def run(argv):
if not should_replace(cpra_to_rsids_trie_fname) and not should_replace(
rsid_to_cpra_trie_fname):
print('tries are up-to-date!')
else:
with open(os.path.join(conf.data_dir, 'sites', 'sites.tsv'),
'rt') as f:
lines = [parse_line(line) for line in f]
print('done loading.')
cpra_to_rsids_trie = marisa_trie.BytesTrie(
lines, order=marisa_trie.LABEL_ORDER)
cpra_to_rsids_trie.save(cpra_to_rsids_trie_fname)
print('done with cpra->rsids trie at ' + cpra_to_rsids_trie_fname)
# TODO: What if several different chrom-pos-ref-alts have the same rsid? Do we only get the first? Or the last?
reversed_lines = ((rsid.decode(), cpra.encode())
for (cpra, rsids) in lines
for rsid in rsids.split(b','))
rsid_to_cpra_trie = marisa_trie.BytesTrie(
reversed_lines, order=marisa_trie.LABEL_ORDER)
rsid_to_cpra_trie.save(rsid_to_cpra_trie_fname)
print('done with rsid->cpra trie at ' + rsid_to_cpra_trie_fname)
def run(argv):
if '-h' in argv or '--help' in argv:
print('Make tries for converting between chr-pos-ref-alt and rsid')
exit(1)
if not should_replace(cpra_to_rsids_trie_filepath) and not should_replace(rsid_to_cpra_trie_filepath):
print('tries are up-to-date!')
else:
# Note: two identical VariantFileReaders are made in order to allow streaming to reduce memory usage.
# a different trie library might allow feeding both tries while reading from a file, but marisa_trie doesn't.
with VariantFileReader(sites_filepath) as reader:
cpras_and_rsids = (('{chrom}-{pos}-{ref}-{alt}'.format(**v), v['rsids'].encode('ascii')) for v in reader)
cpra_to_rsids_trie = marisa_trie.BytesTrie(cpras_and_rsids, order=marisa_trie.LABEL_ORDER)
cpra_to_rsids_trie.save(cpra_to_rsids_trie_filepath)
print('done with cpra->rsids trie at ' + cpra_to_rsids_trie_filepath)
# Note: if several different chrom-pos-ref-alts have the same rsid, then `trie[rsid]` = `[cpra1, cpra2, ...]`.
with VariantFileReader(sites_filepath) as reader:
def get_rsids_and_cpras():
for v in reader:
if v['rsids']:
cpra = '{chrom}-{pos}-{ref}-{alt}'.format(**v).encode('ascii')
for rsid in v['rsids'].split(','):
yield (rsid, cpra)
rsid_to_cpra_trie = marisa_trie.BytesTrie(get_rsids_and_cpras(), order=marisa_trie.LABEL_ORDER)
rsid_to_cpra_trie.save(rsid_to_cpra_trie_filepath)
print('done with rsid->cpra trie at ' + rsid_to_cpra_trie_filepath)
def run(argv):
if not should_replace(cpra_to_rsids_trie_filepath) and not should_replace(
rsid_to_cpra_trie_filepath):
print('tries are up-to-date!')
else:
with VariantFileReader(sites_filepath) as reader:
cpras_and_rsids = [('{chrom}-{pos}-{ref}-{alt}'.format(**v),
v['rsids'].encode('ascii')) for v in reader]
print('done loading.')
cpra_to_rsids_trie = marisa_trie.BytesTrie(
cpras_and_rsids, order=marisa_trie.LABEL_ORDER)
cpra_to_rsids_trie.save(cpra_to_rsids_trie_filepath)
print('done with cpra->rsids trie at ' + cpra_to_rsids_trie_filepath)
# TODO: What if several different chrom-pos-ref-alts have the same rsid? Do we only get the first? Or the last?
rsids_and_cpras = ((rsid, cpra.encode('ascii'))
for (cpra, rsids) in cpras_and_rsids
for rsid in rsids.decode('ascii').split(',')
if rsid)
rsid_to_cpra_trie = marisa_trie.BytesTrie(
rsids_and_cpras, order=marisa_trie.LABEL_ORDER)
rsid_to_cpra_trie.save(rsid_to_cpra_trie_filepath)
print('done with rsid->cpra trie at ' + rsid_to_cpra_trie_filepath)
def create_trie():
keys = []
values = []
start = time.time()
db = Database(None)
print "Connected to DB in", time.time() - start, "secs"
start = time.time()
db.execute("SELECT * FROM `Final_consolidated`")
print "Retrieved Info from table in", time.time() - start, "secs"
for row in db.cursor:
value = []
for i, ele in enumerate(row):
if i == 2:
keys.append(ele.lower())
else:
value.append(ele)
values.append(json.dumps(value))
trie = marisa_trie.Trie()
print "Starting to zip into Marisa trie"
trie = marisa_trie.BytesTrie(zip(keys, values))
print "Completed zipping into Marisa trie"
print "Pickling Trie"
with open('consolidated.pkl', 'wb') as output:
pickle.dump(trie, output, pickle.HIGHEST_PROTOCOL)
print "Total number of Keys:", len(keys)
def test_iteritems(keys, values):
trie = marisa_trie.BytesTrie(zip(keys, values))
assert trie.items() == list(trie.iteritems())
for key in keys:
prefix = key[:5]
assert trie.items(prefix) == list(trie.iteritems(prefix))
def create_from_other(self, compdawg):
d = count_prefixes(compdawg)
d = dict([(k, v.pop()) for k, v in d.iteritems() if len(v) == 1])
self.trie = marisa_trie.BytesTrie(
((k, v.encode("utf-8")) for k, v in d.iteritems()))
self.dawg = dawg.BytesDAWG(
((k, v.encode("utf-8")) for k, v in d.iteritems()))
def _pack_multiple_nodes(keys, nodes):
values = []
for node in nodes:
values.append(bytes(node.pack_to_string(), encoding='utf8'))
binary_trie = mt.BytesTrie(zip(keys, values))
return binary_trie
def test_dumps_loads(data):
trie = marisa_trie.BytesTrie(data)
buf = io.BytesIO()
pickle.dump(trie, buf)
buf.seek(0)
assert trie == pickle.load(buf)
def test_getitem_multiple(self):
data = [
('foo', b'x'),
('fo', b'y'),
('foo', b'a'),
]
trie = marisa_trie.BytesTrie(data)
assert trie['fo'] == [b'y']
assert trie['foo'] == [b'a', b'x']
def test_getitem_fuzzy(self):
data = self.data()
trie = marisa_trie.BytesTrie(data)
for key, value in data:
assert trie[key] == [value]
with pytest.raises(KeyError):
trie['2135']
def test_contains(self):
data = self.data()
trie = marisa_trie.BytesTrie(data)
for key, value in data:
assert key in trie
non_key = '2135'
assert non_key not in trie
def test_getitem_multiple():
data = [
("foo", b"x"),
("fo", b"y"),
("foo", b"a"),
]
trie = marisa_trie.BytesTrie(data)
assert trie["fo"] == [b"y"]
assert trie["foo"] == [b"a", b"x"]
def load_trie(filename):
"""
Load a BytesTrie from the marisa_trie on-disk format.
"""
trie = marisa_trie.BytesTrie()
# marisa_trie raises warnings that make no sense. Ignore them.
with warnings.catch_warnings():
warnings.simplefilter("ignore")
trie.load(filename)
return trie
def test_contains(keys, values, missing_key):
assume(missing_key not in keys)
data = zip(keys, values)
trie = marisa_trie.BytesTrie(data)
for word, value in data:
assert word in trie
assert missing_key not in trie
def __init__(self, phenos):
self._phenos = copy.deepcopy(phenos)
self._preprocess_phenos()
self._cpra_to_rsids_trie = marisa_trie.BytesTrie().load(
common_filepaths['cpra-to-rsids-trie'])
self._rsid_to_cpra_trie = marisa_trie.BytesTrie().load(
common_filepaths['rsid-to-cpra-trie'])
self._gene_alias_trie = marisa_trie.BytesTrie().load(
common_filepaths['gene-aliases-trie'])
self._autocompleters = [
self._autocomplete_variant,
self._autocomplete_rsid,
self._autocomplete_phenocode,
self._autocomplete_gene,
]
if any('phenostring' in pheno for pheno in self._phenos.values()):
self._autocompleters.append(self._autocomplete_phenostring)
def test_iteritems(self):
keys = get_random_words(1000)
values = get_random_binary(1000)
trie = marisa_trie.BytesTrie(zip(keys, values))
assert trie.items() == list(trie.iteritems())
for key in keys:
prefix = key[:5]
assert trie.items(prefix) == list(trie.iteritems(prefix))
def build_trie(file):
keys = []
values = []
with codecs.open(file, 'r', 'utf-8') as f:
for line in f.readlines():
parts = line.split(u'\t')
keys.append(parts[1].strip())
# byte tire needs bytes
values.append(parts[0].strip().encode('utf-8'))
return marisa_trie.BytesTrie(zip(keys, values))
def test_getitem(keys, values, missing_key):
assume(missing_key not in keys)
data = zip(keys, values)
trie = marisa_trie.BytesTrie(data)
for key, value in data:
assert trie[key] == [value]
with pytest.raises(KeyError):
trie[missing_key]
def test_has_keys_with_prefix():
fruit_trie = marisa_trie.BytesTrie([
('apple', b'foo'),
('pear', b'bar'),
('peach', b'baz'),
])
assert fruit_trie.has_keys_with_prefix('')
assert fruit_trie.has_keys_with_prefix('a')
assert fruit_trie.has_keys_with_prefix('pe')
assert fruit_trie.has_keys_with_prefix('pear')
assert not fruit_trie.has_keys_with_prefix('x')
def test_null_bytes_in_values(self):
keys = get_random_words(10000)
values = get_random_binary(10000)
assert any(b'\x00' in p for p in values)
data = zip(keys, values)
trie = marisa_trie.BytesTrie(data)
for key, value in data:
assert trie[key] == [value]
def test_pickling(self):
trie = marisa_trie.BytesTrie([
('foo', b'foo'),
('bar', b'bar'),
])
buf = io.BytesIO()
pickle.dump(trie, buf)
buf.seek(0)
trie2 = pickle.load(buf)
assert trie2['foo'] == [b'foo']
assert trie2['bar'] == [b'bar']
def __init__(self, seq):
self.nw_dst = marisa_trie.BytesTrie(zip(seq.nw_dst, seq.dl_vlan))
self.nw_src = marisa_trie.BytesTrie(zip(seq.nw_src, seq.dl_vlan))
self.metadata = marisa_trie.BytesTrie(zip(seq.metadata, seq.dl_vlan))
self.in_port = marisa_trie.BytesTrie(zip(seq.in_port, seq.dl_vlan))
self.dl_dst = marisa_trie.BytesTrie(zip(seq.dl_dst, seq.dl_vlan))
self.dl_src = marisa_trie.BytesTrie(zip(seq.dl_src, seq.dl_vlan))
def analyze(input):
#만약 사전 컴파일 중이면 5초 기다려주기
if is_compiling():
time.sleep(5)
compile_dic = marisa_trie.BytesTrie().load(COMPILED_NE_DIC_PATH)
morphs, search_iter, search_index, morph_result = gen_prefix_array(input)
# resultlist = []
ne_list = []
ne_tag_list = []
longest = True
for i in range(len(search_iter)):
candi = search_iter[i]
# print (self.searchIndex[idx])
findlist = compile_dic.prefixes(candi)
LOGGER.info('candi={}, findlist={}'.format(candi, findlist))
if len(findlist) == 0:
continue
# 최장일치
if longest:
item = max(findlist, key=len)
# print("item=", item)
# for tag in compile_dic:
if item in compile_dic and item not in ne_list:
# values = [value.decode('utf-8') for value in compile_dic[item]]
# resultlist.append((item, values))
for value in compile_dic[item]:
ne_list.append(item)
ne_tag_list.append(value.decode('utf-8'))
# 최단일치
else:
for item in findlist:
# print("item1=", item)
# for tag in compile_dic:
# resultlist.append((item, tag))
if item in compile_dic:
# values = [value.decode('utf-8') for value in compile_dic[item]]
# resultlist.append((item, values))
for value in compile_dic[item]:
ne_list.append(item)
ne_tag_list.append(value.decode('utf-8'))
# 개체명 분석 결과 구조체
ne_result = {
'ne': ne_list
, 'neTags': ne_tag_list
, 'neNum': len(ne_list)
, 'morphResult': morph_result
}
return ne_result
def test_items(self):
data = [
('fo', b'y'),
('foo', b'x'),
('foo', b'a'),
]
trie = marisa_trie.BytesTrie(data)
assert set(trie.items()) == set(data)
assert set(trie.items('f')) == set(data)
assert set(trie.items('fo')) == set(data)
assert set(trie.items('foo')) == set(data[1:])
assert trie.items('food') == []
assert trie.items('bar') == []
def test_items():
data = [
("fo", b"y"),
("foo", b"x"),
("foo", b"a"),
]
trie = marisa_trie.BytesTrie(data)
assert set(trie.items()) == set(data)
assert set(trie.items("f")) == set(data)
assert set(trie.items("fo")) == set(data)
assert set(trie.items("foo")) == set(data[1:])
assert trie.items("food") == []
assert trie.items("bar") == []
def _read_ipa2xs(self):
path = os.path.join('data', self.ipa2xs_fn)
path = pkg_resources.resource_filename(__name__, path)
pairs = []
with open(path, 'rb') as f:
reader = csv.reader(f, encoding='utf-8')
next(reader)
for ipa, xs, _ in reader:
pairs.append((
ipa,
xs.encode('utf-8'),
))
trie = marisa_trie.BytesTrie(pairs)
return trie
def test_keys(self):
trie = marisa_trie.BytesTrie([
('foo', b'x'),
('fo', b'y'),
('foo', b'a'),
])
# FIXME: ordering?
assert trie.keys() == ['foo', 'foo', 'fo']
assert trie.keys('f') == ['foo', 'foo', 'fo']
assert trie.keys('fo') == ['foo', 'foo', 'fo']
assert trie.keys('foo') == ['foo', 'foo']
assert trie.keys('food') == []
assert trie.keys('bar') == []
def test_keys():
trie = marisa_trie.BytesTrie([
("foo", b"x"),
("fo", b"y"),
("foo", b"a"),
])
# FIXME: ordering?
assert trie.keys() == ["foo", "foo", "fo"]
assert trie.keys("f") == ["foo", "foo", "fo"]
assert trie.keys("fo") == ["foo", "foo", "fo"]
assert trie.keys("foo") == ["foo", "foo"]
assert trie.keys("food") == []
assert trie.keys("bar") == []
def Main(argv):
# ArgParser
parser = argparse.ArgumentParser(usage=PrintHelp())
parser.add_argument("marisa",
help="Cesta ku slovníku vo formáte marisa_trie.")
args = parser.parse_args()
# Nacitanie trie
trie = marisa_trie.BytesTrie()
trie.load(args.marisa)
print("Morphological analyser")
# Vstup
for line in sys.stdin:
for word in line.split():
if (word == "."):
sys.exit()
PrintTrie(trie, word)
def run(argv):
if not os.path.exists(common_filepaths['genes']):
print('Downloading genes')
from . import download_genes
download_genes.run([])
aliases_filepath = common_filepaths['gene-aliases-trie']
if not os.path.exists(aliases_filepath):
print('gene aliases will be stored at {!r}'.format(aliases_filepath))
mapping = get_gene_aliases()
mapping = [(a, cs.encode('ascii')) for a, cs in mapping.items()]
aliases_trie = marisa_trie.BytesTrie(mapping)
aliases_trie.save(aliases_filepath)
else:
print('gene aliases are at {!r}'.format(aliases_filepath))
