Read literal direction from its:dir; requires rdf:version >= 1.2.

gkellogg · gkellogg · commit 38d174bb98d2 · 2025-01-26T16:39:13.000-08:00
diff --git a/Gemfile b/Gemfile
@@ -17,6 +17,7 @@ group :development do
   gem 'sxp',                github: "dryruby/sxp.rb",               branch: "develop"
 
   gem "equivalent-xml"
+  gem 'ostruct', '~> 0.6.1'
 end
 
 group :debug do
diff --git a/Gemfile-pure b/Gemfile-pure
@@ -17,6 +17,7 @@ group :development do
   gem 'sxp',                github: "dryruby/sxp.rb",               branch: "develop"
 
   gem "equivalent-xml"
+  gem 'ostruct', '~> 0.6.1'
 end
 
 group :debug do
diff --git a/lib/rdf/rdfxml.rb b/lib/rdf/rdfxml.rb
@@ -4,6 +4,9 @@
 module RDF
   XML = Class.new(Vocabulary("http://www.w3.org/XML/1998/namespace"))
 
+  # Internationalization Tag Set
+  ITS = Class.new(Vocabulary("http://www.w3.org/2005/11/its"))
+
   ##
   # **`RDF::RDFXML`** is an RDF/XML extension for RDF.rb.
   #
diff --git a/lib/rdf/rdfxml/reader.rb b/lib/rdf/rdfxml/reader.rb
@@ -30,12 +30,16 @@ class EvaluationContext
       attr_accessor :language
       attr_accessor :graph
       attr_accessor :li_counter
+      attr_accessor :direction
+      attr_accessor :version  # RDF Version, mirrored here
 
-      def initialize(base, element, graph, &cb)
+      def initialize(base, element, graph, version: nil, &cb)
         # Initialize the evaluation context, [5.1]
         self.base = RDF::URI(base)
         @uri_mappings = {}
         @language = nil
+        @direction = nil
+        @version = version
         @graph = graph
         @li_counter = 0
 
@@ -47,6 +51,8 @@ def clone(element, **options, &cb)
         new_ec = EvaluationContext.new(@base, nil, @graph)
         new_ec.uri_mappings = self.uri_mappings.clone
         new_ec.language = self.language
+        new_ec.direction = self.direction
+        new_ec.version = self.version
 
         new_ec.extract_from_element(element, &cb) if element
         
@@ -68,6 +74,12 @@ def extract_from_ancestors(el, &cb)
       # Extract Evaluation Context from an element
       def extract_from_element(el, &cb)
         self.language = el.language if el.language
+        # Direction only used in RDF 1.2 or greater
+        self.direction = el.direction if el.direction && self.version.to_s >= "1.2"
+        # Direction requires the appropriate ITS version
+        if el.direction && el.its_version != '2.0'
+          # XXX raise error?
+        end
         if b = el.base
           b = RDF::URI(b)
           self.base = b.absolute? ? b : self.base.join(b)
@@ -100,7 +112,7 @@ def base=(b)
       end
 
       def inspect
-        v = %w(base subject language).map {|a| "#{a}='#{self.send(a).nil? ? 'nil' : self.send(a)}'"}
+        v = %w(base subject language direction).map {|a| "#{a}='#{self.send(a).nil? ? 'nil' : self.send(a)}'"}
         v << "uri_mappings[#{uri_mappings.keys.length}]"
         v.join(",")
       end
@@ -112,6 +124,11 @@ def inspect
     # @return [Module]
     attr_reader :implementation
 
+    # Version of RDF to use. Currently, only "1.2" is defined.
+    # @!attribute [r] version
+    # @return [String]
+    attr_reader :version
+
     ##
     # Initializes the RDF/XML reader instance.
     #
@@ -196,7 +213,11 @@ def each_statement(&block)
         if rdf_nodes.size == 0
           # If none found, root element may be processed as an RDF Node
 
-          ec = EvaluationContext.new(base_uri, root, @graph) do |prefix, value|
+          # Extract RDF version from root
+          @version = root.version
+          add_debug(root, "version: #{@version.inspect}")
+
+          ec = EvaluationContext.new(base_uri, root, @graph, version: @version) do |prefix, value|
             prefix(prefix, value)
           end
 
@@ -206,8 +227,13 @@ def each_statement(&block)
             log_fatal "node must be a proxy not a #{node.class}" unless node.is_a?(@implementation::NodeProxy)
             # XXX Skip this element if it's contained within another rdf:RDF element
 
-            # Extract base, lang and namespaces from parents to create proper evaluation context
-            ec = EvaluationContext.new(base_uri, nil, @graph)
+            # Extract RDF version from node
+            # XXX potentially, one node is processed with version "1.2" and others are parsed without a version.
+            @version = node.version
+            add_debug(root, "version: #{@version.inspect}")
+
+            # Extract base, lang, direction, version and namespaces from parents to create proper evaluation context
+            ec = EvaluationContext.new(base_uri, nil, @graph, version: @version)
             ec.extract_from_ancestors(node) do |prefix, value|
               prefix(prefix, value)
             end
@@ -322,7 +348,11 @@ def nodeElement(el, ec)
         elsif is_propertyAttr?(attr)
           # Attributes not RDF.type
           predicate = attr.uri
-          lit = RDF::Literal.new(attr.value, language: ec.language, validate: validate?, canonicalize: canonicalize?)
+          lit = RDF::Literal.new(attr.value,
+                                 language: ec.language,
+                                 direction: (ec.direction if ec.language),
+                                 validate: validate?,
+                                 canonicalize: canonicalize?)
           add_triple(attr, subject, predicate, lit)
         end
       end
@@ -391,6 +421,8 @@ def nodeElement(el, ec)
             when "nodeID"     then nodeID = attr.value
             else                   attrs[attr] = attr.value
             end
+          elsif attr.namespace.href == RDF::ITS.to_s
+            # No production. Direction already extracted
           else
             attrs[attr] = attr.value
           end
@@ -430,6 +462,7 @@ def nodeElement(el, ec)
             literal_opts[:datatype] = uri(datatype)
           else
             literal_opts[:language] = child_ec.language
+            literal_opts[:direction] = (child_ec.direction if child_ec.language)
           end
           literal = RDF::Literal.new(child.inner_text, **literal_opts)
           add_triple(child, subject, predicate, literal)
@@ -527,7 +560,7 @@ def nodeElement(el, ec)
 
           if attrs.empty? && resourceAttr.nil? && nodeID.nil?
             
-            literal = RDF::Literal.new("", language: ec.language)
+            literal = RDF::Literal.new("", language: ec.language, direction: ec.direction)
             add_triple(child, subject, predicate, literal)
             
             # Reification
@@ -552,7 +585,7 @@ def nodeElement(el, ec)
                 next unless is_propertyAttr?(attr)
 
                 # Attributes not in RDF.type
-                lit = RDF::Literal.new(val, language: child_ec.language)
+                lit = RDF::Literal.new(val, language: child_ec.language, direction: child_ec.direction)
                 add_triple(child, resource, attr.uri, lit)
               end
             end
diff --git a/lib/rdf/rdfxml/reader/nokogiri.rb b/lib/rdf/rdfxml/reader/nokogiri.rb
@@ -54,6 +54,30 @@ def base
           @base == false ? nil : @base
         end
 
+        ##
+        # Element direction
+        #
+        # @return [String]
+        def direction
+          attribute_with_ns("dir", RDF::ITS.to_s)
+        end
+
+        ##
+        # ITS version
+        #
+        # @return [String]
+        def its_version
+          attribute_with_ns("version", RDF::ITS.to_s)
+        end
+
+        ##
+        # RDF version
+        #
+        # @return [String]
+        def version
+          attribute_with_ns("version", RDF.to_uri.to_s)
+        end
+
         ##
         # Monkey patch attribute_with_ns, to insure nil is returned for #null?
         #
diff --git a/lib/rdf/rdfxml/reader/rexml.rb b/lib/rdf/rdfxml/reader/rexml.rb
@@ -74,6 +74,30 @@ def base
           @base == false ? nil : @base
         end
 
+        ##
+        # Element direction
+        #
+        # @return [String]
+        def direction
+          @node.attribute("dir", RDF::ITS.to_s)
+        end
+
+        ##
+        # ITS version
+        #
+        # @return [String]
+        def its_version
+          @node.attribute("version", RDF::ITS.to_s)
+        end
+
+        ##
+        # RDF version
+        #
+        # @return [String]
+        def version
+          @node.attribute("version", RDF.to_uri.to_s)
+        end
+
         def attribute_with_ns(name, namespace)
           @node.attribute(name, namespace)
         end
@@ -120,7 +144,11 @@ def namespaces
         end
 
         def namespace
-          Namespace.new(@node.namespace, @node.prefix) unless @node.namespace.to_s.empty?
+          if @node.namespace
+            Namespace.new(@node.namespace, @node.prefix)
+          else
+            Namespace.new(RDF::XML.to_s, 'rdf')
+          end
         end
 
         def add_namespace(prefix, uri)
diff --git a/rdf-rdfxml.gemspec b/rdf-rdfxml.gemspec
@@ -28,10 +28,11 @@ Gem::Specification.new do |gem|
   gem.required_ruby_version = '>= 3.0'
   gem.requirements          = []
 
+  gem.add_runtime_dependency     'builder',         '~> 3.2', '>= 3.2.4'
   gem.add_runtime_dependency     'rdf',             '~> 3.3'
   gem.add_runtime_dependency     'rdf-xsd',         '~> 3.3'
   gem.add_runtime_dependency     'htmlentities',    '~> 4.3'
-  gem.add_runtime_dependency     'builder',         '~> 3.2', '>= 3.2.4'
+  gem.add_runtime_dependency     'logger',          '~> 1.5'
 
   gem.add_development_dependency 'getoptlong',      '~> 0.2'
   gem.add_development_dependency 'json-ld',         '>= 3.3'
diff --git a/spec/reader_spec.rb b/spec/reader_spec.rb
@@ -224,6 +224,106 @@
         expect(graph).to be_equivalent_graph(expected, logger: logger)
       end
 
+      context :direction do
+        {
+          "Language with no direction": {
+            input: %(<?xml version="1.0" ?>
+              <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+                xmlns:ex="http://example.org/"
+                xml:lang="en"
+                rdf:version="1.2">
+                <rdf:Description rdf:about="http://example.org/joe" ex:name="bar" />
+              </rdf:RDF>),
+            expected: %(
+              <http://example.org/joe> <http://example.org/name> "bar"@en .
+            )
+          },
+          "Language with direction": {
+            input: %(<?xml version="1.0" ?>
+              <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+                xmlns:ex="http://example.org/"
+                xmlns:its="http://www.w3.org/2005/11/its"
+                its:version="2.0"
+                its:dir="ltr"
+                xml:lang="en"
+                rdf:version="1.2">
+                <rdf:Description rdf:about="http://example.org/joe" ex:name="bar" />
+              </rdf:RDF>),
+            expected: %(
+              <http://example.org/joe> <http://example.org/name> "bar"@en--ltr .
+            )
+          },
+          "Language with direction and no RDF version": {
+            input: %(<?xml version="1.0" ?>
+              <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+                xmlns:ex="http://example.org/"
+                xmlns:its="http://www.w3.org/2005/11/its"
+                its:version="2.0"
+                its:dir="ltr"
+                xml:lang="en">
+                <rdf:Description rdf:about="http://example.org/joe" ex:name="bar" />
+              </rdf:RDF>),
+            expected: %(
+              <http://example.org/joe> <http://example.org/name> "bar"@en .
+            )
+          },
+          "Language with direction and no ITS version": {
+            # FIXME: Should this fail, or raise an exception?
+            input: %(<?xml version="1.0" ?>
+              <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+                xmlns:ex="http://example.org/"
+                xmlns:its="http://www.w3.org/2005/11/its"
+                its:dir="ltr"
+                xml:lang="en"
+                rdf:version="1.2">
+                <rdf:Description rdf:about="http://example.org/joe" ex:name="bar" />
+              </rdf:RDF>),
+            expected: %(
+              <http://example.org/joe> <http://example.org/name> "bar"@en--ltr .
+            )
+          },
+          "Language with direction on element directly": {
+            input: %(<?xml version="1.0" ?>
+              <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+                xmlns:ex="http://example.org/"
+                xmlns:its="http://www.w3.org/2005/11/its"
+                rdf:version="1.2">
+                <rdf:Description rdf:about="http://example.org/joe">
+                  <ex:name xml:lang="en" its:version="2.0" its:dir="ltr" >bar</ex:name>
+                </rdf:Description>
+              </rdf:RDF>),
+            expected: %(
+              <http://example.org/joe> <http://example.org/name> "bar"@en--ltr .
+            )
+          },
+          "Direction with no language": {
+            input: %(<?xml version="1.0" ?>
+              <rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
+                xmlns:ex="http://example.org/"
+                xmlns:its="http://www.w3.org/2005/11/its"
+                its:version="2.0"
+                its:dir="ltr"
+                rdf:version="1.2">
+                <rdf:Description rdf:about="http://example.org/joe" ex:name="bar" />
+              </rdf:RDF>),
+            expected: %(
+              <http://example.org/joe> <http://example.org/name> "bar" .
+            )
+          },
+        }.each do |title, properties|
+          it title do
+            if properties[:exception]
+              expect do
+                parse(properties[:input], validate: true)
+              end.to raise_error(RDF::ReaderError)
+            else
+              graph = parse(properties[:input], validate: true)
+              expect(graph).to be_equivalent_graph(properties[:expected], logger: logger)
+            end
+          end
+        end
+      end
+
       context :exceptions do
         it "should raise an error if rdf:aboutEach is used, as per the negative parser test rdfms-abouteach-error001 (rdf:aboutEach attribute)" do
           sampledoc = %q(<?xml version="1.0" ?>

Original file line number	Diff line number	Diff line change
`@@ -4,6 +4,9 @@`
`4`	`4`	`module RDF`
`5`	`5`	`XML = Class.new(Vocabulary("http://www.w3.org/XML/1998/namespace"))`
`6`	`6`
	`7`	`+ # Internationalization Tag Set`
	`8`	`+ ITS = Class.new(Vocabulary("http://www.w3.org/2005/11/its"))`
	`9`	`+`
`7`	`10`	`##`
`8`	`11`	# `RDF::RDFXML` is an RDF/XML extension for RDF.rb.
`9`	`12`	`#`