working_model.py

# -*- coding: utf-8 -*-

##
# \file
# \brief  Model implementing code
# \details This code reads mails individually and then classifies them using pre-trained model.
# \author Sudhanshu Dubey
# \version    1.0
# \date   25/6/19
# \param    mail_file The full address of mail file.
# \warning  Is not suitable for classifying multiple emails. Use fast_multiple.py instead.
# \bug  No known bug.

import sys

from supporters.features import mail_features
import pickle

mail_file = sys.argv[1]
features_matrix = mail_features(mail_file)

ml_model = pickle.load(open('spamfilter.sav', 'rb'))
result = ml_model.predict(features_matrix)
if result == 1:
    print(mail_file + " is a spam!!!")
elif result == 0:
    print(mail_file + " is normal mail.")
else:
    print("something went wrong!!")