litefs-testing

---

name: litefs-testing description: Testing patterns for litefs-py and litefs-django. Use when writing tests, setting up fixtures, understanding test organization, or configuring pytest marks. Triggers: test, pytest, unit test, integration test, property-based testing, hypothesis, fixtures, in-memory adapters.

LiteFS Testing Strategy

Project-specific testing patterns for litefs-py and litefs-django.

When to Use

• Writing tests for LiteFS-related code
• Setting up test fixtures
• Understanding test organization

Test Organization

# By category
pytest -m unit           # Fast, no LiteFS process
pytest -m integration    # With Docker + FUSE
pytest -m property       # Property-based tests

Test Directory Structure

tests/
├── unit/                    # No I/O, no subprocess
│   ├── domain/              # Pure domain logic (settings, config)
│   ├── usecases/            # Use case tests with in-memory adapters
│   └── conftest.py          # In-memory adapter fixtures
├── integration/             # Requires Docker + FUSE
│   ├── test_backend.py
│   └── test_replication.py
└── conftest.py              # Shared fixtures

CI Strategy

• Unit tests: Run on all PRs (fast, no Docker)
• Integration tests: Run on main branch only (requires Docker + FUSE)
• Local testing: Docker Compose setup mirrors integration CI

Test Isolation Rules

Unit tests (tests/unit/) must NOT:

• Spawn subprocesses or call LiteFS binary
• Access filesystem beyond temp directories
• Require FUSE or Docker

Unit tests SHOULD use:

• In-memory adapter fixtures:
  • in_memory_process_runner → InMemoryProcessRunner
  • in_memory_config_writer → InMemoryConfigWriter
  • fake_primary_detector → FakePrimaryDetector

In-Memory Adapter Pattern

# tests/unit/conftest.py
@pytest.fixture
def in_memory_process_runner():
    return InMemoryProcessRunner()

# tests/unit/usecases/test_process_manager.py
def test_start_litefs(in_memory_process_runner):
    manager = ProcessManager(runner=in_memory_process_runner)
    manager.start()
    assert in_memory_process_runner.started is True

Property-Based Testing (Hypothesis)

Use for config generation and settings validation:

uv run pytest -m property -v

When to use:

Pattern	Property to Test	Example
Round-trip	parse(generate(x)) == x	Config YAML generation
Invariants	Bounds always hold	Path validation
Idempotence	f(f(x)) == f(x)	Settings normalization

from hypothesis import given, strategies as st

@pytest.mark.property
@given(mount_path=st.text(min_size=1).filter(lambda x: x.startswith("/")))
def test_config_includes_mount_path(mount_path):
    settings = LiteFSSettings(mount_path=mount_path)
    config = ConfigGenerator().generate(settings)
    assert mount_path in config