ProxyBroker is an open source tool that asynchronously finds public proxies from multiple sources and concurrently checks them (type, level of anonymity, country). Supports HTTP(S) and SOCKS.

• Gathers proxies from 50+ sources, finds ~7000 HTTP(S) and ~500 SOCKS working proxies

  Sources are the websites that publish free public proxy lists daily. And much more: you can add custom sources - websites or a raw data. Detects and recognize proxies in the text (no matter how dirty the data).

• All protocols support

  Proxies can be check for work by HTTP, HTTPS (via CONNECT), SOCKS4 and SOCSK5 protocols.

• Checks proxies on the level of anonymity

  Supports levels: Transparent, Anonymous, High. You can add your own judges.

• Filter proxies by country

  Determines location (country) of the proxy and checks only the specified.

• Is asynchronous

  That helps increase checking speed and decrease waiting time. It's really fast: just in a minute, it will give you ~250 working HTTP proxies.

• Automatically removes duplicate proxies.

• Python 3.5 or higher
• aiohttp
• aiodns
• maxminddb

To install last stable release from pypi:

$ pip install proxybroker

If you use Mac OSX (without XCode) or Microsoft Windows (without Visual Studio or Windows SDK) there may be a problem with the compiling dependencies (pycares). Use this way:

$ git clone https://github.com/constverum/ProxyBroker.git
$ cd ProxyBroker
$ pip install -r requirements.txt proxybroker -f proxybroker/data/wheels

And you can install development version:

$ git clone https://github.com/constverum/ProxyBroker.git
$ cd ProxyBroker
$ python setup.py install

import asyncio
from proxybroker import Broker

loop = asyncio.get_event_loop()

proxies = asyncio.Queue(loop=loop)
broker = Broker(proxies, loop=loop)

loop.run_until_complete(broker.find(limit=4))

while True:
    proxy = proxies.get_nowait()
    if proxy is None: break
    print('Found proxy: %s' % proxy)

As the final result, we get the Proxy objects. And we can get all the information we need through Proxy properties.

Found proxy: <Proxy AU 0.72s [HTTP: Transparent] 1.1.1.1:80>
Found proxy: <Proxy FR 0.33s [HTTP: High, HTTPS] 2.2.2.2:3128>
Found proxy: <Proxy US 1.11s [HTTP: Anonymous, HTTPS] 3.3.3.3:8000>
Found proxy: <Proxy DE 0.45s [SOCKS4, SOCKS5] 4.4.4.4:1080>

import asyncio
from proxybroker import Broker

async def use(proxies):
    while True:
        proxy = await proxies.get()
        if proxy is None:
            break
        elif 'SOCKS5' in proxy.types:  # filter by type
            print('Found SOCKS5 proxy: %s' % proxy)
        else:
            print('Found proxy: %s' % proxy)

async def find(proxies, loop):
    broker = Broker(queue=proxies,
                    timeout=8,
                    attempts_conn=3,
                    max_concurrent_conn=200,
                    judges=['https://httpheader.net/', 'http://httpheader.net/'],
                    providers=['http://www.proxylists.net/', 'http://fineproxy.org/eng/'],
                    verify_ssl=False,
                    loop=loop)

    # only anonymous & high levels of anonymity for http protocol and high for others:
    types = [('HTTP', ('Anonymous', 'High')), 'HTTPS', 'SOCKS4', 'SOCKS5']
    countries = ['US', 'GB', 'DE']
    limit = 10

    await broker.find(types=types, countries=countries, limit=limit)

if __name__ == '__main__':
    loop = asyncio.get_event_loop()
    proxies = asyncio.Queue(loop=loop)
    tasks = asyncio.gather(find(proxies, loop), use(proxies))
    loop.run_until_complete(tasks)

In this example, we explicitly specify the parameters that directly affect on the speed of gathering and checking proxies (see Broker parameters). In most cases it's redundant.

Usually, we want to find:

• a certain number of specific type of proxies
• with a high level of anonymity
• and from specific countries

To do this, we pass the parameters types, countries, and limit to the find method (see Broker methods).

We use two asynchronous functions that execute in parallel:

• find() - gather proxies from the providers, check and pass them to the async queue proxies
• use() - use the checked proxies from proxies without having to wait for the end of the gather

Note: You can start to use the checked proxies for a couple of seconds after the start of the gather. Gather and check of new proxies will continue until the limit is reached or until we not visit all the providers and check all the proxies received from them.

# raw_data.txt
10.0.0.1:80
OK 10.0.0.2:   80 HTTP 200 OK 1.214
10.0.0.3;80;SOCKS5 check date 21-01-02
>>>10.0.0.4@80 HTTP HTTPS status OK

# example.py
# ...
broker = Broker(proxies, loop=loop)

with open('raw_data.txt', 'r') as f:
    data = f.read()

await broker.find(data=data)
# ...

As a source of proxies, instead of the providers, you can use your own source data (it's usual local .txt file). Simply pass your data to the data parameter. Note: At the moment, information about the type of proxy in the raw data is ignored.

# ...
await broker.grab(countries=['US'], limit=100)
# ...

Use the grab method if you want only to gather proxies without a check. Note: The number of found proxies can reach over 40k.

• Check the ping, response time and speed of data transfer
• Check on work with the Cookies/Referrer/POST
• Check site access (Google, Twitter, etc) and even your own custom URL's
• Check proxy on spam. Search proxy ip in spam databases (DNSBL)
• Information about uptime
• Checksum of data returned
• Support for proxy authentication
• Finding outgoing IP for cascading proxy
• The ability to send mail. Check on open 25 port (SMTP)
• The ability to specify the address of the proxy without port (try to connect on defaulted ports)
• The ability to save working proxies to a file (text/json/xml)

• Fork it: https://github.com/constverum/ProxyBroker/fork
• Create your feature branch: git checkout -b my-new-feature
• Commit your changes: git commit -am 'Add some feature'
• Push to the branch: git push origin my-new-feature
• Submit a pull request!

Licensed under the Apache License, Version 2.0

This product includes GeoLite2 data created by MaxMind, available from http://www.maxmind.com.