Async Rust

Chapter 5. Coroutines

At this stage in the book, you should be comfortable with async programming. When you see await syntax in code now, you know what is happening under the hood with futures, tasks, threads, and queues. However, what about the building blocks of async programming? What if we can get our code to pause and resume without having to use an async runtime? Furthermore, what if we can use this pause-and-resume mechanism to test our code by using normal tests? These tests can explore how code behaves under various polling orders and configurations. Our pause-and-resume mechanism can also be the interface between synchronous code and async. This is where coroutines come in.

By the end of this chapter, you should be able to define a coroutine and explain how they can be used. You should be able to integrate coroutines into your own programs to keep memory consumption low for tasks that would require large amounts of memory. You will also be able to mimic async functionality without an async runtime by using coroutines and implement a basic executor. This results in getting async functionality in your main thread without the need for an async runtime. You will also be able to gain fine-grained control over when and in what order your coroutines get polled.

Note

At the time of this writing, we are using the coroutine syntax in nightly Rust. The syntax might have changed, or the coroutine syntax might have made its way to stable Rust. Although changing syntax is annoying, ...