nl: preserve raw bytes in output instead of using from_utf8_lossy

[ChrisDryden]

This bug sent me caused me a bunch of difficulty while working on the locale GNU integration tests, since non utf-8 values were being piped into nl in the GNU integration tests and then coming out as UTF-8 breaking the tests. I am curious to see if this was the root cause for some more integration tests failing in the CI.

Its a bit rough to see that there were integration tests specifically made for this scenario but the tests were implemented with the same "to_string_lossy" which made the tests actually check nothing. Ideally we should add the "matches_gnu()" to all of these integration tests: #9660 so that we can detect these issues preemptively.

[codspeed-hq]

CodSpeed Performance Report

Merging #9673 will improve performances by 75.86%

_{Comparing ChrisDryden:fix-nl-preserve-bytes (93c8d54) with main (ae3c51f)}

Summary

⚡ 2 improvements
✅ 125 untouched
⏩ 6 skipped¹

Benchmarks breakdown

	Benchmark	BASE	HEAD	Change
⚡	nl_large_file[10]	99.9 ms	57.1 ms	+74.82%
⚡	nl_many_lines[100000]	79 ms	44.9 ms	+75.86%

Footnotes

1. 6 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. ↩

[ChrisDryden]

Nice, didn't realize this was also the bottleneck in performance

[github-actions]

GNU testsuite comparison:

Skip an intermittent issue tests/tail/overlay-headers (fails in this run but passes in the 'main' branch)

[cakebaker]

Thanks!