Vietnamese version

A convenient way to implement HTTP requests is using Pythons' requests library. One of requests’ most popular features is simple proxying support. HTTP as a protocol has very well-defined semantics for dealing with proxies, and this contributed to the widespread deployment of HTTP proxies

Proxying is very useful when conducting intensive web crawling/scrapping or when you just want to hide your identity (anonymization).

In this project I am using public proxies to randomise http requests over a number of IP addresses and using a variety of known user agent headers these requests look to have been produced by different applications and operating systems.

Proxies provide a way to use server P (the middleman) to contact server A and then route the response back to you. In more nefarious circles, it's a prime way to make your presence unknown and pose as many clients to a website instead of just one client. Often times websites will block IPs that make too many requests, and proxies is a way to get around this. But even for simulating an attack, you should know how it's done.

Surprisingly, the only thing that tells a server the application triggered the request (like browser type or from a script) is a header called a "user agent" which is included in the HTTP request.

The project code in this repository is crawling five different public proxy websites:

• http://proxyfor.eu/geo.php
• http://free-proxy-list.net
• http://rebro.weebly.com/proxy-list.html
• http://www.samair.ru/proxy/time-01.htm
• https://www.sslproxies.org

After collecting the proxy data and filtering the slowest ones it is randomly selecting one of them to query the target url. The request timeout is configured at 30 seconds and if the proxy fails to return a response it is deleted from the application proxy list. I have to mention that for each request a different agent header is used. The different headers are stored in the /data/user_agents.txt file which contains around 900 different agents.

If you wish to use this module as a CLI tool, install it globally via pip:

  pip install http-request-randomizer

Otherwise, you can clone the repository and use setup tools:

python setup.py install

Clone repo, install requirements, develop and run tests:

pip install -r requirements.txt
tox -e pyDevVerbose

• Command-line interface
• Library API

Assuming that you have http-request-randomizer installed, you can use the commands below:

show help message:

proxyList   -h, --help

specify proxy provider(s) (required):

  -s {proxyforeu,rebro,samair,freeproxy,all} 

Specify output stream (default: sys.stdout), could also be a file:

  -o, --outfile

specify provider timeout threshold in seconds:

  -t, --timeout

specify proxy bandwidth threshold in KBs:

  -bw, --bandwidth

show program's version number:

  -v, --version

To use http-request-randomizer as a library, include it in your requirements.txt file. Then you can simply generate a proxied request using a method call:

import logging
import time
from http_request_randomizer.requests.proxy.requestProxy import RequestProxy

if __name__ == '__main__':

    start = time.time()
    req_proxy = RequestProxy(log_level=logging.ERROR)
    print("Initialization took: {0} sec".format((time.time() - start)))
    print("Size: {0}".format(len(req_proxy.get_proxy_list())))
    print("ALL = {0} ".format(list(map(lambda x: x.get_address(), req_proxy.get_proxy_list()))))

    test_url = 'http://ipv4.icanhazip.com'

    while True:
        start = time.time()
        request = req_proxy.generate_proxied_request(test_url)
        print("Proxied Request Took: {0} sec => Status: {1}".format((time.time() - start), request.__str__()))
        if request is not None:
            print("\t Response: ip={0}".format(u''.join(request.text).encode('utf-8')))
        print("Proxy List Size: {0}".format(len(req_proxy.get_proxy_list())))

        print("-> Going to sleep..")
        time.sleep(10)

The RequestProxy constructor accepts an optional parameter of log_level that can be used to change the level of logging. By default, this is equal to 0, or NOTSET. The python logging levels are documented here. You can either use integers or their equivalent constant in the logging module. (e.g. logging.DEBUG, logging.ERROR, etc)

http-request-randomizer documentation

Many thanks to the open-source community for contributing to this project!

Open an issue here, and be as detailed as possible :)

Feel free to open a ticket! PRs are always welcome!

This project is licensed under the terms of the MIT license.