[MRG] update Index protocol tests to include tests for peek and consume

[ctb]

Test, document, rationalize, and generalize the CounterGather API used for 'gather' optimization.

The main work of this PR is to move tests specific to the CounterGather implementation in src/sourmash/index/__init__.py over to more generic protocol tests that test the API rather than the implementation. As part of this, the PR implements an additional CounterGather compatible class, based on LinearIndex.

The ultimate goals are to -

• nail down the CounterGather API a bit more;
• provide options for alternative implementations that provide specific optimizations;

This PR does the following:

• changes the peek and add methods to use named parameters with a *;
• moves the CounterGather tests from test_index.py into test_index_protocol;
• builds a CounterGather-compatible class wrapper for LinearIndex, for testing purposes;
• moves some common threshold calculation code from the CounterGather class over to search.py;
• adds a new test, test_counter_gather_exact_match;

Fixes #1960

Notes for review

The tests that moved from test_index.py to test_index_protocol.py are unchanged functionally; it's not well represented in the diff, tho, because they moved files.

TODO:

• experiment with adding multiple identical matches, what happens?

[codecov]

Codecov Report

Merging #2111 (c6078a6) into latest (f34bd17) will increase coverage by 7.38%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##           latest    #2111      +/-   ##
==========================================
+ Coverage   84.30%   91.69%   +7.38%     
==========================================
  Files         130       99      -31     
  Lines       15280    11002    -4278     
  Branches     2171     2171              
==========================================
- Hits        12882    10088    -2794     
+ Misses       2095      611    -1484     
  Partials      303      303              

Flag	Coverage Δ
python	91.69% <100.00%> (-0.01%)	⬇️
rust	?

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
src/sourmash/index/__init__.py	96.77% <100.00%> (-0.06%)	⬇️
src/sourmash/search.py	97.94% <100.00%> (+0.03%)	⬆️
src/core/src/ffi/index/revindex.rs
src/core/src/ffi/storage.rs
src/core/src/index/search.rs
src/core/src/cmd.rs
src/core/src/index/mod.rs
src/core/src/errors.rs
src/core/src/sketch/minhash.rs
src/core/src/signature.rs
... and 23 more

---

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 f34bd17...c6078a6. Read the comment docs.

[ctb]

bump @mr-eyes

[ctb]

whoops, thought that I'd already asked for review 😊

@mr-eyes could you review this? it should all be functionality-neutral API changes / test refactoring.

[bluegenes]

just had the one q, otherwise looks good, i think!

[mr-eyes]

@mr-eyes could you review this? it should all be functionality-neutral API changes / test refactoring.

Sorry, it took some time.
Just some minor comments/questions. Otherwise, looks good to me.

[ctb]

thanks to both of you - I was definitely fixated on expanding the tests on this one, and not doing much with the core code, so it's nice that I paired it with #2116 which addresses your concerns 😆