Fix incorrect Group.nmembers for consolidated metadata

[csparker247]

Addresses the issue where Group.nmembers() doesn't honor the max_depth when the group has consolidated metadata. This simply filters flattened_metadata according to the number of / in the sub-elements' paths, which is analogous to checking the depth.

I'm new to this project, so I'm not sure if the implemented test is sufficient.

Closes #3252

TODO:

• Add unit tests and/or doctests in docstrings
• Add docstrings and API docs for any new/modified user-facing classes and functions
• New/modified features documented in docs/user-guide/*.rst
• Changes documented as a new file in changes/
• GitHub Actions have all passed
• Test coverage is 100% (Codecov passes)

[TomAugspurger]

Looks good thanks. Could you add a release note (https://zarr.readthedocs.io/en/stable/developers/contributing.html#changelog)?

Can you also ensure that we have test coverage for the case where we raise a ValueError?

with pytest.raises(ValueError, match="max_depth"):
    await group.nmembers(max_depth=-1)

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 59.67%. Comparing base (b877f89) to head (9df9683).
Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #3287      +/-   ##
==========================================
+ Coverage   59.65%   59.67%   +0.01%     
==========================================
  Files          78       78              
  Lines        8690     8694       +4     
==========================================
+ Hits         5184     5188       +4     
  Misses       3506     3506              

Files with missing lines	Coverage Δ
src/zarr/core/group.py	69.48% <100.00%> (+0.14%)	⬆️

🚀 New features to boost your workflow:

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

[csparker247]

Done. Let me know if there's anything else.

[TomAugspurger]

Thanks!