rdfa_parser 0.1.3 → 0.1.5

data/History.txt

@@ -1,3 +1,10 @@
+=== 0.1.5
+* Gem is now deprecated, see [RDF::RDFa](http://github.org/gkellogg/rdf-rdfa)
+
+=== 0.1.4
+* Reduce dependence on HTML namespace in RDFa parser.
+* Add RDFa tests for HTML4 and HTML5
+
 === 0.1.3
 * Added rdfa-test-suite as sub-module and refactored RDFa tests to use them.
 * Fix but in white-space separated attributes

data/README.rdoc

@@ -4,12 +4,14 @@
 
 == DESCRIPTION:
 
+This Gem deprecated, see rdf-rdfa[http://github.org/gkellogg/rdf-rdfa] gem.
+
 RDFa parser written in pure Ruby. Yields each triple, or generate in-memory graph
 
 == FEATURES/PROBLEMS:
 
-* Most, but not all tests pass.
-* Tests performed by comparison to expected NTriples, not based on SPARQL query
+* Fully compliant RDFa 1.0 parser
+* RDFa tests use SPARQL for most tests due to Rasqal limitataions. Other tests compare directly against N-triples
 * Ultimately, this should be merged in with the Reddy gem to become part of a comprehensive Ruby RDF implementation
   * Support libraries (Graph, Triple, and URIRef are used substantially intact, to facilitate a future merger)
 

data/Rakefile

@@ -4,8 +4,8 @@ begin
   require 'jeweler'
   Jeweler::Tasks.new do |gemspec|
     gemspec.name = "rdfa_parser"
-    gemspec.summary = "RDFa parser written in pure Ruby."
-    gemspec.description = " Yields each triple, or generate in-memory graph"
+    gemspec.summary = "[Deprecated] RDFa parser written in pure Ruby."
+    gemspec.description = "This gem is no longer supported, please see http://rubygems.org/gems/rdf-rdfa"
     gemspec.email = "gregg@kellogg-assoc.com"
     gemspec.homepage = "http://github.com/gkellogg/rdfa_parser"
     gemspec.authors = ["Gregg Kellogg"]
@@ -20,33 +20,3 @@ begin
 rescue LoadError
   puts "Jeweler not available. Install it with: sudo gem install technicalpickles-jeweler -s http://gems.github.com"
 end
-
-require 'spec/rake/spectask'
-Spec::Rake::SpecTask.new(:spec) do |spec|
-  spec.libs << 'lib' << 'spec'
-  spec.spec_files = FileList['spec/**/*_spec.rb']
-end
-
-Spec::Rake::SpecTask.new(:rcov) do |spec|
-  spec.libs << 'lib' << 'spec'
-  spec.pattern = 'spec/**/*_spec.rb'
-  spec.rcov = true
-end
-
-task :spec => :check_dependencies
-
-task :default => :spec
-
-require 'rake/rdoctask'
-Rake::RDocTask.new do |rdoc|
-  if File.exist?('VERSION')
-    version = File.read('VERSION')
-  else
-    version = RdfaParser::VERSION
-  end
-
-  rdoc.rdoc_dir = 'rdoc'
-  rdoc.title = "rdfa_parser #{version}"
-  rdoc.rdoc_files.include('README*', "History.txt")
-  rdoc.rdoc_files.include('lib/**/*.rb')
-end

data/VERSION

@@ -1 +1 @@
-0.1.3
+0.1.5

data/lib/rdfa_parser/literal.rb

@@ -72,9 +72,9 @@ module RdfaParser
       # XMLLiteral and String values are encoding using C-style strings with
       # non-printable ASCII characters escaped.
       def format_as_n3(content, lang)
-        content = c_style(content.to_s)
+        content = escape(content.to_s)
         quoted_content = should_quote? ? "\"#{content}\"" : content
-        "#{quoted_content}^^<#{value}>#{lang ? "@#{lang}" : ""}"
+        "#{quoted_content}^^<#{value}>"
       end
 
       def format_as_trix(content, lang)
@@ -140,10 +140,19 @@ module RdfaParser
         '/'   =>  '/',
       } # :nodoc:
 
-      # Convert a UTF8 encoded Ruby string _string_ to a C-style string, encoded with
+      # Convert a UTF8 encoded Ruby string _string_ to an escaped string, encoded with
       # UTF16 big endian characters as \U????, and return it.
+      #
+      # \\:: Backslash
+      # \':: Single quote
+      # \":: Double quot
+      # \n:: ASCII Linefeed
+      # \r:: ASCII Carriage Return
+      # \t:: ASCCII Horizontal Tab
+      # \uhhhh:: character in BMP with Unicode value U+hhhh
+      # \U00hhhhhh:: character in plane 1-16 with Unicode value U+hhhhhh
       if String.method_defined?(:force_encoding)
-        def c_style(string) # :nodoc:
+        def escape(string) # :nodoc:
           string << '' # XXX workaround: avoid buffer sharing
           string.force_encoding(Encoding::ASCII_8BIT)
           string.gsub!(/["\\\/\x0-\x1f]/) { MAP[$&] }
@@ -163,7 +172,7 @@ module RdfaParser
           string
         end
       else
-        def c_style(string) # :nodoc:
+        def escape(string) # :nodoc:
           string = string.gsub(/["\\\/\x0-\x1f]/) { MAP[$&] }
           string.gsub!(/(
                           (?:
@@ -178,7 +187,12 @@ module RdfaParser
                           s.gsub!(/.{4}/n, '\\\\u\&')
                         }
           string
-       end
+        end
+      end
+      
+      # Reverse operation of escape
+      def unescape(string)
+        string.gsub(/\\([\\\'\"nrt]|u\h{4}|U00\h[6])/) {MAP.invert[$&]}
       end
     end
     
@@ -188,7 +202,7 @@ module RdfaParser
       end
 
       def format_as_n3(content, lang)
-        "\"#{c_style(content)}\"" + (lang ? "@#{lang}" : "")
+        "\"#{escape(content)}\"" + (lang ? "@#{lang}" : "")
         # Perform translation on value if it's typed
       end
 
@@ -229,7 +243,7 @@ module RdfaParser
       end
       
       def format_as_n3(content, lang)
-        "\"#{c_style(content)}\"^^<#{value}>"
+        "\"#{escape(content)}\"^^<#{value}>"
       end
 
       def format_as_trix(content, lang)

data/lib/rdfa_parser/namespace.rb

@@ -66,6 +66,11 @@ module RdfaParser
       end
     end
 
+    def eql?(other)
+      @short == other.short && @uri == other.uri && @fragment == other.fragment
+    end
+    alias_method :==, :eql?
+
     # Output xmlns attribute name
     def xmlns_attr
       short.nil? ? "xmlns" : "xmlns:#{short}"

data/lib/rdfa_parser/rdfa_parser.rb

@@ -50,7 +50,9 @@ module RdfaParser
 
     # Create new parser instance. Options:
     # _graph_:: Graph to parse into, otherwie a new RdfaParser::Graph instance is created
+    # @deprecated Please use rdf-rdfa[http://github.org/gkellogg/rdf-rdfa] instead
     def initialize(options = {})
+      warn "[DEPRECATION] `rdfa_parser` gem is deprecated.  Please use rdf-rb gem instead."
       options = {:graph => Graph.new}.merge(options)
       @debug = []
       BNode.reset # Start sequence anew
@@ -59,7 +61,7 @@ module RdfaParser
       @graph = options[:graph]
     end
   
-    # Parse XHRML+RDFa document from a string or input stream to closure or graph.
+    # Parse XHTML+RDFa document from a string or input stream to closure or graph.
     # _base_ indicates the base URI of the document.
     #
     # Optionally, the stream may be a Nokogiri::HTML::Document or Nokogiri::XML::Document
@@ -117,7 +119,7 @@ module RdfaParser
     # Parsing an RDFa document (this is *not* the recursive method)
     def parse_whole_document(doc, base)
       # find if the document has a base element
-      base_el = doc.xpath('/html:html/html:head/html:base', @namespace.xmlns_hash).first
+      base_el = doc.css('html>head>base').first
       if (base_el)
         base = base_el.attributes['href']
         # Strip any fragment from base

data/spec/literal_spec.rb

@@ -261,6 +261,21 @@ describe "Literals: " do
     end
   end
   
+  describe "an n3 literal" do
+    {
+      "simple literal"  => ["simple literal", nil, nil],
+      "backslash:\\" => ["backslash:\\\\", nil, nil],
+      "dquote:\"" => ["dquote:\\\"", nil, nil],
+      "newline:\n" => ["newline:\\n", nil, nil],
+      "return:\r" => ["return:\\r", nil, nil],
+      "tab:\t" => ["tab:\\t", nil, nil],
+    }.each_pair do |name, args|
+      specify "test #{name}" do
+        Literal.n3_encoded(*args).contents.should == name
+      end
+    end
+  end
+  
 #  it "build_from_language" do
 #    english = Literal.build_from_language("Have a nice day")
 #    english.encoding.should == "en"

data/spec/rdfa-triples/0115.nt

@@ -1,4 +1,4 @@
-<$TCPATH/0115.xhtml> <http://www.example.com/entity> "<string>" .
-<$TCPATH/0115.xhtml> <http://www.example.com/entity> "Ben & Co." .
-<$TCPATH/0115.xhtml> <http://www.example.com/entity> "@" .
-<$TCPATH/0115.xhtml> <http://www.example.com/entity> "@" .
+<$TCPATH/0115.xhtml> <http://www.example.com/entity1> ">" .
+<$TCPATH/0115.xhtml> <http://www.example.com/entity2> "Ben & Co." .
+<$TCPATH/0115.xhtml> <http://www.example.com/entity3> "@" .
+<$TCPATH/0115.xhtml> <http://www.example.com/entity4> "@" .

data/spec/rdfa_helper.rb

@@ -11,6 +11,9 @@ module RdfaHelper
     BASE_MANIFEST_URL = "http://rdfa.digitalbazaar.com/test-suite/"
     BASE_TEST_CASE_URL = "#{BASE_MANIFEST_URL}test-cases/"
     
+    HTMLRE = Regexp.new('([0-9]{4,4})\.xhtml')
+    TCPATHRE = Regexp.compile('\$TCPATH')
+    
     attr_accessor :about
     attr_accessor :name
     attr_accessor :contributor
@@ -98,8 +101,6 @@ module RdfaHelper
       end.compact.join("\n")
 
       namespaces.chop!  # Remove trailing newline
-      htmlre = Regexp.new('([0-9]{4,4})\.xhtml')
-      tcpathre = Regexp.compile('\$TCPATH')
       
       case suite
       when "xhtml"
@@ -107,16 +108,16 @@ module RdfaHelper
         %(<?xml version="1.0" encoding="UTF-8"?>\n) +
         %(<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML+RDFa 1.0//EN" "http://www.w3.org/MarkUp/DTD/xhtml-rdfa-1.dtd">\n) +
         %(<html xmlns="http://www.w3.org/1999/xhtml" version="XHTML+RDFa 1.0"\n)
-        head + "#{namespaces}>\n#{body.gsub(tcpathre, tcpath)}\n</html>"
+        head + "#{namespaces}>\n#{body.gsub(TCPATHRE, tcpath)}\n</html>"
       when "html4"
         head ="" +
         %(<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01//EN" "http://www.w3.org/TR/html4/strict.dtd">\n) +
         %(<html version="XHTML+RDFa 1.0"\n)
-        head + "#{namespaces}>\n#{body.gsub(tcpathre, tcpath).gsub(htmlre, '\1.html')}\n</html>"
+        head + "#{namespaces}>\n#{body.gsub(TCPATHRE, tcpath).gsub(HTMLRE, '\1.html')}\n</html>"
       when "html5"
         head = "<!DOCTYPE html>\n"
         head += namespaces.empty? ? %(<html version="HTML+RDFa 1.0">) : "<html\n#{namespaces}>"
-        head + "\n#{body.gsub(tcpathre, tcpath).gsub(htmlre, '\1.html')}\n</html>"
+        head + "\n#{body.gsub(TCPATHRE, tcpath).gsub(HTMLRE, '\1.html')}\n</html>"
       else
         nil
       end
@@ -124,19 +125,17 @@ module RdfaHelper
     
     # Read in file, and apply modifications reference either .html or .xhtml
     def results
-      htmlre = Regexp.new('([0-9]{4,4})\.xhtml')
-      tcpathre = Regexp.compile('\$TCPATH')
-
       f = self.name + ".sparql"
-      body = File.read(File.join(TEST_DIR, "tests", f)).gsub(tcpathre, tcpath)
+      body = File.read(File.join(TEST_DIR, "tests", f)).gsub(TCPATHRE, tcpath)
       
-      suite == "xhtml" ? body : body.gsub(htmlre, '\1.html')
+      suite == "xhtml" ? body : body.gsub(HTMLRE, '\1.html')
     end
     
     def triples
-      tcpathre = Regexp.compile('\$TCPATH')
       f = self.name + ".nt"
-      File.read(File.join(NT_DIR, f)).gsub(tcpathre, tcpath)
+       body = File.read(File.join(NT_DIR, f)).gsub(TCPATHRE, tcpath)
+      
+      suite == "xhtml" ? body : body.gsub(HTMLRE, '\1.html')
     end
     
     # Run test case, yields input for parser to create triples
@@ -185,7 +184,7 @@ module RdfaHelper
       
       @@test_cases = test_hash.values.map {|statements| TestCase.new(statements, suite)}.
         compact.
-        sort_by{|t| t.about.is_a?(URI) ? t.about.to_s : "zzz"}
+        sort_by{|t| t.name}
     end
   end
 end

data/spec/rdfa_parser_spec.rb

@@ -93,12 +93,12 @@ describe "RDFa parser" do
   end
 
   # W3C Test suite from http://www.w3.org/2006/07/SWD/RDFa/testsuite/
-  %w(xhtml).each do |suite|
+  %w(xhtml html4 html5).each do |suite|
     describe "w3c #{suite} testcases" do
       describe "that are approved" do
         test_cases(suite).each do |t|
           next unless t.status == "approved"
-          #next unless t.name =~ /0131/
+          #next unless t.name =~ /017\d/
           #puts t.inspect
           specify "test #{t.name}: #{t.title}#{",  (negative test)" unless t.expectedResults}" do
             begin
@@ -106,11 +106,7 @@ describe "RDFa parser" do
                 rdfa_parser.parse(rdfa_string, t.informationResourceInput)
               end
             rescue Spec::Expectations::ExpectationNotMetError => e
-              if t.title =~ /XML/
-                pending("XML Tests known to not work propery with Rasqal") {  raise }
-              else
-                raise
-              end
+              raise
             end
           end
         end
@@ -118,15 +114,15 @@ describe "RDFa parser" do
       describe "that are unreviewed" do
         test_cases(suite).each do |t|
           next unless t.status == "unreviewed"
-          #next unless t.name =~ /0154/
+          #next unless t.name =~ /0092/
           #puts t.inspect
           specify "test #{t.name}: #{t.title}#{",  (negative test)" unless t.expectedResults}" do
             begin
               t.run_test do |rdfa_string, rdfa_parser|
                 rdfa_parser.parse(rdfa_string, t.informationResourceInput)
               end
-            #rescue Spec::Expectations::ExpectationNotMetError => e
-            #  pending() {  raise }
+            rescue Spec::Expectations::ExpectationNotMetError => e
+              pending() {  raise }
             end
           end
         end

metadata

@@ -1,7 +1,8 @@
 --- !ruby/object:Gem::Specification 
 name: rdfa_parser
 version: !ruby/object:Gem::Version 
-  version: 0.1.3
+  prerelease: 
+  version: 0.1.5
 platform: ruby
 authors: 
 - Gregg Kellogg
@@ -9,60 +10,65 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2009-11-12 00:00:00 -08:00
+date: 2011-06-11 00:00:00 -07:00
 default_executable: rdfa_parser
 dependencies: 
 - !ruby/object:Gem::Dependency 
   name: addressable
-  type: :runtime
-  version_requirement: 
-  version_requirements: !ruby/object:Gem::Requirement 
+  prerelease: false
+  requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
         version: 2.0.0
-    version: 
+  type: :runtime
+  version_requirements: *id001
 - !ruby/object:Gem::Dependency 
   name: nokogiri
-  type: :runtime
-  version_requirement: 
-  version_requirements: !ruby/object:Gem::Requirement 
+  prerelease: false
+  requirement: &id002 !ruby/object:Gem::Requirement 
+    none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
         version: 1.3.3
-    version: 
+  type: :runtime
+  version_requirements: *id002
 - !ruby/object:Gem::Dependency 
   name: builder
-  type: :runtime
-  version_requirement: 
-  version_requirements: !ruby/object:Gem::Requirement 
+  prerelease: false
+  requirement: &id003 !ruby/object:Gem::Requirement 
+    none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
         version: 2.1.2
-    version: 
+  type: :runtime
+  version_requirements: *id003
 - !ruby/object:Gem::Dependency 
   name: rspec
-  type: :development
-  version_requirement: 
-  version_requirements: !ruby/object:Gem::Requirement 
+  prerelease: false
+  requirement: &id004 !ruby/object:Gem::Requirement 
+    none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
         version: "0"
-    version: 
+  type: :development
+  version_requirements: *id004
 - !ruby/object:Gem::Dependency 
   name: activesupport
-  type: :development
-  version_requirement: 
-  version_requirements: !ruby/object:Gem::Requirement 
+  prerelease: false
+  requirement: &id005 !ruby/object:Gem::Requirement 
+    none: false
     requirements: 
     - - ">="
       - !ruby/object:Gem::Version 
         version: 2.3.0
-    version: 
-description: " Yields each triple, or generate in-memory graph"
+  type: :development
+  version_requirements: *id005
+description: This gem is no longer supported, please see http://rubygems.org/gems/rdf-rdfa
 email: gregg@kellogg-assoc.com
 executables: 
 - rdfa_parser
@@ -72,14 +78,12 @@ extra_rdoc_files:
 - History.txt
 - README.rdoc
 files: 
-- .gitignore
 - .gitmodules
 - History.txt
 - README.rdoc
 - Rakefile
 - VERSION
 - bin/rdfa_parser
-- earl-xhtml1-approved.txt
 - lib/rdfa_parser.rb
 - lib/rdfa_parser/bnode.rb
 - lib/rdfa_parser/exceptions.rb
@@ -224,37 +228,28 @@ homepage: http://github.com/gkellogg/rdfa_parser
 licenses: []
 
 post_install_message: 
-rdoc_options: 
-- --charset=UTF-8
+rdoc_options: []
+
 require_paths: 
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
       version: "0"
-  version: 
 required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
   requirements: 
   - - ">="
     - !ruby/object:Gem::Version 
       version: "0"
-  version: 
 requirements: []
 
 rubyforge_project: 
-rubygems_version: 1.3.5
+rubygems_version: 1.6.2
 signing_key: 
 specification_version: 3
-summary: RDFa parser written in pure Ruby.
-test_files: 
-- spec/bnode_spec.rb
-- spec/graph_spec.rb
-- spec/literal_spec.rb
-- spec/matchers.rb
-- spec/namespace_spec.rb
-- spec/rdfa_helper.rb
-- spec/rdfa_parser_spec.rb
-- spec/spec_helper.rb
-- spec/triple_spec.rb
-- spec/uriref_spec.rb
+summary: "[Deprecated] RDFa parser written in pure Ruby."
+test_files: []
+