class PDF::Reader::Encoding::IdentityH

def to_utf8(str, tounicode = nil)
  array_enc = []
  # iterate over string, reading it in 2 byte chunks and interpreting those
  # chunks as ints
  str.unpack("n*").each do |num|
    # convert the int to a unicode codepoint if possible.
    # without a ToUnicode CMap, it's impossible to reliably convert this text
    # to unicode, so just replace each character with a little box. Big smacks
    # the the PDF producing app.
    if tounicode && (code = tounicode.decode(num))
      array_enc << code
    else
      array_enc << PDF::Reader::Encoding::UNKNOWN_CHAR
    end
  end
  # replace charcters that didn't convert to unicode nicely with something valid
  array_enc.collect! { |c| c ? c : PDF::Reader::Encoding::UNKNOWN_CHAR }
  # pack all our Unicode codepoints into a UTF-8 string
  ret = array_enc.pack("U*")
  # set the strings encoding correctly under ruby 1.9+
  ret.force_encoding("UTF-8") if ret.respond_to?(:force_encoding)
  return ret
end