[datafusion-spark] Optimize hex function

[andygrove]

Is your feature request related to a problem or challenge?

PR #15947 adds a Spark-compatbile hex function. The feedback on the PR includes some suggested performance optimizations. This issue is for tracking implementing that feedback and adding criterion benchmarks.

Describe the solution you'd like

No response

Describe alternatives you've considered

No response

Additional context

No response

[ajita-asthana]

take

[manlikeHB]

Hi @Jefffrey, I'd like to attempt this

[manlikeHB]

take

[Phoenix500526]

Hi, @andygrove Looks like this issue has already been resolved. The two things the issue tracks both seem to have landed:

• perf optimization: perf(spark): use 256-entry byte-pair table in hex encoding #21836 added the 256-entry byte-pair lookup table in spark/src/function/math/hex.rs
• criterion benchmark: datafusion/spark/benches/hex.rs (added in perf: improve performance of spark hex function #19738)

Is there anything still outstanding, or is it good to close?

[shinzoxD]

I'll work on the remaining review item from #15947.

The lookup table (#21836) and criterion benches (#19738) already landed. What's still outstanding from the original review is returning Utf8View instead of Utf8, with an inline fast path for hex strings of at most 12 bytes (the StringView inline limit). I'll keep the existing null-buffer reuse and add coverage for the new physical type.