Is it good to add an arg "use_thread" in asyncify function?

[monchin]

First Check

• I added a very descriptive title here.
• I used the GitHub search to find a similar question and didn't find it.
• I searched the Asyncer documentation, with the integrated search.
• I already searched in Google "How to X in Asyncer" and didn't find any information.
• I already read and followed all the tutorial in the docs and didn't find an answer.
• I already checked if it is not related to Asyncer but to AnyIO.
• I already checked if it is not related to Asyncer but to Trio.

Commit to Help

• I commit to help with one of those options 👆

Example Code

Hi here.
I have read asyncer's doc about asyncify, it's really convenient. But the doc just said

asyncify() takes the sync (blocking) function that you want to call and then returns another async function that takes the actual arguments for the original sync function.
Then you can await that and get the return value that was safely executed in a "worker thread" without blocking the event loop.

If I didn't see the second paragraph above, I would believe that any sync function could be asyncified by asyncify. So I read the code of asyncify

    async def wrapper(
        *args: T_ParamSpec.args, **kwargs: T_ParamSpec.kwargs
    ) -> T_Retval:
        partial_f = functools.partial(function, *args, **kwargs)
        return await anyio.to_thread.run_sync(
            partial_f, cancellable=cancellable, limiter=limiter
        )

It uses anyio.to_thread.run_sync

Description

If the sync function is IO-intensive, asyncify would be great. But if the sync function is CPU-intensive, asyncify would not work because of GIL. For this situation, anyio offers to_process.
So my point is, if it is good to offer a bool arg use_thread in asyncify to make sure whether use to_thread or to_process? Or offer a new function asyncify_to_process? At lease I think now the doc should show that asyncify would be appropriate only when the sync function is IO-intensive.

Operating System

Linux

Operating System Details

No response

asyncer Version

0.0.7

Python Version

3.10.12

Additional Context

No response

[YuriiMotov]

Yes, currently docs don't cover the topic of performance optimization.

But the following statement is not correct

But if the sync function is CPU-intensive, asyncify would not work because of GIL.

It will work - your sync function is CPU-intensive will not block the event loop and this is what asyncify is intended to solve.

It might be a good idea to add a section in docs describing use cases that are better to solve with other approaches

[monchin]

Thank you, I misundeestood the mechanism of asyncify.