def call_and_add(t, fname):
global DATA_TYPE
try:
print('[*] Feature Extracting: {}'.format(fname))
result = extract_features(fname)
print('[*] Adding to db: {}'.format(fname))
client = MongoClient()
db = client[DB_NAME]
result['data_type'] = t
print(result['data_type'])
db['apk'].update({'sha256': result['sha256']}, result, upsert=True)
except Exception as e:
print('[!] Error occured with {}, {}'.format(fname, e))
def analyze():
if request.method == 'POST':
f = request.files.get('file', None)
if f:
fp = os.path.join(app.config['UPLOAD_FOLDER'], f.filename)
f.save(fp)
features = extract_features(fp)
feature_vector = create_vector_single(features)
arr = []
arr.append(feature_vector)
mod1 = pickle.load(open("feature_model.p", "rb"))
mod2 = pickle.load(open("kfold_train_data.p", "rb"))
mod3 = pickle.load(open("decision.p", "rb"))
mod4 = pickle.load(open("random_forest.p", "rb"))
mod5 = pickle.load(open("xgb.p", "rb"))
mod6 = pickle.load(open("pca_decision.p", "rb"))
mod7 = pickle.load(open("pca_ran.p", "rb"))
feature_vector_new = mod1.transform(arr)
feature_vector_dec = mod6.transform(feature_vector_new)
feature_vector_ran = mod7.transform(feature_vector_new)
label = {}
result = mod2.predict(feature_vector_new)
if int(result[0]) == 1:
label["svm"] = "Malware"
else:
label["svm"] = "Goodware"
result = mod5.predict(feature_vector_new)
if int(result[0]) == 1:
label["xgb"] = "Malware"
else:
label["xgb"] = "Goodware"
result = mod4.predict(feature_vector_ran)
if int(result[0]) == 1:
label["random_forest"] = "Malware"
else:
label["random_forest"] = "Goodware"
result = mod3.predict(feature_vector_dec)
if int(result[0]) == 1:
label["Decision_tree"] = "Malware"
else:
label["Decision_tree"] = "Goodware"
else:
label = None
return render_template("index.html", label=label)
def analyze():
if request.method == 'POST':
f = request.files.get('file', None)
if f:
fp = os.path.join(app.config['UPLOAD_FOLDER'], f.filename)
f.save(fp)
features = extract_features(fp)
feature_vector = create_vector_single(features)
arr = []
arr.append(feature_vector)
model = pickle.load(open("xgb.p", "rb"))
mod = pickle.load(open("feature_model.p", "rb"))
arr = mod.transform(arr)
result = model.predict(arr)
label = {}
if int(result[0]) == 1:
label["svm"] = "Malware"
else:
label["svm"] = "Goodware"
else:
label = None
return render_template("index.html", label=label)
import pickle
from feature_Extraction import extract_features, create_vector_single
file_path = "C:/Users/MY PC/PycharmProjects/antivirus/uploads/com.phonepe.app.apk"
features = extract_features(file_path)
feature_vector = create_vector_single(features)
print(len(feature_vector))
arr = []
arr.append(feature_vector)
#model=pickle.load(open("xgb.p", "rb"))
model = pickle.load(open("kfold_train_data.p", "rb"))
mod = pickle.load(open("feature_model.p", "rb"))
arr = mod.transform(arr)
result = model.predict(arr)
if int(result[0]) == 1:
print("Malware")
else:
print("Goodware")