Any thoughts on an async style @lru_cache?

[goodboy]

I can't get my head around the complexity in the equivalent for asyncio:
https://github.com/aio-libs/async-lru/blob/master/async_lru/__init__.py

But we have a small naive implementation (that should be async framework agnostic) that I could put here if anyone is interested?

[belm0]

That async_lru doesn't seem to care about parallel cache misses having identical args, which seems like a significant case for async.

In my project we, don't have a need for async LRU-- though we do have an @async_join decorator for simple side-effect functions (like an action), that will merge parallel calls, without regards to args.