Improved performance of utf8 comparison (1.7x-4x)

[jorgecarleitao]

utf8 2^20               time:   [10.376 ms 10.395 ms 10.424 ms]                      
                        change: [-43.339% -43.191% -43.033%] (p = 0.00 < 0.05)
utf8 scalar 2^20        time:   [1.5213 ms 1.5243 ms 1.5283 ms]                              
                        change: [-75.416% -75.313% -75.200%] (p = 0.00 < 0.05)

The idea is that we can iterate directly over values (&str) ignoring validities, which is evidently faster than iterating over Option<&str>

[codecov]

Codecov Report

Merging #322 (dfe5460) into main (824ad7e) will increase coverage by 0.02%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main     #322      +/-   ##
==========================================
+ Coverage   80.55%   80.58%   +0.02%     
==========================================
  Files         324      324              
  Lines       21399    21453      +54     
==========================================
+ Hits        17239    17288      +49     
- Misses       4160     4165       +5     

Impacted Files	Coverage Δ
src/util/bench_util.rs	0.00% <ø> (ø)
src/compute/comparison/utf8.rs	67.41% <100.00%> (-1.41%)	⬇️
tests/it/io/ipc/common.rs	100.00% <0.00%> (ø)
tests/it/io/ipc/write/stream.rs	100.00% <0.00%> (ø)
src/io/ipc/read/stream.rs	83.76% <0.00%> (+1.48%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 824ad7e...dfe5460. Read the comment docs.