Using oniguruma for splitting strings...

oniguruma.rb

# ¿á you think á eh?

# "스마트폰을 가지고 있나? "
# 
# reg = ORegexp.new( 'р(уби.*)', 'i', 'utf8' )
# matches = reg.match("Text: Ехал Грека Через Реку")
# 
# 5ms)   SHOW TABLES
#   Sphinx   Querying: '스마트폰을 | 가지고 | 있나'
#   Sphinx (0.001932s)   Found  results
#   Sphinx   Sphinx Daemon returned error: index ...: syntax error, unexpected '|' near '스마트폰을 | 가지고 | 있나'
# 
# ThinkingSphinx::SphinxError (
require 'rubygems'
require 'oniguruma'


# query.gsub(/\W+/, ' ').split(/\s+/).select { |s| s.split('').length > 1 }.join(' | ')

a = "스마트폰을 가지고 있나? "

puts Oniguruma::ORegexp.new('\w+', :encoding => Oniguruma::ENCODING_UTF8).scan(a).map { |m| m[0].unpack("U*").map { |e| "U+%04x" % e }.join }.join(' | ')