perf: optimize next_day in datafusion-comet-spark-expr

[andygrove]

This PR was created by an LLM as a draft PR. I will mark it as ready for review after human review.

Which issue does this PR close?

N/A — autonomous exploratory PR.

Rationale for this change

Replace the per-row to_uppercase() heap allocation in day-of-week parsing with an allocation-free ASCII stack-buffer fast path (Unicode to_uppercase fallback only for non-ASCII), avoiding one allocation per row.

What changes are included in this PR?

Replace the per-row to_uppercase() heap allocation in day-of-week parsing with an allocation-free ASCII stack-buffer fast path (Unicode to_uppercase fallback only for non-ASCII), avoiding one allocation per row.

How are these changes tested?

Correctness: unit tests + seeded differential fuzz (bit-identical Arrow output vs main).

Benchmark (criterion):

• next_day_ scalar day-of-week: 19.459% faster (base 209122ns -> cand 168429ns)
• next_day_ array day-of-week: 28.982% faster (base 200938ns -> cand 142701ns)