def run(self):
"""Run androguard to extract static android information
@return: list of static features
"""
self.key = "apkinfo"
apkinfo = {}
if "file" not in self.task["category"]:
return
from androguard.core.bytecodes.apk import APK
from androguard.core.bytecodes.dvm import DalvikVMFormat
from androguard.core.analysis.analysis import uVMAnalysis
from androguard.core.analysis import analysis
f = File(self.task["target"])
if f.get_name().endswith((".zip", ".apk")) or "zip" in f.get_type():
if not os.path.exists(self.file_path):
raise CuckooProcessingError("Sample file doesn't exist: \"%s\"" % self.file_path)
try:
a = APK(self.file_path)
if a.is_valid_APK():
manifest = {}
apkinfo["files"] = self._apk_files(a)
manifest["package"] = a.get_package()
# manifest["permissions"]=a.get_details_permissions_new()
manifest["main_activity"] = a.get_main_activity()
manifest["activities"] = a.get_activities()
manifest["services"] = a.get_services()
manifest["receivers"] = a.get_receivers()
# manifest["receivers_actions"]=a.get__extended_receivers()
manifest["providers"] = a.get_providers()
manifest["libraries"] = a.get_libraries()
apkinfo["manifest"] = manifest
# apkinfo["certificate"] = a.get_certificate()
static_calls = {}
if self.check_size(apkinfo["files"]):
vm = DalvikVMFormat(a.get_dex())
vmx = uVMAnalysis(vm)
static_calls["all_methods"] = self.get_methods(vmx)
static_calls["is_native_code"] = analysis.is_native_code(vmx)
static_calls["is_dynamic_code"] = analysis.is_dyn_code(vmx)
static_calls["is_reflection_code"] = analysis.is_reflection_code(vmx)
# static_calls["dynamic_method_calls"]= analysis.get_show_DynCode(vmx)
# static_calls["reflection_method_calls"]= analysis.get_show_ReflectionCode(vmx)
# static_calls["permissions_method_calls"]= analysis.get_show_Permissions(vmx)
# static_calls["crypto_method_calls"]= analysis.get_show_CryptoCode(vmx)
# static_calls["native_method_calls"]= analysis.get_show_NativeMethods(vmx)
else:
log.warning("Dex size bigger than: %s",
self.options.decompilation_threshold)
apkinfo["static_method_calls"] = static_calls
except (IOError, OSError, zipfile.BadZipfile) as e:
raise CuckooProcessingError("Error opening file %s" % e)
return apkinfo
def run(self):
"""Run androguard to extract static android information
@return: list of static features
"""
self.key = "apkinfo"
apkinfo = {}
if "file" not in self.task["category"] or not HAVE_ANDROGUARD:
return
f = File(self.task["target"])
if f.get_name().endswith((".zip", ".apk")) or "zip" in f.get_type():
if not os.path.exists(self.file_path):
raise CuckooProcessingError(
"Sample file doesn't exist: \"%s\"" % self.file_path)
try:
a = APK(self.file_path)
if a.is_valid_APK():
manifest = {}
apkinfo["files"] = self._apk_files(a)
manifest["package"] = a.get_package()
# manifest["permissions"]=a.get_details_permissions_new()
manifest["main_activity"] = a.get_main_activity()
manifest["activities"] = a.get_activities()
manifest["services"] = a.get_services()
manifest["receivers"] = a.get_receivers()
# manifest["receivers_actions"]=a.get__extended_receivers()
manifest["providers"] = a.get_providers()
manifest["libraries"] = a.get_libraries()
apkinfo["manifest"] = manifest
# apkinfo["certificate"] = a.get_certificate()
static_calls = {}
if self.check_size(apkinfo["files"]):
vm = DalvikVMFormat(a.get_dex())
vmx = uVMAnalysis(vm)
static_calls["all_methods"] = self.get_methods(vmx)
static_calls[
"is_native_code"] = analysis.is_native_code(vmx)
static_calls["is_dynamic_code"] = analysis.is_dyn_code(
vmx)
static_calls[
"is_reflection_code"] = analysis.is_reflection_code(
vmx)
# static_calls["dynamic_method_calls"]= analysis.get_show_DynCode(vmx)
# static_calls["reflection_method_calls"]= analysis.get_show_ReflectionCode(vmx)
# static_calls["permissions_method_calls"]= analysis.get_show_Permissions(vmx)
# static_calls["crypto_method_calls"]= analysis.get_show_CryptoCode(vmx)
# static_calls["native_method_calls"]= analysis.get_show_NativeMethods(vmx)
else:
log.warning("Dex size bigger than: %s",
self.options.decompilation_threshold)
apkinfo["static_method_calls"] = static_calls
except (IOError, OSError, BadZipfile) as e:
raise CuckooProcessingError("Error opening file %s" % e)
return apkinfo
def extract_attributes(sha256):
with NamedTemporaryFile() as f:
f.write(default_storage.open(sha256).read())
f.seek(0)
sign = ApplicationSignature.compute_from_apk(f.name)
package = sign.handle
sign = sign.to_dict()
a = APK(f.name)
sign['uploaded_at'] = datetime.now()
sign['sha256'] = sha256
sign['activities'] = a.get_activities()
sign['features'] = a.get_features()
sign['libraries'] = a.get_libraries()
sign['main_activity'] = a.get_activities()
sign['min_sdk_version'] = a.get_min_sdk_version()
sign['max_sdk_version'] = a.get_max_sdk_version()
sign['target_sdk_version'] = a.get_target_sdk_version()
sign['permissions'] = a.get_permissions()
sign['aosp_permissions'] = a.get_requested_aosp_permissions()
sign[
'third_party_permissions'] = a.get_requested_third_party_permissions(
)
sign['providers'] = a.get_providers()
sign['receivers'] = a.get_receivers()
sign['services'] = a.get_services()
sign['is_valid'] = a.is_valid_APK()
sign['is_signed'] = a.is_signed()
sign['is_signed_v1'] = a.is_signed_v1()
sign['is_signed_v2'] = a.is_signed_v2()
sign['is_signed_v3'] = a.is_signed_v3()
if not es.exists(settings.ELASTICSEARCH_APK_INDEX, id=sha256):
es.index(index=settings.ELASTICSEARCH_APK_INDEX,
id=sha256,
body=sign)
else:
es.update(index=settings.ELASTICSEARCH_APK_INDEX,
id=sha256,
body={'doc': sign},
retry_on_conflict=5)
del a, sign, f
gc.collect()
return package
def extract_features_using_androguard(androguard: APK):
resulat_dict = dict()
count_dict = dict()
_permissions = androguard.get_permissions()
_hardwares = list(androguard.get_features())
np = len(_permissions)
if np > 0:
resulat_dict = {
'defined_perm': dict(Counter(_permissions)),
}
count_dict = {'num_permission': np}
nh = len(_hardwares)
if nh > 0:
resulat_dict.update({'hardware': dict(Counter(_hardwares))})
count_dict.update({'num_hardware': len(_hardwares)})
# List of all intents in apk
_filters = list()
# Apk components
components_dict = {
'activity': androguard.get_activities(),
'service': androguard.get_services(),
'receiver': androguard.get_receivers(),
'provider': androguard.get_providers()
}
# Extract intents
for categorie, names in components_dict.items():
# Components and Count the number of each components in apk
nm = len(names)
if nm > 0:
count_dict.update({'num_' + categorie: nm})
resulat_dict.update(
{categorie: dict(Counter(components_dict[categorie]))})
# Extract intent filter for each component
intents = extract_intent(androguard, categorie, names)
ni = len(intents)
if ni > 0:
_filters.extend(intents)
count_dict.update({'num_intent_' + categorie: ni})
nf = len(_filters)
if nf > 0:
resulat_dict.update({'intent': dict(Counter(_filters))})
count_dict.update({'num_intent': nf})
resulat_dict.update({'component_count': count_dict})
return resulat_dict
def get_data(apkPath, apkName):
a = APK(apkPath)
jsonFile = apkName + ".json"
pro = a.get_providers()
rec = a.get_receivers()
ser = a.get_services()
act = a.get_activities()
per = a.get_permissions()
hwc = get_hardware(a)
data = {'permissions': [], 'hardwareComponent': [], 'components': {}}
data['permissions'] = per
data['hardwareComponent'] = hwc
data['components']['providers'] = pro
data['components']['receivers'] = rec
data['components']['services'] = rec
data['components']['activities'] = act
with open(jsonFile, 'w') as f:
json.dump(data, f)
def lim_features_categories(apk_filepath):
try:
apk = APK(apk_filepath)
info = {
'declared permissions': sorted(apk.get_permissions()),
'activities': apk.get_activities(),
'services': apk.get_services(),
'intent filters': apk.get_intent_filters('receiver', ''),
'content providers': apk.get_providers(),
'broadcast receivers': apk.get_receivers(),
'hardware components': apk.get_features()
}
for category in info:
info[category] = [
feature.replace(".", "_").lower() for feature in info[category]
]
return info
except:
# We just do not process the APK
pass
def extract_features(file_path):
result = {}
try:
a = APK(file_path)
d = DalvikVMFormat(a.get_dex())
dx = Analysis(d)
vm = dvm.DalvikVMFormat(a.get_dex())
#vmx = analysis.uVMAnalysis(vm)
d.set_vmanalysis(dx)
d.set_decompiler(DecompilerDAD(d, dx))
except:
return None
result['android_version_code'] = a.get_androidversion_code()
result['android_version_name'] = a.get_androidversion_name()
result['max_sdk'] = a.get_max_sdk_version()
result['min_sdk'] = a.get_min_sdk_version()
#result['libraries'] = a.get_libraries()
result['filename'] = a.get_filename()
result['target_sdk'] = a.get_target_sdk_version()
result['md5'] = hashlib.md5(a.get_raw()).hexdigest()
result['sha256'] = hashlib.sha256(a.get_raw()).hexdigest()
result['permissions'] = a.get_permissions()
result['activities'] = a.get_activities()
result['providers'] = a.get_providers()
result['services'] = a.get_services()
result['strings'] = d.get_strings()
result['class_names'] = [c.get_name() for c in d.get_classes()]
result['method_names'] = [m.get_name() for m in d.get_methods()]
result['field_names'] = [f.get_name() for f in d.get_fields()]
#result['is_native_code'] = 1 if analysis.is_native_code(dx) else 0
result['is_obfuscation'] = 1 if analysis.is_ascii_obfuscation(d) else 0
#result['is_crypto_code'] = 1 if analysis.is_crypto_code(dx) else 0
'''result['is_dyn_code'] = 1 if analysis.is_dyn_code(dx) else 0
result['is_reflection_code'] = 1 if analysis.is_reflection_code(vmx) else 0'''
result['is_database'] = 1 if d.get_regex_strings(DB_REGEX) else 0
arr = []
s = a.get_elements("action", "name")
for i in s:
arr.append(i)
result['intents'] = arr
s_list = []
s_list.extend(result['class_names'])
s_list.extend(result['method_names'])
s_list.extend(result['field_names'])
result['entropy_rate'] = entropy_rate(s_list)
result['feature_vectors'] = {}
# Search for the presence of api calls in a given apk
result['feature_vectors']['api_calls'] = []
for call in API_CALLS:
status = 1 if dx.get_method(call) else 0
result['feature_vectors']['api_calls'].append(status)
# Search for the presence of permissions in a given apk
result['feature_vectors']['permissions'] = []
for permission in PERMISSIONS:
status = 1 if permission in result['permissions'] else 0
result['feature_vectors']['permissions'].append(status)
#Search for the presence of intents in a given apk
result['feature_vectors']['intents'] = []
n = len(INTENTS)
m = len(result['intents'])
for i in range(n):
stri = INTENTS[i]
flg = False
for j in range(m):
if stri in result['intents'][j]:
flg = True
break
if flg:
status = 1
else:
status = 0
result['feature_vectors']['intents'].append(status)
#Check for special strings in code
result['feature_vectors']['special_strings'] = []
for word in SPECIAL_STRINGS:
status = 1 if d.get_regex_strings(word) else 0
result['feature_vectors']['special_strings'].append(status)
return result
def extract_features(file_path):
result = {}
try:
a = APK(file_path)
d = DalvikVMFormat(a.get_dex())
dx = Analysis(d)
vm = dvm.DalvikVMFormat(a.get_dex())
vmx = analysis.Analysis(vm)
d.set_vmanalysis(dx)
d.set_decompiler(DecompilerDAD(d, dx))
except Exception as e:
print e
return None
result['android_version_code'] = a.get_androidversion_code()
result['android_version_name'] = a.get_androidversion_name()
result['max_sdk'] = a.get_max_sdk_version()
result['min_sdk'] = a.get_min_sdk_version()
result['libraries'] = a.get_libraries()
result['filename'] = a.get_filename()
result['target_sdk'] = a.get_target_sdk_version()
result['md5'] = hashlib.md5(a.get_raw()).hexdigest()
result['sha256'] = hashlib.sha256(a.get_raw()).hexdigest()
result['permissions'] = a.get_permissions()
result['activities'] = a.get_activities()
result['providers'] = a.get_providers()
result['services'] = a.get_services()
result['strings'] = d.get_strings()
result['class_names'] = [c.get_name() for c in d.get_classes()]
result['method_names'] = [m.get_name() for m in d.get_methods()]
result['field_names'] = [f.get_name() for f in d.get_fields()]
# result['is_native_code'] = 1 if analysis.is_native_code(dx) else 0
result['is_obfuscation'] = 1 if analysis.is_ascii_obfuscation(d) else 0
# result['is_crypto_code'] = 1 if analysis.is_crypto_code(dx) else 0
# result['is_dyn_code'] = 1 if analysis.is_dyn_code(dx) else 0
# result['is_reflection_code'] = 1 if analysis.is_reflection_code(vmx) else 0
result['is_database'] = 1 if d.get_regex_strings(DB_REGEX) else 0
s_list = []
s_list.extend(result['class_names'])
s_list.extend(result['method_names'])
s_list.extend(result['field_names'])
result['entropy_rate'] = entropy_rate(s_list)
result['feature_vectors'] = {}
result['feature_vectors']['api_calls'] = []
for call in API_CALLS:
status = 1 if dx.get_method_by_name(".", call, ".") else 0
result['feature_vectors']['api_calls'].append(status)
result['feature_vectors']['permissions'] = []
for permission in PERMISSIONS:
status = 1 if permission in result['permissions'] else 0
result['feature_vectors']['permissions'].append(status)
result['feature_vectors']['special_strings'] = []
for word in SPECIAL_STRINGS:
status = 1 if d.get_regex_strings(word) else 0
result['feature_vectors']['special_strings'].append(status)
result['feature_vectors']['others'] = [
# result['is_reflection_code'],
# result['is_crypto_code'],
# result['is_native_code'],
result['is_obfuscation'],
result['is_database'],
# result['is_dyn_code']
]
return result
def run(self):
"""Run androguard to extract static android information
@return: list of static features
"""
self.key = "apkinfo"
apkinfo = {}
if "file" not in self.task["category"] or not HAVE_ANDROGUARD:
return
f = File(self.task["target"])
#if f.get_name().endswith((".zip", ".apk")) or "zip" in f.get_type():
if not os.path.exists(self.file_path):
raise CuckooProcessingError("Sample file doesn't exist: \"%s\"" %
self.file_path)
try:
a = APK(self.file_path)
if a.is_valid_APK():
manifest = {}
apkinfo["files"] = self._apk_files(a)
manifest["package"] = a.get_package()
apkinfo["hidden_payload"] = []
for file in apkinfo["files"]:
if self.file_type_check(file):
apkinfo["hidden_payload"].append(file)
apkinfo["files_flaged"] = self.files_name_map
manifest["permissions"] = get_permissions(a)
manifest["main_activity"] = a.get_main_activity()
manifest["activities"] = a.get_activities()
manifest["services"] = a.get_services()
manifest["receivers"] = a.get_receivers()
manifest["receivers_actions"] = get_extended_receivers(a)
manifest["providers"] = a.get_providers()
manifest["libraries"] = list(a.get_libraries())
apkinfo["manifest"] = manifest
apkinfo["icon"] = get_apk_icon(self.file_path)
certificate = get_certificate(self.file_path)
if certificate:
apkinfo["certificate"] = certificate
#vm = DalvikVMFormat(a.get_dex())
#strings = vm.get_strings()
strings = self._get_strings(self.file_path)
apkinfo["interesting_strings"] = find_strings(strings)
apkinfo["dex_strings"] = strings
static_calls = {}
if self.options.decompilation:
if self.check_size(apkinfo["files"]):
vm = DalvikVMFormat(a.get_dex())
vmx = Analysis(vm)
vmx.create_xref()
static_calls["all_methods"] = get_methods(vmx)
static_calls[
"permissions_method_calls"] = get_show_Permissions(
vmx)
static_calls[
"native_method_calls"] = get_show_NativeMethods(
vmx)
static_calls["is_native_code"] = bool(
static_calls["native_method_calls"]
) # True if not empty, False if empty
static_calls[
"dynamic_method_calls"] = get_show_DynCode(vmx)
static_calls["is_dynamic_code"] = bool(
static_calls["dynamic_method_calls"])
static_calls[
"reflection_method_calls"] = get_show_ReflectionCode(
vmx)
static_calls["is_reflection_code"] = bool(
static_calls["reflection_method_calls"])
static_calls[
"crypto_method_calls"] = get_show_CryptoCode(vmx)
static_calls["is_crypto_code"] = bool(
static_calls["crypto_method_calls"])
classes = list()
for cls in vm.get_classes():
classes.append(cls.name)
static_calls["classes"] = classes
else:
log.warning(
"Aborted decompilation, static extraction of calls not perforemd",
)
apkinfo["static_method_calls"] = static_calls
except (IOError, OSError, BadZipfile) as e:
raise CuckooProcessingError("Error opening file %s" % e)
return apkinfo
def run(self):
"""Run androguard to extract static android information
@return: list of static features
"""
self.key = "apkinfo"
apkinfo = {}
if "file" not in self.task["category"] or not HAVE_ANDROGUARD:
return
#f = File(self.task["target"])
#if f.get_name().endswith((".zip", ".apk")) or "zip" in f.get_type():
if not os.path.exists(self.file_path):
raise CuckooProcessingError("Sample file doesn't exist: \"%s\"" % self.file_path)
apkinfo["APKiD"] = self._scan_APKiD(self.file_path)
try:
a = APK(self.file_path)
if a.is_valid_APK():
manifest = {}
apkinfo["files"] = self._apk_files(a)
apkinfo["encrypted_assets"] = self.find_encrypted_assets(a)
manifest["package"] = a.get_package()
apkinfo["hidden_payload"] = []
for file in apkinfo["files"]:
if self.file_type_check(file):
apkinfo["hidden_payload"].append(file)
apkinfo["files_flaged"] = self.files_name_map
manifest["permissions"]= get_permissions(a)
manifest["main_activity"] = a.get_main_activity()
manifest["activities"] = a.get_activities()
manifest["services"] = a.get_services()
manifest["receivers"] = a.get_receivers()
manifest["receivers_actions"] = get_extended_receivers(a)
manifest["receivers_info"] = get_receivers_info(a)
manifest["providers"] = a.get_providers()
manifest["libraries"] = a.get_libraries()
apkinfo["manifest"] = manifest
apkinfo["icon"] = get_apk_icon(self.file_path)
certificate = get_certificate(self.file_path)
if certificate:
apkinfo["certificate"] = certificate
#vm = DalvikVMFormat(a.get_dex())
#strings = vm.get_strings()
strings = self._get_strings(self.file_path)
for subdir, dirs, files in os.walk(self.dropped_path):
for file in files:
path = os.path.join(subdir, file)
try:
extra_strings = self._get_strings(path)
strings = list(set(extra_strings + strings))
except:
pass
apkinfo["dex_strings"] = strings
static_calls = {}
if self.options.decompilation:
if self.check_size(apkinfo["files"]):
vm = DalvikVMFormat(a.get_dex())
vmx = uVMAnalysis(vm)
# Be less verbose about androguard logging messages.
logging.getLogger("andro.runtime").setLevel(logging.CRITICAL)
static_calls["all_methods"] = get_methods(vmx)
static_calls["is_native_code"] = analysis.is_native_code(vmx)
static_calls["is_dynamic_code"] = analysis.is_dyn_code(vmx)
static_calls["is_reflection_code"] = analysis.is_reflection_code(vmx)
static_calls["is_crypto_code"] = is_crypto_code(vmx)
static_calls["dynamic_method_calls"] = get_show_DynCode(vmx)
static_calls["reflection_method_calls"] = get_show_ReflectionCode(vmx)
static_calls["permissions_method_calls"] = get_show_Permissions(vmx)
static_calls["crypto_method_calls"] = get_show_CryptoCode(vmx)
static_calls["native_method_calls"] = get_show_NativeMethods(vmx)
classes = list()
for cls in vm.get_classes():
classes.append(cls.name)
static_calls["classes"] = classes
else:
log.warning("Dex size bigger than: %s",
self.options.decompilation_threshold)
apkinfo["static_method_calls"] = static_calls
except (IOError, OSError, BadZipfile) as e:
raise CuckooProcessingError("Error opening file %s" % e)
return apkinfo
def run(self):
"""Run androguard to extract static android information
@return: list of static features
"""
self.key = "apkinfo"
apkinfo = {}
if "file" not in self.task["category"] or not HAVE_ANDROGUARD:
return
f = File(self.task["target"])
#if f.get_name().endswith((".zip", ".apk")) or "zip" in f.get_type():
if not os.path.exists(self.file_path):
raise CuckooProcessingError("Sample file doesn't exist: \"%s\"" % self.file_path)
try:
a = APK(self.file_path)
if a.is_valid_APK():
manifest = {}
apkinfo["files"] = self._apk_files(a)
manifest["package"] = a.get_package()
apkinfo["hidden_payload"] = []
for file in apkinfo["files"]:
if self.file_type_check(file):
apkinfo["hidden_payload"].append(file)
apkinfo["files_flaged"] = self.files_name_map
manifest["permissions"]= get_permissions(a)
manifest["main_activity"] = a.get_main_activity()
manifest["activities"] = a.get_activities()
manifest["services"] = a.get_services()
manifest["receivers"] = a.get_receivers()
manifest["receivers_actions"] = get_extended_receivers(a)
manifest["providers"] = a.get_providers()
manifest["libraries"] = a.get_libraries()
apkinfo["manifest"] = manifest
apkinfo["icon"] = get_apk_icon(self.file_path)
certificate = get_certificate(self.file_path)
if certificate:
apkinfo["certificate"] = certificate
#vm = DalvikVMFormat(a.get_dex())
#strings = vm.get_strings()
strings = self._get_strings(self.file_path)
apkinfo["interesting_strings"] = find_strings(strings)
apkinfo["dex_strings"] = strings
static_calls = {}
if self.options.decompilation:
if self.check_size(apkinfo["files"]):
vm = DalvikVMFormat(a.get_dex())
vmx = uVMAnalysis(vm)
static_calls["all_methods"] = get_methods(vmx)
static_calls["is_native_code"] = analysis.is_native_code(vmx)
static_calls["is_dynamic_code"] = analysis.is_dyn_code(vmx)
static_calls["is_reflection_code"] = analysis.is_reflection_code(vmx)
static_calls["is_crypto_code"] = is_crypto_code(vmx)
static_calls["dynamic_method_calls"] = get_show_DynCode(vmx)
static_calls["reflection_method_calls"] = get_show_ReflectionCode(vmx)
static_calls["permissions_method_calls"] = get_show_Permissions(vmx)
static_calls["crypto_method_calls"] = get_show_CryptoCode(vmx)
static_calls["native_method_calls"] = get_show_NativeMethods(vmx)
classes = list()
for cls in vm.get_classes():
classes.append(cls.name)
static_calls["classes"] = classes
else:
log.warning("Dex size bigger than: %s",
self.options.decompilation_threshold)
apkinfo["static_method_calls"] = static_calls
except (IOError, OSError, BadZipfile) as e:
raise CuckooProcessingError("Error opening file %s" % e)
return apkinfo
def analyze(path):
try:
start = process_time()
hashfunctions = dict(md5=hashlib.md5,
sha1=hashlib.sha1,
sha256=hashlib.sha256,
sha512=hashlib.sha512)
a = APK(path)
certs = set(
a.get_certificates_der_v3() + a.get_certificates_der_v2() +
[a.get_certificate_der(x) for x in a.get_signature_names()])
for cert in certs:
x509_cert = x509.Certificate.load(cert)
issuer = {
'commonName': None,
'organizationName': None,
'organizationalUnitName': None,
'countryName': None,
'stateOrProvinceName': None,
'localityName': None
}
subject = {
'commonName': None,
'organizationName': None,
'organizationalUnitName': None,
'countryName': None,
'stateOrProvinceName': None,
'localityName': None
}
strIssuer = get_certificate_name_string(x509_cert.issuer,
short=False)
strSubject = get_certificate_name_string(x509_cert.subject,
short=False)
arrIssuer = strIssuer.split(',')
for i in arrIssuer:
if i.lstrip().split('=')[0] == 'commonName':
issuer['commonName'] = i.lstrip().split('=')[1]
elif i.lstrip().split('=')[0] == 'organizationName':
issuer['organizationName'] = i.lstrip().split('=')[1]
elif i.lstrip().split('=')[0] == 'organizationalUnitName':
issuer['organizationalUnitName'] = i.lstrip().split('=')[1]
elif i.lstrip().split('=')[0] == 'countryName':
issuer['countryName'] = i.lstrip().split('=')[1]
elif i.lstrip().split('=')[0] == 'stateOrProvinceName':
issuer['stateOrProvinceName'] = i.lstrip().split('=')[1]
elif i.lstrip().split('=')[0] == 'localityName':
issuer['localityName'] = i.lstrip().split('=')[1]
arrSubject = strSubject.split(',')
for i in arrSubject:
if i.lstrip().split('=')[0] == 'commonName':
subject['commonName'] = i.lstrip().split('=')[1]
elif i.lstrip().split('=')[0] == 'organizationName':
subject['organizationName'] = i.lstrip().split('=')[1]
elif i.lstrip().split('=')[0] == 'organizationalUnitName':
subject['organizationalUnitName'] = i.lstrip().split(
'=')[1]
elif i.lstrip().split('=')[0] == 'countryName':
subject['countryName'] = i.lstrip().split('=')[1]
elif i.lstrip().split('=')[0] == 'stateOrProvinceName':
subject['stateOrProvinceName'] = i.lstrip().split('=')[1]
elif i.lstrip().split('=')[0] == 'localityName':
subject['localityName'] = i.lstrip().split('=')[1]
for k, v in hashfunctions.items():
if k == 'md5':
md5 = v(cert).hexdigest()
elif k == 'sha1':
sha1 = v(cert).hexdigest()
elif k == 'sha256':
sha256 = v(cert).hexdigest()
elif k == 'sha512':
sha512 = v(cert).hexdigest()
md5 = md5
appName = a.get_app_name()
fileSize = os.stat(a.get_filename()).st_size
sha1 = sha1
sha256 = sha256
sha512 = sha512
timestamp = time.time()
dateTime = datetime.fromtimestamp(timestamp)
timeOfSubmit = dateTime.strftime("%Y-%m-%d %H:%M:%S")
package = a.get_package()
androidversionCode = a.get_androidversion_code()
androidversionName = a.get_androidversion_name()
minSDKVersion = a.get_min_sdk_version()
maxSDKVersion = a.get_max_sdk_version()
targetSDKVersion = a.get_target_sdk_version()
mainActivity = a.get_main_activity()
attributes = {
'validFrom':
x509_cert['tbs_certificate']['validity']
['not_before'].native.strftime("%Y-%m-%d %H:%M:%S"),
'validTo':
x509_cert['tbs_certificate']['validity']
['not_after'].native.strftime("%Y-%m-%d %H:%M:%S"),
'serialNumber':
hex(x509_cert.serial_number),
'hashAlgorithm':
x509_cert.hash_algo,
'signatureAlgorithm':
x509_cert.signature_algo
}
certificateAttributes = json.dumps(attributes)
certificateIssuer = json.dumps(issuer)
certificateSubject = json.dumps(subject)
declaredPermissions = json.dumps(a.get_declared_permissions())
requestedPermissions = json.dumps(a.get_permissions())
activities = json.dumps(a.get_activities())
services = json.dumps(a.get_services())
receivers = json.dumps(a.get_receivers())
providers = json.dumps(a.get_providers())
stop = process_time()
analysisTime = stop - start
connect = mysql.connect()
cursor = connect.cursor()
sql = "INSERT INTO tbl_apkinfo (md5, appName, fileSize, analysisTime, sha1, sha256, sha512, firstSubmission, lastSubmission, package, androidversionCode, androidversionName, minSDKVersion, maxSDKVersion, targetSDKVersion, mainActivity, certificateAttributes, certificateIssuer, certificateSubject, declaredPermissions, requestedPermissions, activities, services, providers, receivers) VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s)"
param = (md5, appName, fileSize, analysisTime, sha1, sha256, sha512,
timeOfSubmit, timeOfSubmit, package, androidversionCode,
androidversionName, minSDKVersion, maxSDKVersion,
targetSDKVersion, mainActivity, certificateAttributes,
certificateIssuer, certificateSubject, declaredPermissions,
requestedPermissions, activities, services, providers,
receivers)
cursor.execute(sql, param)
connect.commit()
connect.close()
androaxml_main(path,
os.path.join(app.config['OUTPUT_PATH'], md5 + '.xml'))
return True
except:
return False
def extract_features(file_path):
result = {}
try:
a = APK(file_path)
d = DalvikVMFormat(a.get_dex())
dx = VMAnalysis(d)
vm = dvm.DalvikVMFormat(a.get_dex())
vmx = analysis.uVMAnalysis(vm)
d.set_vmanalysis(dx)
d.set_decompiler(DecompilerDAD(d, dx))
except:
return None
result['android_version_code'] = a.get_androidversion_code()
result['android_version_name'] = a.get_androidversion_name()
result['max_sdk'] = a.get_max_sdk_version()
result['min_sdk'] = a.get_min_sdk_version()
result['libraries'] = a.get_libraries()
result['filename'] = a.get_filename()
result['target_sdk'] = a.get_target_sdk_version()
result['md5'] = hashlib.md5(a.get_raw()).hexdigest()
result['sha256'] = hashlib.sha256(a.get_raw()).hexdigest()
result['permissions'] = a.get_permissions()
result['activities'] = a.get_activities()
result['providers'] = a.get_providers()
result['services'] = a.get_services()
#result['strings'] = d.get_strings()
#result['class_names'] = [c.get_name() for c in d.get_classes()]
#result['method_names'] = [m.get_name() for m in d.get_methods()]
#result['field_names'] = [f.get_name() for f in d.get_fields()]
class_names = [c.get_name() for c in d.get_classes()]
method_names = [m.get_name() for m in d.get_methods()]
field_names = [ f.get_name() for f in d.get_fields()]
result['is_native_code'] = 1 if analysis.is_native_code(dx) else 0
result['is_obfuscation'] = 1 if analysis.is_ascii_obfuscation(d) else 0
result['is_crypto_code'] = 1 if analysis.is_crypto_code(dx) else 0
result['is_dyn_code'] = 1 if analysis.is_dyn_code(dx) else 0
result['is_reflection_code'] = 1 if analysis.is_reflection_code(vmx) else 0
result['is_database'] = 1 if d.get_regex_strings(DB_REGEX) else 0
s_list = []
#s_list.extend(result['class_names'])
#s_list.extend(result['method_names'])
#s_list.extend(result['field_names'])
s_list.extend(class_names)
s_list.extend(method_names)
s_list.extend(method_names)
result['entropy_rate'] = entropy_rate(s_list)
result['feature_vectors'] = {}
# Search for the presence of api calls in a given apk
result['feature_vectors']['api_calls'] = []
for call in API_CALLS:
status = 1 if dx.tainted_packages.search_methods(".", call, ".") else 0
result['feature_vectors']['api_calls'].append(status)
# Search for the presence of permissions in a given apk
result['feature_vectors']['permissions'] = []
for permission in PERMISSIONS:
status = 1 if permission in result['permissions'] else 0
result['feature_vectors']['permissions'].append(status)
result['feature_vectors']['special_strings'] = []
for word in SPECIAL_STRINGS:
status = 1 if d.get_regex_strings(word) else 0
result['feature_vectors']['special_strings'].append(status)
opt_seq = []
for m in d.get_methods():
for i in m.get_instructions():
opt_seq.append(i.get_name())
optngramlist = [tuple(opt_seq[i:i+NGRAM]) for i in xrange(len(opt_seq) - NGRAM)]
optngram = Counter(optngramlist)
optcodes = dict()
tmpCodes = dict(optngram)
#for k,v in optngram.iteritems():
# if v>=NGRAM_THRE:
#optcodes[str(k)] = v
# optcodes[str(k)] = 1
tmpCodes = sorted(tmpCodes.items(),key =lambda d:d[1],reverse=True)
for value in tmpCodes[:NGRAM_THRE]:
optcodes[str(value[0])] = 1
result['feature_vectors']['opt_codes'] = optcodes
return result