The OAIS describes a single abstract entity known as an archival information package (AIP), which bundles together inter alia, a byte stream representing original content, representation information by means of which that content information can be recreated, and metadata by which it can be identified, described, managed etc. In OAIS terminology, an AIP comprises content information (CI) and preservation description information (PDI); the former being subdivided into a byte stream and the information necessary to represent that stream, and the latter into four separate components (reference, context, provenance, and fixity).
For CEDARS purposes, it has been agreed that each AIP should have a unique identifier, known as its CEDARS Reference Identifier or CRID. This identifier is specified as a mandatory attribute of the <aip> element; it may also used by a local entity manager to provide the system identifier needed in order to access the underlying data object. The model assumed here also requires explicit specification of an entity name for that purpose (supplied as an attribute of the <content> element); in practice, it is likely that only one mechanism will be needed.
We define XML elements corresponding to each of these basic components as follows:
<!-- * *--> <ELEMENT aip (pdi, content)> <ATTLIST aip crid CDATA #REQUIRED> <ELEMENT pdi (reference, context, provenance, fixity) > <ELEMENT content (ri)> <ATTLIST content dataObject ENTITY #REQUIRED>