making content_length really optinal as intended

[realgam3]

What do these changes do?

Fix issue made on version 3.12.0
Where content_length argument in function write_bytes supposed to be optional but it's not because it's a positional argument and not keyword argument.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.76%. Comparing base (ebf65d9) to head (e577aaa).

Additional details and impacted files

@@           Coverage Diff           @@
##           master   #11325   +/-   ##
=======================================
  Coverage   98.76%   98.76%           
=======================================
  Files         129      129           
  Lines       43374    43374           
  Branches     2323     2323           
=======================================
  Hits        42837    42837           
  Misses        383      383           
  Partials      154      154           

Flag	Coverage Δ
CI-GHA	98.64% <ø> (ø)
OS-Linux	98.38% <ø> (ø)
OS-Windows	96.79% <ø> (-0.02%)	⬇️
OS-macOS	97.68% <ø> (ø)
Py-3.10.11	97.32% <ø> (-0.02%)	⬇️
Py-3.10.18	97.72% <ø> (-0.01%)	⬇️
Py-3.11.13	97.92% <ø> (ø)
Py-3.11.9	97.52% <ø> (ø)
Py-3.12.10	97.62% <ø> (-0.02%)	⬇️
Py-3.12.11	98.02% <ø> (ø)
Py-3.13.3	98.28% <ø> (ø)
Py-3.9.13	97.21% <ø> (ø)
Py-3.9.23	97.61% <ø> (-0.01%)	⬇️
Py-pypy7.3.16	87.36% <ø> (-4.70%)	⬇️
VM-macos	97.68% <ø> (ø)
VM-ubuntu	98.38% <ø> (ø)
VM-windows	96.79% <ø> (-0.02%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[codspeed-hq]

CodSpeed Performance Report

Merging #11325 will not alter performance

_{Comparing realgam3:fix-client_reqrep (e577aaa) with master (ebf65d9)}

Summary

✅ 59 untouched benchmarks

[Dreamsorcerer]

supposed to be optional

I'm not convinced this is correct. Is there a use case where you are calling this function in your code? I was under the impression that users would not access this method..

[realgam3]

supposed to be optional

I'm not convinced this is correct. Is there a use case where you are calling this function in your code? I was under the impression that users would not access this method..

I'm using it with my library:
https://github.com/realgam3/aiohttp-raw

[Dreamsorcerer]

Ah, subclasses, always forget about those..

My thinking is that it is beneficial to be a required parameter to ensure that people don't miss it. However, I guess it's also a breaking change to 3.x, so we should probably make it optional in the 3.12 branch atleast. For your project, you'll want to force the None at all times, otherwise it's likely to break your functionality.

@bdraco Does that make sense?

Also, interesting project, might be useful for some of our own testing when it comes to parsing issues, HTTP request forgery and such.

[bdraco]

@bdraco Does that make sense?

Yes. Seems fine to me