Metadata

Basic Principles

Resources are described and documented by metadata that is defined by work type schemas.
The data dictionary contains all possible fields that resources might have and define the field properties.
Some fields are designated as core fields, all work and collection resources have these fields, regardless of their work type.
Work type metadata schemas pull fields from the data dictionary and define customizations needed for different classes of resources.

During MVP, metadata is either editable (descriptive) or non-editable (administrative, technical, rights, etc.)
External controlled vocabularies will pull from source on regular basis and cached in CHO for MVP. (1.x will use a triplestore)
Local controlled vocabularies will be managed outside of CHO and imported/exported as CSV files. Term must be indexed for a resource to be created.
Metadata for works can be created and updated in batch with CSV spreadsheets. These requirements are in-flux and best documented with the batch file specifications.

If no titles are included (work or file set), the filename is used.
CSV headers must use the data dictionary label for field names.
- Note: alternate_ids = Identifier
batch_id is not a CHO field, but assigned by the Digital Production Team. Most batch_id values should end in YYYY-MM-DD.
repeating fields are entered in the same cell, delimited with a double pipe || -- no spaces
The creator, which is repeating, may also contain creator roles. Roles are delimited from their creator with a single pipe |.
- e.g. Doe, John|au||Doe, Jane|ill
Representative file sets are not explicitly included in CSV metadata or assigned an identifier
As of 2019-06, the access_rights field name cannot be changed without breaking the application. The application uses this field to control access to collections and resources via application roles. This strategy may change in future...