Improve performance of filter_dict

[viirya]

Which issue does this PR close?

Closes #2062.

Rationale for this change

filter context string dictionary (kept 1/2)                                                                                                                                                                                            
                        time:   [21.995 us 22.197 us 22.408 us]                                                                                                                                                                        
                        change: [-2.3630% -1.2909% -0.2122%] (p = 0.02 < 0.05)                                                                                                                                                         
                        Change within noise threshold.                                                                                                                                                                                 
Found 8 outliers among 100 measurements (8.00%)                                                                                                                                                                                        
  8 (8.00%) high mild                                                                                                                                                                                                                  
                                                                                                                                                                                                                                       
filter context string dictionary high selectivity (kept 1023/1024)                                                                                                                                                                     
                        time:   [10.567 us 10.650 us 10.735 us]                                                                                                                                                                        
                        change: [-3.2419% -2.2248% -1.2276%] (p = 0.00 < 0.05)
                        Performance has improved.

filter context string dictionary low selectivity (kept 1/1024)                                                                              
                        time:   [910.91 ns 917.49 ns 924.83 ns]
                        change: [-23.778% -22.819% -21.986%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 2 outliers among 100 measurements (2.00%)
  1 (1.00%) high mild
  1 (1.00%) high severe

filter context string dictionary w NULLs (kept 1/2)                                                                             
                        time:   [53.386 us 53.870 us 54.349 us]
                        change: [-5.8932% -4.8921% -3.8850%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 2 outliers among 100 measurements (2.00%)
  2 (2.00%) high mild

filter context string dictionary w NULLs high selectivity (kept 1023/1024)                                                                              
                        time:   [20.187 us 20.381 us 20.576 us]
                        change: [-2.2931% -1.2510% -0.2547%] (p = 0.02 < 0.05)
                        Change within noise threshold.
Found 6 outliers among 100 measurements (6.00%)
  5 (5.00%) high mild
  1 (1.00%) high severe

filter context string dictionary w NULLs low selectivity (kept 1/1024)                                                                              
                        time:   [1.2670 us 1.2742 us 1.2816 us]
                        change: [-24.367% -23.820% -23.263%] (p = 0.00 < 0.05)
                        Performance has improved.
Found 4 outliers among 100 measurements (4.00%)
  4 (4.00%) high mild

What changes are included in this PR?

Are there any user-facing changes?

[viirya]

cc @sunchao

[sunchao]

LGTM

[codecov-commenter]

Codecov Report

Merging #2063 (3983af9) into master (742a590) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master    #2063   +/-   ##
=======================================
  Coverage   83.55%   83.56%           
=======================================
  Files         222      222           
  Lines       58216    58224    +8     
=======================================
+ Hits        48645    48657   +12     
+ Misses       9571     9567    -4     

Impacted Files	Coverage Δ
arrow/src/array/array_dictionary.rs	92.09% <100.00%> (+0.03%)	⬆️
arrow/src/compute/kernels/filter.rs	88.08% <100.00%> (+0.03%)	⬆️
...row/src/array/builder/string_dictionary_builder.rs	90.64% <0.00%> (-0.72%)	⬇️
arrow/src/ffi.rs	87.52% <0.00%> (+0.04%)	⬆️
parquet/src/column/writer.rs	91.53% <0.00%> (+0.05%)	⬆️
parquet_derive/src/parquet_field.rs	66.21% <0.00%> (+0.45%)	⬆️
arrow/src/datatypes/ffi.rs	76.56% <0.00%> (+1.56%)	⬆️

---

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 742a590...3983af9. Read the comment docs.

[viirya]

Thanks @sunchao

[ursabot]

Benchmark runs are scheduled for baseline = 88e0de5 and contender = 50e285f. 50e285f is a master commit associated with this PR. Results will be available as each benchmark for each run completes.
Conbench compare runs links:
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ec2-t3-xlarge-us-east-2] ec2-t3-xlarge-us-east-2
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on test-mac-arm] test-mac-arm
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ursa-i9-9960x] ursa-i9-9960x
[Skipped ⚠️ Benchmarking of arrow-rs-commits is not supported on ursa-thinkcentre-m75q] ursa-thinkcentre-m75q
Buildkite builds:
Supported benchmarks:
ec2-t3-xlarge-us-east-2: Supported benchmark langs: Python, R. Runs only benchmarks with cloud = True
test-mac-arm: Supported benchmark langs: C++, Python, R
ursa-i9-9960x: Supported benchmark langs: Python, R, JavaScript
ursa-thinkcentre-m75q: Supported benchmark langs: C++, Java