dockerctx

dockerctx is a context manager for managing the lifetime of a docker container.

The main use case is for setting up scaffolding for running tests, where you want something a little broader than unit tests, but less heavily integrated than, say, what you might write using Robot framework.

Install

$ pip install dockerctx

The development-specific requirements will be installed automatically.

Demo

This is taken from one of the tests:

import time
import redis
import pytest
from dockerctx import new_container

# First make a pytest fixture

@pytest.fixture(scope='function')
def f_redis():

    # This is the new thing! It's pretty clear.  The `ready_test` provides
    # a way to customize what "ready" means for each container. Here,
    # we simply pause for a bit.

    with new_container(
            image_name='redis:latest',
            ports={'6379/tcp': 56379},
            ready_test=lambda: time.sleep(0.5) or True) as container:
        yield container

# Here is the test.  Since the fixture is at the "function" level, a fully
# new Redis container will be created for each test that uses this fixture.
# After the test completes, the container will be removed.

def test_redis_a(f_redis):
    # The container object comes from the `docker` python package. Here we
    # access only the "name" attribute, but there are many others.
    print('Container %s' % f_redis.name)
    r = redis.StrictRedis(host='localhost', port=56379, db=0)
    r.set('foo', 'bar')
    assert r.get('foo') == b'bar'

Note that a brand new Redis container is created here, used within the context of the context manager (which is wrapped into a pytest fixture here), and then the container is destroyed after the context manager exits.

In the src, there is another, much more elaborate test which

1. runs a postgres container;
2. waits for postgres to begin accepting connections;
3. creates a database;
4. creates tables (using the SQLAlchemy ORM);
5. performs database operations;
6. tears down and removes the container afterwards.