Account for all loaded namespaces during validation

[rly]

Description

Validation on cached schema currently validates against only the most specific namespaces. For example, if one were to draw a dependency tree:

hdmf-common
   |       \
 core    hdmf-experimental
   |  \
ndx-a  ndx-b

Then validation would occur on the leaves of the tree: ndx-a, ndx-b, and hdmf-experimental. Data types are found in one namespace and not another (e.g., NWBFile is not in hdmf-experimental. NewTypeA that is defined in ndx-a is not in ndx-b.) and this causes validation to fail.

There are also other cross-namespace linking and inheritance considerations to account for. For example, if ndx-a defines a type with a dataset name: column; neurodata_type_inc: VectorData and ndx-b defines a type neurodata_type_def: MyVectorData; neurodata_type_inc: VectorData, then MyVectorData should be allowed for the 'column' dataset in ndx-a.

See also #525 and #542 (comment)

Environment

Python Executable: Conda 
Python Version: Python 3.8
Operating System: Windows
HDMF Version: dev

Checklist

• Have you ensured the bug was not already reported ?
• Have you included a brief and descriptive title?
• Have you included a clear description of the problem you are trying to solve?
• Have you included a minimal code snippet that reproduces the issue you are encountering?
• Have you checked our Contributing document?