This is the overall architecture for the OCDX Repository. The OCDX is a repository is a
- Wikibase is used to create manifests, following the manifest laid out in libbyh/ocdx. A sparkle engine will be used to search for the collections ** Have wikibase be able to launch a notebook, including the full contents of the collection (defined below)
- JupyterHub is used, following a K8S architecure described in
- Each collection contains: ** A Manifest ** A pointer to a dataset (could be multiple files)
- *A Jupyter Hub Notebook Directory containing a set of notebooks