This file comes from glenn adams:
	A while ago, I placed a number of mappings tables from various Asian
	character sets (CNS, BIG5, JIS, GB, KSC, etc) onto anonymous FTP on
	METIS.COM [140.186.33.40].  See /pub/csmaps.  These tables have not
	only the Han character mappings, but the non-Han mappings as well.
---------------------------------------------------------------------------

File Name:		README
Description:		Overview of Character Set Mappings to Unicode
Last Changed:		Fri Oct 16, 1992 at 00:52:33
Modification History:   09/16/92 Created
			10/01/92 Add non-Han mappings to BIG5
			10/16/92 Update/Add KSC5601 mappings after verifying
			--       against IBM Japan mappings; remove bogus
			--       CNS11643 mappings.

Character set mappings to Unicode Version 1.0 are available in the following
files in (UN*X) compressed form:

  big5.Z		- Taiwanese BIG5 Character Set
  cns11643.Z		- Chinese National Standard 11643 (1986)
  gb2312.Z		- PRC Chinese Standard 2312 (1980)
  hproc16.Z		- Hewlett-Packard Series 9000 Taiwanese Character Set
  ibm5550.Z		- IBM Taiwanese Character Set
  jis208.Z		- Japanese Industrial Standard 0X208 (1983)
  ksc5601.Z		- Korean National Standard 5601 (1987)

Notes:

1. The mapping files are compressed using the standard UN*X compress
   program; they must be transferred by FTP in binary mode.  If you
   can't uncompress them, send me mail and I can arrange to uncompress
   them temporarily.
2. The first column is the source charset, the second column is UNICODE.
3. All KSC5601 mappings have been cross checked against a similar mapping
   produced by IBM Japan.  The mappings were verified as identical except
   for the following:

    # KSC        IBM     METIS
    0x2124	0x00b7	0x30fb
    0x2129	0x00ad	0x2013
    0x212a	0x2015	0x2014
    0x212b	0x2225	0x2016
    0x212d	0x223c	0xff5e
    0x214b	0xffe0	0x00a2
    0x214c	0xffe1	0x00a3
    0x214d	0xffe5	0x00a5
    0x216c	0x226a	0x00ab
    0x216d	0x226b	0x00bb
    0x217e	0xffe2	0x00ac
    0x2226	0xff5e	0x02dc
    0x2230	0x02d0	0x2236
    0x2241	0x2299	0x25c9
    0x235c	0xffe6	0x20a9
    0x237e	0xffe3	0x203e

4. The IBM5550 map currently contains only the Han character mappings, and not
   any non-Han character mappings.
5. CNS11643 separately encodes radicals which are not maintained distinctly
   in Unicode; these entries have an ASTERISK in the second column.  They
   could be unified with the non-radical Han characters of the same shape.
   A small number of other characters have not yet been mapped; these too
   have an asterisk in the second column.
6. HPROC16 is used on Hewlett Packard 9000 series in Taiwan; it is
   essentially a transformed Chinese Telegraph Code, containing both level
   one and level two CCTC codes.  This mapping is incomplete.  Only those
   characters which are also present in BIG5 have mappings.  Of the characters
   mapped, three are duplicates, and are marked with a third column beginning
   with an '#' character.
7. The unicode mappings conform to the published volumes; any changes which
   were caused by the 10646 merger have not yet been reflected in these
   tables.

If you use these tables, please inform me of any errors or discrepancies you
encounter.  The process of creating mappings is not always straightforward;
e.g., in CNS11643, there are some pretty questionable characters with little
known semantics.  In these cases, I chose the best Unicode character I could
find at the time.

Glenn Adams
<glenn@metis.com>
