#async #async-await #await #future #macro #api-bindings

macro async-attributes

Experimental language-level polyfills for Async Rust

6 releases (stable)

1.1.2 Jan 23, 2021
1.1.1 Nov 7, 2019
1.0.0 Sep 18, 2019
0.3.0-alpha.5 Sep 18, 2019
0.1.0 Jul 1, 2019

#1658 in Asynchronous

Download history 114158/week @ 2024-07-29 117111/week @ 2024-08-05 123902/week @ 2024-08-12 116953/week @ 2024-08-19 142256/week @ 2024-08-26 134256/week @ 2024-09-02 126594/week @ 2024-09-09 120955/week @ 2024-09-16 139790/week @ 2024-09-23 134518/week @ 2024-09-30 152270/week @ 2024-10-07 148348/week @ 2024-10-14 153282/week @ 2024-10-21 137715/week @ 2024-10-28 163975/week @ 2024-11-04 159315/week @ 2024-11-11

622,527 downloads per month
Used in 1,323 crates (9 directly)

MIT/Apache

12KB
91 lines

async-attributes

crates.io version build status downloads docs.rs docs

Experimental language-level polyfills for Async Rust.

Examples

#[async_attributes::main]
async fn main() {
    println!("Hello, world!");
}

About

Async Rust is a work in progress. The language has enabled us to do some fantastic things, but not everything is figured out yet. This crate exists to polyfill language-level support for async idioms before they can be part of the language.

A great example of this is async fn main, which we first introduced as part of the runtime crate. Its premise is that if async fn is required for every await call, it makes sense to apply that even to fn main. Unfortunately this would require compiler support to enable, so we've provided an experimental polyfill for it in the mean time.

Why isn't this crate part of async-std?

We want to make sure async-std's surface area is stable, and only includes things that would make sense to be part of "an async version of std". Language level support is really important, but not part of the standard library.

This has some distinct benefits: in particular it allows us to version both crates at a different pace. And as features are added to the language (or we decide they weren't a great idea after all), we can incrementally shrink the surface area of this crate.

The other big benefit is that it allows libraries to depend on async-std without needing to pull in the rather heavy syn, quote, and proc-macro2 crates. This should help keep compilation times snappy for everyone.

Installation

$ cargo add async-attributes

Safety

This crate uses #![deny(unsafe_code)] to ensure everything is implemented in 100% Safe Rust.

Contributing

Want to join us? Check out our "Contributing" guide and take a look at some of these issues:

References

License

MIT OR Apache-2.0

Dependencies

~1.5MB
~37K SLoC