How to unit test your Juniper PyEZ project.

Unit Testing does not replace functional testing, however, it does help to improve code quality and reduce time required to access real environments by testing against mocked (faked) environments. Mocking is the replacement of one or more function calls or objects with mock calls or objects.

Unit testing help to prevent regression bugs. Testing with mocked environments enables you to test code paths that are difficult to reproduce with a real environment, such as exception handling.

This project includes a few examples and a small pyez utility called routing_neighbors.py with corresponding unit tests in the directory tests.

The examples are based on the python test framework pytest and travis as continuous integration (CI) service.

You execute the tests with the command: python -m pytest -v --durations=10 --cov="routing_neighbors" as shown in .travis.yml.

pytest auto discovery checks for all files in directory tests starting with tests_ and executes all included functions beginning with the same prefix.

The mock object for your PyEZ device connection (pyez_mock/device.py) allow you to test your PyEZ application against an emulated environment.

This mock object uses the root element tag of the rpc-request to return the corresponding rpc-reply stored in the rpc_reply_dict or as a file in the directory rpc-reply.

For the following rpc-request:

<get-route-information>
    <detail/>
</get-route-information>

The mocked PyEZ device attempts to retrieve the rpc-reply from the rpc_reply_dict using get-route-information as the key; otherwise, it attempts to retrieve the reply from the file tests/rpc-reply/get-route-information.xml relative to working directory. The parameters of the rpc-request will be ignored (e.g. ).

You must format the rpc-reply as a string with a valid XML structure and rpc-reply as root element:

from pyez_mock.device import rpc_reply_dict

rpc_reply_dict['get-alarm-information'] = """
    <rpc-reply>
        <alarm-information>
            <alarm-summary>
                <no-active-alarms/>
            </alarm-summary>
        </alarm-information>
    </rpc-reply>"""

The rpc_reply_dict allows you to generate a rpc-reply dynamically during test execution or to return exceptions (see tests/test_exceptions.py).

from pyez_mock.device import device

dev = device(host="1.2.3.4", user="juniper")
result = device.rpc.get_route_information(detail=True)
assert result.findtext(".//destination-count") == "5"

pytest_device is a pre-defined pytest fixture with module level scope and default credentials. You can create custom device fixtures as shown in tests/test_device.py.

from pyez_mock.device import pytest_device as device

def test_default(device):
    result = device.rpc.get_route_information(detail=True)
    assert result.findtext(".//destination-count") == "5"