Finish 3.3.2

ruby-rdf · Jul 29, 2024 · 6527307 · 6527307
2 parents 85e64aa + bdc66c2
commit 6527307
Show file tree

Hide file tree

Showing 12 changed files with 368 additions and 43 deletions.
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -19,7 +19,7 @@ jobs:
     strategy:
       fail-fast: false
       matrix:
-        ruby: ['3.0', 3.1, 3.2, ruby-head, jruby]
+        ruby: ['3.0', 3.1, 3.2, 3.3, ruby-head, jruby]
     steps:
       - name: Clone repository
         uses: actions/checkout@v3

diff --git a/README.md b/README.md
@@ -603,6 +603,10 @@ To get a local working copy of the development repository, do:
 
     % git clone git://github.com/ruby-rdf/json-ld.git
 
+## Change Log
+
+See [Release Notes on GitHub](https://github.com/ruby-rdf/json-ld/releases)
+
 ## Mailing List
 * <https://lists.w3.org/Archives/Public/public-rdf-ruby/>
 

diff --git a/VERSION b/VERSION
@@ -1 +1 @@
-3.3.1
+3.3.2
diff --git a/example-files/vc1373-2.jsonld b/example-files/vc1373-2.jsonld
@@ -0,0 +1,62 @@
+{
+  "@context": [
+    {
+      "@protected": true,
+      "@vocab": "https://www.w3.org/ns/credentials/issuer-dependent#",
+      "id": "@id",
+      "type": "@type",
+      "VerifiablePresentation": {
+        "@id": "https://www.w3.org/2018/credentials#VerifiablePresentation",
+        "@context": {
+          "@protected": true,
+          "id": "@id",
+          "type": "@type",
+          "holder": {
+            "@id": "https://www.w3.org/2018/credentials#holder",
+            "@type": "@id"
+          },
+          "proof": {
+            "@id": "https://w3id.org/security#proof",
+            "@type": "@id",
+            "@container": "@graph"
+          },
+          "verifiableCredential": {
+            "@id": "https://www.w3.org/2018/credentials#verifiableCredential",
+            "@type": "@id",
+            "@container": "@graph",
+            "@context": null
+          },
+          "termsOfUse": {
+            "@id": "https://www.w3.org/2018/credentials#termsOfUse",
+            "@type": "@id"
+          }
+        }
+      },
+      "issuer": {
+        "@id": "https://www.w3.org/2018/credentials#issuer",
+        "@type": "@id",
+        "@context": {
+          "@protected": true,
+
+          "id": "@id",
+          "type": "@type",
+
+          "description": {
+            "@id": "https://schema.org/description",
+            "@context": {
+              "value": "@value", "lang": "@language", "dir": "@direction"
+            }
+          },
+          "name": {
+            "@id": "https://schema.org/name",
+            "@context": {
+              "value": "@value", "lang": "@language", "dir": "@direction"
+            }
+          }
+        }
+      }
+    }
+  ],
+  "type": "VerifiablePresentation",
+  "verifiableCredential": ["http://university.example/credentials/1872"]
+}
diff --git a/example-files/vc1373.jsonld b/example-files/vc1373.jsonld
@@ -0,0 +1,126 @@
+{
+  "@context": [
+    {
+      "@protected": true,
+      "@vocab": "https://www.w3.org/ns/credentials/issuer-dependent#",
+      "id": "@id",
+      "type": "@type",
+      "VerifiablePresentation": {
+        "@id": "https://www.w3.org/2018/credentials#VerifiablePresentation",
+        "@context": {
+          "@protected": true,
+          "id": "@id",
+          "type": "@type",
+          "holder": {
+            "@id": "https://www.w3.org/2018/credentials#holder",
+            "@type": "@id"
+          },
+          "proof": {
+            "@id": "https://w3id.org/security#proof",
+            "@type": "@id",
+            "@container": "@graph"
+          },
+          "verifiableCredential": {
+            "@id": "https://www.w3.org/2018/credentials#verifiableCredential",
+            "@type": "@id",
+            "@container": "@graph",
+            "@context": null
+          },
+          "termsOfUse": {
+            "@id": "https://www.w3.org/2018/credentials#termsOfUse",
+            "@type": "@id"
+          }
+        }
+      },
+      "issuer": {
+        "@id": "https://www.w3.org/2018/credentials#issuer",
+        "@type": "@id",
+        "@context": {
+          "@protected": true,
+
+          "id": "@id",
+          "type": "@type",
+
+          "description": {
+            "@id": "https://schema.org/description",
+            "@context": {
+              "value": "@value", "lang": "@language", "dir": "@direction"
+            }
+          },
+          "name": {
+            "@id": "https://schema.org/name",
+            "@context": {
+              "value": "@value", "lang": "@language", "dir": "@direction"
+            }
+          }
+        }
+      }
+    }
+  ],
+  "type": "VerifiablePresentation",
+  "verifiableCredential": [
+    {
+      "@context": [
+        {
+          "@protected": true,
+          "@vocab": "https://www.w3.org/ns/credentials/issuer-dependent#",
+          "id": "@id",
+          "type": "@type",
+          "VerifiablePresentation": {
+            "@id": "https://www.w3.org/2018/credentials#VerifiablePresentation",
+            "@context": {
+              "@protected": true,
+              "id": "@id",
+              "type": "@type",
+              "holder": {
+                "@id": "https://www.w3.org/2018/credentials#holder",
+                "@type": "@id"
+              },
+              "proof": {
+                "@id": "https://w3id.org/security#proof",
+                "@type": "@id",
+                "@container": "@graph"
+              },
+              "verifiableCredential": {
+                "@id": "https://www.w3.org/2018/credentials#verifiableCredential",
+                "@type": "@id",
+                "@container": "@graph",
+                "@context": null
+              },
+              "termsOfUse": {
+                "@id": "https://www.w3.org/2018/credentials#termsOfUse",
+                "@type": "@id"
+              }
+            }
+          },
+          "issuer": {
+            "@id": "https://www.w3.org/2018/credentials#issuer",
+            "@type": "@id",
+            "@context": {
+              "@protected": true,
+
+              "id": "@id",
+              "type": "@type",
+
+              "description": {
+                "@id": "https://schema.org/description",
+                "@context": {
+                  "value": "@value", "lang": "@language", "dir": "@direction"
+                }
+              },
+              "name": {
+                "@id": "https://schema.org/name",
+                "@context": {
+                  "value": "@value", "lang": "@language", "dir": "@direction"
+                }
+              }
+            }
+          }
+        }
+      ],
+      "id": "http://university.example/credentials/1872",
+      "type": "VerifiableCredential",
+      "issuer": "https://university.example/issuers/565049"
+    }
+  ]
+}
diff --git a/json-ld.gemspec b/json-ld.gemspec
@@ -37,6 +37,8 @@ Gem::Specification.new do |gem|
   gem.add_runtime_dependency     'multi_json',      '~> 1.15'
   gem.add_runtime_dependency     "rack",            '>= 2.2', '< 4'
   gem.add_runtime_dependency     'rdf',             '~> 3.3'
+  gem.add_runtime_dependency     'rexml',           '~> 3.2'
+  gem.add_development_dependency 'getoptlong',      '~> 0.2'
   gem.add_development_dependency 'jsonlint',        '~> 0.4'  unless is_java
   gem.add_development_dependency 'oj',              '~> 3.15' unless is_java
   gem.add_development_dependency 'rack-test',       '~> 2.1'

diff --git a/lib/json/ld/api.rb b/lib/json/ld/api.rb
@@ -290,8 +290,7 @@ def self.compact(input, context, expanded: false, serializer: nil, **options)
       def self.flatten(input, context, expanded: false, serializer: nil, **options)
         flattened = []
         options = {
-          compactToRelative: true,
-          extractAllScripts: true
+          compactToRelative: true
         }.merge(options)
 
         # Expand input to simplify processing
@@ -518,6 +517,8 @@ def self.frame(input, frame, expanded: false, serializer: nil, **options)
       # @option options (see #initialize)
       # @option options [Boolean] :produceGeneralizedRdf (false)
       #   If true, output will include statements having blank node predicates, otherwise they are dropped.
+      # @option options [Boolean] :extractAllScripts (true)
+      #   If set, when given an HTML input without a fragment identifier, extracts all `script` elements with type `application/ld+json` into an array during expansion.
       # @raise [JsonLdError]
       # @yield statement
       # @yieldparam [RDF::Statement] statement
@@ -638,7 +639,7 @@ def self.loadRemoteDocument(url,
             options[:headers]['Accept'].sub('application/ld+json,',
               "application/ld+json;profile=#{requestProfile}, application/ld+json;q=0.9,")
         end
-        documentLoader.call(url, **options) do |remote_doc|
+        documentLoader.call(url, extractAllScripts: extractAllScripts, **options) do |remote_doc|
           case remote_doc
           when RDF::Util::File::RemoteDocument
             # Convert to RemoteDocument
@@ -758,6 +759,28 @@ class << self
         alias fromRDF fromRdf
       end
 
+      ##
+      # Hash of recognized script types and the loaders that decode them
+      # into a hash or array of hashes.
+      #
+      # @return Hash{type, Proc}
+      SCRIPT_LOADERS = {
+        'application/ld+json' => ->(content, url:, **options) do
+            validate_input(content, url: url) if options[:validate]
+            mj_opts = options.keep_if { |k, v| k != :adapter || MUTLI_JSON_ADAPTERS.include?(v) }
+            MultiJson.load(content, **mj_opts)
+          end
+      }
+
+      ##
+      # Adds a loader for some specific content type
+      #
+      # @param [String] type
+      # @param [Proc] loader
+      def self.add_script_loader(type, loader)
+        SCRIPT_LOADERS[type] = loader
+      end
+
       ##
       # Load one or more script tags from an HTML source.
       # Unescapes and uncomments input, returns the internal representation
@@ -812,47 +835,52 @@ def self.load_html(input, url:,
           element = input.at_xpath("//script[@id='#{id}']")
           raise JSON::LD::JsonLdError::LoadingDocumentFailed, "No script tag found with id=#{id}" unless element
 
-          unless element.attributes['type'].to_s.start_with?('application/ld+json')
+          script_type = SCRIPT_LOADERS.keys.detect {|type| element.attributes['type'].to_s.start_with?(type)}
+          unless script_type
             raise JSON::LD::JsonLdError::LoadingDocumentFailed,
               "Script tag has type=#{element.attributes['type']}"
           end
 
-          content = element.inner_html
-          validate_input(content, url: url) if options[:validate]
-          mj_opts = options.keep_if { |k, v| k != :adapter || MUTLI_JSON_ADAPTERS.include?(v) }
-          MultiJson.load(content, **mj_opts)
+          loader = SCRIPT_LOADERS[script_type]
+          loader.call(element.inner_html, url: url, **options)
         elsif extractAllScripts
           res = []
-          elements = if profile
-            es = input.xpath("//script[starts-with(@type, 'application/ld+json;profile=#{profile}')]")
-            # If no profile script, just take a single script without profile
-            es = [input.at_xpath("//script[starts-with(@type, 'application/ld+json')]")].compact if es.empty?
-            es
-          else
-            input.xpath("//script[starts-with(@type, 'application/ld+json')]")
-          end
-          elements.each do |element|
-            content = element.inner_html
-            validate_input(content, url: url) if options[:validate]
-            mj_opts = options.keep_if { |k, v| k != :adapter || MUTLI_JSON_ADAPTERS.include?(v) }
-            r = MultiJson.load(content, **mj_opts)
-            if r.is_a?(Hash)
-              res << r
-            elsif r.is_a?(Array)
-              res.concat(r)
+
+          SCRIPT_LOADERS.each do |type, loader|
+            elements = if profile
+              es = input.xpath("//script[starts-with(@type, '#{type};profile=#{profile}')]")
+              # If no profile script, just take a single script without profile
+              es = [input.at_xpath("//script[starts-with(@type, '#{type}')]")].compact if es.empty?
+              es
+            else
+              input.xpath("//script[starts-with(@type, '#{type}')]")
+            end
+            elements.each do |element|
+              content = element.inner_html
+              r = loader.call(content, url: url, extractAllScripts: true, **options)
+              if r.is_a?(Hash)
+                res << r
+              elsif r.is_a?(Array)
+                res.concat(r)
+              end
             end
           end
           res
         else
-          # Find the first script with type application/ld+json.
-          element = input.at_xpath("//script[starts-with(@type, 'application/ld+json;profile=#{profile}')]") if profile
-          element ||= input.at_xpath("//script[starts-with(@type, 'application/ld+json')]")
-          raise JSON::LD::JsonLdError::LoadingDocumentFailed, "No script tag found" unless element
+          # Find the first script with a known type
+          script_type, element = nil, nil
+          SCRIPT_LOADERS.keys.each do |type|
+            next if script_type # already found the type
+            element = input.at_xpath("//script[starts-with(@type, '#{type};profile=#{profile}')]") if profile
+            element ||= input.at_xpath("//script[starts-with(@type, '#{type}')]")
+            script_type = type if element
+          end
+          unless script_type
+            raise JSON::LD::JsonLdError::LoadingDocumentFailed, "No script tag found" unless element
+          end
 
           content = element.inner_html
-          validate_input(content, url: url) if options[:validate]
-          mj_opts = options.keep_if { |k, v| k != :adapter || MUTLI_JSON_ADAPTERS.include?(v) }
-          MultiJson.load(content, **mj_opts)
+          SCRIPT_LOADERS[script_type].call(content, url: url, **options)
         end
       rescue MultiJson::ParseError => e
         raise JSON::LD::JsonLdError::InvalidScriptElement, e.message

diff --git a/lib/json/ld/compact.rb b/lib/json/ld/compact.rb
@@ -299,10 +299,10 @@ def compact(element,
                   else
                     index_key = context.expand_iri(index_key, vocab: true)
                     container_key = context.compact_iri(index_key, vocab: true)
-                    map_key, *others = Array(compacted_item[container_key])
+                    map_key, *others = Array(compacted_item.is_a?(Hash) && compacted_item[container_key])
                     if map_key.is_a?(String)
                       case others.length
-                      when 0 then compacted_item.delete(container_key)
+                      when 0 then compacted_item.delete(container_key) if compacted_item.is_a?(Hash)
                       when 1 then compacted_item[container_key] = others.first
                       else        compacted_item[container_key] = others
                       end
@@ -316,15 +316,15 @@ def compact(element,
                   map_key = expanded_item['@language']
                   value?(expanded_item) ? expanded_item['@value'] : compacted_item
                 elsif container.include?('@type')
-                  map_key, *types = Array(compacted_item[container_key])
+                  map_key, *types = Array(compacted_item.is_a?(Hash) && compacted_item[container_key])
                   case types.length
-                  when 0 then compacted_item.delete(container_key)
+                  when 0 then compacted_item.delete(container_key) if compacted_item.is_a?(Hash)
                   when 1 then compacted_item[container_key] = types.first
                   else        compacted_item[container_key] = types
                   end
 
                   # if compacted_item contains a single entry who's key maps to @id, then recompact the item without @type
-                  if compacted_item.keys.length == 1 && expanded_item.key?('@id')
+                  if compacted_item.is_a?(Hash) && compacted_item.keys.length == 1 && expanded_item.key?('@id')
                     compacted_item = compact({ '@id' => expanded_item['@id'] },
                       base: base,
                       property: item_active_property,