uasyncio Query

[pythoncoder]

Normal practice is to keep the event loop running forever, either by issuing run_forever() or by issuing run_until_complete(coro()) where coro() never terminates. Within that environment you can run synchronous functions and methods as well as asynchronous coroutines.

There are rare cases in testing where you issue run_until_complete(coro()) and you have coro() terminate, returning a value. In that case CPython's asyncio allows the value to be retrieved. uasyncio V2 does not. My fast_io version fixes this and other bugs.

I've never actually used that feature except to prove whether it works. My event loops invariably run forever.

[doublevee]

Thanks Peter - that explains the exact issue I was running into. I think I will follow the advice and move everything into an infinite loop (with a run_until_complete just during testing) and call both async and normal functions as and when required. I had it in my head for some reason that you couldn't mix these - which now seems daft - but I'm learning everyday

Do you know if your changes will be pulled into the main code?

[pythoncoder]

...
Do you know if your changes will be pulled into the main code?

They will not, because an entirely new version of uasyncio is imminent. This fixes the bugs addressed in my fork. I'm hoping it will also offer the fast I/O feature and the synchronisation primitives. I have no desire to maintain a version of uasyncio and only did so because bugfix PR's were not getting implemented. From what I've seen the new version is a big improvement.

[doublevee]

Understood.

I look forward to an updated uasyncio library but in the meantime, I now have the information I need to proceed. Thanks again all.

[RWLTOK]

How imminent is imminent? I love the module and am looking forward to the improvements.

Rich

...
Do you know if your changes will be pulled into the main code?

They will not, because an entirely new version of uasyncio is imminent. This fixes the bugs addressed in my fork. I'm hoping it will also offer the fast I/O feature and the synchronisation primitives. I have no desire to maintain a version of uasyncio and only did so because bugfix PR's were not getting implemented. From what I've seen the new version is a big improvement.

[pythoncoder]

I gather it's now @Damien's priority so hopefully soon.