Database Migration Changelog¶
- ???? - The backend added a 'promote' permission to override default
visibility for user-created content. The
scripts/ruby/create_promote_perm.rbscript added this new permission's node to the DB.
- 03/01/2014 - The backend code was altered to make the (3-letter)
language code the per-item description discriminator (along with the
descriptionIdentifierfield) for descriptions, meaning their IDs could be generated deterministically instead of randomly as before. The
scripts/ruby/migrate_description_ids.rbupdated the existing descriptions accordingly.
- 03/01/2014 - The script 'scripts/ruby/migrate_639-1_639-2_langs.rb` updated all language code fields from their 2-letter to 3-letter representations.
- 03/12/2013 - The
scripts/ruby/set_country_repo_child_count.rbwas used to cache the number of 'child' relations a node has on its _CHILD_COUNT meta property. An update to the code henceforth maintains this count on hierarchical relationships.
- 03/12/2013 - Debugging the global and per-item event chains proved
difficult because there were several 'link' used that had no
properties and only served to connect two 'first-class' nodes via an
indirection. The script
scripts/ruby/stamp_link_nodes.rbset debugging properties
_linkTypeon these nodes.
- 20/10/2013 - The initial import of EAD data from ICA-Atom had missed
the archivist's note field. The
scripts/ruby/fix_icaatom_import.rbscript corrected this by setting this data on the graph nodes from a custom CSV dump of the ICA-Atom MySQL DB.
- 20/10/2013 - Fixed an issue (on the frontend) where saving/updating a
repository would erase any maintenance event nodes it was connected
to. The script
scripts/ruby/fix_maintenance_events.rbrestored that maintenance events that were erroneously deleted.
- 10/10/2013 - Fixed an issue where some repositories were imported
without the correct (country) permission scope. This led to their
documents having incorrect IDs. The script
scripts/ruby/reset_doc_ids.rbwas used to set the proper permission scope and recalculate IDs for all items in the hierarchy (country -> repo -> doc)