Skip to content

README.md

Latest commit

 

History

History
147 lines (109 loc) · 5.57 KB

README.md

File metadata and controls

147 lines (109 loc) · 5.57 KB

LRR Integration Testing with Python

This directory contains the API/integration testing package for "aio-lanraragi". It includes tools for setting up and tearing down LRR docker/Windows environments, and creating synthetic archive data.

Integration testing version updates apply only if changes to integration testing code or tests have occurred.

For information on setting up a developer environment for testing, refer to development. Additionally, see resource management for documentation on how resources are managed during test-time.

Usage

Integration testing relies on a deployment environment. Currently two environments (Docker, Windows runfile) are supported. Ensure port range 3010-3020 (LRR testing ports) and 6389-6399 (Redis testing ports) are available.

Docker Deployment

This section assumes you have the following prerequisites:

  • Docker (confirm with docker version) and connection to Docker API (ls /var/run/docker.sock).

Clone and install the project:

git clone https://github.com/psilabs-dev/aio-lanraragi.git
cd aio-lanraragi
pip install -e ".[dev]"
pip install -e ./integration_tests

Enable BuildKit (or prepend it for your tests):

export DOCKER_BUILDKIT=1

All of the following are run within aio-lanraragi/integration_tests/.

Run integration tests on the official Docker image ("difegue/lanraragi"). Docker tests default to Valkey as the cache backend:

pytest tests

Use Redis explicitly if you need the old cache backend:

pytest tests --cache-backend redis

Run integration tests with a custom Docker image:

pytest tests --image myusername/customimage

Run integration tests with a Docker image built off a LANraragi git repo (with an optional ref):

pytest tests --git-url=https://github.com/difegue/LANraragi.git --git-ref=dev

Run integration tests with a Docker image built off a path to a local LANraragi project:

pytest tests --build /path/to/LANraragi/project

Run integration tests from a local project at a specific git ref (commit, branch, or tag):

pytest tests --build /path/to/LANraragi/project --build-ref abc123

Windows Deployment

This section assumes the following prerequisites for your development environment:

  • MSYS2 UCRT64 (ensure all your deps are in UCRT64) bash. To confirm, run echo $MSYSTEM and check that "UCRT64" is returned.
  • Visual Studio Desktop Development with C++ (tested in Community 2026)

However, you may run into further issues; in which case, please consult the LRR Discord.

Clone and install the project:

git clone https://github.com/psilabs-dev/aio-lanraragi.git
cd aio-lanraragi
pip install -e ".[dev]"
pip install -e ./integration_tests

Run the following from the LRR project root to install dependencies:

./tools/build/windows/install-deps.sh
./tools/build/windows/install.sh

Create the LRR distribution:

./tools/build/windows/create-dist.sh

Run integration tests on Windows from a pre-built distribution and an available staging directory (which you need to create):

# use full paths!
pytest tests --win-dist /path/to/win-dist --staging /path/to/staging

Deterministic Testing

By default, random variable sampling (e.g. for tag generation or list shuffling) is induced by seed value 42 via a numpy generator. You may change the seed to something else:

pytest tests/test_auth.py --npseed 43

Playwright UI Testing

Playwright integration tests are experimental. Install with:

pip install playwright
playwright install
pytest tests --playwright

Due to certain event loop quirks with pytest-playwright and compatibility issues with pytest-asyncio, we will only use playwright.

The process of adding UI tests can be broken to the following:

  • outlining the UI steps taken
  • converting those steps to Playwright

The conversion stage may be assisted via codegen, a tool which records browser actions into Python Playwright code:

npx playwright codegen --target=python http://localhost:3001

Logging

To see LRR process logs accompanying a test failure, use the pytest flag --log-cli-level=ERROR:

pytest tests/test_simple.py::test_should_fail --log-cli-level=ERROR
# ------------------------------------------------------- live log call --------------------------------------------------------
# ERROR    tests.conftest:conftest.py:84 Test failed: tests/test_simple.py::test_should_fail
# ERROR    aio_lanraragi_tests.lrr_docker:conftest.py:96 LRR: s6-rc: info: service s6rc-oneshot-runner: starting
# ERROR    aio_lanraragi_tests.lrr_docker:conftest.py:96 LRR: s6-rc: info: service s6rc-oneshot-runner successfully started
# ERROR    aio_lanraragi_tests.lrr_docker:conftest.py:96 LRR: s6-rc: info: service fix-attrs: starting

On test failures, pytest will attempt to collect the service logs from the running LRR process/container before cleaning the environment for the next test.

See pytest docs for more test-related options.

Scope

The scope of this library is limited to perform routine (i.e. not long-running by default) API integration or E2E tests within the "tests" directory. The library tests will confirm one or more of the following points:

  1. That the LRR server deployment was successful;
  2. That the functionality provided by LRR API is correct and according to API documentation;
  3. That the aio-lanraragi client API calls are correct.
  4. That LRR exhibits expected browser-side behavior.