def test_read_valid_file_should_return_non_empty_dictionary():
# arrange
sut = PcapReader()
# act
dir_path = os.path.dirname(os.path.realpath(__file__))
res = sut.read(dir_path + "/files_fingerprint_recognizer/valid_input.pcap")
# assert
assert bool(res)
def test_read_valid_file_should_return_dictionary_of_flows():
# arrange
sut = PcapReader()
# act
dir_path = os.path.dirname(os.path.realpath(__file__))
res = sut.read(dir_path + "/files_fingerprint_recognizer/valid_input.pcap")
# assert
assert all(isinstance(x, Flow) for x in res.values())
def test_read_invalid_path_should_throw_exception():
# arrange
sut = PcapReader()
# act
try:
sut.read("")
assert False
except Exception as e:
print(e)
assert True
def test_read_should_not_throw_os_exception():
# arrange
sut = PcapReader()
dir_path = os.path.dirname(os.path.realpath(__file__))
# act
try:
res = sut.read(dir_path +
"/files_fingerprint_recognizer/valid_input.pcap")
except OSError as ex:
assert False
except:
assert True
def test_tshark_should_be_installed():
# arrange
sut = PcapReader()
dir_path = os.path.dirname(os.path.realpath(__file__))
# act
try:
res = sut.read(dir_path +
"/files_fingerprint_recognizer/valid_input.pcap")
except ModuleNotFoundError as ex:
assert False
except:
assert True
def run(pcaps_paths=[], flows_paths=[], write_flows_dir='', symbols_path='', out_dir='', train_size=0.0, recognizer=None):
# Phase 1: Read flows from pcap file or binary file
flows_per_app = dict()
if flows_paths:
for pcaps_path in flows_paths:
appName = Path(pcaps_path).parts[-2]
executionName = Path(pcaps_path).stem
flows_per_app[appName] = flows_per_app.get(appName, {'all': {}})
print(f"reading binaries from {pcaps_path}")
flows_per_app[appName]['all'][executionName] = BinaryReader.read(pcaps_path)
if write_flows_dir:
write_path = write_flows_dir + '/' + appName + '/' + executionName + '.p'
print(f"writing binaries into {write_path}")
BinaryReader.write(flows_per_app[appName]['all'][executionName], write_path)
elif pcaps_paths:
pcap_reader = PcapReader()
for pcaps_path in pcaps_paths:
appName = Path(pcaps_path).parts[-2]
executionName = Path(pcaps_path).stem
flows_per_app[appName] = flows_per_app.get(appName, {'all': {}})
print(f"reading pcaps from {pcaps_path}")
flows_per_app[appName]['all'][executionName] = pcap_reader.read(pcaps_path)
if write_flows_dir:
write_path = write_flows_dir + '/' + appName + '/' + executionName + '.p'
print(f"writing binaries into {write_path}")
BinaryReader.write(flows_per_app[appName]['all'][executionName], write_path)
else:
raise Exception("There is no input file specified")
# Phase 2: Save flows as binary if output path is speficied
# Phase 3: Load previously saved symbols dictionary
symbols = {}
if symbols_path:
symbols = BinaryReader.read(symbols_path, {})
# Phase 4: Split test/train data if needed
reshape_flows_per_app(flows_per_app, train_size)
# Phase 5: Generate sequence of symbols based on symbols set and flows
fingerprint_recognizer = build_recognizer(symbols, recognizer)
for appName, value in flows_per_app.items():
result = {}
for label, v in value.items():
result[label] = []
for executionName, flows in v.items():
# IT IS THE MAIN PART OF APPLICATION
r = fingerprint_recognizer.recognize(flows)
result[label].append(r)
# Phase 6: Save generated sequence of symbols
if out_dir:
save_app_flows(appName, result, out_dir);
# Phase 7: Save set of unique symbols to use in the future
if symbols_path:
symbols = fingerprint_recognizer.get_symbols()
BinaryReader.write(symbols, symbols_path)