anvil VS kotlin-inject

Handling DI at scale can be a challenging task in avoiding circular dependencies, build bottlenecks, and poor developer experience. To solve these challenges and make it easier for our developers, we adopted Anvil, a compiler plugin that allows us to invert how developers wire, hook up dependencies and keep our implementations loosely coupled. However, before we get into the juicy details of using this new compiler plugin, let's talk about our current implementation and its problems that we are trying to solve.

I would recommend checking out Anvil for multi-module setup as it can provide an easier migration path than Hilt while also providing similar features.

In terms of of what Anvil is I'd really recommend reading the README It's a lot more flexible than Hilt and provides a code generation hook so if you have a codebase specific DI pattern you can also set that up to be auto-generated with a custom annotation if you desire.

Recently I've been looking into Anvil as an alternative to Dagger and realised that it can handily generate factories for you without kapt. That brings some exciting potential for build time improvement so I was eager to try it out. However very quickly I got stuck.

I'm currently using Dagger and Anvil for my DI needs. It's been working really well, especially around what Anvil permits in terms of multibindings defined on the type declaration rather than in a module. For example:

Have you considered something like Dagger + Anvil (https://github.com/square/anvil) ? It works well enough that it gets out of your way. Is it doing insane things in the background at compilation time ? Absolutely. But it's worth it. And to save time in debug, you can shim in dagger-reflect to avoid the cost of having to run annotation processors.

For more details, check the official documentation: Anvil Dagger Factory Generation.

And finally, testing: Anvil offers a replace module feature that is handful to provide new dependencies during tests. For that, we create helper modules called testing and we provide fake dependencies of those replacing the production modules. Developers that include the testing in their test classpath can automatically interact with our testing utilities (or create their own, if required). To be completely honest here, it is more of an ongoing process.

Depending on how you use Anvil, there are definitely things that are missing/not as straightforward, e.g. https://github.com/evant/kotlin-inject/issues/212

I haven't used koin in a couple years since my newer jobs are using hilt and dagger/anvil for the compile time safety. In my personal projects I have been using kmp which when I started two years ago, koin was in alpha which was why I decided to use kodein. Which koin is stable in kmp, although kotlin-inject seems like a promising kmp compile time di in place of dagger.

I would suggest looking at Kotlin-Inject, it's a compile-time DI like Dagger and it's really pleasant to use.

Dagger/ hilt will not work on kmm being that they are java projects. One project that is similar to them though on kmm is kotlin-inject which might be something to look at.

I'd say it's second after this one. That one has multiplatform support in snapshot version, while mine is in alpha)

Interesting! Author of kotlin-inject here, I had originally looked at writing a compiler plugin and using IR but the lack of documentation and a non-stable api scared me away. Cool that you were able to get it to work.