawg(self):
d = simtrie.Set([])
assert list(d.keys()) == []
def empty_dawg(self):
return simtrie.Set()
def dawg(self):
return simtrie.Set(self.keys)