quickproxy

A lightweight, asynchronous, programmable HTTP proxy for python. Built with Tornado.

Use

A simple proxy:

import quickproxy
quickproxy.run_proxy(port=8080)

A reverse proxy:

This proxy will fetch responses from an AWS s3 bucket with the same ID as the request's hostname.

def callback(request):
	request.host = request.host+".s3-website-us-east-1.amazonaws.com"
	request.port = 80
	return request

quickproxy.run_proxy(port=8080, req_callback=callback)

Reference

Quickproxy exposes just one function:

run_proxy(port,
          methods=['GET', 'POST'], 
          req_callback=DEFAULT_CALLBACK,
          resp_callback=DEFAULT_CALLBACK,
          err_callback=DEFAULT_CALLBACK,
          start_ioloop=True)

It runs a proxy on the specified port. You can pass the following parameters to configure quickproxy:

• methods: the HTTP methods this proxy will support

• req_callback: a callback that is passed a RequestObj that it should modify and then return. By default this is the identity function.

• resp_callback: a callback that is given a ResponseObj that it should modify and then return. By default this is the identity function.

• err_callback: in the case of an error, this callback will be called. there's no difference between how this and the resp_callback are used. By default this is the identity function.

• start_ioloop: if True (default), the tornado IOLoop will be started immediately.

Request callback functions

The request callback should receive a RequestObj and return a RequestObj.

request_callback(requestobj)
	return requestobj

The RequestObj is a python object with the following attributes that can be modified before it is returned:

• protocol: either 'http' or 'https'

• host: the destination hostname of the request

• port: the port for the request

• path: the path of the request ('/index.html' for example)

• query: the query string ('?key=value&other=value')

• fragment: the hash fragment ('#fragment')

• method: request method ('GET', 'POST', etc)

• username: always passed as None, but you can set it to override the user

• password: None, but can be set to override the password

• body: request body as a string

• headers: a dictionary of header / value pairs (for example {'Content-Type': 'text/plain', 'Content-Length': 200})

• follow_redirects: true to follow redirects before returning a response

Response callback functions

The response and error callbacks should receive a ResponseObj and return a ResponseObj, similar to the request callback above.

The ResponseObj is a python object with the following attributes that can be modified before it is returned:

• code: response code, such as 200 for 'OK'

• headers: the response headers

• pass_headers: a list or set of headers to pass along in the response. All other headers will be stripped out. By default this includes:

  ('Date', 'Cache-Control', 'Server', 'Content-Type', 'Location')

• body: response body as a string

Credits

Much of this code was adopted from Senko's tornado-proxy:

https://github.com/senko/tornado-proxy

...which is itself based on the code by Bill Janssen posted to: http://groups.google.com/group/python-tornado/msg/7bea08e7a049cf26