def validate_account_stake(snapshot, balances, args):
step('Verifying user account stake')
invalid_stake = 0
print_count = 0
total_accounts = len(balances)
for account in balances:
total = asset2int(balances[account][CSV_EOS_BALANCE])
liquid, net, cpu = get_account_stake(snapshot, account,
args.core_symbol)
# TODO: validate F(stake) ?
if total != liquid + cpu + net:
print_count = print_some(
print_count,
'{0} => TOTAL: [{1}] L:[{2}] C:[{3}] N:[{4}]'.format(
account, total, liquid, cpu, net))
invalid_stake += 1
if invalid_stake > 0:
print "> %d accounts with invalid stake" % (invalid_stake)
else:
success()
return True
def validate_system_accounts(snapshot):
step('Verifying system accounts')
found = []
for name, account in snapshot['accounts'].iteritems():
tick()
if name in system_accounts:
if account['privileged'] != system_accounts[name]['privileged']:
fail()
print "> %s account wrong privileged setting" % (name)
return False
# Verify resignement
if name != "eosio.null" and name != "eosio.prods":
actor = system_accounts[name]['actor']
permission = system_accounts[name]['permission']
if authority_controlled_by_one_actor(account['permissions']["owner"], actor, permission) != True or \
authority_controlled_by_one_actor(account['permissions']["owner"], actor, permission) != True:
fail()
print "> %s account NOT PROPERLY RESIGNED" % (name)
return False
found.append(name)
not_found = set(system_accounts.keys()) - set(found)
if len(not_found):
fail()
print "> missing system accounts %s" % (','.join(not_found))
return False
success()
return True
def monitor_progress(software):
"""Until all nodes are complete, monitor and print the status of the upgrade"""
step("Watch the progress")
errors = 0
with LiveMultilineOutput() as output:
while True:
try:
software.get(fields="state,status_details,elapsed_duration,estimated_duration")
if not software.state == "in_progress":
break
errors = 0
new_list = []
for detail in software.status_details:
new_list.append(
"[%s]: %s - %s"
% (detail.node.name, detail.name, detail.issue.message)
)
percent_complete = (
software.elapsed_duration / software.estimated_duration
) * 100
new_list.append("%.2f%% complete (estimate)" % percent_complete)
output.change(new_list)
except Exception: # pylint: disable=broad-except
errors += 1
if errors > 10:
break
finally:
time.sleep(30)
if errors > 10:
raise RuntimeError("ERROR: No longer could communicate with the server. Is it down?")
def validate_EOS_token(snapshot, args):
step('Verifying EOS token')
symbol = args.core_symbol
key = name_to_string(symbol2int(symbol), False)
sym = str(symbol2int(symbol))
eos_token = snapshot['tables']['eosio.token'][key]['stat'][sym]['data']
ok = True
if asset2int(eos_token['max_supply']) != 100000000000000:
if ok:
warning()
ok = False
print "> EOS max supply != 10000M (%s)" % eos_token['max_supply']
tokens = 0
code = snapshot['tables']['eosio.token']
for s in code:
if 'stat' in code[s]: tokens += 1
if tokens != 1:
if ok:
warning()
ok = False
print "> more than one token found in eosio.token"
if ok:
success()
return True
def visualize(seq, env, env_path, v_t, policy, T, task, goal):
for t in range(T):
# basic info
pos = env.agent_pos
vec = dir2num(env.dir_vec)
front_cell = env.front_pos # Get the cell in front of the agent
print("time:{t}, state: {p},{v}".format(t=t, p=pos, v=vec))
if task == 'Key':
# the key is in front of the agent
if what(front_cell, env) == 'Key':
seq.append(PK)
cost, done = step(env, PK)
print("----get Key----")
return seq
if task == 'Door':
# the door is in front of the agent
if what(front_cell, env) == 'Door':
seq.append(UD)
cost, done = step(env, UD)
seq.append(MF)
cost, done = step(env, MF)
print("----unlock the door----")
return seq
# read the corresponding policy given current position
control = policy[pos[0], pos[1], vec, t]
print("next move: {c}".format(c=c_dict[control]))
seq.append(control)
cost, done = step(env, control)
if done:
print("Reached Goal")
return seq
def validate_genesis(snapshot, genesis):
step('Validating genesis')
ddiff = DictDiffer(snapshot['genesis_state'], genesis)
added = ddiff.added()
removed = ddiff.removed()
changed = ddiff.changed()
if len(added) != 0 or len(removed) != 0 or len(changed) != 0:
fail()
if len(added) != 0:
print "> Snapshot has '%s' and your genesis dont" % (
','.join(added))
if len(removed) != 0:
print "> Your genesis has %s and snapshot dont" % (
','.join(removed))
if len(changed) != 0:
print "> Your genesis and snapshot have different '%s'" % (
','.join(changed))
print "# Snapshot genesis"
print json.dumps(snapshot['genesis_state'], indent=2, sort_keys=True)
print
print "# Your genesis"
print json.dumps(genesis, indent=2, sort_keys=True)
return False
success()
return True
def import_timetable(self):
print()
for unit in self.units:
if not self.import_timetable_for_unit(unit):
return False
else:
utils.step("Przetwarzam plan lekcji klasy {}".format(unit),
state=" OK ")
return True
def load_erc20_snapshot(args):
step('Loading ETH snapshot')
with open(args.csv_balance, 'rb') as csvfile:
balances = {}
for row in csv.reader(csvfile, delimiter=','):
tick()
row = [unicode(i) for i in row]
row[CSV_EOS_BALANCE] = '{0} {1}'.format(row[CSV_EOS_BALANCE],
args.core_symbol)
balances[row[CSV_EOS_ACCOUNT]] = row
success()
return balances
def show_current_cluster_image():
"""Verify our current software status"""
step("Show the packages")
substep("Show the current cluster image")
software = Software()
software.get()
print("Current software package:", software)
substep("Show the available software packages")
print("Available software packages:", list(SoftwarePackage.get_collection()))
return software
def run(self):
"""
Run the network simulation by evaluating model equations at every time
step.
"""
# integration constant
rho_a = 1./(self.stim_len)
# simulation time
time = np.arange(0, self.t_max)
for t in time[1:]:
# stop simulation if target found or max guesses
if self._terminate_condition(t):
break
# ----- Semantic Layer -----
# winning/active units
self.z[t-1] = step(self.w[t-1], self.theta_w)
# unit activities
self.a[t] = self.a[t-1] + rho_a*(mult(self.W,
self.z[t-1],
self.a[t-1]) +
self.I[t-1])
# ----- WTA Layer -----
eta = np.random.random(self.N) - self.noise_offset
# normalise activities to 0-1
divisor = np.max([1., self.a[t-1].max()])
self.w[t] = self.w[t-1] + self.rho_w * \
(self.c1*self.z[t-1] +
self.c2*self.a[t-1]/divisor -
self.c3*self.y[t-1] -
self.c4*self.r[t-1] +
self.c5*eta)
self.y[t] = self.z[t-1].sum()
# ----- Inhibitory Layer -----
# -2 added because of discreete update lags
self.r[t] = (self.r[t-1] +
step(self.z[t-self.stim_len:t].sum(axis=0),
self.stim_len-2)).clip(max=1)
# several winners are not allowed
if self.y[t] > 1:
raise Exception('Multiple winners')
# rectify WTA activities to 0 (if needed)
self.w[t] = self.w[t].clip(min=0)
def run(self):
"""
Run the network simulation by evaluating model equations at every time
step.
"""
# integration constant
rho_a = 1. / (self.stim_len)
# simulation time
time = np.arange(0, self.t_max)
for t in time[1:]:
# stop simulation if target found or max guesses
if self._terminate_condition(t):
break
# ----- Semantic Layer -----
# winning/active units
self.z[t - 1] = step(self.w[t - 1], self.theta_w)
# unit activities
self.a[t] = self.a[t - 1] + rho_a * (
mult(self.W, self.z[t - 1], self.a[t - 1]) + self.I[t - 1])
# ----- WTA Layer -----
eta = np.random.random(self.N) - self.noise_offset
# normalise activities to 0-1
divisor = np.max([1., self.a[t - 1].max()])
self.w[t] = self.w[t-1] + self.rho_w * \
(self.c1*self.z[t-1] +
self.c2*self.a[t-1]/divisor -
self.c3*self.y[t-1] -
self.c4*self.r[t-1] +
self.c5*eta)
self.y[t] = self.z[t - 1].sum()
# ----- Inhibitory Layer -----
# -2 added because of discreete update lags
self.r[t] = (self.r[t - 1] +
step(self.z[t - self.stim_len:t].sum(axis=0),
self.stim_len - 2)).clip(max=1)
# several winners are not allowed
if self.y[t] > 1:
raise Exception('Multiple winners')
# rectify WTA activities to 0 (if needed)
self.w[t] = self.w[t].clip(min=0)
def validate_memory(snapshot):
step('Verifying 64Gb max ram size')
table = snapshot['tables']['eosio']['eosio']['global']
params = table[table.keys()[0]]['data']
max_ram_size = int(params["max_ram_size"]) >> 30
if max_ram_size != 64:
warning()
print "> Max ram size != 64Gb : (%d)" % max_ram_size
else:
success()
return True
def update_cluster_image(software, new_package):
"""Set the new version of the software on the cluster"""
step("Update the cluster to the new image")
software.version = new_package.version
try:
if Node.count_collection() > 1:
# make sure cluster HA is enabled
cli = CLI()
cli.execute("cluster ha modify", body={"configured": "true"})
software.patch(poll_timeout=1200, poll_interval=30, skip_warnings="true")
except NetAppRestError:
print("Current cluster software status:")
software.get()
pprint(software)
def validate_no_privileged_accounts(snapshot):
step('Verifying privileged accounts')
privileged_accounts = []
for sa in system_accounts:
if system_accounts[sa]['privileged']:
privileged_accounts.append(sa)
for name in snapshot['accounts']:
if snapshot['accounts'][name][
'privileged'] and name not in privileged_accounts:
warning()
print "> Invalid privileged account found : %s" % name
else:
success()
return True
def validate_account_permissions(snapshot, balances):
step('Verifying user account permission')
accounts = snapshot['accounts']
invalid_perm = 0
for account in balances:
perms = snapshot['accounts'][account]['permissions']
if authority_controlled_by_one_key(perms["active"], balances[account][CSV_EOS_ADDR]) != True or \
authority_controlled_by_one_key(perms["owner"], balances[account][CSV_EOS_ADDR]) != True:
invalid_perm += 1
if invalid_perm > 0:
warning()
print "> %d accounts with invalid permission" % (with_code)
else:
success()
def MCTS_search(self, state: str, history: list, last_action: tuple):
"""
以state为根节点进行MCTS搜索，搜索历史保存在histoty之中
:param state: 一个字符串代表的当前状态，根节点
:param history: 包含当前状态的一个列表
:param last_action: 上一次的落子位置
:return:
"""
while True:
board = utils.state_to_board(state, self.config.board_size)
game_over, v = utils.is_game_over(board, self.goal) # 落子前检查game over
if game_over:
self.update_tree(v, history=history)
break
if state not in self.tree:
# 未出现过的state，则评估然后展开
v = self.evaluate_and_expand(state, board, last_action) # 落子前进行评估
self.update_tree(v, history=history)
break
sel_action = self.select_action_q_and_u(state) # 根据state选择一个action
history.append(sel_action) # 放进action
board = utils.step(board, sel_action)
state = utils.board_to_state(board)
history.append(state)
last_action = sel_action
def run(self, e=0.25):
"""
对弈一局，获得一条数据，即从初始到游戏结束的一条数据
:return:
"""
state = self.get_init_state()
game_over = False
data = [] # 收集(状态，动作)二元组
value = 0
last_action = None
while not game_over:
policy, action = self.get_action(state, e, last_action)
data.append((state, policy, last_action)) # 装初始局面不装最终局面，装的是动作执行之前的局面
board = utils.step(utils.state_to_board(state, self.config.board_size), action)
state = utils.board_to_state(board)
# self.pruning_tree(board, state) # 走完一步以后，对其他分支进行剪枝，以节约内存；注释掉，以节约时间
game_over, value = utils.is_game_over(board, self.goal)
# assert value != 1.0
last_action = action
self.reset() # 把树重启
turns = len(data)
if turns % 2 == 1:
value = -value
weights = utils.construct_weights(turns, gamma=self.config.gamma)
final_data = []
for i in range(turns):
final_data.append((*data[i], value, weights[i])) # (状态，policy，last_action, value， weight)
value = -value
return final_data
def validate_no_code_in_accounts(snapshot, balances):
step('Verifying user account code')
with_code = 0
for account in balances:
tick()
acnt = snapshot['accounts'][account]
if int('0x'+acnt['code_version'],16) != 0 or \
acnt['abi'] != "" or \
acnt['vm_type'] != 0 or \
acnt['vm_version'] != 0:
with_code += 1
if with_code > 0:
warning()
print "> %d accounts with code set" % (with_code)
else:
success()
def validate_EOS_token(snapshot, balances, args):
step('Verifying EOS token')
symbol = args.core_symbol
key = name_to_string(symbol2int(symbol), False)
sym = str(symbol2int(symbol))
eos_token = snapshot['tables']['eosio.token'][key]['stat'][sym]['data']
ok = True
if asset2int(eos_token['max_supply']) != 100000000000000:
if ok:
warning()
ok = False
print "> EOS max supply != 10000M (%s)" % eos_token['max_supply']
# Calc supply:
def EOS_for_N_buys_of_8k_ram(N):
return int(round(10000.0 * float((10**6) * N / ((2**23) - N))))
# 163928 number of registered accounts in final snapshot
expected_supply = 10000000000000 + EOS_for_N_buys_of_8k_ram(163928)
if abs(asset2int(eos_token['supply']) - expected_supply) > 20000:
if ok:
warning()
ok = False
print "> EOS supply %s != %s" % (eos_token['max_supply'],
int2asset(expected_supply, "EOS"))
tokens = 0
code = snapshot['tables']['eosio.token']
for s in code:
if 'stat' in code[s]: tokens += 1
if tokens != 1:
if ok:
warning()
ok = False
print "> more than one token found in eosio.token"
if ok:
success()
return True
def delete_account():
"""Delete the User"""
step("Delete the limited account and role")
show_account()
accname = input(
"Enter the name of the Account that needs to be deleted:- ")
try:
account = Account.find(name=accname)
except NetAppRestError as error:
print("Exception caught :" + str(error))
substep("Delete our %s account" % accname)
try:
account.delete()
except NetAppRestError as error:
print("Exception caught :" + str(error))
def validate_global_params_against_genesis(snapshot, genesis):
step('Verifying global params vs genesis')
table = snapshot['tables']['eosio']['eosio']['global']
params = table[table.keys()[0]]['data']
diff = []
for gp in genesis['initial_configuration']:
if genesis['initial_configuration'][gp] != params[gp]:
diff.append(gp)
if len(diff):
warning()
print "Global params different from genesis : %s" % diff
else:
success()
return True
def download_new_cluster_image(parsed_args: argparse.Namespace):
"""Start our HTTP server and tell ONTAP to start downloading the new image"""
step("Download a new software package")
substep("Start an HTTP server to serve the file")
image_server = Thread(target=serve_image_download, args=(parsed_args,))
image_server.start()
substep("Have ONTAP download the package")
local_ip = socket.gethostbyname(socket.gethostname())
url = "http://%s:%s/%s" % (local_ip, parsed_args.port, parsed_args.image_path)
download_package = SoftwarePackageDownload(url=url)
download_package.post(poll_timeout=1200, poll_interval=60)
image_server.join()
substep("Show the new package")
packages = list(SoftwarePackage.get_collection())
print("Available software packages:", packages)
return packages[0]
def update_account():
"""Module to update the account details with new role"""
step("Update the account with new roles")
show_account()
accname = input(
"Enter the name of the Account that needs to be updated with the new role:- "
)
try:
account = Account.find(name=accname)
except NetAppRestError as error:
print("Exception caught :" + str(error))
print("======================")
substep("Create the new role")
rolename = input("Enter the name of the Role to be created:- ")
privpath = input(
"Enter the name of the Command Directoey PATH [eg: /api/storage/volume]:- "
)
pathaccess = input(
"Enter the Access for the Command Directory PATH [none/all/readonly]:- "
)
role = Role(
name=rolename,
privileges=[
RolePrivilege(access=pathaccess, path=privpath),
],
)
try:
role.post()
except NetAppRestError as error:
print("Exception caught :" + str(error))
print("New role created: %s" % rolename)
substep("Assign the test account to the test role")
account.role = role
try:
account.patch()
except NetAppRestError as error:
print("Exception caught :" + str(error))
print("======================")
print("Account updated with the new role: %s" % account)
def validate_account_creation(snapshot, balances):
found_users = {}
print_count = 0
step('Verifying user account creation')
for account in balances.keys():
tick()
if account not in snapshot['accounts']:
print_count = print_some(print_count, account)
continue
found_users[account] = balances[account]
not_found = len(balances) - len(found_users)
if not_found:
print "> %d (out of %d) accounts were not found in the EOS snapshot" % (
not_found, len(balances))
else:
success()
return found_users
def load_eos_snapshot(args):
step('Loading EOS snapshot')
with open(args.snapshot) as file:
snapshot = json.load(file)
# Convert list to dictionary
accounts = {}
for acc in snapshot['accounts']:
tick()
accounts[acc['name']] = acc
snapshot['accounts'] = accounts
# Add permissions to account dict
for perm in snapshot['permissions']:
tick()
if perm['id']['_id'] == 0: continue
owner = perm['owner']
if owner not in snapshot['accounts']:
raise ValidationException(
"invalid snapshot: permission owner not found")
if 'permissions' not in snapshot['accounts'][owner]:
snapshot['accounts'][owner]['permissions'] = {}
snapshot['accounts'][owner]['permissions'][perm['name']] = perm
# Tables
tables = {}
for t in snapshot['tables']:
tick()
code = t['tid']['code']
scope = t['tid']['scope']
table = t['tid']['table']
if code not in tables:
tables[code] = {}
if scope not in tables[code]:
tables[code][scope] = {}
tables[code][scope][table] = t['rows']
snapshot['tables'] = tables
success()
return snapshot
def generate_episode(pi, returns, counts, epsilon, start_from=None):
"""
generates an episode in a uniformly chosen random state and random action
then follows policy pi
updates returns and counts
first visit variant
"""
assert (np.array_equal(pi[0, :], np.zeros(n + 1)))
assert (np.allclose(pi.sum(axis=0)[1:], np.ones(n)))
sample_uniform_action = lambda s: np.random.choice(s) + 1 # from 1 to s
if start_from:
s, a = start_from
if a == None:
a = sample_uniform_action(s)
else:
s = np.random.choice(n - 1) + 1 # from 1 to n-1
a = sample_uniform_action(s)
def eps_greedy_sample_action(s):
greedy = np.random.choice(2, p=[epsilon, 1 - epsilon])
if not greedy:
return sample_uniform_action(s)
else:
return np.random.choice(n + 1, p=pi[:, s]) # from greedy policy pi
states = [s]
actions = [a]
rewards = [0]
first_visited = np.ones(n + 1, dtype=int) * (-1)
while s != n and s != 0:
if first_visited[s] == -1:
first_visited[s] = len(states) - 1
s_n, r = step(s, a, p_h, n)
states.append(s_n)
rewards.append(r)
if s_n != 0 and s_n != n:
actions.append(eps_greedy_sample_action(s_n))
else:
actions.append(0)
s = states[-1]
a = actions[-1]
for s, t in enumerate(first_visited):
if t == -1: # never visited
continue
a = actions[t]
returns[a, s] += sum(rewards[t + 1:])
if first_visited[s] == t:
counts[a, s] += 1
def _get_reward(self, offset=3):
golden_standard_db = self.golden_standard_db
data_cur = []
if golden_standard_db[0][0] is None:
print("THE GOLD STANDARD IS MORE LIKE SILVER...[?] HMMM")
#print(self.current_data)
#print(self.golden_standard_db)
try:
sys.exit(-1)
except SystemExit:
os._exit(-2)
else:
tmp = golden_standard_db[0][0].lower().replace(' ', '')
golden_standard_db = [(tmp, golden_standard_db[0][1])]
"""
data_cur.append((tup[0][0].lower().replace(' ', ''), tup[0][1]))AttributeError: 'spacy.tokens.span.Span' object has no attribute 'lower'
"""
for tup in self.current_db:
data_cur.append((str(tup[0][0]).lower().replace(' ',
''), tup[0][1]))
a = set(golden_standard_db)
if len(a) == 0:
print("Well josue, the world is weird")
try:
print("ERROR IN THE FUNCTION _get_reward()")
sys.exit(-1)
except SystemExit:
os._exit(-2)
# TODO: PA: it shouldn't be the extracted NER from the snippet in self.current_data ?
b = set(data_cur)
# Jaccard index - penalty
# penalty = e^(alpha * len(b)) * u(len(b)-offset) + min (edit_distance(A,B)) / len(A_content)
edit_vect = np.array(utils.edit_distance(a, b)) # Range: [0, inf)
penalty = m.pow(
m.e, self.alpha_reward * len(b)) * utils.step(len(b) - offset)
penalty += edit_vect.mean() / utils.len_content(a)
reward_cur = (len(a.intersection(b)) / len(a.union(b))) - penalty
reward = reward_cur - self.reward_prev
self.reward_prev = reward_cur
return reward
def validate_genesis(snapshot, genesis):
step('Validating genesis')
ddiff = DictDiffer(snapshot['genesis_state'], genesis)
added = ddiff.added()
removed = ddiff.removed()
changed = ddiff.changed()
if len(added) != 0 or len(removed) != 0 or len(changed) != 0:
warning()
tmp = tempfile.mkstemp()[1]
with open(tmp, 'w') as out:
if len(added) != 0:
out.write("> Snapshot has '%s' and your genesis dont\n" %
(','.join(added)))
if len(removed) != 0:
out.write("> Your genesis has %s and snapshot dont\n" %
(','.join(removed)))
if len(changed) != 0:
out.write("> Your genesis and snapshot have different '%s'\n" %
(','.join(changed)))
out.write('\n')
out.write("# Snapshot genesis\n")
out.write(
json.dumps(snapshot['genesis_state'], indent=2,
sort_keys=True))
out.write('\n')
out.write("# Your genesis\n")
out.write(json.dumps(genesis, indent=2, sort_keys=True))
out.write('\n')
print "> please check %s for details" % tmp
else:
success()
return True
def main() -> None:
"""Main function"""
arguments = [
Argument("-c",
"--cluster",
"API server IP:port details",
required=True),
Argument(
"-a",
"--cert_account",
"Account which be used for certificate authentication",
default="admin",
),
Argument(
"-o",
"--organization",
"Organization name for the root certificate",
default="MyCompany",
)
]
args = parse_args(
"This script will enable certificate authentication on the provided account (or admin)",
arguments,
)
setup_logging()
setup_connection(args.cluster, args.api_user, args.api_pass)
# make a temp directory for certs to live in
dirpath = tempfile.mkdtemp()
step("Install a root CA for %s" % args.organization)
my_ca = install_cert(args)
step("Sign a CSR for cert_user with our root")
sign_domain(my_ca, dirpath, args)
step("Install signed cert as client-ca")
enable_cert_auth(args)
step("Verify cert auth works for our user")
test_cert_auth(args, dirpath)
def make_preds(dataset, encoder, decoder, dictionary, batch_size,
cuda, max_length):
# Turn on evaluation mode which disables dropout.
encoder.eval()
decoder.eval()
utils.set_gradient(encoder, False)
utils.set_gradient(decoder, False)
pred_corpus_tokens = []
gold_corpus_tokens = []
start_time = time.time()
minibatches = utils.minibatch_generator(batch_size, dataset, cuda)
for n_batch, batch in enumerate(minibatches):
_, pred, _ = utils.step(encoder, decoder, batch, None, False, args,
beam_size=args.beam_size)
# true target
gold = batch[1].data
for i in xrange(pred.size(1)):
# get tokens from the predicted iindices
pred_tokens = [dictionary.idx2word[x] for x in pred[:, i]]
gold_tokens = [dictionary.idx2word[x] for x in gold[:, i]]
# filter out u'<pad>', u'<eos>'
filter_tokens = [u'<pad>', u'<eos>']
pred_tokens = filter(lambda x: x not in filter_tokens, pred_tokens)
gold_tokens = filter(lambda x: x not in filter_tokens, gold_tokens)
pred_corpus_tokens.append(pred_tokens)
gold_corpus_tokens.append(gold_tokens)
if n_batch % args.log_interval == 0 and n_batch > 0:
elapsed = time.time() - start_time
print('| {:5d}/{:5d} batches | ms/batch {:5.2f} |'.format(
n_batch, len(dataset[0]) // args.batch_size,
elapsed * 1000 / args.log_interval))
start_time = time.time()
return pred_corpus_tokens, gold_corpus_tokens
def GetPath(self):
path = []
env = deepcopy(self.env)
i, j, dir, hasKey, is_locked = self.init_state
done = False
while not done:
action = self.policyMap[i][j][dir][hasKey][is_locked]
path.append(action)
_, done = step(env, action)
i, j = env.agent_pos
vec = env.dir_vec
if vec[0] == 0:
dir = 1 + vec[1]
else:
dir = 2 - vec[0]
hasKey = 1 if env.carrying else 0
is_locked = 1 if env.grid.get(
self.info['door_pos'][0],
self.info['door_pos'][1]).is_locked else 0
return path