multicore benchmarks are at https://github.com/rubychan/stringscanner-benchmarks

1 files changed, 0 insertions, 161 deletions

diff --git a/etc/multicore/multi-slice.rb b/etc/multicore/multi-slice.rb
deleted file mode 100644
index 33b8cd6..0000000
--- a/etc/multicore/multi-slice.rb
+++ /dev/null
@@ -1,161 +0,0 @@
-require 'strscan'
-require 'benchmark'
-require 'thread'
-
-class Scanner < StringScanner
-  
-  def initialize code
-    super code
-  end
-  
-  def tokenize encoder = Tokens.new
-    scan_tokens encoder
-    encoder
-  end
-  
-protected
-  
-  def scan_tokens encoder
-    until eos?
-      if matched = scan(/ +/)
-        encoder.text_token matched, :space
-      elsif matched = scan(/\n/)
-        encoder.text_token matched, :space
-      elsif matched = scan(/!/)
-        encoder.text_token matched, :not_going_to_happen
-      elsif matched = scan(/=/)  #/
-        encoder.text_token matched, :not_going_to_happen
-      elsif matched = scan(/%/)
-        encoder.text_token matched, :not_going_to_happen
-      elsif matched = scan(/\d+/)
-        encoder.text_token matched, :number
-      elsif matched = scan(/\w+/)
-        encoder.text_token matched, :word
-      elsif matched = scan(/[,.]/)
-        encoder.text_token matched, :op
-      elsif scan(/\(/)
-        encoder.begin_group :par
-      elsif scan(/\)/)
-        encoder.end_group :par
-      else
-        raise
-      end
-    end
-  end
-  
-end
-
-
-class Tokens < Array
-  alias token push
-  alias text_token push
-  alias block_token push
-  def begin_group kind; push :begin_group, kind end
-  def end_group kind; push :end_group, kind end
-end
-
-
-class Encoder
-  
-  def setup
-    @out = ''
-    @opened = []
-  end
-  
-  def finish
-    @out
-  end
-  
-  def encode scanner
-    setup
-    scanner.tokenize self
-    finish
-  end
-  
-  # def token content, kind
-  #   if content.is_a? ::String
-  #     text_token content, kind
-  #   elsif content.is_a? ::Symbol
-  #     block_token content, kind
-  #   else
-  #     raise 'Unknown token content type: %p' % [content]
-  #   end
-  # end
-  
-  def text_token text, kind
-    # @out <<
-    #   if kind == :space
-    #     text
-    #   else
-    #     text.gsub!(/[)\\]/, '\\\\\0')  # escape ) and \
-    #     "#{kind}(#{text})"
-    #   end
-  end
-  
-  # def block_token action, kind
-  #   case action
-  #   when :begin_group
-  #     begin_group kind
-  #   when :end_group
-  #     end_group kind
-  #   else
-  #     raise
-  #   end
-  # end
-  
-  def begin_group kind
-    # @opened << kind
-    # @out << "#{kind}<"
-  end
-  
-  def end_group kind
-    # @opened.pop
-    # @out << '>'
-  end
-  
-end
-
-size = ((ARGV.first || 1).to_f * 1_000_000).to_i  # size
-
-# generate string
-code = "2011 alpha, beta, (gamma), delta.\n"
-code *= (size.to_f / code.size).ceil
-code.slice! size..-1
-
-slice_size = (ARGV[1] || 100).to_i
-N = 1
-
-100.times do
-  out = Encoder.new.encode(Scanner.new(code))
-end
-
-1000.times do
-2.times do
-  threads = []
-  seconds = Benchmark.realtime do N.times do
-    chunk_offsets = [0]
-    code.lines.each_slice slice_size do |lines|
-      chunk_offsets << chunk_offsets.last + lines.join.bytesize
-    end
-    threads.clear
-    chunk_offsets.each_cons(2) do |this_chunk, next_chunk|
-      threads << Thread.new do
-        Thread.current[:out] = Encoder.new.encode Scanner.new(code[this_chunk...next_chunk])
-      end
-    end
-    threads.each(&:join)
-    out = threads.map { |t| t[:out] }.join
-  end end
-  
-  mb = N * size / 1_000_000.0
-  puts 'Multi-Threaded: %0.1f MB in %0.2fs = %0.1f MB/s @ %d threads' % [mb, seconds, mb / seconds, threads.size]
-end
-2.times do
-  seconds = Benchmark.realtime do N.times do
-    out = Encoder.new.encode(Scanner.new(code))
-  end end
-  
-  mb = N * size / 1_000_000.0
-  puts 'Single-Threaded: %0.1f MB in %0.2fs = %0.1f MB/s' % [mb, seconds, mb / seconds]
-end
-end
\ No newline at end of file