New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
doc/issue-76 #81
doc/issue-76 #81
Conversation
Bumps [pytest-asyncio](https://github.com/pytest-dev/pytest-asyncio) from 0.10.0 to 0.11.0. - [Release notes](https://github.com/pytest-dev/pytest-asyncio/releases) - [Commits](pytest-dev/pytest-asyncio@v0.10.0...v0.11.0) Signed-off-by: dependabot-preview[bot] <support@dependabot.com>
Can you add a test-case? If you think it's possible, in the meantime, I'll check if it is still "working" with 3.6/3.7 versions of python. |
closes #76 |
I was planning to write a test case when I started looking at it but noticed there were no test cases for any of that file. It's a bit more of an undertaking to write one for that change and I don't think I'm the best equipped person to do it, not having enough low level asyncio experience to really write effective tests. |
It's okay @daveoconnor, no problem, I'll do it later then. |
Great! Yes, this is solving my bug on Python 3.8, based on aiohttp-3.6.2 and tartiflette-1.2.0. |
Any idea when the next version might be out? |
Soon:tm: Can you try:
Instead of what you've proposed? I don't undestand why we should create a task on anext, cancel it and then await it with supress(cancellederror) |
That's why I need time to understand the problem. Sadly i'm swamped at work. |
The reason I did it with a task was because according to the thread I linked to earlier, if I'm reading it correctly, there may be an issue with just calling aclose(). https://bugs.python.org/issue38559#msg355163. Double checking the results and how it impacts this bug, aclose() under the suppress seems to work too. It's been a while now but if I remember correctly I'd tried that at the time and was seeing subscriptions continuing to run, but I can't reproduce that now. |
#76 (comment) -> Okay, I haven't had any time to check on this. I'll belive y'll, i'll take some time to merge and deliver a new version sunday evening CEST. Thanks |
Pull request for connection _unsubscribe implementation based on tasks rather than trying to directly close the async generator directly, to fix changes made in python 3.8.
Information to do it this way gotten from: https://bugs.python.org/issue38559
This is way outside of my wheelhouse so if someone with asyncio experience can take a look and make sure this is the correct way to proceed that'd be good.
Thanks