How we test SDKs at RevenueCat

RevenueCat's approach to maintaining a robust iOS SDK involves an extensive testing process that includes unit tests, StoreKit tests, backend integration tests, SDK installation tests, API tests, documentation reference tests, and a testing app for UI-based features. These tests are designed to ensure the SDK's reliability across multiple platforms and OS versions, with continuous integration (CI) playing a significant role in automating the process. Unit tests are run continuously during development to identify bugs quickly, while StoreKit tests verify correct usage of StoreKit, particularly StoreKit 2, which cannot be mocked. Backend integration tests are pivotal in verifying real app interactions with the RevenueCat backend, albeit with less precision than unit tests. API tests prevent unintended changes to the public API, ensuring compatibility for various apps. SDK installation tests focus on the successful integration of the SDK across different package managers, and documentation reference tests ensure that public documentation remains accurate and helpful. RevenueCat also conducts testing parties and gathers feedback from both internal and external app developers to refine user experience and catch any oversights. This comprehensive testing strategy allows RevenueCat to roll out new features confidently, knowing that their SDK is well-tested and reliable across diverse environments, ultimately enhancing user experience and satisfaction.