From: Ernest Ellingson <erne@...>
Date: 2001-12-03T07:50:27+09:00
Subject: [ruby-talk:27298] Re: ANN: REXML 1.1a7

In my haste to put something out about converting files to UTF8, I screwed 
up the ASCII conversion.  In addition the regular expression that detects 
UTF-8 was changed slightly (100 to 200 {1} removed. The code below works 
correctly.

def utf8_enc(str)
      enc=""
      if str=~/^\376\377/
           enc='UTF16'  #unicode big endian
      elsif str=~/^\377\376/
           enc='UNILE' #unicode litle endian
      elsif str=~/([\300-\377][\200-\277])/  #not a guaranteed test won't 
detect random pairs
           enc='UTF8'
      else
           enc='ASC'
      end
      if enc== 'UTF8'
           strUTF8=str
      else
           arrayEnc=[]
           str.each_byte{|b| arrayEnc << b}
           arrayUTF8=[]
           case enc
           when 'ASC'
                arrayEnc.each{|b| arrayUTF8 << [b].pack("U")}
                strUTF8=arrayUTF8.join("")
           when 'UTF16' || 'UNILE'
                if enc=='UTF16'
                     2.step(arrayEnc.size-1, 2){|i| arrayUTF8 << 
[arrayEnc.at(i+1) + arrayEnc.at(i) * 0x100].pack("U")}
                elsif enc=='UNILE'
                     2.step(arrayEnc.size-1, 2){|i| arrayUTF8 << 
[arrayEnc.at(i) + arrayEnc.at(i+1) * 0x100].pack("U")}
                else
                     str.each_byte
                end
                strUTF8=arrayUTF8.join("")
           end
      end
end