Testing

Ferris provides utilities to test your application using `nose <>`_ and `webtest <>`_. Ferris provides a few scripts and utility classes to allow you to focus on writing tests instead of worrying about mocking App Engine.

Running Tests

To run tests you can use the backend_test.sh (or backend_test.bat):

scripts #> ./backend_test.sh

This script will invoke the test runner and discover all tests under app/tests and run them.

Alternatively, you can run ferris/scripts/test_runner.py to run the tests:

ferris/scripts #> python test_runner.py app

Writing Tests for Models

Models and other parts of the application that don’t involve WSGI (such as services) can be tested using WithTestBed:d

class ferris.tests.lib.WithTestBed(methodName='runTest')

Provides a complete App Engine test environment.

Here is a trivial example:

from app.models.cat import Cat
from ferris.tests.lib import WithTestBed

class CatTest(WithTestBed):
    test_herding(self):
        Cat(name="Pickles").put()
        Cat(name="Mr. Sparkles").put()

        assert Cat.query().count() == 2

Writing Tests for Handlers

Handlers, components, and templates usually have to be tested within a full WSGI environment. This is where AppTestCase comes in handy:

class ferris.tests.lib.AppTestCase(methodName='runTest')

Provides a complete App Engine test environment and also automatically routes all application and plugin handlers to testapp.

This allows you to make requests to your application just as you would if the application were running:

from ferris.tests.lib import AppTestCase

class CatsTest(AppTestCase):
    test_herding(self):
        Cat(name="Pickles").put()
        Cat(name="Mr. Sparkles").put()

        r = self.testapp.get('/cats')

        assert "Pickles" in r
        assert "Mr. Sparkles" in r