oamg/leapp-repository/leapp-test-runner

Running Tests and Linters

All non-container commands (make lint, make test_no_lint, make fast_lint, make dev_test_no_lint) require a local virtualenv and REPOSITORIES envar (e.g. REPOSITORIES=common,el9toel10). Create the virtualenv with make install-deps (RHEL/CentOS) or make install-deps-fedora (Fedora) — the Makefile handles activation internally, don't source the venv manually. Set PYTHON_VENV=pythonX.X to choose the Python version (default: python3.6). Container commands are self-contained and recommended as the default.

Container-based testing (CI-equivalent, no venv needed)

Three container options, each matching a target Python version:

Pick container(s) based on which repositories your change touches:

• common actors: test on el8 and el9 (both Python versions must pass).
• el8toel9 only: el8 is sufficient, el9 is a bonus.
• el9toel10 only: el9.

# Full test (lint + unit tests) in container — recommended
TEST_CONTAINER=el9 make test_container
 
# Tests only, skip lint — faster iteration
TEST_CONTAINER=el9 make test_container_no_lint
 
# All containers at once
make test_container_all

Single-actor testing (fastest feedback, requires venv)

Use during development when iterating on one actor:

ACTOR=checkmemory make dev_test_no_lint

Linting

# Lint in container (recommended, no venv needed)
make lint_container
 
# Full lint locally (requires venv): pylint + flake8 + isort
make lint
 
# Lint only local changes relative to main branch (requires venv)
make fast_lint
 
# Auto-fix isort violations locally (requires venv)
make lint_fix

Choosing what to run

Interpreting failures

• pylint / flake8: fix the reported file and line.
• isort: run make lint_fix to auto-sort imports, then verify.
• pytest: read the traceback; check if the test uses monkeypatch.setattr correctly and if mocked models match current schema.