sockets - reading DNS packets in Python

Question

I'm playing around with Python sockets, and decided to see if I could implement a very basic name server (i.e. a lookup table for a domain name to an IP address). So I've set up my server so far to just dump the received data.

#!/usr/bin/python
import socket

s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
host = ''
port = 53
size = 512
s.bind((host, port))
while True:
    data, addr = s.recvfrom(size)
    print repr(data)

When I run the above code and point my DNS to 127.0.0.1 I get something akin to the following:

'Yx04x01x00x00x01x00x00x00x00x00x00x03wwwx06googlex03comx00x00x01x00x01'
'Jxaax01x00x00x01x00x00x00x00x00x00x03wwwx06googlex03comx00x00x1cx00x01'
'Yx04x01x00x00x01x00x00x00x00x00x00x03wwwx06googlex03comx00x00x01x00x01'

I'm assuming that it is something to do with the DNS question packet structure, but I'm not to sure.

A) Are the above escape characters? A specific text encoding? Or simply just bytes?

B) How can I interpret the data and work with it?

EDIT: Changing the socket to take raw instead of datagrams results in the following:

'Ex00$x00xe4x96x00x00@x01x00x00x7fx00x00x01x7fx00x00x01x03x03Xxb6x00x00x00x00Ex00Vx00mx82x00x00xffx11x00x00x7fx00x00x01x7fx00x00x01xf3xe1x005x00Bx00x00'

Answer 1

You could start with something like this:

#!/usr/bin/env python

import pprint
import socket
import struct


def decode_labels(message, offset):
    labels = []

    while True:
        length, = struct.unpack_from("!B", message, offset)

        if (length & 0xC0) == 0xC0:
            pointer, = struct.unpack_from("!H", message, offset)
            offset += 2

            return labels + decode_labels(message, pointer & 0x3FFF), offset

        if (length & 0xC0) != 0x00:
            raise StandardError("unknown label encoding")

        offset += 1

        if length == 0:
            return labels, offset

        labels.append(*struct.unpack_from("!%ds" % length, message, offset))
        offset += length


DNS_QUERY_SECTION_FORMAT = struct.Struct("!2H")

def decode_question_section(message, offset, qdcount):
    questions = []

    for _ in range(qdcount):
        qname, offset = decode_labels(message, offset)

        qtype, qclass = DNS_QUERY_SECTION_FORMAT.unpack_from(message, offset)
        offset += DNS_QUERY_SECTION_FORMAT.size

        question = {"domain_name": qname,
                    "query_type": qtype,
                    "query_class": qclass}

        questions.append(question)

    return questions, offset


DNS_QUERY_MESSAGE_HEADER = struct.Struct("!6H")

def decode_dns_message(message):

    id, misc, qdcount, ancount, nscount, arcount = DNS_QUERY_MESSAGE_HEADER.unpack_from(message)

    qr = (misc & 0x8000) != 0
    opcode = (misc & 0x7800) >> 11
    aa = (misc & 0x0400) != 0
    tc = (misc & 0x200) != 0
    rd = (misc & 0x100) != 0
    ra = (misc & 0x80) != 0
    z = (misc & 0x70) >> 4
    rcode = misc & 0xF

    offset = DNS_QUERY_MESSAGE_HEADER.size
    questions, offset = decode_question_section(message, offset, qdcount)

    result = {"id": id,
              "is_response": qr,
              "opcode": opcode,
              "is_authoritative": aa,
              "is_truncated": tc,
              "recursion_desired": rd,
              "recursion_available": ra,
              "reserved": z,
              "response_code": rcode,
              "question_count": qdcount,
              "answer_count": ancount,
              "authority_count": nscount,
              "additional_count": arcount,
              "questions": questions}

    return result


s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
host = ''
port = 53
size = 512
s.bind((host, port))
while True:
    data, addr = s.recvfrom(size)
    pprint.pprint(decode_dns_message(data))

And then fill in the decoding functions for the remaining records.