vectorizer = joblib.load(path + '\\model\\tf_idf.sav')
#setting up log files
dt = datetime.datetime.now()
log_file = path + '\\logs\\' + str(dt.year) + str(dt.month) + str(
dt.day) + '-' + str(randint(0, 10000)) + '.txt'
logger = logging.getLogger(log_file)
logger.setLevel(logging.INFO)
logging.info('Starting...')
#fetching email contents
#storing email subject along with body
#this is to classify those emails which dont have text but
#just an image embedded in the body
for uid in data:
dict = mail.parse_email(uid)
content = [dict['Subject'] + '\n' + dict['Body']]
#extracting features
#getting predicition from classifier
features = vectorizer.transform(content)
pred = clf.predict(features)
#moving to spam folder (if predicted as spam)
if pred[0] == 1:
mov = mail.move_email(data[1], 'Inbox', '[Gmail]/Spam')
if mov == 'OK':
message = 'Following email moved to spam:\nuid: ' + str(
uid
) + '\nSubject: ' + str(dict['Subject']) + '\nContent: ' + str(
dict['Body']) + '\n----------------------------------'
