Sockets are a cornerstone of network programming, allowing different processes to communicate over the network. In Python, the socket module provides an interface to the Berkeley sockets API, which is a standard API for network communication.
To get started, you need to understand the two main types of sockets: stream sockets and datagram sockets. Stream sockets (TCP) provide a reliable, connection-oriented service, while datagram sockets (UDP) are connectionless and can lose packets. Here’s a basic example of creating a TCP socket:
import socket
# Create a TCP/IP socket
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
# Bind the socket to the address and port
s.bind(('localhost', 12345))
# Listen for incoming connections
s.listen(1)
print("Server is listening on port 12345...")
In this snippet, we create a TCP socket using socket.AF_INET for IPv4 addressing and socket.SOCK_STREAM for the TCP protocol. The bind method associates the socket with a specific address and port, and the listen method starts listening for incoming connections.
Next, you’ll want to accept incoming connections. The accept method will block and wait for an incoming connection, returning a new socket object for the connection along with the address of the client:
conn, addr = s.accept()
print(f"Connection from {addr} has been established.")
Now you have a connection object, conn, that you can use to send and receive data. It’s essential to handle data transmission properly to ensure that both ends can communicate effectively. Here’s how to receive data from the connected client:
data = conn.recv(1024)
print(f"Received data: {data.decode()}")
The recv method retrieves data sent by the client, and the size parameter specifies the maximum amount of data to be received at once. After processing, you usually want to send data back to the client:
conn.sendall(b'Hello, client!')
After you’re done with the connection, it’s a good practice to close the connection to free up resources:
conn.close()
This gives you the basics of setting up a TCP socket in Python. Datagram sockets follow a similar pattern, but with some differences in how data is sent and received. Understanding these foundations will allow you to build more complex network applications.
