Avoid generating an unused Date header fallback

[marcus-campbell]

What do these changes do?

headers.setdefault(hdrs.DATE, rfc822_formatted_time()) evaluates the formatter even when the response already has a Date header, so the clock read (and possible cache refresh) happens and the result is discarded. This switches to a presence check so the formatter only runs when its value is used, and adds a regression test plus a small _prepare_headers() benchmark with and without an explicit Date.

Locally, a paired A/B run showed ~0.45 µs (~2.4%) per response on the explicit-Date path and no measurable change on the generated-Date path.

Follow-up to #9898: that benchmark only constructed Response objects and never reached _prepare_headers(), which is why CodSpeed showed no change.

Are there changes in behavior for the user?

No. A user-supplied Date is still retained.

Is it a substantial burden for the maintainers to support this?

No.

Related issue number

N/A (see #9898 above)

Checklist

• I think the code is well written
• Unit tests for the changes exist
• Documentation reflects the changes (no user-facing documentation change is
  required for this internal performance cleanup)
• If you provide code modification, please add yourself to CONTRIBUTORS.txt
• Add a new news fragment into the CHANGES/ folder

AI assistance

Drafted with help from GPT-5.6 Sol (benchmarking, parts of this description); code by me. I've reviewed all of the code and remain responsible for the submission.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 99.01%. Comparing base (394959f) to head (0993f17).
✅ All tests successful. No failed tests found.

Additional details and impacted files

@@           Coverage Diff           @@
##           master   #13299   +/-   ##
=======================================
  Coverage   99.01%   99.01%           
=======================================
  Files         135      135           
  Lines       50026    50045   +19     
  Branches     2611     2614    +3     
=======================================
+ Hits        49531    49550   +19     
  Misses        371      371           
  Partials      124      124           

Flag	Coverage Δ
Autobahn	22.02% <25.00%> (-0.01%)	⬇️
CI-GHA	98.90% <100.00%> (-0.01%)	⬇️
OS-Linux	98.67% <100.00%> (+<0.01%)	⬆️
OS-Windows	97.00% <100.00%> (+<0.01%)	⬆️
OS-macOS	97.91% <100.00%> (-0.02%)	⬇️
Py-3.10	98.09% <100.00%> (-0.01%)	⬇️
Py-3.11	98.33% <100.00%> (-0.01%)	⬇️
Py-3.12	98.41% <100.00%> (-0.01%)	⬇️
Py-3.13	98.40% <100.00%> (-0.01%)	⬇️
Py-3.14	98.44% <100.00%> (+<0.01%)	⬆️
Py-3.14t	97.51% <100.00%> (-0.01%)	⬇️
Py-pypy-3.11	97.39% <100.00%> (+0.01%)	⬆️
VM-macos	97.91% <100.00%> (-0.02%)	⬇️
VM-ubuntu	98.67% <100.00%> (+<0.01%)	⬆️
VM-windows	97.00% <100.00%> (+<0.01%)	⬆️
cython-coverage	82.90% <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 Harness.
📢 Have feedback on the report? Share it here.

[codspeed-hq]

Merging this PR will degrade performance by 1%

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 1 improved benchmark
❌ 1 (👁 1) regressed benchmark
✅ 90 untouched benchmarks
🆕 2 new benchmarks
⏩ 83 skipped benchmarks¹

Performance Changes

	Benchmark	BASE	HEAD	Efficiency
⚡	test_list_element_pattern_redos_payload[whitespace_run_after_content]	33.9 ms	30.9 ms	+9.91%
🆕	test_prepare_web_response_headers[explicit-date]	N/A	7.6 ms	N/A
🆕	test_prepare_web_response_headers[generated-date]	N/A	8.1 ms	N/A
👁	test_read_base64_part	14.3 ms	16 ms	-10.83%

Tip

Curious why this is faster? Comment @codspeedbot explain why this is faster on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing marcus-campbell:perf-avoid-unused-date-fallback (0993f17) with master (394959f)}

Footnotes

1. 83 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. ↩

[marcus-campbell]

A note on the CodSpeed report above: "will not alter performance" is expected here, because the two benchmarks that exercise this change (test_prepare_web_response_headers[explicit-date] and [generated-date]) are new in this PR. There's no BASE measurement to compare against, so they show N/A rather than an improvement. The 84 untouched benchmarks don't cover _prepare_headers(), which is the gap that also affected #9898's experiment.

The evidence for the change is the local paired A/B run described in the PR body.

[greptile-apps]

Confidence Score: 5/5

The PR appears safe to merge.

No blocking failure remains.

_{Reviews (3): Last reviewed commit: "Merge branch 'master' into perf-avoid-un..." | Re-trigger Greptile}

[patchback]

Backport to 3.15: 💚 backport PR created

✅ Backport PR branch: patchback/backports/3.15/5fa85842951ad68b2cba0bbc511dad1eca90cd3a/pr-13299

Backported as #13543

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

[patchback]

Backport to 3.14: 💚 backport PR created

✅ Backport PR branch: patchback/backports/3.14/5fa85842951ad68b2cba0bbc511dad1eca90cd3a/pr-13299

Backported as #13544

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