Rewrite HED code for version 4.0.0 of hed-validator

[happy5214]

This new version of hed-validator (not yet released) changes the API significantly. This is an early preview of the changes. Tests do not pass with the current release, and they were not attempted with the development branch.

[happy5214]

The test failures are due to actual invalid HED in the examples, and not due to bad code. @VisLab will take care of the example failures.

[effigies]

To be clear, is the invalid HED newly invalid, or is it newly caught by the validator?

[VisLab]

Newly caught. Version 4.0.0 of the hed-validator is a full validator for the HED specification.

[happy5214]

bids-standard/bids-examples#486 should fix the remaining test failures.

[happy5214]

The above-linked PR fixed the remaining HED issues, except for one with eeg_ds003645s_hed_library on the no-network tests. This dataset requires the SCORE v1.0.0 library schema, which is not bundled with the validator and must be downloaded from GitHub. How do I disable that dataset for the no-network tests?

[VisLab]

The above-linked PR fixed the remaining HED issues, except for one with eeg_ds003645s_hed_library on the no-network tests. This dataset requires the SCORE v1.0.0 library schema, which is not bundled with the validator and must be downloaded from GitHub. How do I disable that dataset for the no-network tests?

Alternatively, we could modify that example to use libraries that are bundled. This would cause the test to pass and would test library schemas, but it would not test whether a schema not bundled could be accessed. Let us know how we should proceed.
@effigies @rwblair

[effigies]

How do I disable that dataset for the no-network tests?

Had to look it up myself and it was as easy to fix as to point you to. Pushed dcebc91, if you want to see where.

[happy5214]

Do you want to merge this now, or wait for a 4.0.2 release with the fixes we're discussing in #157?

[VisLab]

@effigies Please merge now. 4.0.1 fixes an edge case bug that is needed for some of our datasets.

[effigies]

This makes sense to me. @rwblair Do you want to review (or merge if you trust me)?