Simplified min_max_string and min_max_binary

[HaoYang670]

Signed-off-by: remzi 13716567376yh@gmail.com

Changes in the PR

1. Replace fold with reduce
2. use macro to abstract the common expression
3. Add test to cover the no null cases.
4. Add benchmark for min utf8
5. Less code is better
6. Functional style

No obvious performance impact

min 2^10 utf8           time:   [33.766 ns 33.786 ns 33.810 ns]                           
                        change: [+0.0167% +0.2385% +0.4298%] (p = 0.02 < 0.05)
                        Change within noise threshold.
Found 7 outliers among 100 measurements (7.00%)
  3 (3.00%) high mild
  4 (4.00%) high severe

min null 2^10 utf8      time:   [33.352 ns 33.370 ns 33.391 ns]                                
                        change: [-4.7561% -4.6289% -4.5317%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 8 outliers among 100 measurements (8.00%)
  4 (4.00%) high mild
  4 (4.00%) high severe

[codecov]

Codecov Report

Merging #1004 (c2332f1) into main (a10db3a) will increase coverage by 0.02%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #1004      +/-   ##
==========================================
+ Coverage   71.42%   71.45%   +0.02%     
==========================================
  Files         356      356              
  Lines       19784    19752      -32     
==========================================
- Hits        14131    14113      -18     
+ Misses       5653     5639      -14     

Impacted Files	Coverage Δ
src/compute/aggregate/min_max.rs	74.24% <100.00%> (+3.51%)	⬆️

---

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 a10db3a...c2332f1. Read the comment docs.

[HaoYang670]

I set the length of each string to be 1 to make the time of comparing each element to be determined.

[jorgecarleitao]

Thanks a lot for this PR.

I went through this and also ran the benches on main vs this PR with different string lengths and array lengths and the performance is equivalent on my machine.

Given that this is less code and more tested, we ship it :)