Mock + Proxy in Automated Testing for an Advertising SDK

Project Background The advertising SDK provides a jar package for app developers to integrate a mobile ad platform, offering functions like ad requests, behavior tracking, error logging, and anti‑fraud.

Team Situation The project is split between a Shanghai backend team and a Beijing client team, leading to synchronization issues, unstable API documentation, frequent protocol changes, lack of a server test environment, and difficulty covering numerous protocol fields with traditional test case design.

Key Problems 1. Insufficient server test environment and realistic ad content. 2. Overly complex protocol fields causing test case omissions, especially for error handling. 3. Test case design heavily impacted by requirement changes, making updates costly. 4. Manual testing is inefficient and prone to missing cases.

Analysis & Approach The article focuses on solving problems 4‑7. For the missing server environment, a mock approach is adopted: the backend pre‑configures fixed online ad resources, allowing the client to be tested without real backend interaction. To address protocol complexity, a reliable test case design strategy is created to cover the maximum scenarios with the fewest cases.

Solution Details - Split test cases into normal‑return and exception handling groups. - For normal flows, create a checklist covering environment, network, download, UI checks, etc., reviewed by the team and product owners. - For exception flows, automatically generate test cases based on defined error codes (e.g., 21 error codes) to achieve near‑100% automation from case generation to execution. - Maintain real‑time updates of test cases as requirements evolve.

Q&A Highlights Q: How to test without a server environment? A: Use mock data with pre‑configured online ad resources. Q: How to handle many protocol fields? A: Adopt a minimal‑case coverage strategy. Q: How to reduce cost of test case changes? A: Separate normal and exception cases, use checklists, and automate exception case generation. Q: How to improve manual test efficiency? A: Automate error‑code coverage and aim for full automation of case execution.

Conclusion: The mock‑proxy method combined with a structured, automated test case design significantly improves testing efficiency and coverage for the advertising SDK.