Re: CCCII and CNS questions

From: Werner LEMBERG (sx0005@sx2.hrz.uni-dortmund.de)
Date: Thu Jan 21 1999 - 17:55:11 EST


> I'm currently working on a CCCII->CNS mapping table by printing
> out CCCII and CNS glyphs side by side (I'll probably post this on
> a server in case someone is interested -- expect a set of *large*
> PS files). As a starting point, I use Christian Wittern's
> mapping data from his cefdb database of the kanjibas project.

   You and I seem to be on amazingly similar paths. I got a copy of
   the very same database file, and have converted it to a dBase
   version. I have been writing a Visual Basic application to display
   the characters side by side to check&correct errors. Using manual
   methods, I've already found a few. Once I have the full
   application up an running, I expect I'll find quite a few more.

I've now checked about 20% of my CCCII->CNS mapping (almost the first
CCCII plane plus its variant glyphs). First result: a systematic
error, affecting about 280 characters! Everything in the CNS range
6-26xx up to 6-28xx is off by one. Example:

  2D3D42 -> 6-2752 should be 2D3D42 -> 6-2753

Except this, I've found few errors (about 20).

My road map:

  .) complete the CCCII->CNS check
  .) try to map all remaining CNS characters to CCCII
  .) try to find all CCCII duplicate glyphs so that I can represent as
     much CCCII as possible with CNS glyphs -- I'm not interested in a
     one-to-one mapping (which isn't possible anyway).

In which results are you interested?

BTW, do you have a Four-Courner table of all CCCII glyphs (not only
the base glyphs)?

    Werner



This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:44 EDT