def get_matrix(mail):
file = get_full_path('data/psocols.txt')
with open(file, 'rb') as f:
lexicon = pickle.load(f)
lex = np.zeros(len(lexicon))
for word in mail:
if word.lower() in lexicon:
index_value = lexicon.index(word.lower())
lex[index_value] += 1
features = np.array(lex, dtype=np.float32)
return features
def predict(body):
file_content = utils.Process(body)
mail = file_content.process()
matrix = np.array([encoder.get_matrix(mail)])
model = load_model(get_full_path('data/model.h5'))
model.compile(loss='categorical_crossentropy',
optimizer='adam',
metrics=['accuracy'])
result = model.predict_classes(matrix)
backend.clear_session()
return predictions.get(result[0])
def upload():
blacklist = open(get_full_path(blacklist_dir)).read()
if request.method == 'POST':
file = request.files['file']
session.clear()
if file and allowed_file(file.filename):
filename = get_file_name(file.filename)
if black_checker(filename):
return render_template(
'base.html',
blacklist=open(blacklist_dir).read(),
message='This sender has been blacklisted!!')
else:
content = file.read()
detect = predict(content)
if detect == 'fraud':
update_blacklist(filename)
return render_template(
'base.html',
content=content,
blacklist=open(blacklist_dir).read(),
message=' This is a fraudulent mail.'
'Therefore,the mail has been updated on the blacklist')
elif detect == 'ham':
render_template('base.html',
blacklist=blacklist,
content=content)
elif detect == 'spam':
return render_template(
'base.html',
blacklist=blacklist,
message=
'This mail is spam, therefore it will be sent to spam folder'
)
redirect(
url_for('upload', blacklist=blacklist, content=content))
return render_template('base.html',
blacklist=blacklist,
content=file.read())
else:
return render_template('base.html', blacklist=blacklist)
from flask import Flask, request, session
from core import predict
from flask_bootstrap import Bootstrap
from flask import url_for, redirect, render_template
from dir import get_full_path
import nltk
app = Flask(__name__)
bootstrap = Bootstrap(app)
app.config['SECRET_KEY'] = 'hard to guess string'
ALLOWED_EXTENSIONS = ['txt', 'csv']
blacklist_dir = get_full_path('data/blacklist.txt')
@app.route('/', methods=['GET', 'POST'])
def upload():
blacklist = open(get_full_path(blacklist_dir)).read()
if request.method == 'POST':
file = request.files['file']
session.clear()
if file and allowed_file(file.filename):
filename = get_file_name(file.filename)
if black_checker(filename):
return render_template(
'base.html',
blacklist=open(blacklist_dir).read(),
message='This sender has been blacklisted!!')
else:
from dir import get_full_path
from nltk.corpus import words
def check(word):
if word in words.words():
return True
else:
return False
def worded(text):
return re.findall(r'\w+', text.lower())
WORDS = Counter(worded(open(get_full_path('data/big.txt')).read()))
def P(word, N=sum(WORDS.values())):
"""Probability of `word`."""
return WORDS[word] / N
def correction(word):
"""Most probable spelling correction for word."""
return max(candidates(word), key=P)
def candidates(word):
"""Generate possible spelling corrections for word."""
return (known([word]) or known(edits1(word)) or known(edits2(word))