Containers are missing fields after reading file without first importing package

[rly]

When an NWBFile containing a LabeledEvents or TTLs type is read using PyNWB without first calling import ndx_events, then it is missing the timestamps dataset which should be inherited from Events. Interestingly, the description attribute is correctly inherited from Events.

Example code:

from pynwb import NWBHDF5IO
# import ndx_events

with NWBHDF5IO('nwbfile.nwb', 'r', load_namespaces=True) as io:
    nwbfile = io.read()
    print(nwbfile.acquisition['LabeledEvents'])

Output without importing ndx_events (missing timestamps field):

LabeledEvents abc.LabeledEvents at 0x1745887195848
Fields:
  data: <HDF5 dataset "data": shape (3,), type "<u4">
  data__labels: ['a' 'b' 'c']
  description: behavioral events of the experimental paradigm

Output with importing ndx_events (contains timestamps field):

LabeledEvents ndx_events.events.LabeledEvents at 0x2565438502472
Fields:
  data: <HDF5 dataset "data": shape (3,), type "<u4">
  description: behavioral events of the experimental paradigm
  labels: ['a' 'b' 'c']
  timestamps: <HDF5 dataset "timestamps": shape (3,), type "<f4">
  unit: seconds

This is likely an issue of within-spec spec resolution when handling load_namespaces=True and has to be addressed within HDMF.

MWE to generate a file with `LabeledEvents`

from pynwb import NWBFile, NWBHDF5IO, validate
import datetime
from ndx_events import LabeledEvents


nwbfile = NWBFile(
    session_description='session_description',
    identifier='identifier',
    session_start_time=datetime.datetime.now(datetime.timezone.utc),
)

labeled_events = LabeledEvents(name='LabeledEvents',
                               description='behavioral events of the experimental paradigm',
                               timestamps=[1, 2, 3],
                               data=[2, 1, 0],
                               labels=['a', 'b', 'c'])
nwbfile.add_acquisition(labeled_events)

filename = 'nwbfile.nwb'

with NWBHDF5IO(filename, 'w') as io:
    io.write(nwbfile)

with NWBHDF5IO(filename, 'r') as io:
    errors = validate(io)
    print(errors)
    nwbfile = io.read()
    print(nwbfile)
    print(nwbfile.acquisition['LabeledEvents'])

[rly]

Thanks to @ttngu207 for the bug report!