darlingtonia 3.0.5 → 3.1.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 6be5042f9ac88ade67585052eae9cc2dfa30855be0356bbee2f2895150c5d9a5
-  data.tar.gz: 0466ed8a2f5fd289c591ca230f987d294b3710cac3becc4a0a3194c6c73ea31f
+  metadata.gz: ee7ce1d03349bd1c624e7ad3bd3c12e92db474f77559444f15a36a34f444a8aa
+  data.tar.gz: bbfcdc4b7b0cd0ac1b9cfc83e30cb03d19d626e8664a05a7be762d7ea3e721ec
 SHA512:
-  metadata.gz: 18b96a5efae6c3f30c5965c88cd35e974db44e1abd155ab01fb19cdd1b53719d24c6ed4e0e29b45c4744f7534bd18fe18a7c674d4e9d1bc2cb92f07e7a200233
-  data.tar.gz: 7a3a556e44d1ca0552b84f3db46abe19ec35260c2c6ddfbd0fe370341748e927e1345162bac1e5aba939a9b78343d8a4b1b85d0b2929991729d36ba07e0d6553
+  metadata.gz: bc8ca923806a61c2b99c2c8bd9d97942464a539afb637c3f4206ce02faf397d9bd1e893f65c89153816eeee7bb5c55a0e9a25f59820462bbf6665606504cb3c0
+  data.tar.gz: 067ffa15cfccfb7f045e33f015f1e7dd3f07526214153e953aab1427723f9436a03e12ab1e927306c05004178f9e62260990d3ba639fd95ca761209a5e918163

data/.rubocop.yml

@@ -21,13 +21,16 @@ Metrics/BlockLength:
     - 'spec/**/*'
     - 'lib/darlingtonia/spec/**/*'
 
+Metrics/ClassLength:
+  Enabled: false
+
 Metrics/CyclomaticComplexity:
   Exclude:
     - lib/darlingtonia/hyrax_basic_metadata_mapper.rb
+    - lib/darlingtonia/hyrax_record_importer.rb
 
 Metrics/LineLength:
-  Exclude:
-    - 'lib/darlingtonia/hyrax_basic_metadata_mapper.rb'
+  Enabled: false
 
 Metrics/MethodLength:
   Exclude:
@@ -35,6 +38,10 @@ Metrics/MethodLength:
     - 'lib/darlingtonia/hyrax_basic_metadata_mapper.rb'
     - lib/darlingtonia/hyrax_record_importer.rb
 
+Metrics/PerceivedComplexity:
+  Exclude:
+    - lib/darlingtonia/hyrax_record_importer.rb
+
 Naming/AccessorMethodName:
   Exclude:
     - lib/darlingtonia/hyrax_record_importer.rb

data/CHANGELOG.md

@@ -1,6 +1,13 @@
+3.1.0 - Tue Feb 26, 2019
+
+New Feature: `HyraxRecordImporter` now accepts a `deduplication_field` in the
+attributes hash it receives  when it is created. If a `deduplication_field`
+is provided, the system will look for existing works with that field and matching
+value and will update the record instead of creating a new record.
+
 3.0.5 - Tue Feb 26, 2019
 
-When setting the depositor, query for user with `find_by_user_key`, which is the Hyrax convention. 
+When setting the depositor, query for user with `find_by_user_key`, which is the Hyrax convention.
 If user_key isn't found, fall back to querying by User.id for backward compatibility.
 
 3.0.4 - Fri Feb 22, 2019

data/README.md

@@ -35,8 +35,16 @@ class MyImporter
   end
 
   def import
+    attrs = {
+      collection_id: collection_id,     # pass a collection id to the record importer and all records will be added to that collection
+      depositor_id: depositor_id,       # pass a Hyrax user_key here and that Hyrax user will own all objects created during this import
+      deduplication_field: 'identifier' # pass a field with a persistent identifier (e.g., ARK) and it will check to see if a record with that identifier already     
+    }                                   # exists, update its metadata if so, and only if it doesn't find a record with that identifier will it make a new object.
+
     file = File.open(@csv_file)
-    Darlingtonia::Importer.new(parser: Darlingtonia::CsvParser.new(file: file), record_importer: Darlingtonia::HyraxRecordImporter.new).import
+    parser = Darlingtonia::CsvParser.new(file: file)
+    record_importer = Darlingtonia::HyraxRecordImporter.new(attributes: attrs)
+    Darlingtonia::Importer.new(parser: parser, record_importer: record_importer).import
     file.close # unless a block is passed to File.open, the file must be explicitly closed
   end
 end

data/lib/darlingtonia/hyrax_record_importer.rb

@@ -17,6 +17,11 @@ module Darlingtonia
     # @return [String] an id number associated with the process that kicked off this import run
     attr_accessor :batch_id
 
+    # @!attribute [rw] deduplication_field
+    # @return [String] if this is set, look for records with a match in this field
+    # and update the metadata instead of creating a new record. This will NOT re-import file attachments.
+    attr_accessor :deduplication_field
+
     # @!attribute [rw] success_count
     # @return [String] the number of records this importer has successfully created
     attr_accessor :success_count
@@ -30,15 +35,21 @@ module Darlingtonia
     #        the CSV/mapper. These are useful for logging
     #        and tracking the output of an import job for
     #        a given collection, user, or batch.
+    #        If a deduplication_field is provided, the system will
+    #        look for existing works with that field and matching
+    #        value and will update the record instead of creating a new record.
     # @example
     #   attributes: { collection_id: '123',
     #                 depositor_id: '456',
-    #                 batch_id: '789' }
+    #                 batch_id: '789',
+    #                 deduplication_field: 'legacy_id'
+    #               }
     def initialize(error_stream: Darlingtonia.config.default_error_stream,
                    info_stream: Darlingtonia.config.default_info_stream,
                    attributes: {})
       self.collection_id = attributes[:collection_id]
       self.batch_id = attributes[:batch_id]
+      self.deduplication_field = attributes[:deduplication_field]
       set_depositor(attributes[:depositor_id])
       @success_count = 0
       @failure_count = 0
@@ -55,12 +66,27 @@ module Darlingtonia
       self.depositor = user
     end
 
+    ##
+    # @param record [ImportRecord]
+    # @return [ActiveFedora::Base]
+    # Search for any existing records that match on the deduplication_field
+    def find_existing_record(record)
+      return unless deduplication_field
+      return unless record.respond_to?(deduplication_field)
+      return if record.mapper.send(deduplication_field).empty?
+      existing_records = import_type.where("#{deduplication_field}": record.mapper.send(deduplication_field).to_s)
+      raise "More than one record matches deduplication_field #{deduplication_field} with value #{record.mapper.send(deduplication_field)}" if existing_records.count > 1
+      existing_records&.first
+    end
+
     ##
     # @param record [ImportRecord]
     #
     # @return [void]
     def import(record:)
-      create_for(record: record)
+      existing_record = find_existing_record(record)
+      create_for(record: record) unless existing_record
+      update_for(existing_record: existing_record, update_record: record) if existing_record
     rescue Faraday::ConnectionFailed, Ldp::HttpError => e
       error_stream << e
     rescue RuntimeError => e
@@ -153,6 +179,34 @@ module Darlingtonia
 
     private
 
+      # Update an existing object using the Hyrax actor stack
+      # We assume the object was created as expected if the actor stack returns true.
+      def update_for(existing_record:, update_record:)
+        info_stream << "event: record_update_started, batch_id: #{batch_id}, collection_id: #{collection_id}, #{deduplication_field}: #{update_record.respond_to?(deduplication_field) ? update_record.send(deduplication_field) : update_record}"
+        additional_attrs = {
+          depositor: @depositor.user_key
+        }
+        attrs = update_record.attributes.merge(additional_attrs)
+        attrs = attrs.merge(member_of_collections_attributes: { '0' => { id: collection_id } }) if collection_id
+        # Ensure nothing is passed in the files field,
+        # since this is reserved for Hyrax and is where uploaded_files will be attached
+        attrs.delete(:files)
+        based_near = attrs.delete(:based_near)
+        attrs = attrs.merge(based_near_attributes: based_near_attributes(based_near)) unless based_near.nil? || based_near.empty?
+        actor_env = Hyrax::Actors::Environment.new(existing_record,
+                                                   ::Ability.new(@depositor),
+                                                   attrs)
+        if Hyrax::CurationConcern.actor.update(actor_env)
+          info_stream << "event: record_updated, batch_id: #{batch_id}, record_id: #{existing_record.id}, collection_id: #{collection_id}, #{deduplication_field}: #{existing_record.respond_to?(deduplication_field) ? existing_record.send(deduplication_field) : existing_record}"
+          @success_count += 1
+        else
+          existing_record.errors.each do |attr, msg|
+            error_stream << "event: validation_failed, batch_id: #{batch_id}, collection_id: #{collection_id}, attribute: #{attr.capitalize}, message: #{msg}, record_title: record_title: #{attrs[:title] ? attrs[:title] : attrs}"
+          end
+          @failure_count += 1
+        end
+      end
+
       # Create an object using the Hyrax actor stack
       # We assume the object was created as expected if the actor stack returns true.
       def create_for(record:)

data/lib/darlingtonia/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Darlingtonia
-  VERSION = '3.0.5'
+  VERSION = '3.1.0'
 end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: darlingtonia
 version: !ruby/object:Gem::Version
-  version: 3.0.5
+  version: 3.1.0
 platform: ruby
 authors:
 - Data Curation Experts