buffer: speed up hex decoding

[anonrig]

Buffer.from(hex, 'hex') and buf.write(hex, 'hex') both go through StringBytes::Write and the scalar nbytes::HexDecode loop. Valid hex is the common case, so decode four output bytes per iteration and keep the existing “stop at the first invalid nibble” behavior (odd trailing nibble still ignored).

This change was implemented with assistance from a closed-source coding agent. I verified the source, official benchmark/buffers/buffer-hex-encode.js results, and the buffer hex tests listed below.

Official benches

benchmark/buffers/buffer-hex-encode.js (Buffer.from(hex, 'hex')), three runs each, median rate:

len	main	this PR	delta
64	9.34M	10.08M	+8%
1024	1.23M	1.54M	+25%

Tests

test-buffer-badhex, test-buffer-from, test-buffer-write, test-buffer-write-fast, test-buffer-tostring, test-buffer-tostring-range.

[codecov]

Codecov Report

❌ Patch coverage is 90.47619% with 4 lines in your changes missing coverage. Please review.
✅ Project coverage is 90.29%. Comparing base (dd5dfb5) to head (6004dda).
⚠️ Report is 38 commits behind head on main.

Files with missing lines	Patch %	Lines
src/string_bytes.cc	90.47%	0 Missing and 4 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #66174      +/-   ##
==========================================
+ Coverage   90.27%   90.29%   +0.02%     
==========================================
  Files         790      790              
  Lines      271651   272080     +429     
  Branches    51842    51941      +99     
==========================================
+ Hits       245228   245684     +456     
+ Misses      16928    16899      -29     
- Partials     9495     9497       +2     

Files with missing lines	Coverage Δ
src/string_bytes.cc	76.80% <90.47%> (+1.05%)	⬆️

... and 70 files with indirect coverage changes

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.