cocina-models 0.94.0 → 0.94.2

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 325cc1afd87f6c47380be3a4858d8858e31661efdfd39c0360c4a62f4541b6f2
-  data.tar.gz: 4539fb9f4a715cc488e50a67df0c1e345ff227be9afe1dbd94c6466d4902a41b
+  metadata.gz: e147c55355cc4c2f11ed358c5091c0942e48168948da875a1fd1ff97f7dfb95a
+  data.tar.gz: fd668019a9193c1620d419b9a3815115a1bc0d8a93e635721bbf1251db80df10
 SHA512:
-  metadata.gz: d37e0ca7edb90caf68d706741e1f04ad8eafa767527af393f72ef3fe44ab55fa056b027848c46d8c43ed4ccdc97369a86a93a7d1d82b60b8d2302307fc1e634b
-  data.tar.gz: 2bca5d3dcf6f5d141a0402545a558b17953af6f64311b29e11a57fc0f0d842af983a901b86e1a1cc8265c61f454db88d5f2a84460bb90e5bd0c0d8031073e913
+  metadata.gz: 3f9485a94c43606870b8d47f7b0d29597660fe82a884fab5128922501e65eccee5787c43a27f796dd35af36992bb04a417eb89e045e2bd74be4c8b30212e2d04
+  data.tar.gz: 0707161a4e5399762878ed9d9215ae59a807d046fa7f3b199e09227f8fb5da69484b32dc0a942df5fb1281aafd8ec29a054b0eae04f40a68ceddc84fda15e406

data/Gemfile.lock

@@ -1,7 +1,7 @@
 PATH
   remote: .
   specs:
-    cocina-models (0.94.0)
+    cocina-models (0.94.2)
       activesupport
       deprecation
       dry-struct (~> 1.0)
@@ -21,7 +21,7 @@ PATH
 GEM
   remote: https://rubygems.org/
   specs:
-    activesupport (7.1.2)
+    activesupport (7.1.3)
       base64
       bigdecimal
       concurrent-ruby (~> 1.0, >= 1.0.2)
@@ -34,14 +34,14 @@ GEM
     ast (2.4.2)
     attr_extras (7.1.0)
     base64 (0.2.0)
-    bigdecimal (3.1.5)
+    bigdecimal (3.1.6)
     byebug (11.1.3)
     committee (5.0.0)
       json_schema (~> 0.14, >= 0.14.3)
       openapi_parser (~> 1.0)
       rack (>= 1.5)
     commonmarker (0.23.10)
-    concurrent-ruby (1.2.2)
+    concurrent-ruby (1.2.3)
     connection_pool (2.4.1)
     deprecation (1.1.0)
       activesupport
@@ -82,7 +82,7 @@ GEM
       multi_json
     language_server-protocol (3.17.0.3)
     mini_portile2 (2.8.5)
-    minitest (5.21.1)
+    minitest (5.21.2)
     multi_json (1.15.0)
     mutex_m (0.2.0)
     nokogiri (1.16.0)
@@ -93,7 +93,7 @@ GEM
     openapi_parser (1.0.0)
     optimist (3.1.0)
     parallel (1.24.0)
-    parser (3.3.0.3)
+    parser (3.3.0.5)
       ast (~> 2.4.1)
       racc
     patience_diff (1.2.0)
@@ -121,7 +121,7 @@ GEM
       rspec-core (>= 2, < 4, != 2.12.0)
     rss (0.3.0)
       rexml
-    rubocop (1.60.0)
+    rubocop (1.60.1)
       json (~> 2.3)
       language_server-protocol (>= 3.17.0)
       parallel (~> 1.10)

data/README.md

@@ -143,7 +143,7 @@ which pushes the gem to rubygems.org.
 
 ### Step 2: Update client gems coupled to the models
 
-Release new versions of [sdr-client](https://github.com/sul-dlss/sdr-client) and [dor-services-client](https://github.com/sul-dlss/dor-services-client/) pinned to use the new cocina-models version because applications such as [Argo](https://github.com/sul-dlss/argo) depend on both of these gems using the same models.
+Release new versions of [sdr-client](https://github.com/sul-dlss/sdr-client), [dor-services-client](https://github.com/sul-dlss/dor-services-client/), and [dor_indexing](https://github.com/sul-dlss/dor_indexing/) pinned to use the new cocina-models version because applications such as [Argo](https://github.com/sul-dlss/argo) depend on both of these gems using the same models.
 
 ### Step 3: Update services directly coupled to the models
 

data/lib/cocina/models/builders/title_builder.rb

@@ -12,7 +12,8 @@ module Cocina
         # @param [Symbol] strategy ":first" is the strategy for selection when primary or display
         #   title are missing
         # @param [Boolean] add_punctuation determines if the title should be formmated with punctuation
-        # @return [String] the title value for Solr
+        # @return [String, Array] the title value for Solr - for :first strategy, a string; for :all strategy, an array
+        #   (e.g. title displayed in blacklight search results vs boosting values for search result rankings)
         def self.build(titles, strategy: :first, add_punctuation: true)
           if titles.respond_to?(:description)
             Deprecation.warn(self,
@@ -27,7 +28,7 @@ module Cocina
         #   we can boost matches on it in search results (boost matching this string higher than matching full title string)
         #   e.g. "The Hobbit" (main_title) vs "The Hobbit, or, There and Back Again (full_title)
         # @param [[Array<Cocina::Models::Title,Cocina::Models::DescriptiveValue>] titles the titles to consider
-        # @return [String] the main title value for Solr
+        # @return [Array<String>] the main title value(s) for Solr - array due to possible parallelValue
         def self.main_title(titles)
           new(strategy: :first, add_punctuation: false).main_title(titles)
         end
@@ -35,9 +36,9 @@ module Cocina
         # the "full title" is the title WITH subtitle, part name, etc.  We want to able able to index it separately so
         #   we can boost matches on it in search results (boost matching this string higher than other titles present)
         # @param [[Array<Cocina::Models::Title,Cocina::Models::DescriptiveValue>] titles the titles to consider
-        # @return [String] the title value for Solr
+        # @return [Array<String>] the full title value(s) for Solr - array due to possible parallelValue
         def self.full_title(titles)
-          new(strategy: :first, add_punctuation: false).build(titles)
+          [new(strategy: :first, add_punctuation: false, only_one_parallel_value: false).build(titles)].flatten.compact
         end
 
         # "additional titles" are all title data except for full_title.  We want to able able to index it separately so
@@ -45,16 +46,28 @@ module Cocina
         # @param [[Array<Cocina::Models::Title,Cocina::Models::DescriptiveValue>] titles the titles to consider
         # @return [Array<String>] the values for Solr
         def self.additional_titles(titles)
-          new(strategy: :all, add_punctuation: false).build(titles) - [full_title(titles)]
+          [new(strategy: :all, add_punctuation: false).build(titles)].flatten - full_title(titles)
         end
 
-        def initialize(strategy:, add_punctuation:)
+        # @param strategy [Symbol] ":first" selects a single title value based on precedence of
+        #   primary, untyped, first occurrence. ":all" returns an array containing all the values.
+        # @param add_punctuation [boolean] whether the title should be formmated with punctuation (think of a structured
+        #   value coming from a MARC record, which is designed for catalog cards.)
+        # @param only_one_parallel_value [boolean] when true, choose one of the parallel values according to precedence
+        #   of primary, untyped, first occurrence.  When false, return an array containing all the parallel values.
+        #   Why? Think of e.g. title displayed in blacklight search results vs boosting values for ranking of search
+        #   results
+        def initialize(strategy:, add_punctuation:, only_one_parallel_value: true)
           @strategy = strategy
           @add_punctuation = add_punctuation
+          @only_one_parallel_value = only_one_parallel_value
         end
 
         # @param [[Array<Cocina::Models::Title>] cocina_titles the titles to consider
-        # @return [String] the title value for Solr
+        # @return [String, Array] the title value for Solr - for :first strategy, a string; for :all strategy, an array
+        #   (e.g. title displayed in blacklight search results vs boosting values for search result rankings)
+        #
+        # rubocop:disable Metrics/PerceivedComplexity
         def build(cocina_titles)
           cocina_title = primary_title(cocina_titles) || untyped_title(cocina_titles)
           cocina_title = other_title(cocina_titles) if cocina_title.blank?
@@ -62,15 +75,23 @@ module Cocina
           if strategy == :first
             extract_title(cocina_title)
           else
-            cocina_titles.map { |ctitle| extract_title(ctitle) }.flatten
+            result = cocina_titles.map { |ctitle| extract_title(ctitle) }.flatten
+            if only_one_parallel_value? && result.length == 1
+              result.first
+            else
+              result
+            end
           end
         end
+        # rubocop:enable Metrics/PerceivedComplexity
 
+        # this is the single "short title" - the title without subtitle, part name, etc.
+        #    this may be useful for boosting and exact matching for search results
+        # @return [Array<String>] the main title value(s) for Solr - can be array due to parallel titles
         def main_title(titles)
           cocina_title = primary_title(titles) || untyped_title(titles)
           cocina_title = other_title(titles) if cocina_title.blank?
 
-          cocina_title = cocina_title.first if cocina_title.is_a?(Array)
           extract_main_title(cocina_title)
         end
 
@@ -79,40 +100,71 @@ module Cocina
         attr_reader :strategy
 
         def extract_title(cocina_title)
+          title_values = if cocina_title.value
+                           cocina_title.value
+                         elsif cocina_title.structuredValue.present?
+                           rebuild_structured_value(cocina_title)
+                         elsif cocina_title.parallelValue.present?
+                           extract_title_parallel_values(cocina_title)
+                         end
+          result = [title_values].flatten.compact.map { |val| remove_trailing_punctuation(val.strip) }
+          result.length == 1 ? result.first : result
+        end
+
+        # stategy :first says to return a single value (default: true)
+        # only_one_parallel_value? says to return a single value, even if that value is a parallelValue (default: false)
+        #
+        # rubocop:disable Metrics/PerceivedComplexity
+        def extract_title_parallel_values(cocina_title)
+          primary = cocina_title.parallelValue.find { |pvalue| pvalue.status == 'primary' }
+          if primary && only_one_parallel_value? && strategy == :first
+            # we have a primary title and we know we want a single value
+            extract_title(primary)
+          elsif only_one_parallel_value? && strategy == :first
+            # no primary value;  algorithm says prefer an untyped value over a typed value for single value
+            untyped = cocina_title.parallelValue.find { |pvalue| pvalue.type.blank? }
+            extract_title(untyped || cocina_title.parallelValue.first)
+          else
+            cocina_title.parallelValue.map { |pvalue| extract_title(pvalue) }
+          end
+        end
+        # rubocop:enable Metrics/PerceivedComplexity
+
+        # @return [Array<String>] the main title value(s) for Solr - can be array due to parallel titles
+        def extract_main_title(cocina_title) # rubocop:disable Metrics/PerceivedComplexity
           result = if cocina_title.value
-                     cocina_title.value
+                     cocina_title.value # covers both title and main title types
                    elsif cocina_title.structuredValue.present?
-                     title_from_structured_values(cocina_title)
+                     main_title_from_structured_values(cocina_title)
                    elsif cocina_title.parallelValue.present?
-                     return build(cocina_title.parallelValue)
+                     primary = cocina_title.parallelValue.find { |pvalue| pvalue.status == 'primary' }
+                     if primary
+                       extract_main_title(primary)
+                     else
+                       cocina_title.parallelValue.map { |pvalue| extract_main_title(pvalue) }
+                     end
                    end
-          remove_trailing_punctuation(result.strip) if result.present?
-        end
+          return [] if result.blank?
 
-        def extract_main_title(cocina_title)
-          if cocina_title.value
-            cocina_title.value # covers both title and main title types
-          elsif cocina_title.structuredValue.present?
-            main_title_from_structured_values(cocina_title)
-          elsif cocina_title.parallelValue.present?
-            main_title(cocina_title.parallelValue)
-          end
+          [result].flatten.compact.map { |val| remove_trailing_punctuation(val) }
         end
 
         def add_punctuation?
           @add_punctuation
         end
 
+        def only_one_parallel_value?
+          @only_one_parallel_value
+        end
+
         # @return [Cocina::Models::Title, nil] title that has status=primary
-        def primary_title(titles)
-          primary_title = titles.find do |title|
-            title.status == 'primary'
-          end
+        def primary_title(cocina_titles)
+          primary_title = cocina_titles.find { |title| title.status == 'primary' }
           return primary_title if primary_title.present?
 
           # NOTE: structuredValues would only have status primary assigned as a sibling, not as an attribute
 
-          titles.find do |title|
+          cocina_titles.find do |title|
             title.parallelValue&.find do |parallel_title|
               parallel_title.status == 'primary'
             end
@@ -149,56 +201,73 @@ module Cocina
         # @return [String] the title value from combining the pieces of the structured_values by type and order
         #   with desired punctuation per specs
         #
+        # - nonsorting characters value is followed by a space, unless the nonsorting
+        #   character count note has a numeric value equal to the length of the
+        #   nonsorting characters value, in which case no space is inserted
+        # - subtitle is preceded by space colon space, unless it is at the beginning
+        #   of the title string
+        # - partName and partNumber are always separated from each other by comma space
+        # - first partName or partNumber in the string is preceded by period space
+        # - partName or partNumber before nonsorting characters or main title is followed
+        #   by period space
+        #
+        # for punctuaion funky town, thank MARC and catalog cards
+        #
+        # rubocop:disable Metrics/AbcSize
         # rubocop:disable Metrics/CyclomaticComplexity
         # rubocop:disable Metrics/MethodLength
         # rubocop:disable Metrics/PerceivedComplexity
-        def title_from_structured_values(title)
-          # parse out the parts
-          main_title = ''
-          subtitle = ''
-          non_sort_value = ''
+        def rebuild_structured_value(cocina_title)
+          result = ''
           part_name_number = ''
-          title.structuredValue.each do |structured_value|
-            # There can be a structuredValue inside a structuredValue.  For example,
+          cocina_title.structuredValue.each do |structured_value| # rubocop:disable Metrics/BlockLength
+            # There can be a structuredValue inside a structuredValue, for example,
             #   a uniform title where both the name and the title have internal StructuredValue
-            return title_from_structured_values(structured_value) if structured_value.structuredValue.present?
+            return rebuild_structured_value(structured_value) if structured_value.structuredValue.present?
 
             value = structured_value.value&.strip
             next unless value
 
-            # additional types:  name, uniform ...
+            # additional types ignored here, e.g. name, uniform ...
             case structured_value.type&.downcase
             when 'nonsorting characters'
-              non_sort_value = "#{value}#{non_sorting_padding(title, value)}"
+              padding = non_sorting_padding(cocina_title, value)
+              result = add_non_sorting_value(result, value, padding)
             when 'part name', 'part number'
-              part_name_number = part_name_number(title.structuredValue) if part_name_number.blank?
-            when 'main title', 'title'
-              main_title = value
-            when 'subtitle'
-              # combine multiple subtitles into a single string
-              subtitle = if !add_punctuation?
-                           if subtitle.present?
-                             [subtitle, value].join(' ')
-                           else
-                             value
-                           end
-                         elsif subtitle.present?
-                           # subtitle is preceded by space colon space, unless it is at the beginning of the title string
-                           "#{subtitle.sub(/[. :]+$/, '')} : #{value.sub(/^:/, '').strip}"
+              if part_name_number.blank?
+                part_name_number = part_name_number(cocina_title.structuredValue)
+                result = if !add_punctuation?
+                           [result, part_name_number].join(' ')
+                         elsif result.present?
+                           # part name/number is preceded by period space, unless it is at the beginning of the title string
+                           "#{result.sub(/[ .,]*$/, '')}. #{part_name_number}. "
                          else
-                           value.sub(/^:/, '').strip
+                           "#{part_name_number}. "
                          end
+              end
+            when 'main title', 'title'
+              # nonsorting characters ending with hyphen, apostrophe or space should be slammed against the main title,
+              #   even if we are not adding punctuation
+              result = if add_punctuation? || result.ends_with?(' ') || result.ends_with?('-') || result.ends_with?('\'')
+                         [result, value].join
+                       else
+                         [remove_trailing_punctuation(result), remove_trailing_punctuation(value)].select(&:presence).join(' ')
+                       end
+            when 'subtitle'
+              result = if !add_punctuation?
+                         [result, value.sub(/^:/, '').strip].select(&:presence).join(' ')
+                       elsif result.present?
+                         # subtitle is preceded by space colon space, unless it is at the beginning of the title string
+                         "#{result.sub(/[. :]+$/, '')} : #{value.sub(/^:/, '').strip}"
+                       else
+                         result = value.sub(/^:/, '').strip
+                       end
             end
           end
 
-          # combine the parts into a single title string
-          if add_punctuation?
-            combine_with_punctuation(non_sort_value: non_sort_value, main_title: main_title, subtitle: subtitle,
-                                     part_name_number: part_name_number)
-          else
-            ["#{non_sort_value}#{main_title}", subtitle, part_name_number].select(&:presence).join(' ')
-          end
+          result
         end
+        # rubocop:enable Metrics/AbcSize
         # rubocop:enable Metrics/CyclomaticComplexity
         # rubocop:enable Metrics/MethodLength
         # rubocop:enable Metrics/PerceivedComplexity
@@ -206,11 +275,14 @@ module Cocina
         # main_title is title.structuredValue.value with type 'main title' (or just title.value)
         # @param [Cocina::Models::Title] title with structured values
         # @return [String] the main title value
-        def main_title_from_structured_values(cocina_title) # rubocop:disable Metrics/MethodLength
+        #
+        # rubocop:disable Metrics/MethodLength
+        # rubocop:disable Metrics/PerceivedComplexity
+        def main_title_from_structured_values(cocina_title)
           result = ''
           # combine pieces of the cocina structuredValue into a single title
           cocina_title.structuredValue.each do |structured_value|
-            # There can be a structuredValue inside a structuredValue.  For example,
+            # There can be a structuredValue inside a structuredValue, for example,
             #   a uniform title where both the name and the title have internal StructuredValue
             return main_title_from_structured_values(structured_value) if structured_value.structuredValue.present?
 
@@ -219,43 +291,36 @@ module Cocina
 
             case structured_value.type&.downcase
             when 'nonsorting characters'
-              non_sort_value = "#{value}#{non_sorting_padding(cocina_title, value)}"
-              result = "#{non_sort_value}#{result}" # non-sorting characters are at the beginning of the title
-            when 'main title'
-              result = "#{result}#{value}"
-            when 'title'
-              result = value
+              padding = non_sorting_padding(cocina_title, value)
+              result = add_non_sorting_value(result, value, padding)
+            when 'main title', 'title'
+              result = if ['\'', '-'].include?(result.last)
+                         [result, value].join
+                       else
+                         [remove_trailing_punctuation(result).strip, remove_trailing_punctuation(value).strip].select(&:presence).join(' ')
+                       end
             end
           end
-          result
-        end
 
-        # Thank MARC and catalog cards for this mess. We need to add punctuation.
-        # rubocop:disable Metrics/MethodLength
-        def combine_with_punctuation(non_sort_value:, main_title:, subtitle:, part_name_number:)
-          result = "#{non_sort_value}#{main_title}"
-          if subtitle.present?
-            result = if result.present?
-                       "#{result.sub(/[. :]+$/, '')} : #{subtitle.sub(/^:/, '').strip}"
-                     else
-                       result = subtitle
-                     end
-          end
-          if part_name_number.present?
-            result = if result.present?
-                       "#{result.sub(/[ .,]*$/, '')}. #{part_name_number}."
-                     else
-                       "#{part_name_number}."
-                     end
-          end
           result
         end
         # rubocop:enable Metrics/MethodLength
+        # rubocop:enable Metrics/PerceivedComplexity
 
+        # Thank MARC and catalog cards for this mess.
         def remove_trailing_punctuation(title)
           title.sub(%r{[ .,;:/\\]+$}, '')
         end
 
+        def add_non_sorting_value(title_so_far, non_sorting_value, padding)
+          non_sort_value = "#{non_sorting_value}#{padding}"
+          if title_so_far.present?
+            [title_so_far.strip, padding, non_sort_value].join
+          else
+            non_sort_value
+          end
+        end
+
         def non_sorting_padding(title, non_sorting_value)
           non_sort_note = title.note&.find { |note| note.type&.downcase == 'nonsorting character count' }
           if non_sort_note

data/lib/cocina/models/version.rb

@@ -2,6 +2,6 @@
 
 module Cocina
   module Models
-    VERSION = '0.94.0'
+    VERSION = '0.94.2'
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: cocina-models
 version: !ruby/object:Gem::Version
-  version: 0.94.0
+  version: 0.94.2
 platform: ruby
 authors:
 - Justin Coyne
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2024-01-19 00:00:00.000000000 Z
+date: 2024-01-26 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activesupport
@@ -562,7 +562,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.3.7
+rubygems_version: 3.4.13
 signing_key:
 specification_version: 4
 summary: Data models for the SDR