This year we’ve been working on building a graph-ready item tree model meant to accurately model all corners of deposited and inferred metadata in a schema-agnostic manner.
Much of the initial work involves a comprehensive review of item types, deciding item type definitions, and fixing extant discrepancies. The first fixes are now live in the REST API and are enumerated below. These changes should go unnoticed by most metadata consumers, and most will only be visible for new or updated works until the next full reindex.
standard-seriestype has been removed [CR-525] - Jira
- Database works now have type=database (previously it was “null”) [CR-527] - Jira
- Edited books now have type=edited-book (previously it was “book”) [CR-526] - Jira
- All reports now have type=report [CR-522] - Jira
- Report components now have type=report-component (previously it was “book-component”) [CR-524] - Jira