Notes from discussion of Identifiers at DukeWorkshop

This information is out of date as of 21 June 2009. See now MapMaker and Aggregator.

Note: information assembled during this discussion will be used to augment/clarify IdentifierMapping

What are the identifiers in each dataset?

Primary keys (the main identifier for a record in each dataset)

• HGV primary key
  • Datensatz Nummer is an auto-assigned unique number but it is not stable, because FMPro reassigns upon every rebuild

• HGV identifier: PublikationL:
  • Example: P.Tebt. III.2 1047
  • Fieldname in Filemaker database: PublikationL
  • Where surfaced in HGV website: field PublikationL
  • Where in export XML from Filemaker? JC TO ADD
  • Where in unaggregated EpiDoc? JC TO ADD
  • Where in aggregated EpiDoc? JC TO ADD

• HGV Identifier: "hgv%s" % TexIDLang
  • Where in unaggregated EpiDoc: <TEI.2 id="hgv10007" (this is assembled by the mapping process from two other (different) fields in the HGV database: texID & texLett)

• HGV Reference Identifier (for DDB content):
  • <TEI.2 n="p.aberd.;;49" (inserted by mapping process)
    • p.aberd. is human-readable version of a numeric value used in HGV and documented in collection.xml and the conversion from numeric to string is done in mapping process; the strings however could be added to HGV filemaker, and then could be exported as same from there in lieu of the numeric values
  • this information derives from the DDBFull field in the HGV fmpro db

• HGV Reference Identifier (for Trismegistos content):
  • <bibl type="Trismegistos"><biblScope type="numbers">10007</biblScope></bibl> (adds a <biblScope type="parts">b</biblScope> if there's a many-to-one relationship between the Trismegistos record and multiple HGV records)
  • HGV db fieldname: texID

• HGV Reference information (for APIS content):
  • Has both in database and offline URLs etc for APIS content; however, these were not used in the IDP1 mapping process

• HGV Reference information (for Perseus versions of DDB content):
  • these are built from DDBFull; there are questions as to where and how these are getting constructed; maybe in various places in runner process, then do we care?
  • action items:
    • do we care about links to Perseus anymore?
    • are there any linking, mapping, tooling dependencies on Duke URLs?

• HGV Reference information (for print publications):
  • AnderePublikation (in HGV): exhaustive list of previous/superseded/deprecated publications of a given document (or portion thereof)

• DDB primary key
  • Example?
  • Where in unaggregated DDB EpiDoc? <TEI.2 n="0001;1;3">
  • Where in aggregated EpiDoc? <TEI.2 n="0001;1;3">

• APIS primary key
  • Example? institution.apis.thenumbertheinstitutionassigns
  • How does APIS data get to PN? Is this method stable? Harvested via the APIS OAI/PMH interface

• Does PN produce identifiers for its views, aggregations, individual components?
  • No

Secondary/Alternative identifiers

• HGV
• DDB
• APIS

Foreign keys (identifiers from other datasets) held in a given dataset

• HGV
• DDB
• APIS

URIs

• Which systems/resources provide stable URLs or URIs for discrete content items?
• Are these cataloged, surfaced, indexed anywhere?

Relevant pages elsewhere in the wiki

• ?