Refactor code to replace hydra/active_fedora functionality with rSolr and ActiveRecord

[candyhazlett]

• Dev task: https://trello.com/c/TPefK5uH
  • Refactor methods that used hydra/activefedora (AF) to use rSolr and bulit-in ActiveRecord interface directly instead, with the goal of (loosely) following the app->Solr process/architecture of the CRRJ-BNDA app to the extent possible.
    • [x] Identify specific classes and methods that used AF
    • https://github.com/NEU-DSG/tapas_rails/issues/85#issuecomment-1958000130
    • https://github.com/NEU-DSG/tapas_rails/issues/85#issuecomment-1958000784
    • [ ] Outline functionality AF handled previously and how
    • [x] review how SolrService are used:
    • [x] https://github.com/NEU-DSG/tapas_rails/issues/85#issuecomment-1958001528
    • [x] Write/refactor code to use rSolr commands for:
    • [x] Connection to Solr core
    • [x] Write (or reimplement) to_solr methods to return hash/json object in the following Classes for Solr index/document CRUD functions:
      • [x] collections
      • [x] communities
      • [x] core files
    • [ ] Update Solr config files to include fields outlined here

[candyhazlett]

useful links/images/etc:

• first pass at removing hydra/activefedora
  • commit #2cafc8
• diagram of hydra architecture

[candyhazlett]

files where hydra/activefedora is used or mentioned:

• lib/cerberus_core/lib/cerberus_core/solr_document_behavior.rb
• lib/cerberus_core/lib/cerberus_core/base_models/core_file.rb
• lib/cerberus_core/lib/cerberus_core/base_models/content_file.rb
• lib/cerberus_core/lib/cerberus_core/base_models/community.rb
• lib/cerberus_core/lib/cerberus_core/base_models/collection.rb
• lib/cerberus_core/lib/cerberus_core.rb
• config/solr.yml: this should be deleted
• config/environments/staging.rb: this one needs more research; its presence implies the Rails app previously persisted data to Fedora even though the app is configured to use MySQL, Solr, and an XML db.
• config/environments/production.rb
• config/environments/development.rb
• config/predicate_mappings.yml

[candyhazlett]

files I refactored previously where activefedora was commented out and/or now use rSolr in place of activefedora, and need review:

• app/jobs/index_job.rb
• app/validators/core_file_validator.rb
• app/models/solr_document.rb
• app/models/concerns/tapas_queries.rb
• app/models/concerns/og_reference.rb

[candyhazlett]

files that don't reference hydra/activefedora explicitly, but need review:

• lib/cerberus_core/lib/cerberus_core/concerns/traversals.rb
  • As of 2/21/24, the QueryServices class does not exist in CerberusCore
• app/views/catalog/_index_default.html.haml
  • This index view partial has two possible Solr queries to retrieve associated entities (collection and community) when one or both if conditions are true; they both use the Blacklight SolrService for the queries so maybe this is the best approach, curious about how efficient it is... No longer applicable, as this view is not rendered as part of the user flow for viewing catalog index
• app/helpers/controller_helper.rb
• app/controllers/catalog_controller.rb
  • exclude_unwanted_models method needs review to decide how to implement

[candyhazlett]

Todo next: list any important notes, thoughts, etc. for future dev work;

• checking current solr queries for CRUD functions, do the fields used in post requests match? are there queries still needed?
• create to_solr methods to return solr hash for indexing records

[candyhazlett]

This was closed by accident, issue still open and in development.