My inner while loop containing msg = client_socket.recv(1024).decode() is unable to handle multiple client requests without having client_socket, addr = server_socket.accept() inside the loop. I'm not sure how can I keep reusing the same connection for all the requests.
from socket import *
import sys
def run_server(port):
# Create server socket
server_socket = socket(AF_INET, SOCK_STREAM)
# Bind the socket to a specific address
server_socket.bind(('localhost', port))
# Argument 1 represents the maximum number of queued connections
server_socket.listen(1)
print(f"Serving on port {port}")
while 1:
# Accept incoming connection from client
client_socket, addr = server_socket.accept()
print(f"Connection established from address {addr}")
while 1:
# Receive data from client
msg = client_socket.recv(1024).decode()
# Extract file name
file_name = msg.split('\n')[0].split()[1].replace('/', '')
# Process client requests
try:
with open(file_name, 'rb') as file:
status_line = "HTTP/1.1 200 OK\r\n"
file_extension = file_name.split('.')[-1].lower()
if file_extension == 'html':
content_type_header = "Content-Type: text/html\r\n\r\n"
elif file_extension == 'png':
content_type_header = "Content-Type: image/png\r\n\r\n"
content = file.read()
# Combine status line, headers and content
http_response = (status_line + "Connection: keep-alive\r\n" + content_type_header).encode() + content
except OSError as _:
# File not found, generate a 404 response
status_line = "HTTP/1.1 404 Not Found\r\n"
content_type_header = "Content-Type: text/html\r\n\r\n"
content = "<h1>404 Not Found</h1><p>The requested file was not found on this server.</p>"
http_response = (status_line + content_type_header + content).encode()
# Send HTTP response back to client
client_socket.sendall(http_response)
# Close the connection after handling multiple requests
client_socket.close()
print("Connection closed")
server_socket.close()
if __name__ == "__main__":
if len(sys.argv) != 2:
print("Usage: python WebServer.py <port>")
sys.exit(1)
# Get port number from command line parameter
port = int(sys.argv[1])
run_server(port)
I assume you are doing this to learn since implementing the HTTP protocol is complex.
TCP is a byte streaming protocol and you wil need to follow a higher-level protocol such as HTTP 1.0 [RFC 1945] to determine how to read the stream and how much to read. Since HTTP requests contain \r\n-terminated lines for the request and headers and a blank line to indicate end-of-headers, wrapping the TCP socket in a file-like object via socket.makefile makes it easier to process the request.
To implement keep-alive, the headers need a Content-Length header to know how much content to read.
Below I've made minimal changes to support your implementation. It is by no means complete. Make a root directory in the same directory as the script below and populated it with a few HTML and PNG files like index.html or icon.png. Point a browser to http://localhost:8080/index.html or http://localhost:8080/icon.png and the files should display in the browser. At least they did for me on Chrome and Edge 😊.
import socket
import sys
import os
from pprint import pprint
def run_server(port):
with socket.socket() as server_socket:
server_socket.bind(('localhost', port))
server_socket.listen()
print(f'Serving on port {port}')
while True:
client_socket, addr = server_socket.accept()
# wrap client socket in a file-like object. Gives access to .readline().
with client_socket, client_socket.makefile('rb') as infile:
print(f'{addr}: connected')
while True:
# Receive HTTP request
line = infile.readline()
if not line: break
request = line.decode('ascii')
print(request)
# Read headers line-by-line and add to dictionary
headers = {}
while line := infile.readline():
if line == b'\r\n': break
key, _, value = line.decode('ascii').partition(':')
headers[key.strip()] = value.strip()
pprint(headers)
op, path, protocol = request.split()
print(op, path, protocol)
# hacky way to look under a directory called root below the webserver script.
file_name = os.path.abspath(os.path.join('root/.' + path))
print(file_name)
# minimally handles a GET for html and png files.
try:
with open(file_name, 'rb') as file:
http_response = b'HTTP/1.1 200 OK\r\n'
file_extension = file_name.split('.')[-1].lower()
if file_extension == 'html':
http_response += b'Content-Type: text/html\r\n'
elif file_extension == 'png':
http_response += b'Content-Type: image/png\r\n'
content = file.read()
except FileNotFoundError as e:
# File not found, generate a 404 response
http_response = b'HTTP/1.1 404 Not Found\r\n'
http_response += b'Content-Type: text/html\r\n'
content = b'<h1>404 Not Found</h1><p>The requested file was not found on this server.</p>'
# Send keepalive and the length of the content.
http_response += b'Connection: keep-alive\r\n'
http_response += f'Content-Length: {len(content)}\r\n'.encode('ascii')
# Send blank line (end of headers) and content.
http_response += b'\r\n' + content
client_socket.sendall(http_response)
print(f'{addr}: disconnected')
if __name__ == '__main__':
if len(sys.argv) != 2:
print('Usage: python WebServer.py <port>')
sys.exit(1)
# Get port number from command line parameter
port = int(sys.argv[1])
run_server(port)