pytest-jubilant 0.8

Add your description here

Pytest plugin for jubilant.

Eases the transition from pytest-operator to jubilant. And some cool stuff on top.

Fixtures

juju

This is a module(and model!)-scoped fixture that, by default, uses a temporary model and tears it down on context exit. Cfr. the --model, --keep-models, and --no-teardown options below for more. Usage:

from jubilant import Juju, all_active

def test_deploy(juju: Juju):
    juju.deploy("./foo.charm", "foo")
    juju.wait(
        lambda status: all_active(status, "foo"),
        timeout=1000
    )

temp_model_factory

This is a module-scoped fixture that manages temporary models for your test runs. It is what the juju fixture is using behind the scenes.

Especially useful if you have test cases that require multiple models.

import pytest
from jubilant import Juju, all_active

@pytest.fixture
def istio(temp_model_factory):
    yield temp_model_factory.get_juju(suffix="istio")
    
    
def test_cmr(juju: Juju, istio: Juju):
    istio.deploy("istio-k8s", "istio")
    istio.wait(
        lambda status: all_active(status, "istio"),
        timeout=1000
    )

    juju.deploy("./foo.charm", "foo")
    juju.wait(
        lambda status: all_active(status, "foo"),
        timeout=1000
    )

    juju.cli("offer", "foo:bar") 
    istio.cli("consume", f"{juju.model}:foo")
    istio.cli("relate", "istio", "foo:bar")

This test will spin up two temporary models, one called test-cmr-<randomhex>, and one called test-cmr-<randomhex>istio, and tear them down on context exit.

This fixture can be used with the options described below:

• pytest tests/test_cmr.py --keep-models will skip model teardown for all generated models.
• pytest tests/test_cmr.py --model test-cmr-<randomhex> will use test-cmr-<randomhex> as base name, and the suffixes you defined in the fixtures will give all generated models predictable names, which means that the tests will reuse the existing models (if found) or create new ones with those names.
• pytest tests/test_cmr.py --switch will switch you to the 'base' model test-cmr-<randomhex> (not to one of the suffixed ones!).

Pytest CLI options

--model

Override the default model name generation (test module + random bits) and use a fixed model name instead. Do note that this model will be torn down at the end of the test run just like any other, so if you're targeting an existing model you care about, don't forget the --no-teardown flag!.

Usage:

pytest ./tests/integration -k test_foo --model "model2" 
# runs the tests on a new 'model2' model and tears it down afterwards

juju add-model model1
pytest ./tests/integration -k test_foo --model "model1" --no-teardown 
# runs the tests on the existing 'model1' model, and keeps it

--switch

Switch to the (randomly-named?) model that is currently in scope, so you can keep an eye on the juju status as the tests progress. (Won't work well if you're running multiple test modules in parallel.)

Usage:

pytest ./tests/integration -k test-something --switch  
# will switch you to the test-something-09i123451 (random bits may differ) model as soon as it's created 

pytest ./tests/integration -k test-something --model mymodel --switch  
# will switch you to the `mymodel` model as soon as it's created 

--keep-models

Skip destroying the newly generated models when the tests are done. Usage:

pytest ./tests/integration --keep-models  

--no-teardown

Skip all tests marked with teardown (and skip destroying the models! this flag implies the --keep-models one). Useful to inspect the state of a model after a (failed) test run.

Usage: pytest ./tests/integration --no-teardown

--no-setup

Skip all tests marked with setup. Especially useful when re-running a test on an existing model which is already set-up, but not torn down. See this article for the idea behind this workflow. Usage:

pytest ./tests/integration --no-teardown # make a note of the temporary model name
pytest ./tests/integration --model <temporary model name> --no-setup 

--dump-logs

Prior to tearing down all models owned by a temp_model_factory (i.e. prior to cleaning up a test module execution), dump the juju debug-log --replay for each model into a directory (default "<CWD>/.logs"). File naming scheme is:

<module name>-<random bits>[-<suffix>]-jdl.txt

Usage:

pytest ./tests/integration ./integration/test_ingress.py --dump-logs=./debug_logs
# once the tests are done, you'll find the logs in 
# ./debug_logs/test-ingress-c372ef49-jdl.txt (random bits may vary).

pytest ./tests/integration ./integration/test_ingress.py --model foo --dump-logs=./debug_logs
# once the tests are done, you'll find the logs in
# ./debug_logs/foo-jdl.txt

pytest ./tests/integration ./integration/test_ingress.py --dump-logs=""
# no logs will be saved

Markers

setup

Marker for tests that prepare (parts of) a model.

Usage:

import pytest

@pytest.mark.setup
def test_deploy(juju):
    juju.deploy("A")
    juju.deploy("B")

@pytest.mark.setup
def test_relate(juju):
    juju.integrate("A", "B")

teardown

Marker for tests that destroy (parts of) a model.

Usage:

import pytest

@pytest.mark.teardown
def test_disintegrate(juju):
    juju.remove_relation("A", "B")

@pytest.mark.teardown
def test_destroy(juju):
    juju.remove_application("A")
    juju.remove_application("B")

Utilities

pack and get_resources

Wrapper around charmcraft pack to build a charm and return the packed charm path, ready to be passed to juju.deploy. get_resources will parse a charmcraft.yaml file and return a mapping from resources to their upstream-source field as is standard convention.

# example /path/to/foo-charm-repo-root-dir/charmcraft.yaml

# [snip] 
resources:
  nginx-image:
    type: oci-image
    description: OCI image for nginx
    upstream-source: ubuntu/nginx:1.24-24.04_beta
  nginx-prometheus-exporter-image:
    type: oci-image
    description: OCI image for nginx-prometheus-exporter
    upstream-source: nginx/nginx-prometheus-exporter:1.1.0

Usage:

from pytest_jubilant import pack, get_resources
import pytest


@pytest.mark.setup
def test_build_deploy_charm(juju):
    charm_root = "/path/to/foo-charm-repo-root-dir/"
    juju.deploy(
        pack(charm_root),
        "foo",
        # the resources can only be inferred from the charm's metadata/charmcraft yaml 
        # if you use the `upstream-source` convention
        resources=get_resources(charm_root),
        num_units=3,
    )

If you have a multiplatform charm, you can use: pack("./", platform="ubuntu@24.04").

DEVELOPERS

To release:

# obtain the current latest version out there
git tag | tail -n 1

new_tag="v0.5"  # for example!
git tag $new_tag -m "new fancy feature"
git push origin head --tag

Once the PR is merged, the release CI will kick in and put the tag in pytest_jubilant.version.py