Optimise decompression size

[Dreamsorcerer]

Raise the compression max_length if we know the user is going to allow a size larger than the default anyway.
Also get rid of AsyncStreamReaderMixin, which just doesn't make any sense.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 98.92%. Comparing base (0c7ce34) to head (8e0877f).
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##           master   #12357   +/-   ##
=======================================
  Coverage   98.92%   98.92%           
=======================================
  Files         133      133           
  Lines       46550    46567   +17     
  Branches     2423     2427    +4     
=======================================
+ Hits        46048    46065   +17     
  Misses        373      373           
  Partials      129      129           

Flag	Coverage Δ
CI-GHA	98.98% <100.00%> (-0.01%)	⬇️
OS-Linux	98.72% <100.00%> (+<0.01%)	⬆️
OS-Windows	96.98% <90.62%> (-0.01%)	⬇️
OS-macOS	97.87% <90.62%> (-0.01%)	⬇️
Py-3.10.11	97.38% <90.62%> (-0.01%)	⬇️
Py-3.10.20	97.86% <90.62%> (+<0.01%)	⬆️
Py-3.11.15	98.10% <90.62%> (+<0.01%)	⬆️
Py-3.11.9	97.64% <90.62%> (-0.01%)	⬇️
Py-3.12.10	97.73% <90.62%> (-0.01%)	⬇️
Py-3.12.13	98.20% <90.62%> (+<0.01%)	⬆️
Py-3.13.12	98.44% <90.62%> (-0.01%)	⬇️
Py-3.14.3	98.50% <90.62%> (-0.01%)	⬇️
Py-3.14.4t	97.51% <90.62%> (-0.01%)	⬇️
Py-pypy3.11.15-7.3.21	97.34% <90.62%> (-0.01%)	⬇️
VM-macos	97.87% <90.62%> (-0.01%)	⬇️
VM-ubuntu	98.72% <100.00%> (+<0.01%)	⬆️
VM-windows	96.98% <90.62%> (-0.01%)	⬇️
cython-coverage	38.22% <75.00%> (+<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]

Merging this PR will improve performance by 55.88%

⚡ 2 improved benchmarks
✅ 59 untouched benchmarks
⏩ 4 skipped benchmarks¹

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
⚡	test_get_request_with_251308_compressed_chunked_payload[zlib_ng.zlib_ng-pyloop]	244 ms	215.5 ms	+13.2%
⚡	test_get_request_with_251308_compressed_chunked_payload[isal.isal_zlib-pyloop]	113.3 ms	72.7 ms	+55.88%

---

_{Comparing optimise-iter-chunked (8e0877f) with master (0c7ce34)}

Footnotes

1. 4 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[Dreamsorcerer]

There's a bit of awkwardness and I'm not sure that's the best approach around .read(). Basically brotlicffi with max_length=0 produces empty bytes and with max_length=sys.maxsize hits a MemoryError as it tries to pre-allocate an entire array of that size. So need to work around that.

[Dreamsorcerer]

stream here has a limit of 1, so after reading 3 bytes there is still 1 left. Previously that would avoid resuming as it didn't exceed the limit. Now the limit gets implicitly raised to 3, so a resume is triggered. This seems like a good idea to me; if the user asked for 3 bytes, there's a good chance they'll want another 3, so might as well buffer them up.