make find_by_source_identifier dynamic

Import a csv with child works. The forming of relationships is not working. Part of the problem is the find_by_bulkrax_identifier call.

From GBH, this used to be find_by_bulkrax_identifier which not all clients will configure as their source identifier. Instead we need to ask for the source identifier and use that for the sql query. This commit goes along with a PR from Hyku which currently has the find_by_source_identifier.rb files defined.

Issue:
- notch8/hykuup_knapsack#128

Co-Authored-By: Kirk Wang <[email protected]>

app/factories/bulkrax/object_factory.rb

@@ -28,10 +28,10 @@ class ObjectFactory # rubocop:disable Metrics/ClassLength
     class_attribute :transformation_removes_blank_hash_values, default: false
 
     define_model_callbacks :save, :create
-    attr_reader :attributes, :object, :source_identifier_value, :klass, :replace_files, :update_files, :work_identifier, :work_identifier_search_field, :related_parents_parsed_mapping, :importer_run_id
+    attr_reader :attributes, :object, :source_identifier_value, :klass, :replace_files, :update_files, :work_identifier, :work_identifier_search_field, :related_parents_parsed_mapping, :importer_run_id, :entry
 
     # rubocop:disable Metrics/ParameterLists
-    def initialize(attributes:, source_identifier_value:, work_identifier:, work_identifier_search_field:, related_parents_parsed_mapping: nil, replace_files: false, user: nil, klass: nil, importer_run_id: nil, update_files: false)
+    def initialize(attributes:, source_identifier_value:, work_identifier:, work_identifier_search_field:, related_parents_parsed_mapping: nil, replace_files: false, user: nil, klass: nil, importer_run_id: nil, update_files: false, entry:)
       @attributes = ActiveSupport::HashWithIndifferentAccess.new(attributes)
       @replace_files = replace_files
       @update_files = update_files
@@ -42,6 +42,7 @@ def initialize(attributes:, source_identifier_value:, work_identifier:, work_ide
       @source_identifier_value = source_identifier_value
       @klass = klass || Bulkrax.default_work_type.constantize
       @importer_run_id = importer_run_id
+      @entry = entry
     end
     # rubocop:enable Metrics/ParameterLists
 

app/factories/bulkrax/valkyrie_object_factory.rb

@@ -30,7 +30,10 @@ def search_by_identifier
       # Query can return partial matches (something6 matches both something6 and something68)
       # so we need to weed out any that are not the correct full match. But other items might be
       # in the multivalued field, so we have to go through them one at a time.
-      match = Hyrax.query_service.custom_queries.find_by_bulkrax_identifier(identifier: source_identifier_value)
+      match = Hyrax.query_service.custom_queries.find_by_source_identifier(
+        work_identifier: entry.work_identifier,
+        source_identifier_value: source_identifier_value
+      )
 
       return match if match
     rescue => err

app/models/concerns/bulkrax/import_behavior.rb

@@ -185,7 +185,8 @@ def factory
                           user: user,
                           klass: factory_class,
                           importer_run_id: importerexporter.last_run.id,
-                          update_files: update_files)
+                          update_files: update_files,
+                          entry: self)
     end
 
     def factory_class

app/parsers/bulkrax/application_parser.rb

@@ -68,15 +68,15 @@ def records(_opts = {})
     # @return [Symbol] the name of the identifying property in the source system from which we're
     # importing (e.g. is *not* this application that mounts *this* Bulkrax engine).
     #
-    # @see #work_identifier
+    # @see #source_identifier
     # @see https://github.com/samvera-labs/bulkrax/wiki/CSV-Importer#source-identifier Bulkrax Wiki regarding source identifier
     def source_identifier
       @source_identifier ||= get_field_mapping_hash_for('source_identifier')&.values&.first&.[]('from')&.first&.to_sym || :source_identifier
     end
 
     # @return [Symbol] the name of the identifying property for the system which we're importing
     #         into (e.g. the application that mounts *this* Bulkrax engine)
-    # @see #source_identifier
+    # @see #work_identifier
     def work_identifier
       @work_identifier ||= get_field_mapping_hash_for('source_identifier')&.keys&.first&.to_sym || :source
     end

app/services/hyrax/custom_queries/find_by_source_identifier.rb

@@ -0,0 +1,35 @@
+# frozen_string_literal: true
+
+module Hyrax
+  module CustomQueries
+    ##
+    # @see https://github.com/samvera/valkyrie/wiki/Queries#custom-queries
+    class FindBySourceIdentifier
+      def self.queries
+        [:find_by_source_identifier]
+      end
+
+      def initialize(query_service:)
+        @query_service = query_service
+      end
+
+      attr_reader :query_service
+      delegate :resource_factory, to: :query_service
+      delegate :orm_class, to: :resource_factory
+
+      ##
+      # @param identifier String
+      def find_by_source_identifier(work_identifier:, source_identifier_value:)
+        sql_query = sql_by_source_identifier
+        query_service.run_query(sql_query, work_identifier, source_identifier_value).first
+      end
+
+      def sql_by_source_identifier
+        <<-SQL
+          SELECT * FROM orm_resources
+          WHERE metadata -> ? ->> 0 = ?;
+        SQL
+      end
+    end
+  end
+end

app/services/wings/custom_queries/find_by_bulkrax_identifier.rb → app/services/wings/custom_queries/find_by_source_identifier.rb

@@ -1,13 +1,16 @@
 # frozen_string_literal: true
+
+# TODO: Make more dynamic. Possibly move to Bulkrax. 
+
 module Wings
   module CustomQueries
-    class FindByBulkraxIdentifier
+    class FindBySourceIdentifier
       # Custom query override specific to Wings
       # Use:
       #   Hyrax.custom_queries.find_bulkrax_id(identifier: identifier, models: [ModelClass])
 
       def self.queries
-        [:find_by_bulkrax_identifier]
+        [:find_by_source_identifier]
       end
 
       attr_reader :query_service
@@ -17,7 +20,7 @@ def initialize(query_service:)
         @query_service = query_service
       end
 
-      def find_by_bulkrax_identifier(identifier:, use_valkyrie: true)
+      def find_by_source_identifier(identifier:, use_valkyrie: true)
         af_object = ActiveFedora::Base.where("bulkrax_identifier_sim:#{identifier}").first
 
         return af_object unless use_valkyrie
@@ -26,4 +29,4 @@ def find_by_bulkrax_identifier(identifier:, use_valkyrie: true)
       end
     end
   end
-end
+end

spec/services/hyrax/custom_queries/find_by_source_identifier_spec.rb

@@ -0,0 +1,22 @@
+require 'spec_helper'
+
+RSpec.describe Hyrax::CustomQueries::FindBySourceIdentifier do
+  describe '.queries' do
+    subject { described_class.queries }
+    let(:query_name) { :find_by_source_identifier }
+
+    it { is_expected.to include(query_name) }
+
+    it 'is registered with the Hyrax.query_service' do
+      expect(Hyrax.query_service.custom_queries).to respond_to(query_name)
+    end
+
+    context ':find_by_source_identifier query' do
+      it 'is valid SQL' do
+        expect do
+          Hyrax.query_service.custom_queries.find_by_source_identifier(work_identifier: 'source', source_identifier_value: "testing-bulkrax-1-2-3")
+        end.not_to raise_error
+      end
+    end
+  end
+end