fix(tests): add mocked dummy tests that run without a database

[AdarshJ173]

Summary

Closes #236

The existing tests/test_dummy.py requires a live database session (dbsession / dbpool fixtures) and fails in the default GitHub Actions environment because no database service is configured in the pytest job.

This PR adds a fully mocked alternative that runs in any plain Python environment — no Docker, no Postgres, no MySQL required.

---

Root Cause

test_creation and test_getting both accept ORM-specific fixtures (dbsession: AsyncSession / dbpool: AsyncConnectionPool) and call real DummyDAO methods that hit the database. The CI pytest job in tests.yml does not start any database service, so these tests always fail outside of the full docker-compose stack.

---

What Changed

tests/test_dummy_mocked.py (new file)

• Rewrites test_creation_mocked and test_getting_mocked using unittest.mock.AsyncMock
• Patches DummyDAO.create_dummy_model and DummyDAO.filter at the module level so no real DB connection is opened
• Still exercises the actual HTTP endpoints via AsyncClient — the API layer is tested for real; only the DAO ↔ DB hop is stubbed
• Uses side_effect on filter to simulate the before/after state (empty list → populated list) without any persistent state
• Works regardless of which ORM (sqlalchemy, psycopg, beanie, ormar) or API type (rest, graphql) is selected

.github/workflows/tests.yml

• Adds a new pytest-mocked job that:
  • Checks out the repo
  • Installs deps with uv sync (no Docker needed)
  • Runs uv run pytest tests/test_dummy_mocked.py -vv directly on the runner
• The existing pytest (docker-compose) job is preserved unchanged — full integration tests still work locally
• Both jobs run on every push

---

Why This Approach

The issue author specifically requested "providing mocked tests [as] an example of how to write mocked tests". Using unittest.mock.AsyncMock (stdlib, no new dependencies) is the idiomatic Python pattern for this and serves as a clear template for users of the generated project.

---

Testing

# No database needed — runs immediately in any Python 3.12+ env:
uv run pytest tests/test_dummy_mocked.py -vv

Expected output:

PASSED tests/test_dummy_mocked.py::test_creation_mocked
PASSED tests/test_dummy_mocked.py::test_getting_mocked
2 passed in <1s

Notes

• No new dependencies — unittest.mock is stdlib
• No changes to existing test files or fixtures
• No schema changes
• No breaking changes — strictly additive
• The pytest-mocked CI job runs in ~30 seconds with no services, making it an ideal fast-feedback check for PRs