Refactor requests and responses to use classvar defaults to avoid multiple __init__s

[bdraco]

I was looking at the benchmarks and was wondering why 3.x was faster creating many objects. This PR re-aligns master to be similar to how these are implemented in 3.x as it was before #3942

Refactor mixins to avoid having to call __init__ by using classvar defaults instead.

The downside is we have to drop __slots__ but the trade-off seems worth it based on the benchmark.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.73%. Comparing base (006fbc3) to head (d609099).
Report is 1 commits behind head on master.

✅ All tests successful. No failed tests found.

Additional details and impacted files

@@            Coverage Diff             @@
##           master   #10037      +/-   ##
==========================================
- Coverage   98.73%   98.73%   -0.01%     
==========================================
  Files         121      121              
  Lines       36771    36750      -21     
  Branches     4392     4391       -1     
==========================================
- Hits        36307    36286      -21     
  Misses        314      314              
  Partials      150      150              

Flag	Coverage Δ
CI-GHA	98.62% <100.00%> (-0.01%)	⬇️
OS-Linux	98.31% <100.00%> (-0.01%)	⬇️
OS-Windows	96.14% <100.00%> (-0.01%)	⬇️
OS-macOS	97.41% <100.00%> (-0.01%)	⬇️
Py-3.10.11	97.26% <100.00%> (-0.01%)	⬇️
Py-3.10.15	97.85% <100.00%> (+0.05%)	⬆️
Py-3.11.10	97.84% <100.00%> (-0.01%)	⬇️
Py-3.11.9	97.31% <100.00%> (-0.01%)	⬇️
Py-3.12.7	98.37% <100.00%> (-0.01%)	⬇️
Py-3.13.0	98.37% <100.00%> (-0.01%)	⬇️
Py-3.9.13	97.18% <100.00%> (-0.01%)	⬇️
Py-3.9.20	97.72% <100.00%> (-0.05%)	⬇️
Py-pypy7.3.16	97.32% <100.00%> (-0.01%)	⬇️
VM-macos	97.41% <100.00%> (-0.01%)	⬇️
VM-ubuntu	98.31% <100.00%> (-0.01%)	⬇️
VM-windows	96.14% <100.00%> (-0.01%)	⬇️

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 #10037 will improve performances by 95.13%

_{Comparing mixin_inits (d609099) with master (006fbc3)}

Summary

⚡ 5 improvements
✅ 39 untouched benchmarks

Benchmarks breakdown

	Benchmark	master	mixin_inits	Change
⚡	test_simple_web_response	671.6 µs	448.4 µs	+49.78%
⚡	test_simple_web_stream_response	483.7 µs	247.9 µs	+95.13%
⚡	test_web_response_with_bytes_body	764.4 µs	541 µs	+41.29%
⚡	test_web_response_with_headers	1,108.2 µs	886.6 µs	+24.99%
⚡	test_web_response_with_text_body	883 µs	661.5 µs	+33.47%

[bdraco]

Wasn't expecting that kind of improvement. I would say the trade off is worth it especially considering we can do it in a few more places

[asvetlov]

I'm neutral to this change.
Such microoptimizations are not visible on macrobenchmarks.
Also, next Python version could change the picture.

[bdraco]

Looks to be ~2.7% speed up in testing

before

after

[asvetlov]

Wow! 2.5% is a high number, I didn't expect this.

[bdraco]

Part of the issue with benchmarks that run coros (not a problem for non coros) is that we have to use loop.run_until_complete which means we benchmark all the asyncio overhead and anything running in the loop (we get both the server and client side) so the benchmark only tends to pickup very significant changes as codspeed is also set to only report changes of 8%+ or more. Which for a client side benchmark is really 8% of asyncio overhead, server request, and the actual client code.

aiohttp/tests/test_benchmarks_client.py

Line 33 in 006fbc3

loop.run_until_complete(run_client_benchmark())

[patchback]

Backport to 3.11: 💔 cherry-picking failed — conflicts found

❌ Failed to cleanly apply 2e369db on top of patchback/backports/3.11/2e369db2d9abcbef41031592b4f044adf5f89f59/pr-10037

Backporting merged PR #10037 into master

1. Ensure you have a local repo clone of your fork. Unless you cloned it
   from the upstream, this would be your origin remote.
2. Make sure you have an upstream repo added as a remote too. In these
   instructions you'll refer to it by the name upstream. If you don't
   have it, here's how you can add it:

   $ git remote add upstream https://github.com/aio-libs/aiohttp.git

3. Ensure you have the latest copy of upstream and prepare a branch
   that will hold the backported code:

   $ git fetch upstream
   $ git checkout -b patchback/backports/3.11/2e369db2d9abcbef41031592b4f044adf5f89f59/pr-10037 upstream/3.11

4. Now, cherry-pick PR Refactor requests and responses to use classvar defaults to avoid multiple __init__s #10037 contents into that branch:

   $ git cherry-pick -x 2e369db2d9abcbef41031592b4f044adf5f89f59

   If it'll yell at you with something like fatal: Commit 2e369db2d9abcbef41031592b4f044adf5f89f59 is a merge but no -m option was given., add -m 1 as follows instead:

   $ git cherry-pick -m1 -x 2e369db2d9abcbef41031592b4f044adf5f89f59

5. At this point, you'll probably encounter some merge conflicts. You must
   resolve them in to preserve the patch from PR Refactor requests and responses to use classvar defaults to avoid multiple __init__s #10037 as close to the
   original as possible.
6. Push this branch to your fork on GitHub:

   $ git push origin patchback/backports/3.11/2e369db2d9abcbef41031592b4f044adf5f89f59/pr-10037

7. Create a PR, ensure that the CI is green. If it's not — update it so that
   the tests and any other checks pass. This is it!
   Now relax and wait for the maintainers to process your pull request
   when they have some cycles to do reviews. Don't worry — they'll tell you if
   any improvements are necessary when the time comes!

🤖 @patchback
I'm built with octomachinery and
my source is open — https://github.com/sanitizers/patchback-github-app.

[patchback]

Backport to 3.12: 💔 cherry-picking failed — conflicts found

❌ Failed to cleanly apply 2e369db on top of patchback/backports/3.12/2e369db2d9abcbef41031592b4f044adf5f89f59/pr-10037

Backporting merged PR #10037 into master

1. Ensure you have a local repo clone of your fork. Unless you cloned it
   from the upstream, this would be your origin remote.
2. Make sure you have an upstream repo added as a remote too. In these
   instructions you'll refer to it by the name upstream. If you don't
   have it, here's how you can add it:

   $ git remote add upstream https://github.com/aio-libs/aiohttp.git

3. Ensure you have the latest copy of upstream and prepare a branch
   that will hold the backported code:

   $ git fetch upstream
   $ git checkout -b patchback/backports/3.12/2e369db2d9abcbef41031592b4f044adf5f89f59/pr-10037 upstream/3.12

4. Now, cherry-pick PR Refactor requests and responses to use classvar defaults to avoid multiple __init__s #10037 contents into that branch:

   $ git cherry-pick -x 2e369db2d9abcbef41031592b4f044adf5f89f59

   If it'll yell at you with something like fatal: Commit 2e369db2d9abcbef41031592b4f044adf5f89f59 is a merge but no -m option was given., add -m 1 as follows instead:

   $ git cherry-pick -m1 -x 2e369db2d9abcbef41031592b4f044adf5f89f59

5. At this point, you'll probably encounter some merge conflicts. You must
   resolve them in to preserve the patch from PR Refactor requests and responses to use classvar defaults to avoid multiple __init__s #10037 as close to the
   original as possible.
6. Push this branch to your fork on GitHub:

   $ git push origin patchback/backports/3.12/2e369db2d9abcbef41031592b4f044adf5f89f59/pr-10037

7. Create a PR, ensure that the CI is green. If it's not — update it so that
   the tests and any other checks pass. This is it!
   Now relax and wait for the maintainers to process your pull request
   when they have some cycles to do reviews. Don't worry — they'll tell you if
   any improvements are necessary when the time comes!

🤖 @patchback
I'm built with octomachinery and
my source is open — https://github.com/sanitizers/patchback-github-app.