url: skip plain prefixes in percent-escape checks

[rozzilla]

decodeFormComponent() already locates the first %, but hasPercentHex() scans from the beginning again. Pass that position to the helper so it skips the plain-text prefix. Replacing + with a space preserves the position; escape validation and decoding behavior are unchanged.

Constructor throughput on Apple M2/macOS, with rebuilt Node v27.0.0-pre binaries
against upstream a7a96c2448ab91ca079de5e64233aa79a400b00e:

Query value	Throughput improvement
32 ASCII characters followed by %20	+17.2%
1,024 ASCII characters followed by %20	+139.4%
1,024 ASCII characters followed by %E2%82%AC	+135.6%

No corrected-significant regressions were detected.

[nodejs-github-bot]

Review requested:

• @nodejs/performance
• @nodejs/url

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 90.47%. Comparing base (a7a96c2) to head (b3259a3).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #66696      +/-   ##
==========================================
- Coverage   90.47%   90.47%   -0.01%     
==========================================
  Files         791      791              
  Lines      277088   277088              
  Branches    53286    53281       -5     
==========================================
- Hits       250703   250696       -7     
+ Misses      16805    16803       -2     
- Partials     9580     9589       +9     

Files with missing lines	Coverage Δ
lib/internal/url.js	93.31% <100.00%> (ø)

... and 33 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.