fix: avoid oob take for primitive

[a10y]

Fix potential out of bounds panic in the scalar Take kernel for PrimitiveArray.

We could definitely do this faster, e.g. by unrolling the loop 8x and dispatching based on the validity of the next 8 indices, but most production platforms will take the AVX2 kernel or the nightly Portable SIMD kernel anyway.

[codspeed-hq]

CodSpeed Performance Report

Merging #5535 will not alter performance

_{Comparing aduffy/fix-prim-scalar-take (7101064) with develop (a63d0ab)}

Summary

✅ 181 untouched
⏩ 1532 skipped¹

Footnotes

1. 1532 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. ↩