loc_mods 0.2.3 → 0.2.5

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 2b206d76220e77a18fd6609c3f7646eaf8fb836ac2d69a7d6da019db5820f24d
-  data.tar.gz: ffd23bb7508ad1fef1e0b165eb07620fc0a2420b6c436ac34cec8208affb6463
+  metadata.gz: a2a4762188e137e5ea3515cbff8dcbfa6e2749c7eb1ca4921d230e20c2c066db
+  data.tar.gz: a510dd995107ef4b33dabcc0c8e9658948e3c4d7474d50fd2cbd8d2eaf1b7e31
 SHA512:
-  metadata.gz: 9f56bbb4bfdab2dd2820ef67d94461ac5caa47d62937762a5cbeacbea8e70497c75968d92b3ce23f5d7e79f6a737c98c43058d6e11b774d0777340f040808fea
-  data.tar.gz: 0ac7feedcebe5df178f6ad313ba811004de126c82adc0f5289f44b43212a26ef2acfa3f09a45afa9cd4b158b7b4ef97976413ac357b4eb11749ab227c801a168
+  metadata.gz: 410d05bac6bde6ad26693359a2331c7b93ba4abcdabed33ad51f3f25dd9c08491be258ff0c9c2f4752f9b323cfd9b60102193af4535d30e92b7297679892439a
+  data.tar.gz: a70f2deb4dff1d8fd8f119e2d78b40330444caffa5f3d47fc6939ba51241a47fcbaf47eda8a64f35b002d7c7660c7f9c97debceda0c9019e2d8c7a13870785e2

data/.rubocop.yml

@@ -1,3 +1,5 @@
+inherit_from: .rubocop_todo.yml
+
 AllCops:
   TargetRubyVersion: 3.0
 

data/.rubocop_todo.yml

@@ -0,0 +1,83 @@
+# This configuration was generated by
+# `rubocop --auto-gen-config`
+# on 2024-08-15 09:25:52 UTC using RuboCop version 1.65.1.
+# The point is for the user to remove these configuration records
+# one by one as the offenses are removed from the code base.
+# Note that changes in the inspected code, or installation of new
+# versions of RuboCop, may require this file to be generated again.
+
+# Offense count: 1
+# Configuration parameters: Severity, Include.
+# Include: **/*.gemspec
+Gemspec/RequiredRubyVersion:
+  Exclude:
+    - 'loc_mods.gemspec'
+
+# Offense count: 4
+# Configuration parameters: AllowedMethods.
+# AllowedMethods: enums
+Lint/ConstantDefinitionInBlock:
+  Exclude:
+    - 'spec/loc_mods/collection_spec.rb'
+
+# Offense count: 8
+# Configuration parameters: AllowedMethods, AllowedPatterns, CountRepeatedAttributes.
+Metrics/AbcSize:
+  Max: 50
+
+# Offense count: 2
+# Configuration parameters: CountComments, CountAsOne, AllowedMethods, AllowedPatterns.
+# AllowedMethods: refine
+Metrics/BlockLength:
+  Max: 66
+
+# Offense count: 1
+# Configuration parameters: CountComments, CountAsOne.
+Metrics/ClassLength:
+  Max: 253
+
+# Offense count: 6
+# Configuration parameters: AllowedMethods, AllowedPatterns.
+Metrics/CyclomaticComplexity:
+  Max: 15
+
+# Offense count: 10
+# Configuration parameters: CountComments, CountAsOne, AllowedMethods, AllowedPatterns.
+Metrics/MethodLength:
+  Max: 43
+
+# Offense count: 5
+# Configuration parameters: AllowedMethods, AllowedPatterns.
+Metrics/PerceivedComplexity:
+  Max: 17
+
+# Offense count: 1
+# Configuration parameters: ForbiddenDelimiters.
+# ForbiddenDelimiters: (?i-mx:(^|\s)(EO[A-Z]{1}|END)(\s|$))
+Naming/HeredocDelimiterNaming:
+  Exclude:
+    - 'bin/update-nist-mods'
+
+# Offense count: 1
+Security/Open:
+  Exclude:
+    - 'bin/update-nist-mods'
+
+# Offense count: 65
+# Configuration parameters: AllowedConstants.
+Style/Documentation:
+  Enabled: false
+
+# Offense count: 1
+# Configuration parameters: AllowedMethods.
+# AllowedMethods: respond_to_missing?
+Style/OptionalBooleanParameter:
+  Exclude:
+    - 'lib/loc_mods/comparable_mapper.rb'
+
+# Offense count: 1
+# This cop supports safe autocorrection (--autocorrect).
+# Configuration parameters: AllowHeredoc, AllowURI, URISchemes, IgnoreCopDirectives, AllowedPatterns.
+# URISchemes: http, https
+Layout/LineLength:
+  Max: 126

data/README.adoc

@@ -28,15 +28,44 @@ LocMods::Collection.from_xml(File.read("reference/allrecords-MODS.xml"))
 
 === Command line interface
 
-LocMods provides a command-line interface (CLI) for various operations. The main
-executable is `loc-mods`.
+LocMods provides a command-line interface (CLI) for various operations.
+
+The main executable is `loc-mods`.
+
+[source,shell]
+----
+Commands:
+  loc-mods detect-duplicates PATH...  # Detect duplicate records in MODS XML files or directories
+  loc-mods help [COMMAND]             # Describe available commands or one specific command
+----
+
+
+
 
 ==== Detect duplicates
 
 The `detect-duplicates` command allows you to find duplicate MODS records based
 on using a "primary ID" that is their DOI (Digital Object Identifier).
 
+NOTE: The library assumes that every record has a DOI. If that is not the case,
+another way to setting the primary key needs to be defined.
+
+Usage:
+
+[source,shell]
+----
 Usage:
+  loc-mods detect-duplicates PATH...
+
+Options:
+  [--show-unchanged], [--no-show-unchanged] # Show unchanged attributes in the diff output
+                                            # Default: false
+  [--highlight-diff], [--no-highlight-diff] # Highlight only the differences
+                                            # Default: false
+  [--color=COLOR]                           # Use colors in the diff output (auto, on, off)
+                                            # Default: auto
+                                            # Possible values: auto, on, off
+----
 
 [source,shell]
 ----
@@ -45,15 +74,28 @@ $ loc-mods detect-duplicates [OPTIONS] <file_or_directory_path>
 
 Options:
 
-* `-r, --recursive`: Search for MODS files recursively in subdirectories.
-* `-v, --verbose`: Display more detailed output.
-* `-o, --output FILE`: Write the results to a file instead of stdout.
+`--show-unchanged`::
+(default: `false`)
+Show attributes of both objects even when they were not changed.
+
+`--highlight-diff`::
+(default: `false`)
+Highlight values only when they differ between two records.
+
+`--color=COLOR`::
+(default: `auto`) Use colors in the diff output. Values:
+
+`auto`::: the CLI will detect whether the terminal supports colors and display
+with colors if it does.
+`on`::: the CLI will always display with colors.
+`off`::: the CLI will never display with colors.
+
 
 Example:
 
 [source,shell]
 ----
-$ loc-mods detect-duplicates -r -v /path/to/mods/files
+$ loc-mods detect-duplicates  /path/to/mods/files
 ----
 
 This command will:

data/lib/loc_mods/abstract.rb

@@ -1,16 +1,16 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 module LocMods
-  class Abstract < BaseMapper
-    attribute :content, Shale::Type::String
-    attribute :display_label, Shale::Type::String
-    attribute :type, Shale::Type::String
-    attribute :shareable, Shale::Type::Value
-    attribute :alt_rep_group, Shale::Type::String
-    attribute :alt_format, Shale::Type::Value
-    attribute :content_type, Shale::Type::String
+  class Abstract < Lutaml::Model::Serializable
+    attribute :content, :string
+    attribute :display_label, :string
+    attribute :type, :string
+    attribute :shareable, :string
+    attribute :alt_rep_group, :string
+    attribute :alt_format, :string
+    attribute :content_type, :string
 
     xml do
       root "abstract"

data/lib/loc_mods/access_condition.rb

@@ -1,18 +1,18 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 module LocMods
-  class AccessCondition < BaseMapper
-    attribute :content, Shale::Type::String
-    attribute :display_label, Shale::Type::String
-    attribute :lang, Shale::Type::String
-    attribute :script, Shale::Type::String
-    attribute :transliteration, Shale::Type::String
-    attribute :type, Shale::Type::String
-    attribute :alt_rep_group, Shale::Type::String
-    attribute :alt_format, Shale::Type::Value
-    attribute :content_type, Shale::Type::String
+  class AccessCondition < Lutaml::Model::Serializable
+    attribute :content, :string
+    attribute :display_label, :string
+    attribute :lang, :string
+    attribute :script, :string
+    attribute :transliteration, :string
+    attribute :type, :string
+    attribute :alt_rep_group, :string
+    attribute :alt_format, :string
+    attribute :content_type, :string
 
     xml do
       root "accessCondition"

data/lib/loc_mods/alternative_name.rb

@@ -1,23 +1,23 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 require_relative "identifier"
 require_relative "name_part"
 require_relative "role"
 
 module LocMods
-  class AlternativeName < BaseMapper
-    attribute :lang, Shale::Type::String
-    attribute :script, Shale::Type::String
-    attribute :transliteration, Shale::Type::String
-    attribute :display_label, Shale::Type::String
-    attribute :alt_type, Shale::Type::String
+  class AlternativeName < Lutaml::Model::Serializable
+    attribute :lang, :string
+    attribute :script, :string
+    attribute :transliteration, :string
+    attribute :display_label, :string
+    attribute :alt_type, :string
     attribute :name_part, NamePart, collection: true
-    attribute :display_form, Shale::Type::String, collection: true
-    attribute :affiliation, Shale::Type::String, collection: true
+    attribute :display_form, :string, collection: true
+    attribute :affiliation, :string, collection: true
     attribute :role, Role, collection: true
-    attribute :description, Shale::Type::String, collection: true
+    attribute :description, :string, collection: true
     attribute :name_identifier, Identifier, collection: true
 
     xml do

data/lib/loc_mods/area.rb

@@ -1,13 +1,13 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 require_relative "hierarchical_part"
 
 module LocMods
-  class Area < BaseMapper
+  class Area < Lutaml::Model::Serializable
     attribute :content, HierarchicalPart
-    attribute :area_type, Shale::Type::Value
+    attribute :area_type, :string
 
     xml do
       root "area"

data/lib/loc_mods/cartographic_extension.rb

@@ -1,11 +1,11 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 module LocMods
-  class CartographicExtension < BaseMapper
-    attribute :content, Shale::Type::String
-    attribute :display_label, Shale::Type::String
+  class CartographicExtension < Lutaml::Model::Serializable
+    attribute :content, :string
+    attribute :display_label, :string
 
     xml do
       root "cartographicExtension"

data/lib/loc_mods/cartographics.rb

@@ -1,17 +1,17 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 require_relative "cartographic_extension"
 
 module LocMods
-  class Cartographics < BaseMapper
-    attribute :authority, Shale::Type::String
-    attribute :authority_uri, Shale::Type::Value
-    attribute :value_uri, Shale::Type::Value
-    attribute :scale, Shale::Type::String
-    attribute :projection, Shale::Type::String
-    attribute :coordinates, Shale::Type::String, collection: true
+  class Cartographics < Lutaml::Model::Serializable
+    attribute :authority, :string
+    attribute :authority_uri, :string
+    attribute :value_uri, :string
+    attribute :scale, :string
+    attribute :projection, :string
+    attribute :coordinates, :string, collection: true
     attribute :cartographic_extension, CartographicExtension, collection: true
 
     xml do

data/lib/loc_mods/city_section.rb

@@ -1,13 +1,13 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 require_relative "hierarchical_part"
 
 module LocMods
-  class CitySection < BaseMapper
+  class CitySection < Lutaml::Model::Serializable
     attribute :content, HierarchicalPart
-    attribute :city_section_type, Shale::Type::Value
+    attribute :city_section_type, :string
 
     xml do
       root "citySection"

data/lib/loc_mods/classification.rb

@@ -1,15 +1,15 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 module LocMods
-  class Classification < BaseMapper
-    attribute :content, Shale::Type::String
-    attribute :edition, Shale::Type::String
-    attribute :display_label, Shale::Type::String
-    attribute :alt_rep_group, Shale::Type::String
-    attribute :usage, Shale::Type::Value
-    attribute :generator, Shale::Type::String
+  class Classification < Lutaml::Model::Serializable
+    attribute :content, :string
+    attribute :edition, :string
+    attribute :display_label, :string
+    attribute :alt_rep_group, :string
+    attribute :usage, :string
+    attribute :generator, :string
 
     xml do
       root "classification"

data/lib/loc_mods/cli.rb

@@ -1,12 +1,15 @@
 # frozen_string_literal: true
 
-# lib/loc_mods/cli.rb
 require "thor"
 require "loc_mods"
 
 module LocMods
   class Cli < Thor
     desc "detect-duplicates PATH...", "Detect duplicate records in MODS XML files or directories"
+    method_option :show_unchanged, type: :boolean, default: false, desc: "Show unchanged attributes in the diff output"
+    method_option :highlight_diff, type: :boolean, default: false, desc: "Highlight only the differences"
+    method_option :color, type: :string, enum: %w[auto on off], default: "auto",
+                          desc: "Use colors in the diff output (auto, on, off)"
 
     def detect_duplicates(*paths)
       all_records = []
@@ -46,12 +49,13 @@ module LocMods
           puts "  Comparison #{index + 1}:"
           puts "  File 1: #{record1[:file]}"
           puts "  File 2: #{record2[:file]}"
-          differences = record1[:record].compare(record2[:record])
-          if differences
-            puts "  ----"
-            print_differences(differences)
-            puts "  ----"
-          end
+          print_differences(
+            record1[:record],
+            record2[:record],
+            options[:show_unchanged],
+            options[:highlight_diff],
+            color_enabled?
+          )
           puts "\n"
         end
       end
@@ -67,87 +71,51 @@ module LocMods
       end
     end
 
-    def print_differences(differences, prefix = "", path = [])
-      if differences.is_a?(Comparison)
-        print_difference(differences, path)
-        return
-      end
-
-      # Differences is a Hash here
-      differences.each do |key, value|
-        # puts "key #{key}, value #{value}"
-        current_path = path + [key]
-        # This is a comparison
-        if value.is_a?(Comparison)
-          print_difference(value, path)
-          next
-        end
+    def print_differences(record1, record2, show_unchanged, highlight_diff, use_colors)
+      diff_score, diff_tree = Lutaml::Model::Serialize.diff_with_score(
+        record1,
+        record2,
+        show_unchanged: show_unchanged,
+        highlight_diff: highlight_diff,
+        use_colors: use_colors,
+        indent: "  ",
+      )
+      similarity_percentage = (1 - diff_score) * 100
+
+      puts "  Differences:"
+      puts diff_tree
+      puts "  Similarity score: #{similarity_percentage.round(2)}%"
+    end
 
-        raise "Differences must be in form of a Hash" unless value.is_a?(Hash)
+    def color_enabled?
+      case options[:color]
+      when "on"
+        true
+      when "off"
+        false
+      else
+        supports_color?
+      end
+    end
 
-        # This is not array, end here
-        next unless value.keys.any? do |k|
-          k.is_a?(Integer) || k == :_array_size_difference
-        end
+    def supports_color?
+      return false unless $stdout.tty?
 
-        # if value[:_array_size_difference]
-        #   puts "  #{format_path(current_path)} [_array_size_difference]:"
-        #   puts "    Record 1: #{format_value(value[:_array_size_difference].original)}"
-        #   puts "    Record 2: #{format_value(value[:_array_size_difference].updated)}"
-        #   puts
-        # end
-
-        value.each do |subkey, subvalue|
-          # if [:self, :other].include?(subkey)
-          #   raise "Subkey is self or other"
-          # end
-          # puts "subkey (#{subkey})"
-          # puts "subvalue #{subvalue}, prefix #{prefix}, current_path + [subkey] #{current_path + [subkey]}"
-
-          if subkey.is_a?(Integer)
-            print_differences(subvalue, prefix, current_path + [subkey])
-          else
-            if subvalue.is_a?(Comparison)
-              print_difference(subvalue, current_path + [subkey])
-              next
-            end
-
-            raise "In an array diff but not an Integer!"
-          end
-        end
+      if RbConfig::CONFIG["host_os"] =~ /mswin|mingw|cygwin/
+        return true if ENV["ANSICON"]
+        return true if ENV["ConEmuANSI"] == "ON"
+        return true if ENV["TERM"] == "xterm"
       end
-    end
 
-    def print_difference(value, current_path)
-      return unless value.original || value.updated
+      return true if ENV["COLORTERM"]
 
-      puts "  #{format_path(current_path)}:"
-      puts "    Record 1: #{format_value(value.original)}"
-      puts "    Record 2: #{format_value(value.updated)}"
-      puts
-    end
+      term = ENV["TERM"]
+      return false if term.nil? || term.empty?
 
-    def format_path(path)
-      path.map.with_index do |part, index|
-        if index.zero?
-          part.to_s
-        elsif part.is_a?(Integer)
-          "[#{part}]"
-        else
-          ".#{part}"
-        end
-      end.join
-    end
+      color_terms = %w[ansi color console cygwin gnome konsole kterm
+                       linux msys putty rxvt screen tmux vt100 xterm]
 
-    def format_value(value)
-      case value
-      when nil, ComparableNil
-        "(nil)"
-      when String
-        "\"#{value}\""
-      else
-        value.to_s
-      end
+      color_terms.any? { |ct| term.include?(ct) }
     end
   end
 end

data/lib/loc_mods/collection.rb

@@ -1,13 +1,13 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 require_relative "record"
 
 module LocMods
-  class Collection < BaseMapper
+  class Collection < Lutaml::Model::Serializable
     attribute :mods, Record, collection: true
-    attribute :schema_location, Shale::Type::String
+    attribute :schema_location, :string
 
     xml do
       root "modsCollection"

data/lib/loc_mods/copy_information.rb

@@ -1,6 +1,6 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 require_relative "enumeration_and_chronology"
 require_relative "form"
@@ -8,11 +8,11 @@ require_relative "item_identifier"
 require_relative "note"
 
 module LocMods
-  class CopyInformation < BaseMapper
+  class CopyInformation < Lutaml::Model::Serializable
     attribute :form, Form
-    attribute :sub_location, Shale::Type::String, collection: true
-    attribute :shelf_locator, Shale::Type::String, collection: true
-    attribute :electronic_locator, Shale::Type::String, collection: true
+    attribute :sub_location, :string, collection: true
+    attribute :shelf_locator, :string, collection: true
+    attribute :electronic_locator, :string, collection: true
     attribute :note, Note, collection: true
     attribute :enumeration_and_chronology, EnumerationAndChronology,
               collection: true

data/lib/loc_mods/date.rb

@@ -1,15 +1,15 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 module LocMods
-  class Date < BaseMapper
-    attribute :content, Shale::Type::String
-    attribute :encoding, Shale::Type::String
-    attribute :qualifier, Shale::Type::String
-    attribute :point, Shale::Type::String
-    attribute :key_date, Shale::Type::Value
-    attribute :calendar, Shale::Type::String
+  class Date < Lutaml::Model::Serializable
+    attribute :content, :string
+    attribute :encoding, :string
+    attribute :qualifier, :string
+    attribute :point, :string
+    attribute :key_date, :string
+    attribute :calendar, :string
 
     xml do
       root "recordChangeDate"

data/lib/loc_mods/date_other.rb

@@ -1,13 +1,13 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 require_relative "date"
 
 module LocMods
-  class DateOther < BaseMapper
-    attribute :content, Shale::Type::String
-    attribute :type, Shale::Type::String
+  class DateOther < Lutaml::Model::Serializable
+    attribute :content, :string
+    attribute :type, :string
 
     xml do
       root "dateOther"

data/lib/loc_mods/detail.rb

@@ -1,14 +1,14 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 module LocMods
-  class Detail < BaseMapper
-    attribute :type, Shale::Type::String
-    attribute :level, Shale::Type::Integer
-    attribute :number, Shale::Type::String, collection: true
-    attribute :caption, Shale::Type::String, collection: true
-    attribute :title, Shale::Type::String, collection: true
+  class Detail < Lutaml::Model::Serializable
+    attribute :type, :string
+    attribute :level, :integer
+    attribute :number, :string, collection: true
+    attribute :caption, :string, collection: true
+    attribute :title, :string, collection: true
 
     xml do
       root "detail"

data/lib/loc_mods/edition.rb

@@ -1,12 +1,12 @@
 # frozen_string_literal: true
 
-require_relative "base_mapper"
+require "lutaml/model"
 
 module LocMods
   # LocMods::StringPlusLanguagePlusSupplied
-  class Edition < BaseMapper
-    attribute :content, Shale::Type::String
-    attribute :supplied, Shale::Type::String
+  class Edition < Lutaml::Model::Serializable
+    attribute :content, :string
+    attribute :supplied, :string
 
     xml do
       root "edition"