red.rb

# encoding: utf-8
unless RUBY_VERSION =~ /^1.9/
  $KCODE = 'u'
  puts "$KCODE == " + $KCODE
else
  puts "Encoding == " + "".encoding.name
end

s = "a *b* c *d*<"

# Extracted from redcloth3
re = /(^|[>\s\(])          # sta
      (?!\-\-)
      (\*\*|\*|\?\?|\-|__|_|%|\+|\^|~|)      # oqs
      (\*)              # qtag
      (\w|[^\s].*?[^\s])    # content
      (?!\-\-)
      \*
      (\*\*|\*|\?\?|\-|__|_|%|\+|\^|~|)      # oqa
      (?=[[:punct:]]|\s|\)|$)/x

s.scan(re) { |m| p m }

# Attempt to simplify the regex
re = /(\*)              # qtag
      ([^\*])    # content
      \*
      (?=[[:punct:]]|\s)/x

s.scan(re) { |m| p m }

running with 1.8.7p249

$ ruby1.8.7 -v red.rb
ruby 1.8.7 (2010-01-10 patchlevel 249) [i686-darwin9.8.0]
$KCODE == UTF8
[" ", "", "*", "b", ""]
[" ", "", "*", "d", ""]
["*", "b"]
["*", "d"]

running with 1.9.2dev r27536

$ ruby1.9 -v red.rb
ruby 1.9.2dev (2010-04-28 trunk 27536) [i386-darwin9.8.0]
Encoding == UTF-8
[" ", "", "*", "b", ""]
["*", "b"]