A novel way of handling malicious requests within a Django application.

Install from PyPi (stable):

pip install django-rickroll

or Github (dev):

pip install -e git://github.com/codeinthehole/django-rickroll.git#egg=django-rickroll

Add 'rickroll' to your INSTALLED_APPS and the following to your MIDDLEWARE_CLASSES:

'rickroll.middleware.HackingAttemptMiddleware'

Wherever you detect a hacking attempt, raise a HackingAttempt exception and the middleware will rickroll the offending user.

For example, in your views.py:

from rickroll.exceptions import HackingAttempt

def some_view(request):
    # ...
    if hacking_detected:
        raise HackingAttempt()

then the user in question will get redirected to http://www.youtube.com/watch?v=dQw4w9WgXcQ. The destination can be overridden using the setting RICKROLL_URL.

If your site ever gets penetration tested, the testing company will likely manipulate every request to your site where there are parameters being passed (query parameters, POST params, cookies etc). When you encounter a parameter which should be an integer but has value ../../../../../../etc/passwd, it's not always clear what the "right" response should be. I thought it would be funny to rickroll.