File Sets

A file set is a group of files that represent the same things. For example, a TIFF, JP2, JPG, PDF, TXT, and Thumbnail JPG of the same scanned document all comprise one file set. Most works include many file sets.

There are two types of file sets: normal and representative.

Normal File Set

• Represents one component of a work (e.g. one page of a book)
• Requires a _preservation or _service file to be present
• Distinguished from other file sets with incremental segments appended to the local identifier

Representative File Set

• Represents an entire work (e.g. a book)
• Requires a _access file to be present
• Distinguished from other file sets with NO incremental segments

Notes

• A file set should:
  • include only one preservation file (though preservation_redacted may be included),
  • apart from sufixes, all filenames should match within a file set (workID_incrementals), and
  • may include a thumbnail.
• Descriptive, Administrative, Technical and Preservation metadata can be attached to any object (collection, work, fileset).
  • This includes access controls (rights).
  • In practice, files should rarely, if ever, have descriptive metadata attached. (PCDM does not strictly allow descriptive metadata at the file-level.)
• File sets can only belong to collection (representative imagery, project documentation) and work objects. File sets cannot belong to file sets.
  • Files can only belong to file sets.
• File sets are derived from file and directory names.
• Each collection and work will have a representative file set. This file set will not be designated in its own subdirectory but can be derived from the filenmame which will match the workID plus sufixes.
• See also Works.