def __init__(self, rules_file, s3_base_url=None, s3_atch_dir=None, aws_access_key_id=None, aws_secret_access_key=None):
with open(rules_file, "rb") as file_handle:
self.rules = json.loads(file_handle.read().decode("utf-8"))
# '__rootdir__' is the root directory in s3 we expect to find the email
# files. save it to the object, then delete it from the rules (since
# it's not actually a rule)
if "__rootdir__" in self.rules:
del self.rules["__rootdir__"]
for dest in self.rules:
rule_sets = to_list(self.rules[dest])
for rule_set in rule_sets:
for key in rule_set:
rule_set[key] = to_list(rule_set[key])
self.rules[dest] = rule_sets
self.client = None
if not s3_base_url:
return
# split the 's3_base_url' into the bucket / directory format that the
# boto library expects.
s3_base_url = s3_base_url.strip("/ ")
if s3_base_url.startswith("s3://"):
s3_base_url = s3_base_url[5:]
parts = s3_base_url.split("/")
self.s3_bucket = parts[0]
self.s3_base_dir = "/".join(parts[1:])
self.attachment_dest_dir = self.s3_base_dir + "/" + s3_atch_dir
#logging.info("email processor", self.s3_bucket, self.s3_base_dir, self.attachment_dest_dir)
self.client = s3.Client(key=aws_access_key_id, secret=aws_secret_access_key)
def _build_project(project, env):
children = []
children += to_list(project.pre_build(env))
children += to_list(project.build(env))
children += to_list(project.post_build(env))
children += to_list(project.install(env))
return children
def index():
bundles = db.fetch("SELECT bundle AS path, COUNT(id) AS count FROM " + db.tbl_image + " GROUP BY bundle ORDER BY path")
current_labels = to_list(request.args.get('labels', None))
current_not_labels = to_list(request.args.get('notlabels', None))
related_labels = get_related_labels(current_labels, current_not_labels)
labels = db.fetch("SELECT id, name, COUNT(image) AS count FROM " + db.tbl_label + " INNER JOIN " +
db.tbl_image_label + " ON (label = id) GROUP BY id ORDER BY name ASC")
return render_template('index.html', config=config, bundles=bundles, labels=labels,
current_labels=current_labels, current_not_labels=current_not_labels,
related_labels=related_labels)
def authorize(auth, group_name, region):
has_from = 'from' in auth
has_to = 'to' in auth
if has_from and has_to:
raise Exception("illegal authorization block: using both 'from' and 'to' is not allowed: " + str(auth))
if not has_from and not has_to:
raise Exception("illegal authorization block: neither 'from' nor 'to' was specified: " + str(auth))
if has_from:
from_addrs = util.to_list(auth['from'])
for from_addr in from_addrs:
do_authorize(from_addr, group_name, auth, region)
else:
to_addrs = util.to_list(auth['to'])
for to_addr in to_addrs:
do_authorize(group_name, to_addr, auth, region)
def __init__(self, manager, key, uselist, callable_, typecallable, trackparent=False, extension=None, **kwargs):
self.manager = manager
self.key = key
self.uselist = uselist
self.callable_ = callable_
self.typecallable= typecallable
self.trackparent = trackparent
self.extensions = util.to_list(extension or [])
def init():
"""
Clear old files and iterate posts and collect basic info
"""
global _post_page_count
global _post_file_names
global _tags
global _categories
# remove old static files
if os.path.exists(_deploy_dir):
for file in os.listdir(_deploy_dir):
if file.startswith('.'):
continue
path = os.path.join(_deploy_dir, file)
if os.path.isdir(path):
shutil.rmtree(path)
else:
os.remove(path)
else:
os.mkdir(_deploy_dir)
file_list = util.get_posts_list()
# get post count
if C.entry_count_one_page == 0:
_post_page_count = 1
else:
_post_page_count = len(file_list) // C.entry_count_one_page + 1
for file in file_list:
# get post file names
_post_file_names.append(os.path.splitext(file)[0])
# get tags and categories
file_path = os.path.join('posts', file)
d = util.render_yaml(util.read_md_file_head(file_path))
if 'categories' in d:
_categories |= set(util.to_list(d['categories']))
if 'tags' in d:
_tags |= set(util.to_list(d['tags']))
return True
def init():
"""
Clear old files and iterate posts and collect basic info
"""
global _post_page_count
global _post_file_names
global _tags
global _categories
# remove old static files
if os.path.exists(_deploy_dir):
for file in os.listdir(_deploy_dir):
if file.startswith('.'):
continue
path = os.path.join(_deploy_dir, file)
if os.path.isdir(path):
shutil.rmtree(path)
else:
os.remove(path)
else:
os.mkdir(_deploy_dir)
file_list = util.get_posts_list()
# get post count
if C.entry_count_one_page == 0:
_post_page_count = 1
else:
total = len(file_list)
_post_page_count = total // C.entry_count_one_page \
+ (1 if total % C.entry_count_one_page != 0 else 0)
for file in file_list:
# get post file names
_post_file_names.append(os.path.splitext(file)[0])
# get tags and categories
file_path = os.path.join('posts', file)
d = util.render_yaml(util.read_md_file_head(file_path))
if 'categories' in d:
_categories |= set(util.to_list(d['categories']))
if 'tags' in d:
_tags |= set(util.to_list(d['tags']))
def do_authorize(from_group, to_group, auth, region):
protocols = [ 'tcp' ]
if 'protocols' in auth:
protocols = util.to_list(auth['protocols'])
account = str(region['aws_account'])
region_name = str(region['EC2_REGION']).strip()
auths = 0
for protocol in protocols:
if protocol == 'icmp':
authorize_icmp(from_group, to_group, auth['icmp_types'], account, region_name)
auths += 1
else:
if 'ports' not in auth:
raise Exception("no ports specified in auth: "+str(auth))
ports = util.to_list(auth['ports'])
for port in ports:
authorize_normal(from_group, to_group, port, protocol, account, region_name)
auths += 1
if auths == 0:
raise Exception("no authorizations made for auth: "+str(auth))
def __init__(self,
prob=1.0,
words_substr_triggers=None,
words_trigger=None,
exact_trigger=None,
substrings_trigger=None,
custom_trigger=None,
bot_only=False,
active_guild=None,
active_channel=None,
active_users=None,
regex_trigger=None,
delete_message=False,
log=False):
if words_trigger is None:
words_trigger = []
if exact_trigger is None:
exact_trigger = []
if substrings_trigger is None:
substrings_trigger = []
if words_substr_triggers is None:
words_substr_triggers = []
if regex_trigger is not None:
regex_trigger = re.compile(regex_trigger)
words_trigger = to_list(words_trigger)
exact_trigger = to_list(exact_trigger)
substrings_trigger = to_list(substrings_trigger)
active_guild = to_list(active_guild)
active_channel = to_list(active_channel)
active_users = to_list(active_users)
words_substr_triggers = to_list(words_substr_triggers)
extra_words, extra_substr = split_words_phrases(words_substr_triggers)
words_trigger += extra_words
substrings_trigger += extra_substr
self.prob = prob
self.words_trigger = words_trigger
self.exact_trigger = exact_trigger
self.substrings_trigger = substrings_trigger
self.custom_trigger = custom_trigger
self.bot_only = bot_only # If False, won't react to bots, if True will react only to bots
self.active_guilds = active_guild # List of guild name in which it can get triggered
self.active_channels = active_channel # List of channel ids in which it can get triggered
self.active_users = active_users
self.regex_trigger = regex_trigger
self.delete_message = delete_message
self.log = log
def run(self, env):
sh = env.shell
def _expand_vars(cmd):
cmd_type = type(cmd)
if cmd_type == str:
cmd = replace_variables(cmd, env.config['variables'])
elif cmd_type == list:
cmd = [
replace_variables(sub, env.config['variables'])
for sub in cmd
]
return cmd
# Interpolate any variables
self.commands = [_expand_vars(cmd) for cmd in self.commands]
# Run each of the commands
children = []
for cmd in self.commands:
cmd_type = type(cmd)
# See if the string is actually an action
if cmd_type == str:
action_cls = Scripts.find_action(cmd)
if action_cls:
cmd = action_cls()
cmd_type = type(cmd)
if cmd_type == str:
result = sh.exec(*cmd.split(' '))
if result.returncode != 0:
print('Command failed, exiting')
sys.exit(12)
elif cmd_type == list:
result = sh.exec(*cmd)
if result.returncode != 0:
print('Command failed, exiting')
sys.exit(12)
elif isinstance(cmd, Action):
Scripts.run_action(cmd, env)
elif callable(cmd):
children += to_list(cmd(env))
else:
print('Unknown script sub command: {}: {}', cmd_type, cmd)
sys.exit(4)
return children
def create_tokenizer():
"""
根据训练数据集中图像名，和其对应的标题，生成一个tokenizer,作为LSTM的输入/输出必须是数字，所以需要我们使用
字典数据类型来存储文字和数字对应关系。
:return: 生成的tokenizer
https://keras-cn.readthedocs.io/en/latest/legacy/preprocessing/text/#tokenizer
"""
train_image_names = util.load_image_names('{}{}{}'.format(current_path, os.sep, 'Flickr_8k.trainImages.txt'))
description_path = '{}{}{}'.format(current_path, os.sep, 'descriptions.txt')
train_descriptions = util.load_clean_captions(description_path, train_image_names)
lines = util.to_list(train_descriptions)
tokenizer = Tokenizer()
tokenizer.fit_on_texts(lines)
return tokenizer
def run_instance(unit, instance_number):
config = unit.config
region_config = unit.region_context.region_config
region_name = unit.region_context.region
args = ["ec2-run-instances", unit.get_ami(),
"--instance-initiated-shutdown-behavior", "terminate",
"--region", region_config['EC2_REGION'],
"-g", unit.env_group_name,
"-g", unit.group_name]
if 'extra_groups' in config:
for group in util.to_list(config['extra_groups']):
args.append("-g")
args.append(group)
args += [ "-t", unit.get_instance_size(), "-k", unit.get_keypair() ]
data = unit.get_user_data()
if data is not None:
data_file = util.write_temp_file("run_"+unit.unit_name+"_"+unit.role_name+"_userdata_", suffix=".json")
args += [ "-f", data_file ]
azone = unit.get_availability_zone()
if azone is not None:
args += [ "-z", azone]
elif 'default_availability_zone' in region_config:
args += [ "-z", region_config['default_availability_zone'] ]
output = util.execute_shell(args)
if util.is_noop():
iid = "i-debug"
else:
iid = util.find_element_on_line_starting_with(output[0], "INSTANCE", 1)
return iid
def bootstrap(unit, host, iid):
node_name = unit.group_name + "-" + iid
region_config = unit.region_context.region_config
# print ">>>>>bootstrap: region_context.region_config="+str(region_config)
knife_args = ["knife", "bootstrap", host,
"--run-list", "role["+unit.role_name+"]",
"--node-name", node_name,
"--ssh-user", "ubuntu",
"-i", util.subst_homedir(region_config['KEY_DIR']) +"/"+region_config['keypair']+".pem",
"--sudo", "--no-host-key-verify",
"--environment", unit.env_name]
util.execute_shell(knife_args)
util.execute_shell(["ec2-create-tags", iid, "--tag", "Name=" + node_name])
util.execute_shell(["create-cname", node_name+".unit", host+".", "86400"])
if POST_BOOTSTRAP in unit.config:
add_to_haproxy(unit, host, iid)
for type in ['sg', 'asg', 'tag']:
key = 'refresh_' + type + '_hosts'
if key in unit.config[POST_BOOTSTRAP]:
for group in util.to_list(unit.config[POST_BOOTSTRAP][key]):
util.execute_shell(["ssh-"+type+"-hosts", group, "sudo", "chef-client"])
file = './datasets/iris/iris.data'
#file = './datasets/pendigits/pendigits.tra'
#file = './datasets/satimage/sat.trn'
#file = './datasets/letter-recognition/letter-recognition.data'
# given
dataset = util.load_data(file, delimiter=',')
def remove_label(data):
return map(lambda x: x[:-1],
data)
data = remove_label(dataset)
data = util.to_number(data)
data = util.to_list(data)
data = util.rescale(data)
def get_label(data):
return map(lambda x: x[-1],
data)
target = get_label(dataset)
target = util.to_list(target)
#cluster_count_list = [26, 52, 104, 208, 416, 832]
cluster_count_list = [3]
for cluster_count in cluster_count_list:
print('cluster count:', cluster_count)
# cluster_count = 8
return tf.transpose(tf.concat([first, hats, last], axis=1))
def net_pred(self, x):
u = NNUtil.neural_net(x, self.weights, self.biases)
return u
if __name__ == "__main__":
working_dir = "./Test_Collection/CB/"
paras = util.load_parameter(working_dir + "paras.txt")
N = int(paras["N"])
width = int(paras["width"])
depth = int(paras["depth"])
max_epoch = int(paras["max_epoch"])
num_tested_per_element = int(paras["num_tested_per_element"])
pivot = util.to_list(paras["pivot"])
upper_bound = pivot[-1]
lower_bound = pivot[0]
num_element = len(pivot) - 1
hidden_layers = [width] * depth
new_dir = "{}_{}_{}/N={}/{}_{}/".format(lower_bound, upper_bound,
num_element, N, depth, width)
dump_dir = working_dir + new_dir
if not os.path.exists(dump_dir):
os.makedirs(dump_dir)
shutil.copy(working_dir + "paras.txt", dump_dir)
model = CBSolutionNN(N, pivot, hidden_layers)
def evaluate(args, model, dev_dataset, data_loader, gold_dict, tokenizer,
device, max_len, use_squad_v2):
dataset, examples, features = dev_dataset.dataset, dev_dataset.examples, dev_dataset.features
model.eval()
all_results = []
with torch.no_grad(), \
tqdm(total=len(data_loader.dataset)) as progress_bar:
for batch in data_loader:
batch = tuple(t.to(device) for t in batch)
inputs = {
"input_ids": batch[0],
"attention_mask": batch[1],
"token_type_ids": batch[2],
}
if args.bidaf:
inputs['char_ids'] = batch[6]
example_indices = batch[3]
batch_size = inputs["input_ids"].size(0)
# Forward
outputs = model(**inputs)
for i, example_index in enumerate(example_indices):
eval_feature = features[example_index.item()]
unique_id = int(eval_feature.unique_id)
output = [to_list(output[i]) for output in outputs]
start_logits, end_logits = output
result = SquadResult(unique_id, start_logits, end_logits)
all_results.append(result)
# Log info
progress_bar.update(batch_size)
if args.dev_logits_save_file is not None:
with open(args.dev_logits_save_file, 'wb') as f:
pickle.dump(all_results, f)
pred_dict = compute_predictions_logits(
examples,
features,
all_results,
1,
args.max_ans_len,
True,
'save/temp/predict_temp.json',
'save/temp/nbest_temp.json',
'save/temp/nlog_odd.log',
False,
args.use_squad_v2,
args.null_score_diff_threshold,
tokenizer,
)
model.train()
results = util.eval_dicts(gold_dict, pred_dict, use_squad_v2)
results_list = [('F1', results['F1']), ('EM', results['EM'])]
if use_squad_v2:
results_list.append(('AvNA', results['AvNA']))
results = OrderedDict(results_list)
return results, pred_dict
def select_only(label, data):
return filter(lambda x: x[4] == label,
data)
data = select_only(label, data)
def remove_label(data):
return map(lambda x: x[:-1],
data)
data = remove_label(data)
data = util.to_number(data)
data = util.to_list(data)
data = util.rescale(data)
# data
x = list(map(lambda x: x[features[0]],
data))
y = list(map(lambda x: x[features[1]],
data))
nullfmt = NullFormatter() # no labels
# definitions for the axes
left, width = 0.1, 0.65
bottom, height = 0.1, 0.65
bottom_h = left_h = left + width + 0.02
import keras.layers as klay import keras.backend as K import keras.callbacks as kclbk import util import polylib import skbasis as skb _epsilon = 1.0e-14 ## ========================================= ## root_dir = "./Test_Collection/" result_dir = root_dir + "CB_keras/" paras = util.load_parameter(root_dir + "paras.txt") # domain contains coordinates of element boundaries domain = util.to_list(paras["pivot"]) depth = int(sys.argv[2]) width = int(sys.argv[3]) max_epoch = int(sys.argv[4]) # domain parameters N_elem = len(domain) - 1 # number of elements upper_bound = domain[-1] lower_bound = domain[0] # C_k continuity CK = 0 # C^k continuity LAMBDA_COEFF = 1.0 # lambda constant aa = 3.0
