[WIP] sparse: use staged cursor and upper_bound for WAND

[sparknack]

Use staged cursor and upper_bound to avoid recalculating the score from cursor 0 every time in the loop. And modify some variable names to be more readable.

Simple benchmark result(time used to search all queries once on the full dataset) with drop ratio build 0.32 and drop ratio search 0.6:

BM25	Before	After	Improvement %
bigann SPLADE vectors of MSMarco	81,980ms	66,205ms	19.2%
Milvus 2.5 Full Text Search Vectors of MSMarco	1,005ms	1,034ms	~

IP	Before	After	Improvement %
bigann SPLADE vectors of MSMarco	85,527ms	75,366ms	12%
Milvus 2.5 Full Text Search Vectors of MSMarco(using IP/TF-IDF metric though, not BM25)	2,839ms	3,191ms	-12%

This optimization seems to have an improvement on learned sparse embedding like SPLADE, which is denser, but leads to a little bit of performance degration on more sparse embeddings.

Hold the review of this PR for now until more investigation.

[sre-ci-robot]

[APPROVALNOTIFIER] This PR is NOT APPROVED

This pull-request has been approved by: sparknack
To complete the pull request process, please assign hhy3 after the PR has been reviewed.
You can assign the PR to them by writing /assign @hhy3 in a comment when ready.

The full list of commands accepted by this bot can be found here.

Needs approval from an approver in each of these files:

• OWNERS

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[sre-ci-robot]

Welcome @sparknack! It looks like this is your first PR to zilliztech/knowhere 🎉

[mergify]

@sparknack 🔍 Important: PR Classification Needed!

For efficient project management and a seamless review process, it's essential to classify your PR correctly. Here's how:

1. If you're fixing a bug, label it as kind/bug.
2. For small tweaks (less than 20 lines without altering any functionality), please use kind/improvement.
3. Significant changes that don't modify existing functionalities should be tagged as kind/enhancement.
4. Adjusting APIs or changing functionality? Go with kind/feature.

For any PR outside the kind/improvement category, ensure you link to the associated issue using the format: “issue: #”.

Thanks for your efforts and contribution to the community!.

[sparknack]

/kind improvement

[sparknack]

issue: #967

[zhengbuqian]

/hold

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 74.39%. Comparing base (3c46f4c) to head (f0d8927).
Report is 262 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff            @@
##           main     #970       +/-   ##
=========================================
+ Coverage      0   74.39%   +74.39%     
=========================================
  Files         0       82       +82     
  Lines         0     6690     +6690     
=========================================
+ Hits          0     4977     +4977     
- Misses        0     1713     +1713     

see 82 files with indirect coverage changes