uc3-dmp-id 0.1.33 → 0.1.35

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: aa5192435943869c90b3aaeb8c9628f91150ea21605589759c8a8dbd7c9f90e3
-  data.tar.gz: 14524cbcdf3708c3239e9973227c9516cbdb3cd7e5c40f62bda7de3d2700c894
+  metadata.gz: 5828afe1afaf7214cdea0e02ed8b834587068be56f1f1a060050c7768c4d19a8
+  data.tar.gz: 10c0bb60320d71414959a9623e435d71ab52b74b758a38fc23cb7fad68f5bd3e
 SHA512:
-  metadata.gz: 329763f4160364093199267eec3f6b4d317e35cd8c092030394db9b2c0fdbd74105749e70ee7acf5eece5e1227926d8ef45fca0804e7f5c4e31ad53a8cb90b41
-  data.tar.gz: df6fb50d5c5249d279411c6023bb299a712d881f4e767a2b0f6ffca6b55d292fbb686cd2e163d93994133c1bf2564ec4cae347c43fed1d852f58376f33e768f4
+  metadata.gz: f4195fb80461a995b42d10e41ff6828aa79692d3cf92c7ce942f71f830c74d2544806a4c0ada2924d0aa7bf8c310c8b8f03a63eeacbd8754d4262aeb343f7dba
+  data.tar.gz: a1108d2311d8c8331b2b176278ae10d9283bdcbb0f4568c0b7d488e22268293b0356c36a51937857f709c64be89edcf2c8a8f98d74d98a66eda3c08f9ef0dea2

data/lib/uc3-dmp-id/comparator.rb

@@ -56,9 +56,7 @@ module Uc3DmpId
       scoring = []
       return scoring unless hash.is_a?(Hash) && !hash['title'].nil?
 
-      @dmps.each do |entry|
-        dmp = entry.fetch('_source', {})
-
+      @dmps.each do |dmp|
         # Compare the grant ids. If we have a match return the response immediately since that is
         # a very positive match!
         response = { confidence: 'None', score: 0, notes: [] }
@@ -79,7 +77,7 @@ module Uc3DmpId
         next if response[:score] <= 2
 
         # Set the confidence level based on the score
-        response[:dmp_id] = entry['_id']
+        response[:dmp_id] = "DMP##{dmp['dmp_id']}"
         response[:confidence] = if response[:score] > 10
                                   'High'
                                 else

data/lib/uc3-dmp-id/finder.rb

@@ -1,6 +1,7 @@
 # frozen_string_literal: true
 
 require 'uc3-dmp-dynamo'
+require 'securerandom'
 
 module Uc3DmpId
   class FinderError < StandardError; end
@@ -75,6 +76,9 @@ module Uc3DmpId
         dmp = resp['dmp'].nil? ? JSON.parse({ dmp: resp }.to_json) : resp
         return nil if dmp['dmp']['PK'].nil?
 
+        # Attach any harvester mods to the JSON
+        dmp['dmp'] = _attach_harvester_mods(client:, p_key:, json: dmp['dmp'], logger:)
+
         dmp = Versioner.append_versions(p_key: dmp['dmp']['PK'], dmp:, client:, logger:) if cleanse
         dmp = _remove_narrative_if_private(json: dmp)
         cleanse ? Helper.cleanse_dmp_json(json: dmp) : dmp
@@ -224,6 +228,47 @@ module Uc3DmpId
         end
         json
       end
+
+      # Fetch any Harvester modifications and attach them to the JSON in the way the DMPTool expects
+      # TODO: eventually just update the rebuilt DMPTool to work with the HARVESTER_MODS records as-is
+      def _attach_harvester_mods(client:, p_key:, json:, logger: nil)
+        # Fetch the `"SK": "HARVESTER_MODS"` record
+        client = Uc3DmpDynamo::Client.new if client.nil?
+        resp = client.get_item(
+          key: { PK: Helper.append_pk_prefix(p_key:), SK: Helper::SK_HARVESTER_MODS }, logger:
+        )
+        return json unless resp.is_a?(Hash)
+
+        mods = []
+        resp.fetch('related_works', {}).each do |key, val|
+          rec = val.dup
+          next if rec['provenance'].nil?
+
+          # Change the name of the `logic` array to `notes`
+          rec['notes'] = rec['logic']
+          rec['score'] = rec['score'].to_s
+          # For `work-type` that equal `outputmanagementplan`, change it to `output_management_plan`
+          rec['work_type'] = 'output_management_plan' if rec['work_type'] == 'outputmanagementplan'
+
+          # The old `dmphub_modifications` array was grouped by provenance
+          prov_array = mods.select { |entry| entry['provenance'] == rec['provenance'] }
+          if prov_array.any?
+            prov_array << rec
+          else
+            mods << {
+              id: "#{Time.now.utc.strftime('%Y-%m-%d')}-#{SecureRandom.hex(4)}",
+              provenance: rec['provenance'],
+              augmenter_run_id: SecureRandom.hex(8),
+              timestamp: rec['discovered_at'].nil? ? Time.now.utc.iso8601 : rec['discovered_at'],
+              dmproadmap_related_identifiers: [rec],
+              funding: []
+            }
+          end
+        end
+        # Add a `dmphub_modifications` array to the JSON
+        json['dmphub_modifications'] = JSON.parse(mods.to_json)
+        json
+      end
     end
   end
 end

data/lib/uc3-dmp-id/helper.rb

@@ -12,6 +12,8 @@ module Uc3DmpId
     SK_DMP_PREFIX = 'VERSION#'
     SK_DMP_REGEX = /VERSION#\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}\+\d{2}:\d{2}/
 
+    SK_HARVESTER_MODS = "HARVESTER_MODS"
+
     # TODO: Verify the assumed structure of the DOI is valid
     DOI_REGEX = %r{[0-9]{2}\.[0-9]{4,}/[a-zA-Z0-9/_.-]+}
     URL_REGEX = %r{(https?://)?([a-zA-Z0-9\-_]\.)+[a-zA-Z0-9\-_]{2,3}(:[0-9]+)?/?}

data/lib/uc3-dmp-id/updater.rb

@@ -43,6 +43,9 @@ module Uc3DmpId
         # Bail if the system trying to make the update is not the creator of the DMP ID
         raise UpdaterError, Helper::MSG_DMP_FORBIDDEN if owner != updater
 
+        # Handle any changes to the dmphub_modifications section
+        version = _process_harvester_mods(client:, p_key:, json: version, logger:)
+
         # Remove the version info because we don't want to save it on the record
         version.delete('dmphub_versions')
 
@@ -177,6 +180,35 @@ module Uc3DmpId
       end
       # rubocop:enable Metrics/AbcSize, Metrics/MethodLength
       # rubocop:enable Metrics/PerceivedComplexity, Metrics/CyclomaticComplexity
+
+      # Fetch any Harvester modifications to the JSON
+      def _process_harvester_mods(client:, p_key:, json:, logger: nil)
+        return json if json.fetch('dmphub_modifications', []).empty?
+
+        # Fetch the `"SK": "HARVESTER_MODS"` record
+        client = Uc3DmpDynamo::Client.new if client.nil?
+        resp = client.get_item(
+          key: { PK: Helper.append_pk_prefix(p_key:), SK: Helper::SK_HARVESTER_MODS }, logger:
+        )
+        return json unless resp.is_a?(Hash) && resp['related_works'].is_a?(Hash)
+
+        # The `dmphub_modifications` array will ONLY ever have things the harvester mods know about
+        # so just find them and update the status accordingly
+        mods = resp.dup
+        json['dmphub_modifications'].each do |entry|
+          next if entry.fetch('dmproadmap_related_identifiers', []).empty?
+
+          entry['dmproadmap_related_identifiers'].each do |related|
+            next if mods['related_works'][related.identifier].nil?
+
+            mods['related_works'][related.identifier]['status'] = related['status']
+          end
+        end
+
+        client.put_item(json: mods, logger:)
+        json.delete('dmphub_modifications')
+        json
+      end
     end
   end
 end

data/lib/uc3-dmp-id/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module Uc3DmpId
-  VERSION = '0.1.33'
+  VERSION = '0.1.35'
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: uc3-dmp-id
 version: !ruby/object:Gem::Version
-  version: 0.1.33
+  version: 0.1.35
 platform: ruby
 authors:
 - Brian Riley
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2024-04-11 00:00:00.000000000 Z
+date: 2024-05-01 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: json