kotlin-context-di

Manual dependency injection in Kotlin can be set up by defining an interface-first AppDependencies contract and wiring components through separate SystemContext and SystemTestContext implementations. This approach keeps initialization transparent without relying on open classes or lazy properties.

The best way to wire Kotlin apps for testability is using interface-first dependency contracts where controllers and route functions accept dependency interfaces rather than framework lifecycle objects. Fresh per-test contexts with in-memory fakes ensure standalone test wiring.

Yes, manual DI works with Ktor by passing dependency interfaces to route functions and using a standalone SystemTestContext. This creates fresh, self-contained test contexts using fakes, completely avoiding inheritance between production and test code.

No, manual dependency injection in Kotlin does not require open classes or lazy initialization. It uses type-safe eager initialization with explicit production wiring, keeping the dependency setup transparent and straightforward.

You can override specific dependencies in a Kotlin test context using delegation-based partial overrides. The SystemTestContext utilizes covariant override inference with inner grouping classes, allowing targeted concrete fake access without rewriting the entire wiring setup.