This library provides the core interface, `IInjector`, as well as a small set of utils to make your life easier. Your patches should implement `IInjector`.
*`inject(ClassNode, MethodNode)`: will be invoked providing you the ClassNode and MethodNode you requested. This is where the actual patching will happen.
There's some more optional methods you *don't have to* implement, but really should because they will make your life considerably easier when it's time to debug:
Finally, you should mark your classes as service providers, by creating a text file called `ftbsc.lll.IInjector` in `src/main/resources/META-INF/services` on your project. Inside, put the fully qualified names of your patches (example: `ftbsc.bscv.asm.patches.TestPatch$TickPatch`).
You are going to need an appropriate loader to use Lillero patches: **this is just a library and does nothing by itself**. You need to make it work by loading services implementing the `IInjector` interface, and by calling their `inject(ClassNode, MethodNode)` methods with the appropriate parameters.
It should also be noted that this library will be required at runtime for your patches to function. You will either have to bundle it using [Shadow](https://github.com/johnrengelman/shadow) or similar, or get this into your runtime classpath by some other means (our [loader](https://github.com/zaaarf/lillero-loader/), for instance, takes care of this for you).
Finally, know that you can spare yourself some trouble, by using this [annotation processor](https://github.com/zaaarf/lillero-processor/) to reduce boilerplate code to a minimum.
* In "raw" environments, you want to be using Notch (fully obfuscated) names whenever you are told to reference a class or method by name, since those are the ones that exist at runtime.
- Use deobfuscated names if you are running from ForgeGradle's or loom's runClient task.
- If you are using our loader (see below), use intermediary (unreadable but unique) names in every place you are told to use a name - ModLauncher will do the rest.