Hackerman's Hacking Tutorials

The knowledge of anything, since all things have causes, is not acquired or complete unless it is known by its causes. - Avicenna

Nov 15, 2020 - 3 minute read - Comments - python

Customizing Python's SimpleHTTPServer

The other day I customized the Python built-in SimpleHTTPServer with some routes. I did not find a lot of info about it (most use it to serve files). This is how I did some basic customization.

This is for Python 3.8.6 (which what I have in my testing VM) but it should work on Python 3.9 (and probably the same for Python 2).

Code is at https://github.com/parsiya/Parsia-Code/tree/master/python-simplehttpserver.

How to Serve Files

python -m http.server 1234 --bind 127.0.0.1.

Custom GET Responses

But I needed to customize the path. Let's start with a simple implementation. We need to create our own BaseHTTPRequestHandler.

from http.server import HTTPServer, BaseHTTPRequestHandler

class MyHandler(BaseHTTPRequestHandler):
    def do_GET(self):
        pass

httpd = HTTPServer(('localhost', 10000), MyHandler)
httpd.serve_forever()

To respond to GET requests we need to add code to do_GET. Let's say we want to return a 200 response that says It works!.

# 01.py
def do_GET(self):
    # send 200 response
    self.send_response(200)
    # send response headers
    self.end_headers()
    # send the body of the response
    self.wfile.write(bytes("It Works!", "utf-8"))
01.py 01.py

Custom Response Headers

Note the server adds some default headers. To modify these we can use send_header before calling end_headers. This is very useful for adding the Content-Type header.

# 02.py
def do_GET(self):
    # send 200 response
    self.send_response(200)
    # add our own custom header
    self.send_header("myheader", "myvalue")
    # send response headers
    self.end_headers()
    # send the body of the response
    self.wfile.write(bytes("It Works!", "utf-8"))
02.py 02.py

To override a header we cannot use send_header because it will just add it as a new header to the response. Based on the documentation it seems like the Date and Server response headers cannot be changed :(.

Read Request Path and Query Strings

The complete path and query strings are in the self.path object inside do_GET and similar methods. First we need to parse it with urllib.parse.urlparse. Then we can get the query string and path from the parsed object's fields query and path, respectively.

from urllib.parse import urlparse

def do_GET(self):
    # first we need to parse it
    parsed = urlparse(self.path)
    # get the query string
    query_string = parsed.query
    # get the request path, this new path does not have the query string
    path = parsed.path

Read Request Headers

I needed to read the incoming request headers. These are stored in the headers object. It is of type http.client.HTTPMessage which is a subclass of email.message.Message.

We can get the first value of a header by name with headers.get("header name") . To get all values for a specific header (because headers can be repeated) use headers.get_all("header name").

# 03.py
def do_GET(self):
    # get the value of the "Authorization" header and echo it.
    authz = self.headers.get("authorization")
    # send 200 response
    self.send_response(200)
    # send response headers
    self.end_headers()
    # send the body of the response
    self.wfile.write(bytes(authz, "utf-8"))

Note: Header names are not case-sensitive in HTTP (or in this module).

03.py 03.py

Reading The Body of POST Requests

To handle POST requests we need to implement do_POST (surprise). To read the body of the POST request we:

  1. Read the Content-Length header in the incoming request.
  2. Read that many bytes from self.rfile.
    1. I could not find a way to read "all bytes" in rfile. I had to rely on the Content-Length header.
Reading the body
 1
 2
 3
 4
 5
 6
 7
 8
 9
10
def do_POST(self):
    # read the content-length header
    content_length = int(self.headers.get("Content-Length"))
    # read that many bytes from the body of the request
    body = self.rfile.read(content_length)

    self.send_response(200)
    self.end_headers()
    # echo the body in the response
    self.wfile.write(body)
04.py 04.py

Server Over TLS

First, you need to create a private key and certificate in pem format. To create a self-signed certificate/key in one line with OpenSSL:

openssl req -newkey rsa:2048 -nodes -keyout key.pem -x509 -days 365 -out certificate.pem

Then modify the last lines of the original script to:

httpd = HTTPServer(('localhost', 443), MyHandler)
httpd.socket = ssl.wrap_socket(httpd.socket, server_side=True, certfile="certificate.pem", keyfile="key.pem")
httpd.serve_forever()

The Same-Origin Policy Gone Wild The $15000 PlayStation Bounty

comments powered by Disqus