Batching

Overview
API
Example

Uni Task provides an easy way to achive multi-threading using UniTask.RunOnThreadPool().

CBSL extension for UniTask provides batching API to split tasks on multiple threads.

Batch Scheduler#

BatchScheduler.Process() takes in a IEnumerable<I> Where I represents input data, then it divides the data into batches and processes each batch on seprate threads.

The process method has 2 overloads :-

public static async UniTaskVoid Process<I>(IEnumerable<I> input, int size, Action<I> process, Action<Batch<I>>? preProcess, Action<Batch<I>>? postProcess)

public static async UniTask<IEnumerable<O>> Process<I, O>(IEnumerable<I> input, int size, Func<I, O> process, Action<Batch<I>>? preProcess, Action<Batch<I>, O[]>? postProcess)

the difference between the two is in the for each process, one takes a Func<I, O> and one takes an Action<I>. Use the Func overload if you want your tasks to return something, the result is also an array corresponding to each input.

PreProcess & PostProcess callbacks are run on main thread for each batch and can be null.