def test_random_suffix_links(self):
T, n, A = 100, 200, ['a', 'b']
for _ in range(T):
t = rand.random_word(n, A)
reference = TestSuffixTrees.get_suffix_links(
*suffix_tree.mccreight(t, n))
self.check_suffix_links(t, n, reference)
def check_suffix_array(self, t, n, reference):
for algorithm in SUFFIX_ARRAY_ALGORITHMS:
self.assertEqual(algorithm(t, n), reference,
'Algorithm: {0}'.format(algorithm.__name__))
self.assertEqual(
suffix_array.from_suffix_tree(suffix_tree.mccreight(t, n)[0], n),
reference, 'Suffix tree from suffix array')
def test_random_suffix_tree(self, _, algorithm, links):
T, n, A = 100, 200, ['a', 'b']
for _ in range(T):
t = rand.random_word(n, A)
reference = (suffix_tree.naive(t, n),
get_suffix_links(*suffix_tree.mccreight(t, n))[1])
self.check_suffix_trees(t, n, reference, algorithm, links)
def test_all_suffix_links(self):
N, A = 10, ['a', 'b']
for n in range(2, N + 1):
for t in itertools.product(A, repeat = n):
t = '#' + ''.join(t)
reference = TestSuffixTrees.get_suffix_links(
*suffix_tree.mccreight(t, n))
self.check_suffix_links(t, n, reference)
def test_all_suffix_tree(self, _, algorithm, links):
N, A = 10, ['a', 'b']
for n in range(2, N + 1):
for t in itertools.product(A, repeat=n):
t = '#' + ''.join(t)
reference = (suffix_tree.naive(t, n),
get_suffix_links(*suffix_tree.mccreight(t, n))[1])
self.check_suffix_trees(t, n, reference, algorithm, links)
def check_suffix_links(self, t, n, reference):
self.assertEqual(
TestSuffixTrees.get_suffix_links(*suffix_tree.mccreight(t, n)),
reference, 'Algorithm: mccreight')
self.assertEqual(
TestSuffixTrees.get_suffix_links(*suffix_tree.ukkonen(t, n)),
reference, 'Algorithm: ukkonen')
self.assertEqual(
TestSuffixTrees.get_backward_suffix_links(
*suffix_tree.weiner(t, n)), reference, 'Algorithm: weiner')
def check_lcp_array(self, t, n, reference):
self.assertEqual(
suffix_array.lcp_from_suffix_array(
suffix_array.prefix_doubling(t, n), t, n), reference,
'LCP array from suffix array')
self.assertEqual(
suffix_array.lcp_from_suffix_tree(suffix_tree.mccreight(t, n)[0]),
reference, 'LCP array from suffix tree')
self.assertEqual(
suffix_array.lcp_kasai(suffix_array.prefix_doubling(t, n), t, n),
reference, 'Algorithm: kasai')
def fast_on_average(text, word, n, m):
ST, _ = suffix_tree.mccreight(word, m)
i, r = m, min(2 * math.ceil(math.log(m, 2)), m - 1)
while i <= n:
if ST.find_node(text[(i - r):(i + 1)], r + 1) is not None:
subtext = text[0] + text[(i - m + 1):(i + m - r)]
subn = min(i + m - r, n + 1) - (i - m + 1)
yield from [
v + i - m
for v in forward.knuth_morris_pratt(subtext, word, subn, m)
]
i = i + m - r
['Boyer-Moore-Galil', backward.boyer_moore_galil],
['Turbo-Boyer-Moore', backward.turbo_boyer_moore],
['bad shift heuristic', backward.bad_shift_heuristic],
['quick search heuristic', backward.quick_search],
[
'Boyer-Moore-Apostolico-Giancarlo',
backward.boyer_moore_apostolico_giancarlo
],
['Horspool', backward.horspool],
['Karp-Rabin', other.karp_rabin],
['fast-on-average', other.fast_on_average],
['two-way constant space', other.two_way],
[
'suffix tree',
lambda t, w, n, m: suffix_tree.contains(
suffix_tree.mccreight(t, n)[0], t, w, n, m),
],
[
'suffix array',
lambda t, w, n, m: suffix_array.contains(
suffix_array.prefix_doubling(t, n), t, w, n, m),
],
['lcp-lr array', lcp_lr_contains],
]
class TestExactStringMatching(unittest.TestCase):
run_large = unittest.skipUnless(os.environ.get('LARGE', False),
'Skip test in small runs')
def check_first_exact_match(self, t, w, n, m, reference, algorithm):
from generator import rand
from string_indexing import farach, lcp, suffix_array, suffix_tree
LCP_ARRAY_ALGORITHMS = [
[
'Farach',
farach.lcp_array,
],
[
'Kasai',
lambda t, n: lcp.kasai(suffix_array.prefix_doubling(t, n), t, n),
],
[
'from suffix tree',
lambda t, n: lcp.from_suffix_tree(suffix_tree.mccreight(t, n)[0]),
],
[
'from suffix array',
lambda t, n: lcp.from_suffix_array(suffix_array.prefix_doubling(t, n),
t, n),
]
]
class TestLcpArrays(unittest.TestCase):
run_large = unittest.skipUnless(os.environ.get('LARGE', False),
'Skip test in small runs')
def check_lcp_array(self, t, n, reference, algorithm):
self.assertEqual(algorithm(t, n), reference)
import parameterized
from generator import rand
from string_indexing import farach, suffix_array, suffix_tree
SUFFIX_ARRAY_ALGORITHMS = [[
'Karp-Miller-Rosenberg', suffix_array.prefix_doubling
], ['Karkkainen-Sanders', suffix_array.skew], ['Farach', farach.suffix_array],
['Larsson-Sadakane', suffix_array.larsson_sadakane],
['Ko-Aluru', suffix_array.small_large],
['Zhang-Nong-Chan', suffix_array.induced_sorting],
[
'from suffix tree',
lambda t, n: suffix_array.from_suffix_tree(
suffix_tree.mccreight(t, n)[0], n),
]]
class TestSuffixArrays(unittest.TestCase):
run_large = unittest.skipUnless(os.environ.get('LARGE', False),
'Skip test in small runs')
def check_suffix_array(self, t, n, reference, algorithm):
self.assertEqual(algorithm(t, n), reference,
'Algorithm {}, text {}'.format(algorithm.__name__, t))
@parameterized.parameterized.expand(SUFFIX_ARRAY_ALGORITHMS)
def test_examples(self, _, algorithm):
self.check_suffix_array('#abaaba', 6, [7, 6, 3, 4, 1, 5, 2], algorithm)
self.check_suffix_array('#banana', 6, [7, 6, 4, 2, 1, 5, 3], algorithm)
