Archivematica

Rebuilding the Elasticsearch index needs to be updated for recent changes around the creation and storing of AIPs

• AIPs are stored through the storage service, and not all put in /var/archivematica/sharedDirectory/www/AIPsStore/ on the local filesystem
• AIPs can be a .7z or a .tar.bz2 - and potentially other packages, or even uncompressed, in the future

Additionally, the number of AIPs stored in an AIC (#5803) needs to be indexed. This is stored in the Dublin Core tag 'extent' in the AIC METS file, in the metadata/ directory.

• A directory is provided as a parameter, and all files in there are considered to be AIPs that need to be re-indexed (Possibly filtering on file extensions)
  Advantage: flexible, doesn't need to be updated for storage service changes
  Disadvantage: if AIPs are not already stored in one location (eg some are in LOCKSS), could be hard to get them all in one place

• Query the storage service for all AIPs it knows of, and index those
  Advantage: Uses existing system that is supposed to track this, handles files scattered across different storage locations
  Disadvantages: Relies on storage service knowing all the AIPs (would not handle storage service recreation), needs to be updated for SS API changes