Skip to content

dcermak/pytest_container

Repository files navigation

Pytest container

https://codecov.io/gh/dcermak/pytest_container/branch/main/graph/badge.svg?token=D16Q2PGL67 https://app.fossa.com/api/projects/git%2Bgithub.com%2Fdcermak%2Fpytest_container.svg?type=shield PyPI

Find the latest documentation on dcermak.github.io/pytest_container.

pytest_container is a pytest plugin to test container images via pytest fixtures and testinfra. It takes care of all the boring tasks, like spinning up containers, finding free ports and cleaning up after tests, and allows you to focus on implementing the actual tests.

The plugin automates the following tasks:

  • pull, launch, and stop containers
  • build containers using a Dockerfile
  • wait for containers to become healthy before executing tests
  • bind exposed container ports to free ports on the host
  • mount volumes via temporary directories
  • parallel test execution through pytest-xdist
  • build dependent container images in the correct order
  • run the same test on as many container images as necessary
  • create, launch and destroy podman pods

pytest_container provides four fixtures that give you everything you need for testing containers. Spinning up a container image can be as simple as instantiating a Container and parametrizing a test function with the container fixture:

TW = Container(url="registry.opensuse.org/opensuse/tumbleweed:latest")

@pytest.mark.parametrize("container", [TW], indirect=["container"])
def test_etc_os_release_present(container: ContainerData):
    assert container.connection.file("/etc/os-release").exists

The fixture automatically pulls and spins up the container, stops it and removes it after the test is completed. Your test function receives an instance of ContainerData with the ContainerData.connection attribute. The ContainerData.connection attribute is a testinfra connection object. It can be used to run basic tests inside the container itself. For example, you can check whether files are present, packages are installed, etc.

Use cases

  1. Run functional tests on operating system container images
  2. Verify your software on multiple operating systems

About

Collection of pytest helpers and fixtures to test container images

Topics

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages