shuf: don't truncate -o output when input or random source fails

[sylvestre]

No description provided.

[codspeed-hq]

Merging this PR will not alter performance

✅ 323 untouched benchmarks
⏩ 46 skipped benchmarks¹

---

_{Comparing sylvestre:shuf (94dfa6b) with main (4c35af0)}

Footnotes

1. 46 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. ↩

[github-actions]

GNU testsuite comparison:

Skip an intermittent issue tests/date/date-locale-hour (fails in this run but passes in the 'main' branch)
Skip an intermittent issue tests/tail/retry (fails in this run but passes in the 'main' branch)
Skip an intermittent issue tests/tail/symlink (fails in this run but passes in the 'main' branch)
Skipping an intermittent issue tests/date/resolution (passes in this run but fails in the 'main' branch)
Note: The gnu test tests/dd/no-allocate is now being skipped but was previously passing.

[sylvestre]

he CodSpeed regression wasn't extra work - it was a lost inlining. Reproduced with cachegrind: the open_output closure plus three separate output-creation sites bloated uumain past the inliner's threshold, so shuf_exec got emitted out-of-line instead of inlined into uumain. Instruction/cache-miss counts were otherwise identical. Moved open_output to a plain helper and open the output once before the match (still after reading the input, so -o isn't truncated on failure)