Python Cookbook

Converting IP Addresses

Credit: Alex Martelli, Greg Jorgensen

Problem

You need to convert IP addresses from dotted quads to long integers and back, and extract network and host portions of such addresses.

Solution

The socket and struct modules let you easily convert long integers to dotted quads and back:

import socket, struct

def dottedQuadToNum(ip):
    "convert decimal dotted quad string to long integer"
    return struct.unpack('>L',socket.inet_aton(ip))[0]

def numToDottedQuad(n):
    "convert long int to dotted quad string"
    return socket.inet_ntoa(struct.pack('>L',n))

To split an IP address into network and host portions, we just need to apply a suitable binary mask to the long integer form of the IP address:

def makeMask(n):
    "return a mask of n bits as a long integer"
    return (2L<<n-1)-1

def ipToNetAndHost(ip, maskbits):
    "return tuple (network, host) dotted-quad addresses given IP and mask size"
    # by Greg Jorgensen

    n = dottedQuadToNum(ip)
    m = makeMask(maskbits)

    host = n & m
    net = n - host

    return numToDottedQuad(net), numToDottedQuad(host)

Discussion

The format we use for the struct.pack and struct.unpack calls must start with a '>', which specifies big-endian byte order. This is the network byte order used by the socket.inet_aton and socket.inet_ntoa functions. If you omit the '>', struct instead uses the native byte order of the machine the code is running on, while the socket module still uses big-endian byte order.

The network part of an IP address used to be expressed as an explicit bit ...