Re: How do the Koreans do it?

From: Ken Lunde (lunde@mv.us.adobe.com)
Date: Wed Oct 26 1994 - 13:37:13 EDT


        Here is how I understand it:

o The leading bit in the two octets (16 bits) is set.

o The remaining 15 bits are split into three groups of 5 bits each.

o A pre-combined hangul character can be composed of up to three hangul
  elements, each one representing a letter in their hangul alphabet.

o These three 5-bit chunks are used to encode the the hangul elements that
  compose a pre-combined hangul.

Think of it as a three-dimensional matrix whereby each axis represents the
three possible positions for hangul elements within a pre-combined character,
and each axis is presented by five bits (32 values).
        The thing that disturbs me is that systems *must* treat the 16 bits
as a single unit. One of these 5-bit units spans the first and second octet,
which, in my opinion, is dangerous from an electronic transmission point
of view.

-- Ken Lunde
   Project Manager for CJK Font Development
   Adobe Systems Incorporated
   lunde@mv.us.adobe.com
   http://jasper.ora.com/lunde/ (my WWW Home Page)



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:30 EDT