[codex] Promote specialized attributes of included typed datasets

[ehennestad]

Summary

This PR exposes attributes that are added at the inclusion site of a typed dataset on the containing MATLAB class API using the dataset_attribute naming pattern.

Why

The NWB schema allows a typed dataset reused via neurodata_type_inc to gain additional attributes when it is included into another type. matnwb was not surfacing those specialized attributes consistently in the generated API. This showed up most clearly for Units.spike_times.resolution.

What changed

• added support for promoting inclusion-site attributes from typed datasets onto the containing generated class
• made the promotion rule schema-driven: promote only attributes added on the included dataset instance, not attributes that already belong to the included dataset type and are merely refined there
• kept read/export behavior aligned with the promoted properties

User impact

Users can access specialized included-dataset attributes directly from the containing type, e.g. spike_times_resolution, without leaking those attributes onto unrelated VectorData instances.

Root cause

The generator handled untyped dataset attributes as dependent properties on the containing class, but typed reused datasets did not have an equivalent schema-driven path. A temporary hidden-property-based heuristic also turned out to be too tied to legacy VectorData patching.

Validation

• regenerated core classes for schema 2.9.0
• ran tests.system.UnitTimesIOTest.testRoundTrip
• verified that Units exposes spike_times_resolution
• verified that SimultaneousRecordingsTable does not expose recordings_table
• checked support for typed unnamed included datasets; no additional generated API changes were triggered by the current core schema

[codecov]

Codecov Report

❌ Patch coverage is 92.13483% with 7 lines in your changes missing coverage. Please review.
✅ Project coverage is 95.45%. Comparing base (df83711) to head (8ae7459).

Files with missing lines	Patch %	Lines
+file/processClass.m	85.10%	7 Missing ⚠️

Additional details and impacted files

@@                      Coverage Diff                      @@
##           refactor-io-parse-dataset     #797      +/-   ##
=============================================================
- Coverage                      95.50%   95.45%   -0.05%     
=============================================================
  Files                            191      191              
  Lines                           6961     7047      +86     
=============================================================
+ Hits                            6648     6727      +79     
- Misses                           313      320       +7     

☔ 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.