Fix #775: handle empty datasets in get_type with informative error

[sejalpunwatkar]

This PR addresses issue #775 where validation of datasets with empty shapes caused an IndexError.

Changes:

• Updated get_type in validator.py to check for empty/null data before indexing.
• Replaced the generic IndexError with a more informative EmptyArrayError.
• Added a unit test in tests/unit/test_validator.py to verify the fix and prevent regressions.

Fixes #775

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 92.76%. Comparing base (0897b93) to head (42aa600).

Additional details and impacted files

@@            Coverage Diff             @@
##              dev    #1421      +/-   ##
==========================================
- Coverage   92.85%   92.76%   -0.09%     
==========================================
  Files          41       41              
  Lines        9989     9993       +4     
  Branches     2054     2056       +2     
==========================================
- Hits         9275     9270       -5     
- Misses        436      441       +5     
- Partials      278      282       +4     

☔ View full report in Codecov by Sentry.
📢 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.

[sejalpunwatkar]

Hi @rly , I've updated the PR to fix #775. I added the safety check for empty datasets in get_type and included a corresponding unit test.
Note on file changes: You will notice a large number of line changes (1018 insertions, 845 deletions) in the modified files. This is because I ran ruff format to resolve the linting failures in CI and align these specific files with the project's updated style standards. I also simplified the str and bytes checks in get_type to maintain the required complexity score (C901).
I see that some environment-specific tests are still failing; I am looking into those logs now and will provide an update shortly.

[rly]

The optional and zarr tests are expected to fail until the next pynwb release, so those are OK.

The ruff / style changes are significant and should be moved to a separate PR so that this one is easy to review and self-contained. However, I don't think ruff should be complaining about changing single quote to double quote if you are running it from the repo, which should use the configuration specified in pyproject.toml.

Could you please limit this PR to just the changes relevant to the fix?

[sejalpunwatkar]

@rly, I've cleaned up the PR. It now contains only the fix for #775 and the new test (16 lines total). I've avoided the bulk formatting changes to keep this PR self-contained. The Zarr and optional test failures are the ones you mentioned are expected to fail.