Remove ArrowSignedNumericType to Simplify and reduce code duplication in arithmetic kernels

[jhorstmann]

Which issue does this PR close?

Closes #1160.

Rationale for this change

Reduces some code duplication and hopefully makes the whole file a bit easier to understand.

What changes are included in this PR?

• unified most of the logic for the simd division / modulo kernels
• removed ArrowSignedNumericType which was only introduced for a single kernel
• simplified generic type bounds to require the needed numeric ops

Are there any user-facing changes?

The removal of ArrowSignedNumericType is a breaking change but I don't think there were any usages outside of the arrow crate or even the arithmetic.rs file.

cc @alamb @viirya

[viirya]

The comment needs to be updated too?

[codecov-commenter]

Codecov Report

Merging #1161 (60e9479) into master (884c6a6) will increase coverage by 0.04%.
The diff coverage is 69.23%.

@@            Coverage Diff             @@
##           master    #1161      +/-   ##
==========================================
+ Coverage   82.55%   82.59%   +0.04%     
==========================================
  Files         173      173              
  Lines       50673    50612      -61     
==========================================
- Hits        41833    41804      -29     
+ Misses       8840     8808      -32     

Impacted Files	Coverage Δ
arrow/src/compute/kernels/arithmetic.rs	90.19% <69.23%> (+5.54%)	⬆️
arrow/src/array/transform/mod.rs	85.43% <0.00%> (-0.27%)	⬇️
parquet_derive/src/parquet_field.rs	66.21% <0.00%> (+0.22%)	⬆️
arrow/src/datatypes/datatype.rs	66.80% <0.00%> (+0.42%)	⬆️

---

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 884c6a6...60e9479. Read the comment docs.

[alamb]

This looks great @jhorstmann -- I am sorry I ran out of time today to review Arrow PRs -- i will give this one a careful look tomorrow

[alamb]

Marking as an API change due to ArrowSignedNumericType though I agree that it is not likely to matter in practice

One datapoint: the only place I can find it seems to be in a copy of the arrow-rs codebase:

https://sourcegraph.com/search?q=context:global+ArrowSignedNumericType&patternType=literal

Likewise github shows results only from forks / copies of Arrow as well: https://github.com/search?q=ArrowSignedNumericType&type=code

[alamb]

This looks a lot simpler to me. 👍 very nice -- thank you for the contribution @jhorstmann ❤️

FWIW we are going to try releasing arrow 8.0.0 next week from master next week (as it contains some other breaking changes)