DynamicTable Extensions and Bugfixes

This release includes various bugfixes and performance fixes along with updated tutorial scripts.
Some (but not all) changes include:

• Updates to the Tutorials to use new DataStub features
• Fixes to the new DynamicTable getRow and addRow functions.
• Implement support for native MATLAB table objects when returned by getRow.

Various Ergonomic Improvements

DataStub is now indexable

• Instead of using DataStub.load(...) DataStubs now utilizes natural MATLAB indexing behavior: DataStub(...).
• DataStub.load() behavior is retained for backwards compatibility.
• DataStubs are now Sealed which means they cannot be subclassed.

DynamicTable objects

• new getRow() and addRow functions which allows for adding to a dynamic table by row.
• For more information, take a look at the convertTrials tutorial.

Miscellaneous

• Various bugfixes for DataStub bugs
• Basic Usage and Trial Conversion tutorials updated to use DynamicTable and DataStub methods.

Submodule Revision Hotfix

This release fixes a remnant submodule issue where hdmf-common for nwb schema version 2.2.5 was still somehow a submodule reference, which may've caused errors with generateCore.

Revert Subdirectories

This release reverts the change introduced by v2.2.5.1 wherein the nwb schema was stored as a set of submodules. This change was done due to the realization that some MatNWB tooling which do not depend on a valid git installation may break as that is the only way to download a git repository with submodules.

This change also fixes certain tests that may be indicative of changes to MATLAB's 2020b release.

Submoduled Schema Release

This release adds a new schema selection feature and alters generateCore, generateExtension, and nwbRead accordingly.

• Schemas are now submodules, so those who are using git should call git submodule update --init --recursive to be able generate the core schema again. Furthermore, those cloning with git should use git clone --recurse-submodules <url>.
• When nwbRead is called on a file without a schema, it will attempt to call generateCore with the embedded version string automatically. A schema error will only be thrown if the embedded schema does not exist.
• generateCore is now called with only one argument, which is a char array designating the desired schema version string (i.e. '2.2.5'). When called without arguments, the most recent nwb schema version is generated (right now, '2.2.5'). The old behavior where generateCore could be given multiple extension namespace paths has been moved to generateExtension.
• 'generateExtensioncan now take multiple arguments. This is a replacement for deprecating thegenerateCore` behavior.

Support Schema 2.2.5

Supports NeurodataWithoutBorders/nwb-schema@aa7702e

Support Schema 2.2.4

https://nwb-schema.readthedocs.io/en/latest/format_release_notes.html

Schema 2.2.3 was skipped as it was unusable without 2.2.4

Compatible with Schema 2.2.2

Updated schema

Include changes to schema version.

Compatible with Schema 2.2.1

Fixed DataPipe not setting Dimensions correctly

DataPipe originally did not account for an initial allocation case, causing
errors when appending.

This has been resolved.

Typecheck bugfix Release

Type correcting was broken in v0.2.0. This release fixes the issue.