def test_names():
auto = Automaton()
auto.add_all(NAMES)
auto.update_automaton()
auto_matches = [(m.start, m.end) for m in auto.get_matches(TEXT)]
with TemporaryDirectory() as tmpdir:
#tmpdir = ''
fnm = os.path.join(tmpdir, 'test.aca')
auto.save_to_file(fnm)
auto2 = Automaton()
auto2.load_from_file(fnm)
auto2_matches = [(m.start, m.end) for m in auto2.get_matches(TEXT)]
assert list(auto.items()) == list(auto2.items())
assert list(auto.prefixes()) == list(auto2.prefixes())
assert auto_matches == auto2_matches
auto3 = Automaton()
auto3.load_from_string(auto2.save_to_string())
auto3_matches = [(m.start, m.end) for m in auto3.get_matches(TEXT)]
assert list(auto.items()) == list(auto2.items())
assert list(auto.prefixes()) == list(auto2.prefixes())
assert auto_matches == auto3_matches
def test_with_updating():
auto = Automaton()
auto.add('hers')
matches = auto.get_matches('ushers')
assert len(matches) == 1
auto.add('us')
matches = auto.get_matches('ushers')
assert len(matches) == 2
def test_automaton_with_words():
auto = Automaton()
for token in ['he', 'she', 'his', 'hers']:
auto.add(token)
expected_all_matches = [
Match(1, 4, 'Y'), Match(2, 4, 'Y'),
Match(2, 6, 'Y')
]
all_matches = auto.get_matches('ushers', exclude_overlaps=False)
print(all_matches)
assert expected_all_matches == all_matches
expected_nonoverlap_matches = [Match(2, 6, 'Y')]
nonoverlap_matches = auto.get_matches('ushers', exclude_overlaps=True)
assert expected_nonoverlap_matches == nonoverlap_matches
def test_names():
auto = Automaton(NAMES)
auto.add_all(NAMES)
print (auto)
matches = set(' '.join(match.elems) for match in auto.get_matches(TEXT.split()))
names = set(' '.join(name) for name in NAMES)
assert names == matches
def test_with_words():
auto = Automaton()
auto.add(['funderbeam'])
auto.add(['mattermark'])
auto.add(['500', 'startups'])
txt = 'funderbeam and mattermark along with 500 startups'.split()
expected = [Match(0, 1, 'Y'), Match(2, 3, 'Y'), Match(5, 7, 'Y')]
actual = auto.get_matches(txt)
assert expected == actual
def test_lemmas():
auto = Automaton()
auto.add(['sunlabob'], 'CO')
auto.add(['renewable'], 'CO')
lemmas = [
'sunlabob', 'renewable', 'energy', 'receive', '$', '2.1', 'million',
'investment'
]
print(auto.str())
matches = auto.get_matches(lemmas)
assert len(matches) == 2
def test_map_interface():
auto = Automaton()
auto['us'] = 'USA'
auto['her'] = 'EUROPE'
assert auto['us'] == 'USA'
assert auto['her'] == 'EUROPE'
matches = auto.get_matches('usher')
assert len(matches) == 2
assert matches[0].label == 'USA'
assert matches[1].label == 'EUROPE'
assert 'us' in auto
assert 'his' not in auto
# -*- coding: utf-8 -*-
from __future__ import unicode_literals, print_function, absolute_import
# create a new AC automaton
from aca import Automaton
automaton = Automaton()
# instead of plain strings, you can also use lists of tokens
names = [
(['Yuri', 'Artyukhin'], 'developer'),
(['Tom', 'Anderson', 'Jr'], 'designer'),
]
automaton.add_all(names)
# you can add an item like this as well
automaton[['Tom', 'Anderson']] = 'manager'
# if you are not using plain strings, make sure you tokenize the text as well
text = 'Tom Anderson Jr and Yuri Artyukhin work on my project'.split()
print('matches that maximize the number of matched words')
for match in automaton.get_matches(text):
print(match.start, match.end, match.elems, match.label)
print('all matches')
for match in automaton.get_matches(text, exclude_overlaps=False):
print(match.start, match.end, match.elems, match.label)
# -*- coding: utf-8 -*-
from __future__ import unicode_literals, print_function, absolute_import
# create a new AC automaton
from aca import Automaton
automaton = Automaton()
# add a dictionary of words to the automaton
painkillers = ['paracetamol', 'ibuprofen', 'hydrocloride']
automaton.add_all(painkillers)
# match the dictionary on a text
text = 'paracetamol and hydrocloride are a medications to relieve pain and fever. paracetamol is less efficient than ibuprofen'
for match in automaton.get_matches(text):
print(match.start, match.end, match.elems)