rdf-turtle 0.1.0 → 0.1.1

data/lib/rdf/turtle/reader.rb

@@ -96,7 +96,7 @@ module RDF::Turtle
       next unless phase == :finish
       prefix = current[:prefix]
       iri = current[:resource]
-      callback.call(:trace, "prefixID", "Defined prefix #{prefix.inspect} mapping to #{iri.inspect}")
+      callback.call(:trace, "prefixID", lambda {"Defined prefix #{prefix.inspect} mapping to #{iri.inspect}"})
       reader.prefix(prefix, iri)
     end
     
@@ -104,19 +104,17 @@ module RDF::Turtle
     production(:base) do |reader, phase, input, current, callback|
       next unless phase == :finish
       iri = current[:resource]
-      callback.call(:trace, "base", "Defined base as #{iri}")
+      callback.call(:trace, "base", lambda {"Defined base as #{iri}"})
       reader.options[:base_uri] = iri
     end
     
     # [9] verb ::= predicate | "a"
     production(:verb) do |reader, phase, input, current, callback|
-      next unless phase == :finish
       input[:predicate] = current[:resource] if phase == :finish
     end
 
     # [10] subject ::= IRIref | blank
     production(:subject) do |reader, phase, input, current, callback|
-      next unless phase == :finish
       input[:subject] = current[:resource] if phase == :finish
     end
 
@@ -127,7 +125,7 @@ module RDF::Turtle
         # Part of an rdf:List collection
         input[:object_list] << current[:resource]
       else
-        callback.call(:trace, "object", "current: #{current.inspect}")
+        callback.call(:trace, "object", lambda {"current: #{current.inspect}"})
         callback.call(:statement, "object", input[:subject], input[:predicate], current[:resource])
       end
     end
@@ -176,6 +174,10 @@ module RDF::Turtle
     ##
     # Initializes a new reader instance.
     #
+    # Note, the spec does not define a default mapping for the empty prefix,
+    # but it is so commonly used in examples that we define it to be the
+    # empty string anyway, except when validating.
+    #
     # @param  [String, #to_s]          input
     # @param  [Hash{Symbol => Object}] options
     # @option options [Hash]     :prefixes     (Hash.new)
@@ -199,6 +201,7 @@ module RDF::Turtle
     def initialize(input = nil, options = {}, &block)
       super do
         @options = {:anon_base => "b0", :validate => false}.merge(options)
+        @options = {:prefixes => {nil => ""}}.merge(@options) unless @options[:validate]
 
         debug("base IRI") {base_uri.inspect}
         
@@ -228,16 +231,21 @@ module RDF::Turtle
     def each_statement(&block)
       @callback = block
 
-      parse(@input, START.to_sym, @options.merge(:branch => BRANCH, :follow => FOLLOW)) do |context, *data|
+      parse(@input, START.to_sym, @options.merge(:branch => BRANCH,
+                                                 :first => FIRST,
+                                                 :follow => FOLLOW)
+      ) do |context, *data|
+        loc = data.shift
         case context
         when :statement
-          add_triple(*data)
+          add_statement(loc, RDF::Statement.from(data))
         when :trace
-          debug(*data)
+          debug(loc, *data)
         end
       end
     rescue RDF::LL1::Parser::Error => e
-      error("each_statement", e.message, :backtrace => e.backtrace)
+      debug("Parsing completed with errors:\n\t#{e.message}")
+      raise RDF::ReaderError, e.message if validate?
     end
     
     ##
@@ -262,8 +270,7 @@ module RDF::Turtle
     # @param [URI, Node, Literal] object:: the object of the statement
     # @return [Statement]:: Added statement
     # @raise [RDF::ReaderError]:: Checks parameter types and raises if they are incorrect if parsing mode is _validate_.
-    def add_triple(node, subject, predicate, object)
-      statement = RDF::Statement.new(subject, predicate, object)
+    def add_statement(node, statement)
       if statement.valid?
         debug(node) {"generate statement: #{statement}"}
         @callback.call(statement)
@@ -334,28 +341,21 @@ module RDF::Turtle
       @bnode_cache[value.to_s] ||= RDF::Node.new(value)
     end
 
-    # @param [String] str Error string
-    # @param [Hash] options
-    # @option options [URI, #to_s] :production
-    # @option options [Token] :token
-    def error(node, message, options = {})
-      if !validate? && !options[:fatal]
-        debug(node, message, options)
-      else
-        raise RDF::ReaderError, message, options[:backtrace]
-      end
-    end
-
     ##
     # Progress output when debugging
     # @param [String] node relative location in input
     # @param [String] message ("")
     # @yieldreturn [String] added to message
-    def debug(node, message = "", options = {})
+    def debug(*args)
       return unless @options[:debug] || RDF::Turtle.debug?
+      options = args.last.is_a?(Hash) ? args.pop : {}
       depth = options[:depth] || self.depth
-      message += yield if block_given?
-      str = "[#{@lineno}]#{' ' * depth}#{node}: #{message}"
+      message = args.pop
+      message = message.call if message.is_a?(Proc)
+      args << message if message
+      args << yield if block_given?
+      message = "#{args.join(': ')}"
+      str = "[#{@lineno}]#{' ' * depth}#{message}"
       @options[:debug] << str if @options[:debug].is_a?(Array)
       $stderr.puts(str) if RDF::Turtle.debug?
     end

data/lib/rdf/turtle/terminals.rb

@@ -14,6 +14,7 @@ module RDF::Turtle
                            [\\uF900-\\uFDCF]|[\\uFDF0-\\uFFFD]|[\\u{10000}-\\u{EFFFF}]
                          EOS
       U_CHARS2         = Regexp.compile("\\u00B7|[\\u0300-\\u036F]|[\\u203F-\\u2040]")
+      IRI_RANGE        = Regexp.compile("[[^<>\"{}|^`\\\\]&&[^\\x00-\\x20]]")   # [^<>\"{}|^`\\] - [#x00-#x20]
     else
       ##
       # UTF-8 regular expressions for Ruby 1.8.x.
@@ -49,33 +50,42 @@ module RDF::Turtle
                            \\xCC[\\x80-\\xBF]|\\xCD[\\x80-\\xAF]|             (?# [\\u0300-\\u036F]|)
                            \\xE2\\x80\\xBF|\\xE2\\x81\\x80                    (?# [\\u203F-\\u2040])
                          EOS
+      IRI_RANGE        = Regexp.compile(<<-EOS.gsub(/\s+/, ''))
+                           \\x21|                                             (?# ")
+                           [\\x23-\\x3b]|\\x3d|                               (?# < & >)
+                           [\\x3f-\\x5b]|\\x5d|\\x5f|                         (?# \ ^ `)
+                           [\\x61-\\x7a]|                                     (?# { } |)
+                           [\\x7e-\\xff]
+                         EOS
     end
     UCHAR                = RDF::LL1::Lexer::UCHAR
 
     WS                   = / |\t|\r|\n  /                                         # [93s]
+    PERCENT              = /%[0-9A-Fa-f]{2}/                                      # [162s]
+    PN_LOCAL_ESC         = /\\[_~\.\-\!$\&'\(\)\*\+,;=:\/\?\#@%]/                 # [163s]                                      # [163s]
+    PLX                  = /#{PERCENT}|#{PN_LOCAL_ESC}/                           # [160s]
     PN_CHARS_BASE        = /[A-Z]|[a-z]|#{U_CHARS1}|#{UCHAR}/                     # [95s]
     PN_CHARS_U           = /_|#{PN_CHARS_BASE}/                                   # [96s]
     PN_CHARS             = /-|[0-9]|#{PN_CHARS_U}|#{U_CHARS2}/                    # [98s]
     PN_CHARS_BODY        = /(?:(?:\.|#{PN_CHARS})*#{PN_CHARS})?/
-    PN_LOCAL             = /(?:[0-9]|#{PN_CHARS_U})#{PN_CHARS_BODY}/              # [100s]
+    PN_LOCAL_BODY        = /(?:(?:\.|#{PN_CHARS}|#{PLX})*(?:#{PN_CHARS}|#{PLX}))?/
+    PN_LOCAL             = /(?:[0-9]|#{PN_CHARS_U}|#{PLX})#{PN_LOCAL_BODY}/       # [100s]
 
     EXPONENT             = /[eE][+-]?[0-9]+/                                      # [86s]
                                                                                   
     ANON                 = /\[#{WS}*\]/                                           # [94s]
     BLANK_NODE_LABEL     = /_:#{PN_LOCAL}/                                        # [73s]
-    DECIMAL              = /(?:[0-9]+\.[0-9]*|\.[0-9]+)/                          # [78s]
-    DECIMAL_NEGATIVE     = /\-(?:[0-9]+\.[0-9]*|\.[0-9]+)/                        # [83s]
-    DECIMAL_POSITIVE     = /\+(?:[0-9]+\.[0-9]*|\.[0-9]+)/                        # [81s]
-    DOUBLE               = /(?:[0-9]+\.[0-9]*|\.[0-9]+|[0-9]+)#{EXPONENT}/        # [79s]
-    DOUBLE_NEGATIVE      = /\-(?:[0-9]+\.[0-9]*|\.[0-9]+|[0-9]+)#{EXPONENT}/      # [79s]
-    DOUBLE_POSITIVE      = /\+(?:[0-9]+\.[0-9]*|\.[0-9]+|[0-9]+)#{EXPONENT}/      # [79s]
+    DECIMAL              = /(?:[0-9]+\.[0-9]+|\.[0-9]+)/                          # [78s]
+    DECIMAL_NEGATIVE     = /\-(?:[0-9]+\.[0-9]+|\.[0-9]+)/                        # [83s]
+    DECIMAL_POSITIVE     = /\+(?:[0-9]+\.[0-9]+|\.[0-9]+)/                        # [81s]
+    DOUBLE               = /(?:[0-9]+\.[0-9]+|\.[0-9]+|[0-9]+)#{EXPONENT}/        # [79s]
+    DOUBLE_NEGATIVE      = /\-(?:[0-9]+\.[0-9]+|\.[0-9]+|[0-9]+)#{EXPONENT}/      # [79s]
+    DOUBLE_POSITIVE      = /\+(?:[0-9]+\.[0-9]+|\.[0-9]+|[0-9]+)#{EXPONENT}/      # [79s]
     ECHAR                = /\\[tbnrf\\"']/                                        # [91s]
     INTEGER              = /[0-9]+/                                               # [77s]
     INTEGER_NEGATIVE     = /\-[0-9]+/                                             # [83s]
     INTEGER_POSITIVE     = /\+[0-9]+/                                             # [80s]
-    # Spec confusion: the EBNF definition of IRI_REF seems malformed, and has no
-    # provision for \^, as discussed elsewhere in the spec.
-    IRI_REF              = /<(?:[^<>"{}|^`\\\x00-\x20]|#{U_CHARS1})*>/            # [70s]
+    IRI_REF              = /<(?:#{IRI_RANGE}|#{UCHAR})*>/                         # [70s]
     LANGTAG              = /@[a-zA-Z]+(?:-[a-zA-Z0-9]+)*/                         # [76s]
     PN_PREFIX            = /#{PN_CHARS_BASE}#{PN_CHARS_BODY}/                     # [99s]
     PNAME_NS             = /#{PN_PREFIX}?:/                                       # [71s]

data/lib/rdf/turtle/writer.rb

@@ -105,7 +105,7 @@ module RDF::Turtle
     end
 
     ##
-    # Addes a statement to be serialized
+    # Adds a statement to be serialized
     # @param  [RDF::Statement] statement
     # @return [void]
     def write_statement(statement)
@@ -113,7 +113,7 @@ module RDF::Turtle
     end
 
     ##
-    # Addes a triple to be serialized
+    # Adds a triple to be serialized
     # @param  [RDF::Resource] subject
     # @param  [RDF::URI]      predicate
     # @param  [RDF::Value]    object
@@ -135,7 +135,7 @@ module RDF::Turtle
 
       self.reset
 
-      debug {"\nserialize: graph: #{@graph.size}"}
+      debug("\nserialize") {"graph: #{@graph.size}"}
 
       preprocess
       start_document
@@ -166,14 +166,16 @@ module RDF::Turtle
       when u = @uri_to_prefix.keys.detect {|u| uri.index(u.to_s) == 0}
         # Use a defined prefix
         prefix = @uri_to_prefix[u]
-        prefix(prefix, u) unless u.to_s.empty? # Define for output
-        debug {"get_pname: add prefix #{prefix.inspect} => #{u}"}
-        uri.sub(u.to_s, "#{prefix}:")
+        unless u.to_s.empty?
+          prefix(prefix, u) unless u.to_s.empty?
+          debug("get_pname") {"add prefix #{prefix.inspect} => #{u}"}
+          uri.sub(u.to_s, "#{prefix}:")
+        end
       when @options[:standard_prefixes] && vocab = RDF::Vocabulary.each.to_a.detect {|v| uri.index(v.to_uri.to_s) == 0}
         prefix = vocab.__name__.to_s.split('::').last.downcase
         @uri_to_prefix[vocab.to_uri.to_s] = prefix
         prefix(prefix, vocab.to_uri) # Define for output
-        debug {"get_pname: add standard prefix #{prefix.inspect} => #{vocab.to_uri}"}
+        debug("get_pname") {"add standard prefix #{prefix.inspect} => #{vocab.to_uri}"}
         uri.sub(vocab.to_uri.to_s, "#{prefix}:")
       else
         nil
@@ -208,7 +210,7 @@ module RDF::Turtle
         prop_list << prop.to_s
       end
       
-      debug {"sort_properties: #{prop_list.join(', ')}"}
+      debug("sort_properties") {prop_list.join(', ')}
       prop_list
     end
 
@@ -246,7 +248,7 @@ module RDF::Turtle
     # @return [String]
     def format_uri(uri, options = {})
       md = relativize(uri)
-      debug {"relativize(#{uri.inspect}) => #{md.inspect}"} if md != uri.to_s
+      debug("relativize") {"#{uri.inspect} => #{md.inspect}"} if md != uri.to_s
       md != uri.to_s ? "<#{md}>" : (get_pname(uri) || "<#{uri}>")
     end
     
@@ -263,11 +265,9 @@ module RDF::Turtle
     protected
     # Output @base and @prefix definitions
     def start_document
-      @started = true
-      
       @output.write("#{indent}@base <#{base_uri}> .\n") unless base_uri.to_s.empty?
       
-      debug {"start_document: #{prefixes.inspect}"}
+      debug("start_document") {prefixes.inspect}
       prefixes.keys.sort_by(&:to_s).each do |prefix|
         @output.write("#{indent}@prefix #{prefix}: <#{prefixes[prefix]}> .\n")
       end
@@ -307,7 +307,7 @@ module RDF::Turtle
       # Add distinguished classes
       top_classes.each do |class_uri|
         graph.query(:predicate => RDF.type, :object => class_uri).map {|st| st.subject}.sort.uniq.each do |subject|
-          debug {"order_subjects: #{subject.inspect}"}
+          debug("order_subjects") {subject.inspect}
           subjects << subject
           seen[subject] = true
         end
@@ -339,7 +339,7 @@ module RDF::Turtle
     # prefixes.
     # @param [Statement] statement
     def preprocess_statement(statement)
-      #debug {"preprocess: #{statement.inspect}"}
+      #debug("preprocess") {statement.inspect}
       references = ref_count(statement.object) + 1
       @references[statement.object] = references
       @subjects[statement.subject] = true
@@ -370,12 +370,9 @@ module RDF::Turtle
     def reset
       @depth = 0
       @lists = {}
-      @namespaces = {}
       @references = {}
       @serialized = {}
       @subjects = {}
-      @shortNames = {}
-      @started = false
     end
 
     ##
@@ -392,28 +389,36 @@ module RDF::Turtle
       end
     end
 
-    private
+    protected
     
     ##
     # Add debug event to debug array, if specified
     # @param [String] message ("")
     # @yieldreturn [String] added to message
-    def debug(message = "", options = {})
+    def debug(*args)
       return unless @options[:debug] || RDF::Turtle.debug?
-      message += yield if block_given?
+      options = args.last.is_a?(Hash) ? args.pop : {}
+      depth = options[:depth] || @depth
+      message = args.pop
+      message = message.call if message.is_a?(Proc)
+      args << message if message
+      args << yield if block_given?
+      message = "#{args.join(': ')}"
       @options[:debug] << message if @options[:debug].is_a?(Array)
       $stderr.puts(message) if RDF::Turtle.debug?
     end
 
+    private
+
     # Checks if l is a valid RDF list, i.e. no nodes have other properties.
-    def is_valid_list(l)
-      #debug {"is_valid_list: #{l.inspect}"}
+    def is_valid_list?(l)
+      #debug("is_valid_list?") {l.inspect}
       return RDF::List.new(l, @graph).valid?
     end
     
     def do_list(l)
       list = RDF::List.new(l, @graph)
-      debug {"do_list: #{list.inspect}"}
+      debug("do_list") {list.inspect}
       position = :subject
       list.each_statement do |st|
         next unless st.predicate == RDF.first
@@ -425,8 +430,8 @@ module RDF::Turtle
     end
     
     def p_list(node, position)
-      return false if !is_valid_list(node)
-      #debug {"p_list: #{node.inspect}, #{position}"}
+      return false if !is_valid_list?(node)
+      #debug("p_list") {"#{node.inspect}, #{position}"}
 
       @output.write(position == :subject ? "(" : " (")
       @depth += 2
@@ -444,7 +449,7 @@ module RDF::Turtle
     def p_squared(node, position)
       return false unless p_squared?(node, position)
 
-      #debug {"p_squared: #{node.inspect}, #{position}"}
+      #debug("p_squared") {"#{node.inspect}, #{position}"}
       subject_done(node)
       @output.write(position == :subject ? '[' : ' [')
       @depth += 2
@@ -456,16 +461,16 @@ module RDF::Turtle
     end
     
     def p_default(node, position)
-      #debug {"p_default: #{node.inspect}, #{position}"}
+      #debug("p_default") {"#{node.inspect}, #{position}"}
       l = (position == :subject ? "" : " ") + format_value(node)
       @output.write(l)
     end
     
     def path(node, position)
-      debug do
-        "path: #{node.inspect}, " +
+      debug("path") do
+        "#{node.inspect}, " +
         "pos: #{position}, " +
-        "[]: #{is_valid_list(node)}, " +
+        "[]: #{is_valid_list?(node)}, " +
         "p2?: #{p_squared?(node, position)}, " +
         "rc: #{ref_count(node)}"
       end
@@ -473,7 +478,7 @@ module RDF::Turtle
     end
     
     def verb(node)
-      debug {"verb: #{node.inspect}"}
+      debug("verb") {node.inspect}
       if node == RDF.type
         @output.write(" a")
       else
@@ -482,7 +487,7 @@ module RDF::Turtle
     end
     
     def object_list(objects)
-      debug {"object_list: #{objects.inspect}"}
+      debug("object_list") {objects.inspect}
       return if objects.empty?
 
       objects.each_with_index do |obj, i|
@@ -499,7 +504,7 @@ module RDF::Turtle
       end
 
       prop_list = sort_properties(properties) - [RDF.first.to_s, RDF.rest.to_s]
-      debug {"predicate_list: #{prop_list.inspect}"}
+      debug("predicate_list") {prop_list.inspect}
       return if prop_list.empty?
 
       prop_list.each_with_index do |prop, i|
@@ -515,13 +520,13 @@ module RDF::Turtle
     end
     
     def s_squared?(subject)
-      ref_count(subject) == 0 && subject.is_a?(RDF::Node) && !is_valid_list(subject)
+      ref_count(subject) == 0 && subject.is_a?(RDF::Node) && !is_valid_list?(subject)
     end
     
     def s_squared(subject)
       return false unless s_squared?(subject)
       
-      debug {"s_squared: #{subject.inspect}"}
+      debug("s_squared") {subject.inspect}
       @output.write("\n#{indent} [")
       @depth += 1
       predicate_list(subject)
@@ -539,7 +544,7 @@ module RDF::Turtle
     end
     
     def statement(subject)
-      debug {"statement: #{subject.inspect}, s2?: #{s_squared?(subject)}"}
+      debug("statement") {"#{subject.inspect}, s2?: #{s_squared?(subject)}"}
       subject_done(subject)
       s_squared(subject) || s_default(subject)
       @output.puts

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: rdf-turtle
 version: !ruby/object:Gem::Version
-  version: 0.1.0
+  version: 0.1.1
   prerelease: 
 platform: ruby
 authors:
@@ -9,11 +9,11 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2011-09-27 00:00:00.000000000 Z
+date: 2011-12-22 00:00:00.000000000Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rdf
-  requirement: &2161375780 !ruby/object:Gem::Requirement
+  requirement: &2166587480 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -21,21 +21,32 @@ dependencies:
         version: 0.3.4
   type: :runtime
   prerelease: false
-  version_requirements: *2161375780
+  version_requirements: *2166587480
 - !ruby/object:Gem::Dependency
-  name: yard
-  requirement: &2161375180 !ruby/object:Gem::Requirement
+  name: open-uri-cached
+  requirement: &2166586740 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
       - !ruby/object:Gem::Version
-        version: 0.6.0
+        version: 0.0.4
   type: :development
   prerelease: false
-  version_requirements: *2161375180
+  version_requirements: *2166586740
+- !ruby/object:Gem::Dependency
+  name: spira
+  requirement: &2166586000 !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 0.0.12
+  type: :development
+  prerelease: false
+  version_requirements: *2166586000
 - !ruby/object:Gem::Dependency
   name: rspec
-  requirement: &2161374480 !ruby/object:Gem::Requirement
+  requirement: &2166585320 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -43,10 +54,21 @@ dependencies:
         version: 2.5.0
   type: :development
   prerelease: false
-  version_requirements: *2161374480
+  version_requirements: *2166585320
+- !ruby/object:Gem::Dependency
+  name: rdf-isomorphic
+  requirement: &2166584640 !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ! '>='
+      - !ruby/object:Gem::Version
+        version: 0.3.4
+  type: :development
+  prerelease: false
+  version_requirements: *2166584640
 - !ruby/object:Gem::Dependency
   name: rdf-n3
-  requirement: &2161373540 !ruby/object:Gem::Requirement
+  requirement: &2166577480 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -54,10 +76,10 @@ dependencies:
         version: 0.3.5
   type: :development
   prerelease: false
-  version_requirements: *2161373540
+  version_requirements: *2166577480
 - !ruby/object:Gem::Dependency
   name: rdf-spec
-  requirement: &2161372320 !ruby/object:Gem::Requirement
+  requirement: &2166576800 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
@@ -65,18 +87,18 @@ dependencies:
         version: 0.3.4
   type: :development
   prerelease: false
-  version_requirements: *2161372320
+  version_requirements: *2166576800
 - !ruby/object:Gem::Dependency
-  name: rdf-isomorphic
-  requirement: &2161347600 !ruby/object:Gem::Requirement
+  name: yard
+  requirement: &2166576000 !ruby/object:Gem::Requirement
     none: false
     requirements:
     - - ! '>='
       - !ruby/object:Gem::Version
-        version: 0.3.4
+        version: 0.6.0
   type: :development
   prerelease: false
-  version_requirements: *2161347600
+  version_requirements: *2166576000
 description: RDF::Turtle is an Turtle reader/writer for the RDF.rb library suite.
 email: public-rdf-ruby@w3.org
 executables: []
@@ -93,7 +115,6 @@ files:
 - lib/rdf/ll1/scanner.rb
 - lib/rdf/turtle/format.rb
 - lib/rdf/turtle/meta.rb
-- lib/rdf/turtle/patches.rb
 - lib/rdf/turtle/reader.rb
 - lib/rdf/turtle/terminals.rb
 - lib/rdf/turtle/version.rb
@@ -125,3 +146,4 @@ signing_key:
 specification_version: 3
 summary: Turtle reader/writer for Ruby.
 test_files: []
+has_rdoc: false