Use internalRepresentation option in fromRdf to preserve raw RDF literals as values.

Author: gkellogg

lib/json/ld/api.rb

@@ -66,6 +66,7 @@ class API
     # @param [String, #read, Hash, Array, JSON::LD::Context] context
     #   An external context to use additionally to the context embedded in input when expanding the input.
     # @param  [Hash{Symbol => Object}] options
+    # @option options [Symbol] :adapter used with MultiJson
     # @option options [RDF::URI, String, #to_s] :base
     #   The Base IRI to use when expanding the document. This overrides the value of `input` if it is a _IRI_. If not specified and `input` is not an _IRI_, the base IRI defaults to the current document IRI if in a browser context, or the empty string if there is no document context. If not specified, and a base IRI is found from `input`, options[:base] will be modified with this value.
     # @option options [Boolean] :compactArrays (true)
@@ -74,10 +75,10 @@ class API
     #   Creates document relative IRIs when compacting, if `true`, otherwise leaves expanded.
     # @option options [Proc] :documentLoader
     #   The callback of the loader to be used to retrieve remote documents and contexts. If specified, it must be used to retrieve remote documents and contexts; otherwise, if not specified, the processor's built-in loader must be used. See {documentLoader} for the method signature.
-    # @option options [Boolean] :lowercaseLanguage
-    #   By default, language tags are left as is. To normalize to lowercase, set this option to `true`.
     # @option options [String, #read, Hash, Array, JSON::LD::Context] :expandContext
     #   A context that is used to initialize the active context when expanding a document.
+    # @option options [Boolean] :extendedRepresentation (false)
+    #   Use the extended internal representation.
     # @option options [Boolean] :extractAllScripts
     #   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.
     # @option options [Boolean, String, RDF::URI] :flatten
@@ -86,19 +87,19 @@ class API
     #   When set, this has the effect of inserting a context definition with `@language` set to the associated value, creating a default language for interpreting string values.
     # @option options [Symbol] :library
     #   One of :nokogiri or :rexml. If nil/unspecified uses :nokogiri if available, :rexml otherwise.
+    # @option options [Boolean] :lowercaseLanguage
+    #   By default, language tags are left as is. To normalize to lowercase, set this option to `true`.
+    # @option options [Boolean] :ordered (true)
+    #   Order traversal of dictionary members by key when performing algorithms.
     # @option options [String] :processingMode
     #   Processing mode, json-ld-1.0 or json-ld-1.1.
-    #   If `processingMode` is not specified, a mode of `json-ld-1.0` or `json-ld-1.1` is set, the context used for `expansion` or `compaction`.
-    # @option options [Boolean] rdfstar      (false)
+    # @option options [Boolean] :rdfstar      (false)
     #   support parsing JSON-LD-star statement resources.
     # @option options [Boolean] :rename_bnodes (true)
     #   Rename bnodes as part of expansion, or keep them the same.
     # @option options [Boolean]  :unique_bnodes   (false)
     #   Use unique bnode identifiers, defaults to using the identifier which the node was originally initialized with (if any).
-    # @option options [Symbol] :adapter used with MultiJson
     # @option options [Boolean] :validate Validate input, if a string or readable object.
-    # @option options [Boolean] :ordered (true)
-    #   Order traversal of dictionary members by key when performing algorithms.
     # @yield [api]
     # @yieldparam [API]
     # @raise [JsonLdError]
@@ -562,6 +563,7 @@ def self.fromRdf(input, useRdfType: false, useNativeTypes: false, serializer: ni
 
       API.new(nil, nil, **options) do
         result = from_statements(input,
+          extendedRepresentation: options[:extendedRepresentation],
           useRdfType: useRdfType,
           useNativeTypes: useNativeTypes)
       end

lib/json/ld/from_rdf.rb

@@ -14,9 +14,11 @@ module FromRDF
     # @param [Boolean] useRdfType (false)
     #   If set to `true`, the JSON-LD processor will treat `rdf:type` like a normal property instead of using `@type`.
     # @param [Boolean] useNativeTypes (false) use native representations
+    # @param extendedRepresentation (false)
+    #   Use the extended internal representation for native types.
     #
     # @return [Array<Hash>] the JSON-LD document in normalized form
-    def from_statements(dataset, useRdfType: false, useNativeTypes: false)
+    def from_statements(dataset, useRdfType: false, useNativeTypes: false, extendedRepresentation: false)
       default_graph = {}
       graph_map = {'@default' => default_graph}
       referenced_once = {}
@@ -41,9 +43,9 @@ def from_statements(dataset, useRdfType: false, useNativeTypes: false)
         default_graph[name] ||= {'@id' => name} unless name == '@default'
 
         subject = statement.subject.statement? ?
-          resource_representation(statement.subject, useNativeTypes)['@id'].to_json_c14n :
+          resource_representation(statement.subject, useNativeTypes, extendedRepresentation)['@id'].to_json_c14n :
           statement.subject.to_s
-        node = node_map[subject] ||= resource_representation(statement.subject, useNativeTypes)
+        node = node_map[subject] ||= resource_representation(statement.subject, useNativeTypes, extendedRepresentation)
 
         # If predicate is rdf:datatype, note subject in compound literal subjects map
         if @options[:rdfDirection] == 'compound-literal' && statement.predicate == RDF.to_uri + 'direction'
@@ -53,10 +55,10 @@ def from_statements(dataset, useRdfType: false, useNativeTypes: false)
         # If object is an IRI, blank node identifier, or statement, and node map does not have an object member, create one and initialize its value to a new JSON object consisting of a single member @id whose value is set to object.
         unless statement.object.literal?
           object = statement.object.statement? ?
-            resource_representation(statement.object, useNativeTypes)['@id'].to_json_c14n :
+            resource_representation(statement.object, useNativeTypes, extendedRepresentation)['@id'].to_json_c14n :
             statement.object.to_s
           node_map[object] ||=
-            resource_representation(statement.object, useNativeTypes)
+            resource_representation(statement.object, useNativeTypes, extendedRepresentation)
         end
 
         # If predicate equals rdf:type, and object is an IRI or blank node identifier, append object to the value of the @type member of node. If no such member exists, create one and initialize it to an array whose only item is object. Finally, continue to the next RDF triple.
@@ -66,7 +68,7 @@ def from_statements(dataset, useRdfType: false, useNativeTypes: false)
         end
 
         # Set value to the result of using the RDF to Object Conversion algorithm, passing object, rdfDirection, and use native types.
-        value = resource_representation(statement.object, useNativeTypes)
+        value = resource_representation(statement.object, useNativeTypes, extendedRepresentation)
 
         merge_value(node, statement.predicate.to_s, value)
 
@@ -171,18 +173,18 @@ def from_statements(dataset, useRdfType: false, useNativeTypes: false)
 
     RDF_LITERAL_NATIVE_TYPES = Set.new([RDF::XSD.boolean, RDF::XSD.integer, RDF::XSD.double]).freeze
 
-    def resource_representation(resource, useNativeTypes)
+    def resource_representation(resource, useNativeTypes, extendedRepresentation)
       case resource
       when RDF::Statement
         # Note, if either subject or object are a BNode which is used elsewhere,
         # this might not work will with the BNode accounting from above.
-        rep = {'@id' => resource_representation(resource.subject, false)}
+        rep = {'@id' => resource_representation(resource.subject, false, extendedRepresentation)}
         if resource.predicate == RDF.type
           rep['@id'].merge!('@type' => resource.object.to_s)
         else
           rep['@id'].merge!(
             resource.predicate.to_s =>
-            as_array(resource_representation(resource.object, useNativeTypes)))
+            as_array(resource_representation(resource.object, useNativeTypes, extendedRepresentation)))
         end
         rep
       when RDF::Literal
@@ -197,6 +199,8 @@ def resource_representation(resource, useNativeTypes)
           rescue ::JSON::ParserError => e
             raise JSON::LD::JsonLdError::InvalidJsonLiteral, e.message
           end
+        elsif useNativeTypes && extendedRepresentation
+          res['@value'] = resource  # Raw literal
         elsif resource.datatype.start_with?("https://www.w3.org/ns/i18n#") && rdfDirection == 'i18n-datatype' && @context.processingMode('json-ld-1.1')
           lang, dir = resource.datatype.fragment.split('_')
           res['@value'] = resource.to_s

spec/from_rdf_spec.rb

@@ -268,6 +268,73 @@
           it(title) {do_fromRdf(processingMode: "json-ld-1.1", **params)}
         end
       end
+
+      context "extendedRepresentation: true" do
+        {
+          "true": {
+            output: [{
+               "@id" => "http://example.org/vocab#id",
+               "http://example.org/vocab#bool" => [{"@value" => RDF::Literal(true)}]
+             }],
+            input:%(
+              @prefix ex: <http://example.org/vocab#> .
+              @prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
+              ex:id ex:bool true .
+            )
+          },
+          "false": {
+            output: [{
+               "@id" => "http://example.org/vocab#id",
+               "http://example.org/vocab#bool" => [{"@value" => RDF::Literal(false)}]
+             }],
+            input: %(
+              @prefix ex: <http://example.org/vocab#> .
+              @prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
+              ex:id ex:bool false .
+            )
+          },
+          "double": {
+            output: [{
+               "@id" => "http://example.org/vocab#id",
+               "http://example.org/vocab#double" => [{"@value" => RDF::Literal(1.23E0)}]
+             }],
+            input: %(
+              @prefix ex: <http://example.org/vocab#> .
+              @prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
+              ex:id ex:double 1.23E0 .
+            )
+          },
+          "double-zero": {
+            output: [{
+               "@id" => "http://example.org/vocab#id",
+               "http://example.org/vocab#double" => [{"@value" => RDF::Literal(0, datatype: RDF::XSD.double)}]
+             }],
+            input: %(
+              @prefix ex: <http://example.org/vocab#> .
+              @prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
+              ex:id ex:double 0.0E0 .
+            )
+          },
+          "integer": {
+            output: [{
+               "@id" => "http://example.org/vocab#id",
+               "http://example.org/vocab#integer" => [{"@value" => RDF::Literal(123)}]
+             }],
+            input: %(
+              @prefix ex: <http://example.org/vocab#> .
+              @prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .
+              ex:id ex:integer 123 .
+            )
+          },
+        }.each do |title, params|
+          params[:input] = RDF::Graph.new << RDF::Turtle::Reader.new(params[:input])
+          it(title) {
+            do_fromRdf(processingMode: "json-ld-1.1",
+            useNativeTypes: true,
+            extendedRepresentation: true,
+            **params)}
+        end
+      end
     end
 
     context "anons" do