Add some boilerplate support for rdfstar.

Author: gkellogg

.travis.yml

@@ -57,9 +57,6 @@ Full documentation available on [Rubydoc.info][RDF/XML doc])
 ### Additional vocabularies
 * {RDF::XML}
 
-## TODO
-* Consider a SAX-based parser for improved performance
-
 ## Resources
 * [RDF.rb][RDF.rb]
 * [RDF/XML][RDF/XML]
@@ -109,4 +106,5 @@ see <https://unlicense.org/> or the accompanying {file:UNLICENSE} file.
 [YARD]:         https://yardoc.org/
 [YARD-GS]:      https://rubydoc.info/docs/yard/file/docs/GettingStarted.md
 [PDD]:              https://unlicense.org/#unlicensing-contributions
-[RDF/XML doc]:  https://rubydoc.info/github/ruby-rdf/rdf-rdfxml/master/frames
+[RDF/XML doc]:  https://rubydoc.info/github/ruby-rdf/rdf-rdfxml/master/frames
+[RDF-star]:         https://w3c.github.io/rdf-star/rdf-star-cg-spec.html

README.md

@@ -5,9 +5,9 @@ require "bundler/setup"
 require 'rdf/rdfxml'
 require 'rdf/turtle'
 require 'rdf/ntriples'
+require 'rdf/ordered_repo'
 require 'getoptlong'
 require 'logger'
-require 'open-uri'
 
 def run(input, **options)
   reader_class = RDF::Reader.for(options[:input_format].to_sym)
@@ -17,10 +17,12 @@ def run(input, **options)
   num = 0
 
   if options[:output_format] == :ntriples || options[:quiet]
-    reader_class.new(input, **options).each do |statement|
+    reader_class.new(input, **options[:parser_options]).each do |statement|
       num += 1
-      if options[:quiet]
-        #print "."
+      if options[:errors] && statement.invalid?
+        $stderr.puts "Invalid statement #{statement.inspect}"
+      elsif options[:quiet]
+        print "." if options[:quiet] == 1
       else
         options[:output].puts statement.to_ntriples
       end
@@ -31,63 +33,108 @@ def run(input, **options)
       options[:output].puts statement.inspect
     end
   else
-    r = reader_class.new(input, **options)
-    g = RDF::Repository.new << r
+    r = reader_class.new(input, **options[:parser_options])
+    g = RDF::OrderedRepo.new << r
     num = g.count
-    STDERR.puts "graph: #{g.dump(:ttl)}" if options[:verbose]
-    options[:output].puts g.dump(options[:output_format], options.merge(prefixes: r.prefixes))
+    options[:output].puts g.dump(options[:output_format], prefixes:  r.prefixes, **options[:writer_options])
   end
-  STDERR.puts
+  $stderr.puts
   secs = Time.new - start
-  STDERR.puts "Parsed #{num} statements in #{secs} seconds @ #{num/secs} statements/second."
-rescue
+  $stderr.puts "Parsed #{num} statements in #{secs} seconds @ #{num/secs} statements/second."
+rescue Exception => e
   fname = input.respond_to?(:path) ? input.path : "-stdin-"
-  STDERR.puts("Error in #{fname}")
-  raise
+  $stderr.puts("Error in #{fname}: #{e.message}")
+  $stderr.puts "Backtrace: " + e.backtrace.join("\n  ")
+  raise e
 end
 
 logger = Logger.new(STDERR)
 logger.level = Logger::WARN
 logger.formatter = lambda {|severity, datetime, progname, msg| "#{severity}: #{msg}\n"}
 
+parser_options = {
+  base_uri:  nil,
+  validate:  false,
+  logger: logger,
+}
+
+writer_options = {
+  base_uri:  nil,
+  standard_prefixes:  true,
+  logger: logger,
+}
+
 options = {
+  parser_options:  parser_options,
+  writer_options:  writer_options,
   input_format: :rdfxml,
-  output: STDOUT,
   output_format: :ntriples,
+  output: STDOUT,
   strict: true,
   validate: false,
   verbose: false,
-  logger: logger
 }
+
 input = nil
 
-opts = GetoptLong.new(
-  ["--dbg", GetoptLong::NO_ARGUMENT],
-  ["--execute", "-e", GetoptLong::REQUIRED_ARGUMENT],
-  ["--format", GetoptLong::REQUIRED_ARGUMENT],
-  ["--input-format", GetoptLong::REQUIRED_ARGUMENT],
-  ["--output", "-o", GetoptLong::REQUIRED_ARGUMENT],
-  ["--quiet", GetoptLong::NO_ARGUMENT],
-  ["--template", GetoptLong::REQUIRED_ARGUMENT],
-  ["--uri", GetoptLong::REQUIRED_ARGUMENT],
-  ["--validate", GetoptLong::NO_ARGUMENT],
-  ["--verbose", GetoptLong::NO_ARGUMENT],
-)
+OPT_ARGS = [
+  ["--canonicalize", GetoptLong::NO_ARGUMENT, "Canonize all terms"],
+  ["--debug", GetoptLong::NO_ARGUMENT, "Debugging output"],
+  ["--errors", GetoptLong::NO_ARGUMENT, "Display invalid statements"],
+  ["--execute", "-e", GetoptLong::REQUIRED_ARGUMENT, "Run against source in argument"],
+  ["--format", GetoptLong::REQUIRED_ARGUMENT, "Output format, any RDF format symbol, sxp, or inspect"],
+  ["--help", "-?", GetoptLong::NO_ARGUMENT, "print this message"],
+  ["--input-format", GetoptLong::REQUIRED_ARGUMENT, "Format of the input file, defaults to ttl"],
+  ["--info", GetoptLong::NO_ARGUMENT, "Show progress on execution"],
+  ["--output", "-o", GetoptLong::REQUIRED_ARGUMENT, "Save output to file"],
+  ["--quiet", GetoptLong::NO_ARGUMENT, "Do not show parser output"],
+  ["--stream", GetoptLong::NO_ARGUMENT, "Use streaming writer"],
+  ["--template", GetoptLong::REQUIRED_ARGUMENT, "Haml template for writer"],
+  ["--uri", GetoptLong::REQUIRED_ARGUMENT, "Default base URI"],
+  ["--validate", GetoptLong::NO_ARGUMENT, "Run parser in strict validation mode"],
+  ["--verbose", GetoptLong::NO_ARGUMENT, "Verbose output"],
+]
+
+def usage
+  STDERR.puts %{
+    RDF::RDFXML version #{RDF::RDFXML::VERSION}
+    
+    Usage: #{$0} [options] file ...
+  }.gsub(/^    /, '')
+  width = OPT_ARGS.map do |o|
+    l = o.first.length
+    l += o[1].length + 2 if o[1].is_a?(String)
+    l
+  end.max
+  OPT_ARGS.each do |o|
+    s = "  %-*s  " % [width, (o[1].is_a?(String) ? "#{o[0,2].join(', ')}" : o[0])]
+    s += o.last
+    STDERR.puts s
+  end
+  exit(1)
+end
+
+opts = GetoptLong.new(*OPT_ARGS.map {|o| o[0..-2]})
+
 opts.each do |opt, arg|
   case opt
-  when '--dbg'            then logger.level = Logger::DEBUG
-  when '--execute'        then input = arg
-  when '--format'         then options[:output_format] = arg.to_sym
-  when '--input-format'   then options[:input_format] = arg.to_sym
-  when '--output'         then options[:output] = File.open(arg, "w")
+  when '--debug'        then logger.level = Logger::DEBUG
+  when '--canonicalize' then parser_options[:canonicalize] = true
+  when '--errors'       then options[:errors] = true
+  when '--execute'      then input = arg
+  when '--format'       then options[:output_format] = arg.to_sym
+  when "--help"         then usage()
+  when '--input-format' then options[:input_format] = arg.to_sym
+  when '--output'       then options[:output] = File.open(arg, "w")
+  when '--rdfstar'      then parser_options[:rdfstar] = true
   when '--quiet'
-    options[:quiet] = true
+    options[:quiet] = options[:quiet].to_i + 1
     logger.level = Logger::FATAL
-  when '--template'       then options[:haml] = arg.to_sym
-  when '--uri'            then options[:base_uri] = arg
-  when '--validate'       then options[:validate] = true
-  when '--verbose'        then options[:verbose] = true
-  when '--version'        then options[:version] = arg
+  when '--template'     then options[:haml] = arg.to_sym
+  when '--uri'          then parser_options[:base_uri] = writer_options[:base_uri] = arg
+  when '--validate'     then parser_options[:validate] = true
+  when '--verbose'      then options[:verbose] = true
+  when '--version'      then options[:version] = arg
   end
 end
 
@@ -99,3 +146,4 @@ else
     run(Kernel.open(test_file), **options)
   end
 end
+puts