regex_unicode_kcode.rb

latin = "a"
arabic = [0xfeb6].pack('U') # Sheen, ش
cjk = [0x3333].pack('U') # HUIITO, ㌳

kcodes = [ 'NONE', 'u', 's' ]
chars = [latin, arabic, cjk]
patterns = [ /\w/, /\W/, /[[:punct:]]/ ]

kcodes.each do |kcode|
  $KCODE = kcode
  puts "KCODE = #{kcode}"
  patterns.each do |pattern|
    puts "  Pattern: #{pattern}"
    chars.each do |char|
      puts "    #{char}: #{!!char.match(pattern)}"
    end
  end
end

# Output:
#
#KCODE = NONE
#  Pattern: (?-mix:\w)
#    a: true
#    ﺶ: false
#    ㌳: false
#  Pattern: (?-mix:\W)
#    a: false
#    ﺶ: true
#    ㌳: true
#  Pattern: (?-mix:[[:punct:]])
#    a: false
#    ﺶ: false
#    ㌳: false
#KCODE = u
#  Pattern: (?-mix:\w)
#    a: true
#    ﺶ: true
#    ㌳: true
#  Pattern: (?-mix:\W)
#    a: false
#    ﺶ: false
#    ㌳: false
#  Pattern: (?-mix:[[:punct:]])
#    a: false
#    ﺶ: false
#    ㌳: false
#KCODE = s
#  Pattern: (?-mix:\w)
#    a: true
#    ﺶ: true
#    ㌳: true
#  Pattern: (?-mix:\W)
#    a: false
#    ﺶ: true
#    ㌳: true
#  Pattern: (?-mix:[[:punct:]])
#    a: false
#    ﺶ: false
#    ㌳: false