def make_wordlist(filepath):
with open(filepath, 'r') as f:
wordlist = Automaton()
for idx, word in enumerate(set(Base().encode(t) for t in f.read().split())):
wordlist.add_word(word, (idx, word))
wordlist.make_automaton()
return wordlist
def test_add_concepts():
data_path = prepare_data(
path.join(path.dirname(__file__), "..", "data", "raw",
"vocabularies-tiny.zip"))
dataframe = pd.read_csv(path.join(data_path, "CONCEPT.csv"),
sep="\t").dropna(subset=["concept_name"])
automaton = Automaton()
automaton = add_concepts(
automaton, zip(dataframe["concept_name"], dataframe["concept_id"]))
automaton.make_automaton()
assert len(tuple(automaton.keys())) == 15791
first_keys = sorted(automaton.keys())[:10]
assert first_keys == [
'% REF', '(1-6)-alpha-glucomannan', '1 alpha-hydroxyergocalciferol',
"1,1',1'',1'''-(ethylenedinitrilo)tetra-2-propanol",
'1,1,1-trichloro-2,2,2-trifluoroethane', '1,1-difluoroethane',
'1,10-decanediol', '1,10-phenanthroline', '1,2,6-hexanetriol',
'1,2-Dipalmitoylphosphatidylcholine'
]
first_concept_id, first_concept_name = automaton.get(first_keys[0])
assert (first_concept_id, first_concept_name) == (8514, '% REF')
def update_automaton(dataframe,
automaton_filename=path.join(PROCESSED_DATA_PATH,
"vocabulary_automaton.pkl")):
# Assert we have the same amount of concept names and ids.
assert len(dataframe["concept_name"] == dataframe["concept_id"])
try:
with open(automaton_filename, "rb") as automaton_file:
automaton = pickle.load(automaton_file)
logging.info("Loaded previous automaton from path '{}'.".format(
automaton_filename))
except FileNotFoundError:
logging.info("Created new automaton.")
automaton = Automaton()
automaton = add_concepts(
automaton, zip(dataframe["concept_name"], dataframe["concept_id"]))
automaton.make_automaton()
with open(automaton_filename, "wb") as automaton_file:
pickle.dump(automaton, automaton_file)
logging.info(
"Updated automaton under path '{}'.".format(automaton_filename))
return automaton
def __init__( self, bc_to_id: Dict[str, str], len_linker: int, len_primer: int, *, max_mm: int = 1, use_stats: bool = True ): self.bc_to_id = bc_to_id self.len_linker = len_linker self.len_primer = len_primer self.stats = None if not use_stats else dict( n_only_primer=0, n_multiple_bcs=0, n_no_barcode=0, n_regular=0, n_barcode_mismatch=0, n_junk=0, ) self.automaton = Automaton() all_barcodes, self.blacklist = get_all_barcodes(bc_to_id.keys(), max_mm=max_mm) for pattern, barcode in all_barcodes.items(): self.automaton.add_word(pattern, barcode) self.automaton.make_automaton()
def build_automata(vocab):
# Build Aho-Corasick matching automata for vocabulary items
# grouped by length.
from ahocorasick import Automaton
start_time = datetime.now()
info('start building automata at {}'.format(
start_time.strftime("%H:%M:%S")))
strings = list(vocab)
max_len = max(len(s) for s in strings)
strings.sort(key=lambda s: len(s))
strings_by_len = defaultdict(list)
for k, g in groupby(strings, lambda s: len(s)):
strings_by_len[k] = list(g)
automata_by_len = {}
for i in range(1, max_len + 1):
if i not in strings_by_len:
continue
a = Automaton()
for s in strings_by_len[i]:
a.add_word(s, i)
a.make_automaton()
automata_by_len[i] = a
end_time = datetime.now()
info('finish building automata at {} (delta {})'.format(
end_time.strftime("%H:%M:%S"), end_time - start_time))
return automata_by_len
def build_automaton(self):
q = Entity.all()
q = q.filter(Entity.schema.in_(self.TYPES.keys()))
matches = {}
for entity in q:
tag = self.TYPES.get(entity.schema)
if tag is None:
continue
for name in entity.names:
if name is None or len(name) > 120:
continue
match = self.match_form(name)
if match is None:
continue
if match in matches:
matches[match].append((name, tag))
else:
matches[match] = [(name, tag)]
if not len(matches):
return
automaton = Automaton()
for term, entities in matches.iteritems():
automaton.add_word(term, entities)
automaton.make_automaton()
return automaton
def _generate(self):
latest = Entity.latest()
if latest is None:
return
if self.latest is not None and self.latest >= latest:
return
self.latest = latest
matches = {}
q = Entity.all()
q = q.filter(Entity.state == Entity.STATE_ACTIVE)
for entity in q:
for term in entity.regex_terms:
if term in matches:
matches[term].append(entity.id)
else:
matches[term] = [entity.id]
if not len(matches):
self.automaton = None
return
self.automaton = Automaton()
for term, entities in matches.iteritems():
self.automaton.add_word(term.encode('utf-8'), entities)
self.automaton.make_automaton()
log.info('Generated automaton with %s terms', len(matches))
def build_automata(vocab):
# Build Aho-Corasick matching automata for vocabulary items
# grouped by length. The wordpiece convention is inverted for
# matching: continuations are unmarked (instead of "##") and
# string start is marked by "^^".
from ahocorasick import Automaton
start_time = datetime.now()
info('start building automata at {}'.format(
start_time.strftime("%H:%M:%S")))
strings = [v[2:] if v.startswith('##') else '^^' + v for v in vocab]
max_len = max(len(s) for s in strings)
strings.sort(key=lambda s: len(s))
strings_by_len = defaultdict(list)
for k, g in groupby(strings, lambda s: len(s)):
strings_by_len[k] = list(g)
automata_by_len = {}
for i in range(1, max_len + 1):
if i not in strings_by_len:
continue
a = Automaton()
for s in strings_by_len[i]:
a.add_word(s, i)
a.make_automaton()
automata_by_len[i] = a
end_time = datetime.now()
info('finish building automata at {} (delta {})'.format(
end_time.strftime("%H:%M:%S"), end_time - start_time))
return automata_by_len
def load_vocab(self, paths, add=False, separator='\t'):
if add and hasattr(self, 'automaton'):
pass
else:
self.automaton = Automaton()
vocab = defaultdict(list)
tags = set()
if isinstance(paths, str):
paths = [paths]
for path in paths:
name_space = os.path.split(path)[-1]
print('read %s' % path)
output = os.popen('wc -l ' + path)
total = int(output.readline().split()[0])
with open(path, 'r') as r_f:
print('vocab file Examples:')
for n, line in enumerate(r_f):
print(line.strip())
if n >= 10:
break
r_f.seek(0)
for line in tqdm(r_f, desc='read file', total=total):
if random.random() > self.keep_vocab_pro:
continue
splits = line.strip().split(separator)
try:
if len(splits) == 2:
word, tag = splits
value = True
elif len(splits) == 3:
word, tag, value = splits
value = char2num(value)
elif len(splits) == 1:
word = splits[0]
value = True
tag = 'WORD'
else:
continue
if self.vocab_name_space:
tag = name_space + '_' + tag
vocab[word].append((tag, value))
if tag not in tags:
tags.add(tag)
except Exception as e:
print('vocab error: path-%s, line %s' % (path, line),
e)
continue
self.tags = tags if not hasattr(self, 'tags') else self.tags | tags
for word, value in tqdm(vocab.items(), desc='add words'):
self.automaton.add_word(word, (len(word), word, value))
print('总共有%s个词' % len(vocab))
self.automaton.make_automaton()
def _make_kwtree(keywords):
if keywords:
kwtree = Automaton()
for keyword in keywords:
kwtree.add_word(keyword, keyword)
kwtree.make_automaton()
else:
kwtree = None
return kwtree
def __init__(self, identifier_mapper, identifiers):
self.identifier_mapper = identifier_mapper
self.identifiers = identifiers
self.automaton = Automaton()
for identifier in identifiers:
mapped = identifier_mapper(identifier)
self.automaton.add_word(identifier, (len(identifier), mapped))
self.automaton.make_automaton()
self.dest_dirs = set()
def __init__(self, gaze_file=data_path):
self.locations = {}
self.vocab_to_location = {}
self.automaton = Automaton()
with open(gaze_file) as cin:
self.load_gazes(cin)
self.automaton.make_automaton()
def _get_keyword_processor(self, custom_vocab: List[str]):
keyword_processor = Automaton()
for i, keyword in enumerate(custom_vocab):
if len(keyword) > 1:
keyword_processor.add_word(keyword, (i, keyword))
keyword_processor.make_automaton()
return keyword_processor
def initialize_ac_automaton(kmers: pd.DataFrame):
A = Automaton()
for idx, kmer in enumerate(set(kmers['kmer'])):
A.add_word(kmer, (idx, kmer))
A.make_automaton()
return A
def benchmark_pyahocorasick(LINE):
from ahocorasick import Automaton, STORE_INTS
automaton = Automaton()
for i, key in enumerate(KEYS):
automaton.add_word(key, key)
automaton.make_automaton()
print(list(automaton.iter(LINE)))
benchmark("list(automaton.iter(LINE))", locals())
def build_automaton(vocab):
# Build Aho-Corasick matching automaton for vocabulary items
from ahocorasick import Automaton
start_time = datetime.now()
info('start building automaton at {}'.format(
start_time.strftime("%H:%M:%S")))
a = Automaton()
for v in vocab:
a.add_word(v, len(v))
a.make_automaton()
end_time = datetime.now()
info('finish building automata at {} (delta {})'.format(
end_time.strftime("%H:%M:%S"), end_time - start_time))
return a
def init_automaton(scheme_fasta):
"""Initialize Aho-Corasick Automaton with kmers from SNV scheme fasta
Args:
scheme_fasta: SNV scheme fasta file path
Returns:
Aho-Corasick Automaton with kmers loaded
"""
A = Automaton()
for header, sequence in parse_fasta(scheme_fasta):
A.add_word(sequence, (header, sequence, False))
A.add_word(revcomp(sequence), (header, sequence, True))
A.make_automaton()
return A
def _get_automaton(normalizer):
with compiler_lock:
if normalizer in AUTOMATA:
return AUTOMATA.get(normalizer)
aho = Automaton()
count = 0
for place in iter_places():
name = place.get('name')
norm = normalizer(name)
value = (place.get('code'), place.get('country'))
aho.add_word(norm, value)
count += 1
log.debug("Country automaton: %d places", count)
aho.make_automaton()
AUTOMATA[normalizer] = aho
return aho
def init_automaton(scheme_fasta):
"""Initialize Aho-Corasick Automaton with kmers from SNV scheme fasta
Args:
scheme_fasta: SNV scheme fasta file path
Returns:
Aho-Corasick Automaton with kmers loaded
"""
A = Automaton()
for header, sequence in parse_fasta(scheme_fasta):
kmer_list = expand_degenerate_bases(sequence)
for seq in kmer_list:
A.add_word(seq, (header, seq, False))
A.add_word(revcomp(seq), (header, seq, True))
A.make_automaton()
return A
def test_match_text():
data_path = prepare_data(
path.join(path.dirname(__file__), "..", "data", "raw",
"vocabularies-tiny.zip"))
dataframe = pd.read_csv(path.join(data_path, "CONCEPT.csv"),
sep="\t").dropna(subset=["concept_name"])
automaton = Automaton()
automaton = add_concepts(
automaton, zip(dataframe["concept_name"], dataframe["concept_id"]))
automaton.make_automaton()
matches = list(generate_matches(automaton=automaton, text=dummy_abstract))
match_soll_values = [(54, (46257025, 'ethyl acetate')),
(653, (45616149, 'formic acid')),
(785, (8512, 'day'))]
assert matches == match_soll_values
def create_automaton(dict_dir, vocab_suffix, min_word_len=3):
assert isinstance(min_word_len, int) or isinstance(min_word_len, dict)
automaton = Automaton()
if os.path.isdir(dict_dir):
dicts_path = [
os.path.join(dict_dir, i) for i in os.listdir(dict_dir)
if i.endswith(vocab_suffix)
]
else:
dicts_path = [dict_dir]
for path in dicts_path:
tag = os.path.split(path)[-1].strip(vocab_suffix)
vocab = set(readfile(path, deal_func=lambda x: x.strip()))
tag_min_word_len = min_word_len if isinstance(
min_word_len, int) else min_word_len[tag]
for word in vocab:
word_len = len(word)
if word_len >= tag_min_word_len:
automaton.add_word(word, (word_len, word, tag))
automaton.make_automaton()
return automaton
def __create_automaton(self):
paths = [
('Brand', os.path.join(Path.dictionary, 'Brand.txt')),
('Car', os.path.join(Path.dictionary, 'Car.txt')),
('Train', os.path.join(Path.dictionary, 'Train.txt')),
('Predicate', os.path.join(Path.dictionary, 'config.txt'))
]
automaton = Automaton()
for tag, path in paths:
with open(path, 'r') as r_f:
for line in r_f:
line = line.rstrip('\n')
_, *words = line.split('\t')
for word in words:
word = re.sub('\(.*?\)', '', word.lower())
_, tag_set = automaton.get(word, (word, set()))
tag_set.add(tag)
automaton.add_word(word, (word, tag_set))
automaton.make_automaton()
return automaton
def _generate(self):
latest = Entity.latest()
if self.latest is not None and self.latest >= latest:
return
self.latest = latest
matches = defaultdict(set)
q = Entity.all()
q = q.options(joinedload('other_names'))
q = q.filter(Entity.state == Entity.STATE_ACTIVE)
for entity in q:
for term in entity.regex_terms:
matches[term].add(entity.id)
if not len(matches):
self.automaton = None
return
self.automaton = Automaton()
for term, entities in matches.items():
self.automaton.add_word(term.encode('utf-8'), entities)
self.automaton.make_automaton()
log.info('Generated automaton with %s terms', len(matches))
async def _update_links_automaton(self):
"""
Fetch the latest version of the links from the table, build an automaton.
"""
logger.info(
"_update_links_automaton: fetching links from table %s",
self._links_table,
)
try:
links = await self._api.run_db_interaction(
"Fetch links from the table", _db_fetch_links,
self._links_table)
logger.info("_update_links_automaton: we received %d links",
len(links))
new_link_automaton = Automaton(ahocorasick.STORE_LENGTH)
for link in links:
new_link_automaton.add_word(link)
await make_deferred_yieldable(
deferToThread(new_link_automaton.make_automaton))
self._link_automaton = new_link_automaton
except Exception as e:
logger.exception("_update_links_automaton: could not update")
raise e
def default_phrase_mapper():
"""Return a phrase mapper initialized with the default phrases."""
return PhraseMapper(Automaton(), DEFAULT_PHRASES).setup()
def location_phrase_mapper():
"""Return a phrase mapper initialized with the location phrases."""
return PhraseMapper(Automaton(), LOCATION_PHRASES).setup()
from ahocorasick import Automaton
from pickle import load, dump
auto = Automaton()
auto.add_word('abc', 'abc')
auto.add_word('def', 'def')
with open('automaton-wee.pickle', 'wb') as dest:
dump(auto, dest)
import threading
from ahocorasick import Automaton
# python dictionary with trie patterns
file_words = json.load(open("../static_ioc_sample_30k.txt", "r"))
words_to_search = list()
trie_words = list()
total_words_to_search = 1000
total_words_added = 0
t = list()
patterns = dict()
total_initial_words = 0
total_iterations = 10 # CHANGE the number of iterations to perform: +/- 30k patterns per iteration.
A = Automaton()
for x in range(0, total_iterations):
print("In iteration ", x)
for key in file_words:
for value in file_words[key]:
value_random = value + str(random.randint(10000, 500000))
if total_words_to_search != total_words_added:
words_to_search.append(value)
total_words_added += 1
if x == 0:
total_initial_words += 1
A.add_word(value_random, value)
print(f"Initial words {total_initial_words}")
print(f"Total patterns on AC trie: {total_initial_words*total_iterations+1}")
A.make_automaton()
def __init__(self, *args, **kwargs):
super(AhoCorasick, self).__init__(*args, **kwargs)
# Use the Aho-Corasick search algorithm to speed up phrase lookups
self.automaton = Automaton()
def __init__(self):
self.latest = None
self.automaton = Automaton()
self.matches = {}
