Add UTF-8 BOM support for package.json files

[Boshen]

Summary

• Add UTF-8 BOM (Byte Order Mark) detection and preservation when sorting package.json files
• Optimized implementation to avoid performance penalties
• Comprehensive test coverage with dedicated BOM fixture

Motivation

Some tools and editors (like those used in Vite playground) create package.json files with UTF-8 BOM markers. When sorting these files, the BOM should be preserved to maintain compatibility with tools that expect it.

Reference: https://github.com/vitejs/vite/blob/main/playground/resolve/utf8-bom-package/package.json

Implementation

1. BOM Detection: Uses strip_prefix() to detect and remove BOM (\u{FEFF}) before JSON parsing
2. Preservation: If BOM was present in input, prepends it to the sorted output
3. Performance: Pre-allocates exact capacity and builds output string efficiently (avoids O(n) insert(0) operation)
4. No Side Effects: Files without BOM remain unchanged

Test Coverage

• BOM preservation when present
• No BOM added when not present
• Idempotency (sorting twice produces same result)
• New fixture: tests/fixtures/package-bom.json with actual UTF-8 BOM

Test Plan

• All existing tests pass
• New BOM test passes
• Benchmarks show no performance regression
• Manually verified with Vite example

🤖 Generated with Claude Code

[codspeed-hq]

Merging this PR will not alter performance

✅ 4 untouched benchmarks

---

_{Comparing feat/utf8-bom-support (81a848c) with main (89d649f)¹}

Footnotes

1. No successful run was found on main (6f66639) during the generation of this report, so 89d649f was used instead as the comparison base. There might be some changes unrelated to this pull request in this report. ↩