Allow contexts to be HTML documents, with preference towards script elements of type application/ld+json;profile=http://www.w3.org/ns/json-ld#context`.

For w3c/json-ld-syntax#66.

Author: gkellogg

lib/json/ld/api.rb

@@ -384,9 +384,12 @@ def self.frame(input, frame, expanded: false, **options)
       when IO, StringIO then MultiJson.load(frame.read)
       when String
         remote_doc = options[:documentLoader].call(frame)
-        case remote_doc.document
-        when String then MultiJson.load(remote_doc.document)
-        else remote_doc.document
+        if remote_doc.content_type == 'text/html'
+          load_html(remote_doc.document, url: context.to_s, profile: 'http://www.w3.org/ns/json-ld#frame')
+        elsif remote_doc.document.is_a?(String)
+          MultiJson.load(remote_doc.document)
+        else
+          remote_doc.document
         end
       end
 
@@ -587,22 +590,22 @@ class << self
       alias :fromRDF :fromRdf
     end
 
-    private
-    def validate_input(input, url:)
-      return unless defined?(JsonLint)
-      jsonlint = JsonLint::Linter.new
-      input = StringIO.new(input) unless input.respond_to?(:read)
-      unless jsonlint.check_stream(input)
-        raise JsonLdError::LoadingDocumentFailed, "url: #{url}\n" + jsonlint.errors[''].join("\n")
-      end
-      input.rewind
-    end
-
     ##
     # Load one or more script tags from an HTML source.
     # Unescapes and uncomments input, returns the internal representation
     # Yields document base
-    def load_html(input, url:, library: nil, extractAllScripts: false, **options)
+    # @param [String] input
+    # @param [String] url   Original URL
+    # @param [:nokogiri, :rexml] library (nil)
+    # @param [Boolean] extractAllScripts (false)
+    # @param [Boolean] profile (nil) Optional priortized profile when loading a single script by type.
+    # @param [Hash{Symbol => Object}] options
+    def load_html(input, url:,
+                         library: nil,
+                         extractAllScripts: false,
+                         profile: nil,
+                         **options)
+
       if input.is_a?(String)
         library ||= begin
           require 'nokogiri'
@@ -647,7 +650,8 @@ def load_html(input, url:, library: nil, extractAllScripts: false, **options)
         MultiJson.load(content, options)
       elsif extractAllScripts
         res = []
-        input.xpath("//script[starts-with(@type, 'application/ld+json')]").each do |element|
+        elements = input.xpath("//script[starts-with(@type, 'application/ld+json')]")
+        elements.each do |element|
           content = element.inner_html
           validate_input(content, url: url) if options[:validate]
           r = MultiJson.load(content, options)
@@ -660,7 +664,8 @@ def load_html(input, url:, library: nil, extractAllScripts: false, **options)
         res
       else
         # Find the first script with type application/ld+json.
-        element = input.at_xpath("//script[starts-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')]")
         content = element ? element.inner_html : "[]"
         validate_input(content, url: url) if options[:validate]
         MultiJson.load(content, options)
@@ -669,6 +674,25 @@ def load_html(input, url:, library: nil, extractAllScripts: false, **options)
       raise JSON::LD::JsonLdError::InvalidScriptElement, e.message
     end
 
+    # Use from a differnet location
+    # @see {#load_html}
+    def self.load_html(input, **options)
+      self.new([], nil).load_html(input, **options)
+    end
+
+    ##
+    # Validate JSON using JsonLint, if loaded
+    private
+    def validate_input(input, url:)
+      return unless defined?(JsonLint)
+      jsonlint = JsonLint::Linter.new
+      input = StringIO.new(input) unless input.respond_to?(:read)
+      unless jsonlint.check_stream(input)
+        raise JsonLdError::LoadingDocumentFailed, "url: #{url}\n" + jsonlint.errors[''].join("\n")
+      end
+      input.rewind
+    end
+
     ##
     # A {RemoteDocument} is returned from a {documentLoader}.
     #

lib/json/ld/context.rb

@@ -486,10 +486,14 @@ def parse(local_context, remote_contexts: [], from_term: nil)
               context_opts.delete(:headers)
               @options[:documentLoader].call(context.to_s, context_opts) do |remote_doc|
                 # 3.2.5) Dereference context. If the dereferenced document has no top-level JSON object with an @context member, an invalid remote context has been detected and processing is aborted; otherwise, set context to the value of that member.
-                jo = case remote_doc.document
-                when String then MultiJson.load(remote_doc.document)
-                else remote_doc.document
+                jo = if remote_doc.content_type == 'text/html'
+                  API.load_html(remote_doc.document, url: context.to_s, profile: 'http://www.w3.org/ns/json-ld#context')
+                elsif remote_doc.document.is_a?(String)
+                  MultiJson.load(remote_doc.document)
+                else
+                  remote_doc.document
                 end
+
                 raise JsonLdError::InvalidRemoteContext, "#{context}" unless jo.is_a?(Hash) && jo.has_key?('@context')
                 context = jo['@context']
               end
@@ -701,6 +705,10 @@ def create_term_definition(local_context, term, defined, from_term: nil, protect
           raise JsonLdError::InvalidIRIMapping, "non-absolute @reverse IRI: #{definition.id} on term #{term.inspect}" unless
             definition.id.is_a?(RDF::URI) && definition.id.absolute?
 
+          if term.include?(':') && (term_iri = expand_iri(term)) != definition.id
+            raise JsonLdError::InvalidIRIMapping, "term #{term} expands to #{definition.id}, not #{term_iri}"
+          end
+
           warn "[DEPRECATION] Blank Node terms deprecated in JSON-LD 1.1." if (processingMode || "json-ld-1.1") >= "json-ld-1.1" && definition.id.start_with?("_:")
 
           # If value contains an @container member, set the container mapping of definition to its value; if its value is neither @set, @index, @type, @id, an absolute IRI nor null, an invalid reverse property error has been detected (reverse properties only support set- and index-containers) and processing is aborted.
@@ -723,6 +731,10 @@ def create_term_definition(local_context, term, defined, from_term: nil, protect
           raise JsonLdError::InvalidKeywordAlias, "expected value of @id to not be @context on term #{term.inspect}" if
             definition.id == '@context'
 
+          if term.match?(/:[^:]/) && (term_iri = expand_iri(term)) != definition.id
+            raise JsonLdError::InvalidIRIMapping, "term #{term} expands to #{definition.id}, not #{term_iri}"
+          end
+
           warn "[DEPRECATION] Blank Node terms deprecated in JSON-LD 1.1." if (processingMode || "json-ld-1.1") >= "json-ld-1.1" && definition.id.start_with?("_:")
 
           # If id ends with a gen-delim, it may be used as a prefix for simple terms

spec/context_spec.rb

@@ -32,7 +32,9 @@ def containers
         "homepage": {"@id": "http://xmlns.com/foaf/0.1/homepage", "@type": "@id"},
         "avatar": {"@id": "http://xmlns.com/foaf/0.1/avatar", "@type": "@id"}
       }
-    }), base_uri: "http://example.com/context")
+    }),
+    base_uri: "http://example.com/context",
+    headers: {content_type: "application/ld+json"})
   end
   subject {context}
 
@@ -78,6 +80,90 @@ def containers
         }, logger)
       end
 
+      it "retrieves and parses a remote context document in HTML using the context profile" do
+        remote_doc =
+          RDF::Util::File::RemoteDocument.new(%q(
+            <html><head>
+            <script>Not This</script>
+            <script type="application/ld+json">
+            {
+              "@context": {
+                "homepage": {"@id": "http://example.com/this-would-be-wrong", "@type": "@id"},
+                "avatar": {"@id": "http://example.com/this-would-be-wrong", "@type": "@id"}
+              }
+            }
+            </script>
+            <script type="application/ld+json;profile=http://www.w3.org/ns/json-ld#context">
+            {
+              "@context": {
+                "xsd": "http://www.w3.org/2001/XMLSchema#",
+                "name": "http://xmlns.com/foaf/0.1/name",
+                "homepage": {"@id": "http://xmlns.com/foaf/0.1/homepage", "@type": "@id"},
+                "avatar": {"@id": "http://xmlns.com/foaf/0.1/avatar", "@type": "@id"}
+              }
+            }
+            </script>
+            <script type="application/ld+json;profile=http://www.w3.org/ns/json-ld#context">
+            {
+              "@context": {
+                "homepage": {"@id": "http://example.com/this-would-also-be-wrong", "@type": "@id"},
+                "avatar": {"@id": "http://example.com/this-would-also-be-wrong", "@type": "@id"}
+              }
+            }
+            </script>
+            </head></html>
+            ),
+            base_uri: "http://example.com/context",
+            headers: {content_type: "text/html"})
+          JSON::LD::Context::PRELOADED.clear
+        expect(JSON::LD::API).to receive(:documentLoader).with("http://example.com/context", anything).and_yield(remote_doc)
+        ec = subject.parse("http://example.com/context")
+        expect(ec.send(:mappings)).to produce({
+          "xsd"      => "http://www.w3.org/2001/XMLSchema#",
+          "name"     => "http://xmlns.com/foaf/0.1/name",
+          "homepage" => "http://xmlns.com/foaf/0.1/homepage",
+          "avatar"   => "http://xmlns.com/foaf/0.1/avatar"
+        }, logger)
+      end
+
+      it "retrieves and parses a remote context document in HTML" do
+        remote_doc =
+          RDF::Util::File::RemoteDocument.new(%q(
+            <html><head>
+            <script>Not This</script>
+            <script type="application/ld+json">
+            {
+              "@context": {
+                "xsd": "http://www.w3.org/2001/XMLSchema#",
+                "name": "http://xmlns.com/foaf/0.1/name",
+                "homepage": {"@id": "http://xmlns.com/foaf/0.1/homepage", "@type": "@id"},
+                "avatar": {"@id": "http://xmlns.com/foaf/0.1/avatar", "@type": "@id"}
+              }
+            }
+            </script>
+            <script type="application/ld+json">
+            {
+              "@context": {
+                "homepage": {"@id": "http://example.com/this-would-also-be-wrong", "@type": "@id"},
+                "avatar": {"@id": "http://example.com/this-would-also-be-wrong", "@type": "@id"}
+              }
+            }
+            </script>
+            </head></html>
+            ),
+            base_uri: "http://example.com/context",
+            headers: {content_type: "text/html"})
+          JSON::LD::Context::PRELOADED.clear
+        expect(JSON::LD::API).to receive(:documentLoader).with("http://example.com/context", anything).and_yield(remote_doc)
+        ec = subject.parse("http://example.com/context")
+        expect(ec.send(:mappings)).to produce({
+          "xsd"      => "http://www.w3.org/2001/XMLSchema#",
+          "name"     => "http://xmlns.com/foaf/0.1/name",
+          "homepage" => "http://xmlns.com/foaf/0.1/homepage",
+          "avatar"   => "http://xmlns.com/foaf/0.1/avatar"
+        }, logger)
+      end
+
       it "notes non-existing @context" do
         expect {subject.parse(StringIO.new("{}"))}.to raise_error(JSON::LD::JsonLdError::InvalidRemoteContext)
       end
@@ -376,6 +462,16 @@ def containers
         "@prefix string" => {"foo" => {"@id" => 'http://example.org/', "@prefix" => "str"}},
         "@prefix array" => {"foo" => {"@id" => 'http://example.org/', "@prefix" => []}},
         "@prefix object" => {"foo" => {"@id" => 'http://example.org/', "@prefix" => {}}},
+        "IRI term expands to different IRI" => {
+          "ex" => "http://example.com/",
+          "ex2" => "http://example.com/2/",
+          "ex:foo" => "ex2:foo"
+      },
+        "IRI term expands to different IRI (reverse)" => {
+          "ex" => "http://example.com/",
+          "ex2" => "http://example.com/2/",
+          "ex:foo" => {"@reverse" => "ex2:foo"}
+        }
       }.each do |title, context|
         it title do
           expect {