pennmarc 1.0.8 → 1.0.9

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 2dbb99dde790dda81ab9af2e7fdcc003a6ebc06a5ff447816705ef0836594a30
-  data.tar.gz: 75ab2803be0bc910f62ecb177490ab4de09c13cadabbf5b55921197f7e713ef9
+  metadata.gz: 652039b781e2ce7442f628ca722f4fda59207d1d8171bfd2fad84d2c67f3c06f
+  data.tar.gz: 166bbf8498f673b36373671db47031ef8a2f4d5496ed1fddeee89718bd23e09c
 SHA512:
-  metadata.gz: 755bce4545eedf08220d746d7a01a0d2e2d5a0abfc777cd94aff4a2e33d7f34a457f7bd181031bc934c8c2435a5c8e8fe5a20a301ba363974ce2dcafa6d81783
-  data.tar.gz: 8613afeaf7fbe166fc1ce62acd792bcc85f89537ad0f515da140b811b17c2a62e3136fbeb8335758068ae672fb3a579f847cc6c12fcd2a696fa3358f85a10524
+  metadata.gz: 68df667344adaff013da3eab5417bdd44748726a113190d20c44cc1cf4d061269619265e3de1edc9c7c767ffff5a4548a9fcb3ed2039b28a2608ec62a6e024d1
+  data.tar.gz: 8df6d32dbc84ff3851e4b35576e24c96c42b536330ded1f0f1678de3d1aa0729547a71b0c6ccea15e7c949ee15e068d1c0f19b18a6b414167cb736effa2b01dd

data/lib/pennmarc/enriched_marc.rb

@@ -32,5 +32,16 @@ module PennMARC
     # a subfield code NOT used by the MARC 21 spec for 852 holdings records.
     # we add this subfield during preprocessing to store boundwith record IDs.
     SUB_BOUND_WITH_ID = 'y'
+
+    # MARC enrichment originating from Alma Api
+    # @see https://developers.exlibrisgroup.com/alma/apis/docs/bibs/R0VUIC9hbG1hd3MvdjEvYmlicy97bW1zX2lkfQ==/ Alma docs
+    module AlmaApi
+      TAG_PHYSICAL_INVENTORY = 'AVA'
+      TAG_DIGITAL_INVENTORY = 'AVA'
+      TAG_ELECTRONIC_INVENTORY = 'AVE'
+
+      SUB_PHYSICAL_CALL_NUMBER = 'd'
+      SUB_PHYSICAL_CALL_NUMBER_TYPE = 'k'
+    end
   end
 end

data/lib/pennmarc/helpers/classification.rb

@@ -0,0 +1,95 @@
+# frozen_string_literal: true
+
+module PennMARC
+  # Generates library of congress and dewey classifications using call number data.
+  class Classification < Helper
+    # Subfield value that identifies Library of Congress call number
+    LOC_CALL_NUMBER_TYPE = '0'
+
+    # Subfield value that identifies Dewey call number
+    DEWEY_CALL_NUMBER_TYPE = '1'
+
+    # Hash that maps call number type to the appropriate mapper
+    CLASSIFICATION_MAPS = {
+      LOC_CALL_NUMBER_TYPE => Mappers.loc_classification,
+      DEWEY_CALL_NUMBER_TYPE => Mappers.dewey_classification
+    }.freeze
+
+    # Enriched MARC tags that hold classification data
+    TAGS = [EnrichedMarc::TAG_ITEM, EnrichedMarc::AlmaApi::TAG_PHYSICAL_INVENTORY].freeze
+
+    class << self
+      # Parse classification values for faceting. We retrieve classification values from enriched MARC fields 'itm' or
+      # 'AVA' originating respectively from the Alma publishing process or from the Alma Api. We return the
+      # highest level LOC or Dewey classifications from each available call number, joining the class code with
+      # its title in a single string. See {PennMARC::EnrichedMarc} and {PennMARC::EnrichedMarc::AlmaApi} for more
+      # information on the enriched MARC fields.
+      # @see https://developers.exlibrisgroup.com/alma/apis/docs/bibs/R0VUIC9hbG1hd3MvdjEvYmlicy97bW1zX2lkfQ==/ AVA docs
+      # @param [MARC::Record] record
+      # @return [Array<String>] array of classifications
+      def facet(record)
+        record.fields(TAGS).flat_map do |field|
+          call_number_type = subfield_values(field, call_number_type_sf(field))&.first
+          call_numbers = subfield_values(field, call_number_sf(field))
+
+          call_numbers.filter_map do |call_number|
+            class_code = call_number[0]
+            title = translate_classification(class_code, call_number_type)
+            next if title.blank?
+
+            format_facet(class_code, call_number_type, title)
+          end
+        end
+      end
+
+      private
+
+      # Retrieve subfield code that stores the call number on enriched marc field
+      # @param [MARC::DataField] field
+      # @return [String]
+      def call_number_sf(field)
+        return EnrichedMarc::SUB_ITEM_CALL_NUMBER if field.tag == EnrichedMarc::TAG_ITEM
+
+        EnrichedMarc::AlmaApi::SUB_PHYSICAL_CALL_NUMBER
+      end
+
+      # Retrieve subfield code that stores call number type on enriched marc field
+      # @param [MARC::DataField] field
+      # @return [String]
+      def call_number_type_sf(field)
+        return EnrichedMarc::SUB_ITEM_CALL_NUMBER_TYPE if field.tag == EnrichedMarc::TAG_ITEM
+
+        EnrichedMarc::AlmaApi::SUB_PHYSICAL_CALL_NUMBER_TYPE
+      end
+
+      # retrieve title of classification based on single char classification code and call number type
+      # @param[String] class_code classification code
+      # @param[String] call_number_type value from call number type subfield
+      # @return [String, NilClass]
+      def translate_classification(class_code, call_number_type)
+        map = CLASSIFICATION_MAPS[call_number_type]
+
+        return if map.blank?
+
+        translate_relator(class_code, map)
+      end
+
+      # format classification facet by joining single character classification code with its corresponding title.
+      # Our Dewey mapping codes are single digit, so we must concatenate '00' to the class code to accurately reflect
+      # Dewey class codes.
+      # @return [String]
+      def format_facet(class_code, call_number_type, title)
+        return [class_code, title].join(' - ') if loc_call_number_type?(call_number_type)
+
+        ["#{class_code}00", title].join(' - ')
+      end
+
+      # Determine whether call number type is library of congress
+      # @param [String] call_number_type value from call number type subfield
+      # @return [Boolean]
+      def loc_call_number_type?(call_number_type)
+        call_number_type == '0'
+      end
+    end
+  end
+end

data/lib/pennmarc/mappers.rb

@@ -23,6 +23,16 @@ module PennMARC
         @relator ||= load_map('relator.yml')
       end
 
+      # @return [Hash]
+      def loc_classification
+        @loc_classification ||= load_map('loc_classification.yml')
+      end
+
+      # @return [Hash]
+      def dewey_classification
+        @dewey_classification ||= load_map('dewey_classification.yml')
+      end
+
       # @param [String] filename of mapping file in config directory, with file extension
       # @return [Hash] mapping as hash
       def load_map(filename)

data/lib/pennmarc/mappings/dewey_classification.yml

@@ -0,0 +1,11 @@
+---
+'0': Computer science, information & general works
+'1': Philosophy & psychology
+'2': Religion
+'3': Social sciences
+'4': Language
+'5': Science
+'6': Technology
+'7': Arts & recreation
+'8': Literature
+'9': History & geography

data/lib/pennmarc/mappings/loc_classification.yml

@@ -0,0 +1,22 @@
+---
+A: General Works
+B: Philosophy, Psychology, Religion
+C: 'History: Auxiliary Sciences'
+D: 'History:  General & European'
+E: 'History:  United States'
+F: 'History:  Western Hemisphere'
+G: Geography, Anthropology, Recreation
+H: Social Science, Economics, Sociology
+J: Political Science
+K: Law
+L: Education
+M: Music
+N: Fine Arts
+P: Literature & Languages
+Q: Science
+R: Medicine
+S: Agriculture
+T: Technology
+U: Military Science
+V: Naval Science
+Z: Bibliography & Library Science

data/lib/pennmarc/version.rb

@@ -1,5 +1,5 @@
 # frozen_string_literal: true
 
 module PennMARC
-  VERSION = '1.0.8'
+  VERSION = '1.0.9'
 end

data/spec/lib/pennmarc/helpers/classification_spec.rb

@@ -0,0 +1,41 @@
+# frozen_string_literal: true
+
+describe 'PennMARC::Classification' do
+  include MarcSpecHelpers
+
+  let(:helper) { PennMARC::Classification }
+  let(:record) do
+    marc_record fields: [marc_field(tag: tag,
+                                    subfields: { call_number_type_sf => '0', call_number_sf => 'TA683 .B3 1909b' }),
+                         marc_field(tag: tag,
+                                    subfields: { call_number_type_sf => '0', call_number_sf => 'QL756 .S643' }),
+                         marc_field(tag: tag,
+                                    subfields: { call_number_type_sf => '1', call_number_sf => '691.3 B2141' }),
+                         marc_field(tag: tag,
+                                    subfields: { call_number_type_sf => '1', call_number_sf => '378.748 POS1952.29' })]
+  end
+
+  describe '.facet' do
+    context 'with enrichment via the Alma publishing process' do
+      let(:tag) { PennMARC::EnrichedMarc::TAG_ITEM }
+      let(:call_number_type_sf) { PennMARC::EnrichedMarc::SUB_ITEM_CALL_NUMBER_TYPE }
+      let(:call_number_sf) { PennMARC::EnrichedMarc::SUB_ITEM_CALL_NUMBER }
+
+      it 'returns expected values' do
+        expect(helper.facet(record)).to contain_exactly('T - Technology', '600 - Technology',
+                                                        '300 - Social sciences', 'Q - Science')
+      end
+    end
+
+    context 'with enrichment with availability info via Alma Api' do
+      let(:tag) { PennMARC::EnrichedMarc::AlmaApi::TAG_PHYSICAL_INVENTORY }
+      let(:call_number_type_sf) { PennMARC::EnrichedMarc::AlmaApi::SUB_PHYSICAL_CALL_NUMBER_TYPE }
+      let(:call_number_sf) { PennMARC::EnrichedMarc::AlmaApi::SUB_PHYSICAL_CALL_NUMBER }
+
+      it 'returns expected values' do
+        expect(helper.facet(record)).to contain_exactly('T - Technology', '600 - Technology',
+                                                        '300 - Social sciences', 'Q - Science')
+      end
+    end
+  end
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: pennmarc
 version: !ruby/object:Gem::Version
-  version: 1.0.8
+  version: 1.0.9
 platform: ruby
 authors:
 - Mike Kanning
@@ -10,7 +10,7 @@ authors:
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2023-12-07 00:00:00.000000000 Z
+date: 2023-12-08 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activesupport
@@ -91,6 +91,7 @@ files:
 - lib/pennmarc/heading_control.rb
 - lib/pennmarc/helpers/access.rb
 - lib/pennmarc/helpers/citation.rb
+- lib/pennmarc/helpers/classification.rb
 - lib/pennmarc/helpers/creator.rb
 - lib/pennmarc/helpers/database.rb
 - lib/pennmarc/helpers/date.rb
@@ -109,8 +110,10 @@ files:
 - lib/pennmarc/helpers/subject.rb
 - lib/pennmarc/helpers/title.rb
 - lib/pennmarc/mappers.rb
+- lib/pennmarc/mappings/dewey_classification.yml
 - lib/pennmarc/mappings/iso639-2-languages.yml
 - lib/pennmarc/mappings/iso639-3-languages.yml
+- lib/pennmarc/mappings/loc_classification.yml
 - lib/pennmarc/mappings/locations.yml
 - lib/pennmarc/mappings/relator.yml
 - lib/pennmarc/parser.rb
@@ -120,6 +123,7 @@ files:
 - spec/fixtures/marcxml/test.xml
 - spec/lib/pennmarc/helpers/access_spec.rb
 - spec/lib/pennmarc/helpers/citation_spec.rb
+- spec/lib/pennmarc/helpers/classification_spec.rb
 - spec/lib/pennmarc/helpers/creator_spec.rb
 - spec/lib/pennmarc/helpers/database_spec.rb
 - spec/lib/pennmarc/helpers/date_spec.rb