cocina-models 0.81.0 → 0.82.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA256:
-  metadata.gz: 0d8392fadece0bb4ec12ed93e958c99abf04e29ad0a78c14ee8651ec8ecd1e37
-  data.tar.gz: db4a5de344bb53d98b2b71f725e8b1d1b9e0d5e0b28416989ad55f19df826376
+  metadata.gz: ebc24ff8dbe8eb1fe740eb977abb73abc504409aa510905395e6660adcd05b3b
+  data.tar.gz: 62ad4e7cafce859250323e92cc2173257957b16775e767224a4de2e9fd639bae
 SHA512:
-  metadata.gz: 4880e4eb86935116e0bec017926aeab5e5aa12d8792ee5066d0aeef5f514e6f13f63c8bc68c03035fdff2748f95aa7215f236384e109a8e532bd76e9677ae53c
-  data.tar.gz: 22154804ee4b103bdbd3aaaee449327bb4a6c9bc1dce9431a5311ed06c5e76039ccb420634d275f56d644a60ffa529bd9cb822f9c549be057f4972d791a34bf7
+  metadata.gz: 7a3b0ed8ab383f945450d42343dbaed6e709854ff0f3deca1f62b93c79883d3157b42f14cdaafd9392ef858620a0c58ae6867c16037c627499efddf3cb33880b
+  data.tar.gz: 9e1084ea73c6f547eab7b8a076d9e2a61a46361c9da1541effca8529d48d032621b6cd5868bd2886308aea1f834b4c586ae1058ca6f67f0315393e4eb632bbb9

data/.rubocop_todo.yml

@@ -1,6 +1,6 @@
 # This configuration was generated by
 # `rubocop --auto-gen-config --auto-gen-only-exclude`
-# on 2022-04-27 19:20:35 UTC using RuboCop version 1.28.2.
+# on 2022-06-08 23:20:02 UTC using RuboCop version 1.28.2.
 # 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
@@ -35,7 +35,7 @@ Lint/UnusedMethodArgument:
     - 'lib/cocina/models/mapping/from_mods/subject.rb'
     - 'lib/cocina/models/mapping/to_mods/event.rb'
 
-# Offense count: 95
+# Offense count: 96
 # Configuration parameters: IgnoredMethods, CountRepeatedAttributes.
 Metrics/AbcSize:
   Max: 40
@@ -56,16 +56,17 @@ Metrics/ParameterLists:
 RSpec/DescribeClass:
   Enabled: false
 
-# Offense count: 89
+# Offense count: 87
 # Configuration parameters: CountAsOne.
 RSpec/ExampleLength:
   Max: 128
 
-# Offense count: 9
+# Offense count: 10
 # Configuration parameters: Max.
 RSpec/NestedGroups:
   Exclude:
     - 'spec/cocina/models/mapping/normalizers/mods/origin_info_normalizer_spec.rb'
+    - 'spec/cocina/models/validators/date_time_validator_spec.rb'
 
 # Offense count: 1
 # Configuration parameters: MinBodyLength, AllowConsecutiveConditionals.
@@ -87,7 +88,7 @@ Style/MultilineBlockChain:
     - 'lib/cocina/models/mapping/to_mods/form.rb'
     - 'lib/cocina/models/mapping/to_mods/subject.rb'
 
-# Offense count: 206
+# Offense count: 223
 # This cop supports safe auto-correction (--auto-correct).
 # Configuration parameters: AllowHeredoc, AllowURI, URISchemes, IgnoreCopDirectives, AllowedPatterns, IgnoredPatterns.
 # URISchemes: http, https

data/cocina-models.gemspec

@@ -28,12 +28,15 @@ Gem::Specification.new do |spec|
   spec.add_dependency 'deprecation'
   spec.add_dependency 'dry-struct', '~> 1.0'
   spec.add_dependency 'dry-types', '~> 1.1'
+  spec.add_dependency 'edtf' # used for date/time validation
   spec.add_dependency 'equivalent-xml' # for diffing MODS
+  spec.add_dependency 'jsonpath' # used for date/time validation
   spec.add_dependency 'nokogiri'
   spec.add_dependency 'openapi3_parser' # Parsing openapi doc
   # Match these version requirements to what committee wants,
   # so that our client (non-committee) users have the same dependencies.
   spec.add_dependency 'openapi_parser', '>= 0.11.1', '< 1.0'
+  spec.add_dependency 'rss' # used for date/time validation
   spec.add_dependency 'super_diff'
   spec.add_dependency 'thor'
   spec.add_dependency 'zeitwerk', '~> 2.1'

data/lib/cocina/models/mapping/escape_html.rb

@@ -0,0 +1,23 @@
+# frozen_string_literal: true
+
+module Cocina
+  module Models
+    module Mapping
+      # Escaps HTML entities as CDATA for MODS since HTML is not permitted in MODS
+      class EscapeHtml
+        # @param [String] data
+        # @param [Nokogiri::XML::Builder]
+        def self.with_cdata(data, builder)
+          tokens = data.split(%r{(</?(?:i|cite)>)})
+          tokens.map do |token|
+            if /\A<.+>\z/.match? token
+              builder.cdata(token)
+            else
+              builder.text(token)
+            end
+          end
+        end
+      end
+    end
+  end
+end

data/lib/cocina/models/mapping/to_mods/note.rb

@@ -68,7 +68,9 @@ module Cocina
                     else
                       note.value
                     end
-            xml.public_send tag_name, value, attributes
+            xml.public_send tag_name, attributes do |builder|
+              EscapeHtml.with_cdata(value, builder) if value
+            end
           end
 
           def write_basic(note)

data/lib/cocina/models/validators/date_time_validator.rb

@@ -0,0 +1,102 @@
+# frozen_string_literal: true
+
+require 'edtf'
+require 'jsonpath'
+require 'rss'
+
+module Cocina
+  module Models
+    module Validators
+      # Validates that dates of known types are type-valid
+      class DateTimeValidator
+        VALIDATABLE_TYPES = %w[edtf iso8601 w3cdtf].freeze
+
+        def self.validate(_clazz, attributes)
+          new(attributes).validate
+        end
+
+        def initialize(attributes)
+          @attributes = attributes
+        end
+
+        def validate
+          return unless meets_preconditions?
+
+          return if invalid_dates.empty?
+
+          raise ValidationError, "Invalid date(s) for #{druid}: #{invalid_dates}"
+        end
+
+        private
+
+        attr_reader :attributes
+
+        def meets_preconditions?
+          attributes.key?(:description)
+        end
+
+        def invalid_dates
+          @invalid_dates ||= validatable_dates.filter_map do |date_hash|
+            code = date_hash.dig('encoding', 'code')
+            bad_values = JsonPath.new('$..value').on(date_hash.to_json).reject do |value|
+              send("valid_#{code}?", value)
+            end
+
+            next if bad_values.empty?
+
+            [*bad_values, code]
+          end
+        end
+
+        def validatable_dates
+          # Why is the `uniq` needed below? Odd behavior when handling highly nested use cases:
+          #
+          # > JsonPath.new("$..date..[?(@.encoding.code =~ /#{VALIDATABLE_TYPES.join('|')}/)]").on(attributes[:description].to_json)
+          # > [
+          #     {"structuredValue"=>[{"value"=>"1996", "type"=>"start"}, {"value"=>"1998", "type"=>"end"}], "encoding"=>{"code"=>"iso8601"}},
+          #     {"structuredValue"=>[{"value"=>"1996", "type"=>"start"}, {"value"=>"1998", "type"=>"end"}], "encoding"=>{"code"=>"iso8601"}}
+          # ]
+          #
+          # Notice how the JSONPath expression returns the *same exact* structure twice despite only being present once in the data.
+          JsonPath
+            .new("$..date..[?(@.encoding.code =~ /#{VALIDATABLE_TYPES.join('|')}/)]")
+            .on(attributes[:description].to_json)
+            .uniq
+        end
+
+        def valid_edtf?(value)
+          Date.edtf!(value)
+          true
+        rescue StandardError
+          false
+        end
+
+        def valid_iso8601?(value)
+          DateTime.iso8601(value)
+          true
+        rescue StandardError
+          false
+        end
+
+        def valid_w3cdtf?(value)
+          Time.w3cdtf(value)
+          true
+        rescue StandardError
+          # NOTE: the upstream W3CDTF implementation in the `rss` gem does not
+          #       allow two patterns that should be valid per the specification:
+          #
+          # * YYYY
+          # * YYYY-MM
+          #
+          # So we catch the false positives from the upstream gem and allow
+          # these two patterns to validate
+          /\A\d{4}(-0[1-9]|1[0-2])?\Z/.match?(value)
+        end
+
+        def druid
+          @druid ||= attributes[:externalIdentifier]
+        end
+      end
+    end
+  end
+end

data/lib/cocina/models/version.rb

@@ -2,6 +2,6 @@
 
 module Cocina
   module Models
-    VERSION = '0.81.0'
+    VERSION = '0.82.0'
   end
 end

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: cocina-models
 version: !ruby/object:Gem::Version
-  version: 0.81.0
+  version: 0.82.0
 platform: ruby
 authors:
 - Justin Coyne
 autorequire:
 bindir: exe
 cert_chain: []
-date: 2022-06-01 00:00:00.000000000 Z
+date: 2022-06-17 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: activesupport
@@ -66,6 +66,20 @@ dependencies:
     - - "~>"
       - !ruby/object:Gem::Version
         version: '1.1'
+- !ruby/object:Gem::Dependency
+  name: edtf
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: equivalent-xml
   requirement: !ruby/object:Gem::Requirement
@@ -80,6 +94,20 @@ dependencies:
     - - ">="
       - !ruby/object:Gem::Version
         version: '0'
+- !ruby/object:Gem::Dependency
+  name: jsonpath
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: nokogiri
   requirement: !ruby/object:Gem::Requirement
@@ -128,6 +156,20 @@ dependencies:
     - - "<"
       - !ruby/object:Gem::Version
         version: '1.0'
+- !ruby/object:Gem::Dependency
+  name: rss
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: '0'
 - !ruby/object:Gem::Dependency
   name: super_diff
   requirement: !ruby/object:Gem::Requirement
@@ -377,6 +419,7 @@ files:
 - lib/cocina/models/location_based_access.rb
 - lib/cocina/models/location_based_download_access.rb
 - lib/cocina/models/mapping/error_notifier.rb
+- lib/cocina/models/mapping/escape_html.rb
 - lib/cocina/models/mapping/from_mods/access.rb
 - lib/cocina/models/mapping/from_mods/admin_metadata.rb
 - lib/cocina/models/mapping/from_mods/alt_rep_group.rb
@@ -461,6 +504,7 @@ files:
 - lib/cocina/models/validators/associated_name_validator.rb
 - lib/cocina/models/validators/catalog_links_validator.rb
 - lib/cocina/models/validators/dark_validator.rb
+- lib/cocina/models/validators/date_time_validator.rb
 - lib/cocina/models/validators/description_types_validator.rb
 - lib/cocina/models/validators/description_values_validator.rb
 - lib/cocina/models/validators/open_api_validator.rb
@@ -492,7 +536,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
-rubygems_version: 3.3.9
+rubygems_version: 3.3.7
 signing_key:
 specification_version: 4
 summary: Data models for the SDR