def post(self):
"""
Store sent file
:return: Success or fail
"""
# Get file from request content
fin = request.files[Const.FILE + '[0]']
file_name = str(fin).split('\'')[1].split('\'')[0]
fu.log(fu.get_current_time() + '[' + Const.SEND_CSV + ' ' +
request.method + '] Received request to store csv file:' +
file_name + '\n')
# Store file
result = csvh.add_file(fin, file_name)
return result[0], result[1]
def post(self):
"""
Try to execute the given query
:return: A float number representing the query result or an error message
"""
# Decrypt data received from Cloud Provider
content = request.get_json()
user_id = content[Const.ID]
file_name = content[Const.FILE]
query = content[Const.QUERY]
epsilon = float(content[Const.EPSILON])
fu.log(fu.get_current_time() + '[' + Const.QUERY + ' ' +
request.method + '] Received from ' + user_id +
' request for query:' + query + ' from ' + file_name +
' with epsilon = ' + str(epsilon) + '\n')
result = qh.check_query(user_id, file_name, query, epsilon)
return result[0], result[1]
def exec_query_operation(operation, epsilon, budget, lower, upper):
os.chdir(Const.ROOT_PATH + Const.DIFF_PRIV_MASTER_PATH)
try:
subprocess.check_output(['%s %s%s:priv_%s -- %f %f %f %f' % (Const.BAZEL_RUN, Const.DIFF_PRIV_PATH,
Const.OPERATIONS_PATH, operation, epsilon, budget,
lower, upper)], shell=True, stderr=subprocess.STDOUT)
except subprocess.CalledProcessError:
# Execution failed
os.chdir(Const.PARENT_DIR)
fu.log(fu.get_current_time() + Const.NO_RESULT[0] + '\n')
return Const.NO_RESULT
os.chdir(Const.PARENT_DIR)
if os.path.exists(Const.RESULT_PATH):
data = pd.read_csv(Const.RESULT_PATH, header=None)
# Extract results
true_value = str(data[0].values[0])
anon_value = str(data[1].values[0])
fu.log(fu.get_current_time() + 'Result of ' + operation + ' has true value = ' + true_value +
' and anonymized value = ' + anon_value + '\n')
# Remove result file
os.remove(Const.RESULT_PATH)
return anon_value, Const.OK
# Execution failed
fu.log(fu.get_current_time() + Const.NO_RESULT[0] + '\n')
return Const.NO_RESULT
def add_file(fin, file_name):
# Check if file is a csv
if not file_name.split('.')[1] in Const.SUPPORTED_FILE_FORMAT:
fu.log(fu.get_current_time() + 'File ' + file_name +
' is not supported\n')
return Const.INVALID_FILE
# Check if given file already exists
if os.path.exists(Const.ROOT_PATH + Const.CSV_FILES_PATH + file_name):
fu.log(fu.get_current_time() + 'File already exists in directory ' +
Const.ROOT_PATH + Const.CSV_FILES_PATH +
' and it can\'t be stored\n')
return Const.FILE_EXIST
# Save file in csv files directory
fin.save(Const.ROOT_PATH + Const.CSV_FILES_PATH + file_name)
# Check if given file has numeric columns
if not fu.has_numeric_columns(Const.ROOT_PATH + Const.CSV_FILES_PATH +
file_name):
os.remove(Const.ROOT_PATH + Const.CSV_FILES_PATH + file_name)
fu.log(fu.get_current_time() +
'File has no numeric columns and it can\'t be stored\n')
return Const.NO_NUMERIC
fu.log(fu.get_current_time() + 'File stored in directory ' +
Const.ROOT_PATH + Const.CSV_FILES_PATH + '\n')
return 'File stored!', Const.OK
def check_query(user, file_name, query, epsilon):
# Check if file for query exists
if not os.path.exists(Const.ROOT_PATH + Const.CSV_FILES_PATH + file_name):
fu.log(fu.get_current_time() + 'User ' + user + ' made a query on ' + file_name + ' that does not exist\n')
return Const.FILE_NOT_EXIST
# Parse the query
statement, operation, column, lower, upper = fu.parse_query(query)
operation = operation.lower()
# Check if operation is valid
if not check_operation(statement.upper(), operation):
fu.log(fu.get_current_time() + Const.INVALID_OPERATION[0] + '\n')
return Const.INVALID_OPERATION
# Check given bounds and clamp if they are out of limits
lower, upper = clamp(lower, upper)
if lower == Const.INVALID_BOUNDS[0]:
fu.log(fu.get_current_time() + Const.INVALID_BOUNDS[0] + '\n')
return Const.INVALID_BOUNDS
# Extract data
full_data = pd.read_csv(Const.ROOT_PATH + Const.CSV_FILES_PATH + file_name, header=0)
# Check if column exists
if column not in full_data.columns:
fu.log(fu.get_current_time() + Const.INVALID_COLUMN[0] + '\n')
return Const.INVALID_COLUMN
# Get data according the given column
data = full_data[column]
# Check if data to use is numeric
if fu.is_numeric(data):
df = full_data[[full_data.columns[0], column]]
# df = df.loc[(df[column] > lower) & (df[column] < upper)]
df.to_csv(Const.ROOT_PATH + Const.DIFF_PRIV_MASTER_PATH + Const.DIFF_PRIV_PATH + Const.OPERATIONS_PATH + '/'
+ Const.TMP_FILE_PATH, header=False, index=False)
# Check if users list file exists
if os.path.exists(Const.ROOT_PATH + Const.USERS_LIST_PATH + Const.USERS):
data = pd.read_csv(Const.ROOT_PATH + Const.USERS_LIST_PATH + Const.USERS, header=0)
# Verify if the given user made previous queries and check its remaining budget
if user in data[[Const.ID]].values:
row = data[data[Const.ID] == user].index.tolist()[0]
# User has not enough remaining budget
if data.iloc[row][1] < Const.QUERY_BUDGET:
fu.log(
fu.get_current_time() + 'User ' + user + ' has not enough budget (' + str(data.iloc[row][1]) +
') to compute query\n')
return Const.NO_BUDGET
fu.log(fu.get_current_time() + 'User ' + user + ' found with budget ' + str(data.iloc[row][1]) +
', enough to compute query\n')
# User has enough budget, then decrease it
budget = data.iloc[row][1]
data.iat[row, 1] -= Const.QUERY_BUDGET
fu.log(fu.get_current_time() + 'User ' + user + ' budget updated to ' + str(data.iloc[row][1]) + '\n')
else:
# User not found, then add a new one
df = pd.DataFrame({Const.ID: [user], Const.BUDGET: [(Const.STARTING_BUDGET - Const.QUERY_BUDGET)]},
columns=[Const.ID, Const.BUDGET])
data = data.append(df, ignore_index=True)
budget = Const.STARTING_BUDGET
fu.log(fu.get_current_time() + 'User ' + user + ' not found. Created a new one with budget ' + str(
budget) +
'\n')
fu.log(
fu.get_current_time() + 'User ' + user + ' budget updated to ' + str(budget - Const.QUERY_BUDGET) +
'\n')
# Overwrite users list file
data.to_csv(Const.ROOT_PATH + Const.USERS_LIST_PATH + Const.USERS, index=False)
else:
# File does not exist, create a new one
df = pd.DataFrame({Const.ID: [user], Const.BUDGET: [Const.STARTING_BUDGET - Const.QUERY_BUDGET]},
columns=[Const.ID, Const.BUDGET])
df.to_csv(Const.ROOT_PATH + Const.USERS_LIST_PATH + Const.USERS, index=False)
budget = Const.STARTING_BUDGET
fu.log(fu.get_current_time() + 'File ' + Const.ROOT_PATH + Const.USERS_LIST_PATH + Const.USERS +
' not found. Created a new one and added user ' + user + ' with budget ' + str(budget) + '\n')
fu.log(fu.get_current_time() + 'User ' + user + ' budget updated to ' + str(
budget - Const.QUERY_BUDGET) + '\n')
# Execute the query
return exec_query_operation(operation, epsilon, budget, lower, upper)
fu.log(fu.get_current_time() + Const.NO_NUMERIC_QUERY[0] + '\n')
return Const.NO_NUMERIC_QUERY
