PythonScrapyBasicSetup
Basic setup with random user agents and proxy addresses for Python Scrapy Framework.
Setup
1. Install Scrapy Framework
pip install Scrapy
2. Install Beautiful Soup 4 for proxy middleware based on proxydocker lists
pip install beautifulsoup4
3. Install Tor, Stem (controller library for Tor), and Privoxy (HTTP proxy server).
apt-get install tor python-stem privoxy
Hash a password with Tor:
tor --hash-password secretPassword
Then copy a hashed password and paste it with control port to /etc/tor/torrc:
ControlPort 9051
HashedControlPassword 16:72C8ADB0E34F8DA1606BB154886604F708236C0D0A54557A07B00CAB73
Restart Tor:
sudo /etc/init.d/tor restart
Enable Privoxy forwarding by adding next line to /etc/privoxy/config:
forward-socks5 / localhost:9050 .
Restart Privoxy:
sudo /etc/init.d/privoxy restart
Both Tor and Privoxy should be up & running (check netstat -l). If you used different password or control port, update settings.py.
If you get some errors regarding the pyOpenSSL (check this issue), try to downgrade the Twisted engine:
pip install Twisted==16.4.1
Usage
To see what it does just:
python run.py
Project contains three middleware classes in middlewares directory. ProxyMiddleware downloads IP proxy addresses and before every process request chooses one randomly. TorMiddleware has a similar purpose, but it relies on Tor network. RandomUserAgentMiddleware downloads user agent strings and saves them into 'USER_AGENT_LIST' settings list. It also selects one UA randomly before every process request. Middlewares are activated in settings.py file.
This project also contains two spiders just for testing purposes, spiders/iptester.py and spiders/uatester.py. You can run them individually:
scrapy crawl UAtester
scrapy crawl IPtester
run.py file is a also good example how to include and run your spiders sequentially from one script.
If you have any questions or problems, feel free to create a new issue. Scrap responsibly!