rdf-trig 1.0.1 → 1.0.3

data/lib/rdf/trig/reader.rb

@@ -71,16 +71,19 @@ module RDF::TriG
       case token.value
       when 'A', 'a'           then input[:resource] = RDF.type
       when 'true', 'false'    then input[:resource] = RDF::Literal::Boolean.new(token.value)
-      when '@base', '@prefix' then input[:lang] = token.value[1..-1]
+      when '.'                then input[:terminated] = true
       else                         input[:string] = token.value
       end
     end
 
+    terminal(:GRAPH,      /graph/i) do |prod, token, input|
+      input[:string_value] = token.value
+    end
     terminal(:PREFIX,      PREFIX) do |prod, token, input|
-      input[:string_value] = token.value.downcase
+      input[:string_value] = token.value
     end
     terminal(:BASE,      BASE) do |prod, token, input|
-      input[:string_value] = token.value.downcase
+      input[:string_value] = token.value
     end
 
     terminal(:LANGTAG,              LANGTAG) do |prod, token, input|
@@ -88,47 +91,72 @@ module RDF::TriG
     end
 
     # Productions
-    # [3g] graph defines the basic creation of context
-    start_production(:graph) do |input, current, callback|
+    # [2g] block defines the basic creation of context
+    start_production(:block) do |input, current, callback|
       callback.call(:context, "graph", nil)
     end
-    production(:graph) do |input, current, callback|
+    production(:block) do |input, current, callback|
       callback.call(:context, "graph", nil)
     end
 
-    # [4g] graphName
-    # Normally, just returns the resource, but if called from [3g], also
-    # sets the context for triples defined within that graph
-    production(:graphName) do |input, current, callback|
+    # [7g] labelOrSubject
+    # Sets the context for triples defined within that graph
+    production(:labelOrSubject) do |input, current, callback|
       # If input contains set_graph_iri, use the returned value to set @context
-      debug("graphName") {"Set graph context to #{current[:resource]}"}
-      callback.call(:context, "graphName", current[:resource])
+      debug(":labelOrSubject") {"Set graph context to #{current[:resource]}"}
+      callback.call(:context, "labelOrSubject", current[:resource])
+      input[:resource] = current[:resource]
     end
 
+    # _triplesOrGraph_2 ::= predicateObjectList '.'
+    start_production(:_triplesOrGraph_2) do |input, current, callback|
+      # Default graph after all
+      callback.call(:context, "graph", nil)
+      debug("_triplesOrGraph_2") {"subject: #{current[:resource]}"}
+      current[:subject] = input[:resource]
+    end
 
     # Productions
     # [4] prefixID defines a prefix mapping
     production(:prefixID) do |input, current, callback|
       prefix = current[:prefix]
       iri = current[:resource]
+      lexical = current[:string_value]
+      terminated = current[:terminated]
       debug("prefixID") {"Defined prefix #{prefix.inspect} mapping to #{iri.inspect}"}
+      if lexical.start_with?('@') && lexical != '@prefix'
+        error(:prefixID, "should be downcased")
+      elsif lexical == '@prefix'
+        error(:prefixID, "directive not terminated") unless terminated
+      else
+        error(:prefixID, "directive should not be terminated") if terminated
+      end
       prefix(prefix, iri)
     end
 
     # [5] base set base_uri
     production(:base) do |input, current, callback|
       iri = current[:resource]
+      lexical = current[:string_value]
+      terminated = current[:terminated]
       debug("base") {"Defined base as #{iri}"}
+      if lexical.start_with?('@') && lexical != '@base'
+        error(:base, "should be downcased")
+      elsif lexical == '@base'
+        error(:base, "directive not terminated") unless terminated
+      else
+        error(:base, "directive should not be terminated") if terminated
+      end
       options[:base_uri] = iri
     end
 
-    # [6] triples
-    start_production(:triples) do |input, current, callback|
+    # [52s]  triplesBlock
+    start_production(:triplesBlock) do |input, current, callback|
       # Note production as triples for blankNodePropertyList
       # to set :subject instead of :resource
       current[:triples] = true
     end
-    production(:triples) do |input, current, callback|
+    production(:triplesBlock) do |input, current, callback|
       # Note production as triples for blankNodePropertyList
       # to set :subject instead of :resource
       current[:triples] = true
@@ -139,7 +167,7 @@ module RDF::TriG
       input[:predicate] = current[:resource]
     end
 
-    # [10] subject ::= IRIref | BlankNode | collection
+    # [10] subject ::= iri | blank
     start_production(:subject) do |input, current, callback|
       current[:triples] = nil
     end

data/lib/rdf/trig/streaming_writer.rb

@@ -0,0 +1,42 @@
+module RDF::TriG
+  ##
+  # Streaming writer interface
+  # @author [Gregg Kellogg](http://greggkellogg.net/)
+  module StreamingWriter
+    ##
+    # Write out a statement, retaining current
+    # `subject` and `predicate` to create more compact output
+    # @return [void] `self`
+    def stream_statement(statement)
+      if statement.context != @streaming_context
+        stream_epilogue
+        if statement.context
+          @output.write "#{format_term(statement.context)} {"
+        end
+        @streaming_context, @streaming_subject, @streaming_predicate = statement.context, statement.subject, statement.predicate
+        @output.write "#{format_term(statement.subject)} "
+        @output.write "#{format_term(statement.predicate)} "
+      elsif statement.subject != @streaming_subject
+        @output.write " .\n#{indent(@streaming_subject ? 1 : 0)}"
+        @streaming_subject, @streaming_predicate = statement.subject, statement.predicate
+        @output.write "#{format_term(statement.subject)} "
+        @output.write "#{format_term(statement.predicate)} "
+      elsif statement.predicate != @streaming_predicate
+        @streaming_predicate = statement.predicate
+        @output.write ";\n#{indent(@streaming_subject ? 2 : 1)}#{format_term(statement.predicate)} "
+      else
+        @output.write ",\n#{indent(@streaming_subject ? 3 : 2)}"
+      end
+      @output.write("#{format_term(statement.object)}")
+    end
+
+    ##
+    # Complete open statements
+    # @return [void] `self`
+    def stream_epilogue
+      @output.puts " }" if @streaming_context
+    end
+
+    private
+  end
+end

data/lib/rdf/trig/writer.rb

@@ -1,4 +1,5 @@
 require 'rdf/turtle'
+require 'rdf/trig/streaming_writer'
 
 module RDF::TriG
   ##
@@ -34,6 +35,13 @@ module RDF::TriG
   #     end
   #   end
   #
+  # @example Serializing RDF statements to a string in streaming mode
+  #   RDF::TriG::Writer.buffer(:stream => true) do |writer|
+  #     repo.each_statement do |statement|
+  #       writer << statement
+  #     end
+  #   end
+  #
   # The writer will add prefix definitions, and use them for creating @prefix definitions, and minting QNames
   #
   # @example Creating @base and @prefix definitions in output
@@ -48,6 +56,7 @@ module RDF::TriG
   #
   # @author [Gregg Kellogg](http://greggkellogg.net/)
   class Writer < RDF::Turtle::Writer
+    include StreamingWriter
     format RDF::TriG::Format
     
     class ContextFilteredRepo
@@ -103,6 +112,8 @@ module RDF::TriG
     #   Maximum depth for recursively defining resources, defaults to 3
     # @option options [Boolean]  :standard_prefixes   (false)
     #   Add standard prefixes to @prefixes, if necessary.
+    # @option options [Boolean] :stream (false)
+    #   Do not attempt to optimize graph presentation, suitable for streaming large repositories.
     # @option options [String]   :default_namespace (nil)
     #   URI to use as default namespace, same as `prefixes\[nil\]`
     # @yield  [writer] `self`
@@ -111,6 +122,7 @@ module RDF::TriG
     # @yield  [writer]
     # @yieldparam [RDF::Writer] writer
     def initialize(output = $stdout, options = {}, &block)
+      reset
       super do
         # Set both @repo and @graph to a new repository.
         # When serializing a context, @graph is changed
@@ -125,45 +137,74 @@ module RDF::TriG
       end
     end
 
+
+    ##
+    # Adds a statement to be serialized
+    # @param  [RDF::Statement] statement
+    # @return [void]
+    def write_statement(statement)
+      case
+      when @options[:stream]
+        stream_statement(statement)
+      else
+        super
+      end
+    end
+
+    ##
+    # Write out declarations
+    # @return [void] `self`
+    def write_prologue
+      case
+      when @options[:stream]
+        stream_prologue
+      else
+        super
+      end
+    end
+
     ##
     # Outputs the TriG representation of all stored triples.
     #
     # @return [void]
     # @see    #write_triple
     def write_epilogue
-      @max_depth = @options[:max_depth] || 3
-      @base_uri = RDF::URI(@options[:base_uri])
+      case
+      when @options[:stream]
+        stream_epilogue
+      else
+        @max_depth = @options[:max_depth] || 3
+        @base_uri = RDF::URI(@options[:base_uri])
 
-      reset
+        reset
 
-      debug {"\nserialize: repo: #{@repo.size}"}
+        debug {"\nserialize: repo: #{@repo.size}"}
 
-      preprocess
-      start_document
+        preprocess
+        start_document
 
-      order_contexts.each do |ctx|
-        debug {"context: #{ctx.inspect}"}
-        reset
-        @depth = 2
+        order_contexts.each do |ctx|
+          debug {"context: #{ctx.inspect}"}
+          reset
+          @depth = 2
 
-        if ctx
-          @output.write("\n#{format_value(ctx)} {")
-        else
-          @output.write("\n{")
-        end
+          if ctx
+            @output.write("\n#{format_value(ctx)} {")
+          end
 
-        # Restrict view to the particular context
-        @graph = ContextFilteredRepo.new(@repo, ctx)
+          # Restrict view to the particular context
+          @graph = ContextFilteredRepo.new(@repo, ctx)
 
-        # Pre-process statements again, but in the specified context
-        @graph.each {|st| preprocess_statement(st)}
-        order_subjects.each do |subject|
-          unless is_done?(subject)
-            statement(subject)
+          # Pre-process statements again, but in the specified context
+          @graph.each {|st| preprocess_statement(st)}
+          order_subjects.each do |subject|
+            unless is_done?(subject)
+              statement(subject)
+            end
           end
-        end
 
-        @output.puts("}")
+          @output.puts("}") if ctx
+        end
       end
     end
 

metadata

@@ -1,14 +1,14 @@
 --- !ruby/object:Gem::Specification
 name: rdf-trig
 version: !ruby/object:Gem::Version
-  version: 1.0.1
+  version: 1.0.3
 platform: ruby
 authors:
 - Gregg Kellogg
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-05-30 00:00:00.000000000 Z
+date: 2013-09-09 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rdf
@@ -44,14 +44,14 @@ dependencies:
     requirements:
     - - ! '>='
       - !ruby/object:Gem::Version
-        version: 1.0.7
+        version: 1.0.9
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
     - - ! '>='
       - !ruby/object:Gem::Version
-        version: 1.0.7
+        version: 1.0.9
 - !ruby/object:Gem::Dependency
   name: open-uri-cached
   requirement: !ruby/object:Gem::Requirement
@@ -164,6 +164,7 @@ files:
 - lib/rdf/trig/format.rb
 - lib/rdf/trig/meta.rb
 - lib/rdf/trig/reader.rb
+- lib/rdf/trig/streaming_writer.rb
 - lib/rdf/trig/version.rb
 - lib/rdf/trig/writer.rb
 - lib/rdf/trig.rb
@@ -187,7 +188,7 @@ required_rubygems_version: !ruby/object:Gem::Requirement
       version: '0'
 requirements: []
 rubyforge_project: rdf-trig
-rubygems_version: 2.0.3
+rubygems_version: 2.0.5
 signing_key: 
 specification_version: 4
 summary: TriG reader/writer for Ruby.