Add an accumulator type to the Absorb trait

[rs-sac]

When applying several operations to the same value, certain operations may be faster to compute once at the end rather than incrementally after each change. E.g. if values are Vec<u64>, it would be faster to compute an average or to sort once, after all updates are applied, rather than incrementally after each update.

Add an associated type to the trait for the accumulator, pass it to the absorb methods, and call a new finalize method at the end of each absorb loop. If not needed, the accumulator can be set to () and the finalize method can be left with its default no-op implementation.

---

This change is 

[codecov]

Codecov Report

❌ Patch coverage is 83.33333% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 58.8%. Comparing base (a9758d4) to head (22adb1f).
⚠️ Report is 22 commits behind head on main.

Files with missing lines	Patch %	Lines
src/lib.rs	33.3%	2 Missing ⚠️

Additional details and impacted files

Files with missing lines	Coverage Δ
src/write.rs	71.9% <100.0%> (+2.0%)	⬆️
src/lib.rs	86.6% <33.3%> (+8.0%)	⬆️

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[jonhoo]

@rs-sac After we discussed this offline, do you still think you need this?

For anyone reading this, the quick summary was to change the Op type to instead be "batches" of operations, which would then allow you do this special handling directly in absorb, rather than needing to pass state between absorb calls.

[rs-sac]

We don't need this for now. The jury may still be a little out, but I implemented batching and merging outside of left-right.