def encodeCntTimeout(src, symbol, dst, encoding, discardDelim):
'''encodeCntTimeout(src, symbol, dst, encoding, discardDelim) -> [Transition]
Encodes a counter timeout in the place of the transition.
'''
newTransitions = []
if encoding in {
CounterEncoding.unary, CounterEncoding.binaryLittleEndian,
CounterEncoding.binaryBigEndian
}:
zeroState = dst + "Y0"
endState = zeroState + "_end"
newTransitions.append(Automaton.makeEpsTrans(src, zeroState))
newTransitions.append(
Automaton.makeTrans(zeroState, SYMBOL_ZERO, zeroState))
newTransitions.append(
Automaton.makeTrans(zeroState, SYMBOL_ZERO, endState))
else:
raise Exception("Invalid encoding: " + str(encoding))
if discardDelim:
newTransitions.append(Automaton.makeEpsTrans(endState, dst))
else:
newTransitions.append(
Automaton.makeTrans(endState, SYMBOL_ENABLED, dst))
return newTransitions
def test_nfa_nfa_1(self):
"""Convertion from NFA to DFA."""
q1, q2, q3 = map(State, ['q1', 'q2', 'q3'])
q1.add_transition('b', q2)
q1.add_transition('$', q3)
q2.add_transition('a', [q2, q3])
q2.add_transition('b', q3)
q3.add_transition('a', q1)
M = Automaton('M', [q1, q2, q3], ['a', 'b'], q1, q1)
D = M.to_dfa()
self.assertTrue('q1' in D.Q[0].ready)
self.assertTrue('q3' in D.Q[0].ready)
self.assertEqual(len(D.Q[0].ready), 2)
self.assertTrue('q2' in D.Q[1].ready)
self.assertEqual(len(D.Q[1].ready), 1)
self.assertTrue('q2' in D.Q[2].ready)
self.assertTrue('q3' in D.Q[2].ready)
self.assertEqual(len(D.Q[2].ready), 2)
self.assertTrue('q3' in D.Q[3].ready)
self.assertEqual(len(D.Q[3].ready), 1)
self.assertTrue('q1' in D.Q[4].ready)
self.assertTrue('q2' in D.Q[4].ready)
self.assertTrue('q3' in D.Q[4].ready)
self.assertEqual(len(D.Q[4].ready), 3)
self.assertEqual(len(D.Q[5].ready), 0)
def convert_graph_fmt(filename, fmt, output_dir=None):
automaton = Automaton(fromfile=filename)
filename_dir, filename_name = os.path.split(filename)
if output_dir:
filename_dir = output_dir
new_filename = os.path.join(filename_dir,
".".join([filename_name.split('.')[0], fmt]))
automaton.export(filename=new_filename)
def create_word_wfsa(corpus):
wfsa = Automaton()
for word in corpus:
prob = corpus[word]
state = "".join(word) + "_0"
wfsa.emissions[state] = word
wfsa.m_emittors[word].add(state)
wfsa.m["^"][state] = math.log(prob)
wfsa.m[state]["$"] = 0 # log(1)
return wfsa
def eliminate_e_trans(
automate: Automaton
): # methode pour enlever les espsilon transitions d'un automate
for l in range(len(automate.transitions)):
while "%" in automate.transitions[l][1]:
list_of_ep_trans = stock_of_trans(automate, "%")
for i in range(len(list_of_ep_trans)):
etatq = list_of_ep_trans[i][0]
etatk = list_of_ep_trans[i][2]
list_des_trans = []
for k in range(len(automate.transitions)):
if automate.transitions[k][0] == etatk:
list_des_trans.append(automate.transitions[k])
for j in range(len(list_des_trans)):
automate.add_transition(etatq, list_des_trans[j][1],
list_des_trans[j][2])
automate.remove_transition(etatq, "%", etatk)
if etatk in automate.acceptstates:
automate.make_accept(etatq)
for state in automate.statesdict.values():
while function_de_transition(automate, state, "%") is not None:
eliminate_e_trans(automate)
automate.remove_unreachable()
automate.name = automate.name + " sans e trans"
return automate
def intersect(automaton1: Automaton, automaton2: Automaton) -> Automaton:
automaton1 = automaton1.as_DFA()
automaton2 = automaton2.as_DFA()
states = get_states_cross_product(automaton1, automaton2)
initial = automaton1.initial + automaton2.initial
accept = set()
for a1 in automaton1.accept:
for a2 in automaton2.accept:
accept.add(a1+a2)
alphabet = automaton1.alphabet | automaton2.alphabet
return Automaton(states, initial, accept, alphabet)
def add_transitions(automaton: Automaton, og_state: State, character: str,
dest_state: State):
if character == EPSILON:
if dest_state.name in automaton.acceptstates:
automaton.make_accept(og_state.name)
for char in dest_state.transitions:
for dest in dest_state.transitions.get(char).keys():
add_transitions(automaton, og_state, char, dest)
elif (og_state.name, character, dest_state.name) not in a.transitions:
a.add_transition(og_state.name, character, dest_state.name)
return
def check(rules, tests, parser_args={}):
print "-" * 70
g = Grammar(rules)
a = Automaton(g)
fd = open(join(testdir, "tmp.py"), "w")
a.write_parser(fd)
fd.close()
del a, g
import tmp
reload(tmp)
p = tmp.Parser(**parser_args)
EOF.set_real_eof(p.EOF)
for input, e_tree, e_err in tests:
e_err = [(x[0], frozenset(x[1])) for x in e_err]
print "input: " + repr(input)
try:
tree = p.parse((x, k) for k, x in enumerate(input))
err = []
except p.ParseErrors, e:
tree = e.tree
err = e.errors
err = [(x[0], frozenset(x[1])) for x in err]
success = True
for e in e_err:
if e not in err:
print " missed error: " + repr(e)
success = False
for e in err:
if e not in e_err:
print " unexpected error: " + repr(e)
success = False
if e_tree != ignore and tree != e_tree:
print " unexpected result:"
print " expected: " + repr(e_tree)
print " got: " + repr(tree)
success = False
if success:
print " success"
else:
print " failure"
global errors
errors += 1
def check(rules, tests, parser_args={}):
print "-" * 70
g = Grammar(rules)
a = Automaton(g)
fd = open(join(testdir, "tmp.py"), "w")
a.write_parser(fd)
fd.close()
del a, g
import tmp
reload(tmp)
p = tmp.Parser(**parser_args)
EOF.set_real_eof(p.EOF)
for input, e_tree, e_err in tests:
e_err = [(x[0], frozenset(x[1])) for x in e_err]
print "input: " + repr(input)
try:
tree = p.parse((x, k) for k, x in enumerate(input))
err = []
except p.ParseErrors, e:
tree = e.tree
err = e.errors
err = [(x[0], frozenset(x[1])) for x in err]
success = True
for e in e_err:
if e not in err:
print " missed error: " + repr(e)
success = False
for e in err:
if e not in e_err:
print " unexpected error: " + repr(e)
success = False
if e_tree != ignore and tree != e_tree:
print " unexpected result:"
print " expected: " + repr(e_tree)
print " got: " + repr(tree)
success = False
if success:
print " success"
else:
print " failure"
global errors
errors += 1
def union_to_automaton(self, first_aut, second_aut):
""" Converte uma expressao de uniao em seu respectivo automato """
final_state = self.create_state()
finals = {final_state}
initial = self.create_state()
states = first_aut.states.union(second_aut.states).union(
{initial, final_state})
alphabet = first_aut.alphabet.union(second_aut.alphabet)
resulting_function = {}
resulting_function.update(first_aut.function)
resulting_function.update(second_aut.function)
resulting_function[initial] = {
'&': {initial, first_aut.initial, second_aut.initial}
}
resulting_function[final_state] = {'&': finals}
for final in first_aut.finals.union(second_aut.finals):
resulting_function[final]['&'].add(final_state)
# print(f"resulting_function = {resulting_function}")
return Automaton(states, alphabet, resulting_function, finals, initial)
def parseAut(problem, it):
'''parseAut(problem, it) -> Automaton
Parses an automaton (resp. transducer) representation into an instance of the
Automaton class. Modifies problem, it.
'''
# initialize
aut = Automaton()
for line in it:
if (line == "}"): # end of automaton
return aut
elif (line[0:2] == "//"): # comments
pass
elif (line == ""): # empty string
pass
elif (reStartStates.match(line)): # start states
Parser.parseStartStates(aut, line)
elif (reAcceptStates.match(line)): # accepting states
Parser.parseAcceptStates(aut, line)
elif (reEpsTrans.match(line)): # epsilon transition
Parser.parseEpsTrans(aut, line)
elif (reAutTrans.match(line)) or (reTransdTrans.match(line)):
Parser.parseTrans(problem, aut, line)
else:
raise Exception("Syntax error: " + line)
def closure_to_automaton(self, automaton):
""" Converte o fechamento para um automato """
final_state = self.create_state()
finals = {final_state}
initial = self.create_state()
states = automaton.states.union({initial, final_state})
alphabet = automaton.alphabet
resulting_function = {}
resulting_function.update(automaton.function)
resulting_function[initial] = {
'&': {initial, automaton.initial, final_state}
}
resulting_function[final_state] = {'&': finals}
finals_copy = automaton.finals.copy()
for final in finals_copy:
resulting_function[final]['&'].add(automaton.initial)
resulting_function[final]['&'].add(final_state)
# print(f"resulting_function = {resulting_function}")
return Automaton(states, alphabet, resulting_function, finals, initial)
def main():
args = parser.parse_args()
automaton = Automaton.from_file(args.path_to_automaton)
automaton = automaton_to_buchi(automaton)
formulas = get_formulas_from_file(args.path_to_ltl)
for formula in formulas:
neg = formula.normal_form(negation=True)
closure = neg.get_closure()
if any(map(lambda x: isinstance(x, UntilFormula), closure)):
ltl_automaton = LabeledGeneralisedBuchiAutomaton.from_formula(neg, aps=automaton.aps)
ltl_automaton = BuchiAutomaton.from_labeled(ltl_automaton)
results = verification(automaton, ltl_automaton)
if not results[0]:
print("Formula \"{}\" is valid\n---".format(formula.get_text()))
else:
print("Formula \"{}\" is invalid, here goes the counter example :\nStart:\n{};\nCycle:\n{};\n---".format(
formula.get_text(),
"\n".join(map(lambda x: "\t" + " ".join(sorted(map(lambda y: y.name, x))), results[1])),
"\n".join(map(lambda x: "\t" + " ".join(sorted(map(lambda y: y.name, x))), results[2]))
))
else:
print("Formula \"{}\" has not temporal operator from (U, F, G, R). \n"
"So nothing to verify really\n---".format(formula.get_text()))
def run_uniform_exp(self, quantizer, distance, emissions, state_bits, entropy):
exp_name = "{0}-{1}-{2}-{3}-{4}".format(
quantizer.levels,
abs(quantizer.neg_cutoff),
'm',
emissions,
distance[0])
logging.info("Running {0}".format(exp_name))
learnt_wfsa_filename = "{0}/{1}".format(self.workdir,
"learnt_{0}.wfsa".format(exp_name))
corpus = (self.morpheme_corpus if emissions == "m" else
self.unigram_corpus)
# read Automaton or learn it and dump it finally
if os.path.exists(learnt_wfsa_filename):
# read already learnt automaton
learnt_wfsa = Automaton.create_from_dump(open(learnt_wfsa_filename))
learnt_wfsa.quantizer = quantizer
learnt_wfsa.round_and_normalize()
else:
# create and learn new automaton
alphabet = get_alphabet(corpus)
numbers_per_letters = dict([(letter, 1)
for letter in alphabet])
#print numbers_per_letters
wfsa = Automaton.create_uniform_automaton(numbers_per_letters)
wfsa.finalize()
wfsa.quantizer = quantizer
wfsa.round_and_normalize()
cp = lambda *x: checkpoint_dump(wfsa,
"{0}/cp_{1}".format(self.workdir, exp_name), *x)
logging.info('learning starts here')
learnt_wfsa = learn_wfsa(wfsa, corpus, distance, cp)
# dump
with open(learnt_wfsa_filename, "w") as of:
learnt_wfsa.dump(of)
# encode automaton
encoder = Encoder(entropy)
bits_a, bits_e, bits_t, err, hq, tc = encode_wfsa(
learnt_wfsa, corpus, encoder)
return [exp_name, bits_a, bits_e, bits_t, err, hq, tc]
def determinise(automate: Automaton):
automateDet = Automaton("deterministe")
automateDet.initial = automate.initial
listeEtats = []
listeEtats.append(set([automate.initial.name]))
listeEtatsNew = listeEtats
print(listeEtats)
while len(listeEtats) != 0:
elt1 = listeEtats[0]
listeEtats = listeEtats[1:]
for elt11 in elt1:
for x in automate.alphabet:
EtatDest = []
for (source, symbol, destination) in automate.transitions:
if source == str(elt11) and symbol == x:
EtatDest.append(destination)
if (len(EtatDest) != 0) and EtatDest not in listeEtatsNew:
listeEtats.append(EtatDest)
listeEtatsNew.append(EtatDest)
automateDet.add_transition(str(elt1), x, str(EtatDest))
for etat in automate.acceptstates:
for elt1 in listeEtatsNew:
if etat in elt1:
automateDet.make_accept(str(elt1))
return automateDet
def enabledToFair(autEnabled):
'''enabledToFair(autEnabled) -> Automaton
Encodes fairness into the Enabled automaton, thus obtaining the language of all
possible configurations with enabledness encoded.
'''
result = Automaton()
return result
def test_app():
passwords = [
("x", "", "1"),
("want", "a", "epuu7Aeja9"),
("emerge", "bamboo", "zexae2Moo2"),
("question", "predict", "dahTho9Thai5yiasie1c"),
("quick fiber estate ripple phrase", "topic", "huu4aeju2gooth1iS6ai")
]
auto = Automaton()
client = Client(auto)
# Test password insertion
assert client.get_size() == 0
for i, (name, login, password) in enumerate(passwords):
auto.actions = "rb"
client.add(name, login, password)
assert client.get_size() == i+1
subtest_password_list(client, passwords)
subtest_has_name(client, passwords)
subtest_password_retrieval(client, auto, passwords)
# Export in plain text and also in encrypted form
# Do this before password removal testing
auto.actions = "brb"
export_plain = client.export(encrypt=False)
auto.actions = "b"
export_encrypted = client.export()
# Test password removal
removal_order = [name for name, _, _ in passwords]
random.shuffle(removal_order)
names = set(removal_order)
for name in removal_order:
auto.actions = "rb"
client.delete_by_name(name)
names.remove(name)
assert set(client.get_names()) == names
assert client.get_size() == 0
# Test import plain
subtest_clear(client, auto, passwords)
auto.actions = ";b"
client.import_("1.1.0", export_plain, encrypted=False)
subtest_password_list(client, passwords)
subtest_password_retrieval(client, auto, passwords)
# Test import encrypted
subtest_clear(client, auto, passwords)
auto.actions = ";b"
client.import_("1.1.0", export_encrypted, encrypted=True)
subtest_password_list(client, passwords)
subtest_password_retrieval(client, auto, passwords)
def createPlayer1(autEnabled):
'''createPlayer1(autEnabled) -> Automaton
Synthesises a transducer for Player 1 from the Enabled automaton autEnabled.
Player 1 nondeterministically chooses one enabled transition (if such exists).
'''
enabledTransd = autEnabled.toTransducer()
# the transducer for "no transition enabled"
noEnabledTransTransd = enabledTransd.renameStates(lambda s: s + '_noEn')
noEnabledTransTransd = noEnabledTransTransd.filterTrans(
lambda trans: (trans[1], trans[2]) != (SYMBOL_ENABLED, SYMBOL_ENABLED))
noEnabledTransTransd = noEnabledTransTransd.removeUseless()
rename1 = lambda state: state + "_en1"
rename2 = lambda state: state + "_en2"
# the transducer for chosing one enabled transition
oneChosenTransTransd1 = enabledTransd.renameStates(rename1)
oneChosenTransTransd2 = enabledTransd.renameStates(rename2)
oneChosenTransTransd1 = oneChosenTransTransd1.clearAcceptStates()
oneChosenTransTransd2 = oneChosenTransTransd2.clearStartStates()
oneChosenTransTransdMerged = Automaton.autUnion(oneChosenTransTransd1, oneChosenTransTransd2)
for trans in enabledTransd.transitions:
fstSymb = Automaton.getSymbol1(trans)
sndSymb = Automaton.getSymbol2(trans)
if (fstSymb, sndSymb) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
oneChosenTransTransdMerged.addTransTransd(
rename1(Automaton.getSrcState(trans)),
SYMBOL_ENABLED,
SYMBOL_CHOSEN,
rename2(Automaton.getTgtState(trans)))
# result = Automaton.autUnion(oneChosenTransTransdMerged, noEnabledTransTransd)
result = oneChosenTransTransdMerged
result = result.singleStartState(PLAYER1_START_STATE)
return result
def kleene(a1: Automaton) -> Automaton:
a2 = a1.deepcopy()
for state in a2.acceptstates:
a2.add_transition(state, "%", a2.initial.name)
new_initial_state = nouvel_etat(a1)
a2.add_transition(new_initial_state, EPSILON, a1.initial.name)
a2.initial = a2.statesdict[new_initial_state]
a2.make_accept(new_initial_state)
a2.name = "kleen"
return a2
def compute_entropy(probs, quantizer):
dist = 0.0
modeled_sum = sum([math.exp(quantizer.representer(math.log(prob))) for prob in probs])
for prob in probs:
prob_q = math.exp(quantizer.representer(math.log(prob)))
if prob_q == 0.0:
prob_q = Automaton.eps
prob_q /= modeled_sum
dist += Automaton.kullback(prob, prob_q)
#print prob, prob_q, Automaton.kullback(prob, prob_q)
return dist
def main():
wfsa = Automaton.create_from_dump(open(sys.argv[1]))
if len(sys.argv) > 2:
remaining = float(sys.argv[2])
lang = wfsa.language(remaining)
else:
lang = wfsa.language()
for w in lang:
di = wfsa.state_indices["$"]
prob = math.exp(lang[w][di])
print "{0} {1}".format("".join(w), prob)
def concat(a1: Automaton, a2: Automaton) -> Automaton:
a1star_a2 = a1.deepcopy()
a1star_a2.name = "a1star_a2"
nom_nouvel_etat = nouvel_etat(a1star_a2)
for s in a2.states:
if s in a1star_a2.states: # l'état de a2 existe dans la copie de a1star
a2.rename_state(
s,
nom_nouvel_etat) # ici on modifie a2 directement -> à éviter
nom_nouvel_etat = str(int(nom_nouvel_etat) +
1) # incrémente le compteur
for (s, a, d) in a2.transitions:
a1star_a2.add_transition(s, a, d)
a1star_a2.make_accept(a2.acceptstates)
for ac in a1.acceptstates:
a1star_a2.add_transition(ac, EPSILON, a2.initial.name)
a1star_a2.make_accept(a1.acceptstates,
accepts=False) # transforme en états non acceptants
return a1star_a2
def kleene_star(automaton1: Automaton) -> Automaton:
states = automaton1.states
newInitial = State("kleen_initial")
states.add(newInitial)
for state in automaton1.states:
if state.name in automaton1.accept:
state.add_transition(newInitial.name, EPSILON_SYMBOL)
automaton1.accept.add(newInitial.name)
newInitial.add_transition(automaton1.initial, EPSILON_SYMBOL)
return Automaton(states, newInitial.name, automaton1.accept, automaton1.alphabet)
def main():
# read automaton
wfsa = Automaton.create_from_dump(open(sys.argv[1]))
# read corpus
corpus = read_corpus(open(sys.argv[2]), separator=sys.argv[3], skip=[sys.argv[4]])
normalize_corpus(corpus)
# call distance_from_corpus
distances = {}
dist = wfsa.distance_from_corpus(corpus, Automaton.kullback, distances=distances)
# print out result
for k, v in distances.iteritems():
print k, v
def main():
path = 'entrada.txt'
input_file = open_input(path)
transitions_raw = get_transitions(input_file)
intial_or_final = initial_and_final(input_file)
states_dict = get_states(transitions_raw, intial_or_final)
word = get_word(input_file)
states = load_states(states_dict)
transitions = load_transitions(states, transitions_raw)
a = Automaton('Automato Finito', word, transitions)
inicial = a.get_initials_list()
for i in inicial:
a.start_reach(word, i)
word = get_word(input_file)
#print(a.get_steps())
gif = 1
for s in a.get_steps():
save_dot_and_png(a, s)
save_gif(gif)
gif += 1
def concat(a1: Automaton, a2: Automaton) -> Automaton:
"""
Concatenate 2 automaton and return the result
"""
a1a2 = a1.deepcopy()
a1a2.name = a1.name + a2.name
new_state_name = nouvel_etat(a1a2)
for s in a2.states:
if s in a1a2.states:
while new_state_name in a2.states:
new_state_name = str(int(new_state_name) + 1)
a2.rename_state(s, new_state_name)
new_state_name = str(int(new_state_name) + 1)
for (s, a, d) in a2.transitions:
a1a2.add_transition(s, a, d)
a1a2.make_accept(a2.acceptstates)
for ac in a1.acceptstates:
a1a2.add_transition(ac, EPSILON, a2.initial.name)
a1a2.make_accept(a1.acceptstates, accepts=False)
return a1a2
def main():
automaton = Automaton.create_from_dump(open(sys.argv[1]))
corpus = read_corpus(open(sys.argv[2]))
normalize_corpus(corpus)
entropy = float(sys.argv[3])
string_bits = "u"
if len(sys.argv) > 4:
string_bits = sys.argv[4]
q = LogLinQuantizer(10, -20)
automaton.quantizer = q
encoder = Encoder(entropy, string_bits)
print encoder.encode(automaton, corpus)
def encodeCntTimeout(src, symbol, dst, encoding, discardDelim):
'''encodeCntTimeout(src, symbol, dst, encoding, discardDelim) -> [Transition]
Encodes a counter timeout in the place of the transition.
'''
newTransitions = []
if encoding in {CounterEncoding.unary,
CounterEncoding.binaryLittleEndian, CounterEncoding.binaryBigEndian}:
zeroState = dst + "Y0"
endState = zeroState + "_end"
newTransitions.append(Automaton.makeEpsTrans(src, zeroState))
newTransitions.append(Automaton.makeTrans(zeroState, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTrans(zeroState, SYMBOL_ZERO, endState))
else:
raise Exception("Invalid encoding: " + str(encoding))
if discardDelim:
newTransitions.append(Automaton.makeEpsTrans(endState, dst))
else:
newTransitions.append(Automaton.makeTrans(endState, SYMBOL_ENABLED, dst))
return newTransitions
def main():
parser = argparse.ArgumentParser(
description="Operation for automatons"
)
parser.add_argument("--automaton1", "-a", dest="automaton1", type=str, required=True, help="File of the first automaton description")
parser.add_argument("--automaton2", "-b", dest="automaton2", type=str, help="File of the second automaton description. Not required for operation kleene_star")
parser.add_argument("--operation", "-o", dest="operation", type=str, required=True, choices=OPS, help="Operation to execute")
args = parser.parse_args()
automaton1 = Automaton(from_file=args.automaton1)
check_errors(automaton1.errors)
if args.operation != "kleene_star":
automaton2 = None
if args.automaton2:
automaton2 = Automaton(from_file=args.automaton2)
else:
print("Need to define --automaton2")
sys.exit(1)
check_errors(automaton2.errors)
result = getattr(operations, args.operation)(automaton1, automaton2)
else:
result = operations.kleene_star(automaton1)
print(result.render())
def get_rule_30(steps=None):
"""for get rule 30
Args:
steps (int): number of steps to run the rule 30
Returns:
cell_json(list): python object notation of """
# cells = [[0,0,0,0,1,0,0,0,0],
# [0,0,0,1,1,1,0,0,0],
# [0,0,1,1,0,0,1,0,0],
# [0,1,1,0,1,1,1,1,0],
# [1,1,0,0,1,0,0,0,1]]
cells = Automaton(steps, rule=30).string()
cell_json = [{"z": cells, "y": list(range(1, steps))}]
# cell_json = [{"z":cells}]
logging.info("\n" + str(Automaton(20, rule=30).string()))
return cell_json
def union(a1: Automaton, a2: Automaton) -> Automaton:
"""
Unite 2 automaton and return the result
"""
a1ora2 = a1.deepcopy()
a1ora2.name = a1.name + "+" + a2.name
new_state_name = nouvel_etat(a1ora2)
for s in a2.states:
if s in a1ora2.states:
while new_state_name in a2.states:
new_state_name = str(int(new_state_name) + 1)
a2.rename_state(s, new_state_name)
new_state_name = str(int(new_state_name) + 1)
for (s, a, d) in a2.transitions:
a1ora2.add_transition(s, a, d)
a1ora2.make_accept(a2.acceptstates)
new_state_name = nouvel_etat(a1ora2)
a1ora2.add_transition(new_state_name, EPSILON, a1ora2.initial.name)
a1ora2.add_transition(new_state_name, EPSILON, a2.initial.name)
a1ora2.initial = a1ora2.statesdict[new_state_name]
return a1ora2
def kleene(a1: Automaton) -> Automaton:
"""
Return kleene star of the automaton passed in the argument
"""
a1star = a1.deepcopy()
a1star.name = "a1star"
for state in a1star.acceptstates:
a1star.add_transition(state, EPSILON, a1star.initial.name)
new_state_name = nouvel_etat(a1star)
a1star.add_transition(new_state_name, EPSILON, a1star.initial.name)
a1star.initial = a1star.statesdict[new_state_name]
a1star.make_accept(new_state_name)
return a1star
def __init__(self, examples=None, automaton=None):
"""
:args:
automaton - (optional) instance of Automaton class
examples - list of strings, this can include the empty string ('')
"""
self.A = automaton if automaton else Automaton()
self.S = sorted(set(examples))
self.Σ = {char for s in self.S for char in s}
# If empty string is accepted, make root accepting state
self.root = self.A.add_node(
is_initial=True, is_final=True if self.S[0] == '' else False)
# Drop empty string from list of examples
if self.S[0] == '':
self.S = self.S[1:]
self.verbose = False
def run_3state_exp(self, quantizer, distance, harant, emissions,
state_bits):
aut_name = "{0}-{1}-{2}-{3}-{4}".format(
quantizer.levels,
abs(quantizer.neg_cutoff),
"_".join(("@".join(h) if type(h) == tuple else h) for h in harant),
emissions,
distance[0])
exp_name = "{0}-{1}".format(aut_name, state_bits)
logging.info("Running {0}".format(exp_name))
learnt_wfsa_filename = "{0}/{1}".format(self.workdir,
"learnt_{0}.wfsa".format(aut_name))
corpus = (self.morpheme_corpus if emissions == "m" else
self.unigram_corpus)
# read Automaton or learn it and dump it finally
if os.path.exists(learnt_wfsa_filename):
# read already learnt automaton
learnt_wfsa = Automaton.create_from_dump(open(learnt_wfsa_filename))
learnt_wfsa.quantizer = quantizer
learnt_wfsa.round_and_normalize()
else:
# create and learn new automaton
wfsa = create_new_three_state_fsa(self.morpheme_corpus,
harant, emissions)
wfsa.finalize()
wfsa.quantizer = quantizer
wfsa.round_and_normalize()
cp = lambda *x: checkpoint_dump(wfsa,
"{0}/cp_{1}".format(self.workdir, aut_name), *x)
learnt_wfsa = learn_wfsa(wfsa, corpus, distance, cp)
# dump
with open(learnt_wfsa_filename, "w") as of:
learnt_wfsa.dump(of)
# encode automaton
encoder = (self.morpheme_encoder if emissions=="m" else
self.unigram_encoder)
encoder.state_bits = state_bits
bits_a, bits_e, bits_t, err, hq, tc = encode_wfsa(
learnt_wfsa, corpus, encoder)
return [exp_name, bits_a, bits_e, bits_t, err, hq, tc]
def concatenation_to_automaton(self, first_aut, second_aut): #pylint: disable=R0201
""" Converte uma expressao de concatenacao em seu respectivo automato """
states = first_aut.states.union(second_aut.states)
finals = second_aut.finals
initial = first_aut.initial
alphabet = first_aut.alphabet.union(second_aut.alphabet)
resulting_function = {}
resulting_function.update(first_aut.function)
resulting_function.update(second_aut.function)
finals_copy = first_aut.finals.copy()
for final in finals_copy:
resulting_function[final]['&'].add(second_aut.initial)
return Automaton(states, alphabet, resulting_function, finals, initial)
def union(a1: Automaton, a2: Automaton) -> Automaton:
Union = a1.deepcopy()
nom_nouvel_etat = nouvel_etat(a2)
for s in a2.states:
if s in Union.states:
Union.rename_state(
s,
nouvel_etat(a2)) # ici on modifie a3 directement -> à éviter
nom_nouvel_etat = str(int(nom_nouvel_etat) + 1)
for (s, a, d) in a2.transitions:
Union.add_transition(s, a, d)
Union.make_accept(a2.acceptstates)
Union.add_transition(nom_nouvel_etat, EPSILON, a1.initial.name)
Union.add_transition(nom_nouvel_etat, EPSILON, a2.initial.name)
Union.initial.name = nom_nouvel_etat
Union.remove_transition(Union.initial.name, EPSILON, Union.initial.name)
return Union
def concat(automaton1: Automaton, automaton2: Automaton) -> Automaton:
states = set()
for state in sorted(automaton1.states):
new_state = State(f"A1{state.name}")
for trans in state.transitions:
new_state.add_transition(f"A1{trans.target}", trans.symbol)
if state.name in automaton1.accept:
new_state.add_transition(f"A2{automaton2.initial}", EPSILON_SYMBOL)
states.add(new_state)
for state in sorted(automaton2.states):
new_state = State(f"A2{state.name}")
for trans in state.transitions:
new_state.add_transition(f"A2{trans.target}", trans.symbol)
states.add(new_state)
initial = f"A1{automaton1.initial}"
accept = set()
for state_name in automaton2.accept:
accept.add(f"A2{state_name}")
alphabet = automaton1.alphabet | automaton2.alphabet
return Automaton(states, initial, accept, alphabet)
def create_automaton(self, current_node=None):
""" Cria o automato que corresponde a expressao """
if current_node is None:
current_node = self.expression_tree
# print(f"Current node is {current_node.get_value()}")
if current_node.get_left_child(
) is None and current_node.get_right_child() is None:
state = self.create_state()
final = self.create_state()
function = {
state: {
'&': {state},
current_node.get_value(): {final}
},
final: {
'&': {final}
}
}
aut = Automaton({state, final}, {current_node.get_value()},
function, {final}, state)
return aut
if current_node.get_left_child() is not None:
left_aut = self.create_automaton(current_node.get_left_child())
if current_node.get_right_child() is not None:
right_aut = self.create_automaton(current_node.get_right_child())
if current_node.get_value() == '.':
return self.concatenation_to_automaton(left_aut, right_aut)
if current_node.get_value() == '+':
return self.union_to_automaton(left_aut, right_aut)
# O valor eh *
if current_node.get_left_child() is not None:
return self.closure_to_automaton(left_aut)
return self.closure_to_automaton(right_aut)
def main():
parser = argparse.ArgumentParser(
description='display 1-D cellular automaton')
parser.add_argument('rule', type=int, help='wolframe rule (0-255)')
parser.add_argument('steps', type=int, help='number of steps')
parser.add_argument('seed', help='bin(seed) -> first line')
parser.add_argument('--debug',
action='store_true',
help='for debug purpose')
parser.add_argument('--raw', action='store_true', help='display with 0,1')
args = parser.parse_args()
if args.debug:
logging.basicConfig(level=logging.DEBUG)
automaton = Automaton(args.rule, args.seed)
automaton.display()
automaton.print_steps(args.steps, args.raw)
def createPlayer2(autPlay2, autEnabled):
'''createPlayer2(autPlay2, autEnabled) -> Automaton
Encode enabledness into the transducer autPlay2 for transitions of Player 2
using autEnabled defining enabledness.
'''
#############################################################################
def matchDelimChosenToEnDis(lhs, rhs): #
if lhs == SYMBOL_DELIM and rhs in {SYMBOL_ENABLED, SYMBOL_DISABLED}: #
return True #
elif lhs == SYMBOL_CHOSEN and rhs == SYMBOL_ENABLED: #
return True #
else: #
return lhs == rhs #
#############################################################################
#############################################################################
def matchDelimEnDis(lhs, rhs): #
if lhs == SYMBOL_DELIM and rhs in {SYMBOL_ENABLED, SYMBOL_DISABLED}: #
return True #
else: #
return lhs == rhs #
#############################################################################
#####################################
def pickChosenOrRight(lhs, rhs): #
if lhs == SYMBOL_CHOSEN: #
return lhs #
else: #
return rhs #
#####################################
result = Automaton.transdAutAutProd(
autPlay2,
autEnabled, matchDelimChosenToEnDis, pickChosenOrRight,
autEnabled, matchDelimEnDis, lambda x, y: y
)
return result
def main(options):
if not options.automaton_file:
raise Exception("Automaton \"option\" (-a) is mandatory")
automaton = Automaton.create_from_dump(open(options.automaton_file))
if options.quantizer:
automaton.quantizer = AbstractQuantizer.read(open(options.quantizer))
automaton.round_and_normalize()
input_ = sys.stdin
if options.corpus:
input_ = open(options.corpus)
corpus = read_corpus(input_, options.separator)
corpus = normalize_corpus(corpus)
learner = Learner.create_from_options(automaton, corpus, options)
learner.main()
output = sys.stdout
if options.output:
output = open(options.output, "w")
learner.automaton.dump(output)
def encodeEnablednessAut(aut, autEnabled):
'''encodeEnablednessAut(aut, autEnabled) -> Automaton
Encodes enabledness (given by autEnabled) into an automaton aut. This is done
by performing a product of aut and autEnabled, while treating 'delim' as
{'enabled', 'disabled'}.
'''
def funDelimMatchEnDis(lhs, rhs):
if ((lhs == SYMBOL_DELIM) and
((rhs == SYMBOL_ENABLED) or (rhs == SYMBOL_DISABLED))):
return True
else:
return lhs == rhs
def funTakeRhs(lhs, rhs):
return rhs
aut2 = aut.removeEpsilon()
result = Automaton.generalIntersection(aut2, autEnabled,
funDelimMatchEnDis, funTakeRhs)
return result
def create_three_state_fsa(corpus):
prefixes, suffixes = get_morpheme_frequencies(corpus)
fsa = Automaton()
for morpheme in prefixes.keys()+suffixes.keys()+['@', 'O']:
state = morpheme+'_0'
fsa.emissions[state] = morpheme
fsa.m_emittors[morpheme].add(state)
total_prefix_freq = sum(prefixes.values())
total_suffix_freq = sum(suffixes.values())
for prefix, p_freq in prefixes.iteritems():
fsa.m['^'][prefix+'_0'] = math.log(p_freq/total_prefix_freq)
fsa.m[prefix+'_0']['@_0'] = 0.0
for suffix, s_freq in suffixes.iteritems():
fsa.m['@_0'][suffix+'_0'] = math.log(s_freq/total_suffix_freq)
fsa.m[suffix+'_0']['$'] = 0.0
return fsa
def regexp_to_automaton(re: str) -> Automaton:
"""
Moore's algorithm: regular expression `re` -> non-deterministic automaton
"""
postfix = RegExpReader(re).to_postfix()
stack: List[Automaton] = []
for c in postfix:
if c in (".", "+", "*"):
right = stack.pop()
left = None
if c != "*":
left = stack.pop()
result = operation(left, right, c)
stack.append(result)
else:
automate = Automaton(c)
automate.add_transition("0", c, "1")
automate.make_accept("1")
stack.append(automate)
return stack[0]
def check(g, params):
a = Automaton(g, params)
a.check()
return a
def main():
automaton = Automaton.create_from_dump(open(sys.argv[1]))
corpus = read_corpus(open(sys.argv[2]), "#")
dc = DistanceCache(automaton, corpus)
dc.build_paths()
def autPlay1ToFair(aut):
'''autPlay1ToFair(aut) -> Automaton
Encodes fairness into aut for Player 1.
'''
###########################################
def encodeCntChooseTrans(src, symb1, symb2, dst, encoding, discardDelim):
'''encodeCntChooseTrans(src, symb1, symb2, dst, encoding, discardDelim) -> [Transition]
Encodes a transition for choosing process.
'''
cntState = dst + "_" + symb1 + "_" + symb2
newTransitions = []
newTransitions.append(Automaton.makeEpsTrans(src, cntState))
if encoding == CounterEncoding.unary:
endState = cntState + "_end"
if (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ZERO, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ONE, SYMBOL_ONE, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ZERO, endState))
elif (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_CHOSEN):
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ONE, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ONE, SYMBOL_ONE, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ONE, endState))
else:
raise Exception("Invalid delimiter symbol \'" + symb1 + "/" +
symb2 + "\' (only 'enabled' and 'chosen' are allowed)")
elif encoding in {CounterEncoding.binaryLittleEndian,
CounterEncoding.binaryBigEndian}:
cntState2 = cntState + "_2"
endState = cntState2
if (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ONE, SYMBOL_ONE, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ZERO, cntState2))
newTransitions.append(Automaton.makeTransTransd(cntState2, SYMBOL_ZERO, SYMBOL_ZERO, cntState2))
newTransitions.append(Automaton.makeTransTransd(cntState2, SYMBOL_ONE, SYMBOL_ONE, cntState2))
elif (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_CHOSEN):
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ONE, SYMBOL_ONE, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ONE, cntState2))
newTransitions.append(Automaton.makeTransTransd(cntState2, SYMBOL_ZERO, SYMBOL_ONE, cntState2))
newTransitions.append(Automaton.makeTransTransd(cntState2, SYMBOL_ONE, SYMBOL_ONE, cntState2))
else:
raise Exception("Invalid delimiter symbol \'" + symb1 + "/" +
symb2 + "\' (only 'enabled' and 'chosen' are allowed)")
else:
raise Exception("Invalid encoding: " + str(encoding))
# transition from endState
if discardDelim:
newTransitions.append(Automaton.makeEpsTrans(endState, dst))
else:
newTransitions.append(Automaton.makeTransTransd(endState, symb1, symb2, dst))
return newTransitions
######################################################
output = Automaton()
output.startStates = aut.startStates[:]
output.acceptStates = aut.acceptStates[:]
# for trans in aut.transitions:
# if Automaton.isEpsilonTrans(trans):
# # epsilon transitions
# output.addTrans(transition = trans)
# elif (Automaton.getSymbol1(trans) in ENCODING_ALPHABET):
# # delimiters
# (src, symb1, symb2, tgt) = trans
# cntState = tgt + "_" + symb1 + "_" + symb2
# output.addTrans(transition = (src, cntState))
# output.addTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ZERO, cntState)
# output.addTransTransd(cntState, SYMBOL_ONE, SYMBOL_ONE, cntState)
# output.addTransTransd(cntState, symb1, symb2, tgt)
# else:
# # other transitions
# output.addTrans(transition = trans)
for trans in aut.transitions:
if Automaton.isEpsilonTrans(trans):
# epsilon transitions
output.addTrans(transition = trans)
elif (Automaton.getSymbol1(trans) in ENCODING_ALPHABET):
# delimiters
(src, symb1, symb2, tgt) = trans
counterTransitions = encodeCntChooseTrans(
src, symb1, symb2, tgt, options.encoding, options.discardDelimiter)
output.addTransitions(counterTransitions)
# cntState = tgt + "_" + symb1 + "_" + symb2
# endState = cntState + "_end"
# output.addTrans(transition = (src, cntState))
# if (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
# # TODO: these are bad
# output.addTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ZERO, cntState)
# output.addTransTransd(cntState, SYMBOL_ONE, SYMBOL_ONE, cntState)
# output.addTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ZERO, endState)
# elif (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_CHOSEN):
# # TODO: these are bad
# output.addTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ONE, cntState)
# output.addTransTransd(cntState, SYMBOL_ONE, SYMBOL_ONE, cntState)
# output.addTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ONE, endState)
# else:
# raise Exception("Invalid delimiter symbol \'" + symb1 + "/" +
# symb2 + "\' (only 'enabled' and 'chosen' are allowed)")
#
# # transition from endState
# if options.discardDelimiter:
# output.addTrans(transition =
# Automaton.makeEpsTrans(endState, tgt))
# else:
# output.addTransTransd(endState, symb1, symb2, tgt)
else:
# other transitions
output.addTrans(transition = trans)
output.transitions = list(set(output.transitions)) # kill duplicates
return output
def encodeCntTrans(src, dst, symbol, encoding, discardDelim, allowZero):
newTransitions = []
if encoding == CounterEncoding.unary:
oneState = dst + "_" + symbol + "_" + SYMBOL_ONE
zeroState = dst + "_" + symbol + "_" + SYMBOL_ZERO
symbState = dst + "_" + symbol
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTrans(oneState, SYMBOL_ONE, oneState))
if allowZero:
newTransitions.append(
Automaton.makeEpsTrans(oneState, zeroState))
else:
newTransitions.append(
Automaton.makeTrans(oneState, SYMBOL_ONE, zeroState))
newTransitions.append(
Automaton.makeTrans(zeroState, SYMBOL_ZERO, zeroState))
newTransitions.append(
Automaton.makeTrans(zeroState, SYMBOL_ZERO, symbState))
if discardDelim:
newTransitions.append(Automaton.makeEpsTrans(symbState, dst))
else:
newTransitions.append(Automaton.makeTrans(symbState, symbol, dst))
elif encoding in {CounterEncoding.binaryLittleEndian,
CounterEncoding.binaryBigEndian}:
cntState = dst + "_cnt"
newTransitions.append(Automaton.makeEpsTrans(src, cntState))
newTransitions.append(Automaton.makeTrans(cntState, SYMBOL_ZERO, cntState))
newTransitions.append(Automaton.makeTrans(cntState, SYMBOL_ONE, cntState))
if allowZero:
endState = cntState
else:
cntState2 = cntState + "2"
endState = cntState2
newTransitions.append(Automaton.makeTrans(cntState, SYMBOL_ZERO, cntState2))
# newTransitions.append(Automaton.makeTrans(cntState, SYMBOL_ONE, cntState2))
newTransitions.append(Automaton.makeTrans(cntState2, SYMBOL_ZERO, cntState2))
newTransitions.append(Automaton.makeTrans(cntState2, SYMBOL_ONE, cntState2))
# how to treat the delimiter
if discardDelim:
newTransitions.append(Automaton.makeEpsTrans(endState, dst))
else:
newTransitions.append(Automaton.makeTrans(endState, symbol, dst))
pass
else:
raise Exception("Invalid type of counter encoding")
return newTransitions
def encodeCounter(aut, encoding, discardDelim, allowZero):
'''encodeCounter(aut, encoding, discardDelim, allowZero) -> Automaton
Encode counters using a given encoding into aut, yielding a new automaton. In
particular, every delimiter 'D' is substituted with a sub-automaton accepting
the language of all valid counter configurations (depending on the encoding).
Depending on discardDelim, 'D' is kept or discarded. The allowZero
parameter is a Boolean flag that when set to True allows the value of the
counter to be zero.
'''
##########################################################
def encodeCntTrans(src, dst, symbol, encoding, discardDelim, allowZero):
newTransitions = []
if encoding == CounterEncoding.unary:
oneState = dst + "_" + symbol + "_" + SYMBOL_ONE
zeroState = dst + "_" + symbol + "_" + SYMBOL_ZERO
symbState = dst + "_" + symbol
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTrans(oneState, SYMBOL_ONE, oneState))
if allowZero:
newTransitions.append(
Automaton.makeEpsTrans(oneState, zeroState))
else:
newTransitions.append(
Automaton.makeTrans(oneState, SYMBOL_ONE, zeroState))
newTransitions.append(
Automaton.makeTrans(zeroState, SYMBOL_ZERO, zeroState))
newTransitions.append(
Automaton.makeTrans(zeroState, SYMBOL_ZERO, symbState))
if discardDelim:
newTransitions.append(Automaton.makeEpsTrans(symbState, dst))
else:
newTransitions.append(Automaton.makeTrans(symbState, symbol, dst))
elif encoding in {CounterEncoding.binaryLittleEndian,
CounterEncoding.binaryBigEndian}:
cntState = dst + "_cnt"
newTransitions.append(Automaton.makeEpsTrans(src, cntState))
newTransitions.append(Automaton.makeTrans(cntState, SYMBOL_ZERO, cntState))
newTransitions.append(Automaton.makeTrans(cntState, SYMBOL_ONE, cntState))
if allowZero:
endState = cntState
else:
cntState2 = cntState + "2"
endState = cntState2
newTransitions.append(Automaton.makeTrans(cntState, SYMBOL_ZERO, cntState2))
# newTransitions.append(Automaton.makeTrans(cntState, SYMBOL_ONE, cntState2))
newTransitions.append(Automaton.makeTrans(cntState2, SYMBOL_ZERO, cntState2))
newTransitions.append(Automaton.makeTrans(cntState2, SYMBOL_ONE, cntState2))
# how to treat the delimiter
if discardDelim:
newTransitions.append(Automaton.makeEpsTrans(endState, dst))
else:
newTransitions.append(Automaton.makeTrans(endState, symbol, dst))
pass
else:
raise Exception("Invalid type of counter encoding")
return newTransitions
##############################################################
result = Automaton()
result.startStates = aut.startStates[:]
result.acceptStates = aut.acceptStates[:]
for trans in aut.transitions:
(src, symb, tgt) = trans
assert symb != SYMBOL_DISABLED
if symb == SYMBOL_ENABLED:
# for end-of-subword transitions
counterTransitions = encodeCntTrans(
src, tgt, symb, encoding, discardDelim, allowZero)
result.addTransitions(counterTransitions)
else:
# for ordinary transitions
result.addTrans(transition = trans)
result.transitions = list(set(result.transitions)) # kill duplicates
return result.removeUseless()
def encodeCntDecTrans(src, symb1, symb2, dst, encoding, discardDelim):
'''encodeCntDecTrans(src, symb1, symb2, dst, encoding, discardDelim) -> [Transition]
Encodes the counter decrement transitions.
'''
newTransitions = []
if encoding == CounterEncoding.unary:
if (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
oneState = dst + "_enXenX1"
zeroState = dst + "_enXenX0"
endState = zeroState + "_end"
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ONE, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ZERO, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ZERO, SYMBOL_ZERO, endState))
elif (symb1, symb2) == (SYMBOL_CHOSEN, SYMBOL_ENABLED):
chEnOneState = dst + "_chXenX1"
endState = chEnOneState + "_end"
newTransitions.append(Automaton.makeEpsTrans(src, chEnOneState))
newTransitions.append(Automaton.makeTransTransd(chEnOneState, SYMBOL_ONE, SYMBOL_ONE, chEnOneState))
newTransitions.append(Automaton.makeTransTransd(chEnOneState, SYMBOL_ONE, SYMBOL_ZERO, endState))
else:
raise Exception("Unexpected symbols: (" + symb1 + ", " + symb2 + ")")
elif encoding == CounterEncoding.binaryLittleEndian:
if (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
oneState = dst + "_enXenX1"
zeroState = dst + "_enXenX0"
endState = zeroState
# binary decrement
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ZERO, SYMBOL_ONE, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ZERO, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ONE, SYMBOL_ONE, zeroState))
elif (symb1, symb2) == (SYMBOL_CHOSEN, SYMBOL_ENABLED):
oneState = dst + "_chXenX1"
zeroState = dst + "_chXenX0"
endState = zeroState
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ONE, SYMBOL_ONE, zeroState))
else:
raise Exception("Unexpected symbols: (" + symb1 + ", " + symb2 + ")")
elif encoding == CounterEncoding.binaryBigEndian:
if (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
oneState = dst + "_enXenX1"
zeroState = dst + "_enXenX0"
endState = zeroState
# binary decrement
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ZERO, SYMBOL_ZERO, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ONE, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ZERO, SYMBOL_ONE, zeroState))
elif (symb1, symb2) == (SYMBOL_CHOSEN, SYMBOL_ENABLED):
oneState = dst + "_chXenX1"
zeroState = dst + "_chXenX0"
endState = zeroState
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ONE, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ZERO, zeroState))
else:
raise Exception("Unexpected symbols: (" + symb1 + ", " + symb2 + ")")
else:
raise Exception("Invalid encoding: " + str(encoding))
# deal with the delimiter
if options.discardDelimiter:
newTransitions.append(Automaton.makeEpsTrans(endState, dst))
else:
newTransitions.append(Automaton.makeTransTransd(endState, symb1, symb2, dst))
return newTransitions
def encodeCntChooseTrans(src, symb1, symb2, dst, encoding, discardDelim):
'''encodeCntChooseTrans(src, symb1, symb2, dst, encoding, discardDelim) -> [Transition]
Encodes a transition for choosing process.
'''
cntState = dst + "_" + symb1 + "_" + symb2
newTransitions = []
newTransitions.append(Automaton.makeEpsTrans(src, cntState))
if encoding == CounterEncoding.unary:
endState = cntState + "_end"
if (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ZERO, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ONE, SYMBOL_ONE, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ZERO, endState))
elif (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_CHOSEN):
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ONE, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ONE, SYMBOL_ONE, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ONE, endState))
else:
raise Exception("Invalid delimiter symbol \'" + symb1 + "/" +
symb2 + "\' (only 'enabled' and 'chosen' are allowed)")
elif encoding in {CounterEncoding.binaryLittleEndian,
CounterEncoding.binaryBigEndian}:
cntState2 = cntState + "_2"
endState = cntState2
if (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ONE, SYMBOL_ONE, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ZERO, cntState2))
newTransitions.append(Automaton.makeTransTransd(cntState2, SYMBOL_ZERO, SYMBOL_ZERO, cntState2))
newTransitions.append(Automaton.makeTransTransd(cntState2, SYMBOL_ONE, SYMBOL_ONE, cntState2))
elif (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_CHOSEN):
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ONE, SYMBOL_ONE, cntState))
newTransitions.append(Automaton.makeTransTransd(cntState, SYMBOL_ZERO, SYMBOL_ONE, cntState2))
newTransitions.append(Automaton.makeTransTransd(cntState2, SYMBOL_ZERO, SYMBOL_ONE, cntState2))
newTransitions.append(Automaton.makeTransTransd(cntState2, SYMBOL_ONE, SYMBOL_ONE, cntState2))
else:
raise Exception("Invalid delimiter symbol \'" + symb1 + "/" +
symb2 + "\' (only 'enabled' and 'chosen' are allowed)")
else:
raise Exception("Invalid encoding: " + str(encoding))
# transition from endState
if discardDelim:
newTransitions.append(Automaton.makeEpsTrans(endState, dst))
else:
newTransitions.append(Automaton.makeTransTransd(endState, symb1, symb2, dst))
return newTransitions
def create_new_three_state_fsa(corpus, harant="a", emissions="m"):
"""This fsa will not make use of an "O" state and will have edges from
the start state to all word-final morphemes
@harant - transitions to the second level
- "a": for all suffix morphemes
- list: morphemes and morpheme igrams that will be harant'ed
- else: No transitions here
@emissions - morpheme or character emissions
- "m": morpheme
- "c": character
"""
prefixes, suffixes = get_morpheme_frequencies(corpus)
fsa = Automaton()
morphemes_with_index = zip(len(prefixes) * [0], prefixes.keys())
morphemes_with_index += zip(len(suffixes) * [2], suffixes.keys())
for index, morpheme in morphemes_with_index:
state = "{0}_{1}".format(morpheme, index)
if emissions == "m":
fsa.emissions[state] = (morpheme,)
fsa.m_emittors[morpheme].add(state)
elif emissions == "c":
fsa.emissions[state] = tuple(morpheme)
fsa.m_emittors[tuple(morpheme)].add(state)
total_suffix_freq = sum(suffixes.values())*0.5
total_prefix_freq = sum(prefixes.values())+total_suffix_freq
#half of suffix frequencies should appear on edges from the epsilon state,
#the other half on edges from the starting state
for prefix, p_freq in prefixes.iteritems():
fsa.m['^'][prefix+'_0'] = math.log(p_freq/total_prefix_freq)
fsa.m[prefix+'_0']['EPSILON_1'] = 0.0
fsa.m['^']['EPSILON_0'] = -5.0
fsa.m['EPSILON_0']['EPSILON_1'] = 0.0
for suffix, s_freq in suffixes.iteritems():
fsa.m['EPSILON_1'][suffix+'_2'] = math.log(
(0.5*s_freq)/total_suffix_freq)
#the two 0.5s cancel each other out, which reflects that once
#we got as far as the epsilon state, it doesn't matter anymore
#whether we allowed suffixes to follow the start state or not
if harant == "a":
fsa.m['^'][suffix+'_2'] = math.log((0.5*s_freq)/total_prefix_freq)
elif type(harant) == list and suffix in harant:
fsa.m['^'][suffix+'_2'] = math.log((0.5*s_freq)/total_prefix_freq)
fsa.m[suffix+'_2']['$'] = 0.0
for prefix, _ in prefixes.iteritems():
for suffix, _ in suffixes.iteritems():
if type(harant) == list and (prefix, suffix) in harant:
state_name = "".join(prefix) + "".join(suffix) + "_3"
fsa.m["^"][state_name] = -20.0
fsa.m[state_name]["$"] = 0.0
fsa.emissions[state_name] = (prefix,suffix)
fsa.m_emittors[(prefix,suffix)].add(state_name)
fsa.round_and_normalize()
return fsa
def autPlay2ToFair(aut):
'''autPlay2ToFair(aut) -> Automaton
Encodes fairness into aut for Player 2.
'''
###########################################
def encodeCntDecTrans(src, symb1, symb2, dst, encoding, discardDelim):
'''encodeCntDecTrans(src, symb1, symb2, dst, encoding, discardDelim) -> [Transition]
Encodes the counter decrement transitions.
'''
newTransitions = []
if encoding == CounterEncoding.unary:
if (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
oneState = dst + "_enXenX1"
zeroState = dst + "_enXenX0"
endState = zeroState + "_end"
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ONE, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ZERO, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ZERO, SYMBOL_ZERO, endState))
elif (symb1, symb2) == (SYMBOL_CHOSEN, SYMBOL_ENABLED):
chEnOneState = dst + "_chXenX1"
endState = chEnOneState + "_end"
newTransitions.append(Automaton.makeEpsTrans(src, chEnOneState))
newTransitions.append(Automaton.makeTransTransd(chEnOneState, SYMBOL_ONE, SYMBOL_ONE, chEnOneState))
newTransitions.append(Automaton.makeTransTransd(chEnOneState, SYMBOL_ONE, SYMBOL_ZERO, endState))
else:
raise Exception("Unexpected symbols: (" + symb1 + ", " + symb2 + ")")
elif encoding == CounterEncoding.binaryLittleEndian:
if (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
oneState = dst + "_enXenX1"
zeroState = dst + "_enXenX0"
endState = zeroState
# binary decrement
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ZERO, SYMBOL_ONE, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ZERO, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ONE, SYMBOL_ONE, zeroState))
elif (symb1, symb2) == (SYMBOL_CHOSEN, SYMBOL_ENABLED):
oneState = dst + "_chXenX1"
zeroState = dst + "_chXenX0"
endState = zeroState
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ONE, SYMBOL_ONE, zeroState))
else:
raise Exception("Unexpected symbols: (" + symb1 + ", " + symb2 + ")")
elif encoding == CounterEncoding.binaryBigEndian:
if (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
oneState = dst + "_enXenX1"
zeroState = dst + "_enXenX0"
endState = zeroState
# binary decrement
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ZERO, SYMBOL_ZERO, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ONE, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTransTransd(zeroState, SYMBOL_ZERO, SYMBOL_ONE, zeroState))
elif (symb1, symb2) == (SYMBOL_CHOSEN, SYMBOL_ENABLED):
oneState = dst + "_chXenX1"
zeroState = dst + "_chXenX0"
endState = zeroState
newTransitions.append(Automaton.makeEpsTrans(src, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ONE, oneState))
newTransitions.append(Automaton.makeTransTransd(oneState, SYMBOL_ONE, SYMBOL_ZERO, zeroState))
else:
raise Exception("Unexpected symbols: (" + symb1 + ", " + symb2 + ")")
else:
raise Exception("Invalid encoding: " + str(encoding))
# deal with the delimiter
if options.discardDelimiter:
newTransitions.append(Automaton.makeEpsTrans(endState, dst))
else:
newTransitions.append(Automaton.makeTransTransd(endState, symb1, symb2, dst))
return newTransitions
#################################################
output = Automaton()
output.startStates = aut.startStates[:]
output.acceptStates = aut.acceptStates[:]
for trans in aut.transitions:
if Automaton.isEpsilonTrans(trans):
output.addTrans(transition = trans)
else:
(src, symb1, symb2, tgt) = trans
# if (symb1, symb2) == (SYMBOL_DISABLED, SYMBOL_DISABLED):
# disState = tgt + "_disXdis"
# output.addTrans(transition = (src, disState))
# output.addTransTransd(disState, SYMBOL_ZERO, SYMBOL_ZERO, disState)
# output.addTransTransd(disState, SYMBOL_ONE, SYMBOL_ONE, disState)
# output.addTransTransd(disState, SYMBOL_DISABLED, SYMBOL_DISABLED, tgt)
# elif (symb1, symb2) == (SYMBOL_ENABLED, SYMBOL_ENABLED):
# oneState = tgt + "_enXenX1"
# zeroState = tgt + "_enXenX0"
# output.addTrans(transition = (src, oneState))
# output.addTransTransd(oneState, SYMBOL_ONE, SYMBOL_ONE, oneState)
# output.addTransTransd(oneState, SYMBOL_ONE, SYMBOL_ZERO, zeroState)
# output.addTransTransd(zeroState, SYMBOL_ZERO, SYMBOL_ZERO, zeroState)
# output.addTransTransd(zeroState, SYMBOL_ENABLED, SYMBOL_ENABLED, tgt)
# elif (symb1, symb2) == (SYMBOL_DISABLED, SYMBOL_ENABLED):
# # NOTE: this case determines the particular notion of fairness, right?
# oneState = tgt + "_starXenX1"
# disEnState = tgt + "_disXen"
# output.addTrans(transition = (src, oneState))
# output.addTransTransd(oneState, SYMBOL_ZERO, SYMBOL_ONE, oneState)
# output.addTransTransd(oneState, SYMBOL_ONE, SYMBOL_ONE, oneState)
# output.addTransTransd(oneState, SYMBOL_ZERO, SYMBOL_ZERO, disEnState)
# output.addTransTransd(disEnState, SYMBOL_DISABLED, SYMBOL_ENABLED, tgt)
# elif (symb1, symb2) == (SYMBOL_CHOSEN, SYMBOL_DISABLED):
# chDisState = tgt + "_chXdis"
# output.addTrans(transition = (src, chDisState))
# output.addTransTransd(chDisState, SYMBOL_ZERO, SYMBOL_ZERO, chDisState)
# output.addTransTransd(chDisState, SYMBOL_ONE, SYMBOL_ONE, chDisState)
# output.addTransTransd(chDisState, SYMBOL_CHOSEN, SYMBOL_DISABLED, tgt)
# elif (symb1, symb2) == (SYMBOL_CHOSEN, SYMBOL_ENABLED):
# chEnOneState = tgt + "_chXenX1"
# chEnState = tgt + "_chXen"
# output.addTrans(transition = (src, chEnOneState))
# output.addTransTransd(chEnOneState, SYMBOL_ZERO, SYMBOL_ONE, chEnOneState)
# output.addTransTransd(chEnOneState, SYMBOL_ONE, SYMBOL_ONE, chEnOneState)
# output.addTransTransd(chEnOneState, SYMBOL_ZERO, SYMBOL_ZERO, chEnState)
# output.addTransTransd(chEnState, SYMBOL_CHOSEN, SYMBOL_ENABLED, tgt)
# else:
# assert symb1 not in FAIR_ENCODING_ALPHABET
# assert symb2 not in FAIR_ENCODING_ALPHABET
#
# output.addTrans(transition = trans)
if symb1 in FAIR_ENCODING_ALPHABET and symb2 in FAIR_ENCODING_ALPHABET:
counterTransitions = encodeCntDecTrans(src, symb1, symb2, tgt,
options.encoding, options.discardDelimiter,)
output.addTransitions(counterTransitions)
else:
assert symb1 not in FAIR_ENCODING_ALPHABET
assert symb2 not in FAIR_ENCODING_ALPHABET
output.addTrans(transition = trans)
output.transitions = list(set(output.transitions)) # kill duplicates
return output
def autFinalToFair(aut, autEnabled, options):
'''autFinalToFair(aut, autEnabled, options) -> Automaton
Encodes fairness into an automaton representing final configurations of
a system, w.r.t. enabled transitions given by autEnabled. The final states are
states given by aut with fairness encoded, or states corresponding to one
enabled process's counter reaching zero. The options parameter contains
command line arguments.
'''
###########################################
def encodeCntTimeout(src, symbol, dst, encoding, discardDelim):
'''encodeCntTimeout(src, symbol, dst, encoding, discardDelim) -> [Transition]
Encodes a counter timeout in the place of the transition.
'''
newTransitions = []
if encoding in {CounterEncoding.unary,
CounterEncoding.binaryLittleEndian, CounterEncoding.binaryBigEndian}:
zeroState = dst + "Y0"
endState = zeroState + "_end"
newTransitions.append(Automaton.makeEpsTrans(src, zeroState))
newTransitions.append(Automaton.makeTrans(zeroState, SYMBOL_ZERO, zeroState))
newTransitions.append(Automaton.makeTrans(zeroState, SYMBOL_ZERO, endState))
else:
raise Exception("Invalid encoding: " + str(encoding))
if discardDelim:
newTransitions.append(Automaton.makeEpsTrans(endState, dst))
else:
newTransitions.append(Automaton.makeTrans(endState, SYMBOL_ENABLED, dst))
return newTransitions
#############################################
aut1 = autEnabled.renameStates(lambda x: x + "Y1")
aut1 = aut1.clearAcceptStates()
aut2 = autEnabled.renameStates(lambda x: x + "Y2")
aut2 = aut2.clearStartStates()
aut3 = Automaton.autUnion(aut1, aut2)
for (src, symb, tgt) in autEnabled.transitions:
if symb == SYMBOL_ENABLED:
aut3.addTrans(src + "Y1", SYMBOL_ENABLED_TIMEOUT, tgt + "Y2")
aut4 = encodeCounter(aut3,
encoding = options.encoding,
discardDelim = options.discardDelimiter,
allowZero = True
)
aut5 = Automaton()
aut5.startStates = aut4.startStates[:]
aut5.acceptStates = aut4.acceptStates[:]
# encode the timeout condition
for trans in aut4.transitions:
if (not Automaton.isEpsilonTrans(trans) and
Automaton.getSymbol(trans) == SYMBOL_ENABLED_TIMEOUT):
(src, symb, tgt) = trans
# zeroState = tgt + "Y0"
# aut5.addTrans(transition = (src, zeroState))
# aut5.addTrans(zeroState, SYMBOL_ZERO, zeroState)
# aut5.addTrans(zeroState, SYMBOL_ZERO, tgt)
# for the special timeout symbol
counterTransitions = encodeCntTimeout(src, symb, tgt, options.encoding, options.discardDelimiter)
aut5.addTransitions(counterTransitions)
else:
# for other symbols
aut5.addTrans(transition = trans)
autB = encodeCounter(aut,
encoding = options.encoding,
discardDelim = options.discardDelimiter,
allowZero = True
)
output = Automaton.autUnion(aut5, autB)
output = output.singleStartState(FINAL_START_STATE)
output.transitions = list(set(output.transitions)) # kill duplicates
return output.removeUseless()