json-ld 3.1.2 → 3.1.7

data/lib/json/ld/format.rb

@@ -58,9 +58,10 @@ module JSON::LD
             out = options[:output] || $stdout
             out.set_encoding(Encoding::UTF_8) if RUBY_PLATFORM == "java"
             options = options.merge(expandContext: options.delete(:context)) if options.has_key?(:context)
+            options[:base] ||= options[:base_uri]
             if options[:format] == :jsonld
               if files.empty?
-                # If files are empty, either use options[:execute]
+                # If files are empty, either use options[:evaluate] or STDIN
                 input = options[:evaluate] ? StringIO.new(options[:evaluate]) : STDIN
                 input.set_encoding(options.fetch(:encoding, Encoding::UTF_8))
                 JSON::LD::API.expand(input, validate: false, **options) do |expanded|
@@ -93,9 +94,10 @@ module JSON::LD
             raise ArgumentError, "Compacting requires a context" unless options[:context]
             out = options[:output] || $stdout
             out.set_encoding(Encoding::UTF_8) if RUBY_PLATFORM == "java"
+            options[:base] ||= options[:base_uri]
             if options[:format] == :jsonld
               if files.empty?
-                # If files are empty, either use options[:execute]
+                # If files are empty, either use options[:evaluate] or STDIN
                 input = options[:evaluate] ? StringIO.new(options[:evaluate]) : STDIN
                 input.set_encoding(options.fetch(:encoding, Encoding::UTF_8))
                 JSON::LD::API.compact(input, options[:context], **options) do |compacted|
@@ -126,7 +128,7 @@ module JSON::LD
               control: :url2,
               use: :required,
               on: ["--context CONTEXT"],
-              description: "Context to use when compacting.") {|arg| RDF::URI(arg)},
+              description: "Context to use when compacting.") {|arg| RDF::URI(arg).absolute? ? RDF::URI(arg) : StringIO.new(File.read(arg))},
           ]
         },
         flatten: {
@@ -137,9 +139,10 @@ module JSON::LD
           lambda: ->(files, **options) do
             out = options[:output] || $stdout
             out.set_encoding(Encoding::UTF_8) if RUBY_PLATFORM == "java"
+            options[:base] ||= options[:base_uri]
             if options[:format] == :jsonld
               if files.empty?
-                # If files are empty, either use options[:execute]
+                # If files are empty, either use options[:evaluate] or STDIN
                 input = options[:evaluate] ? StringIO.new(options[:evaluate]) : STDIN
                 input.set_encoding(options.fetch(:encoding, Encoding::UTF_8))
                 JSON::LD::API.flatten(input, options[:context], **options) do |flattened|
@@ -162,7 +165,16 @@ module JSON::LD
                 end
               end
             end
-          end
+          end,
+          options: [
+            RDF::CLI::Option.new(
+              symbol: :context,
+              datatype: RDF::URI,
+              control: :url2,
+              use: :required,
+              on: ["--context CONTEXT"],
+              description: "Context to use when compacting.") {|arg| RDF::URI(arg)},
+          ]
         },
         frame: {
           description: "Frame JSON-LD or parsed RDF",
@@ -173,9 +185,10 @@ module JSON::LD
             raise ArgumentError, "Framing requires a frame" unless options[:frame]
             out = options[:output] || $stdout
             out.set_encoding(Encoding::UTF_8) if RUBY_PLATFORM == "java"
+            options[:base] ||= options[:base_uri]
             if options[:format] == :jsonld
               if files.empty?
-                # If files are empty, either use options[:execute]
+                # If files are empty, either use options[:evaluate] or STDIN
                 input = options[:evaluate] ? StringIO.new(options[:evaluate]) : STDIN
                 input.set_encoding(options.fetch(:encoding, Encoding::UTF_8))
                 JSON::LD::API.frame(input, options[:frame], **options) do |framed|
@@ -207,7 +220,7 @@ module JSON::LD
               control: :url2,
               use: :required,
               on: ["--frame FRAME"],
-              description: "Frame to use when serializing.") {|arg| RDF::URI(arg)}
+              description: "Frame to use when serializing.") {|arg| RDF::URI(arg).absolute? ? RDF::URI(arg) : StringIO.new(File.read(arg))}
           ]
         },
       }

data/lib/json/ld/from_rdf.rb

@@ -14,16 +14,14 @@ module JSON::LD
     # @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 [Boolean] ordered (true)
-    #   Ensure output objects have keys ordered properly
+    #
     # @return [Array<Hash>] the JSON-LD document in normalized form
-    def from_statements(dataset, useRdfType: false, useNativeTypes: false, ordered: false)
+    def from_statements(dataset, useRdfType: false, useNativeTypes: false)
       default_graph = {}
       graph_map = {'@default' => default_graph}
       referenced_once = {}
 
       value = nil
-      ec = @context
 
       # Create an entry for compound-literal node detection
       compound_literal_subjects = {}
@@ -34,7 +32,7 @@ module JSON::LD
       dataset.each do |statement|
         #log_debug("statement") { statement.to_nquads.chomp}
 
-        name = statement.graph_name ? ec.expand_iri(statement.graph_name).to_s : '@default'
+        name = statement.graph_name ? @context.expand_iri(statement.graph_name, base: @options[:base]).to_s : '@default'
 
         # Create a graph entry as needed
         node_map = graph_map[name] ||= {}
@@ -42,30 +40,29 @@ module JSON::LD
 
         default_graph[name] ||= {'@id' => name} unless name == '@default'
 
-        subject = ec.expand_iri(statement.subject, as_string: true)
-        node = node_map[subject] ||= {'@id' => subject}
+        subject = statement.subject.to_s
+        node = node_map[subject] ||= resource_representation(statement.subject,useNativeTypes)
 
         # If predicate is rdf:datatype, note subject in compound literal subjects map
         if @options[:rdfDirection] == 'compound-literal' && statement.predicate == RDF.to_uri + 'direction'
           compound_literal_subjects[name][subject] ||= true
         end
 
-        # If object is an IRI or blank node identifier, 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.
-        node_map[statement.object.to_s] ||= {'@id' => statement.object.to_s} unless
-          statement.object.literal?
+        # 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?
+          node_map[statement.object.to_s] ||=
+            resource_representation(statement.object, useNativeTypes)
+        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.
+        # XXX JSON-LD* does not support embedded value of @type
         if statement.predicate == RDF.type && statement.object.resource? && !useRdfType
           merge_value(node, '@type', statement.object.to_s)
           next
         end
 
         # Set value to the result of using the RDF to Object Conversion algorithm, passing object, rdfDirection, and use native types.
-        value = ec.expand_value(nil,
-                                statement.object,
-                                rdfDirection: @options[:rdfDirection],
-                                useNativeTypes: useNativeTypes,
-                                log_depth: @options[:log_depth])
+        value = resource_representation(statement.object, useNativeTypes)
 
         merge_value(node, statement.predicate.to_s, value)
 
@@ -147,11 +144,11 @@ module JSON::LD
       end
 
       result = []
-      default_graph.keys.opt_sort(ordered: ordered).each do |subject|
+      default_graph.keys.opt_sort(ordered: @options[:ordered]).each do |subject|
         node = default_graph[subject]
         if graph_map.has_key?(subject)
           node['@graph'] = []
-          graph_map[subject].keys.opt_sort(ordered: ordered).each do |s|
+          graph_map[subject].keys.opt_sort(ordered: @options[:ordered]).each do |s|
             n = graph_map[subject][s]
             n.delete(:usages)
             node['@graph'] << n unless node_reference?(n)
@@ -163,5 +160,31 @@ module JSON::LD
       #log_debug("fromRdf") {result.to_json(JSON_STATE) rescue 'malformed json'}
       result
     end
+
+    private
+    def resource_representation(resource, useNativeTypes)
+      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)}
+        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)))
+        end
+        rep
+      when RDF::Literal
+        @context.expand_value(nil,
+                              resource,
+                              rdfDirection: @options[:rdfDirection],
+                              useNativeTypes: useNativeTypes,
+                              base: @options[:base])
+      else
+        {'@id' => resource.to_s}
+      end
+    end
   end
 end

data/lib/json/ld/reader.rb

@@ -1,5 +1,5 @@
 # -*- encoding: utf-8 -*-
-# frozen_string_literal: true
+
 module JSON::LD
   ##
   # A JSON-LD parser in Ruby.
@@ -7,6 +7,7 @@ module JSON::LD
   # @see https://www.w3.org/TR/json-ld11-api
   # @author [Gregg Kellogg](http://greggkellogg.net/)
   class Reader < RDF::Reader
+    include StreamingReader
     format Format
 
     ##
@@ -19,7 +20,7 @@ module JSON::LD
           control: :url2,
           datatype: RDF::URI,
           on: ["--expand-context CONTEXT"],
-          description: "Context to use when expanding.") {|arg| RDF::URI(arg)},
+          description: "Context to use when expanding.") {|arg| RDF::URI(arg).absolute? ? RDF::URI(arg) : StringIO.new(File.read(arg))},
         RDF::CLI::Option.new(
           symbol: :extractAllScripts,
           datatype: TrueClass,
@@ -46,6 +47,12 @@ module JSON::LD
           control: :select,
           on: ["--rdf-direction DIR", %w(i18n-datatype compound-literal)],
           description: "How to serialize literal direction (i18n-datatype compound-literal)") {|arg| RDF::URI(arg)},
+        RDF::CLI::Option.new(
+          symbol: :stream,
+          datatype: TrueClass,
+          control: :checkbox,
+          on: ["--[no-]stream"],
+          description: "Optimize for streaming JSON-LD to RDF.") {|arg| arg},
       ]
     end
 
@@ -63,13 +70,11 @@ module JSON::LD
       options[:base_uri] ||= options[:base]
       super do
         @options[:base] ||= base_uri.to_s if base_uri
-        begin
-          # Trim non-JSON stuff in script.
-          @doc = if input.respond_to?(:read)
-            input
-          else
-            StringIO.new(input.to_s.sub(%r(\A[^{\[]*)m, '').sub(%r([^}\]]*\Z)m, ''))
-          end
+        # Trim non-JSON stuff in script.
+        @doc = if input.respond_to?(:read)
+          input
+        else
+          StringIO.new(input.to_s.sub(%r(\A[^{\[]*)m, '').sub(%r([^}\]]*\Z)m, ''))
         end
 
         if block_given?
@@ -85,7 +90,11 @@ module JSON::LD
     # @private
     # @see   RDF::Reader#each_statement
     def each_statement(&block)
-      JSON::LD::API.toRdf(@doc, **@options, &block)
+      if @options[:stream]
+        stream_statement(&block)
+      else
+        API.toRdf(@doc, **@options, &block)
+      end
     rescue ::JSON::ParserError, ::JSON::LD::JsonLdError => e
       log_fatal("Failed to parse input document: #{e.message}", exception: RDF::ReaderError)
     end
@@ -95,7 +104,7 @@ module JSON::LD
     # @see   RDF::Reader#each_triple
     def each_triple(&block)
       if block_given?
-        JSON::LD::API.toRdf(@doc, **@options) do |statement|
+        each_statement do |statement|
           yield(*statement.to_triple)
         end
       end

data/lib/json/ld/streaming_reader.rb

@@ -0,0 +1,578 @@
+# -*- encoding: utf-8 -*-
+require 'json/ld'
+require 'json/ld/expand'
+require 'json/ld/to_rdf'
+
+module JSON::LD
+  ##
+  # A streaming JSON-LD parser in Ruby.
+  #
+  # @see http://json-ld.org/spec/ED/20110507/
+  # @author [Gregg Kellogg](http://greggkellogg.net/)
+  module StreamingReader
+    include Utils
+    include JSON::LD::ToRDF # For value object conversion
+
+    # The base URI to use when resolving relative URIs
+    # @return [RDF::URI]
+    attr_reader :base
+    attr_reader :namer
+
+    def self.format; JSON::LD::Format; end
+
+    ##
+    # @see   RDF::Reader#each_statement
+    def stream_statement(&block)
+      unique_bnodes, rename_bnodes = @options[:unique_bnodes], @options.fetch(:rename_bnodes, true)
+      # FIXME: document loader doesn't stream
+      @base = RDF::URI(@options[:base] || base_uri)
+      value = MultiJson.load(@doc, **@options)
+      context_ref = @options[:expandContext]
+      #context_ref = @options.fetch(:expandContext, remote_doc.contextUrl)
+      context = Context.parse(context_ref, **@options)
+
+      @namer = unique_bnodes ? BlankNodeUniqer.new : (rename_bnodes ? BlankNodeNamer.new("b") : BlankNodeMapper.new)
+      # Namer for naming provisional nodes, which may be determined later to be actual
+      @provisional_namer = BlankNodeNamer.new("p")
+
+      parse_object(value, nil, context, graph_is_named: false) do |st|
+        # Only output reasonably valid triples
+        if st.to_a.all? {|r| r.is_a?(RDF::Term) && (r.uri? ? r.valid? : true)}
+          block.call(st)
+        end
+      end
+    rescue ::JSON::ParserError, ::JSON::LD::JsonLdError => e
+      log_fatal("Failed to parse input document: #{e.message}", exception: RDF::ReaderError)
+    end
+
+  private
+
+    # Parse a node object, or array of node objects
+    #
+    # @param [Array, Hash] input
+    # @param [String] active_property
+    #   The unexpanded property referencing this object
+    # @param [Context] context
+    # @param [RDF::Resource] subject referencing this object
+    # @param [RDF::URI] predicate the predicate part of the reference
+    # @param [Boolean] from_map
+    #   Expanding from a map, which could be an `@type` map, so don't clear out context term definitions
+    # @param [Boolean] graph_is_named
+    #   Use of `@graph` implies a named graph; not true at the top-level.
+    # @param [RDF::URI] extra_type from a type map
+    # @param [String] language from a language map
+    # @param [RDF::Resource] node_id from an id map
+    # @return [void]
+    def parse_object(input, active_property, context,
+                     subject: nil, predicate: nil, from_map: false,
+                     extra_type: nil, language: nil, node_id: nil,
+                     graph_is_named: true, &block)
+
+      # Skip predicates that look like a BNode
+      if predicate.to_s.start_with?('_:')
+        warn "[DEPRECATION] Blank Node properties deprecated in JSON-LD 1.1."
+        return
+      end
+
+      if input.is_a?(Array)
+        input.each {|e| parse_object(e, active_property, context, subject: subject, predicate: predicate, from_map: from_map, &block)}
+        return
+      end
+
+      # Note that we haven't parsed an @id key, so have no subject
+      have_id, node_reference, is_list_or_set = false, false, false
+      node_id ||= RDF::Node.new(@provisional_namer.get_sym)
+      # For keeping statements not yet ready to be emitted
+      provisional_statements = []
+      value_object = {}
+
+      # Use a term-specific context, if defined, based on the non-type-scoped context.
+      property_scoped_context = context.term_definitions[active_property].context if active_property && context.term_definitions[active_property]
+
+      # Revert any previously type-scoped term definitions, unless this is from a map, a value object or a subject reference
+      # FIXME
+      if input.is_a?(Hash) && context.previous_context
+        expanded_key_map = input.keys.inject({}) do |memo, key|
+          memo.merge(key => context.expand_iri(key, vocab: true, as_string: true, base: base))
+        end
+        revert_context = !from_map &&
+          !expanded_key_map.values.include?('@value') &&
+          !(expanded_key_map.values == ['@id'])
+        context = context.previous_context if revert_context
+      end
+
+      # Apply property-scoped context after reverting term-scoped context
+      context = context.parse(property_scoped_context, base: base, override_protected: true) unless
+        property_scoped_context.nil?
+
+      # Otherwise, unless the value is a number, expand the value according to the Value Expansion rules, passing active property.
+      unless input.is_a?(Hash)
+        input = context.expand_value(active_property, input, base: base)
+      end
+
+      # Output any type provided from a type map
+      provisional_statements << RDF::Statement(node_id, RDF.type, extra_type) if
+        extra_type
+
+      # Add statement, either provisionally, or just emit
+      add_statement = Proc.new do |st|
+        if have_id || st.to_quad.none? {|r| r == node_id}
+          block.call(st)
+        else
+          provisional_statements << st
+        end
+      end
+
+      # Input is an object (Hash), parse keys in order
+      state = :await_context
+      input.each do |key, value|
+        expanded_key = context.expand_iri(key, base: base, vocab: true)
+        case expanded_key
+        when '@context'
+          raise JsonLdError::InvalidStreamingKeyOrder,
+                "found #{key} in state #{state}" unless state == :await_context
+          context = context.parse(value, base: base)
+          state = :await_type
+        when '@type'
+          # Set the type-scoped context to the context on input, for use later
+          raise JsonLdError::InvalidStreamingKeyOrder,
+                "found #{key} in state #{state}" unless [:await_context, :await_type].include?(state)
+
+          type_scoped_context = context
+          as_array(value).sort.each do |term|
+            raise JsonLdError::InvalidTypeValue,
+                  "value of @type must be a string: #{term.inspect}" if !term.is_a?(String)
+            term_context = type_scoped_context.term_definitions[term].context if type_scoped_context.term_definitions[term]
+            context = context.parse(term_context, base: base, propagate: false) unless term_context.nil?
+            type = type_scoped_context.expand_iri(term,
+              base: base,
+              documentRelative: true,
+              vocab: true)
+
+            # Early terminate for @json
+            type = RDF.JSON if type == '@json'
+            # Add a provisional statement
+            provisional_statements << RDF::Statement(node_id, RDF.type, type)
+          end
+          state = :await_type
+        when '@id'
+          raise JsonLdError::InvalidSetOrListObject,
+                "found #{key} in state #{state}" if is_list_or_set
+          raise JsonLdError::CollidingKeywords,
+                "found #{key} in state #{state}" unless [:await_context, :await_type, :await_id].include?(state)
+
+          # Set our actual id, and use for replacing any provisional statements using our existing node_id, which is provisional
+          raise JsonLdError::InvalidIdValue,
+                "value of @id must be a string: #{value.inspect}" if !value.is_a?(String)
+            node_reference = input.keys.length == 1
+          expanded_id = context.expand_iri(value, base: base, documentRelative: true)
+          next if expanded_id.nil?
+          new_node_id = as_resource(expanded_id)
+          # Replace and emit any statements including our provisional id with the newly established node (or graph) id
+          provisional_statements.each do |st|
+            st.subject = new_node_id if st.subject == node_id
+            st.object = new_node_id if st.object == node_id
+            st.graph_name = new_node_id if st.graph_name == node_id
+            block.call(st)
+          end
+
+          provisional_statements.clear
+          have_id, node_id = true, new_node_id
+
+          # if there's a subject & predicate, emit that statement now
+          if subject && predicate
+            st = RDF::Statement(subject, predicate, node_id)
+            block.call(st)
+          end
+          state = :properties
+
+        when '@direction'
+          raise JsonLdError::InvalidStreamingKeyOrder,
+                "found @direction in state #{state}" if state == :properties
+          value_object['@direction'] = value
+          state = :await_id
+        when '@graph'
+          # If `@graph` is at the top level (no `subject`) and value contains no keys other than `@graph` and `@context`, add triples to the default graph
+          # Process all graph statements
+          parse_object(value, nil, context) do |st|
+            # If `@graph` is at the top level (`graph_is_named` is `false`) and input contains no keys other than `@graph` and `@context`, add triples to the default graph
+            relevant_keys = input.keys - ['@context', key]
+            st.graph_name = node_id unless !graph_is_named && relevant_keys.empty?
+            if st.graph_name && !st.graph_name.valid?
+              warn "skipping graph statement within invalid graph name: #{st.inspect}"
+            else
+              add_statement.call(st)
+            end
+          end
+          state = :await_id unless state == :properties
+        when '@included'
+          # Expanded values must be node objects
+          have_statements = false
+          parse_object(value, active_property, context) do |st|
+            have_statements ||= st.has_subject?
+            block.call(st)
+          end
+          raise JsonLdError::InvalidIncludedValue, "values of @included must expand to node objects" unless have_statements
+          state = :await_id unless state == :properties
+        when '@index'
+          state = :await_id unless state == :properties
+          raise JsonLdError::InvalidIndexValue,
+                "Value of @index is not a string: #{value.inspect}" unless value.is_a?(String)
+        when '@language'
+          raise JsonLdError::InvalidStreamingKeyOrder,
+                "found @language in state #{state}" if state == :properties
+          raise JsonLdError::InvalidLanguageTaggedString,
+                "@language value must be a string: #{value.inspect}" if !value.is_a?(String)
+          if value !~ /^[a-zA-Z]{1,8}(-[a-zA-Z0-9]{1,8})*$/
+            warn "@language must be valid BCP47: #{value.inspect}"
+            return
+          end
+          language = value
+          state = :await_id
+        when '@list'
+          raise JsonLdError::InvalidSetOrListObject,
+                "found #{key} in state #{state}" if
+            ![:await_context, :await_type, :await_id].include?(state)
+          is_list_or_set = true
+          if subject
+            node_id = parse_list(value, active_property, context, &block)
+          end
+          state = :properties
+        when '@nest'
+          nest_context = context.term_definitions[active_property].context if context.term_definitions[active_property]
+          nest_context = if nest_context.nil?
+            context
+          else
+            context.parse(nest_context, base: base, override_protected: true)
+          end
+          as_array(value).each do |v|
+            raise JsonLdError::InvalidNestValue, v.inspect unless
+              v.is_a?(Hash) && v.keys.none? {|k| nest_context.expand_iri(k, vocab: true, base: base) == '@value'}
+              parse_object(v, active_property, nest_context, node_id: node_id) do |st|
+                add_statement.call(st)
+              end
+          end
+          state = :await_id unless state == :properties
+        when '@reverse'
+          as_array(value).each do |item|
+            item = context.expand_value(active_property, item, base: base) unless item.is_a?(Hash)
+            raise JsonLdError::InvalidReverseValue, item.inspect if value?(item)
+            raise JsonLdError::InvalidReversePropertyMap, item.inspect if node_reference?(item)
+            raise JsonLdError::InvalidReversePropertyValue, item.inspect if list?(item)
+            has_own_subject = false
+            parse_object(item, active_property, context, node_id: node_id, predicate: predicate) do |st|
+              if st.subject == node_id
+                raise JsonLdError::InvalidReversePropertyValue, item.inspect if !st.object.resource?
+                # Invert sense of statements
+                st = RDF::Statement(st.object, st.predicate, st.subject)
+                has_own_subject = true
+              end
+              add_statement.call(st)
+            end
+
+            # If the reversed node does not make any claims on this subject, it's an error
+            raise JsonLdError::InvalidReversePropertyValue, item.inspect unless has_own_subject
+          end
+          state = :await_id unless state == :properties
+        when '@set'
+          raise JsonLdError::InvalidSetOrListObject,
+                "found #{key} in state #{state}" if
+                ![:await_context, :await_type, :await_id].include?(state)
+          is_list_or_set = true
+          value = as_array(value).compact
+          parse_object(value, active_property, context, subject: subject, predicate: predicate, &block)
+          node_id = nil
+          state = :properties
+        when '@value'
+          raise JsonLdError::InvalidStreamingKeyOrder,
+                "found @value in state #{state}" if state == :properties
+          value_object['@value'] = value
+          state = :await_id
+        else
+          state = :await_id unless state == :properties
+          # Skip keys that don't expand to a keyword or absolute IRI
+          next if expanded_key.is_a?(RDF::URI) && !expanded_key.absolute?
+          parse_property(value, key, context, node_id, expanded_key) do |st|
+            add_statement.call(st)
+          end
+        end
+      end
+
+      # Value object with @id
+      raise JsonLdError::InvalidValueObject,
+            "value object has unknown key: @id" if
+            !value_object.empty? && (have_id || is_list_or_set)
+
+      # Can't have both @id and either @list or @set
+      raise JsonLdError::InvalidSetOrListObject,
+            "found @id with @list or @set" if
+            have_id && is_list_or_set
+
+      type_statements = provisional_statements.select {|ps| ps.predicate == RDF.type && ps.graph_name.nil?}
+      value_object['@language'] = (@options[:lowercaseLanguage] ? language.downcase : language) if language
+      if !value_object.empty? &&
+         (!value_object['@value'].nil? ||
+          (type_statements.first || RDF::Statement.new).object == RDF.JSON)
+
+        # There can be only one value of @type
+        case type_statements.length
+        when 0 then #skip
+        when 1
+          raise JsonLdError::InvalidTypedValue,
+                "value of @type must be an IRI or '@json': #{type_statements.first.object.inspect}" unless
+                type_statements.first.object.valid?
+          value_object['@type'] = type_statements.first.object
+        else
+          raise JsonLdError::InvalidValueObject,
+                "value object must not have more than one type"
+        end
+
+        # Check for extra keys
+        raise JsonLdError::InvalidValueObject,
+              "value object has unknown keys: #{value_object.inspect}" unless
+              (value_object.keys - Expand::KEYS_VALUE_LANGUAGE_TYPE_INDEX_DIRECTION).empty?
+
+        # @type is inconsistent with either @language or @direction
+        raise JsonLdError::InvalidValueObject,
+              "value object must not include @type with either " +
+              "@language or @direction: #{value_object.inspect}" if
+              value_object.keys.include?('@type') && !(value_object.keys & %w(@language @direction)).empty?
+
+        if value_object.key?('@language') && !value_object['@value'].is_a?(String)
+          raise JsonLdError::InvalidLanguageTaggedValue,
+                "with @language @value must be a string: #{value_object.inspect}"
+        elsif value_object['@type'] && value_object['@type'] != RDF.JSON
+          raise JsonLdError::InvalidTypedValue,
+                "value of @type must be an IRI or '@json': #{value_object['@type'].inspect}" unless
+                value_object['@type'].is_a?(RDF::URI)
+        elsif value_object['@type'] != RDF.JSON
+          case value_object['@value']
+          when String, TrueClass, FalseClass, Numeric then # okay
+          else
+            raise JsonLdError::InvalidValueObjectValue,
+                  "@value is: #{value_object['@value'].inspect}"
+          end
+        end
+        literal = item_to_rdf(value_object, &block)
+        st = RDF::Statement(subject, predicate, literal)
+        block.call(st)
+      elsif !provisional_statements.empty?
+        # Emit all provisional statements, as no @id was ever found
+        provisional_statements.each {|st| block.call(st)}
+      end
+
+      # Use implicit subject to generate the relationship
+      if value_object.empty? && subject && predicate && !have_id && !node_reference
+        block.call(RDF::Statement(subject, predicate, node_id))
+      end
+    end
+
+    def parse_property(input, active_property, context, subject, predicate, &block)
+      container = context.container(active_property)
+      if container.include?('@language') && input.is_a?(Hash)
+        input.each do |lang, lang_value|
+          expanded_lang = context.expand_iri(lang, vocab: true)
+          if lang !~ /^[a-zA-Z]{1,8}(-[a-zA-Z0-9]{1,8})*$/ && expanded_lang != '@none'
+            warn "@language must be valid BCP47: #{lang.inspect}"
+          end
+
+          as_array(lang_value).each do |item|
+            raise JsonLdError::InvalidLanguageMapValue,
+                  "Expected #{item.inspect} to be a string" unless item.nil? || item.is_a?(String)
+            lang_obj = {'@value' => item}
+            lang_obj['@language'] = lang unless expanded_lang == '@none'
+            lang_obj['@direction'] = context.direction(lang) if context.direction(lang)
+            parse_object(lang_obj, active_property, context, subject: subject, predicate: predicate, &block)
+          end
+        end
+      elsif container.include?('@list')
+        # Handle case where value is a list object
+        if input.is_a?(Hash) &&
+           input.keys.map do |k|
+              context.expand_iri(k, vocab: true, as_string: true, base: base)
+           end.include?('@list')
+          parse_object(input, active_property, context,
+                       subject: subject, predicate: predicate,  &block)
+        else
+          list = parse_list(input, active_property, context, &block)
+          block.call(RDF::Statement(subject, predicate, list))
+        end
+      elsif container.intersect?(JSON::LD::Expand::CONTAINER_INDEX_ID_TYPE) && input.is_a?(Hash)
+        # Get appropriate context for this container
+        container_context = if container.include?('@type') && context.previous_context
+          context.previous_context
+        elsif container.include?('@id') && context.term_definitions[active_property]
+          id_context = context.term_definitions[active_property].context if context.term_definitions[active_property]
+          if id_context.nil?
+            context
+          else
+            context.parse(id_context, base: base, propagate: false)
+          end
+        else
+          context
+        end
+
+        input.each do |k, v|
+          # If container mapping in the active context includes @type, and k is a term in the active context having a local context, use that context when expanding values
+          map_context = container_context.term_definitions[k].context if
+            container.include?('@type') && container_context.term_definitions[k]
+          unless map_context.nil?
+            map_context = container_context.parse(map_context, base: base, propagate: false)
+          end
+          map_context ||= container_context
+
+          expanded_k = container_context.expand_iri(k, vocab: true, as_string: true, base: base)
+          index_key = context.term_definitions[active_property].index || '@index'
+
+          case
+          when container.include?('@index') && container.include?('@graph')
+            # Index is ignored
+            as_array(v).each do |item|
+              # Each value is in a separate graph
+              graph_name = RDF::Node.new(namer.get_sym)
+              parse_object(item, active_property, context) do |st|
+                st.graph_name ||= graph_name
+                block.call(st)
+              end
+              block.call(RDF::Statement(subject, predicate, graph_name))
+
+              # Add a property index, if appropriate
+              unless index_key == '@index'
+                # Expand key based on term
+                expanded_k = k == '@none' ?
+                  '@none' :
+                  container_context.expand_value(index_key, k, base: base)
+
+                # Add the index property as a property of the graph name
+                index_property = container_context.expand_iri(index_key, vocab: true, base: base)
+                emit_object(expanded_k, index_key, map_context, graph_name,
+                            index_property, from_map: true, &block) unless
+                            expanded_k == '@none'
+              end
+            end
+          when container.include?('@index')
+            if index_key == '@index'
+              # Index is ignored
+              emit_object(v, active_property, map_context, subject, predicate, from_map: true, &block)
+            else
+              # Expand key based on term
+              expanded_k = k == '@none' ?
+                '@none' :
+                container_context.expand_value(index_key, k, base: base)
+
+              index_property = container_context.expand_iri(index_key, vocab: true, as_string: true, base: base)
+
+              # index_key is a property
+              as_array(v).each do |item|
+                item = container_context.expand_value(active_property, item, base: base) if item.is_a?(String)
+                raise JsonLdError::InvalidValueObject,
+                  "Attempt to add illegal key to value object: #{index_key}" if value?(item)
+                # add expanded_k as value of index_property in item
+                item[index_property] = [expanded_k].concat(Array(item[index_property])) unless expanded_k == '@none'
+                emit_object(item, active_property, map_context, subject, predicate, from_map: true, &block)
+              end
+            end
+          when container.include?('@id') && container.include?('@graph')
+            graph_name = expanded_k == '@none' ?
+               RDF::Node.new(namer.get_sym) : 
+              container_context.expand_iri(k, documentRelative: true, base: base)
+            parse_object(v, active_property, context) do |st|
+              st.graph_name ||= graph_name
+              block.call(st)
+            end
+            block.call(RDF::Statement(subject, predicate, graph_name))
+          when container.include?('@id')
+            expanded_k = container_context.expand_iri(k, documentRelative: true, base: base)
+            # pass our id
+            emit_object(v, active_property, map_context, subject, predicate,
+                        node_id: (expanded_k unless expanded_k == '@none'),
+                        from_map: true,
+                        &block)
+          when container.include?('@type')
+            emit_object(v, active_property, map_context, subject, predicate,
+                        from_map: true,
+                        extra_type: as_resource(expanded_k),
+                        &block)
+          end
+        end
+      elsif container.include?('@graph')
+        # Index is ignored
+        as_array(input).each do |v|
+          # Each value is in a separate graph
+          graph_name = RDF::Node.new(namer.get_sym)
+          parse_object(v, active_property, context) do |st|
+            st.graph_name ||= graph_name
+            block.call(st)
+          end
+          block.call(RDF::Statement(subject, predicate, graph_name))
+        end
+      else
+        emit_object(input, active_property, context, subject, predicate, &block)
+      end
+    end
+
+    # Wrapps parse_object to handle JSON literals and reversed properties
+    def emit_object(input, active_property, context, subject, predicate, **options, &block)
+      if context.coerce(active_property) == '@json'
+        parse_object(context.expand_value(active_property, input), active_property, context,
+                     subject: subject, predicate: predicate, **options, &block)
+      elsif context.reverse?(active_property)
+        as_array(input).each do |item|
+          item = context.expand_value(active_property, item, base: base) unless item.is_a?(Hash)
+          raise JsonLdError::InvalidReverseValue, item.inspect if value?(item)
+          raise JsonLdError::InvalidReversePropertyValue, item.inspect if list?(item)
+          has_own_subject = false
+          parse_object(item, active_property, context, subject: subject, predicate: predicate, **options) do |st|
+            if st.subject == subject
+              raise JsonLdError::InvalidReversePropertyValue, item.inspect if !st.object.resource?
+              # Invert sense of statements
+              st = RDF::Statement(st.object, st.predicate, st.subject)
+              has_own_subject = true
+            end
+            block.call(st)
+          end
+
+          # If the reversed node does not make any claims on this subject, it's an error
+          raise JsonLdError::InvalidReversePropertyValue,
+                "@reverse value must be a node: #{value.inspect}" unless has_own_subject
+        end
+      else
+        as_array(input).flatten.each do |item|
+          # emit property/value
+          parse_object(item, active_property, context,
+                       subject: subject, predicate: predicate, **options, &block)
+        end
+      end
+    end
+
+    # Process input as an ordered list
+    # @return [RDF::Resource] the list head
+    def parse_list(input, active_property, context, &block)
+      # Transform all entries into their values
+      # this allows us to eliminate those that don't create any statements
+      fake_subject = RDF::Node.new
+      values = as_array(input).map do |entry|
+        if entry.is_a?(Array)
+          # recursive list
+          entry_value = parse_list(entry, active_property, context, &block)
+        else
+          entry_value = nil
+          parse_object(entry, active_property, context, subject: fake_subject, predicate: RDF.first) do |st|
+            if st.subject == fake_subject
+              entry_value = st.object
+            else
+              block.call(st)
+            end
+          end
+          entry_value
+        end
+      end.compact
+      return RDF.nil if values.empty?
+
+      # Construct a list from values, and emit list statements, returning the list subject
+      list = RDF::List(*values)
+      list.each_statement(&block)
+      return list.subject
+    end
+  end
+end