Close unclosed client sessions properly

[hynek]

What do these changes do?

ClientSession.__del__ uses a bare self.close() which causes a warning. This fixes that warning. I think this should be fine like that?

Are there changes in behavior for the user?

One warning less.

Related issue number

n/a

Checklist

• I think the code is well written
• Unit tests for the changes exist
• Documentation reflects the changes
• If you provide code modification, please add yourself to CONTRIBUTORS.txt
  • The format is <Name> <Surname>.
  • Please keep alphabetical order, the file is sorted by names.
• Add a new news fragment into the CHANGES folder
  • name it <issue_id>.<type> for example (588.bug)
  • if you don't have an issue_id change it to the pr id after creating the pr
  • ensure type is one of the following:
    • .feature: Signifying a new feature.
    • .bugfix: Signifying a bug fix.
    • .doc: Signifying a documentation improvement.
    • .removal: Signifying a deprecation or removal of public API.
    • .misc: A ticket has been closed, but it is not of interest to users.
  • Make sure to use full sentences with correct case and punctuation, for example: "Fix issue with non-ascii contents in doctest text files."

[asvetlov]

It's a complex question.
__del__ is called on session destruction.
It could be even interpreter finalization stage if the session has global scope (we discourage this approach but still support it).

If the loop is closed before ClientSession.__del__ -- nothing to do.

On other hand the session could be dropped in coroutine context:

async def f():
    session = aiohttp.ClientSession()
    ...
    del session  # or just implicit decref on frame destruction

In this case the loop is running, loop.run_until_complete() is prohibited.

[hynek]

Yeah, I suspected it’s not clear cut.

I’d suggest to either drop the close altogether or wrap it in a with suppress(Exception): for an best-effort try. As it’s now it’s kind useless, isn’t it?

[asvetlov]

I think dropping is just fine.
Let's keep a warning report part but don't try to close the session: it is not very reliable.

[codecov-io]

Codecov Report

Merging #2523 into master will decrease coverage by <.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #2523      +/-   ##
==========================================
- Coverage   97.13%   97.13%   -0.01%     
==========================================
  Files          39       39              
  Lines        8066     8065       -1     
  Branches     1414     1414              
==========================================
- Hits         7835     7834       -1     
  Misses         99       99              
  Partials      132      132

Impacted Files	Coverage Δ
tests/autobahn/client.py	94.91% <0%> (-0.02%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 48a704d...d573a63. Read the comment docs.

[asvetlov]

Thanks!
Backported to 2.3, aiohttp 2.3.3 released

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a [new issue] for related bugs.
If you feel like there's important points made in this discussion, please include those exceprts into that [new issue].
[new issue]: https://github.com/aio-libs/aiohttp/issues/new