Cleanup Partition Nullability

[viceroypenguin]

This PR adds to #803. It uses nullable TKey values in the implementation to more accurately express the code.

[codecov]

Codecov Report

Merging #881 (2c8b409) into master (c3a7094) will increase coverage by 0.20%.
The diff coverage is 100.00%.

❗ Current head 2c8b409 differs from pull request most recent head 26b9656. Consider uploading reports for the commit 26b9656 to get more accurate results

@@            Coverage Diff             @@
##           master     #881      +/-   ##
==========================================
+ Coverage   92.38%   92.58%   +0.20%     
==========================================
  Files         110      108       -2     
  Lines        3441     3440       -1     
  Branches     1020     1024       +4     
==========================================
+ Hits         3179     3185       +6     
+ Misses        200      194       -6     
+ Partials       62       61       -1     

Impacted Files	Coverage Δ
MoreLinq/Partition.cs	100.00% <100.00%> (+1.75%)	⬆️
MoreLinq/Pad.cs	100.00% <0.00%> (ø)
MoreLinq/FallbackIfEmpty.cs	88.67% <0.00%> (ø)
MoreLinq/Assume.cs
MoreLinq/Debug.cs
MoreLinq/Lookup.cs	64.12% <0.00%> (+2.25%)	⬆️

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[atifaziz]

Like PR #882, this make unrelated changes (doc and nullability revisions). Please submit one overall idea/feature per PR.

[atifaziz]

Thanks, but the bulk of changes seem to be unrelated (and I've highlighted a few):

• Formatting changes
• More eager argument validation
• Annotation of some arguments with parameter names.

uses nullable TKey values in the implementation to more accurately express the code.

This is the only change that should be in here, which is towards the end of Partition.cs.

I would avoid the temptation to bundle unrelated changes as it has the following downsides:

• The diff/patch is larger than it needs to be.
• As a result of the above point, it takes longer to review the PR.
• The PR subject line or the commit message won't reflect the actual change (lost in the noise).
• The final commit won't be atomic so, e.g., if someone wants to revert it in the future, it will undo more work than necessary.

[atifaziz]

Thank you!