Reliable Typed Python Applications

PYTHON PRO: PRODUCTION CHECKLIST

CODE STYLE
✓ Type hints: strict mode enabled (mypy --strict on all files)
✓ All functions annotated: params + return types
✓ Dataclasses used for models (frozen=True for immutable)
✓ Black formatting applied (line length 88)
✓ Ruff checks pass (all categories)

ASYNC PATTERNS
✓ Async functions: use async/await, no blocking calls in event loop
✓ Dependency injection: accept clients as params, never global singletons
✓ Error handling: custom exceptions (inherit from base exception class)
✓ Logging: structured JSON logs (include context, trace_id)

TEST SUITE
✓ Pytest fixtures: isolated test database, mock clients
✓ Mocking: external APIs mocked (requests-mock), no live calls
✓ Coverage: 85% minimum, gaps flagged in CI
✓ Parametrized tests: test multiple inputs in 1 test function
✓ Slow tests: marked with @pytest.mark.slow, skipped in fast mode

SCRIPT EXAMPLE OUTPUT
# models.py
@dataclass(frozen=True)
class User:
    id: int
    name: str
    email: str

async def fetch_user(user_id: int, db: Database) -> User:
    """Fetch user from database. Raises UserNotFoundError if missing."""
    result = await db.query("SELECT * FROM users WHERE id = ?", user_id)
    if not result:
        raise UserNotFoundError(f"User {user_id} not found")
    return User(**result)

# tests/test_models.py
@pytest.mark.asyncio
async def test_fetch_user_success(mock_db):
    mock_db.query.return_value = {"id": 1, "name": "Alice", "email": "alice@x.com"}
    user = await fetch_user(1, mock_db)
    assert user.id == 1