refactor(lightningcss): replace parcel source map with rspack source map

[intellild]

Summary

Use the new generic Lightning CSS source map writer API so Rspack can generate rspack_sources::SourceMap directly instead of going through parcel_sourcemap.

This updates the workspace lightningcss dependency to the git revision 61aee3776df2563e7c0d2982fcf1e7d3ec51571b from intellild/lightningcss, removes direct parcel_sourcemap usage from the loader and minimizer, and keeps source map output wired through local RspackSourceMapWriter adapters.

Validation

• cargo check -p rspack_loader_lightningcss -p rspack_plugin_lightning_css_minimizer
• cargo tree -i parcel_sourcemap returns package not found
• cargo fmt --check -p rspack_loader_lightningcss -p rspack_plugin_lightning_css_minimizer
• pnpm exec rstest Config.part1.test.js -t source-map-absolute-sources
• pnpm exec rstest Config.part3.test.js -t verify-bundle-css-minify-lightning

---

by OpenAI Codex

[github-actions]

📦 Binary Size-limit

Comparing 4169ecd to docs: v2.1 release blog (#14527) by Gengkun

🎉 Size decreased by 392.00KB from 67.82MB to 67.44MB (⬇️0.56%)

[codspeed-hq]

Merging this PR will not alter performance

⚠️ 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 regressed benchmark
✅ 39 untouched benchmarks
⏩ 47 skipped benchmarks¹

Warning

Please fix the performance issues or acknowledge them on CodSpeed.

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
❌	Simulation	rust@runtime_requirements	10.4 ms	10.6 ms	-2.37%
⚡	Simulation	rust@create_module_hashes	8.6 ms	8.5 ms	+2.15%

Tip

Investigate this regression by commenting @codspeedbot fix this regression on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing codex/lightningcss-sourcemap-writer (4169ecd) with main (42f39a8)}

Footnotes

1. 47 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. ↩