Add AnyIO CapacityLimiter docs

[aminalaee]

Related to #1724

Summary

Checklist

• I understand that this PR may be closed in case there was no previous discussion. (This doesn't apply to typos!)
• I've added a test for each change that was introduced, and I tried as much as possible to make a single atomic change.
• I've updated the documentation accordingly.

[aminalaee]

I think it's not that easy to find this in the docs, if you don't know that you should look for the threadpool docs.

[adriangb]

Agreed but can't think of a better name right now. @Kludex any ideas?

[Kludex]

Concurrency?

[Kludex]

Concurrency.

[agronholm]

"Sync" threadpools, as opposed to what?

[adriangb]

Suggested change

	# Set number of threads in threadpool to 100
	anyio.to_thread.current_default_thread_limiter().total_tokens = 100
	# Set number of threads in threadpool to 100
	# Note that this increases the limit for anything using AnyIO, not just Starlette
	anyio.to_thread.current_default_thread_limiter().total_tokens = 100

[adriangb]

Devil's advocate: if we document this would it then be breaking change if we start using our own capacity limiter(s), meaning this stops working? I think not, it's okay to break docs to some extent.

[aminalaee]

We can update the docs when we add the CapacityLimiter to explain this.

[adriangb]

Agreed but can't think of a better name right now. @Kludex any ideas?

[Kludex]

Doesn't this needs to be in the lifespan if using multiple workers?

[aminalaee]

I think each worker will do this, so there’s no difference.

but I can do a quick test to confirm

[agronholm]

Each worker runs in a separate process, right?

[Kludex]

@aminalaee Do you still want to work on this?

[Kludex]

I'm closing this as stale. Please reopen if you still want to work on it, or if someone else wants to take it over, go ahead.