Bulkrax PR housekeeping

[rococodogs]

odds and ends that seemed better suited for a supplementary PR:

• cleans up Dockerfile and enable corepack (which in turns enables yarn) for all web targets
• bumps the Hyrax.class.noid_minter_class instance in development to prevent a flurry of failures on the first batch ingest
• installs experimental ldr-development-seeds gem to add seed items to the repository
• configures Bulkrax.default_field_mapping to split values on '\|'
• reorganized the Bulkrax field_mappings config to be organized by type (bibliographic vs date vs internal) and alphabetical for easier use
• Importers form: removes the input field to choose between upload a csv (now the only option) and choose one server-side (which won't work in our case)
• Importers form: in cases where only one Parser is available for Importers, select that one by default
• in Bulkrax field_mapping config, merge all defined field configurations with Bulkrax.default_field_mapping to ensure the defined defaults are applied
• removes previous Darlingtonia code used for CSV ingesting 🫡