1.20Question 3

Prompt: You need to verify that a soft-delete feature works correctly. The API returns 204 on DELETE, and the user no longer appears in the GET /users list. How would you verify this at the database level, and what edge cases would you check?

What a strong answer should cover:

• Confirming deleted_at is set to a recent timestamp (not NULL)
• Confirming the row still exists (it is a soft delete, not a hard delete)
• Checking that other fields (name, email) remain intact
• Edge cases: double-delete, re-activation, cascade effects on related records

Example answer:

• I would query SELECT id, email, deleted_at FROM users WHERE email = '<test_email>' and assert the row still exists with deleted_at set to a timestamp within the last few seconds.
• I would verify the row is excluded from the application's "active users" query by checking WHERE deleted_at IS NULL returns no match, while WHERE deleted_at IS NOT NULL does.
• Edge cases I would test: calling DELETE twice on the same user (idempotency), attempting to re-activate a soft-deleted user, and verifying that the user's orders and profiles are still intact (soft delete should not cascade to related records unless the business rules specify otherwise).