Accept zarr v3 arrays in array-valued fields; fix ElectricalSeries len()

[rly]

Motivation

Fixes #2234.

Reading NWB files through the zarr backend (hdmf-zarr's zarr v3 migration) passes lazily-read datasets into container constructors as raw zarr.Array objects. A zarr v3 Array implements neither __iter__ nor __len__, which broke pynwb in two places:

1. Fields declared with docval type=collections.abc.Iterable rejected zarr arrays, because isinstance(zarr_array, Iterable) is False (the ABC only recognizes __iter__; iteration otherwise works via the __getitem__ sequence protocol).
2. ElectricalSeries.__init__ called len(electrodes.data) directly, raising TypeError: object of type 'Array' has no len().

hdmf already accepts zarr v3 arrays in its own type system (it registers zarr.Array in the array_data docval macro), so both issues are pynwb-side.

Changes

• Retype array-valued fields from Iterable to ('array_data', 'data'): TimeSeries.control/control_description, ImageSeries.dimension/starting_frame, AbstractFeatureSeries.features/feature_units, and the deprecated Clustering.peak_over_rms, ClusterWaveforms.waveform_mean/waveform_sd (still constructed on read).
• Leave NWBFile.electrode_groups as Iterable: it holds ElectrodeGroup objects, not array data, and is never backed by a dataset.
• In ElectricalSeries.__init__, derive the electrode count via get_data_shape(...)[0] instead of len(electrodes.data), matching the pattern already used for data.
• Add tests/unit/test_array_data_fields.py guarding the retype (array accepted, non-array iterable rejected) and the electrode_groups exclusion.

Behavior change

('array_data', 'data') accepts zarr.Array but does not accept non-array iterables (str, set, range, generators) for these fields, which Iterable previously allowed. For these array-valued fields that is a desirable tightening. Noted in the changelog.

How tested

• Full unit + integration suite: 795 passed, 8 skipped, 6683 subtests passed.
• End to end against zarr v3: with hdmf-zarr's zarr v3 migration (no zarr.Array monkey-patches) and this branch, hdmf-zarr's zarr-v2-backward-compat read suite goes from 3 failed + 57 errors to 61 passed.

Checklist

• Did you update CHANGELOG.md with your changes?
• Did you add/update tests?

🤖 Generated with Claude Code

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 96.02%. Comparing base (eb08f14) to head (9ccb271).

Additional details and impacted files

@@            Coverage Diff             @@
##              dev    #2235      +/-   ##
==========================================
- Coverage   96.02%   96.02%   -0.01%     
==========================================
  Files          30       30              
  Lines        2997     2995       -2     
  Branches      436      436              
==========================================
- Hits         2878     2876       -2     
  Misses         67       67              
  Partials       52       52              

Flag	Coverage Δ
integration	74.55% <66.66%> (-0.02%)	⬇️
unit	86.17% <100.00%> (-0.01%)	⬇️

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

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

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