Simple Python UDP echo server

reply-to-empty-udp.py

#!/usr/bin/env python2
# -*- coding: utf-8 -*-

# Author: David Manouchehri <[email protected]>
# This script will always echo back data on the UDP port of your choice.
# Useful if you want nmap to report a UDP port as "open" instead of "open|filtered" on a standard scan.
# Works with both Python 2 & 3.

import socket

sock = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)

server_address = '0.0.0.0'
server_port = 31337

server = (server_address, server_port)
sock.bind(server)
print("Listening on " + server_address + ":" + str(server_port))

while True:
	payload, client_address = sock.recvfrom(1)
	print("Echoing data back to " + str(client_address))
	sent = sock.sendto(payload, client_address)

@Manouchehri that doesn't seem to be the case. According to the documentation, which you can see with help(socket.socket.recv) in the Python interpreter, it is stated that the first argument is the buffersize and calling the recv method will "receive up to buffersize bytes from the socket." This would mean that, even if a Python socket only received 1 byte in a UDP packet whilte it was receiving data, it would return that 1 byte. I believe the point of the buffersize is to avoid receiving more data than the program can handle (a potential buffer overflow).

I just tested it and it works fine with any size of message. If you change the buffersize to 1024 and then run this client program, you'll see it print out Test passed.

#!/usr/bin/env python3
import socket

client_socket = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
server_address = '0.0.0.0'
server_port = 31337
client_socket.connect((server_address, server_port))

message = 'Hello World'
client_socket.send(message.encode())
response = client_socket.recv(1024).decode()
print("Test passed" if message == response else "Test failed")