With the purpose of learning Hilt I started "migrating" my multi-module Kotlin app from using classes with static methods as helpers to Hilt injection.
After a lot of headaches, now everything's going fine, and I could say I'm happy and in the 80% of the migration, but suddenly a question have come to my mind and stopped coding.
I have some modules, like "Core", "Common", "Repository", etc, and each have its own helpers (although I have to say that most of them are in "Common"), the other are "Repositories", "DAO", "Services" and so) and I wonder to know if the best practice (if there is one regarding this) is to do injection for everything (like a screen helper with methods to convert sp to px and so, language helper to get string from resources, etc), or injection should be only used for context, database, repos, services and that kind of classes.
I'm happy with my progress, and I'm quite sure I'm going in the right direction to make a complete decoupling, but wanted to know what do you have to say about that.
Thanks in advance.