Whose ID for What Entities?

IDs are needed for

  • the root-level document identifier
  • organizations (rights holders, contributors…)
  • users (archivists, burrito creators…)

Each ID identifies the user or organization within a server ecosystem. There are many possible ecosystems and several of these ecosystems may provide IDs within a single metadata document.

ID Syntax

This is surprisingly challenging since

  • we need to allow for a wide range of ids from diverse systems
  • we need to be able to distinguish prefixes from the start of an unqualified idServer

The syntax follows the Universal Resource Name scheme as documented in RFC2141

The syntax is

urn:<namespace ID>:<namespace-specific string>


  • “namespace ID” is case-insensitive and matches
  • “namespace-specific string” is case-sensitive and matches (as defined in RFC2141 §2.2, with rationale in RFC1737)
[0-9A-Za-z\(\)\,\-\.\+\*:[email protected];$_!']

IDs in this format can be tested for prefixedness (?!) by searching for urn:([a-zA-Z0-9\-]+):.*.