Ruby-marc about half as fast as marc4j doing MARC21 and MARC-XML under Traject

results.txt

jruby 1.7.18 (1.9.3p551) 2014-12-22 625381c on Java HotSpot(TM) 64-Bit Server VM 1.8.0-b132 +jit [darwin-x86_64]



Rehearsal ------------------------------------------------
marc4j w/bin  19.640000   1.000000  20.640000 (  8.313000)
ruby w/bin    22.470000   0.600000  23.070000 ( 14.930000)

marc4j w/xml  12.950000   1.130000  14.080000 (  6.189000)
ruby w/xml    19.970000   0.390000  20.360000 ( 12.946000)
-------------------------------------- total: 78.150000sec

                   user     system      total        real
marc4j w/bin   8.430000   0.970000   9.400000 (  6.782000)
ruby w/bin    15.590000   0.500000  16.090000 ( 13.988000)

marc4j w/xml   7.800000   1.090000   8.890000 (  5.717000)
ruby w/xml    11.720000   0.220000  11.940000 ( 10.790000)

xmlbench.rb

$:.unshift  "#{File.dirname(__FILE__)}/../lib"
require 'traject'
require 'traject/marc_reader'
require 'traject/marc4j_reader'
require 'traject/null_writer'
require 'benchmark'

xmlfilename = '10k.xml'
mrcfilename = '10k.mrc'

indexer = Traject::Indexer.new(
  "log.level" => "error"
)

indexer.writer_class = Traject::NullWriter
indexer.to_field('constant') do |rec, acc|
  acc << 'Constant'
end


puts "\n\n#{RUBY_DESCRIPTION}"

Benchmark.bmbm do |x|

  x.report('marc4j w/bin') do
    indexer.settings['marc_source.type'] = 'binary'
    indexer.settings['marc4j_reader.permissive'] = true
    indexer.reader_class = Traject::Marc4JReader
    indexer.process(File.open(mrcfilename))
  end
  x.report('ruby w/bin  ') do
    indexer.settings['marc_source.type'] = 'binary'
    indexer.reader_class = Traject::MarcReader
    indexer.process(File.open(mrcfilename))
  end
  
  puts "\n"

  x.report('marc4j w/xml') do
    indexer.settings['marc_source.type'] = 'xml'    
    indexer.reader_class = Traject::Marc4JReader
    indexer.process(File.open(xmlfilename))
  end
  x.report('ruby w/xml  ') do
    indexer.settings['marc_source.type'] = 'xml'    
    indexer.reader_class = Traject::MarcReader
    indexer.process(File.open(xmlfilename))
  end
  
  puts "\n\n"
  
end